Engdyn 2014.1 Manual

RD.14/185701.1

Ricardo Software Engine Dynamics Simulation

ENGDYN DOCUMENTATION/USER MANUAL VERSION 2014.1 May 2014

Contents

Contents WHAT’S NEW IN ENGDYN 2014.1? .............................................................................................. V A. 1 2

B. 1

2

3

4

5

6 7

KNOWLEDGE CENTRE ........................................................................................................... 1 W HAT IS ENGDYN? ................................................................................................................... 1 TUTORIALS .................................................................................................................................. 3 2.1 Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis .......................... 3 2.2 Tutorial 2: Crankshaft Dynamic Analysis ........................................................................43 2.3 Tutorial 3: NVH Analysis .................................................................................................79 2.4 Tutorial 4: Block Stress Analysis ...................................................................................125 2.5 Tutorial 5: Crankshaft Stress Analysis ..........................................................................149 2.6 Tutorial 6: EHL Big End Bearing Analysis .....................................................................183 HELP .....................................................................................................................................213 USING ENGDYN ....................................................................................................................213 1.1 Overview .......................................................................................................................213 1.2 Getting Started ..............................................................................................................213 1.3 Description of the Main Panel .......................................................................................214 ENGDYN MODELS .................................................................................................................223 2.1 Overview .......................................................................................................................223 2.2 ENGDYN Co-ordinate System ......................................................................................223 2.3 Finite Element Model Data ............................................................................................224 2.4 Crank Train Model .........................................................................................................225 2.5 Cylinder Block Model ....................................................................................................231 2.6 Journal Bearing Oil Film Model .....................................................................................232 2.7 In-Cylinder Model ..........................................................................................................235 2.8 Gas Cylinder Pressure ..................................................................................................236 MODEL GENERATION ...............................................................................................................237 3.1 General ..........................................................................................................................237 3.2 Configure Engine ..........................................................................................................237 3.3 Define Models ...............................................................................................................246 3.4 Editing the Crank Train Model .......................................................................................253 3.5 Editing the Cylinder Block Model ..................................................................................324 SOLUTION ...............................................................................................................................353 4.1 General ..........................................................................................................................353 4.2 Lubrication .....................................................................................................................354 4.3 Loading..........................................................................................................................359 4.4 Evaluate Solution ..........................................................................................................374 POST-PROCESSING .................................................................................................................391 5.1 General ..........................................................................................................................391 5.2 Selecting Loadcases .....................................................................................................391 5.3 Selecting Crank Train Modes ........................................................................................392 5.4 Plotting EHL Results .....................................................................................................393 5.5 Plotting Results .............................................................................................................402 5.6 Exporting Results ..........................................................................................................407 5.7 Backsubstitution ............................................................................................................413 5.8 Animate Results ............................................................................................................413 CRANK SHAFT STRESS ANALYSIS .............................................................................................419 6.1 Overview .......................................................................................................................419 6.2 Using the Graphical User Interface ...............................................................................434 CYLINDER BLOCK ANALYSIS .....................................................................................................463 7.1 Overview .......................................................................................................................463

i

KNOWLEDGE CENTRE What is ENGDYN?

7.2 Frequency Response and Acoustic Analysis ............................................................... 463 7.3 Quasi-Static Analysis .................................................................................................... 468 7.4 Using the Graphical User Interface .............................................................................. 476 8 ADDITIONAL POST-PROCESSING .............................................................................................. 499 8.1 Overview ....................................................................................................................... 499 8.2 Quasi-Static Analysis of a Piston .................................................................................. 499 8.3 Applying Loads to a Finite Element Model of a Connecting Rod ................................. 500 C.

COMMAND FILES ............................................................................................................... 505

1

PRE-PROCESSOR COMMANDS ................................................................................................. 506 1.1 AXES ............................................................................................................................ 507 1.2 BEARING ...................................................................................................................... 508 1.3 BLOCK .......................................................................................................................... 515 1.4 CHILD ........................................................................................................................... 519 1.5 CONROD ...................................................................................................................... 521 1.6 CRANK ......................................................................................................................... 525 1.7 CYLINDER .................................................................................................................... 530 1.8 DAMPER ...................................................................................................................... 534 1.9 DIRECTION .................................................................................................................. 539 1.10 DRIVE ........................................................................................................................... 540 1.11 ELEMENT ..................................................................................................................... 541 1.12 ENGINE ........................................................................................................................ 548 1.13 IN_CYLINDER .............................................................................................................. 549 1.14 LINK .............................................................................................................................. 550 1.15 LOADING ...................................................................................................................... 552 1.16 LUBRICANT ................................................................................................................. 557 1.17 MASS ............................................................................................................................ 559 1.18 MATERIAL .................................................................................................................... 564 1.19 MOUNT ......................................................................................................................... 565 1.20 NODE............................................................................................................................ 567 1.21 OPEN ............................................................................................................................ 568 1.22 PROFILE ...................................................................................................................... 569 1.23 TITLE ............................................................................................................................ 572 2 SOLVER COMMANDS ............................................................................................................... 573 2.1 BEARING ...................................................................................................................... 574 2.2 BLOCK .......................................................................................................................... 575 2.3 COUPLING ................................................................................................................... 576 2.4 DAMPING ..................................................................................................................... 578 2.5 LUBRICANT ................................................................................................................. 580 2.6 OPEN ............................................................................................................................ 581 2.7 SOLUTION ................................................................................................................... 582 3 POST-PROCESSOR COMMANDS ............................................................................................... 592 3.1 CRANK_ANALYSIS...................................................................................................... 592 3.2 BLOCK_ANALYSIS ...................................................................................................... 606 D.

THEORY ............................................................................................................................... 616

1 2

INTRODUCTION........................................................................................................................ 616 STATIC SOLUTIONS ................................................................................................................. 617 2.1 Determinate Solution .................................................................................................... 617 2.2 Indeterminate Solution .................................................................................................. 617 3 DYNAMIC SOLUTIONS .............................................................................................................. 618 4 BEARING MODEL ..................................................................................................................... 619 4.1 Mobility Method ............................................................................................................. 621 4.2 Solution of Reynolds Equation on a Computational Mesh ........................................... 629

ii

Contents

4.3 4.4 4.5 4.6 4.7 4.8 4.9 E.

Oil Viscosity ...................................................................................................................641 Boundary Lubrication Model .........................................................................................643 Integration of Bearing Forces and Moments .................................................................645 Bearing Thermal Balance Solution ...............................................................................645 Compliant Model for EHD Bearing ................................................................................650 Nomenclature ................................................................................................................653 References ....................................................................................................................655

APPENDICES .......................................................................................................................657

1

CALCULATION OF CRANKSHAFT STIFFNESS’ USING THE FINITE ELEMENT METHOD .......................657 1.1 Overview .......................................................................................................................657 1.2 Calculation of Crankshaft Web Stiffnesses ...................................................................657 1.3 Calculation of Crankshaft Element Stiffnesses .............................................................660 2 A METHOD OF EVALUATING THE MASS AND STIFFNESS PROPERTIES OF A FLYWHEEL .................663 2.1 Overview .......................................................................................................................663 2.2 Modelling Approach ......................................................................................................663 2.3 Calculation of Mass and Stiffness Properties ...............................................................663 3 CALCULATION OF MATERIAL STRENGTH PROPERTIES FOR CRANKSHAFT STRESS ANALYSIS .......669 3.1 Overview .......................................................................................................................669 3.2 Base Strengths ..............................................................................................................669 3.3 Elevated Strengths ........................................................................................................669 4 CALCULATION OF STRESS CONCENTRATION FACTORS FOR CRANKSHAFT STRESS ANALYSIS ......671 4.1 Overview .......................................................................................................................671 4.2 Nomenclature ................................................................................................................671 4.3 Journal Fillets ................................................................................................................671 4.4 Oil Hole Breakouts ........................................................................................................673 5 TREATMENT OF NOTCH SENSITIVITY IN CRANKSHAFT STRESS ANALYSIS ....................................677 5.1 Overview .......................................................................................................................677 5.2 Treatment in Finite Element Analysis ...........................................................................678 6 USING OUTPUT FROM THE RICARDO PROGRAM VALDYN AND TVFORCED IN CRANKSHAFT STRESS CALCULATIONS ...................................................................................................................681 6.1 Using VALDYN for Crankshaft Stress Calculations ......................................................681 6.2 Using TVFORCED for Crankshaft Stress Calculations ................................................684 7 SAFETY FACTOR CALCULATIONS ..............................................................................................687 7.1 The Goodman Diagram .................................................................................................687 7.2 Goodman Diagram Construction...................................................................................688 7.3 Equivalent Stress Options .............................................................................................689 7.4 Multi-axial Fatigue Safety Factor...................................................................................691 7.5 Dang Van Fatigue Safety Factor ...................................................................................692 7.6 LinearSWT Safety Factor ..............................................................................................695 7.7 Cycles to Failure Calculation.........................................................................................696 8 RADIATED NOISE CALCULATIONS .............................................................................................699 8.1 Overview .......................................................................................................................699 8.2 Acoustic Equations and the Rayleigh Integral ..............................................................699 8.3 Radiation Efficiency .......................................................................................................703 8.4 Multiple Face Sets .........................................................................................................703 9 MODAL FREQUENCY RESPONSE CALCULATIONS .......................................................................705 9.1 Overview .......................................................................................................................705 9.2 Calculation of Modal Contributions and Vibration Response ........................................705 10 APPLYING LOADS PREDICTED BY VALDYN TO THE CYLINDER BLOCK AND CRANKSHAFT MODELS 707 10.1 Overview .......................................................................................................................707 10.2 Writing Force Profile Data From VALDYN ....................................................................707 10.3 Applying Force Profile Data to the ENGDYN Model .....................................................710

iii


10.4 Applying Force Profile Data to the Cylinder Block Model ............................................. 710 11 ASCII DATA FILE FORMAT READ BY SDF_READ_ASCII() ........................................................ 715 11.1 Introduction ................................................................................................................... 715 11.2 Format........................................................................................................................... 715 12 MATUTIL .......................................................................................................................... 720 12.1 Overview ....................................................................................................................... 720 12.2 Using matutil ................................................................................................................. 720 12.3 Theory ........................................................................................................................... 727

iv

What’s New in ENGDYN 2014.1?

What’s New in ENGDYN 2014.1? Enhancements Feature

Documentation

Updated NVH Analysis Tutorial

Knowledge Tutorial 3

Centre

Updated Block Stress Analysis Tutorial

Knowledge Tutorial 4

Centre

Various bug fixes

Release Notes

v


A. KNOWLEDGE CENTRE 1 What is ENGDYN? ENGDYN is a computer program for analysing the dynamics of the engine and in particular the dynamics of the crank train and its interaction with the cylinder block. ENGDYN provides a number of different solution techniques for predicting engine dynamics using models of varying degrees of sophistication. The crank train and cylinder block models can either be defined as rigid, compliant and dynamic. In its simplest form ENGDYN can be used to perform a statically-determinate solution, whilst in its most sophisticated form it can be used to predict the time-domain response of the 3-dimensional vibration of the coupled crank train and cylinder block system with nonlinear oil films at each of the main journal bearings. This flexibility enables the user to generate an engine model and to perform a solution to meet his particular needs. ENGDYN provides a Graphical User Interface (GUI) to enable the user to perform the model generation, solution and results presentation phases within an easy to use graphical environment. The GUI contains a built-in unit’s converter and automatically converts parameters from their defined units to SI units. Alternatively, for the more experienced user and to provide compatibility with models that were generated before the development of the GUI, ENGDYN also provides a non-graphical environment that uses command data files for model generation, solution and results presentation. ENGDYN uses a Ricardo binary standard data file to store both the model and the results of the solution. When saving, ENGDYN stores the model at the current position and as the model is built the file is appended to. Similarly, once a simulation has been executed, the results are appended to the file. Crank train simulation using ENGDYN consists of three stages. Firstly, the engine model must be generated and consists of the crank train and cylinder block. The GUI is designed such that the user builds the model in a sequential order by using a series of forms. Once the key engine parameters have been defined ENGDYN draws the reduced model of the crank train that the user can pick to edit particular items. Finite element models of the crank train and cylinder block can be viewed with the reduced models. ENGDYN performs its own finite element solution, to evaluate for example crankshaft web stiffness’. Once the model has been generated, the GUI can be used to generate the input command file for performing a solution. This file contains the engine conditions to be simulated, the type of solution to be performed, the models to be included in the solution and the solution parameters for controlling the solution. The solver performs the simulation and appends the results to the binary standard data file. An ASCII file is also written summarising the results of the solution for each load case. Finally, once the simulation has finished running, the user is able within the GUI to select load cases for post-processing. The interface lists all the load cases within the file and for each load case gives the solution parameters defined for that case. The interface currently allows simulation results to be plotted, animated and exported to an ASCII file. 1


The interface has a built in library of pre-defined plots that enables the user to rapidly view the results. The plots can either be printed or saved to file. The user is also able to export results to an ASCII file to allow further data manipulation by the user not currently supported by ENGDYN. In addition, for solutions in which the displacement or vibratory motion of the engine is predicted, the results can be animated interactively in both the time and frequency domains. In addition, the loads calculated by ENGDYN can be used to evaluate the stresses and fatigue safety factors at the critical locations of the crankshaft and can also be applied to the cylinder block either in the frequency domain to perform modal frequency response and acoustic analyses or in the time domain to perform quasi-static analyses. The user can then perform further analyses if required. A new solution can be executed and on completion the results are appended to the binary standard data file.

2

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis

2 Tutorials 2.1 Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft

Analysis 2.1.1

Introduction

Objective:  Gain a general understanding of how to use ENGDYN GUI and carry out the simplest analysis using user defined (non-FE) data.  The user will perform statically determinate analysis to calculate loads for both a Mobility and Hydrodynamic (HD) bearing analysis  User will also gain experience in using plotting and animation features. Items covered:    

Building the engine model Bearing analysis using the Mobility Method Bearing analysis using the Hydrodynamic Method (Finite Volume Solver) Plotting and Animation

Estimated duration:  <0.5 days (overall including performing solutions) Engine:      

Audi V6 TDI V6 High speed direct injection diesel 90 degree vee-angle 30 degree crank-pin offset 4 main bearings 6 big end bearings

3


 The data for this tutorial was obtained from a benchmark study of this engine by Ricardo. Required Files: ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\v6\cp700.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\v6\cp1000.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\v6\cp2000.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\v6\cp4000.PRES

2.1.2

Getting Started

 Copy the necessary files from the example directory to a working directory and ensure that you have write permissions for all the files.  Start engdyn  On Unix or Linux platforms simply type engdyn  On windows click on the shortcut, otherwise go to Start>Programs>Ricardo Software>2014.1>Mechanical Suite>ENGDYN>ENGDYN

2.1.3

Building engine model

2.1.3.1 Configure Engine  Select ‘Configure Engine’ from the buttons on the left side of the Main Panel

 Complete the Engine Configuration Panel as shown

4

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Note that warning messages will pop up as long as all the needed data is not properly entered. If you close the Engine Configuration Window while warning messages still appear, all the data will be lost. Therefore, please make sure all the data is properly entered before closing this window.

5


6


 To enter the main bearings non-uniform information, the user must use the mouse to click on the red dimensions in the diagram in the panel.  On completion select ‘Apply’ which will display

 Press OK button. The crankshaft model will appear in the Main Panel as shown  Use CTRL + middle mouse in order to zoom in and zoom out  Use SHIFT + middle mouse in order to rotate

2.1.3.2 Define Models  Select ‘Define Models’ from the buttons on the left side of the GUI.

7


 Complete the model definitions for the crankshaft, cylinder block, in-cylinder and connecting rod as shown

2.1.3.3 Edit Cranktrain  Select ‘Edit Cranktrain’ from the buttons on the left side of Main Panel.

 Items are edited by selecting from the list on the left of the panel  Many items shown in the panel below need not be defined due to the type of model being edited.  Highlight ‘Crank Web’ as shown  This will then display each of the web elements in green  Click on ‘Select All’

8

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Each of the web elements will then turn red

 Click on ‘Edit Selected’ to display the Web Panel  Complete the ‘Web’ panel as shown below

 The web thickness and journal lengths are taken from the data entered in the ‘configure engine’ section  The Web panel allows counterweight data to be entered for the crank webs  The counterweight geometry is defined only for balancing the crankshaft at a later stage, NOT for defining the mass properties. As we shall be defining the mass properties explicitly in this example (rather than using an FE model), we do not require any counterweight data to be added  Select ‘OK’  Highlight ‘Big End Bearing’  This will then display each of the pin journal nodes in green  Click on ‘Select All’  Each of the main journal nodes will then turn red

9

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Click on ‘Edit Selected’ to display the Pin Bearing Panel  Complete the ‘Bearing’ panel as shown below

 The oil hole angular position can be defined either using the Height column (as in this case) or using the Position column. The value not supplied is calculated from the journal diameter.  The remaining tabs (Mesh, Material, Profile) need not and cannot be edited because the model type is ‘Mobility’. These tabs are only required with Hydrodynamic or Elastohydrodynamic models  Select ‘OK’  Highlight ‘Main Bearing’  Click on ‘Select All’ to highlight all main bearing nodes  Click on ‘Edit Selected’ to display the Main Bearing Panel  Complete the Main Bearing Panel as shown

10


 The ‘Oilholes’ tab cannot be edited because the bearing is fed by a groove.  The remaining tabs (Mesh, Material, Profile, Stiffness) need not and cannot be edited because the model type is ‘Mobility’. These tabs are only required with Hydrodynamic or Elastohydrodynamic models  Select ‘OK’  Highlight ‘Connecting Rod’  Click on ‘Select All’ to select all pin journal nodes  Click on ‘Edit Selected’  Complete the ‘Conrod’ panel as shown.

 Select ‘OK’  Highlight ‘Piston’

11

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Click on ‘Select All’ to select all pin journal nodes  Click on ‘Edit Selected’  Complete the ‘Piston’ panel as shown.

 Highlight ‘Cranknose Assembly’  In this instance since there is only one cranknose there is no need to use Select All or to pick using the mouse. The program automatically selects the cranknose and displays it in red  Select ‘Edit Selected’ to display the Cranknose Assembly Panel  Complete the Cranknose Assembly Panel as shown

 The ‘Element Length’ field does not have to be entered accurately because the crankshaft is not modelled using an FE model. This data will only be used for showing a representative model on the screen.  Select ‘OK’  Highlight ‘Flywheel Assembly’

12

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Again there is no need to use Select All or to pick using the mouse. The program automatically selects the cranknose and displays it in red  Select ‘Edit Selected’ to select the Flywheel Assembly Panel  Complete the ‘Flywheel Assembly’ panel as shown

 Select on ‘OK’  Highlight ‘Lumped Masses’ 13

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  The nodes are numbered from the front of the crankshaft, number one representing the damper hub  For a rigid or compliant crankshaft model it is not necessary to define masses at all the nodes  Select node 3 using the left mouse button.  This is the web node of the first web on the crankshaft axis.  Alternatively nodes can be picked by dragging the mouse with the left button held down  To make multiple selections, hold the SHIFT key down whilst selecting nodes  Select Edit Selected to display the Lumped Mass Panel as shown  Complete the panel as shown

 Data can either be defined with known properties or using a geometric shape and can be with respective to a Cartesian or Polar coordinate system.  Select the Add or Update button to add the data  Multiple masses can be added to each node  Select OK  The mass will be shown in white

14

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Repeat this process defining the masses as listed in the table below. Node

3 5 8 10 12 14 17 19 21 23 26 28

Description

Web 1 counterweight Pin 1 mass Pin 2 mass Web 2 counterweight Web 3 counterweight Pin 3 mass Pin 4 mass Web 4 counterweight Web 5 counterweight Pin 5 mass Pin 6 mass Web 6 counterweight

Mass [kg] 2.610 0.408 0.408 1.500 1.760 0.408 0.408 1.760 1.500 0.408 0.408 2.630

x [mm] 0.470 -0.875 0.875 -0.075 0.300 -0.875 0.875 -0.050 0.297 -0.875 0.875 -0.051

Offset y [mm] -23.220 0.000 0.000 -20.460 13.090 0.000 0.000 12.280 7.215 0.000 0.000 10.140

z [mm] 10.270 0.000 0.000 7.247 12.280 0.000 0.000 13.100 -20.510 0.000 0.000 -22.720

When all lumped masses have been entered the model should appear as shown

 Click on ‘Define Material’ to display the Crankshaft Material Properties Panel

15

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Complete the Crankshaft Material Properties Panel as shown

 Select on ‘OK’  Click on ‘Calculate Masses’  The message ‘Calculation of mass properties completed successfully, Balance Not Set’ should appear in the message box at the bottom of the Main Panel.  Click on ‘Set Balance’ to display the Primary Balance Panel as shown

 No additional balancing is required because the Lumped Masses include the effect of balancing the crankshaft.  Click on ‘Assemble Model’  The message ‘Model assembly completed successfully, Block not assembled’ should appear in the message box at the bottom of the Main Panel  Click on ‘OK’  Click on the ‘Dismiss’ button of the Edit Cranktrain panel

16

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  The message ‘Ready to Perform Solution’ should appear in the message box at the bottom of the Engdyn screen  Save the model

2.1.4

Bearing analysis using the Mobility Solution

2.1.4.1 Define Lubricant Properties  Click on ‘Lubrication’ button from the buttons on the left hand side of the Main Panel

 Use the Browse button to select the lubricant SAE5W30 from the database  By default the program should initially select the database directory at ..\Ricardo\2014.1\Common\Materials\Lubricants  This database contains the most common lubricants  Use either Add or Update to add the lubricant

 Click on OK 2.1.4.2 Define Loading Conditions The cylinder pressure diagram and any additional loadings (e.g., loads from Valdyn and gravity forces) are entered in this step.

17

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Click on the ‘Loading’ button on the Main Panel

This will display the Loading Definition Panel as shown

 A number of different loading maps can be defined, Full Load, Part Load and No Load.  The solver will interpolate at speeds between those defined using this panel  Type in a speed of 750 rev/min  Position the mouse over the File Name column and use the right button to display the pop-up menu. Use ‘Select Pressure file’ to select the file cp700.PRES from your working directory. The panel will appear as shown.

 You may wish to remove the pathname in front of the file.

18

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  The pressure file can contain just pressures (at equal intervals) or as in this case pressures and angles.  You may wish to inspect and understand the file using an appropriate editor  Define the Ambient and Crankcase Pressure as 1.0 bar  The crankcase pressure is only currently used for Hydrodynamic and Elastohydrodynamic bearing models to define the boundary condition at the edge of the bearing  Add an extra line to the table by positioning the mouse over the left column (line number) and using the right button to display a pop-up. Select ‘Insert Row After’  Complete the panel as shown

 The remaining tabs, Force Profile, Force Equation and Distortion need not be completed for this tutorial.  When a model has a piston pin offset or crank offset, the cylinder pressure diagram is often redefined by ENGDYN internally to have the interval angle reduced to 0.25 degrees (sometimes lower) by a process of interpolation. This is because the offsets cause the TDC angle to change slightly and it is important to have an accurate cylinder pressure definition at TDC to achieve good results.  Use the Plot button to display the applied loads and to plot for example the Indicated Torque as shown

19


 This diagram is not particular smooth. If the indicated torque curve is known then you can factor each diagram using the Factor column.  Enter 400[N.m]/T at 3000 rev/min and see the graph change. Set back to 1.0 before proceeding  Click on ‘OK’  Save the model using the File menu from the top of the Main Panel 2.1.4.3 Evaluate Solution We now have sufficient data to proceed with an analysis.  Click on the ‘Evaluate Solution’ button on the Main Panel

20


 The panel will appear as shown. No changes are required  We can only perform a Determinate solution because we have only built a rigid model  Select the Cases tab  Click on ‘Select’ button to add an arithmetic speed sweep series of engine speeds to the loadcases

21

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Select the Bearing Model tab  Complete the panel as shown

 The remaining tabs do not need to be completed for a solution with a rigid block and crankshaft.  Select Solve Directly.

 The analysis should take a few seconds to run.  On completion of the analysis the summary file .EDSUM will be written.  This file contains summary data for the solution. Open this file with an appropriate editor to view the results. Now that the analysis had been completed, the results can be plotted. The next two steps show how to plot the results of the completed analysis. 2.1.4.4 Select Loadcases

22

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Click on the ‘Select Loadcases’ button on the Main Panel

 Select the loadcases as shown

 This table displays the solution parameters defined for each loadcase  Loadcases are grouped by Solution Type and by Loading 2.1.4.5 Plot Results  Click on the ‘Plot results’ button on the Main Panel

 Select ‘Journal Bearing’ from the Model list and highlight the Subset, Plot and Results.

23


 Move the panel so that the crankshaft can be seen on the screen as shown  As with editing the cranktrain each of the main journal nodes are displayed in green  Select main bearings 1 and 2 by dragging the mouse whilst keeping the left button held down  Each of the main journal nodes will then turn red  Alternatively select the nodes by using the Select All button or by selecting each node individually and using the SHIFT key  Click on the Apply button to display the Graph Panel.  Use the Page Up and Page Down buttons to view the following graphs

24


25


26


27


28


29

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis 2.1.4.6 Exporting the data

In addition to the pre defined plots ENGDYN also has the option of exporting formatted data into ASCII files  Click on the Export Results button on the left hand side of the software and the following panel appears

As an example, we are going to export the journal bearing load at the main bearing 1.  Select JOURNAL_BEARING_LOADS in the dataset and Main bearings on the subset  Select the bearing 1 in the 3D viewport  Name the ASCII file with something suitable, such as ‘bearing_load’  Change format to block and select only Y & Z directions  Click Apply

30

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis Two new .EDRES files, for loadcase 1 & 4, were written to your directory. Their content can be reviewed through an appropriate text editor.

2.1.5

Bearing analysis using the Hydrodynamic Solution

Main bearing number 1 will be modified to solve using the Hydrodynamic model 2.1.5.1 Copy the ENGDYN Model to a new filename  Use the File menu on the Main Panel to ‘Copy Design’  This will copy all model and loading data 2.1.5.2 Set up Main Bearing 1 as a Hydrodynamic Model  Click on the ‘Edit Cranktrain’ button on the Main Panel to display the Edit Cranktrain Panel

 Highlight ‘Main Bearing’ as shown

31


 This will then display each of the main journal nodes in green  Select main bearings 1 by dragging the mouse whilst keeping the left button held down or by picking the node.  The main journal nodes will turn red  Select ‘Edit Selected’ to display the Main Bearing Panel Change the Model Type to Hydrodynamic as shown

32

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis For a Hydrodynamic Model Type it is necessary to use the Mesh and Material tabs to define additional data for this model.  In this tutorial we will assume the bearing and journal are circular, and therefore it is not necessary to define a profile using the Profile tab.  Select the Mesh tab, and define a mesh 11 x 73  Select the Material tab

 This material is used for the boundary lubrication model  It is necessary to define the journal material and the material of the bearing lining  Click on Define adjacent to Bearing Material, to display the Material Properties Panel.  Complete the panel as shown

 These data values are typical for a bearing surface.  These data are only used by the boundary lubrication model. 33

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Asperity RMS Height h, Density  and Asperity radius  can be calculated from measured data using the Ricardo MATUTIL program supplied with the Ricardo Software installation which is described in Appendix 12.  The Select button is used to select materials that have previously been defined or that are in the SFE file of finite element model defining the bearing.  Select OK  Click on Define adjacent to Journal Material, to display the Material Properties Panel as shown.

 The journal material defaults to the crankshaft material STEEL which was defined in Section 2.1.3.3  Complete the panel as shown  The asperity data can be calculated from surface profile measurements using the MATUTIL program supplied with the Ricardo Software installation which is described in Appendix 12.

 Select OK and enter the wear and friction coefficients for the bearing surface as shown

34


 Select OK  Click on ‘Define Material’ to display the Crankshaft Material Properties Panel   Select OK  No edits are required since we have defined the material previously.  Click on ‘Calculate Masses’  The message ‘Calculation of mass properties completed successfully, Balance Not Set’ should appear in the message box at the bottom of the Main Panel.  Click on ‘Set Balance’  No additional balancing is required because the Lumped Masses include the effect of balancing the crankshaft.  Click on ‘Assemble Model’  The message ‘Model assembly completed successfully, Block not assembled’ should appear in the message box at the bottom of the Main Panel  Click on ‘OK’  Click on the ‘Dismiss’ button of the Edit Cranktrain panel  The message ‘Ready to Perform Solution’ should appear in the message box at the bottom of the Engdyn screen  The model should appear as shown.

35


 Save the model using the File menu from the top of the Main Panel 2.1.5.3 Evaluate Solution We now have sufficient data to proceed with an analysis.  Click on the ‘Evaluate Solution’ button on the Main Panel

The Evaluate Solution Panel will appear.

36


 Set the End Angle to 1440 deg  This is equivalent to two and a half cycles and should be sufficient to obtain a converged solution of the main bearing HD solution  Select the Cases tab and define a single speed of 3000 rev/min as shown

 Select the Bearing Model tab and complete the panel as shown

37


 The remaining tabs do not need to be completed for this solution since the other parameters are related to indeterminate or dynamic analyses.  Select ‘Define Oil Temps’ to display the Bearing Oil Temperatures Panel

 We will assume that the temperature of the oil in the bearing is at the inlet temperature.  Select OK  Select Solve Directly on the Evaluate Solution Panel.

38

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  On completion of the analysis the summary file .EDSUM will be written.  This file contains summary data for the solution. Open this file with an appropriate editor to view the results.  The solution time for this solution on a Pentium M 1.4 GHz Laptop with 512 Mbytes RAM is approximately 9 minutes.  Solution time will be dependent on how heavily loaded the bearing is and whether there is any boundary lubrication occurring.  This model at this engine condition has some asperity contact which can be viewed by selecting Contact Pressure when animating (See 2.1.5.5)  The memory usage is 67 Mbytes. 2.1.5.4 Select Loadcases  Click on the ‘Select Loadcases’ button on the Main Panel

 Highlight the only speed selectable (3000 revs/min) and then select OK 2.1.5.5 Animate Results  Click on the ‘Animate Results’ button on the Main Panel

 Complete the ‘Animation Results’ panel as shown

39


 Zoom onto main bearing 1  Hold down both the ‘Ctrl’ button on the keyboard and the middle mouse button whilst dragging the mouse.  Change the angle of the view  Hold down both the ‘Shift’ button on the keyboard and the middle mouse button whilst dragging the mouse.  Click on the ‘Play’ button of the animation control toolbar to animate the results as shown.

40

KNOWLEDGE CENTRE Tutorials Tutorial 1: An Introduction to ENGDYN – Concept Crankshaft Analysis  Experiment with looking at different results by selecting various parameters from the Animation Results Panel  Close the animation panels and exit the program.

41

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis

2.2 Tutorial 2: Crankshaft Dynamic Analysis 2.2.1

Introduction

Objective:  To build an ENGDYN model suitable for performing a dynamic analysis of a crankshaft.  Introducing FE models of the crankshaft  The results of this tutorial are used as input to the NVH and Crankshaft Stress tutorials Items Covered:  Building the engine model  Introducing an FE model of the crankshaft  Matrix Reduction  Dynamic solution Estimated duration:  0.5 day (model preparation)  1 day (overall including performing solutions) Engine:  Inline 4 gasoline engine  75.0 x 44.75 mm  1.6 litre Required Files: ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4_crank.inp ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4_2000.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4_3000.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4_4500.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4_5000.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4_5500.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4_6000.PRES Finite element model requirements:

43

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis  The crankshaft FE model should be modelled as described in the “Geometry of the Finite Element Model” Section (Section B.2.4.5.2 in the “Help” part of the manual”)

2.2.2

Getting Started


2.2.3

Building the Model

2.2.3.1 Configure Engine In order to build an ENGDYN model it is first necessary to define the major dimensions and features of the engine. This is done using the Engine Configuration Panel.  Select ‘Configure Engine’ from the buttons on the left side of the Main Panel.



Complete the Engine Configuration Panel as shown.

44


 The Firing Order can be selected using the Options button or supplied as a hyphen separated list as shown  Journal lengths are the lengths of the journals including the journal fillets NOT the bearing lengths which are defined later  All data in this example are uniform. In cases where data are non-uniform is selected. This will display in red on the 2D graphic those items that can be edited by selecting with the mouse. Items that are not shown are edited using the Edit button.  On completion select ‘Apply’ which will display

 Select OK. The crankshaft model will appear in the Main Panel as shown

45


 Note that the dimensions of the cranknose and flywheel assemblies on screen are only representations as no data has yet been entered for these. These are defined in 2.2.3.3. 2.2.3.2 Define Models An ENGDYN model consists of a number of sub-models which are defined using the Model Definitions Panel.  Select ‘Define Models’ from the buttons on the left side of the Main Panel.

 Complete the model definition for the Crankshaft

46


 Use default values for Model, Data and Matrix Formulation  The crankshaft in this example is already in the same co-ordinate system and orientation as defined in the 2.2.3.1. No transformation is therefore necessary  Click on Browse button to display the ‘Model Translation’ Panel as shown

 This panel is used to translate FE models to Ricardo SFE format  Models can alternatively be translated outside the graphical interface using the appropriate FEARCE translator. If the user does this then Origin is left as Ricardo-SFE.  The crankshaft in this example is already in the same co-ordinate system and orientation as defined in the 2.2.3.1. No transformation is therefore necessary

>

47

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis  Set origin to ABAQUS and Units to mm  Units is the units of the ABAQUS model  During the translation the model will be translated so that the SFE file is in SI units.  Use the Browse button to select the file il4_crank.inp from the working directory

 The program will automatically set the output name to il4_crank.SFE although this can be changed  Click Translate to translate the model from the ABAQUS format into SFE format  Select OK to close the panel.  On completion the message Translation successful will appear at the bottom of the Main Panel.  Then there will be a SFE file named il4_crank.SFE in the current directory.  Go to the current directory, find the file il4_crank.SFE  Double-click on this file and the RICARDO FE viewing and interface tool (RDESK) will launch showing our crankshaft FE model

Note that on UNIX/LINUX systems, type rdesk into the shell at the working directory.

48

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis The R-Desk viewport can be controlled in a similar way to ENGDYN’s. To move the model in the panel,  Middle mouse and move = translation  + middle mouse and move = rotation  middle mouse scroll wheel = zoom in/out  Select the Cylinder Block tab and define the Model as Rigid with User Defined data as shown.

 Click on OK. This will display the model as shown.

 By default only the edges of the FE model of the crankshaft will be displayed as shown. To change the appearance of the model select the Model Appearance Panel from the View Menu.  Note that the webs and journals align with the reduced model (green), but that the cranknose and flywheel assemblies are not yet correctly defined. This will be addressed in the next step

49

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis 2.2.3.3 Editing the Cranktrain The next step is to define the data related to the cranktrain.  Select ‘Edit Cranktrain’ from the buttons on the left side of the Main Panel to display the Cranktrain Tool Panel

 Given we defined the crankshaft as Dynamic (See 2.2.3.2) it will be necessary to define everything listed on the left hand side of the panel except Lumped Mass and Mechanical Links  Lumped Mass is used to define any additional masses that are not included in the FE model or defined using Flywheel Assembly or Cranknose Assembly.  Mechanical Links is used to define links for co-simulation. Highlight ‘Crank Web’ as shown  This will then display each of the crank web numbers and elements in green.  Select webs 1, 2, 7 and 8 by dragging the mouse with the left mouse button held down.  Each of the selected webs will then turn red as shown

50


 Select ‘Edit Selected’ to display the Web Panel  Set Counterweight to Present and complete the counterweight data as shown

51


 The counterweight geometry is defined only for the purposes of balancing the crankshaft in 2.2.3.4 NOT for defining its mass properties.  The angles are with respect to the Engdyn coordinate system not with respect to the web.  The journal and thickness data has been calculated from the data supplied in 2.2.3.1.  The stiffness data is calculated in 2.2.3.5  Select OK  Select webs 3, 4, 5 and 6 by dragging the mouse with the left mouse button held down.  Select ‘Edit Selected’ to display the Web Panel  Set Counterweight to Present and complete the counterweight data as shown

52


 Select OK  Highlight Pin Journal Element  This will then display each of the pin journal elements in green  Select ‘Select All’  Each of the pin journal elements will then turn red  Select ‘Edit Selected’ to display the Element Panel as shown

53

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis  This panel allows the internal diameter of a hollow journal to be defined. The pin journals of this crankshaft are solid.  The Outside Diameter was defined during 2.2.3.1.  Select OK  Highlight Main Journal Element  This will then display each of the main journal elements in green  Select ‘Select All’  Each of the main journal elements will then turn red  Select ‘Edit Selected’ to display the Element Panel as shown

 As with the pin journals this panel allows the internal diameter of a hollow journal to be defined. The main journals of this crankshaft are solid.  The Outside Diameter was defined during 2.2.3.1.  Select OK  Highlight Big End Bearing  This will then display each of the cylinder numbers and pin journals in green.  Select ‘Select All’  Each of the pin journals will then turn red  Select ‘Edit Selected’ to display the Bearing Panel  Complete the Bearing Panel as shown

54


 Given the Model Type is Mobility and the bearings are plain it is not necessary to enter any other data than shown.  These bearing are feed from the journal via a feed from the adjacent main bearings.  The oil hole angular position can be defined either using the Position column (as in this case) or using the Height column. The height is calculated from the angle and vice-versa.  The remaining tabs (Mesh, Material, Profile) need not and cannot be edited because the model type is ‘Mobility’. These tabs are only required with Hydrodynamic or Elastohydrodynamic models  Select OK  Highlight Main Bearing  This will then display each of the bearing numbers and main journals in green.  Select ‘Select All’  Each of the main journals will then turn red  Select ‘Edit Selected’ to display the Main Bearing Panel  Complete the Main Bearing Panel as shown

55


 Given the Model Type is Mobility and the bearings are partial grooved it is not necessary to enter any other data than shown.  Any journal oil holes of a grooved bearing are not currently considered.  The remaining tabs (Mesh, Material, Profile) need not and cannot be edited because the model type is ‘Mobility’. These tabs are only required with Hydrodynamic or Elastohydrodynamic models  Select OK  Highlight Thrust Bearing  This will then display the single main journal node in red corresponding to the thrust bearing defined in 2.2.3.1. There is therefore no need to ‘Select All’  Select Edit Selected to display the Thrust Bearing Panel  Complete the panel as shown

56


 Supply geometric extent of the axial thrust face and the bearing clearance  ENGDYN will calculate appropriate stiffness/damping based upon the system material properties  Select OK  Highlight ‘Cylinder’  Click on ‘Select All’ to select all pin journal nodes  Click on ‘Edit Selected’  Complete the ‘Cylinder’ panel as shown.

 The height Head Height is used to derive a node for the cylinder block model in 2.2.3.6.  Select ‘OK’  Highlight ‘Connecting Rod’  Click on ‘Select All’ to select all pin journal nodes  Click on ‘Edit Selected’  Complete the ‘Conrod’ panel as shown.

57

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis  Note the units for the connecting rod inertia!  Select ‘OK’  Highlight ‘Piston’  Click on ‘Select All’ to select all pin journal nodes  Click on ‘Edit Selected’  Complete the ‘Piston’ panel as shown.

 The skirt and top ring data has been set to 0 since this data are only required for block stress analysis and is therefore not required for this tutorial.  Select OK  Highlight Cranknose Assembly  This will then display the cranknose assembly in red. Therefore there is no need to ‘Select All’  Select Edit Selected to display the Cranknose Assembly Panel  Complete the panel as shown

58


 The Element Length effectively positions the first node of the reduced model of the crankshaft. The position of this node must correspond to a plane of nodes in the FE model.  The natural frequency of the vibration damper is 123.3 Hz.  The seismic mass of the damper defined here must not be included in the FE model of the crankshaft The hub of the vibration damper may be included in the FE model or as in this case defined using this panel.  Any other additional masses, for example gears, sprockets and pulleys, not included in the FE model can be defined by selecting Lumped Mass on the Edit Cranktrain Panel. For the purposes of this exercise we will assume there is no additional masses.  Select ‘Edit’ button adjacent to Hub Data to display the Lumped Mass Panel  Complete the Panel as shown.

59


 This defines the mass and inertia of the hub.  Given X offset is 0 this lumped mass is assumed to be at the node (defined by the element length on the Cranknose Assembly Panel)  Click on OK  Click on OK on the Cranknose Assembly Panel  Highlight Flywheel Assembly  This will then display the flywheel assembly in red. Therefore there is no need to ‘Select All’  Select Edit Selected to display the Flywheel Assembly Panel  Complete the panel as shown

60


 The distance X1 effectively positions the last node of the reduced model of the crankshaft. The position of this node must correspond to a plane of nodes in the FE model.  In this example, by setting Type to Conventional, we are assuming the flywheel to be rigid.  The Offset is with respect to the node defined by X1.  The inertia data is with respect to its centroid.  Click on the Clutch tab and enter the mass and inertia data for the clutch.

 The other tabs to not apply to a conventional flywheel assembly  Select OK There are no additional lumped masses to be defined or mechanical links to be defined.  Select Define Material to display the Crankshaft Material Properties Panel as shown

61


 The material data will be read from il4_crank.SFE.  Use select to list the materials in the SFE file and select the material STEEL from the list as shown.

 This table lists all the material in the SFE file.  The selected material is the material of the web overlap section.  Materials used by ENGDYN and those stored in the .SFE are identified by a single unique name.  Select OK and the material is added to the panel as shown

 These data may be edited if required  Select OK  These data are now stored by ENGDYN and also written to the .SFE file.  Previous data will be overwritten.  The message “Mass Not Calculated’ will be written to the Main Panel.

62

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis 2.2.3.4 Calculating the Mass Properties and Balancing the Crankshaft  Click on ‘Calculate Masses’ to calculate lumped mass properties at each reduced node  The progress bar as shown will be displayed

 The program derives an element set (as described 2.4.5.3 of the manual) for each reduced node. These sets can be viewed using FEVIEWER. The mass and inertia of each set is then assigned to the appropriate node.  On completion the message ‘Balance Not Set’ will be written to the Main Panel.  Click on ‘Set Balance’ to display the Primary Balance Panel as shown

63

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis  This panel shows the primary imbalance of the model as given by ‘As modelled’. This is the imbalance of the FE model and any additional lumped masses.  The imbalance will invariably be dependent on the accuracy of the finite element model.  For an in-line 4 engine we would expect the crankshaft to have primary balance.  This panel allows the user to simulate balancing the crankshaft by drilling the counterweights.  Set Calculation to ‘Using Web data’  Use select adjacent to Web Numbers to display the Select Web Panel as shown

 This displays all webs that have counterweights as defined in 2.2.3.3  Select webs 3, 4, 5, and 6 as shown

 Select OK  Use Define adjacent to Material Name to display the Counterweight Material Properties as shown

64


 This by default will display the crankshaft material properties as defined in 2.2.3.3.  Select can be used to select a different material for the counterweight material.  Click on OK.  The Primary Balance Panel will appear as shown.

 For an in-line 4 engine the ‘Required’ balance is 0  Click on Apply to balance the crankshaft  The message ‘Balance calculation completed successfully’ will be written to the Main Panel.  Click on OK  The message ‘Balance Set, Stiffness Not Calculated’ will be written to the Main Panel.

65

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis  Scroll up on the messages on the Main Panel to see the output related to the balance drillings.

2.2.3.5 Matrix Reduction  Click on ‘Calculate Stiffnesses’ to calculate the crank web and element stiffness using the finite element method.  2 solvers are available. The default Vectorized Sparse Solver (VSS) or the Symmetric Conjugent Gradient  Click on ‘No’ to select the VSS solver  The program splits the FE model of the crankshaft into a number of submodels. FE calculations are performed on each model.  These calculations take circa 5 minutes on am Intel I7-2620M 2.70 GHz Laptop with 4 GB RAM.  On completion of the finite element solutions the reduced mass and stiffness matrices of the crankshaft are derived and an eigenvalue solution performed.  The message ‘Stiffness Calculated, Model Not Assembled’ will be written to the Main Panel on completion.  Inspect the contents of your working directory to see the files generated.  Click on ‘Calculate Stiffnesses’ again  This time a query panel will be displayed as shown

 The program checks whether finite element solutions for each web and element already exist.  Click on No  No finite element solutions will be performed  The reduced mass and stiffness matrices of the crankshaft are derived and an eigenvalue solution performed. Click on Assemble Model  This verifies the validity of the data  The message ‘Model Assembly completed successfully. Block not assembled’ will be written to the Main Panel if assembly was successful.  Click on Dismiss on the Cranktrain Tool Panel.  The main panel will now appear as shown

66


2.2.3.6 Assembling the Cylinder Block Model The next step is to assembly the cylinder block model.  Select ‘Edit Block’ from the buttons on the left side of the Main Panel to display the Cylinder Block Tool Panel  The Main Panel will appear as shown

67


 The program creates nodes at each main bearing and at the reversal positions of the small end of the connecting rod (Thrust and Anti-Thrust) and at the centre of the cylinder head for each cylinder. These nodal positions are derived from data defined in 2.2.3.1 and 2.2.3.3.  Simply click on Assemble Model.  Click on OK  The model is now completed and should appear as shown

68

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis  Save the model using the File menu from the top of Main Panel

2.2.4

Solution

2.2.4.1 Define Lubricant Properties  Click on ‘Lubrication’ button from the buttons on the left hand side of the Main Panel

 Use the Browse button to select the lubricant SAE5W30 from the database  By default the program will initially select the database directory at ..\Ricardo\2014.1\Common\Materials\LubricantsThis database contains the most common lubricants  Use either Add or Update to add the lubricant



Click on OK

2.2.4.2 Define Loading Conditions The cylinder pressure diagram and any additional loadings (for example loads from Valdyn and gravity forces) are entered in this step.  Click on the ‘Loading’ button on the Main Panel

69

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis This will display the Loading Definition Panel as shown

 A number of different loading maps can be defined, Full Load, Part Load and No Load.  The solver will interpolate at speeds between those defined using this panel.  Type in a speed of 2000 rev/min  Position the mouse over the File Name column and use the right button to display the pop-up menu. Use ‘Select Pressure file’ to select the file il4_2000.PRES from your working directory. The panel will appear as shown.

 You may wish to remove the pathname in front of the file.  You may wish to inspect and understand the selected file using an appropriate editor  Define the Ambient and Crankcase Pressure as 1.0 bar

70

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis  The crankcase pressure is only currently used for Hydrodynamic and Elastohydrodynamic bearing models to define the boundary condition at the edge of the bearing  Position the mouse over the column number and with the right button depressed display the Row Operations pop-up. Select Insert After. The Panel will appear as shown

 Define the pressure diagram at 3000 rev/min as shown

 Complete the panel as shown.

71


 The remaining tabs, Force Profile, Force Equation and Distortion need not be completed for this tutorial.  Use the Plot button to display the applied pressure loading as shown

 Save the model using the File menu from the top of the Main Panel We now have sufficient data to proceed with an analysis 2.2.4.3 Evaluate Dynamic Solution  Click on the ‘Evaluate Solution’ button on the Main Panel

72

KNOWLEDGE CENTRE Tutorials Tutorial 2: Crankshaft Dynamic Analysis The Evaluate Solution Panel will appear.  Set the solution Type to Dynamic and edit the solution tolerances as shown

 Select the Cases tab and click on Select to define a speed sweep at full load between 2000 and 6000 rev/min in steps of 250 rev/min as shown

73


 Click on OK to populate the table on the Evaluate Solution Panel as shown

 Select the Model Options tab and change the Cylinder Damping to 500 N.s/m and Frequency to 50 Hz as shown.

74


 The cylinder damping parameter is used to tune the torsional response of the crankshaft  The coupling is a torsional spring and damper to prevent rigid motion of the crankshaft about its axis. This is analogous to the stiffness and damping of a dynamometer connected to the crankshaft.  The damping value of the coupling can be used to tune the torsional response of the crankshaft.  The natural frequency of the coupling must be significantly lower than the first flexible torsion mode of the crankshaft.  Select the Bearing Model tab and complete the panel as shown

75


 Select ‘Define’ to display the Bearing Oil Temperatures Panel

 We will assume that the temperature of the oil in the bearing is at the inlet temperature  Select OK  Select Solve Directly on the Evaluate Solution Panel.

76


 On completion of the analysis the summary file .EDSUM will be written.  This file contains summary data for the solution. Open this file with an appropriate editor to view the results

77

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis

2.3 Tutorial 3: NVH Analysis 2.3.1

Overview

Objective:  To use ENGDYN to perform vibration and acoustic analysis on a powertrain  The analysis uses finite element (FE) models of both the powertrain assembly and the crankshaft  The tutorial requires the ENGDYN model from the Crankshaft Dynamic Crankshaft Tutorial Items Covered:  Dynamic cylinder block model  Engine mount modelling  Component Mode Synthesis (CMS) matrix reduction  Vibration Analysis  Acoustic Analysis  Rayleigh Integral Method  Indirect Boundary Element Method (BEM) Estimated duration:  0.5 day (model preparation)  1 day (overall including performing solutions) Engine:  1.6 litre Inline 4 gasoline engine  75.0 x 44.75 mm Required Files: ..\Ricardo\2014.1\Products\ENGDYN/Tutorials\il4\IL4_BLOCK.SFE ..\Ricardo\2014.1\Products\ENGDYN/Tutorials\il4\il4_block_noise.VPT ..\Ricardo\2014.1\Products\ENGDYN/Tutorials\il4\il4_block_vibration.VPT ..\Ricardo\2014.1\Products\ENGDYN/Tutorials\il4\IL4_BLOCK_BoundaryElementM odel.SFE The .EDSF file from Crankshaft Dynamic Analysis Tutorial, so all the files used in tutorial 2 are needed here. Finite element model requirements:  For dynamic and acoustic analysis of the powertrain it is necessary to have a fully dressed finite element model with engine mount brackets to support the engine

79

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis mounts. The model used in this tutorial is not representative but is sufficient for demonstrating the analysis steps.

2.3.2

Getting Started

 Go to the working directory in which you completed the Crankshaft Dynamic Analysis Tutorial. Copy the file IL4_BLOCK.SFE from the Tutorials directory in the installation to this working directory and ensure that you have write permissions for this file.  Start engdyn  On Unix or Linux platforms simply type engdyn  On windows click on the shortcut, otherwise go to Start>Programs>Ricardo Software>2014.1>Mechanical Suite>ENGDYN>ENGDYN In the Crankshaft Dynamic Analysis Tutorial the cylinder block (or powertrain) was modelled as rigid. Obviously for analysing the vibration and acoustic behaviour of the powertrain it is necessary to model the powertrain as dynamic. This section describes the steps in updating the model derived in the previous tutorial.  Open the .EDSF file used for the Crankshaft Dynamic Analysis Tutorial using the File menu at the top of the Main Panel  From the same menu ‘Copy Design’ to a new file

2.3.3

Updating the Model

2.3.3.1 Define Models In this step we need to redefine the cylinder block model as dynamic rather than rigid.  Select ‘Define Models’ from the buttons on the left side of the Main Panel to display the Model Definitions Panel.

 Select the ‘Cylinder Block’ tab and define the model as Dynamic as shown.

80


 Click on Browse button to display the ‘Model Translation’ Panel as shown

 Use the Browse button to select the file IL4_BLOCK.SFE from the working directory.

 The program will automatically set the output name to IL4_BLOCK.SFE although this can be changed  Click on the Transformation tab to define the translation vector as shown to transform the finite element model of the cylinder block.

81


 Click on OK to dismiss the panel  In this case no translation is performed since the model has already been translated into SFE format using the appropriate translator.  Click on OK  Note how the previously defined reduced model of the cylinder block has disappeared.  By default the finite element model of the cylinder block is not displayed. The model may be displayed using the Model Appearance Panel from the View Menu at the top of the Main Panel. 2.3.3.2 Reassembling the Crankshaft Model The program forces you to reassemble the crankshaft model because the cylinder block model has changed.  Select ‘Edit Cranktrain’ from the buttons on the left side of the Main Panel to display the Cranktrain Tool Panel

82

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis  We do not need to redefine any data. However the program will force us to re-edit the main and thrust bearing data.  Click on Assemble Model  An error message will be displayed in the pop-up as shown.

 Dismiss the error message  Highlight ‘Main Bearing’, click on ‘Select All’ followed by ‘Edit Selected’ to display the Main Bearing Panel.  Simply select OK since we do not want to re-edit the data.  Click on Assemble Model again  An error message will be displayed in the pop-up as shown.

 Dismiss the error message  Highlight ‘Thrust Bearing’ and click on ‘Edit Selected’ to display the Thrust Bearing Panel  Simply select OK since we have no data to edit.  This is an error in the current release of the program since it should not force you to edit the thrust bearing since the stiffness is later calculated from the cylinder block model.  Click on Define Material and OK the panel.  Follow Step 4 and Step 5 of the Crankshaft Dynamic Analysis Tutorial  On completion the message ‘Stiffness Calculated, Block Not Assembled’ should be displayed in the message area of the Main Panel.

2.3.3.3 Edit Cylinder Block - Model Definition

83

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis It is necessary to define a reduced model of the cylinder block. The reduced model is a number of nodes and degrees of freedom that are a subset of the complete model. We need to define a number of sets.  A constrained node set for each main bearing. For this set the bearing is defined by a single node with 6 degrees of freedom whose movement is the average of the nodes on the bearing surface.  Four node sets for each cylinder to define the upper and lower reversal points of the small end of the connecting rod on the thrust and anti-thrust sides of the cylinder bore. Each node set will contain a single structural node.  A constrained node set for the head gas face of each cylinder. For this set the gas face is defined by a single node with 3 translational degrees of freedom whose movement is the average of the nodes on the gas face.  A constrained node set that defines the attachment point of each engine mount. For this set the mount is defined by a single node with 3 translational degrees of freedom whose movement is the average of the nodes on the bracket.  Select ‘Edit Block’ from the buttons on the left side of the Main Panel to display the Cylinder Block Tool Panel The following Query Panels will be displayed in succession.

 These are displayed because we have already defined a cylinder block model and in doing so have defined a number of sets   Click on Yes in each case  In most cases you want to answer No to the question. You only want to answer Yes if either the block model type has changed (as in this case) or the main bearing or cylinder geometry has changed. Messages will appear in the bottom of the Main Panel and the Cylinder Block Tool Panel will appear as shown

84


 In this action the program has transformed the finite element model of the cylinder block, calculated its mass properties and attempted to automatically define the sets of the reduced model using the data supplied previously  The reduced model is shown in orange.  The sets defining each main bearing are incomplete and nothing has been defined for the cylinders. This is because the nodes are outside the geometric tolerance.  It is necessary to define these sets using the Define Model Panel.  Click on ‘Define Model’ to display the Define Model Panel as shown

Consider firstly the constrained node sets defining each main bearing  Change Type to Constrained Node and Name to Main Bearing

85

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis  The Cyl and Val fields will now be ghosted and the ID field will be editable.  Click on ‘Select’ adjacent to the ID field and select 1 from the list so that the panel appears as shown

 ID 1 denotes main bearing 1  Alternatively you can simply type 1 in the field  Select the Definition tab

 Select each tab and understand the default values  Each set is ‘clipped’ based on geometric shape  Each shape is defined by a Centre, Axis, Extent and Diameter  Select the Tolerance tab

86


 The default linear tolerance is set to half the minimum distance between any two adjacent finite element nodes.  Each set has its own tolerance values  Change the Linear Tolerance to 0.5 mm and press ‘Clip Set’.

 The ‘clipped’ mesh is shown in pink  The program only clips the external surfaces of the model.  This set is OK so click on ‘Add Set’ to add the set to the reduced model as shown

87


 Repeat this procedure for each of the remaining main bearing by changing the ID on the Name tab and the linear tolerance on the Tolerance tab. On completion the panel should appear as shown.

Secondly consider the node sets defining the thrust and anti-thrust nodes for each cylinder.

88

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis  Select the Name tab  Change Type to Node and Name to cylinder as shown

 Click on ‘Select’ adjacent to the Cyl field and select all the cylinders  Click on ‘Select’ adjacent to the ID field and select lowerAntiThrust so that the panel appears as shown

 Select the Definition tab and again understand the default data  The data for cylinder, valve seat and valve spring are with respect to a cylinder local coordinate system with the origin at the top of the cylinder. This enables sets to be defined for multiple cylinders in a single action.  The shape is set to Sphere in this case to find the nearest relative to the specified centre.  Select the Tolerance tab and again change the linear tolerance to 0.5 mm and click on ‘Clip Set’

89


 The node numbers are displayed  The coordinates of each node are displayed in the message area with respect to the cylinder and the global system.  Select ‘Add Set’ to add the nodes to the reduced model.  Repeat this procedure for each of the remaining cylinder nodes lowerThrust, upperAntiThrust and upperThrust by changing the ID on the Name tab and the linear tolerance on the Tolerance tab. On completion the panel should appear as shown.

90


Finally add sets defining the attachment points of the engine mounts. Again these sets will be from the SFE file.  In this example there are no engine mount brackets. Arbitrary sets have been defined to attach the engine mounts  Select the engine mount sets from the list and click on Read Set.

91


 Select ‘Add Set to obtain the following

92


 Use the mouse and cursor keys to rotate and zoom the model if you haven’t already done so.  OK the Define Model Panel 2.3.3.4 Edit Cylinder Block - Engine Mount Definition  Highlight Elastomeric Mounts in the list on the Cylinder Block Tool Panel  This will then display each of the engine mount nodes in green  Select the front engine mount using the mouse  The node will turn red  Select ‘Edit Selected’ to display the Elastomeric Engine Mount Panel  Complete the properties for the three degrees of freedom as shown

 The stiffness’ are with respect to the local coordinate system  The default orientation is with X vertical.  These stiffness values are typical values for an automotive application. However the rigid body modes using these values will be higher than you would normally expect given the mass of the block assembly in this case is lower given it is not a fully dressed model.  Click on OK  Repeat for the right engine mount

93


 Repeat for the left engine mount

 Repeat for the rear engine mount

94


2.3.3.5 Edit Cylinder Block – Model Reduction The model is reduced using Component Mode Synthesis. This technique ensures that the modal behaviour is the same as the complete model for a specified number of modes. The model is run free-free to give 6 rigid body modes which are latter restrained during model assembly by the engine mounts.  Select ‘Matrix Reduction’ to display the Matrix Reduction Panel

 For this tutorial we will use the FEARCE Vectorized Sparse Solver (VSS)  Alternatively MSC Nastran can be used  If you have limited memory then this option can be used to limit the memory during forward elimination and backsubstitution. This will result in a longer solution time.  Set the Number of Modes to 50  For this example this will obtain modes up to 3.5 kHz.  The number of modes will determine the size of reduced matrix and therefore the speed of the subsequent dynamic solution.  In this case the model has 102 reduced freedoms, so the size of the reduced matrices will be 152 square. 

Select the Default output name  The output name specifies the name of the FEARCE .FRC command file that will be written.  Selecting the Default button will give the output name the same name as the .SFE file

 Switch on the Limited Memory toggle and specify a memory requirement of 1000 Mb or less if your machine has less available memory

95


 Select Solve. The following Query Panel will be displayed.

 Select OK. Another Query Panel will be displayed.

 Select ‘Direct’. This will execute the command file and start the solution. A Progress Panel will be displayed.  This model has approximately 190000 degrees of freedom which are reduced to 102 plus 50 modes for the reduced model to be used by ENGDYN. On completion the message ‘Run completed successfully’ is written to the Main Panel and the following Information Panel is displayed.

96

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis 2.3.3.6 Edit Cylinder Block – Model Assembly  Select ‘Assemble Model’

 This will assemble the cylinder block model  An eigenvalue solution of the reduced model is performed and the axial stiffness of the thrust bearing housing is calculated  Click on OK to dismiss the Cylinder Block Tool Panel  Save the model using the File menu from the top of the Main Panel

2.3.4

Solution

 Repeat the dynamic analysis as described in the Crankshaft Dynamic Analysis Tutorial (Section 2.2.4.3) but for a single speed 2000 rev/min and with the block set to Dynamic.

97

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis 2.3.5 Vibration Analysis The dynamic analysis will predict the forces and displacements (if the block is defined as dynamic) at each of the nodes of the reduced model. This section describes how we can backsubstitute using a forced response analysis using the predicted forces at the reduced nodes to predict the displacements at any point on the cylinder block model. We will predict the response on one side of the block and at each engine mount at a single speed up to 1 kHz. 2.3.5.1 Selecting the Required Loadcase  Click on ‘Select Loadcases’ from the buttons on the left side of the Main Panel.  Select the loadcase at 2000 rev/min full load as shown

 Click on OK to dismiss the panel

2.3.5.2 Performing the Forced Response Analysis  Click on ‘Block Analysis’ from the buttons on the left side of the Main Panel to display the Cylinder Block FEA Panel

 Click on ‘Select Output’ to display the Cylinder Block FEA Output Panel  Highlight ‘Nodal Vibrations’ from the list as shown

98


 Frequency Response Loading will automatically be highlighted.  Click on OK to dismiss the panel.  Click on ‘Select Model’ to display the Cylinder Block Model Panel

 The default will be the FE model used to derive the reduced model.  This defines the model we wish to excite.  The selected file must have modal data and must have reduced nodes at the same location as the reduced model.  Click on OK to dismiss the panel.  Click on ‘Define Output’ to display the Frequency Response and Acoustic Analysis Panel.  Set the Maximum Frequency to 1 kHz as shown and click on Select to list the node sets in the model.

99

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis  This will calculate the response at all the harmonic frequencies up to 1kHz for each selected engine speed. In this case, given we have selected a case at 2000 rev/min and this is a 4 stroke engine we will have results at 16.67, 33.33, 50.0 …Hz  Click on ‘Select’ and select the sets LEFTSIDE, engineMount:ID_left, engineMount:ID_right, engineMount:ID_front and engineMount:ID_rear from the node set list as shown

 The set LEFTSIDE was defined using FEVIEWER  Select the Output tab  Change Export to “Sets Only” and change the default output name to that shown

 The output name is the base name of all the output files.

100

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis  Setting Export to Sets will create an SFE file that contains only the geometry of LEFTSIDE and each of the engine mounts and the reduced nodes. The results will also be written to this file..  Click on OK to dismiss the panel.  Click on ‘Generate Output’ to perform the frequency response analysis A message report will be displayed giving the output from the FEARCE program.

On completion the message ‘Run completed successfully’ is written to the Message Panel and the following Information Panel is displayed

 OK this panel and the Message Panel On completion the following files will be written  IL4_BLOCK_VIBRATION.FRC command file  IL4_BLOCK_VIBRATION.FRL containing the Frequency Response Loading  LC001.ALPHA containing the Modal Contribution factors for load case 1  LC001.TOTALS containing the total loads applied to the model for load case 1  IL4_BLOCK_VIBRATION_SUBSET.SFE containing the results for LEFTSIDE and each of the engine mounts and all the reduced nodes. 2.3.5.3 Plotting Results Go in the working directory and double-click on IL4_BLOCK_VIBRATION_SUBSET.SFE in order to open the .SFE file in R-Desk. The model should appear as shown in the screen shot below.

101


Go to the FE Graph tab as shown on the screen shot below:

102


103

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis In the FE Graph section, click on New -> NVH -> Vibration graph, and the following panel will apppear on the right.

Select Displacement in the left panel, as shown below:

104


Then select the mount sets in the Edit panel on the right (use CTRL + left mouse), and then click “Finish” at the bottom at the edit panel.

105


Finally, click “Plot” in the left hand panel, and you will obtain the R-Plot Graph shown below.

106


 You can explore the process with other quantities, for example velocity.  You can plot several plot at the same time, by highlighting the wanted plots simultaneously (CTRL + left mouse), and the click on “Plot”.

107


2.3.5.4 Viewing Results  Go to the “Data” tab at the bottom of the left panel as shown in the screen shot below.

108


109


 In the “Deformation” tab, select “Displacement” for a frequency of 133.3 Hz, as shown below. This corresponds to the 4th order.

Select “Displacement” in the “Contour” tab. The screen will then look as shown below.

110


Go to the Animation tab that can be selected at the bottom of the middle panel, and animate the results by clicking on the green run bottom.

Experiment with other quantities, like for example velocity and acceleration. To rotate: SHIFT + move mouse with middle bottom pressed To zoom in/out: CTRL + move mouse with middle bottom pressed

111


2.3.6

Noise Prediction

This section describes how we perform an Acoustic Analysis using the Rayleigh Integral Method. The Rayleigh calculation uses a simple method treating each node as a piston in a baffle assuming that the vibration is the same over the surface of the piston. This is normally acceptable for models where the mesh density is sufficient to give a good representation of the mode shapes, and the wavelength is large compared to the element size. We will predict the radiated sound power and sound intensity from each side of the block at a single speed up to 3 kHz. 2.3.6.1 Performing the Acoustic Analysis  Return to the Cylinder Block FEA Panel  Click on ‘Select Output’ to display the Cylinder Block FEA Output Panel  Highlight ‘Radiated Noise’ from the list as shown

 Frequency Response Loading and Nodal Vibrations will be automatically highlighted.  Click on OK to dismiss the panel.  Click on ‘Select Model’ to display the Cylinder Block Model Panel

 Click on OK to dismiss the panel.  Click on ‘Define Output’ to display the Frequency Response and Acoustic Analysis Panel.

112

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis  Set the Maximum Frequency to 3 kHz as shown

 Select the Acoustic tab and Options tab and switch off Delete Vibrations as shown.

 If we are performing this calculation for a large number of speeds and frequency it is usually prudent to leave Delete Vibrations on in order to minimise the size of the resultant files.  Click on the Boundary tab and ‘Radiating Sets’ and select the sets LEFTSIDE and RIGHTSIDE from the face set list as shown

113


 These sets were generated using R-DESK VIEWER.  Select the Output and change the output name to that shown

 Click on OK to dismiss the panel.  Click on ‘Generate Output’ to perform the frequency response and acoustic analysis A message panel will be displayed giving the output from the FEARCE program.

114


On completion the message ‘Run completed successfully’ is written to the Message Panel and the following Information Panel is displayed

 OK this panel and the Message Panel On completion the following files will be written  IL4_BLOCK_NOISE.FRC command file  IL4_BLOCK_NOISE.FRL containing the Frequency Response Loading  LC001.ALPHA containing the Modal Contribution factors for load case 1  LC001.TOTALS containing the total loads applied to the model for load case 1  IL4_BLOCK_NOISE.L001.RES is an ASCII file containing the sound power for load case 1  IL4_BLOCK_NOISE_SUBSET.SFE containing the results for LEFTSIDE and RIGHTSIDE. The size of this file is 63.5 Mbytes.

2.3.6.2 Viewing and plotting Results Open in R-Desk Viewer IL_BLOCK_NOISE.SFE. The process for viewing the results is the same as that described in the Vibration Analysis section (2.3.5). 1. In the Deformation Tab, select Displacement 2. Select “All” 3. In the Frequency tab, select Case 1, and then 1016.67 Hz 4. In the Contour Tab (Data panel on the left), select Sound Intensity.

115

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis 5. Select dBA below Your screen should look similar to the screen shot below.

Note that you can modify the background colour from the tools many in the top left corner. Click

Then Select Options -> Preferences -> Background Colours. You can animate the results using the Animation Tab (at the bottom of the middle panel) In order to plot the results, use the FE Graph tool similarly to what was done in the Vibration analysis section. 1. In the FE Graph panel, when you click “New” and the select “NVH”, chose the “Acoustic Graph” option. 2. In the “Edit” panel on the right, chose “Sound Power – Spectra – XY” 3. Select the desired sets (for example LEFTSIDE and RIGHTSIDE) 4. Click finish 5. Finally click “Plot” in the FE Graph panel. You will obtain a plot similar to the one below.

116


117


2.3.7

Noise Prediction Using Boundary Element Method

This section describes how we perform an Acoustic Analysis using the indirect Boundary Element Method. This method is a more rigorous method and calculates the noise generated by a closed volume. The need for the solution of a set of simultaneous equations, of order N (where N is the number of boundary elements), at each frequency, means that the BEM is substantially slower than the Rayleigh method. 2.3.7.1 Performing the Acoustic Analysis Using BEM  Return to the Cylinder Block FEA Panel  Click on ‘Define Output’, select Acoustic tab and choose indirect BEM in the Method field

 Choose the Boundary tab, Click on Browse button to load SFE file named IL4_BLOCK_BoundaryElementModel.SFE

 Choose Acoustic Mesh tab and select Defined in the Mesh field

118


 Click on the Define button that has become active, with the Acoustic Mesh panel

 Enter a suitable filename for the new mesh, such as Acoustic_Mesh, and select Sphere in the Shape field

119

KNOWLEDGE CENTRE Tutorials Tutorial 3: NVH Analysis We shall define the centre of the sphere at the centre of the gravity of the engine model. We can calculate the centre of gravity of the model using the viewer tool  Double click on the IL4_BLOCK.SFE model.  Right click the Plot item under the model name in the Plots panel  Select Calculate from options list, then select Centre of Gravity

The result will be displayed in the ‘Information’ panel at the bottom of the canvas. The values are returned in SI units and are (176.19, 109.85, 2.81) mm. We have to account for the ENGDYN system. We have translated the model in X direction by 172.0 mm when importing. This gives a Centre_of_Gravity in the ENGDYN system as approximately (4.0, 110.0, 3.0) mm.  Apply the Centre_of_Gravity to the panel as shown

120


 Click on the Radius tab, define the sphere to be 1 meter in radius.  Set the Number of Elements to be 40, which gives a suitable refined mesh

 Click Generate when the panel is complete A new SFE file named Acoustic_Mesh.SFE is created in the current working directory. Double click on the SFE file to look at the model in R-Desk Viewer.

121


The mesh density can be easily changed by changing the number of elements in the Acoustic Mesh tab.  Click OK to confirm the use of acoustic model that has been defined, then the File Name field has been updated as shown below.

 Click on the Fluid tab, define the Fluid Density and Speed of Sound as below

122


 Click on the Options tab, set the Integration Increment to be zero. This field allows the user to increase the order of integration to improve the calculation accuracy, but at the expense of run time. We leave it as zero, meaning the solver will automatically use the optimum integration order.

 Return to the Boundary tab, and the acoustic mesh is fully defined.  Select the Frequency Response tab and set the maximum frequency to 2000 Hz  Click on the Output tab and set the panel as below

123


 Click OK  Click Generate Output to run the noise calculation in the Cylinder Block FEA panel Note that the solution will take approximately 8 hours on a laptop – a 2.5 GHz processor and 4.0 GB of RAM. Rather than running through the FEARCE NVH solver automatically, we can run the calculation through a batch file. On completion the following files will be written  IL4_BLOCK_NOISE_BEM.FRC – FEARCE command file  IL4_BLOCK_NOISE_BEM.FRL – containing the frequency response loading  IL4_BLOCK_NOISE_BEM.L001.RES – containing the sound power for loadcase 1  IL4_BLOCK_NOISE_BEM.OUT – an output file from FEARCE  IL4_BLOCK_NOISE_BEM_EXTERNAL.SFE – containing the vibration results which are interpolated onto the BE model  IL4_BLOCK_BoundaryElementModel.SFE – our BE model containing sound power results  Acoustic_Mesh.SFE – containing sound pressure results  LC001.ALPHA containing the modal contribution factors for loadcase 1  LC001.TOTALS containing the total loads applied to the model for loadcase 1 2.3.7.2 Plotting and viewing results Open the IL4_BLOCK_BoundaryElementModel.SFE by double-clicking this file. Also open Acoustic_Mesh.SFE. In order to open it is the same session, right click in the top left Plots panel and then select “Open Model”. The results can be viewed and analyzed following the same procedure as that described earlier for the Rayleigh method. The sound intensity results are available in IL4_BLOCK_BoundaryElementModel.SFE and the sound pressure results are available in Acoustic_Mesh.SFE.

124

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis

2.4 Tutorial 4: Block Stress Analysis

2.4.1

Overview

Objective:  Build an ENGDYN model suitable for performing a stress analysis of the cylinder block  The analysis uses finite element (FE) models of both the powertrain assembly and the crankshaft  This tutorial requires the ENGDYN model from Tutorial 2 (Crankshaft Dynamic Crankshaft) Items Covered:  Compliant cylinder block model  Restraining the system  Static condensation reduction  Quasi static stress analysis Estimated duration:  0.5 day Required Files: ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\COMPLIANT_BLOCK.SFE Finite element model requirements:  For a stress analysis of the cylinder block it is recommended to have a powertrain assembly that includes all of the major components contributing to the stiffness and hence load paths of the system Method:  In many cases a quasi-static stress analysis of the powertrain is a suitable method for analysing the in service behaviour of the cylinder block  The quasi-static method has a number of advantages over a fully dynamic analysis  The method is computationally less intense allowing iterations to be carried out swiftly  The loads are fully resolved statically and can be applied to both linear and non-linear FE solutions

125

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis  Thermal loads can be combined with the mechanical loads generated by ENGDYN providing a full thermal mechanical assessment  For cases where a known dynamic stress problem exists, ENGDYN can also be used to provide a solution and this is covered in a separate tutorial

2.4.2

Getting Started

The model generated in tutorial 2 (dynamic crankshaft analysis) is used in this tutorial, so we will make a copy of the model created in tutorial 2.  Start engdyn  On Unix or Linux platforms simply type engdyn  On windows click on the shortcut, otherwise go to Start>Programs>Ricardo Software>2014.1>Mechanical Suite>ENGDYN>ENGDYN  Open the model from tutorial 2  Use the File > Open command from the top menu bar  Select Copy Design from the File menu on the top bar  Select a new working directory in the browser and save the model under a new name

Note that saving the file in this way keeps the references to the component files. In this case, it refers to the crank FE model (and reduced components) and the lubrication and pressure data.

126

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis 2.4.3

Updating the Model

2.4.3.1 Define Models In this step, we need to redefine the level of our component cylinder block model  Select ‘Define Models’ from the buttons on the left side of the Main Panel to display the Model Definitions Panel

 Select the ‘Cranktrain’ tab and define the parameters as follows

 Select the ‘Cylinder Block’ tab and set out model level to compliant as shown

 Click on the Select button in the File name field and refer to the cylinder block FE model COMPLANT_BLOCK.SFE

127

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis  Click on Transformation tab, and translate the model by 172 mm in the X direction

 Click OK 2.4.3.2 Reassembling the Crankshaft Model  Click on the Edit Cranktrain button Note that we do not need to redefine any data as we haven’t changed the crankshaft. However the program will force us to re-edit the main and thrust bearing data. This is because the thrust bearing stiffness will now be calculated from the FE model.  Select the Trust Bearing item from the Cranktrain tool panel  Click Edit Selected  Click on Stiffness tab, and set 1 in the Housing field  Click OK

 Select the Main Bearing item from the menu  Click on Select All, and then Edit Selected  Click Ok in the bearing panel as we do not wish to alter the values

128

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis  Run through the crank assembly steps, including ‘Define Material’, ‘Calculate Masses’, and ‘Set Balance’ as we did in tutorial 2 (dynamic crankshaft analysis)

 Click on the Calculate Stiffness button

 Select No as we haven’t changed the crankshaft model in any way 2.4.3.3 Defining the Reduced Model We now need to define a ‘reduced model’ of the FE cylinder block that we have imported. The reduced model is a number of nodes and degrees of freedom that are a subset of the complete FE model. In order to still be able to interact with our FE model once we have performed the reduction, we need to define the degrees of freedom (nodes) whose details we wish to retain. At a minimum this needs the definition at each main bearing, the definition for each cylinder bore and the definition for each cylinder head gas face. We will need to restrain the FE model in space before performing the model reduction. Suitable restraint points need to be selected - using engine mount locations would provide sensible load paths. For consistency of solution these restraints should be the same as used in the final FE stress analysis.  Click on the Edit Block button, and the following query appears

129


 Click Yes as we have changed the block model type A secondary query panel will appear as below relating to the cylinder definitions

 Click YES as we have changed our block model type, and then the Cylinder Block Tool panel appears

 Click on the Define Model button in the Cylinder Block Tool panel, and the Define Model panel appears as below

The three main tools are: 1) Clip: This allows us to adjust the geometric definitions of the required regions and tune the search of the FE model for the required nodes 2) Read: This allows us to read pre-defined ‘sets’ of data directly from the FE model to use in our reduced node definition

130

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis 3) Delete: This allows us to delete any definitions which are not wanted  Click on the Delete tab and select the Constrained Sets tab

 Highlight all five items in the window and click Delete

 Click on the Clip tab (we are going to tune the definition of the main bearings to try and find the complete faces.)  Under the name tab, define the parameters as shown below

131


 Click on Select button next to the ID field and choose main bearing 1 from the list of options.

 Click OK  Click on Definition tab This section reveals the geometric definition that ENGDYN considers to be the inner surface of main bearing 1

132

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis  Click on the Tolerance tab ENGDYN has calculated this tolerance by reading the FE model and determining the smallest length of an element side. The tolerance is half of that smallest length. But if some elements are very small, then this tolerance will be too small to correctly locate the surface that we require.  Change the tolerance to 0.5 mm and click on Clip Set button. ENGDYN finds the complete bearing surface shown in the viewport.

 Click Add Set and the viewport will update to show the complete bearing defined in orange

 Repeat this procedure for each of the five main bearings  In each case, change the bearing ID in the Name tab, then change the geometric tolerance to 0.5mm, Click Clip set and Click Add Set On completion the viewport should like as shown below. The next job is to correctly define the nodes required to represent the four cylinders.

133


 Return to the Name tab in the Clip section of our Define Model panel  Select Node in the Type field and select Cylinder in the Name field, as shown

 Click on the Select button in the ID field and select upperThrust as shown

134

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis  Click on the Select button in the Cyl field and click Select All button in the Cylinders field

 Click OK and the Define Model panel is shown as below

 Click on the Tolerance Tab and set 0.5 in the Linear field

 Click on Clip Set and click Add Set

135


 Repeat this step for each of the remaining cylinder nodes. You should be able to clip for all four cylinders at once each time  upperAntiThrust  lowerThrust  lowerAntiThrust  Set the tolerance to 0.5 mm  Remember to clip the set to locate the nodes in the FE model before adding them

We will define the next sets using the Read option. This is used to select regions already included as named sets on the FE model. Because of the complexity of the shape of the flame face it is often easier to define them in this way than to try and create a shape to automatically clip from. So we can prepare our models using an FE package (either a third party package or Ricardo’s R-Desk VIEWER tool) by ‘painting’ required regions and naming these as face sets.

136

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis  Select the Read tab on the Define Model panel and select type as Constrained Sets You will see a list of sets that exist on the model which we can choose from.

To add these all we need to do is highlight the ones we want, click Read Set, then Click Add Set.  Highlight all of the sets and add them to the model  Click on OK in the Define Model panel. The viewport should update to look similar to that shown below

Next, we can perform the matrix reduction.  Go back to the Cylinder Block Tool panel

137


 Select the Matrix Reduction option  Select FEARCE option from the Solver list, as shown below

The Solver field allows us to choose which solver we would like to use for the reduction. Currently ENGDYN can support MSC/NASTRAN, ABAQUS or its own internal solver FEARCE. A translator can be supplied to export an appropriate analysis deck to NASTRAN or ABAQUS. If FEARCE is selected, no external licenses are required. The Output Name field determines what these files will be called and where they will be located on your system.  Click the Default button next to Output Name field, which means the files will be placed in the working directory and named after the powertrain FE model The Element Check allows a check of the FE model to be performed.  Click Element Check and limited memory.  Set a memory limit approximate to your machine as shown below

138


 Press Solve and the following dialogue panel appears

This dialogue is telling us that the command file to set up the reduction has been written. Choosing to continue will execute this, causing the following actions. 1) The reduced nodes will be added to the FE model; 2) The FEARCE solver will then perform the reduction. If NASTRAN was chosen for the solver, the NASTRAN input deck will be written out at this stage.  Choose Ok  You will then be asked whether you want to perform the reduction immediately or later (in batch)  Select the Direct option  The command file will execute and the reduction will begin This powertrain model has about 190000, and this is to be reduced to about 100.  Click Assemble Model when the reduction is complete  Close the Cylinder Block Tool panel 2.4.3.4 Performing ENGDYN solution We are now ready to perform the ENGDYN solution. The lubrication and cylinder pressure was defined when the original model was set up for the dynamic crank analysis. Hence we can go straight to the Evaluate Solution option.  Select the Indeterminate option from the solution type list

139

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis As we are not considering the dynamics of the system, the solution convergence is usually straight forward.  Make sure that the cylinder block model is set to Compliant as shown  Set up other parameters as follows

 Move to the Cases tab  Add a single loadcase of 5000 rev/min in the speed table as shown

 Click on the Bearing Model tab

140

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis  Select the thermal balance option  This will calculate the rise in temperature across the bearings due to work  This will not slow an indeterminate solution too much so is worthwhile  Finally select the required oil properties file The solution should only take a few minutes to complete with the message shown below

 Close the message window 2.4.3.5 Block Stress Analysis First step is to select what loading condition we require.  Click on the Select Loadcase button  Make sure that the Type of results is Statically Indeterminate  Dynamic results will not be fully resolved statically and are inappropriate  We only have one set of results, so select this in the list

141


 Click Ok when done  Select Block Analysis form the post processing menu The cylinder block FEA panel will appear. This provides us with the workflow required to set up the FE analysis.

 Click on Select Output  Choose the Quasi-Static loading option from the list, shown as below

 Click OK

142

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis  Click on the Select Model option and the Cylinder Block Model appears

 Select the block required for the stress analysis shown as follows

The model transformation used in the ENGDYN solution will be loaded by default. We can change this as before in using the panel under the Transformation tab.

 When done, click OK  Click on OK on the Cylinder Block Model panel  Click on the define the output

143


We shall select which FE solver we want to use.  Choose Ricardo-SFE. This will automatically apply the loads to the FE model.  Click ‘Select’ next to the Crank Angle field

 Select the ‘Use Interval’ option and set the interval to 10 degrees  This will give us 72 load cases from ENGDYN

144


 Click OK when complete The final main field that we can edit is the ‘Start Loadcase’ section. This defines what the first loadcase number that the ENGDYN loads are written to is. For a complete analysis we would usually add assembly and thermal loads to the system in addition to the ENGDYN loads. These would often be added as loadcases 1 and 2 and be added to the model as part of the preparation before ENGDYN. Hence we could start the ENGDYN loads from loadcase 3.  In this example we will use the default as we do not have other loads The Output Name field determines where the files are written that control the FE solution.

 Keep this as default and click OK  click on Generate Output

145

KNOWLEDGE CENTRE Tutorials Tutorial 4: Block Stress Analysis ENGDYN will create a FEARCE run script that will apply the loads onto the FE model and run the solution or set up an appropriate third party solver deck. With the SFE option chosen for loading, this will be executed automatically. This will take several minutes to complete. The STRESS_BLOCK.SFE model is now loaded and ready for solution.

These loads can be viewed in R-Desk viewer Use the Data > Contours > Force option

To implement the FEARCE solver a small script file needs to be created. The following file has been included named Run_stress.FRC

146


This can be executed from a command line in the appropriate directory. Simply type FEARCE and the file name shown as follows.

Finally, open the STRESS_BLOCK.SFE file in order to view the results. The process is similar to the one described in the NVH tutorial for viewing the vibration results (section 2.3.5.4). Please note that this tutorial model in only provided for the purpose of illustrating the process and the used elements are not appropriate for stress analysis. A plot similar to the one below can be obtained by adjusting the scales.

147


148

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis

2.5 Tutorial 5: Crankshaft Stress Analysis 2.5.1

Overview

Objective:  To use ENGDYN to perform a stress and durability analysis of a crankshaft  The crankshaft utilises a finite element (FE) model.  The cylinder block model is assumed rigid in this example. Items Covered:       

Preparing the Engdyn model Geometry Definition Assigning Material Properties Stress Concentration and Fatigue Notch Factors Classical Stress Method Finite Element Method Plotting and Viewing Results

Estimated duration:  0.5 day (model preparation)  1 day (overall including performing solutions) Engine:  Inline 4 gasoline engine Required Files: ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4_crank.SFE ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\IL4_STRESS_CRANK.SFE ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\il4*.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\il4\dynamic_crank.EDSF Finite element model requirements:  Crankshaft FE model should be split into its individual web sections as described in Finite Element Model section of the manual

149

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis 2.5.2 Preparing the Engdyn Model 2.5.2.1 Performing the Solution A pre prepared .EDSF file is provided for this tutorial. This is a copy of the Engdyn model as prepared in the ‘Crankshaft Dynamic Analysis’ tutorial. The reader is invited to review this tutorial to familiarise themselves with the methods of preparation. We shall begin this tutorial by performing a solution suitable for the crankshaft stress analysis  Open Engdyn and click on the File button on the main toolbar  Click on Open from the drop down menu  Select the ‘dynamic_crank’ .EDSF from the tutorial folder  The display should look like the figure below

 The lubrication should have been defined, but check by clicking on the Lubrication button and ensuring that the file SAE5W30.MAT is being referenced

150

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  If not click on the Browse button and reference the file so that the lubrication panel looks like the figure below

 Click on OK in the Lubrication Definitions panel  Click on the Loading button  Ensure that the In-Cylinder panel is set as shown

 Click on OK  Click on the Evaluate Solution button  Configure the Solution panel as shown

151


 We shall carry out a fully dynamic solution, but will also save the statically indeterminate results – ensure that the ‘Save Static Results’ box is checked  Configure the Cases panel as shown – setting a speed range of 2500-5500 rev/min in 500 rev/min intervals

 We will use the default settings for the cylinder block, crankshaft, thrust bearing and coupling panels  Click on the Bearing Model panel and configure it as shown below

152


 To save time in this example, we shall explicitly define the oil temperatures at each bearing – check that these are all 110 OC using the Define Oil Temps button  Click on Solve Directly.

 When the solution has completed, check that it has converged for every loadcase We are now ready to carry out the crank stress analysis 2.5.2.2 Geometry Definition To obtain stress results at specific points on the crankshaft, we must define the geometry at these locations in more detail. This is done through the Crankshaft Stress Analysis panel, accessed by clicking on the Crank Analysis button.  Click on the Crank Analysis button to display the following panel

153


 We shall calculate stresses at the main and pin journal fillets at the rear of the crank  To do this we must also define the geometry of the surrounding webs  We shall regard all the webs as being uniform and so define all components at once  Highlight ‘Crankshaft webs’ in the left of the panel  Click on Select All  The webs will show in red on the display as shown

 Click on Edit Selected to display the Crankshaft Webs panel

154


 Fill in the ‘white’ boxes as shown  Click on Apply  The equivalent width box will update  Click on OK  Highlight ‘Pin journal fillets’ in the left hand of the panel  Click on Select All  Click on Edit Selected to display the Pin journal Fillets panel  Set the Radius to 1.5  Click on Apply  The panel should update as shown

155


 Click on OK  Highlight ‘Main journal fillets’ in the left hand of the panel  Click on Select All  Click on Edit Selected to display the Main journal Fillets panel  Set the Radius to 1.5  Click on Apply  The panel should update as shown

156


 Click on OK The geometric data required for this tutorial has now been applied. The next step is to define the material properties of the crankshaft. 2.5.2.3 Defining the Material Properties At this stage we are not referencing an FE model, so we will define the material properties from the Engdyn database. The procedure that will be shown can also be used to enter the value explicitly.  Click on the Define Material button as shown to display the Crankshaft Material Properties panel

157


 In the Base Properties tab, ensure that the panel is configured as shown below

158

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  Only be concerned with the upper half of the panel (shown above) for the moment. The data in the lower half will be updated when all of the properties have been added  Click on the Base Strengths tab

 We will use properties for Steel 38MnS6 from the Engdyn database  Click on Select (circled) to access the Engdyn materials database  Highlight the STEEL38MnS6 properties as shown then click OK  The base properties will be updated in the upper part of the Base Strengths panel

159


 Now click on the Elevated Strengths tab

160


 Select Calculated in the Strengths field  Click on Define next to Size Factor to display the following panel

 Ensure Default is selected as the Definition of the size factor

161

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  Click Apply, then click OK  Back in the Crankshaft Material Properties panel, select Cold Rolled in the pin and main journal fillet fields  The panel should appear as shown below

 Click on OK 2.5.2.4 Defining Stress Concentration and Fatigue Notch Factors Engdyn will take into account any stress increase or relief due to geometric effects by applying stress concentration and fatigue notch factors to the journal fillets. These factors are applied by altering the material strengths at the locations. To apply these factors we must access the fatigue notch factors panel.  Click on the Notch Factors button as shown

162


 We will apply the Modified Lowell equation to calculate the required factors  Ensure that Modified Lowell is selected in the Characteristic field  Click on Apply, so that the panel appears as shown

 Click on OK 2.5.2.5 Classical Evaluation of Stress and Durability Now that we have defined the geometry and material properties of the regions of interest on the crankshaft, we are in a position to carry out a classical evaluation of the stresses and durability. First we must select which loadcases we are interested in for the solution.

163

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  Click on the Select Loadcases button on the main panel to bring up the following sub panel

 We shall perform a speed sweep across all defined engine speeds which have dynamic results  Display the dynamic results by setting the Solution Type field to dynamic  Click on the Select All button to highlight all speeds  Click on OK  Click on the Define Output button on the Crankshaft Stress Analysis panel  Set the Classical Stress Analysis sub panel as shown  Note: the output name should reference the current Engdyn model

164


 Click OK  Click on Solution on the Crankshaft Stress Analysis panel  Engdyn will now perform the stress solutions at each part of the crankshaft for which geometry has been defined and for each loadcase  When this is complete we can plot results 2.5.2.6 Plotting Classical Stress Results We can now view results at any part of the crank for which the geometry has been defined. As an example we shall look at results for the rear pin journal fillet on the fourth pin.  Highlight the pin journal fillets on the left hand of the panel

165


 The pin journal fillets that have been defined will turn green on the reduced model  Highlight the rear fillet on pin journal four by clicking on it with the left mouse button

 The fillet will turn red  Now click on the Plot Results button

166


 The following panel will appear

 We will look first at the stress history at the fillet  Highlight History, Stress and Goodman-Standard-General options as shown

 Click on Apply  The graph shown overleaf will appear

167

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  The plot can be printed or closed through the File tab on the top menu of the plot

 Now we will look at the predict durability safety factors at the fillet  Highlight Summary, Safety Factor and Goodman-Standard-General as shown

 Note: options that are not applicable disappear from the list of choices  Click on Apply to display the plot

168


 The top graph shows the maximum and minimum (alternating) stresses acting on the fillet  The lower graph shows the lowest predicted durability safety factors on the filler across the speed range  The red line is the design factor which can be changed in the Plot Results panel

169

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis 2.5.2.7 Performing a Finite Element Solution Now we have carried out a classical analysis, we are ready to move to the next level of complexity and perform an FE analysis on the crankshaft. The first step is to create the boundary conditions so that we can perform the solution. These are applied to specific sets that should already have been added to the crankshaft FE model. Details of these sets can be found in section 7.1.2.2 of the manual. A brief description of the sets required for loading follows.  A layer of low stiffness (approx. 1/10th crankshaft stiffness) and massless elements need to be added to each end of the crankshaft to ameliorate the stresses applied to each end (shown in red below). The length of these should be about ¼ of a typical bearing for the crankshaft

 A layer of massless shells (all other properties the same as the crankshaft) should be overlaid onto the low stiffness elements on the cranknose end  Face sets need to be added to all of the Pin and Main journals. These represent the oil contact areas of the bearings  The sets need to be named with the convention shown in the figures below

170


Once the crankshaft has been properly prepared we are ready to generate the boundary conditions  Click on the Select Loadcases from the main panel  For an FE solution we must select a static loadcase – hence the statically indeterminate solutions were saved  Click on the Select Loadcases from the main panel

171

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  For an FE solution we must select a static loadcase – hence the statically indeterminate solutions were saved.

 Select 5500 rev/min from the statically indeterminate results  Click on OK  Now click on the Crank Analysis button in the main panel  Set the Stress Analysis sub panel to FE Quasi Static method and Unit Loads output as shown

172


 Reference the stress quality model by clicking on the Select Model button on the right of the panel

 Ensure that the model is in the same orientation as the Engdyn model by supplying the correct transformation from the appropriate tab  Warnings may appear in the message box of the Engdyn panel if the defined geometry does not match up to the FE model  Care must be taken to ensure that the FE model aligns correctly with the defined geometry  Click on Apply, then OK  Now click on the Define Output button to bring up the panel shown below  Set the desired Solver and ensure that the Centrifugal Load and Unit Moments boxes are checked  The output name will define what the .FRC file is called

 Click on OK

173

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  Click on Solution  Engdyn will create the .FRC file to load the crankshaft. When complete the following box will appear prompting as to whether the .FRC file should be run

 Click Yes  The .FRC file will run, and where appropriate the input deck for the solver will be written out  If FEARCE(VSS) solver was chosen, then the solution will begin automatically The FE solution should now be performed. If a solver other than the FEARCE(VSS) solver was used, then the results need to be appended onto the stress FE model so that they can be post-processed and viewed. 2.5.2.8 Post Processing the FE Results Once the FE stress analysis has run, ensure that the results have been appended to the stress model .SFE where appropriate. We can now post-process and view the results.  Click on Select Loadcases  Ensure that the statically indeterminate loadcase that has been run is selected >

 Click OK

174

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  Click on the Crank Analysis button and select FE Quasi Static method and Stress and Safety Factors as output (as shown below)

 From the Select Model button, ensure that the correct model is referenced (stress model with results) and that it is in the correct orientation  We shall again just consider the rear pin fillet of pin journal four  Highlight Pin Journal Fillets in the left hand side of the panel (the pin fillets on the reduced model turn green)  Select the rear journal fillet on pin four by clicking on it with the left mouse button (it will turn red)

 Click on Edit Selected

175

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  As we have previously defined the geometry we do not need to change the entry

 Click OK on the Pin Journal Fillets panel  Click on the Define Material button  if the material defined in the FE model is not identical to that in the Engdyn model (for example hydrostatic fatigue factors may not be defined in the FE model), the following panel will appear

 Click on Select to display the panel below

 Highlight the ENGDYN properties (in this case) and click on OK  Click on the Notch Factors button and check that the definitions are correct

176


 Click OK in the Notch Factors panel We now need to define the type of algorithm for the durability safety factor calculation  Click on Define Output

 Click on the Select button next to the Node Sets field  Highlight REAR_PIN_FILLET_04 from the sub panel and click OK

177

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  The Node Set sub panel will display all of the relevant sets defined on the FE model – we can only select sets that are also have their geometry defined in the Crankshaft Stress Analysis panel  Click on the Safety tab  For this example we shall only look at the Alternative Goodman criteria  Highlight as shown and click OK  Check that the rest of the FE Stress Analysis panel is configured as above and click OK  The Export field will determine how much of the FE model is exported for post-processing. The alternatives are ALL, EXTERNAL or SETS. The smallest model will be just the sets The post-processing is now ready to run. Engdyn will first create a new .SFE of the chosen subset of the model (in this case just the selected sets). It will then combine and factor the unit loadcase stress results before finally performing the durability calculation. The results can then be plotted in Engdyn, or colour contours viewed with R-Desk.  Click on Solution  Engdyn will write out an .FPO file to perform the post-processing. When this is done the following panel will appear summarising the forces in each separate loadcase >

 Click on OK  Click on OK in the Query panel (below) to perform the post-processing

178


 When the post-processing is complete the following panel will appear

 Click on OK

2.5.2.9 Viewing and Plotting the FE Results If the post-processing completed successfully a new .SFE file would have been written to the directory containing the original stress quality FE model. This will have the word ‘STRESSED’ appended to the original model name, i.e. _STRESSED.SFE If we click on the Plot Results button in Engdyn, we can generate graphs from this new .SFE (referenced automatically) in exactly the same way as detailed in the classic analysis section. In addition, we can also view 3 dimensional contour plots of the areas that were selected using R-Desk.  Click Crank Analysis on the left side of the software

179


 Select the pin journal fillets item from the menu list and select the rear pin fillet on the crank model in the 3D viewport  We have calculated FE results for this item so can now plot the results  Click Plot results  We get the same panel as we used in the classic analysis

 Have a look at a selection of the plots

180


Next let us view one of the FE models

181

KNOWLEDGE CENTRE Tutorials Tutorial 5: Crankshaft Stress Analysis  Double click on the file crank_stress_L0009.SFE The sub model is just the section that we exported  There is now an extensive amount of data stored on the model  A lot of this is resulting from sub-steps in the post processing and can be discarded  Select the Fatigue tab on the data panel You will see three data sets, all labelled 09 for the load case number 9  QS09, Durability for quasi-static loads only  QST09, Durability for quasi-static loads plus torsion vibration  QSV09, Durability for quasi-static loads plus torsion and bending vibration

 Explore the different results and views available

182

KNOWLEDGE CENTRE Tutorials Tutorial 6: EHL Big End Bearing Analysis

2.6 Tutorial 6: EHL Big End Bearing Analysis 2.6.1

Introduction

Objective:  To build an ENGDYN model suitable for performing a static EHL analysis of a big end bearing.  The crankshaft and cylinder block models are both assumed to be rigid.  This tutorial can be applied to a single cylinder of a multi-cylinder engine Items Covered:     

Building the engine model Connecting rod model definition Static Matrix Reduction Static analysis of a big end bearing Plotting and animating results

Estimated duration:  0.5 day (model preparation)  1 day (overall including performing solutions) Engine:  Single Cylinder HYDRA Research Engine  80.25 x 88.9 mm Required Files: ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\hydra\ROD.SFE ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\hydra\SMALLEND.SFE ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\hydra\BIGEND.SFE ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\hydra\pcyl2000.PRES ..\Ricardo\2014.1\Products\ENGDYN\Tutorials\hydra\RotellaX15W40.MAT Finite element model requirements:  Model should contain bearing shells and big end cap bolts  A material should be defined for each bearing (if different) to define the lining of the bearing shell

183


2.6.2

Getting Started


2.6.3

Building the Model

2.6.3.1 Configure Engine In order to build an ENGDYN model it is first necessary to define the major dimensions and features of the engine. This is done using the ‘Configure Engine’ Panel.  Select ‘Configure Engine’ from the buttons on the left side of the Main Panel.

 Complete the Engine Configuration Panel as shown.

 On completion select ‘Apply’ which will display

184


 Select OK. The crankshaft model will appear in the Main Panel as shown.

2.6.3.2 Define Models An ENGDYN model consists of a number of sub-models which are defined using the Model Definitions Panel.  Select ‘Define Models’ from the buttons on the left side of the Main Panel.

 Complete the model definitions for the Crankshaft, Cylinder Block and In-Cylinder as shown.

185


 The crankshaft can be defined as Rigid (Massless) since for the purposes of this static calculation the loads at the big end bearing will not be influenced by the crankshaft mass. This will mean the minimum amount of data will be required subsequently.  Similarly the Cylinder Block can be defined as rigid.  Define the Connecting Rod model type by positioning the mouse over the centre column and selecting the right mouse button. This will display the pop-up as shown.

 Different models can be defined for each cylinder (when there are multiple cylinders)  Compliant Big End, Compliant Small End and Compliant are applicable to static calculations only  Rigid and Dynamic are applicable to dynamic calculations only.  If nothing is defined or None is selected then the solution for that cylinder is equivalent to v3.0 and earlier  Complaint Big End and Compliant Small End are used where a complete model of the connecting rod is not available. Examples of these models are SMALLEND.SFE and BIGEND.SFE. (You may wish to look at these models using FEVIEWER)  The Complaint Big End and Compliant Small End require a cut-plane somewhere along the shank of the rod.

186


 Select Compliant

187

KNOWLEDGE CENTRE Tutorials Tutorial 6: EHL Big End Bearing Analysis  Select the finite model by positioning the mouse over the centre column and selecting the right mouse button. This will display the pop-up as shown. Releasing the button will pop-up the model translation panel as shown.

 Use the Browse button to select the file ROD.SFE from the working directory  Define the transformation vectors to transform the model from the finite element model coordinate system to the ENGDYN coordinate system

 These vectors are with respect to the ENGDYN coordinate system  The connecting rod of each cylinder has its own local coordinate system such that the origin is at the centre of the big end bearing, Y is up the connecting rod, and X runs along the bearing axis  Select OK on both the Model Translation and the Define Models Panels. This will display the connecting rod model as shown.  The program determines the number of unique connecting rod models and writes a message at the bottom of the Main Panel as shown

188


2.6.3.3 Edit Cranktrain  Select ‘Edit Cranktrain’ from the buttons on the left side of the Main Panel to display the Cranktrain Tool Panel

 Given we defined the crankshaft as Rigid (Massless) (See 2.6.3.2) it will only be necessary to define Main Bearings, Big End Bearings, Small End Bearings, Cylinder, Connecting Rod and Piston.

189


Highlight ‘Main Bearing’ as shown  This will then display each of the main journal nodes in green

 Select ‘Select All’  Each of the main journal nodes will then turn red  Select ‘Edit Selected’ to display the Bearing Panel  Complete the Bearing Panel as shown

 Given the Model Type is Mobility and the bearings are fully grooved it is not necessary to enter any other data than shown  Select Apply to display the panel showing the bearing grooves

190


 Select OK  Highlight ‘Big End Bearing’  This will then display the big end journal node in red since there is only one cylinder. There is therefore no need to select ‘Select All’  Select ‘Edit Selected’ to display the Bearing Panel  Complete the Bearing Panel as shown

 This bearing is feed from the journal via a feed from the adjacent second main bearing  The oil hole angular position can be defined either using the Position column (as in this case) or using the Height column For an EHD Model Type it is necessary to use the Mesh and Material tabs to define additional data for this model.  In this first exercise we will assume the bearing and journal are circular, and therefore it is not necessary to define a profile using the Profile tab.  Select the Mesh tab, and define a mesh 9 x 55

191


 This mesh is the computational mesh for the bearing model  Note how the mesh is refined to resolve the oil hole  Select the Material tab

 This material data is used for the boundary lubrication model.  It is necessary to define the journal material and the material of the bearing lining.  Click on Define adjacent to Bearing Material, to display the Material Properties Panel

192


 In this case the material property data will be read from ROD.SFE  It is necessary to have defined a material in the finite element model that corresponds to the bearing lining, although this material will not be assigned to any finite elements  Use Select to list the materials and select the material BEARING from the list as shown

 This table shows all the materials in the SFE file, relevant to the material being edited, together with any materials stored by ENGDYN. (In this case there will be no ENGDYN materials in since we have not yet added any materials)  Materials used by ENGDYN and those stored in the .SFE are identified by a single unique name.  These data have been defined previously outside ENGDYN, although they may be and can be zero.  Select OK and the material is added to the panel as shown.

193


 These data values are typical for a bearing surface and can be edited using these panel  These data are only used by the boundary lubrication model  Asperity RMS Height h, Density  and Asperity radius  can be calculated from measured data using the Ricardo MATUTIL program which is described in Appendix 12.  Select OK  These data are now stored by ENGDYN and also written to the .SFE file.  Previous data will be overwritten.  Click on Define adjacent to Journal Material, to display the Material Properties Panel  There is no FE model of the crankshaft so this time don’t use the Select button (since it will only list the previously defined material called BEARING.)  Complete the Material Properties Panel as shown

 Select OK and enter the wear and friction coefficients for the bearing interface

194


 These data are only used by the boundary lubrication model  The wear coefficient is not used for the solution, but only as a scalar for postprocessing.  Select OK to complete entering data for the big end bearing.  Highlight ‘Small End Bearing’  This will then display the big end journal node in red since there is only one cylinder. There is therefore no need to select ‘Select All’  It is not strictly necessary to define data for the small end bearing but by doing so helps defining the connecting rod reduced model in 2.6.3.4.  A bearing solution of this bearing will be performed.  Select ‘Edit Selected’ to display the Bearing Panel  Complete the Bearing Panel as shown

195


 Given the Model Type is Mobility and the bearing has a single oil hole feed it is not necessary to enter any other data than shown  Select Apply to display the panel showing the bearing oil hole  Select OK  Highlight ‘Connecting Rod’ as shown

 This will then display the big end journal node in red since there is only one cylinder. There is therefore no need to select ‘Select All’  Note that the Define Model and Matrix Reduction buttons can now be selected. These will be used in 2.6.3.4 and 2.6.3.5 respectively.

196

KNOWLEDGE CENTRE Tutorials Tutorial 6: EHL Big End Bearing Analysis >

 Select ‘Edit Selected’ to display the Conrod Panel  Complete the Conrod Panel as shown

 Note that we have changed the units of Inertia to kg.mm^2  In this case because we have a complete model of the connecting rod this data will be overwritten with that of the FE when we perform 2.6.3.6.  Select OK  Highlight ‘Piston’  This will then display the big end journal node in red since there is only one cylinder. There is therefore no need to select ‘Select All’  Select ‘Edit Selected’ to display the Piston Panel  Complete the Piston Panel as shown

 Select OK.

197


This completes the process of defining the cranktrain data other than defining the connecting rod reduced model and performing the matrix reduction for that model. This is dealt with in the next two steps. 2.6.3.4 Edit Cranktrain – Connecting Rod Model Definition It is necessary to define a reduced model of the connecting rod. The reduced model is a number of nodes and degrees of freedom that are a subset of the complete model. We need to define two sets.  A face set that defines the big end bearing, since this is bearing is an EHD model. For this set all the degrees of freedom of the bearing surface are included in the reduced model.  In addition we require a constrained node set that defines the small end bearing. For this set the bearing is defined by a single node with 6 degrees of freedom whose movement is the average of the nodes on the bearing surface.  Highlight ‘Connecting Rod’ as before

 Select ‘Define Model’. Messages will appear in the bottom of the Main Panel and the Define Model Panel will appear as shown

198


 In this action the program has attempted to automatically define the two sets using the data supplied previously – bearing position, diameter and length.  The reduced model is shown in orange  The set defining the big end bearing is incomplete and nothing has been defined for the small end bearing. This is because the nodes are outside tolerance.  It is necessary to define these sets using the Define Model Panel shown on the right. Consider firstly the face set defining the big end bearing  Select the Definition tab

 Select each tab and understand the default values  Each set is ‘clipped’ based on a geometric shape

199

KNOWLEDGE CENTRE Tutorials Tutorial 6: EHL Big End Bearing Analysis  Each shape is defined by a Centre, Axis, Extent and Diameter  Select the Tolerance tab

 The default linear tolerance is set to half the minimum distance between any two adjacent finite element nodes.  Each set has its own tolerance values  Change the Linear Tolerance to 0.5 mm and press ‘Clip Set’.

 Select ‘Add Set’ to add the set to the reduced model

200


Secondly consider the constrained node set defining the small end bearing  Select the Name tab  Change Type to Constrained Node and Name to Small End Bearing

 Select the Definition tab and again understand the default data  Select the Tolerance tab and again change the linear tolerance to 0.5 mm  Select ‘Clip Set’ followed by ‘Add Set’ to obtain the following

201


 Use the mouse and cursor keys to rotate and zoom the model if you haven’t already done so  OK the Define Model Panel This completes the process of defining the reduced model of the connecting rod

202


2.6.3.5 Edit Cranktrain – Connecting Rod Model Reduction  Select ‘Matrix Reduction’ to display the Matrix Reduction Panel

 For this tutorial we will use the FEARCE Vectorized Sparse Solver (VSS)  If you have limited memory then this option can be used to limit the memory during forward elimination and backsubstitution. This will result in a longer solution time. 

Select the Default output name  The output name specifies the name of the FEARCE .FRC command file that will be written.  Selecting the Default button will give the output name the same name as the .SFE file  Note: the include field is for adding extra commands directly to the run script – we will not need to use this feature in this example

 Switch on the Limited Memory toggle and specify a memory requirement of 250 Mb or less if your machine has less available memory

 Select Solve. The following Query Panel will be displayed.

203


 Loadcases 1 to 6 are reserved for 6 body loads in each of the six global directions.  Any additional loads the user may which to apply, such as bolt loads, must be applied using loadcase 7 and higher. Any number of loadcases may be used.  Select OK. Another Query Panel will be displayed.

 Select ‘Direct’. This will execute the command file and start the solution. A Progress Panel will be displayed.

 This model has approximately 100000 degrees of freedom which are reduced to 372 for the reduced model to be used by ENGDYN. On completion the message ‘Run completed successfully’ is written to the Main Panel and the following Information Panel is displayed.

204

KNOWLEDGE CENTRE Tutorials Tutorial 6: EHL Big End Bearing Analysis  OK this panel and the Matrix Reduction Panel 2.6.3.6 Edit Cranktrain – Model Assembly  Select ‘Assemble Model’  The following panel will appear as we are using an FE model for the connecting rod

 Click no as the FE model does not contain bolts etc. and so the values entered in the setting up of the model are more accurate

 This will assemble the connecting rod model.  Messages will be written to the Main Panel as shown  Save the model using the File menu from the top of the Main Panel 2.6.4

Solution

2.6.4.1 Define Lubricant Properties

205

KNOWLEDGE CENTRE Tutorials Tutorial 6: EHL Big End Bearing Analysis  Click on ‘Lubrication’ button from the buttons on the left hand side of the Main Panel

 Use the Browse button to select the lubricant RotellaX15W40.MAT from the working directory  By default on Unix the program will initially select the database directory at ../Ricardo/engdyn/3.0/database/Fluid  This database contains the most common lubricants  Use either Add or Update to add the lubricant



Click on OK

2.6.4.2 Define Loading Conditions The cylinder pressure diagram and any additional loadings (for example loads from Valdyn and gravity forces) are entered in this step.  Click on the ‘Loading’ button on the Main Panel

This will display the Loading Definition Panel as shown

206


 A number of different loading maps can be defined, Full Load, Part Load and No Load.  The solver will interpolate at speeds between those defined using this panel.  For this tutorial we will define a cylinder pressure at a single speed. Any solutions performed at other speeds will use this pressure diagram.  Type in a speed of 2000 rev/min  Set the Interval to 10 deg  The pressure file can contain just pressures at equal intervals as in this case at an interval of 10 deg or as an array of pressures and angles (which may be at unequal intervals)  Position the mouse over the File Name column and use the right button to display the pop-up menu. Use ‘Select Pressure file’ to select the file pcyl2000.PRES from your working directory. The panel will appear as shown.

 You may wish to remove the pathname in front of the file.  You may wish to inspect and understand the selected file using an appropriate editor  Define the Ambient and Crankcase Pressure as 1.0 bar  The crankcase pressure is only currently used for Hydrodynamic and Elastohydrodynamic bearing models to define the boundary condition at the edge of the bearing

207


 The completed panel will appear as shown

 The remaining tabs, Force Profile, Force Equation and Distortion need not be completed for this tutorial.  Use the Plot button to display the applied pressure loading as shown

 Save the model using the File menu from the top of the Main Panel  We now have sufficient data to proceed with an analysis

208


2.6.4.3 Evaluate Solution  Click on the ‘Evaluate Solution’ button on the Main Panel

The Evaluate Solution Panel will appear.

 Set the End Angle to 1080 deg  This is equivalent to three cycles and should be sufficient to obtain a converged solution of the big end bearing EHL solution  Define the Connecting Rod Model as Compliant  This ensures we run an EHL solution rather than a rigid bearing solution  Select the Cases tab and define a single speed of 2000 rev/min as shown

209


 Select the Bearing Model tab and complete the panel as shown

 The Half-Sommerfeld, Mass Conserving (Recommended) model is recommended for EHL calculations  This solver is mass conserving but without oil film history via the Reynolds cavitation condition.  The remaining tabs do not need to be completed for this solution since the other parameters are related to indeterminate or dynamic analyses.  Select ‘Define Oil Temps’ to display the Bearing Oil Temperatures Panel

210


 We will assume that the temperature of the oil in the bearing is at the inlet temperature  Select OK Select Solve Directly on the Evaluate Solution Panel.

 On completion of the analysis the summary file .EDSUM will be written.  This file contains summary data for the solution. Open this file with an appropriate editor to view the results.

211

HELP Using ENGDYN

B. HELP 1 Using ENGDYN 1.1 Overview The ENGDYN Graphical User Interface provides the user with complete control of the model generation, solution and results presentation phases from within an easy to use graphical environment. Complex models can be rapidly generated, solved and the results viewed, without the need to generate input files by hand. The ENGDYN model can be saved at any point during construction. ENGDYN uses a Ricardo binary standard data file to store both the model and the results of the solution and has an .EDSF suffix. When saving, ENGDYN stores the model at the current position and as the model is built up, so the filename.EDSF file is appended to. Similarly, once the simulation has been run, so the results are appended to this file.

1.2 Getting Started To start ENGDYN enter the following command: engdyn [-V ] [-debug ]

where: version

is the particular version that you wish to run. argument then the default version will be run.

debug

is used to control the console output. If you omit this argument then the output option selected using the Console Output Menu shown in Figure B-5 will be used. This argument can have the following options: 0 1 2 3

If you omit this

None Progress

Print Nothing Print messages that show the progress of the program. Verbose Print additional information Debug Print information to assist with debugging the program

filename is the name of the ENGDYN data file with an .EDSF suffix. If the suffix is omitted the program adds the suffix. Omit this argument if you wish to select or name a file, using the file browser. As an example, to run ENGDYN version 2013.1 with a data file named TEST.EDSF, the command line is: engdyn –V 2014.1 TEST.EDSF

213

HELP Using ENGDYN

1.3 Description of the Main Panel On starting ENGDYN the panel shown in Figure B-1 will be displayed. To the right it contains a display area that will be used to display a reduced model of the crankshaft being simulated. At the top of the panel is a menu-bar with four buttons, File, Options, View and Help. On the left hand side of the panel is a column of control buttons, and at the bottom a message area.

Figure B-1 Main Panel

1.3.1

Menu System

The main panel menu bar contains four menu buttons. Each of these buttons will display a menu if activated.

214

HELP Using ENGDYN 1.3.1.1 File File displays the menu shown in Figure B-2.

Figure B-2 File Menu It has the options: New

Clears the current model ready to start a new one

Open

Save As

Allows an existing .EDSF file to be opened. A file selection box will be displayed to allow a selection to be made Saves the current model to the current file name as shown in the top panel border Allows the current model to be saved with a new file name.

Plot Bitmap Print

Saves the main canvas display as a picture in jpeg, bmp, png or rgb formats Allows the drawing area to be printed

Delete Results Copy Design Exit

Discards the current results set

Save

Copies the model data without the results to a new file Quits the program

1.3.1.2 Options Options displays the menu shown in Figure B-3 and has 4 options.

Figure B-3 Options Menu Units Each time a fresh ENGDYN model is commenced, default units are used as displayed in

215

HELP Using ENGDYN the Default Units Entry panel shown in Figure B-4. Additionally, the Units option enables the user to edit the current default units that will be supplied when initially displaying the data entry panels for this particular model. It is important that any changes to the defaults should be made before the Engine Configuration Panel is selected. Units are defined as expressions in (more or less) the same way as expressions in C or FORTRAN. The operators +, -, * and / all have their conventional meanings and the same precedence as in C/FORTRAN. Parentheses (brackets) can be used to control the order of evaluation, which is otherwise from left to right. As well as using *, multiplication can be indicated by a . or by a space between two constants. Thus Newton metre can be written as N*m, N.m or N m. Exponentiation can be indicated by ^, ** or by placing the numeric exponent immediately after a unit name with no intervening white space. Thus metre squared can be written as. m^2, m**2 or m2. Unit names may be prefixed by an SI multiplier. M, mega, milli, m are all examples. Reset resets the units to their default values.

Figure B-4 Default Units Entry Console Output This displays the menu shown in Figure B-5. This menu controls the amount of information that is printed to the terminal window during the execution of the program. The default is Progress.

Figure B-5 Console Output Menu

None

Print nothing

216

HELP Using ENGDYN Progress

Print messages that show the progress of the program

Verbose

Print additional information.

Debug

Print information to assist with debugging the program

Paper Size This displays the menu shown in Figure B-6. This menu controls the size of paper for plotted output. The default is A4.

Figure B-6 Paper Size Menu Tolerances This displays the panel shown in Figure B-7. This panel allows setting linear and angular tolerances for FE model reduction.

Figure B-7 Tolerances Panel Titles This is used to define the titles used for post-processing, e.g. plotting of results. It will pop up the panel shown in Figure B-8. If Reset is pressed, the default titles found in the .EDSF file will be reloaded.

Figure B-8 Titles Panel Toggle Logo This is used to toggle on and off ENGDYN logo on the display area.

217

HELP Using ENGDYN 1.3.1.3 View View displays the menu shown in Figure B-9 and has 1 option.

Figure B-9 View Menu Model Appearance This option pops up the panel shown in Figure B-10 and is used to control the display of the various Engdyn models.

Figure B-10 Model Appearance Panel It is divided into 5 parent tabs, each with a differing number of child tabs, for instance the Crankshaft tab has child tabs for Reduced Model, Stiffness Model and Stress Model. The applicability of these tabs depends on the analysis type and what the stage the model definition/analysis is at, e.g. the Crankshaft, Reduced Model is available once the engine has been configured and the Cylinder Block, Reduced Model is available once the Block has been configured. The Bearing controls the display of the lubrication mesh defined for each of the journal bearings described in sections 3.4.1.4 and 3.4.1.5. The Set refers to sets defined during the Edit Block process. The contents of all tabs are the same: Visible This toggles the visibility of the relevant model. Outline This controls how the outlines of the external element faces are drawn. The options are as follows: None

No outlines.

Edges

Outlines where two external element faces meet at a sharp angle.

218

HELP Using ENGDYN Mesh

Outlines of all the external element faces.

Surfaces This controls whether surfaces are drawn. Mid-side If the model contains parabolic elements the selection of this option will mean that the mid-side nodes are used in the drawing. Switching this option on will slow the drawing of the model but will improve its appearance. Colour These options control the colour of the lines or surfaces of the finite element model. Lighting This option gives the appearance of lights shining onto the model. The surfaces will change in brightness depending on their position and angle. Lighting will only make a visible difference if the element surfaces are drawn. Transparency This option gives the appearance of transparency to a model. The transparency is varied via the slider bar. Transparency will only make a visible difference if the element surfaces are drawn. Clip Plane This enables a clip plane to be defined in the global YZ, XZ or XY plane at a userdefined distance from the origin using the option menus and position field 1.3.1.4 Help Help displays the menu shown in Figure B-11 and has only 2 options.

Figure B-11 Help Menu Overview This opens an internet browser to display html based ENGDYN help. Online Support This opens Ricardo Software support webpage About This displays a panel with information about the program.

219

HELP Using ENGDYN 1.3.2

Message Box

At the bottom of the main panel is a message box. This box is used to display the status of the model.

1.3.3

Display Area

The models of the crankshaft and cylinder block will be displayed once either an existing file has been loaded, or once data has been successfully entered during model generation. The displayed elements and nodes of the reduced models of the crankshaft and cylinder block can be individually selected by clicking on them with the mouse. This enables the user to edit the input parameters and view the results at those nodes or elements. The currently displayed models as defined using the View menu button can be manipulated using the mouse or keyboard, as defined in Table B-1 and Table B-2. Button Left Left (and drag) Left Left (and drag)

Key Shift Shift

Function Pick Pick Pick Pick

Middle Middle

Shift

Translate Rotate

Middle Right

Ctrl -

Zoom None

Description Replaces the currently picked item(s) Replaces the currently picked item(s) Adds to the currently picked item(s) Adds to the currently picked item(s) if more than one item is in the drag box. If only one item is in the drag box that item is toggled. The model follows the mouse movements Near the centre gives up/down or left/right rotation. Near the edge gives clockwise/anti-clockwise rotation Moving up zooms in, moving down zooms out.

Table B-1 Mouse Functions Key(s) x SHIFT and x y SHIFT and y z SHIFT and z r i o

Function Rotates the model anti-clockwise about its x-axis Rotates the model clockwise about its x-axis Rotates the model anti-clockwise about its y-axis Rotates the model clockwise about its y-axis Rotates the model anti-clockwise about its z-axis Rotates the model clockwise about its z-axis Resets the model to its original orientation Zooms in Zooms out Table B-2 Keyboard Functions

220

HELP Using ENGDYN 1.3.4

Control Buttons

The control buttons on the left of the main panel are divided into three groups. The first group controls the model building functions, the second group controls the solution and the third group controls the presentation of results. The graphical interface has been designed in a way such that a complete analysis will involve a natural progression through each of the control buttons in turn. In a number of cases, one panel may have to be completed before progression to a subsequent panel is allowed. If the user tries to progress before completing the dependant panels, a warning message will be displayed highlighting what actions are required.

221

HELP ENGDYN Models

2 ENGDYN Models 2.1 Overview The ENGDYN model consists of two components, the crank train and the cylinder block. These components are defined using models of varying complexity dependent on the application. Input loads due to the reciprocating motion of the piston and connecting rod, the rotating motion of the connecting rod and the force due to the cylinder gas pressure are applied at each crank pin and at each cylinder of the engine. Reaction forces at each of the main journal bearings and at the thrust bearing are calculated using a number of solution methods.

2.2 ENGDYN Co-ordinate System Two co-ordinate systems are used by ENGDYN, one that is fixed relative to the cylinder block and the other that is relative to the crankshaft and rotates with it. At 0 degree CA these co-ordinate systems are coincident. The fixed co-ordinate system is defined as shown in Figure B-12.

223

HELP ENGDYN Models

Figure B-12 ENGDYN Co-ordinate System The x-axis lies on the crankshaft axis and in the direction corresponding to clockwise rotation. The y-axis is aligned with the cylinders for an in-line configured engine and is aligned to be equidistant between the two banks for a Vee-configured engine. The zaxis completes a right-handed axis set.

2.3 Finite Element Model Data The Ricardo Standard Data File (SDF) used by ENGDYN is also used to store any finite element models that are used. Each finite element model including its geometry, loads, constraints and results is translated from the finite element package being used to a SDF known as the Standard Finite Element File (SFE). ENGDYN then uses this file rather than files native to the finite element package to read and write data for analysis. The main advantage of this system is that the reading and writing of data to and from different finite element packages is greatly simplified.

224

HELP ENGDYN Models 2.4 Crank Train Model 2.4.1

Definition

The crank train model consists of the crankshaft and the flywheel and crank nose assemblies. Nodes each with six degrees of freedom interconnected by elements define the crank train from its free-end to its output to the drive train. Currently, the Graphical User Interface (GUI) allows only crank train models to be defined that are de-coupled from the rest of the drive train. For more experienced users the command file data file environment1 provides a means of modelling a crank train that includes the drive train. For example, an engine that drives a generator set through a fixed drive.

2.4.2

Model Types

The crank train model may be defined using one of five model types and these are listed below:  Rigid Massless  Rigid  Compliant Massless  Compliant  Dynamic

 The elements are rigid and the crank train is assumed to have no mass  The elements are rigid and include lumped masses  The crank train is flexible defined by a stiffness matrix but has no mass  The crank train is flexible defined by a stiffness matrix and includes lumped masses  The crank train is dynamic defined by mass and stiffness matrices

It is recommended that a finite element model be used to derive the lumped mass properties and the mass and stiffness matrices of the crank train particularly when defining a model that is either compliant or dynamic. Alternatively mass properties and stiffness’ can be entered interactively. This is particularly useful when defining a rigid model with mass moments included. Crank train Model Type Rigid Massless Rigid Compliant Massless Compliant Dynamic

Solution Method Static Determinate Indeterminate          

Dynamic     

Table B-3 Crank Train Model Type and Solution Method Table B-3 shows the solution methods that are possible for each of the model types. The solution methods are as follows:  Statically-determinate

- The reactions at each of the main journal bearings are

225

HELP ENGDYN Models calculated assuming the crankshaft is pin-jointed at each of the bearings. The journal orbits are calculated from these reactions using the Mobility Method.  Staticallyindeterminate

- The crankshaft and cylinder block (if included) are assumed to be compliant. The reactions and journal orbits at each of the main journal bearings and the quasistatic displaced shape of the crankshaft (and cylinder block) are calculated using the Mobility Method at each main bearing. This method accounts for load sharing between bearings.

 Dynamic

- The crankshaft is assumed to be dynamic and includes the non-linear gyroscopic effects. The reactions and journal orbits at each of the main journal bearings and the 3-dimensional vibratory behaviour of the crankshaft are calculated. The cylinder block can be either modelled as rigid, compliant or dynamic.

2.4.3

Matrix Formulation

The mass and stiffness matrices used for a dynamic model of the crank train are derived by Static Condensation of the finite element model. The techniques used to condense the crank train model were first used by Hodgetts 2345 in the development of 3dimensional crankshaft vibration methods, and subsequently applied to various types of engine by Ricardo 67. The crank train is simplified using a discreet mass-elastic system, idealised to rigid, massive nodes interconnected by flexible, massless elements. The stiffness matrix used for a compliant model is the same as that used for a dynamic model.

2.4.4

Geometry

The geometry of the engine is defined using the GUI. The interface positions the nodes of the reduced crank train model such that;  there is a node at the centre of each journal bearing,  there are two nodes in each crank web on the mid-plane equidistant from the facing lands of each journal of the web, and  there is a node at the crank nose and at the flywheel. It is assumed that the cylinder axis is coincident with the centre of the journal bearing at each crank pin.

226

HELP ENGDYN Models 2.4.5

Finite Element Model

2.4.5.1 General A stiffness model of the crank train (excluding any journal fillet or oil hole definition) is sufficient in defining the mass and stiffness characteristics of the crank train. An example model for a V8 crank train together with its equivalent ENGDYN reduced model is shown in Figure B-13.

Figure B-13 Example of a V8 Crank Train The finite element model is used to derive the lumped mass properties and the mass and stiffness matrices of the crank train dependent on the model type. This is summarised in Table B-4. Crank Train Model Type Rigid Massless Rigid Compliant Massless Complaint Dynamic

Lumped Mass Properties     

Mass Matrix     

Table B-4 Use of Finite Element Model

227

Stiffness Matrix     

HELP ENGDYN Models

Figure B-14 Example Finite Element Models of a V8 Crank Train Figure B-14 shows four different finite element models of the V8 crank train. For each model the equivalent ENGDYN reduced model is shown. The finite element model must consist of at least the crankshaft as shown in view 1, but may consist of the complete crank train including the crank nose and flywheel assemblies as shown in view 4. Components of the crank train not included in the finite element model, for example, the clutch plate assembly are defined separately. These components are treated as lumped masses and do not contribute any stiffness. The finite element model must not include the stiffness element and seismic mass of any vibration dampers that may be included in the crank train since these are defined separately. 2.4.5.2 Geometry of the Finite Element Model The finite element model of the crank train must satisfy a number of conditions and these are as follows:  A cut-plane in the yz-plane must exist at the centre of each main and pin journal bearing. The mesh at these cut-planes may be discontinuous.

228

HELP ENGDYN Models  At each pin journal cutplane three nodes must exist so as to apply restraints to each crank web for performing the stiffness calculations. The positions of these nodes are defined in Figure B-15 for both a solid and hollow journal. The position of the nodes do not have to be exact but as near as possible to those positions shown in the figure. For a hollow journal the node n1 is taken as the node nearest the journal centre.

Figure B-15 Restraint nodes on Crank Pin Journal Cut-Planes

 At the front and rear main journal cut-planes three nodes must exist so as to apply restraints to the shafts for performing the stiffness calculations. The positions of these nodes are defined in Figure B-16 for both a solid and hollow journal. The positions of the nodes do not have to be exact but as near as Figure B-16 Restraint nodes on Front and Rear Main Journal Cut-Planes possible to those positions shown in the figure. For a hollow journal the node n1 is taken as the node nearest the journal centre.  A cut-plane in the yz-plane must exist at the crank nose node. The position of this node; and therefore the cut-plane, should be either at the centre of the vibration damper if fitted, or, alternatively, if no vibration damper is fitted, at a position at the centre of a pulley at which a belt, chain or gear is driven.  A cut-plane in the yz-plane must exist at the flywheel node. If the flywheel assembly is included in the finite element model as shown in view 4 of Figure B-14, the position of the node and therefore the cut-plane should be at the face at which the clutch plate is attached. Conversely, if the flywheel is not included in the finite element model as shown in view 1 of Figure B-12, the position of this node should be at the face at which the flywheel is attached.

229

HELP ENGDYN Models  The finite element should not contain any constraint equations other than those joining the model at the cut-planes described above. 2.4.5.3 Derivation of Mass Properties For those crank train models with mass the reduced ENGDYN model requires lumped masses at each of the crank train nodes. ENGDYN divides the finite element model into a number of element sets where each set corresponds to the lumped mass at a crank train node. This is shown for the V8 example in Figure B-17. The mass properties of each of these sets are calculated and assigned to the appropriate node.

Figure B-17 Element Sets Defining Lumped Mass at each Crank Train Node 2.4.5.4 Derivation of Stiffness Properties ENGDYN calculates for compliant and dynamic crank train models the stiffness for each of the crankshaft webs and for the elements at the ends, performing a finite element solution in each case using the procedures described in Appendix 1. A Symmetric Conjugant Gradient solution algorithm is used to perform the solution. Three stiffness values, torsion, in-plane bending and out-of-plane bending are calculated for each crank web. The elements at the ends of the crank train are assumed to be axisymmetric, and for each of these elements three stiffness values, torsion, bending and axial are calculated. ENGDYN sub-divides the finite element model into separate files that each contains the finite element model of a crank web or an end element. If the file name for the V8 example above is V8.SFE 22 files are created and these are as follows: V8_WEB1B.SFE V8_WEB1T.SFE

V8_WEB5B.SFE V8_WEB5T.SFE

230

V8_ELEM01A.SFE V8_ELEM01B.SFE

HELP ENGDYN Models V8_WEB2B.SFE V8_WEB2T.SFE V8_WEB3B.SFE V8_WEB3T.SFE V8_WEB4B.SFE V8_WEB4T.SFE

V8_WEB6B.SFE V8_WEB6T.SFE V8_WEB7B.SFE V8_WEB7T.SFE V8_WEB8B.SFE V8_WEB8T.SFE

V8_ELEM01T.SFE V8_ELEM30A.SFE V8_ELEM30B.SFE V8_ELEM30T.SFE

Multiple files are required for each crank web and element because the constraints for each stiffness calculation are different. For each crank web two files are created and for web number 1 these are V8_WEB1B.SFE and V8_WEB1T.SFE. These are used to calculate the bending and torsional stiffness’ respectively. On completion of the stiffness calculations each file contains the constraints and displacement loads as described in Appendix 1 and the nodal displacements and reactions due to the applied loads and constraints. Each crank web model is orientated such that the main bearing centre is at the origin and the crank pin lies in the xy-plane. ENGDYN calculates the mass properties of each crank web to determine webs that are identical to each other. The stiffness calculations are then only performed on those webs that are unique. For each element at the ends three files are created and for element number 30 these are V8_ELEM30A.SFE, V8_ELEM30B.SFE and V8_ELEM30T.SFE. These are used to calculate the axial, bending and torsional stiffness’ respectively. On completion of the stiffness calculations each file contains the constraints and displacement loads as described in Appendix 1 and the nodal displacements and reactions due to the applied loads and constraints. Each element model is orientated such that one of the cut-planes is at the origin.

2.5 Cylinder Block Model 2.5.1

Overview

The cylinder block model may be represented in a number of ways either using a finite element model or as simple stiffnesses at each of the main bearings or may instead be excluded from the ENGDYN model. If a finite element model is used the model may be defined as a compliant or dynamic model and is defined by lower-triangular stiffness matrix and for a dynamic model by a lower-triangular mass matrix.

2.5.2


The finite element model of the cylinder block may be a model of the cylinder block assembly including a representation of the cylinder head or of the complete power train including the transmission and any ancillary components. The finite element model is included in the ENGDYN model as a reduced system defined by a reduced set of degrees of freedom. Matrices are derived for this reduced system using an appropriate finite-element package using either static condensation or Component Mode Synthesis (CMS)

231

HELP ENGDYN Models

2.5.3

Definition of Model Types

2.5.3.1 Rigid The cylinder block is assumed to be rigid. In this case only the axial stiffness to ground of the thrust bearing is considered as part of the dynamic solution, although the geometry of a reduced model can be defined if required. 2.5.3.2 Compliant Bearing The cylinder block is assumed to have compliance at each bearing where each bearing is represented by simple stiffness’ to ground. 2.5.3.3 Compliant The cylinder block is derived using a finite element model as described in 2.5.2 and is defined by a reduced stiffness matrix evaluated using an appropriate finite-element package. 2.5.3.4 Compliant Crankcase The cylinder block is derived using a finite element model of the lower crankcase that does not include the cylinders. Otherwise the model is as described in 2.5.2 and is defined by a reduced stiffness matrix evaluated using an appropriate finite-element package in a similar manner to that described above for the Compliant model. 2.5.3.5 Dynamic The cylinder block is derived using a finite element model as described in 2.5.2 and is defined by reduced stiffness and mass matrices evaluated using an appropriate finiteelement package either by Component Mode Synthesis. The advantage of including a dynamic model of the cylinder block in the ENGDYN model is to include the influence of the dynamic modes of the structure within the solution. It is therefore necessary to include a model that correctly models the dynamics of the cylinder block and as a consequence it is more likely that the user will include a model of the power train rather than of the cylinder block assembly only.

2.6 Journal Bearing Oil Film Model 2.6.1

General

Journal bearings support the rotating crankshaft and therefore predicting the oil flow and temperature of within the bearings is essential to calculate their behavior. The oil film affects both the pressures and temperatures applied to the bearing. The shear stress in the oil film generates heat which is transported to the bearings whilst the flow of oil

232

HELP ENGDYN Models through the bearings aids their cooling. Furthermore, high hydrodynamic pressure in the oil film layer deforms the bearing shell. ENGDYN offers two methods to simulate the bearing oil film, a simplified and a complex model. The simplified model is based on the Mobility Method and is described in Section 2.6.2. This model enables to model the journal bearings relatively quickly. On the other hand the bearing profile is neglected. A more complex model is based on the direct numerical solution of the oil flow in the bearing using the Reynolds equation. This method is described in Section 2.6.3. ENGYN allows for the effective oil temperature in the bearing to be user-defined or derived by considering the energy dissipated by the oil through the bearing action as described in described in Section 2.6.42.6.4. ENGDYN assumes that the bearings and journals are perfectly aligned initially when the engine is not running. Misalignment of the bearing and journal should not be included in the crank train and cylinder block models. ENGDYN requires for each journal bearing the following data:  Bearing type (Plain, Partially-Grooved or Fully-Grooved)  Radial clearance  Effective bearing shell length and bearing diameter  Bearing shell groove width (if the bearing is grooved)  Circumferential extent of the groove (if the bearing is grooved)  Oil feed hole diameter and circumferential position (if the bearing is plain) ENGDYN also requires the viscosity-temperature characteristic of the oil used in the engine together with the temperature and pressure of the oil as it is supplied to the bearing. 2.6.2

Mobility Based Solution

The journal bearing oil film model used by ENGDYN is a fully non-linear model based on methods by Booker8 and Kikuchi9. For radial motion (shaft axis parallel to bearing axis) of the shaft in the bearing the oil film model is based on the mobility method of Booker8, using the short bearing approximation (Ocvirk Solution). In its usual form, the method is used to determine the shaft eccentricity and attitude angle for a known force. This is the method used for the quasistatic solution methods and for the dynamic solution when the cylinder block model is complaint. However when the cylinder block is rigid or dynamic for a dynamic solution the method is inverted so that the oil film load and its direction can be determined from a known shaft position and velocity. For rotational motion (shaft axis tilting relative to bearing axis) the short bearing approximation is also used to determine the stiffness and damping of the oil film. This is based on the work of Kikuchi9. Kikuchi modelled the oil film as a set of radial and

233

HELP ENGDYN Models rotational springs and dampers. The rotational springs and dampers resist changes of inclination and rate of inclination. Kikuchi derived coefficients for these springs and dampers assuming the short bearing approximation, and based on work that considered an unbalanced rotating shaft system with a static load acting at each bearing. ENGDYN uses the coefficients of the rotational springs and dampers to provide resistance to inclination at each main journal bearing. The instantaneous oil flow rate through the bearing and the power dissipated by the oil through the bearing action are calculated by ENGDYN for each journal bearing10. These instantaneous values are integrated numerically to estimate the overall cyclic oil flow rate and power loss.

2.6.3

Solution of Reynold’s Equation on Computational Mesh

The hydrodynamics of the thin oil layer is described by the Reynold’s equation. This partial differential equation is derived from Navier-Stokes equations describing the general fluid flow. The Reynold’s Equation is solved numerically on a computational mesh called the lubrication mesh. In the Engdyn solver, the Reynold’s equation can be solved for with or without accounting for the surface roughnesses. The extension to include the surface roughnesses is called the Average Flow Model. In order to treat cavitation within the journal bearing oilflim, various boundary conditions have been implemented. The methodologies vary in their complexity and the oilfilm behavior that they capture. An increase in complexity generally leads to an increase in computational time. The Half-Sommerfeld Boundary Condition is the simplest option and is equivalent to the ENGDYN 4.0 ‘Finite Difference’ method. This robust method does not rely on numerical iterations and therefore the solution is obtained in a relatively short time. This method is capable of modeling the presence of holes and grooves in the bearing. The Half-Sommerfeld Boundary Condition, Mass Conserving model is an extension of the standard Half-Sommerfeld Boundary Condition and will be further extended in future versions of ENGDYN to simulate oil feed holes or grooves that move across the computational mesh allowing for oil transport between bearings. The Reynold’s Boundary Condition is more advanced. The cavitated area is determined iteratively and therefore the solution requires a longer computational time. The Reynold’s Boundary Conditions with Jakobsson Floberg Olsson Condition is the most advanced method. It combines the Reynolds equation with the Reynolds boundary condition and the continuity equation. In the cavitated area, the continuity equation is solved iteratively and the Reynolds equation is solved in the rest of the bearing. This detailed approach incurs a higher CPU time.

234

HELP ENGDYN Models 2.6.4

Thermal Balance Solution

ENGDYN allows alternative methods for determining the oil temperature used to calculate the viscosity for each journal bearing. Firstly, the oil temperature may be specified directly by the user. Secondly the user may specify the oil supply temperature and pressure and ENGDYN will calculate the temperature rise due to the bearing action from the calculated power loss and oil mass outflows and inflows.The solution proceeds until the oil temperature at each journal bearing has converged to a user specified tolerance. ENGDYN offers several thermal balance solutions. The models compatible with the Mobility Method are: Cycle Averaged Equilibrium Temperature - an iterative algorithm searches for a cycle averadged equilibrium oil temperature of oil. The temperature is averaged across the whole cycle and the entire bearing. Using this method temperature all the heat inflows and outflows over the cycle are in equilibrium. T0D Oil Model – Averaged Temperature over Whole Bearing – an iterative algorithm searches for the equilibrium temperature in each timestep. The effect of heat accumulation in the oil held in the bearing is included. The models compatible with Reynold’s equation solution are: T0D Oil Model – Averaged Temperature over Whole Bearing – an iterative algorithm searches for the equilibrium temperature in each timestep. The effect of heat accumulation in the oil held in the bearing is included. T1D Oil Model - Temperatures Averaged Circumferentialy - the iterative algorithm searches for the equilibrium temperature in each timestep and for each axial slice of the bearing. This model simulates the behavior of the oil flowing from the feed holes or grooves located in bearing center to the bearing axial edges. The effect of heat accumulation in the oil held in each axial slice of the bearing is included.

2.7 In-Cylinder Model The loads applied to each cylinder and crank pin due to the gas cylinder pressure and the acceleration of the piston and connecting rod may either be derived assuming the mechanism behaves simply as a slider crank, or alternatively the loads can be calculated using the Ricardo program PISDYN11. In this program the secondary dynamics of the piston are considered and the loads are imported in to ENGDYN from a Ricardo Standard Data File (sdf). If it is assumed that the piston and connecting rod behave as a slider crank then ENGDYN requires the mass of the piston assembly and the mass and polar inertia of the connecting rod. It also requires the length of the connecting rod and the position of its centre of mass. It is assumed that the connecting rod is not articulated. If loads are derived using PISDYN and then this data is read from the sdf file.

235

HELP ENGDYN Models 2.8 Gas Cylinder Pressure ENGDYN creates gas cylinder pressure maps that are defined by the user using a series of cylinder pressure diagrams against speed. The cylinder pressure diagram may be different for each cylinder. The data is either read from ASCII files or if in-cylinder loads are being derived using PISDYN as described above then this data is read from the sdf file. Each map may be named as Full Load, Part Load or No Load. The cylinder pressure maps are stored as absolute pressure but may be supplied as absolute cylinder pressure diagrams with an ambient condition. If loads have not been derived using PISDYN then during solution ENGDYN evaluates cylinder pressures at a given engine speed and load by interpolation from the appropriate map.

1

Ricardo Report DP97/2300 ‘ENGDYN Revision 1.0 Engine Dynamics Simulation Program User Manual’

2

I.Mech.E C99/71, 1971 ‘Vibrations of a Crankshaft’. D.Hodgetts, Advanced School of Automobile Engineering, Cranfield.

3

PhD Thesis (University of London) ‘The Vibrations of Crankshafts’. D.Hodgetts, Cranfield Institute of Technology.

4

I.Mech.E C216/76,1976 ‘The Whirl Modes of a Crankshaft’, D.Hodgetts, Cranfield Institute of Technology.

5

FISITA,1986 ‘The Dynamic Response of Crankshafts and Camshafts’, D.Hodgetts, Cranfield Institute of Technology.

6

I.Mech.E AD Autotech Congress 1987 ‘The Measurement and Prediction of Flywheel Whirl’. A.R.Heath, Ricardo.

7

I.Mech.E C14/87, 1987 ‘Computers in Analysis Techniques for Reciprocating Engine Design’. D.J.Lacy, Ricardo.

8

Transactions of the ASME, journal of Basic Engineering, Sept. 1965 ‘Dynamically Loaded Journal Bearings - Mobility Method Of Solution’. J.F.Booker , Cornell University

9

Bulletin of the JSME, 1970 ‘Analysis of Unbalance Vibration of Rotating Shaft System with Many Bearings and Disks’. K.Kikuchi.

10

Ricardo Report DP96/1722 ‘Summary of the Journal Bearing Oil Film Thermal Balance Calculation Methods used in ENGDYN and OILFILM’

11

Ricardo Software, April 2009, ‘PISDYN Manual – Documentation / User’s Manual Version 5.1’

236

HELP Model Generation

3 Model Generation 3.1 General This section describes how the engine model can be generated using the Graphical User Interface.

3.2 Configure Engine When generating a new model, the first step is to define the dimensions of the engine. This is done by selecting the Configure Engine button from the top left of the Main Panel. This opens the Engine Configuration Panel, as shown in Figure B-18. This panel must be completed before progressing to Define Models. The panel consists of a number of tabs that are described below. The first image of Figure B-18 shows the Engine tab selected. As data are entered, a drawing of the engine will be displayed showing the current number of Cylinders and Main Journals, together with their spacing. It should be noted that this drawing is for the user’s aid only, and is not drawn to scale. It is possible to pan and zoom this drawing using the mouse functions defined in Table B-1. It is assumed that the engine rotates in a clockwise direction and the connecting rod is not articulated. The geometry of the crankshaft, between the front and rear main bearings, is calculated when OK is pressed. If the data are invalid then the user will be warned, and the panel will remain displayed. The co-ordinate system used by ENGDYN is defined in 2.2 and shown in Figure B-12. Two titles may be assigned to the model in the field at the top of the panel. These titles will be displayed on the main panel, and will appear in all text and on all graphical output. These can be changed, if required, during post-processing.

Figure B-18 Engine Configuration Panel

237

HELP Model Generation 3.2.1

Engine

No. of Cyls This specifies how many cylinders the engine has. No. of Mains This specifies how many main journal bearings the engine has. Type This is used to specify whether the engine is an In-Line or Vee configuration. Vee Angle For a Vee engine, this is used to specify the angle between the banks. Cycle This is used to specify the engine cycle, 2-stroke or 4-stroke. Numbering This menu is used to define whether the numbering system starts at the nose of the crankshaft (Ascending) or at the flywheel (Descending). Convention This menu which only applies to Vee engines defines the numbering pattern of the cylinders. Whenever the numbering pattern is modified, any defined firing order will be automatically updated. Front Bank This menu only applies to Vee engines and is used to specify which cylinder (Left or Right) is at the front of the engine (nearest the nose of the crankshaft) Firing Order This is used to define the firing order for the engine (starting at cylinder 1). Firing orders are entered as a series of numbers separated by dashes. For example, a valid firing order for a V8 engine might be: 1-8-5-4-7-2-3-6 Selecting Options will display the Firing Order Entry Panel shown in Figure B-19.

238


Figure B-19 Firing Order Entry Panel This displays the appropriate firing orders for that No. of Cyls and engine configuration Type. The user can either select the desired firing order from this list, or type in their own firing order as a dash separated list. The firing orders will be automatically updated whenever the cylinder numbering system is changed. Alternatively a list of comma separated firing angles may be entered for each cylinder. For the V8 example given above the list of equivalent firing angles could be entered as 45.0,495.0,585.0,315.0,225.0,675.0,405.0,135.0. Rotation This is used to define the direction that the engine rotates, either Clockwise or AntiClockwise.

3.2.2

Cylinders

The second image of Figure B-18 shows the Engine Configuration Panel with the Cylinders tab selected. As data are entered, the cylinders will be displayed showing the axial datum and bank offset (if the engine is a Vee engine) together with the cylinder spacing. Axial Datum This is used to define the absolute position of the first cylinder along the crankshaft from the crank nose. Bank Offset This is used to define the distance between the leading and the trailing banks for a Vee engine. It must have a positive value. The following data relate to the geometry of each cylinder. Inclination Selecting the Edit will display the Cylinder Data Panel shown in Figure B-20.

239


Figure B-20 Cylinder Data Panel Displaying Inclination Angles This is used, if required, to edit the angle of inclination of each cylinder about the x-axis from y as defined in 2.2. These angles are reset if either the Vee Angle or numbering Convention is changed. Data This is used to specify whether the data for each cylinder is Uniform or Non-Uniform. In the case of a Vee engine an additional option Uniform Bank may also be selected. If Data is set to Uniform, then the values of Bore, Crank Throw, Cylinder Offset, Piston Offset and Conrod Length are defined for the engine as described below. If Data is set to Non-Uniform then these data may be edited for each cylinder by selecting the Edit button, which will display the Cylinder Data Panel shown in Figure B-21. If the data is non-uniform, then it is recommended that the user start with Uniform to define the dominant dimensions before switching to Non-Uniform or Uniform Bank to make any required changes.

Figure B-21 Cylinder Data Panel Displaying Non-Uniform Data Similarly if Data is set to Uniform Bank and Edit is selected then the Cylinder Data Panel is displayed with a row for each bank as shown in Figure B-22.

240


Figure B-22 Cylinder Data Panel Displaying Data Uniform for Each Bank Bore This is the cylinder bore. Crank Throw This is the crank throw radius r. Cylinder Offset This is the perpendicular distance of the cylinder axis c from the crankshaft axis. Bearing Offset This is the axial offset of the big-end bearing centre from the cylinder centre and is typically equal to zero. The offset is measured from the cylinder centre axis. A positive value denotes that the bearing is to the rear of the cylinder, whilst a negative value denotes that the bearing is in front of the cylinder. If an offset is specified then there will be moment with respect to the big end bearing centre due to the offset cylinder forces. For reasons of packaging Vee-configured engines often have a small offset. For one bank the big end bearing offset will be to the rear of the cylinder centre (+ve), whilst for the other bank the offset will be in front of the cylinder (-ve). Piston Offset This is the piston pin offset, p measured from the piston axis and is positive if the pin centre is towards the anti-thrust side. Conrod Length This is the distance between the bearing centres of the connecting rod. The piston stroke, s for each cylinder is calculated from the above data given by the equation

s where

l = r = c= p=

l  r 2  c  p 2  l  r 2  c  p 2

Length of the connecting rod between the bearing centres Crank throw Cylinder offset Piston pin offset

241

HELP Model Generation Spacing This is used to define the axial distance between the centres of adjacent cylinders centres. If Data is set to Non-Uniform or Uniform bank, then clicking on the value in the drawing (with the left mouse button) will bring up the Value Entry Panel, so the values can be edited individually. If the data is non-uniform, then it is recommended that the user start with Uniform to define the dominant dimensions before switching to NonUniform to make any required changes.

3.2.3

Main Journals

The third image in Figure B-18 shows the Engine Configuration Panel with the Main Journals tab selected. As data are entered, the main journals will be displayed showing the axial datum and the diameter and length of each journal together with the journal spacing. The thrust bearing is also annotated. Axial Datum This is the absolute position of the centre of the main journal farthest from the flywheel relative to the origin. It should be noted that the origin can be anywhere along the crankshaft axis and, for convenience, is often taken to be at the same location as on the crankshaft drawing (normally at a thrust face). Thrust Brg No. This is used to specify which of the main journals is a thrust bearing. Data This is used to specify whether the data for each main journal is Uniform or Non-Uniform. If Data is set to Uniform, then the values of Journal Length, Journal Diameter and Spacing are defined for the engine as described below. If this option is set Non-Uniform then clicking on the value in the drawing (with the left mouse button) will bring up the Value Entry Panel, so the values can be edited individually. If the data is non-uniform, then it is recommended that the user start with Uniform to define the dominant dimensions before switching to Non-Uniform to make any required changes. Journal Length This the length of the main journal between the facing lands of the adjacent crankshaft webs, as shown in Figure B-23. This figure also shows the definition of the main journal length for the journal nearest the nose of the crankshaft. The centre-lines shown in the figure are the centres of the bearing shells.

242


Figure B-23 Main Journal Length Journal Diameter This is the diameter of the main journal, as shown in Figure B-23. Spacing This is the spacing between the centres of the main journals. Offset By default the bearing centre is assumed to be equidistant between the facing lands of the two adjacent webs and the offset will be equal to half the Journal Length. This is shown the above figure. This offset allows the bearing centre to be offset, and would be used typically when one of the adjacent webs, has a shoulder.

3.2.4

Big End Journals

The fourth image in Figure B-18 shows the Engine Configuration Panel with the Big End Journals tab selected. As data are entered, the pin journals will be displayed showing the diameter and length of each journal together with the journal centre offset. Crankpin Offset For a Vee engine, this is used to specify the angle between the pin journals used by the opposing banks of a single bay. If this is set to zero then the crankshaft will have a single crank-pin journal for each bay. Data This is used to specify whether the data for each pin journal is Uniform or Non-Uniform. If Data is set to Uniform, then the values of Journal Length, Journal Diameter and Offset are defined for the engine as described below. If this option is set Non-Uniform then clicking on the value in the drawing (with the left mouse button) will bring up the Value Entry Panel, so the values can be edited individually. If the data is non-uniform, then it is recommended that the user start with Uniform to define the dominant dimensions before switching to Non-Uniform to make any required changes.

243

HELP Model Generation Diameter This is used to define the diameter of the pin journal. Length This is used to define the length of the pin journal between the facing lands of the adjacent crankshaft webs. For a Vee engine in which the Crankpin Offset is zero this field is ghosted. In this case the overall length of the journal is derived from the sum of the two offsets and the bank offset. Offset This is used to define the position of the facing land of the adjacent crankshaft web. An offset that is not half the journal length allows the bearing centre to be offset from the journal centre.

3.2.5

Generating the Reduced Cranktrain Model

Once all the engine data have been entered, the crankshaft geometry can be generated by clicking OK or APPLY. The geometry of the crankshaft model is then displayed in the Main Panel, as shown in Figure B-24.

244


Figure B-24 Main Panel The nodes of the reduced crankshaft model are positioned such that;  there is a node at the centre of each journal bearing,  there are two nodes in each crank web on the mid-plane equidistant from the facing lands of each journal of the web, and  there is a node at the crank nose and at the flywheel. The nodes at the crank nose and at the flywheel are positioned half a main journal length from the nodes at the adjacent main journals. The user later defines the exact positions of these nodes whilst editing the crank train as described in Section 3.4. The size of the circles representing the crank nose pulley and the flywheel are intended as representations only and are drawn to have diameters equal to the crank throw and stroke respectively.

245

HELP Model Generation 3.3 Define Models Once the crank train geometry has been specified, the crankshaft, cylinder block, incylinder and connecting rod model types must be defined by selecting the Define Models button from the Main Panel. The Model Definitions Panel will be opened, as shown in Figure B-25. This panel must be completed before progressing to Edit Cranktrain.

Figure B-25 Model Definitions Panel

3.3.1

Cranktrain

3.3.1.1 Stiffness Model Five model types are available as listed in Table B-5. The choice of model type will affect both the subsequent data entry requirements and the solutions that may subsequently be performed. Model Type Rigid Massless Rigid Compliant Massless Compliant Dynamic

Stiffness Rigid Rigid Included Included Included

Mass None Included for Static Loading None Included for Static Loading Included

Table B-5 Crank Train Model Types Data This is used to specify whether the model data for the cranktrain are to be Loaded from FE file, or User-defined.

246

HELP Model Generation Matrix Formulation This is used to define for a Compliant, Compliant Massless or Dynamic model how the matrices of the crank train model are to be formulated. At present, the matrices are formulated using the Hodgetts method. The mass and stiffness matrices used for a dynamic model of the crank train are derived by Static Condensation of the finite element model. The techniques used to condense the crank train model were first used by Hodgetts 1234 in the development of 3-dimensional crankshaft vibration methods, and subsequently applied to various types of engine by Ricardo 56. The crank train is simplified using a discreet mass-elastic system, idealised to rigid, massive nodes interconnected by flexible, massless elements. The stiffness matrix used for a compliant model is the same as that used for a dynamic model. File Name If the crank train model data are to be Loaded from FE file, File Name must be defined. The field cannot be edited; the user must click on the Browse button to open the panel shown in Figure B-26 Model Transformation and Translation PanelFigure B-26.

3.3.1.2 Thermal Data Specify if thermal loading from FE file should be applied. File Name If thermal crank train data are to be Loaded from FE file, File Name must be defined. The field cannot be edited; the user must click on the Browse button to open the panel shown in Figure B-26.

3.3.2

Model Translation

The Model Transformation and Translation Panel shown in Figure B-26 is used to select and if necessary translate a finite element model file from a specified third-party package to a Ricardo Standard File (SDF) format known as the Standard Finite Element File (SFE) from which ENGDYN reads and writes data for analysis.

247


Figure B-26 Model Transformation and Translation Panel

Origin This is used to specify from which finite-element package the model is being translated. The available options and possible formats are summarised in Table B-6. Origin RICARDO-SFE ABAQUS IDEAS FEMGEN PAFEC MSC-NASTRAN ANSYS

ASCII .SFE .inp .unv N/A N/A .dat .cdb

Binary .SFE .fil N/A .G40 .BS .f12 .rst

Table B-6 Finite Element Model Formats Units This is used to specify the units of the finite element model. The available options and the units of length, force, density and mass for each of the options are summarised in Table B-7. File Units SI MM METRIC FT INCH

Length m mm m ft inch

Force N N Kgf Lbf Lbf

Density Kg/m3 kg/mm3 kg/m3/9.81 lbm/ft3/386.4 lbm/in3/386.4

Table B-7 Finite-Element Model Unit Systems

248

Mass kg kg kg/9.81 lbm/386.4 lbm/386.4

HELP Model Generation If the File Origin is specified as IDEAS or RICARDO-SFE then this option is not required. The unit system is always assumed to be defined in the IDEAS Universal file and the SFE file is always assumed to be in the SI unit system. Input Name This is used to specify the name of the file containing the finite-element model in a format defined by the File Origin Output Name This is used to specify the name of the SFE file containing the finite-element model.

3.3.3

Model Transformation

The tab shown in the second image of Figure B-26 can be used to define any necessary transformation from the co-ordinate system used by the FE model of the cylinder block to that used by ENGDYN. All vectors are in the ENGDYN global axes. The Mirror vector specifies a normal vector to the mirror, the plane of which passes through the origin at the crankshaft co-ordinate system. If zero is entered for the X, Y or Z values of a transformation, then that transformation will not occur. The transformation matrix is built up in the following order; mirror, rotation about Z, then Y and finally about X and then the translation. The FE model must be orthogonal to the ENGDYN co-ordinate system. When Apply or OK is selected, if the File Origin is not RICARDO-SFE then the input file will be translated to an SFE file with the given output name.

3.3.4

Cylinder Block

The second image in Figure B-25 shows the Model Definitions Panel with the Cylinder Block tab selected. It is similar to the Crankshaft tab. Model Five cylinder block model types are available as listed in Table B-8 and described in more detail in 2.5.3. The choice of model type will affect both the subsequent data entry requirements and the solutions that may subsequently be performed. Model Type Rigid Compliant Bearing Compliant Crankcase Compliant Dynamic

Stiffness Rigid Included (User Supplied) Included (From FE Model) Included (From FE Model) Included (From FE Model)

Mass None None None None Included (From FE Model)

Table B-8 Cylinder Block Model Types Data This is used to specify whether the model data for the cylinder block are to be Loaded from FE file or User-defined. The latter can only be selected if the Model is Rigid or

249

HELP Model Generation Compliant Bearing. In this case nodes are created at each main bearing and cylinder dependant on the data defined using the Engine Configuration Panel and Cylinder. Loaded from FE file can only be selected if the Model is Rigid, Compliant Crankcase, Compliant or Dynamic. If the cylinder block is defined as Rigid and no geometry is to be created then this should be set to None. Matrix Formulation This defines for a Dynamic cylinder block model how the mass and stiffness matrices are formulated. These can be generated either by Static Condensation or Component Mode Synthesis techniques using an appropriate finite-element analysis package. These techniques are described in more detail in 2.5.3. If the cylinder block model data are to be Loaded from FE file, File Name must be defined. The field cannot be edited, the user must click on the Browse button to open the Model Transformation and Translation Panel panel, as shown in Figure B-26 and described in sections 3.3.2 and 3.3.3. 3.3.5

In-Cylinder

The second image in Figure B-25 shows the Model Definitions Panel with the In-Cylinder Block tab selected. Model The loads applied to each cylinder and crank pin due to the gas cylinder pressure and the acceleration of the piston and connecting rod may either be derived assuming the mechanism behaves simply as a Slider Crank, or alternatively the loads can be calculated using the Ricardo program PISDYN7 in which the secondary dynamics of the piston are considered. In this case Secondary Dynamic should be selected. Method This is used to select the method by which the piston side loads are applied to the cylinder bore when using loads derived using PISDYN and when Model is selected as Secondary Dynamic. At present only one method, Simplified is available. In this case the hydrodynamic and asperity pressures are integrated to obtain a force and moment at each time step. These forces and moments are added to the crown forces and moments and are applied to the 4 nodes of the cylinder bore as defined in Figure B-109. File This is used to specify the name of the sdf file containing the in-cylinder model data when using loads derived using PISDYN and when Model is selected as Secondary Dynamic. This file will be written either by PISDYN7 or by the Ricardo Program PDTOSF8. The user may click on the Browse button to open a Select Panel. The contents of the sdf file can be viewed using the Ricardo program SDFBROWSER9 and should contain the following data: g:cylinder:bore g:cylinder:crankThrow g:cylinder:length g:cylinder:height g:cylinder:offset g:conrod:mass

g:conrod:length g:conrod:inertia g:piston:mass g:piston:skirtTop g:piston:topRingHeight g:piston:pinOffset

250

HELP Model Generation g:conrod:cgOffset

g:piston:skirtBottom

The values of bore and crank throw as stored above in g:cylinder:bore and g:cylinder:crankThrow are checked when the OK or Apply buttons are selected against those values defined using Engine Configuration Panel shown in Figure B-18. If either of the values is different an error message is displayed.

3.3.6

Connecting Rod

Figure B-27 shows the Model Definitions Panel with the Connecting Rod tab selected.

Figure B-27 Model Definitions Connecting Rod Tab The cylinder number column is filled in automatically and cannot be edited. Define the Connecting Rod model type by positioning the mouse over the centre column and selecting the right mouse button. This will display the pop-up as shown in Figure B-28

Figure B-28 Connecting Rod Model Type Selection Compliant Big End, Compliant Small End and Compliant are applicable to static calculations only. Rigid and Dynamic are applicable to dynamic calculations only.

251

HELP Model Generation Complaint Big End and Compliant Small End are used where a complete model of the connecting rod is not available, and require a cut-plane somewhere along the shank of the rod, as shown in Figure B-29.

Figure B-29 Compliant Small End and Big End Bearing Models Different models can be defined for each cylinder (when there are multiple cylinders). If Apply to All is selected, all connecting rods will be assigned the same type. If Delete Reduced Model is selected, the reduced model associated with the connecting rod will be deleted. To select the SFE file name position the mouse over the SFE name column and press the right mouse button. This will display the pop-up as shown in Figure B-30.

252


Figure B-30 Connecting Rod SFE Name Selection If Select File is selected, the Model Transformation and Translation Panel panel shown in Figure B-26 will pop up. This panel is the same as the one used for the Crankshaft model described in sections 3.3.2 and 3.3.3. When defining the transformation, note that the connecting rod of each cylinder has its own local coordinate system such that the origin is at the centre of the big end bearing, Y is up the connecting rod, and X runs along the bearing axis If Delete File is selected, the SFE Name will be blanked out and the type reset to None. If Delete Reduced Model is selected, the reduced model associated with the connecting rod will be deleted. If Apply to All is selected, all connecting rods will be assigned the same model. 3.4 Editing the Crank Train Model The definition of the individual components of the cranktrain, for example the geometry of the main and pin journal bearings and the assembly of the cranktrain model itself is initiated by selecting the Edit Cranktrain button from the main tool panel. This will open the Cranktrain Tool Panel shown in Figure B-31. This panel must be completed before progressing to Edit Block.

253


Figure B-31 Cranktrain Tool Panel 3.4.1

Editing the Crank Train Components

The area on the left of the Cranktrain Tool Panel shown in Figure B-31 and Figure B-32 lists the individual components of the crank train that can be edited. In this case, the Crankshaft Web is shown highlighted. Other components from the list can be highlighted by selecting them with the mouse. These may be edited in any order, but they all must be completed before progressing to Assemble Model. By clicking on the Select All button, all components of that type will be selected, and will be shown in red in the main drawing area. Alternatively, individual components may be either selected, or deselected using the mouse on the relevant component shown in the main drawing, as shown in Figure B-32. De-selected components will be shown in green. All selected components may be de-selected by clicking the Deselect All button. Where relevant, numbers will be drawn on the canvas to show more information, e.g. webs 1 and 2.

254


Figure B-32 Editing Crank Train Components If the data are non-uniform for a particular part then it is recommended that the user start by selecting all to define the dominant dimensions before selecting components individually with the mouse to make any required changes. Having selected the particular components that you wish to define, they may be edited by clicking on the Edit Selected button. This will open up the relevant panel. In each panel a drawing area will display a drawing of the current component, defining the model parameters. 3.4.1.1 Crankshaft Web The Crankshaft Web Panel is shown in Figure B-33. This panel is used to define one or more of the crankshaft webs. If the cranktrain Data, as defined in the Model Definition panel, has been set to User Defined, the crankshaft web stiffnesses are entered here for each web. Otherwise, if the cranktrain Data has been set to Loaded from FE file the stiffnesses are evaluated by finite element analysis when the Calculate Stiffnesses

255

HELP Model Generation button is selected as described in 3.4.7. The stiffness fields will appear greyed out as shown in Figure B-33. Once the stiffnesses have been calculated by selecting the Calculate Stiffnesses button the stiffness values will appear.

Figure B-33 Crankshaft Web Panel The Thickness of the web and the journal lengths as defined in the figure are calculated from data entered in the Engine Panel and so their values are displayed, but cannot be edited. The following parameters are only required if the Cranktrain Model is Compliant or Dynamic and the Data, as defined in the Model Definition Panel, is User Defined. In plane stiffness This parameter is used to specify the in-plane stiffness of the crankshaft web as defined in Appendix 1 Out of Plane Stiffness This parameter is used to specify the out-of-plane stiffness of the crankshaft web as defined in Appendix 1. Torsion This parameter is used to specify the torsional stiffness of the crankshaft web as defined in Appendix 1.

256

HELP Model Generation The following parameters are used to define the extent of the counterweight at a web. These data are only subsequently used when the crankshaft is balanced by selecting the Balance button as described in 3.4.6. If the chosen method for obtaining primary balance is to simulate the drilling of the counterweights, then these data are used to determine the location of the balance drillings. These data are not used to calculate the mass properties of the counterweights, since they are evaluated either directly from the finite element model of cranktrain when the Calculate Masses button is selected as described in 3.4.5, or entered using the Lumped Mass Panel as described in 3.4.1.13. These parameters are not required if the Cranktrain Model is Rigid Massless or Compliant Massless Counterweight The user can specify the counterweights at each crankshaft web as being Present or Absent. Counterweight Radius This is used to specify the radius from the crankshaft axis to the outer edge of the counterweight. Counterweight Start This is used to specify the angular position corresponding to the start of the counterweight measured clockwise relative to the Y-axis of the ENGDYN Co-ordinate System as defined in 2.2. This angle should lie between 0 and 360°. Counterweight End This parameter is used to specify the angular position corresponding to the end of the counterweight measured clockwise relative to the Y-axis of the ENGDYN Co-ordinate System as defined in 2.2. This angle should lie between 0 and 360°. 3.4.1.2 Big End Journal Element The Element Panel for the crank pin journals is shown in Figure B-34 and is used to define the geometry of the pin journals.

257


Figure B-34 Element Panel for Big End Journals The Outside Diameter of the big end journals has been previously defined using the Engine Panel and so its value is displayed, but may not be edited. Similarly, the Element Length is calculated from data entered at the Engine panel, and may not be edited. Inside Diameter This defines the inner diameter of the journal and should be set to zero if no hole exists. This is used either to calculate the stiffness of the journal element if the Cranktrain Model is defined as Dynamic, Compliant or Compliant Massless, or the section modulus of the journal when calculating crankshaft stresses as described in Chapter 6. If the crank train Data has been set to Loaded from FE file and the selected element is between two adjacent crank big end nodes the stiffness’ of the element are evaluated using the finite element method if the Inside Diameter is non-zero. These are calculated when the Calculate Stiffnesses button is selected as described in 3.4.7. 3.4.1.3 Main Journal Element The Element Panel for the crank main journals is shown in Figure B-35 and is used to define the geometry of the main journals.

258


Figure B-35 Element Panel for Main Journals The Outside Diameter of the main journals has been previously defined using the Engine Panel and so its value is displayed, but may not be edited. Similarly, the Element Length is calculated from data entered at the Engine panel, and may not be edited. Inside Diameter This defines the inner diameter of the journal and should be set to zero if no hole exists. This is used either to calculate the stiffness of the journal element if the Cranktrain Model is defined as Dynamic, Compliant or Compliant Massless, or the section modulus of the journal when calculating crankshaft stresses as described in Chapter 6. If the crank train Data has been set to Loaded from FE file and the selected element is between two adjacent crank pin nodes the stiffness’ of the element are evaluated using the finite element method if the Inside Diameter is non-zero. These are calculated when the Calculate Stiffnesses button is selected as described in 3.4.7.

3.4.1.4 Main Bearing The Main Bearing Panel shown in Figure B-37 is used to define the geometry and bearing model of the crankshaft main journal bearings. When data values change, the 3D drawing area will also change. The contents of this drawing area depend on which tab is currently selected, e.g. when the groove tab is selected it will show the groove, when the oilholes tab is selected it will show the oilholes. The Model may be orientated in the same way as in the main panel drawing area, described in section B.1.3.1.3.

259


Figure B-36 Main Bearing Panel 3.4.1.4.1

Main Bearing Geometry Definition

The bearing axes for the main journal bearings are defined as being identical to the fixed ENGDYN Co-ordinate System described in 2.2. The main bearing geometry and bearing model is defined using the top half of the Main Bearing Panel shown in Figure B-36 Journal Diameter This has been previously defined using the Engine Configuration Panel and so its value is displayed, but may not be edited. Shell Length This defines the effective length of the bearing shell as shown in Figure B-36. The length should exclude the chamfers on the edge of the bearing shell. If the length is not less than the Journal Length corresponding to the same main journal as defined in the Engine Configuration Panel then an error is displayed. Radial Clearance This defines the nominal ‘cold’ radial clearance of the bearing. The ‘hot’ radial clearance of the bearing can be defined using the Distortion tab of the Loading Definition Panel described in 4.3.4. Model Type This is used to define the lubrication solution scheme and has the options Mobility, Hydrodynamic and Elastohydrodynamic. If the cylinder block model as defined using the Model Definitions Panel is neither Compliant, Compliant Crankcase or Dynamic then the latter cannot be selected.

260

HELP Model Generation Feed From This is used to specify whether the oil hole feed to the main bearing is from the Bearing or from an oil gallery in the crankshaft in which case Journal is selected. Changing this option will change the drawing displayed on the panel. 3.4.1.4.2

Main Bearing and Main Journal Groove Definition

The Groove tab of the Main Bearing Panel shown in Figure B-37 is used to define a groove in both the bearing and the journal. Groove type may be either None or Single for the journaI and either None, Single or Mulitple for the bearing. Multiple grooves are only relevant for HD/EHD bearings. If set to Single then a groove may be entered in the table.

Figure B-37 Main Bearing Groove Tab Width This defines the effective width of the bearing groove. The width should include the chamfers of the groove Start of Groove This defines the angular position corresponding to the start of the bearing groove measured with respect to the bearing axes when the crankshaft is at 0CA. This angle should lie between 0 and 360°. End of Groove This defines the angular position corresponding to the end of the bearing groove measured with respect to the bearing axes when the crankshaft is at 0 CA. This angle should lie between 0 and 360°. Offset This defines the axial offset of the groove from the bearing centre Angle This is used to define the grooves with an angle.

261


Figure B-38 Bearing Groove with an Angle

3.4.1.4.3 Main Bearing and Main Journal Oilhole Definition If the main bearing is a plain bearing then the journal or bearing oil feed holes are defined using the Feed From and Oil Hole Type option menus and using the table which is part of the Oilholes tab shown in Figure B-36. The oil holes are used in the solution to calculate the oil mass flow rate for each bearing. The Oilholes tab of the Main Bearing Panel shown in Figure B-39 is used to define the oilholes.

Figure B-39 Main Bearing Panel Oilholes Tab

Oil Hole Type This is used to specify whether the oil feed to the bearing is through a Single (Leading) or Single (Trailing) drilling or through Twin, or Multiple drillings. Changing this option will change the drawing displayed on the panel. Multiple drillings can only be defined if the oil is fed from the bearing. The diameter and position of each oil hole are defined using the table. The diameter should include the chamfer of the hole. The position of the breakout of each oil hole may be defined by its angular position or its height with respect to the bearing axes. The angular position  is measured clockwise from the y-axis and should lie between 0 and 360, whilst the height is the distance in y and is equal to r cos  where r is the journal radius. Journal oil hole breakout positions are defined with respect to the bearing axes when the crankshaft is at 0CA. If the bearing is fed with a Single (Leading) oil hole then the angle should lie between 0 and 180°. Conversely, if the oil hole is trailing then the

262

HELP Model Generation angle should be between 180 and 360. In the case of the bearing having Multiple oil holes additional oil holes are entered in the. Rows are added and deleted using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and Delete Row options. 3.4.1.4.4 Main Bearing Lubrication Mesh Definition The Mesh tab of the Main Bearing Panel shown in Figure B-40 is used to define the lubrication mesh when the Model Type set is to Hydrodynamic or Elastohydrodynamic.

Figure B-40 Main Bearing Panel Mesh Tab The topology of the mesh depends on the presence of oil feed holes/grooves. Provisions have been made within ENGDYN to internally adjust the hydrodynamic mesh such that the oil hole/groove boundaries are clearly demarcated and this procedure may lead to a non-uniform mesh. Size of the mesh is particularly crucial to the execution of the program from two aspects: (a) CPU time, a larger mesh size implies higher run time; and, (b) non-convergence behaviour (sometimes a larger mesh typically cures the problem). As a result, an optimum mesh size must be selected to address both issues, 11 by 61 has been found to give good a good compromise. No. of Axial Nodes This is used to specify the number of nodes along the axis of the bearing. If an insufficient number of nodes are entered to describe the mesh then this number will be changed automatically by ENGDYN. No. of Circumferential Nodes This is used to specify the number of nodes around the circumference of bearing. If an insufficient number of nodes are entered to describe the mesh then this number will be changed automatically by ENGDYN.

3.4.1.4.5 Main Bearing and Main Journal Material Definition The Material tab of the Main Bearing Panel shown in Figure B-41 is used to define the bearing and journal materials when the Model Type is set Hydrodynamic or Elastohydrodynamic.

263


Figure B-41 Main Bearing Panel Material Tab Bearing Material Selecting Define will display the Material Properties Panel shown in Figure B-42.

Figure B-42 Bearing Material Definition Material Name This is used to define the name of the bearing material. Pressing Select will display the Material Properties Panel shown in Figure B-43 containing the currently available materials. The asperity data can be calculated from surface profile measurements using the MATUTIL program supplied with the Ricardo Software installation which is described in Appendix 12.

Figure B-43 Bearing Material Properties

264

HELP Model Generation Selecting one of the rows and pressing OK will cause the values to be entered into the panel. Young’s Modulus This is used to define the Young’s modulus of the material. Hardness This is used to define the Brinnell Hardness number of the material. Asperity Density This is used to define the number of asperities per unit area of the material. Asperity RMS Height This is used to define the Root Mean Square of asperity heights for the material. Asperity Mean Summit Height This is used to define the Mean Summit Height of asperity heights for the material. Asperity Standard Deviation Summit Height This is used to define the Standard Deviation of asperity peaks for the material. Asperity Radius of Curvature This is used to define the tip radius of asperities for the material. Asperity Ratio The ratio represents the surface conductivity to oil flow in a particular direction.

Journal Material Selecting Define will display the Material Properties Panel shown in Figure B-42 to allow the journal material to be defined. The panel is used the same manner as for the bearing material. Wear Coefficient This value is not easily obtained and is a function lubrication conditions, lubricity and other factors. However, users need not be concerned about the validity of the wear coefficient value, unless that is the focus of their study. Wear coefficient values do not impact the simulation and are used only for post-processing. A table of wear coefficients for adhesive wear is provided in Table B-910.

Lubrication None Poor Good Excellent

Identical 1500 x 10-6 300 x 10-6 30 x 10-6 1 x 10-6

Metal-on-metal Soluble Intermediate 500 x 10-6 100 x 10-6 100 x 10-6 20 x 10-6 -6 10 x 10 2 x 10-6 -6 0.3 x 10 0.1 x 10-6

Insoluble 15 x 10-6 3 x 10-6 0.3 x 10-6 0.03 x 10-6

Table B-9 Example Wear Coefficients

265

Metal-on non-metal 3 x 10-6 1.5 x 10-6 1 x 10-6 0.5 x 10-6

HELP Model Generation Friction Coefficient This value is the friction coefficient for the contacting surfaces.

3.4.1.4.6 Main Bearing Stiffness Definition If the cylinder block Model type is set to Compliant bearing in the Model Definitions Panel then the bearing stiffness’ must be entered for each bearing using the Stiffness tab shown in Figure B-44

Figure B-44 Main Bearing Stiffness Definition Y This is used to specify the stiffness K y in the Y direction in the ENGDYN co-ordinate system. Z This is used to specify the stiffness K z in the Z direction in the ENGDYN co-ordinate system. Phi Y This is used to specify the stiffness K y in the Y direction in the ENGDYN co-ordinate system. Phi Z This is used to specify the stiffness K z in the Z direction in the ENGDYN co-ordinate system

3.4.1.4.7 Main Bearing and Main Journal Profile Definition The Profiles tab of the Main Bearing Panel shown in Figure B-40 is used to define the cold manufactured profiles of the journal and bearing when the Model Type is set Hydrodynamic or Elastohydrodynamic. The hydrodynamic and boundary lubrication models are dependent on clearance values. The clearance profile between the bearing and journal surfaces determines the oil film and contact pressures, which are generated in this region. This profile is based on the nodal clearance values. Any change in the clearance profile due to thermal loading can be defined using the Distortion tab of the Loading Definition Panel described in 4.3.4.

266


Figure B-45 Main Bearing Profile Tab This tab has two identical tabs, Bearing and Journal, which are used to define the profile of the bearing and journal. When the Profile tab is active, it will also display the Contour Panel shown in Figure B-46.

Figure B-46 Contour Panel This is used to control the display of the contours on the lubrication mesh on the 3D panel. The option menu specifies the Value that is displayed. The options are Off, Journal Profile, Bearing Profile and Clearance. Type This is used to define the type of profile. The options are Cylindrical, Circular and NonCircular. If the journal is Cylindrical then is assumed to be a cylinder with a diameter equal to the Nominal Diameter. If the bearing is Cylindrical then is assumed to be a cylinder with a diameter equal to the Nominal Diameter plus twice the Radial Clearance. If Type is Circular or Non-Circular then the journal or bearing profile is defined by a number of slices along the length of the bearing. If only one slice is defined, it is assumed to be the same along the whole length of the bearing or journal. If Type is Circular then the table becomes active and has the appearance shown in Figure B-45. The table is used to define any number of concentric, circular slices of different diameters along the length of the bearing or journal. Rows are added and deleted using the right mouse button with the cursor positioned over the row. This displays a Row Operations Pop-Up menu with Insert Before, Insert after and Delete. An example of a Circular profile would be a barrelled-shaped journal as shown in Figure B-47.

267


Figure B-47 Barrel-Shaped Journal

If Type is Non-Circular then the table has the appearance shown in Figure B-48.

Figure B-48 Main Bearing Profile Tab In this case the table is used to define any number of non-circular slices along the length of the bearing or journal. Profiles are defined or edited using the right hand mouse with the cursor positioned over an appropriate cell in the Profile Number column. This will display a pop-up menu with Edit and Define options and any existing profile numbers. It is therefore possible to Define a new profile, Edit an existing profile, or simply select an existing profile. If Edit or Define are selected then Edit Profile Panel shown in Figure B-49 is displayed.

268


Figure B-49 Edit Profile Panel This panel allows a number of different profile types to be defined or for a profile to be read from a file. The profiles are defined with respect to the bearing axes when the crankshaft is at 0CA. Number This shows the current profile number and is only editable if defining a new profile. Type This is used to specify the profile type. The options are Circular, Elliptical, Sinusoidal, Equation and From File. The availability of certain fields depends on the Type. Diameter This is used to define the diameter of the profile. Y Offset This is used to define an offset of the profile centre in the Y direction. Z Offset This is used to define an offset of the profile centre in the Z direction. Angular Offset This is used to define the angular offset (i.e. rotation) of the profile about the centre. Radial Deviation This is used to define the maximum radial deviation from the Diameter. Number of Lobes This is used to define how many lobes the profile has and can only be edited if the profile Type is sinusoidal. Equation This is used to define an equation as a function of D, R, N, THETA and PHI where these variables correspond to Diameter, Radial Deviation, Number of Lobes, Angle and

269

HELP Model Generation Angular Offset. The expression must have consistent units where the resultant has a unit of length. File Name This is used to define the name of the ASCII file from which the profile is read. The name may be entered directly or by using the File Panel that is displayed by pressing Browse. Each file contains either a single profile or multiple profiles of radii or radial deviations varying with angle, in which the data is defined in block or column free format. The data do not necessarily have to be at equally spaced angle intervals. The data in the file must be either in blocks or columns, not both, with each block labelled using the names ‘ANGLE’ and ‘RADIUS’ or ‘RADIAL DEVIATION’ as appropriate. Each block or column must be defined with an appropriate unit type. An example of a file in which a single profile is defined using block format is as follows: BLOCK

NUMBER=1 NAME='ANGLE' UNITS='deg'

0.00000 80.0000 160.000 240.000 320.000 BLOCK

10.0000 90.0000 170.000 250.000 330.000

20.0000 100.000 180.000 260.000 340.000

30.0000 110.000 190.000 270.000 350.000

40.0000 120.000 200.000 280.000

50.0000 130.000 210.000 290.000

60.0000 140.000 220.000 300.000

70.0000 150.000 230.000 310.000

NUMBER=2 NAME='RADIAL DEVIATION' UNITS='micron'

1.01086 0.85086 1.04215 0.45371 0.49962

1.00000 0.84090 1.03976 0.45462 0.50857

0.99790 0.87401 1.02032 0.48842 0.52708

0.88764 0.95214 0.98457 0.49955 0.62381

0.83746 0.98992 0.90603 0.49111

0.82952 1.02439 0.78639 0.49418

0.82177 1.03059 0.60423 0.49719

0.85677 1.04399 0.50758 0.51053

The same data in column format would be as follows: COLUMN COLUMN 0.00000 10.0000 20.0000 30.0000 40.0000 50.0000 60.0000 . . 340.000 350.000

NUMBER=1 NAME='CRANK ANGLE' UNITS='deg' NUMBER=2 NAME='RADIAL DEVIATION' UNITS='micron' 1.01086 1.00000 0.99790 0.88764 0.83746 0.82952 . . . 0.52708 0.62381

If multiple profiles are defined in the same file then each block or column is differentiated using LEGEND = ‘PROFILE 1’. For example, using column format the data would be of the form: COLUMN COLUMN COLUMN COLUMN COLUMN

NUMBER=1 NUMBER=2 NUMBER=3 NUMBER=4 NUMBER=5

0.00000 10.0000 20.0000 30.0000 40.0000

1.8583 2.5757 2.1159 2.1354 1.4201

NAME='ANGLE' NAME='RADIAL NAME='RADIAL NAME='RADIAL NAME='RADIAL 0.6543 0.6783 0.2000 0.2500 0.4500

UNITS='deg' DEVIATION' UNITS='micron' DEVIATION' UNITS='micron' DEVIATION' UNITS='micron' DEVIATION' UNITS='micron'

0.1000 0.6200 0.5300 0.3456 .

0.5689 0.4367 0.9876 . .

270

LEGEND=’PROFILE LEGEND=’PROFILE LEGEND=’PROFILE LEGEND=’PROFILE

1’ 2’ 3’ 4’

HELP Model Generation 50.0000 1.4879 60.0000. .

. .

. .

. .

Deformation Factor This is only used to exaggerate the visual difference between the profile and the nominal bearing or journal in the plotting area and has no effect on the generated profile. It is possible to zoom in and out in the plotting area by pressing the Ctrl key and middle mouse key together whilst simultaneously moving the mouse up and down. The result of changing the profiles is shown in the 3D panel, Figure B-50.

Figure B-50 Profile The Contour panel shown in Figure B-51 is used to control the display of the contours on the lubrication mesh in the 3D panel. The option menu specifies the Value that is displayed. The options are Off, Journal Profile, Bearing Profile and Clearance.

Figure B-51 Contour

3.4.1.5 Big End Bearing

271

HELP Model Generation The Big End bearing panel behaves in exactly the same way as for the main bearings described in section 3.2.3. The only differences are: 1) There’s no Stiffness tab. 2) There’s an extra Feed From option, from Mains, which is used to specify that the oil hole feed to the big end bearing is from an adjacent main bearing. 3.4.1.5.1

Big End Bearing Geometry Definition

This is defined in the same way as for the main bearings as described in 3.4.1.4.1. The bearing axes for the big end journal bearings are defined with respect to the connecting rod as shown in Figure B-52.

Figure B-52 Big End Bearing Axes 3.4.1.5.2 Big End Bearing and Big End Journal Groove Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.2 3.4.1.5.3 Big End Bearing and Big End Journal Oilhole Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.3. The diameter and position of each oil hole are defined using the table. The diameter should include the chamfer of the hole.

272


Figure B-53 Big End Bearing Oil hole Position The position of the breakout of each oil hole may be defined by its angular position as shown in Figure B-53 or its height as shown on the panel. These dimensions are with respect to the bearing axes (as defined in Figure B-52) when the connecting rod and crank pin are at their TDC Position. The angular position, , is measured clockwise from the y-axis and should lie between 0 and 360, whilst the height is the distance in y and is equal to r cos  where r is the journal radius. If the bearing is fed with a Single (Leading) oil hole then the angle should lie between 0 and 180°. Conversely, if the oil hole is trailing then the angle should be between 180 and 360. If the bearing is fed with a Cross drilling then it is only necessary to define one of the oil hole breakouts since it is assumed that the hole breakouts are diametrically opposed. 3.4.1.5.4 Big End Bearing Lubrication Mesh Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.4. 3.4.1.5.5 Big End Bearing and Big End Journal Material Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.5. 3.4.1.5.6 Big End Bearing and Big End Journal Profile Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.7. The profiles are defined with respect to the bearing axes (as defined in Figure B-52) when the connecting rod and crank pin are at their TDC Position.

273

HELP Model Generation 3.4.1.6 Small End Bearing The Small End bearing panel behaves in a similar manner to the Main bearings panel described in section 3.2.3. The only difference is that there is no Stiffness tab. 3.4.1.6.1 Small End Bearing Geometry Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.1. The bearing axes for the small end journal bearings are defined with respect to the connecting rod as shown in Figure B-52.

Figure B-54 Small End Bearing Axes 3.4.1.6.2

Small End Bearing and Small End Journal Groove Definition

This is defined in the same way as for the main bearings as described in 3.4.1.4.2 3.4.1.6.3 Small End Bearing and Small End Journal Oilhole Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.3. The diameter and position of each oil hole are defined using the table. The diameter should include the chamfer of the hole.

274


Figure B-55 Small End Bearing Oil hole Position The position of the breakout of each oil hole may be defined by its angular position as shown in Figure B-53 or its height as shown on the panel. These dimensions are with respect to the bearing axes (as defined in Figure B-52) when the connecting rod and crank pin are at their TDC Position. The angular position, , is measured clockwise from the y-axis and should lie between 0 and 360, whilst the height is the distance in y and is equal to r cos  where r is the journal radius. If the bearing is fed with a Single (Leading) oil hole then the angle should lie between 0 and 180°. Conversely, if the oil hole is trailing then the angle should be between 180 and 360. If the bearing is fed with a Cross drilling then it is only necessary to define one of the oil hole breakouts since it is assumed that the hole breakouts are diametrically opposed. 3.4.1.6.4 Small End Bearing Lubrication Mesh Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.4. 3.4.1.6.5 Small End Bearing and Small End Journal Material Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.5. 3.4.1.6.6 Small End Bearing and Small End Journal Profile Definition This is defined in the same way as for the main bearings as described in 3.4.1.4.7. The profiles are defined with respect to the bearing axes (as defined in Figure B-52) when the connecting rod and small end bearing are at their TDC Position.

275

HELP Model Generation 3.4.1.7 Thrust Bearing The Thrust Bearing Panel is shown in Figure B-56 and is used to specify the data relating to the thrust bearing model. The available Model options are Linear Stiffness,. Lash Stiffness and Squeeze. This data is only required if the Crankshaft as defined in the Model Definitions Panel is Compliant, Compliant (Massless) or Dynamic. If Linear Stiffness is selected then the thrust bearing oilfilm is represented by a linear stiffness and viscous damper, as defined by the data shown below.

Figure B-56 Linear Stiffness Thrust Bearing Model Bearing Model This is the linear stiffness of the thrust bearing oilfilm Housing This is the axial stiffness of the bearing housing and is only required as input if the cylinder block model as defined in the Model Definitions Panel is Compliant Bearing, and should have a value greater than 0.0. Otherwise, if the cylinder block Model is defined as Compliant, Compliant Crankcase or Dynamic the stiffness will be evaluated from the supplied stiffness matrix of the cylinder block. This field will therefore be ghosted. Once the stiffness has been calculated by selecting Edit Block as described in 3.5 the stiffness value will appear. Damping The linear viscous damper may be defined as a Coefficient or as a Ratio. If Ratio is selected then during solution the program will derive an equivalent damping coefficient based on the mass and stiffness of a single-mass elastic system as defined below. Ratio This is the ratio of critical damping to derive an equivalent viscous damping coefficient c based on the equation

Ratio 

c c  cc 2m.k 0.5

for a single-mass elastic system where m is the total mass of the crankshaft and k is the stiffness of the thrust bearing oilfilm as defined above by Bearing Model.

276

HELP Model Generation Coefficient This is the viscous damping coeffient of the thrust bearing oilfilm if Damping is set to Coefficient If Lash Stiffness is selected then the thrust bearing oilfilm is represented by a linear stiffness and viscous damper with an axial clearance. The data is the same as for the Linear Stiffness but with an additional input Clearance as shown below.

Figure B-57 Lash Stiffness Thrust Bearing Model Clearance This is the axial clearance of the thrust bearing. If Squeeze is selected then the thrust bearing is modeled using an analytical solution of the Reynold’s equation for rotary symmetry with pure squeeze in the axial direction. The hydrodynamic force is given by

F

where

D 32 

o

3h ht  lnDD/DD   D 2

2

 Di

2

2

o

3



i

o

o

Di

= inner diameter of the bearing

Do

= outer diameter of the bearing

h

= clearance



= dynamic viscosity of the lubricant

i

2

2   Di  



The data for this model is defined by the data of the Geometry tab as shown below.

277


Figure B-58 Thrust Bearing Squeeze Model Clearance This is the axial clearance of the thrust bearing. Inner Diameter This is the inner diameter of the thrust bearing and will be greater than the diameter of the adjacent main bearing. Outer Diameter This is the outer diameter of the thrust bearing As with the Linear Stiffness and Lash Stiffness models the Housing stiffness, defining the axial stiffness of the bearing housing may also be required as input for this model if the cylinder block model as defined in the Model Definitions Panel is Compliant Bearing.

3.4.1.8 Cylinder The Cylinder Panel is shown in Figure B-59 and is used to specify the data for the cylinder.

Figure B-59 Cylinder Panel If the Cylinder Model as defined using the Model Definitions Panel is set to Secondary Dynamic then all the data of this panel (except the value for Head Height) is read from the sdf file supplied using the Model Definitions Panel.

278

HELP Model Generation Head Height This is the distance head from the crankshaft centre to the centre of the cylinder head. If the cylinder head has a flat roof chamber then this height will be the same as Cylinder Height. If the head has a pent-roof chamber then this height will be greater than Cylinder Height. This is only required if the cylinder block Data, as defined on the Model Definitions Panel, is User Defined. Otherwise if Data is Loaded from FE file the height is calculated from the FE model when the cylinder block model is assembled and is then displayed here. Cylinder Height This is the distance from the crankshaft centre to the distance of the top of the cylinder bore. This value can never be edited and is only ever read from the .sdf file supplied using the Model Definitions Panel if the In-Cylinder Model as defined using that panel is Secondary Dynamic. The variable g:cylinder:height is read from the sdf file. Length This is the height of the cylinder bore. This value can never be edited and is only ever read from the .sdf file supplied using the Model Definitions Panel if the In-Cylinder Model as defined using that panel is Secondary Dynamic. The variable g:cylinder:length is read from the sdf file.

3.4.1.9 Connecting Rod The Conrod Panel is shown in Figure B-60 and is used to specify the data for the conrod.

Figure B-60 Conrod Panel Inertia This is the connecting-rod polar moment of inertia about its centre of gravity and will be the inertia of the complete rod assembly including bolts and shells. The value for g:conrod:inertia is used if the data is read from the sdf file.

279

HELP Model Generation Length This is the distance between the centres of the connecting rod and is defined using the Engine Configuration Panel and is displayed on this panel for information only. The value for g:conrod:length is used if the data is read from the sdf file. Bearing Offset This is the axial offset of the big-end bearing centre from the cylinder centre and is typically equal to zero. It is defined using the Engine Configuration Panel and is displayed on this panel for information only. The offset is measured from the cylinder centre axis. A positive value denotes that the bearing is to the rear of the cylinder, whilst a negative value denotes that the bearing is in front of the cylinder. If an offset is specified then there will be moment with respect to the big end bearing centre due to the offset cylinder forces. For reasons of packaging Vee-configured engines often have a small offset. For one bank the big end bearing offset will be to the rear of the cylinder centre (+ve), whilst for the other bank the offset will be in front of the cylinder (-ve). Mass This is the total mass of the connecting rod including bolts and bearing shells. The value for g:conrod:mass is used if the data is read from the sdf file. Centre of Gravity This is the distance of the centre of gravity from the big-end centre. The z-co-ordinate of g:conrod:cgOffset is used if the data is read from the sdf file. Recip. Mass Ratio For all dynamic solutions, where the loads due to the connecting rod and piston are applied to the crank-pin, it is necessary to add the effective mass due to the rotating and reciprocating masses of the connecting rod and piston to the crankpin node. The effective mass is given by

M effective  M rot  f .M recip where f is a ratio typically equal to 0.5 and is ratio specified here. This is only required as an input if the crankshaft model is dynamic. 3.4.1.10

Piston

280


Figure B-61 Piston Panel Mass This is the mass of the piston assembly. The value for g:piston:mass is used if the data is read from the sdf file. Offset This is the piston pin offset and is positive if the pin centre is on the anti-thrust side. It is defined using the Engine Configuration Panel and is displayed on this panel for information only. The value for g:piston:pinOffset is used if the data is read from the sdf file. The parameters Top of Skirt, Bottom of Skirt and Top Ring are only used to apply loads to the cylinder bore when performing a quasi-static cylinder block analysis using the direct method as described in 7.3.4. If the user is not intending to perform this type of analysis then these parameters can all be set to zero. If any of these parameters are non-zero then they should all be positive. Top of Skirt and Bottom of Skirt are used to define the extent of the pressure distribution when applying the piston side load. Top Ring is used to define the position above which the cylinder gas pressure is applied to the cylinder bore. Top of Skirt This is used to specify the distance of the top of the piston skirt from the piston pin centre and must be a positive value. The value for g:piston:skirtTop is used if the data is read from the sdf file. Bottom of Skirt This is used to specify the distance of the bottom of the piston skirt from the piston pin centre and must be positive value. The value for g:piston:skirtBottom is used if the data is read from the sdf file. Top Ring This is used to specify the distance of the top face of the top ring from the piston pin centre and must be greater than the Top of Skirt value. The value for g:piston:topRingHeight is used if the data is read from the sdf file.

281

HELP Model Generation Recip. Mass Ratio For all dynamic solutions, where the loads due to the connecting rod and piston are applied to the crank-pin, it is necessary to add the effective mass due to the rotating and reciprocating masses of the connecting rod and piston to the crankpin node. The effective mass is given by

M effective  M rot  f .M recip where f is a ratio typically equal to 0.5 and is ratio specified here. This is only required as an input if the crankshaft model is dynamic. If the connecting rod model for the selected piston is Rigid or Dynamic as defined by the Model Definitions Panel then the panel will have some additional inputs for the piston rings related to the piston friction model as shown in the figure below

Figure B-62 Piston Panel with Ring Data Up to 3 piston rings can be included, typically 2 compression rings and an oil contol ring. For each piston ring the following data is required

282

Cylinder Centre

Axial Thickness


Radial Thickness

Figure B-63 Piston Ring Geometry

Radial Thickness This the width of the piston ring Axial Thickness This is the height of the piston ring Tip Load This is the tangential load required to close the ring gap Contact Pressure This is the contact pressure given by

P

where

2.Ft b.h

Ft

= tangential load required to close the ring gap

b

= cylinder bore

h

= axial thickness of the ring

3.4.1.11 Cranknose Assembly The Cranknose Assembly Panel is shown in Figure B-64 and is used to specify data relating to the crankshaft at the crank nose. No data entry is required for this panel if the crank train Model, as defined in the Model Definitions Panel, is Rigid Massless.

283


Figure B-64 Cranknose Assembly Panel

3.4.1.11.1 Crankshaft If the crank train Data, as defined in the Model Definitions Panel, has been set to User Defined, the stiffness values of the crank nose assembly are entered here. Otherwise, if the crank train Data has been set to Loaded from FE file the stiffness values are evaluated by finite element analysis when the Calculate Stiffnesses button is selected as described in 3.4.7. The stiffness fields will appear ghosted as shown in Figure B-64. Once the stiffness values have been calculated by selecting the Calculate Stiffnesses button the stiffness values will appear. The drawing at the top right of the panel shows, shaded in blue, the part of the crank currently assumed to be included in the FE model of the crankshaft. This drawing is updated depending on what option is selected for Hub Data. Element Length This is used to specify the length of the element that defines the crank nose assembly and effectively defines the position of a node relative to the node at the adjacent main journal bearing as shown in Figure B-64. The position of this node should be either at the centre of the vibration damper if fitted or alternatively if no vibration damper is fitted at a position at the centre of a pulley at which a belt, chain or gear is driven. Initially, the length defaults so that the node is positioned half a main journal length from the adjacent bearing. If the crank train Data, as defined in the Model Definition panel, has been set to Loaded from FE file, then the position of the node that is defined by this length must correspond to a cut-plane in the yz-plane in the finite-element model of the crank train. Otherwise, once the Calculate Stiffnesses button is selected, an error will be reported when the stiffness values are being calculated. If the crank train Data has been set to User Defined, it is important that this length is equivalent to the crank nose stiffness’ that

284

HELP Model Generation are entered using this panel. If the crank train Model is Rigid then the length of this element is not critical. Hub Data If the crank train Model, as defined in the Model Definitions Panel, is Rigid, Compliant or Dynamic then this is used to specify how the mass properties at the crank nose are to be derived. If the Hub Data is Included in FE model it is assumed that the finite element model contains the definition of all the mass at the crank nose. In this case the mass properties are evaluated from the finite-element model of the crank train when the Calculate Masses button is selected as described in 3.4.5. If the hub data is User Defined, it is assumed that there are some additional masses that require definition. In this case the mass properties of the additional mass may be defined by selecting the Edit button. This will open the Lumped Mass Configuration panel, which is described in 3.4.1.13. If the hub data is None then there are no additional masses to define. Changing this option will change the drawing displayed on the panel. The following parameters are only required if the crank train Model, as defined in the Model Definitions Panel, is Compliant or Dynamic and that Data is User Defined. Bending Stiffness This is used to specify the bending stiffness of the element defining the crank nose assembly as defined in Appendix 1. Torsional Stiffness This is used to specify the torsional stiffness of the element defining the crank nose assembly as defined in Appendix 1. Axial Stiffness This is used to specify the axial stiffness of the element defining the crank nose assembly as defined in Appendix 1. 3.4.1.12 Flywheel Assembly The Flywheel Assembly Panel is shown in Figure B-65 and is used to specify data for the flywheel and the adjacent part of the crankshaft. No data entry is required for this panel if the crank train Model, as defined in the Model Definitions Panel, is Rigid Massless If the crank train Data, as defined in the Model Definition panel, has been set to User Defined, the stiffness values of the flywheel assembly are entered here. Otherwise, if the crank train Data has been set to Loaded from FE file the stiffness values are evaluated by finite element analysis when the Calculate Stiffnesses button is selected as described in 3.4.7. The stiffness fields will appear ghosted as shown in Figure B-64. Once the stiffness’ have been calculated by selecting the Calculate Stiffnesses button the stiffness values will appear. The drawing at the top right of the panel shows, shaded in blue, the part of the crank currently assumed to be included in the FE model of the crankshaft. This drawing is updated depending on what option is selected for Type and Primary Mass.

285


Figure B-65 Flywheel Assembly Panel Type This is used to specify the type of flywheel. The options are Conventional, Twin Mass, or Torque Converter. The selection of Type determines what data are entered in the lower part of the panel. If Conventional is selected, then Primary and Clutch data are required. If Twin Mass is selected, then Primary, Secondary, Clutch and Flexible Coupling data are required. Finally, if Torque Converter is selected then Primary data are required. It is intended currently that the Twin Mass option is used either to model a Dual Mass Flywheel (DMF) in which both flywheels are assumed rigid, connected by stiffness’ or to model the flexibility of a conventional flywheel. A method for evaluating the equivalent mass and stiffness data to model the latter is described in Appendix 2. Primary Flywheel The primary flywheel may either be Included in FE model or Excluded from FE model. If the primary flywheel is Included in FE model, then no further mass property data for the primary flywheel are required and the table of data titled Primary will be ghosted. Changing this option will change the drawing displayed on the panel. Clutch This is used to specify to which flywheel the clutch is attached when a Twin-Mass flywheel is being modelled.

286

HELP Model Generation X1 This is used to specify the length of the element that defines the flywheel assembly and effectively defines the position of a node relative to the node at the adjacent main journal bearing as shown in Figure B-64. If the Primary Mass is Included in FE model as defined below the position of the node should be at the face at which the clutch plate is attached. Conversely, if the Primary Mass is Excluded from FE model the position should be at the face at which the flywheel is attached. Initially, the length defaults so that the node is positioned half a main journal length from the adjacent bearing. If the crank train Data, as defined in the Model Definitions Panel, has been set to Loaded from FE file, then the position of the node that is defined by this length must correspond to a cut-plane in the yz-plane in the finite-element model of the crank train. Otherwise, once the Calculate Stiffnesses button is selected, an error will be reported when the stiffness’ are being calculated. If the crank train Data has been set to User Defined, it is important that this length is equivalent to the flywheel stiffness’ that are entered using this panel. If the crank train Model is Rigid then the length of this element is not critical. Bearing This is used to specify the location of a roller element bearing if one exists as part of flywheel assembly. The bearing is modelled using linear stiffness’ and damping coefficients, which are defined using the Bearing tab. The bearing can either exist At Primary Mass or At Secondary Mass. If either of these options is selected then the node representing the selected mass should be located at the bearing centre using either X1 or X2 as appropriate. If no bearing exists then None should be selected. X2 This is used to specify the position of the node representing the secondary flywheel when modelling a Twin-Mass flywheel. X2 is the axial distance relative to the node defined by X1. The following parameters are only required if the crank train Model, as defined using the Model Definitions Panel, is Compliant or Dynamic and that Data, is defined as User Defined. Bending Stiffness This is used to specify the bending stiffness of the element defining the flywheel assembly as defined in Appendix 1. Torsional Stiffness This is used to specify the torsional stiffness of the element defining the flywheel assembly as defined in Appendix 1. Axial Stiffness This is used to specify the axial stiffness of the element defining the flywheel assembly as defined in Appendix 1. A combination of the following mass property data is required for the Primary, Secondary and Clutch masses, dependant on the flywheel Type and whether the Primary Mass is Included in FE model. If the clutch is included then the mass property data of the clutch plate assembly should be included.

287


Figure B-66 Flywheel Primary Mass Tab Offset This is used to specify the distance along the crankshaft X-axis of the centre of gravity of the mass from the node defined by the Element Length. Mass This is used to specify the mass. Inertia XX This is used to specify the inertia about X at the centre of gravity of the mass in the ENGDYN co-ordinate System. Inertia YY This is used to specify the inertia about Y at the centre of gravity of the mass in the ENGDYN co-ordinate System. Inertia ZZ This is used to specify the inertia about Z at the centre of gravity of the mass in the ENGDYN co-ordinate system. The following stiffness and damping data is required to define the flexible coupling when modelling a Twin-Mass flywheel and is entered using the Flexible Coupling tab as shown in Figure B-67. Each of the stiffness values must be  0 whilst each of the damping values must be  0 .

Figure B-67 Flywheel Assembly Flexible Coupling Tab

288

HELP Model Generation Kx This is used to specify the stiffness K x in the X direction in the ENGDYN co-ordinate system. Ky This is used to specify the stiffness K y in the Y direction in the ENGDYN co-ordinate system. Kz This is used to specify the stiffness K z in the Z direction in the ENGDYN co-ordinate system. Kphix This is used to specify the stiffness K x in the X direction in the ENGDYN co-ordinate system. Kphiy This is used to specify the stiffness K y in the Y direction in the ENGDYN co-ordinate system. Kphiz This is used to specify the stiffness K z in the Z direction in the ENGDYN co-ordinate system. Cx This is used to specify the damping coefficient Cx in the X direction in the ENGDYN coordinate system. Cy This is used to specify the damping coefficient Cy in the Y direction in the ENGDYN coordinate system. Cz This is used to specify the damping coefficient Cz in the Z direction in the ENGDYN coordinate system. Cphix This is used to specify the damping coefficient Cx in the X direction in the ENGDYN coordinate system. Cphiy This is used to specify the damping coefficient Cy in the Y direction in the ENGDYN coordinate system. Cphiz This is used to specify the damping coefficient Cz in the Z direction in the ENGDYN coordinate system.

289

HELP Model Generation The following data is required to define the bearing and is entered using the Linear Bearing Definition Tab as shown in Figure B-68. This data is only required if Bearing is At Primary Mass or At Secondary Mass. Each of the stiffness values must be  0 whilst each of the damping values must be  0 .

Figure B-68 Linear Bearing Definition Tab The parameters Length and Diameter are defined in the Linear Bearing Definition Tab and are only used if the cylinder block Data, as defined on the Model Definitions Panel, is Loaded from FE file or User-Defined. The following coefficients define the stiffness and damping of the bearing unless the cylinder block Model, as defined using the Model Definitions Panel, is Rigid in which case these coefficients represent the net stiffness and damping of the bearing and its housing. Ky This is used to specify the stiffness K y in the Y direction in the ENGDYN co-ordinate system. Kz This is used to specify the stiffness K z in the Z direction in the ENGDYN co-ordinate system. Kphiy This is used to specify the stiffness K y in the Y direction in the ENGDYN co-ordinate system. Kphiz This is used to specify the stiffness K z in the Z direction in the ENGDYN co-ordinate system. Cy This is used to specify the damping coefficient Cy in the Y direction in the ENGDYN coordinate system. Cz This is used to specify the damping coefficient Cz in the Z direction in the ENGDYN coordinate system.

290

HELP Model Generation Cphiy This is used to specify the damping coefficient Cy in the Y direction in the ENGDYN coordinate system. Cphiz This is used to specify the damping coefficient Cz in the Z direction in the ENGDYN coordinate system.

3.4.1.13

Lumped Mass

The Lumped Mass Panel as shown in Figure B-69 allows lumped masses to be defined at a particular node of the reduced model of the crank train. No data entry is required for this panel if the crank train Model, as defined in the Model Definitions Panel, is Rigid Massless or Compliant Massless. If the crank train Data, as defined in the Model Definitions Panel, has been set to Loaded from FE file, then these masses will be additional masses that are not included in the finite-element model of the crank train or defined as part of the Cranknose Assembly Panel or Flywheel Assembly Panel. If the crank train Data has been set to User Defined then this panel is used to define the lumped masses of the crank train model. This panel is also used by the Cranknose Assembly Panel to define additional masses at the crank nose. Any number of lumped masses can be defined at any one node. Item The Add button is used to add a mass to the list. The mass is assigned a reference number (displayed in the Item field), when it is added to the list. The user can move between masses by clicking on the and buttons. Additionally, the currently displayed mass may be deleted from the list by clicking on the Del button or the data updated by clicking on the Update button. If a single node is selected the co-ordinate of the node is displayed as shown in Figure B-69 Type This is used to select how the current lumped mass is to be defined. The mass properties of the lumped mass may either be defined by selecting Cylindrical and supplying the geometry of a cylindrical prism and its position relative to the node, or alternatively by selecting Properties and entering the mass properties of the lumped mass directly. These methods of defining the mass properties are described in 3.4.1.13.1 and 3.4.1.13.2 respectively. Changing this option will change the drawing displayed on the panel.

291


Figure B-69 Lumped Mass Panel

3.4.1.13.1 Cylindrical Mass Properties If Type is defined as Cylindrical the following parameters defining the geometry, position and material properties of the cylindrical prism are required, as shown in Figure B-70.

Figure B-70 Lumped Mass Cylindrical Properties Frame

292

HELP Model Generation Offset This is used to specify the axial distance in x of the centre of gravity of the cylindrical prism relative to the current node. Length This is used to specify the length of the cylindrical prism. Outside Diameter This is used to specify the outside diameter of the cylindrical prism Inside Diameter This is used to specify the inside diameter of the cylindrical prism Material This is used to define the material name of the cylindrical prism. The material properties are defined using the Crankshaft Material Properties Panel, as shown in Figure B-99, and as described in 3.4.2 3.4.1.13.2 Mass Properties If Type is defined as Properties the parameters shown in Figure B-71 defining the mass properties of the lumped mass are required.

Figure B-71 Lumped Mass Properties Frame Axis This is used to specify which co-ordinates are to be used to define the offset of the centre of gravity of the lumped mass relative to the current node. The options are Polar and Cartesian and changing this option will change the drawing displayed on the panel. Mass This is used to specify the mass of the lumped mass Inertia XX This is used to specify the inertia about X at the centre of gravity of the lumped mass in the ENGDYN co-ordinate System Inertia YY

293

HELP Model Generation This is used to specify the inertia about Y at the centre of gravity of the lumped mass in the ENGDYN co-ordinate System Inertia ZZ This is used to specify the inertia about Z at the centre of gravity of the lumped mass in the ENGDYN co-ordinate System Inertia XY This is used to specify the product of inertia Ixy at the centre of gravity of the lumped mass in the ENGDYN co-ordinate System Inertia YX This is used to specify the product of inertia Iyx at the centre of gravity of the lumped mass in the ENGDYN co-ordinate System Inertia ZX This is used to specify the product of inertia Izx at the centre of gravity of the lumped mass in the ENGDYN co-ordinate System X Offset This is used to specify the distance in x of the centre of gravity of the lumped mass relative to the current node. If Axis is defined as Polar the following parameters are required to define the offset of the centre of gravity of the lumped mass in polar co-ordinates. Radius This is used to specify the radial distance of the centre of gravity of the lumped mass relative to the current node. Angle This is used to specify the angular position of the centre of gravity of the lumped mass measured from the y-axis at the current node. If Axis is defined as Cartesian the following parameters are required to define the offset of the centre of gravity of the lumped mass in Cartesian co-ordinates. Y Offset This is used to specify the distance in y of the centre of gravity of the lumped mass relative to the current node. Z Offset This is used to specify the distance in z of the centre of gravity of the lumped mass relative to the current node. 3.4.1.13.3 Display of Lumped Masses Once all the lumped masses at the selected node(s) have been defined and the OK button is selected on the Lumped Mass Panel the Main Panel is updated to display the

294

HELP Model Generation position of the lumped masses. Each lumped mass is displayed in white in which a line is drawn connecting the centre of gravity of the mass to the appropriate node. 3.4.1.14 Damper If the crankshaft has any vibration dampers and the model as defined in the Model Definition panel is Dynamic then data for the dampers is defined using the following panels. Dampers can be defined at any node on the crankshaft axis. Each damper can be defined as either a 1-D torsional vibration damper or as general 3-D damper that includes bending, torsional and axial stiffness and damping.

Figure B-72 Damper Panel The Damper Panel as shown above allows multiple dampers to be defined at a given node. A damper is added using the Add button which displays the Damper Properties Panel. An existing damper can be edited by selecting the damper in the list and using the Edit button to display the same panel. Selecting a damper in the list and using the Delete button will delete the damper.

Figure B-73 Damper Properties Panel

295

HELP Model Generation The Damper Properties Panel consists of a tab for each direction together with a Temperature tab to define an optional temperature map for the selected damper. For each direction the damper can either be Elastomeric or Viscous for that direction. Each direction tab has a number of tabs defining the Mass, Stiffness and Damping data for the damper in that direction with the Overview tab providing a summary as shown. Double clicking on a row will take you to the appropriate tab for that data to edit the data. Selecting TV Damper (1D) will enable the PhiX tab only and the following data can be defined. The mass is defined using the Mass tab as shown below

Figure B-74 Mass tab of Damper Properties Panel Siesmic Mass This is the siesmic mass of the vibration damper for the selected direction and will be a mass or inertia dependent on direction. Stiffness can either be constant, vary with either temperature or frequency or both as shown below. Frequency Dependent data is only used when exporting the model to VALDYN and using the Linear Frequency Domain (LFD) solver within that program. (This export option is not currently available)

296


Figure B-75 Stiffness tab of Damper Properties Panel Stiffness This is the constant stiffness of the vibration damper for the selected direction and will be a translational or rotational stiffness dependent on direction. (If a time domain solution is performed and frequency dependent data has been defined then this constant stiffness value will be used by the solver in this instance) Stiffness File This is the file containing the temperature and/or frequency dependent stiffness data if either Temperature Dependent or Frequency Dependent are selected. The file can be edited or plotted using the adjacent buttons. Selecting Use Current will use any existing stiffness data already stored in the .EDSF file. As with stiffness, damping can either be constant, vary with either temperature or frequency or both as shown below.

297


Figure B-76 Damping tab of Damper Properties Panel If the damper for the given direction is defined as Viscous then the following inputs are required Damping This is the constant damping coefficient of the vibration damper for the selected direction and will be a translational or rotational damper dependent on direction. (If a time domain solution is performed and frequency dependent data has been defined then this constant damping value will be used by the solver in this instance) Damping File This is the file containing the temperature and/or frequency dependent damping data if either Temperature Dependent or Frequency Dependent are selected. The file can be edited or plotted using the adjacent buttons. Selecting Use Current will use any existing damping data already stored in the .EDSF file. If the damper for the given direction is defined as Elastomeric then a quasi-elastomeric model is used in the time domain that approximates the damping characteristic of a rubber spring which has a constant dynamic magnifier across the frequency range. Following the method described by Tonar11 the force in the mount in each direction is given by 1

y2 K F  x  v  v.  v M

[ B-1]

298

HELP Model Generation where K

=

Dynamic Stiffness

M

=

Dynamic Magnifier

x

=

Displacement of crankshaft degree of freedom

v

=

Velocity of crankshaft degree of freedom

and where y is a state variable defined by

y  h  y  xt  t where h is a filter frequency. If we choose x(t )  sin t  , the solution of this differential equation is

yt  

h  sin t     sin t  h2   2

which, if h is small compared to  , gives

yt  

  cost  v(t )  2 2 

It can be seen that the damping term in equation [B-1] approximates to K   M  , which is a form commonly used when modelling rubber elements in the frequency domain. The value of h should be chosen to be much smaller than, but not negligible with respect to, the first natural frequency of the system of the crankshaft. Tests using the Ricardo program VALDYN using the QSTIFF element, which has the same formulation as this model, suggest that setting h to 10-20% of the first natural frequency gives results that correlate well with frequency domain results. The following inputs are required for this model.

299

HELP Model Generation Dynamic Magnifier This is used to specify the dynamic magnifier of the vibration damper for the selected direction. This is related to proportion of critical damping c co and to the loss factor

tan  as follows:

c 1 1  tan   co 2 2M

Damping Magnifier File This is the file containing the temperature dependent damping magnifier data if Temperature Dependent is selected. The file can be edited or plotted using the adjacent buttons. Selecting Use Current will use any existing damping data already stored in the .EDSF file. Filter Frequency This is used to define the filter frequency. The Geometry tab as shown below is used to define data to calculate by the solver the Maximum Shear Stress, Power Loss and Power Density for the vibration damper in the selected direction.

Figure B-77 Geometry tab of Damper Properties Panel If None is selected then no data is required and the Maximum Shear Stress, Power Loss and Power Density values will not be calculated by the solver. Conversely, Dimensions or Specify Values can be used to define the data. If Specify Values is selected then the following inputs are required

300

HELP Model Generation Shear Area This is used to specify the effective shear area in the selected direction so as to calculate the Maximum Shear Stress. Volume This is use to specify the volume of the elastomer or the oil so as to calculate the power density. Selecting General Damper (3D) will enable all the direction tabs and the data as defined above can be defined for ach direction. A given direction can be disabled by setting both the stiffness and damping for that direction to zero. The Temperature tab as shown applies to all the directions of a given damper and is used to define the temperature of the elastomer or oil of the damper. The temperature can either be constant or vary with speed.

Figure B-78 Temperature tab of Damper Properties Panel Temperature This is the constant temperature of the vibration damper if Constant is selected. Temperature File This is the file containing the speed dependent temperature data if Speed Dependent is selected. The file can be edited or plotted using the adjacent buttons. Selecting Use Current will use any existing temperature data already stored in the .EDSF file. Temperature This is the initial temperature of the vibration damper for a thermal balance calculation of the damper. (This has not yet been implemented in the solver)

301

HELP Model Generation 3.4.1.15

Mechanical Links

The Mechanical Link Panel is shown in Figure B-79 and is used to define mechanical links and their associated simulation children at a particular node on the crankshaft.

Figure B-79 Mechanical Link Panel Node This is used to display the selected crankshaft node at which the links are defined and cannot be edited. Children Selecting the LIST button adjacent to Children will display the Simulation Children Panel shown in Figure B-80.

Figure B-80 Simulation Children Panel This lists the currently defined simulation children. Child number zero is always defined as being the parent process. Links Selecting the LIST button adjacent to Links will display the Crankshaft Mechanical Links Panel shown in Figure B-81.

302


Figure B-81 Crankshaft Mechanical Links Panel This lists all the currently defined mechanical links attached to the crankshaft. Directions Selecting the EDIT button adjacent to Directions will display the Edit Direction Panel shown in Figure B-82 if local directions have been defined.

Figure B-82 Edit Direction Panel This allows the direction vector to be edited for each local direction Mechanical links are entered using the table where each row defines a unique link. Rows are added to and deleted using the right hand mouse button with the mouse positioned over the row number. This displays a Row Operations Pop-Up menu with Insert Before, Insert After and Delete options. All the links currently defined at the selected node can be deleted using the Delete All button. 3.4.1.15.1 Defining a Mechanical Link A link is defined or edited using the right hand mouse positioned over the appropriate Link ID cell. This will display a pop-up menu with Edit and Define options. Edit allows existing links at that node to be edited whilst Define allows a new link to be defined. In each case Define Mechanical Link Panel shown in Figure B-83 is displayed.

303


Figure B-83 Define Mechanical Link Panel Link ID This used to specify the unique identification number of the mechanical link shared with the child process. Type This is used to specify the type of the link. Currently the only option is Force. Dimension This is used to specify the dimension of the link. The possible options are, 1DTranslational, 1D-Rotational, 3D-Translational, 3D-Rotational and 3D-6DOF. Axis This is used to specify the co-ordinate system of the crankshaft axis in the connecting application pertaining to the link. The options are either Fixed or Rotating. If connecting to VALDYN then Fixed would be selected. Relative To This is used to specify the co-ordinate system of the crankshaft axis in the connecting application pertaining to the link. The options are Node or Origin. If connecting to VALDYN then Origin would be selected. Axis This is used to specify the co-ordinate system of the crankshaft axis in the connecting application pertaining to the link. The options are either Local or Global. Global This must be specified if Axis is set to Global. The options are X, Y or Z.

304

HELP Model Generation Local This must be specified if Axis is set to Local. An existing local direction can be selected using the Select button, which will display the Link Directions Panel shown in Figure B-84

Figure B-84 Link Directions Panel A new local direction is defined using the Define button, which will display Define Direction Panel shown in Figure B-85.

Figure B-85 Define Direction Panel This allows a local direction vector to be defined in the co-ordinate system of the connecting Application. 3.4.1.15.2 Defining a Simulation Child A simulation child is defined or selected using the right hand mouse positioned over the appropriate Simulation Child cell. This will display a pop-up menu with Define and a list of current simulation child numbers. Define allows a new child to be defined, displaying the Define Simulation Child Panel shown in Figure B-86. This has two tabs, Properties and Transformation.

305


Figure B-86 Define Simulation Child Panel Figure B-86 shows the Properties tab selected. Number This is used to specify the number of the Simulation Child number where zero is reserved for the parent process. Application This is used to define the Simulation Child Process. It can entered be either directly or by clicking on the Select button adjacent to the Application box, which will display the Applications Menu shown in Figure B-87.

Figure B-87 Applications Menu Arguments This is used to specify the arguments to the Application. For example, if the simulation child is a VALDYN simulation then this would be set to –s .dat. Host This is used to define the host on which the Application is to run on. At present all the processes have to run from the same host and so this is set to local. The Transformation tab is shown selected in Figure B-88.

306


Figure B-88 Define Simulation Child Panel Transformation Tab This is used to define any necessary transformation from the co-ordinate system used by the Application to that used by ENGDYN. All vectors are in the ENGDYN global axes. The Mirror vector specifies a normal vector to the mirror, the plane of which passes through the origin at the crankshaft co-ordinate system. If zero is entered for the X, Y or Z values of a transformation, then that transformation will not occur. The transformation matrix is built up in the following order; mirror, rotation about Z, then Y and finally about X and then the translation.

3.4.2

Model Definition

When a connecting rod is selected on the Cranktrain Tool Panel and the Matrix Reduction button is subsequently selected, the Define Model Panel shown in Figure B-89 will be displayed.

Figure B-89 Define Model Panel

307

HELP Model Generation This panel is used to define the nodes and elements that are to be included in the reduced model of the connecting rod. Each node and element of the reduced model is included in a unique set which is then referenced when applying loads. Each set can either be a node, constrained node or face set. A node set consists of a single finite element node. The x, y and z degrees of freedom of the node are included in the reduced freedoms of the model. If the set defines a bearing then the φx, φy and φz degrees of freedom are also included. A constrained node set is defined by a node set and a face set. The face set consists of a number of external element faces of the finite element model. Each element face is defined as an element in the reduced model. The node set consists of two coincident nodes at the centre co-ordinate of the set. The first node is constrained to have the average movement of all the nodes defining the face set. (This average constraint has a similar formulation to that of an RBE3 element used in Nastran.) The second node is connected to the first by a stiff spring. The x, y and z freedoms of this node are included in the reduced freedoms of the model. If the set defines a bearing then the φx, φy and φz degrees of freedom are also included. The degrees of freedom of the first node and the nodes defining the element faces are not included. A face set is similar to a constrained node set, but in this case the x, y and z degrees of freedom of each node of the element faces are also included in the reduced freedoms. A face set is required for all those bearings which are modelled using the Elasto-hydrodynamic (EHD) bearing model. A convention for naming each set has been adopted. For example, names for big end and small end bearings are bigEndBearing smallEndBearing Each set would be stored in the .EDSF and .SFE files with either a NODE_SET or FACE_SET prefix dependent on the set type. If a finite element model of the connecting rod exists (as defined using the Model Definitions Panel) then each set can be defined by either clipping based on a geometric shape within a given tolerance, using the Clip tab, or can be read directly from the SFE file using the Read tab. Sets can also be deleted from the reduced model using the Delete tab. The name of each set is case sensitive. The minimum requirement of the reduced model is that there must be sets defining nodes (and elements) at big end bearing and small end bearing (or only one of these bearings if the appropriate compliant connecting rod model was selected). The geometric definition and name of each of these sets are created when the Define Model Panel is displayed. The program will then attempt to create these sets by clipping. Big end bearing requires a set called bigEndBearing and small end bearing requires a set called smallEndBearing

308

HELP Model Generation A bearing modelled using the Mobility or Hydrodynamic (HD) bearing model (as defined using the Bearing Panel) can either be defined as a node or as a constrained Node set. If the bearing is modelled using the Elasto-Hydrodynamic (EHD) bearing model then it is necessary for this set to be a face set since all the degrees of freedom on the bearing surface need to be included in the reduced set. The definition of node, constrained node and face sets is described below. In cases where there is no finite element model of the connecting rod then nodes and corresponding node sets are created by the program to define each bearing as described above. 3.4.2.1 Creating a Set Sets can be defined, to add nodes and elements to the reduced model, by clipping based on a geometric shape within a given tolerance, using the Clip tab. Figure B-89 shows the Define Model Panel with the Clip tab selected. This has three tabs; Name, Definition and Tolerance. Selecting Clip Set will clip the external faces of the finite element model based on the supplied geometric data within the supplied tolerances. If successful the clipped sets will be displayed. (The viewing parameters for the clipped sets can be defined using the Model Appearance Panel using the Set tab.) The sets can then be added to the reduced model by selecting Add Set. 3.4.2.1.1

Set Name

Figure B-89 shows the Name tab selected. Type This is used to specify the set type. The options are Node, Constrained Node and Face as defined below. Name This is used to specify the base name of the set. The options are Accelerometer, Big End Bearing, Small End Bearing, Restraint and Constraint. 3.4.2.1.2

Clip Definition

Figure B-90 shows the Definition tab selected. This is used to define the boundaries of a shape in which the required set exists within a given tolerance. It has four tabs; Centre, Axis, Extent and Diameter.

309


Figure B-90 Define Model Panel Clip Definition Tab Shape Different clip shapes may be defined depending on the shape and complexity of the set being clipped. The options are Plane, Cylinder and Sphere. The chosen shape will change the input data requirements of the Axis, Extent and Diameter tabs. Centre This is used to define the X, Y and Z centre coordinates of the set. Axis This is used to specify X, Y and Z which define the axis of the shape. Extent This is used to define two lengths A and B, which define the extent of the set along the cylinder axis if the shape is a Cylinder. Diameter This is used to define diameter DA used to define a Cylinder or Sphere. 3.4.2.1.2.1 Defining a Cylinder

A cylinder as shown in Figure B-91 is defined by an axis, a diameter DA and two lengths LA and LB which define the extent of the set from the centre along the axis. The length LA defines the length in the positive direction of the axis whilst defines LB the length in the opposite direction of the axis and will be negative. External element faces which fall within the DA ± t and within LB – t and LA + t will be included in the set, where t is the linear tolerance.

310


DB LB DA

LA

C

Figure B-91 Cylinder Definition

3.4.2.1.2.2 Defining a Sphere

A sphere as shown in is defined simply by the diameter DA.. External element faces, which fall within DA ± the linear tolerance, will be included in the set.

DA C

Figure B-92 Sphere Definition 3.4.2.1.3

Clip Tolerance

Figure B-93 shows the Tolerance tab selected. This is used to define the tolerances within which the clip shape is clipped.

311


Figure B-93 Define Model Panel Clip Tolerance Tab Linear This is used to specify the linear tolerance within which the shape is clipped. The default value will be half the distance between any two adjacent nodes of the finite element model. Angular This is used to specify the angular tolerance within which the shape is clipped and is only used for clipping using a plane shape. 3.4.2.2 Reading Sets Existing sets can be read from the .SFE file, to add nodes and elements to the reduced model, using the Read tab. Figure B-94 shows the Define Model Panel with the Read tab selected.

312


Figure B-94 Define Model Panel Read Tab This has three tabs; Node Sets, Face Sets and Constrained Sets. Each tab consists of a list of sets that exist within the finite element model, allowing the user to select those sets that he wishes to include in the reduced model. Only those sets that follow the naming convention will be listed. Selecting Read Set will read the selected sets from the .SFE file and the nodes and elements defining the sets will be displayed. (The viewing parameters for the read sets can be defined using the Model Appearance Panel using the Set tab.) The sets can then be added to the reduced model by selecting Add Set. When face sets are added for small end or big end bearings and the bearing model type is elastohydrodynamic, lubrication mesh node sets are checked against corresponding FE model face sets. Any lubrication mesh node lying outside the FE model face set will be marked by a red cross and a warning for unmatched nodes will be printed. Unmatched lubrication mesh nodes are always made inactive (removed) during model assembly. Note that unmatched nodes are detected based on an internal tolerance. The detection of unmatched nodes is also influenced by the FE model element size. Therefore, some lubrication mesh nodes outside (but close to) the corresponding FE model face set may not be determined to be unmatched.

313


Figure B-95 Unmatched Lubrication Mesh Nodes for Small End Bearing Removing unmatched nodes can be used to define lubrication meshes for non-trivial bearing shapes which can not be defined on the bearing panel. Figure B-95 shows an example of a small end bearing lubrication mesh (in gray) with nodes outside the FE model face set (in orange). These nodes will be removed during model assembly and the resulting lubrication mesh will have the bearing edge set according to the FE model face set. 3.4.2.3 Deleting Sets Sets and the nodes and elements defining each set can be deleted from the reduced model, using the Delete tab. Figure B-96 shows the Define Model Panel with the Delete tab selected.

314


Figure B-96 Define Model Panel Delete Tab This has three tabs; Node Sets, Face Sets and Constrained Sets. Each tab consists of a list of sets that exist within the reduced model, allowing the user to select those sets that he wishes to delete. Selecting Delete Set will delete the selected sets and the nodes and elements defining the sets will be deleted from the reduced model. 3.4.3

Matrix Reduction

When a connecting rod is selected on the Cranktrain Tool Panel and Matrix Reduction button is clicked, Matrix Reduction Panel shown in Figure B-97 will pop up. This panel will only be displayed if all the sets (as described below) defining big end and/or small end bearing have been defined.

315


Figure B-97 Matrix Reduction Panel This panel is used to perform a finite element solution to derive the reduced mass and stiffness matrices. This is only required if the connecting rod model (as defined using the Model Definitions Panel) is Compliant, Compliant Big End, Compliant Small End or Dynamic, including symmetric variants of compliant models. If the model is dynamic then both mass and stiffness matrices are derived otherwise only a stiffness matrix is evaluated. The matrices can either be evaluated using the imbedded Vectorised Sparse Solver (VSS) or using MSC/NASTRAN or ABAQUS (when evaluating the stiffness matrix.) Solver This is used to specify the solver. Options are Fearce (if the Vectorised Sparse Solver is to be used), MSC/Nastran and Abaqus. Matrix Formulation This is used to specify the reduction method. If the model is dynamic then the option menu is set to Component Mode Synthesis (CMS), otherwise it is set to Static Condensation. No. of Modes This is used to specify the number of modes to be included in the CMS analysis. Output Name This is used to specify the name of the .FRC command file and any analysis data file that is written. Selecting the Default button will set the output name to be the name of the connecting rod .SFE file name as defined using the Model Definitions Panel. Element Check This toggle is used to switch on element checking. Limited Memory The memory used by the Vectorised Sparse Solver can be limited using this toggle Memory

316

HELP Model Generation This is used to specific the maximum memory to be used by the Vectorised Sparse Solver (VSS). Selecting Solve will write the .FRC command file that contains commands to define the sets (defined using the Define Model Panel) and to execute the solution. A prompt is then given allowing the user to proceed with the solution. If Solver is set to Fearce then the solution of the finite element model will be performed and a progress bar will be displayed (as shown in Figure B-98), otherwise a data file will be written that can be executed by the selected FE solver outside ENGDYN.

Figure B-98 Connecting Rod Reduction Progress Panel 3.4.4

Defining Material Properties

The Crankshaft Material Properties Panel, as shown in Figure B-99, is used to define the material properties of the crankshaft.

Figure B-99 Crankshaft Material Properties Panel

317

HELP Model Generation Pressing Select will cause the Material Properties Panel as shown in Figure B-100 to appear.

Figure B-100 Material Properties Panel This panel contains a table listing all of the currently defined crankshaft materials. To select a material, click on the relevant row. When OK is pressed, the values in the calling panel will be updated.

3.4.5

Calculating the Mass Properties of the Crank Train

The Calculate Masses button on the Cranktrain Tool Panel as shown in Figure B-31 is used to calculate the lumped mass properties at each of the nodes of the reduced crank train model from the finite element model of the crank train (if Data is defined as Loaded from FE file) and from any user-defined lumped masses entered as described in 0, 3.4.1.12 and 3.4.1.13. This button is only selected if the crank train Model, as defined in the Model Definitions Panel, is Rigid, Compliant or Dynamic. It is selected once the crank train model has been edited and assembled as described in 3.4.1 and 3.4.1.15. The mass properties must be calculated before proceeding to Define Balance. During calculation, a panel containing various progress bars will be displayed to show degree of completion, as shown in Figure B-101.

Figure B-101 Calculate Masses Progress Panel The finite element model of the crank train is divided into a number of element sets where each set corresponds to the lumped mass at a crank train node. This is shown for

318

HELP Model Generation a V8 example in Figure B-102. The mass properties of each of these sets are calculated and assigned to the appropriate node.

Figure B-102 Element Sets Defining Lumped Masses at each Cranktrain Node

3.4.6

Calculating the Primary Balance of the Crank Train Model

The Set Balance button on the Cranktrain Tool Panel as shown in Figure B-31 is used to calculate the primary balance of the crank train model as modelled. It also allows additional mass properties to be derived so that a user-specified primary balance of the crank train can be obtained. This is achieved either by simulating the drilling of specified crankshaft webs, or by defining balance weights. The Primary Balance Panel, as shown in Figure B-103, is displayed when the Set Balance button is selected. The intention of this panel is not to design the balance weight arrangement of the crankshaft but to provide a means firstly for checking that the primary balance of the model is correct and secondly to correct the model to obtain the designed primary balance. The latter is particularly important when performing an analysis to investigate first order vibration at say an engine mount for example.

319


Figure B-103 Primary Balance Panel At the bottom of the Primary Balance Panel the ‘As Modelled’ mass moments of the cranktrain are displayed. These are the total mass moments ,M.y, ,M.z, ,M.y.x and ,M.z.x of the cranktrain calculated from the lumped mass properties of the model and using the currently defined balancing rate. Also displayed at the bottom of the panel are the ‘Required Balance’ mass moments. These are specified by the user when Calculation is set to Using web data or Defined by m,r,x and are the required total mass moments of the cranktrain. If the Apply or OK buttons are selected the ‘As Modelled’ mass moments of the crank train will be re-calculated using the currently defined balancing rate and a calculation, if requested, will be performed to obtain the ‘Required Balance’. If it is not possible to obtain the ‘Required Balance’ for any reason an error will be displayed. Calculation Initially, this will be displayed as None and the ‘Required Balance’ fields will be greyed out. If OK is selected with this option selected no balance calculation is performed and any subsequent analysis will use the crank train model ‘As Modelled’ with a primary balance as indicated. If it is necessary to correct the primary balance of the crank train model then either Using Web Data or Defined by m,r,x should be selected. The required mass moments should then be supplied in ‘Required Balance’. If Using Web Data is selected, the crank train is balanced by simulating drilling the selected webs using Web

320

HELP Model Generation Numbers. If Defined by m,r,x is selected, the crank train is balanced by evaluating the angular position of two balance weights at user-specified axial positions.

3.4.6.1 Obtaining Primary Balance by Simulating the Drilling of Crankshaft Webs Web Numbers If Calculation is defined as Using Web Data then this is used to specify the numbers of the crankshaft webs at which the crankshaft is to be drilled to obtain the ‘Required Balance’ mass moments. An even number of webs must be specified. Each selected crankshaft web must have a counterweight and its geometry must be defined using the Crank Web Panel shown in Figure B-33 and described in 3.4.1. Pressing the Select button will open the Select Crank Web Panel, as shown in Figure B-104.

Figure B-104 Select Crank Web Panel This panel shows a list of the crankshaft webs. Appropriate webs can be selected from the displayed list. An even number of webs should be selected. Material Name If Calculation is defined as Using Web Data then this is used to specify the counterweight material. Pressing the Define button will open the Material Properties Panel, as shown in Figure B-100, from which a material may be selected 3.4.6.2 Obtaining Primary Balance by Defining Balance Weights If Calculation is defined as Defined by m,r,x the following parameters are required to define two balance weights at user-specified axial positions so as to obtain the ‘Required Balance’ mass moments. Dependant Variable This is used to specify the dependant variable for the balance calculation. The user may either define the mass of each balance weight or the radial position of each weight. If the dependant variable is Mass, then the user must supply the radial position using Radius. Conversely if the dependant variable is Radius then the user must supply the mass using Mass.

321

HELP Model Generation Mass If the Dependant Variable is defined as Radius then this is used to specify the mass of each balance weight Radius If the Dependant Variable is defined as Mass then this is used to specify the radial position of each balance weight Distance, x This is used to define the axial position x of each balance weight. Angle The angular position of each balance weight is calculated when the Apply or OK buttons are selected. The calculated values are displayed, and may not be edited.

3.4.7

Calculating the Stiffness’ of the Crank Train Using the Finite Element Model

The Calculate Stiffnesses button on the Cranktrain Tool Panel as shown in Figure B-31 is used to calculate stiffness values of each of the crankshaft webs and of the crank nose and flywheel assemblies using the finite element model of the crank train. This button is only selected if the crank train Model, as defined in the Model Definitions Panel, is Compliant, Compliant Massless or Dynamic and Data is defined as Loaded from FE file. It is selected once the primary balance of the crank train has been calculated, as described in 3.4.6. ENGDYN sub-divides the finite element model into separate files that each contains the finite element model of a crank web or an end element. If the file name for the V8 example shown in Figure B-102 is V8.SFE 22 files will be created and these are as follows: V8_WEB1B.SFE V8_WEB1T.SFE V8_WEB2B.SFE V8_WEB2T.SFE V8_WEB3B.SFE V8_WEB3T.SFE V8_WEB4B.SFE V8_WEB4T.SFE

V8_WEB5B.SFE V8_WEB5T.SFE V8_WEB6B.SFE V8_WEB6T.SFE V8_WEB7B.SFE V8_WEB7T.SFE V8_WEB8B.SFE V8_WEB8T.SFE

V8_ELEM01A.SFE V8_ELEM01B.SFE V8_ELEM01T.SFE V8_ELEM30A.SFE V8_ELEM30B.SFE V8_ELEM30T.SFE

Multiple files are required for each crank web and element because the constraints for each stiffness calculation are different. ENGDYN calculates the mass properties of each crank web to determine webs that are identical to each other. The stiffness calculations are then only performed on those webs that are unique. For each of the unique crank webs and for the elements at the ends, a finite element solution is performed in each case using the procedures described in Appendix 1. A

322

HELP Model Generation Symmetric Conjugant Gradient solution algorithm is used to perform the solution. Three stiffness values, torsion, in-plane bending and out-of-plane bending are calculated for each crank web. The elements at the ends of the crank train are assumed to be axisymmetric, and for each of these elements three stiffness values, torsion, bending and axial are calculated. Before performing the stiffness calculations ENGDYN checks firstly for the existence of each of the files defining the crank webs and end elements and secondly whether the files for each unique crank web and for the elements at the ends contain the nodal displacement and reaction results of the finite element solution. If all the files and the appropriate results exist the Query Panel shown in Figure B-105 is displayed. This gives the user the option, by answering Yes, of overwriting the existing files and re-performing the finite element solutions. If the user replies with No the existing finite element solutions are used to evaluate the stiffness values.

Figure B-105 Stiffness Results Query Panel

Figure B-106 FE Solutions Query Panel For each crank web two files are created and for web number 1 these are V8_WEB1B.SFE and V8_WEB1T.SFE. These are used to calculate the bending and torsional stiffness values respectively. On completion of the stiffness calculations each file contains the constraints and displacement loads as described in Appendix 1 and the nodal displacements and reactions due to the applied loads and constraints. Each crank web model is orientated such that the main bearing centre is at the origin and the crank pin lies in the xy-plane. For each element at the ends three files are created and for element number 30 these are V8_ELEM30A.SFE, V8_ELEM30B.SFE and V8_ELEM30T.SFE. These are used to calculate the axial, bending and torsional stiffness values respectively. On completion of the stiffness calculations each file contains the constraints and displacement loads as described in Appendix 1 and the nodal displacements and reactions due to the applied loads and constraints. Each element model is orientated such that one of the cut-planes is at the origin. On successful completion of the stiffness calculation, the message ‘Stiffnesses Calculated’ will be displayed. The calculated stiffness values can be viewed in the

323

HELP Model Generation Crankshaft Web Panel, Cranknose Assembly Panel and Flywheel Assembly Panel as shown in Figure B-33, Figure B-64 and Figure B-65 respectively.

3.4.8

Model Assembly

Selecting the Assemble Model button will check various data and perform an eigenvalue solution of the reduced model of the crankshaft, if the crankshaft model is dynamic. If there are any dynamic connecting rod models then an eigenvalue solution of these models will also be performed By default a node set will be created for each node of the crankshaft model during assembly. The name of each set will be of the form crankNode:ID_1 where the ID tag value is the crankshaft node number. If the user wants to have explicitly named sets, to apply multiple loads at the same node for example, as described in Section 0 then a file called CRANKSHAFT.SETS is required in the current working directory. This file consists of 2 columns, the first column is the named sets and the second column is the crankshaft node corresponding to each set. An example of such a file is shown below crankNode:ID_gear crankNode:ID_belt crankNode:ID_chain

1 1 1

3.5 Editing the Cylinder Block Model 3.5.1

Overview

The assembly of the cylinder block model is initiated by selecting the Edit Block button from the Main Panel, which pops up the Cylinder Block Tool Panel shown in Figure B-107.

Figure B-107 Cylinder Block Tool Panel

3.5.2

Model Definition

324

HELP Model Generation Selecting the Define Model button on the Cylinder Block Tool Panel will pop up the Define Model Panel shown in Figure B-108.

Figure B-108 Define Model Panel This panel is used to define the nodes and elements that are to be included in the reduced model of the cylinder block. Each node and element of the reduced model is included in a unique set which is then referenced when applying loads (Sections 4.3.2 and 4.3.3) or when defining an engine mount or a mechanical link (as described below) For example it is necessary to define an engine mount set (which includes the node of the engine side bracket) before defining the parameters of the mount itself as described in 3.5.3.1. Each set can either be a node, constrained node or face set. A node set consists of a single finite element node. The x, y and z degrees of freedom of the node are included in the reduced freedoms of the model. If the set defines a bearing then the φx, φy and φz degrees of freedom are also included. A constrained node set is defined by a node set and a face set. The face set consists of a number of external element faces of the finite element model. Each element face is defined as an element in the reduced model. The node set consists of two coincident nodes at the centre co-ordinate of the set. The first node is constrained to have the average movement of all the nodes defining the face set. (This average constraint has a similar formulation to that of an RBE3 element used in Nastran.) The second node is connected to the first by a stiff spring. The x, y and z freedoms of this node are included in the reduced freedoms of the model. If the set defines a bearing then the φx, φy and φz degrees of freedom are also included. The degrees of freedom of the first node and the nodes defining the element faces are not included. A face set is similar to a constrained node set, but in this case the x, y and z degrees of freedom of each node of the element faces are also included in the reduced freedoms. A face set is required for all those bearings which are modelled using the Elasto-hydrodynamic (EHD) bearing model. A convention for naming each set has been adopted. Each set name consists of a case sensitive base name with a number of tags. The tags CYL and VAL are used to differentiate between cylinders and valves. An identification tag is also used. Example set names would be

325

HELP Model Generation engineMount:ID_right valveSeat:CYL_1:VAL_E1 Each set would be stored in the .EDSF and .SFE files with either a NODE_SET or FACE_SET prefix dependent on the set type. If a finite element model of the cylinder block exists (as defined using the Model Definitions Panel) then each set can be defined by either clipping based on a geometric shape within a given tolerance, using the Clip tab, or can be read directly from the SFE file using the Read tab. Sets can also be deleted from the reduced model using the Delete tab. The name of each set is case sensitive. The minimum requirement of the reduced model is that there must be sets defining nodes (and elements) at each cylinder and at each main bearing. The geometric definition and name of each of these sets are created when the Cylinder Block Tool Panel is displayed. The program will then attempt to create these sets by clipping. Each main bearing requires a set called mainBearing:ID_1 where the Identification ID tag refers to the bearing number. For a bearing modelled using the Mobility or Hydrodynamic (HD) bearing model (as defined using the Main Bearing Panel) can either be defined as a node or as a constrained Node set. If the bearing is modelled using the Elasto-Hydrodynamic (EHD) bearing model then it is necessary for this set to be a face set since all the degrees of freedom on the bearing surface need to be included in the reduced set. The definition of node, constrained node and face sets is described below. Each cylinder of the reduced model is defined with 5 finite element nodes. The position of the nodes is summarised in Table B-10 Cylinder Bore Definition and illustrated in Figure B-109. If the nodes in the figure cannot easily be defined then the nearest nodes are sufficient although it is advisable to ensure the nodes lie in the range (l-r, l+r) and that each pair of lower and upper reversal nodes are the same distance from the crankshaft axis.

326


Figure B-109 Cylinder Bore Definition Nodal Position

Perpendicular Distance from Crankshaft Axis l-r l+r l+r l-r

Small end lower reversal (Thrust-side) Small end upper reversal (Thrust-side) Cylinder head (ideally centre of bore) Small end upper reversal (Anti-thrust side) Small end lower reversal (Anti-thrust side) l - connecting rod length r - crank throw

Table B-10 Cylinder Bore Definition Each node is included in a set and these are cylinder:CYL_1:ID_upperAntiThrust cylinder:CYL_1:ID_lowerAntiThrust cylinder:CYL_1:ID_upperThrust cylinder:CYL_1:ID_lowerThrust cylinder:CYL_1:ID_headGasFace The cylinder CYL tag refers to the cylinder number and the identification ID tag refers to the location of the node. All of these sets are defined as node sets.

327


Figure B-110 Display of Reduced Cylinder Block Model Figure B-110 shows the reduced model of a cylinder block model for an in-line 4 engine. The element faces defining each face or constrained node are shown shaded orange. Those nodes whose degrees of freedom are included in the reduced model are shown in green. The engine mounts and main bearings are defined as constrained node sets. In cases where there is no finite element model of the cylinder block then nodes and corresponding node sets are created by the program to define each main bearing and cylinder as described above. In the case of a Compliant Crankcase model additional nodes are defined to define the nodes of the cylinder. 3.5.2.1 Creating a Set Sets can be defined, to add nodes and elements to the reduced model, by clipping based on a geometric shape within a given tolerance, using the Clip tab. Figure B-108 shows the Define Model Panel with the Clip tab selected. This has three tabs; Name, Definition and Tolerance. Selecting Clip Set will clip the external faces of the finite element model based on the supplied geometric data within the supplied tolerances. If successful the clipped sets will be displayed. (The viewing parameters for the clipped

328

HELP Model Generation sets can be defined using the Model Appearance Panel using the Set tab.) The sets can then be added to the reduced model by selecting Add Set. 3.5.2.1.1

Set Name

Figure B-108 shows the Name tab selected. Type This is used to specify the set type. The options are Node, Constrained Node and Face as defined above. Name This is used to specify the base name of the set. The options are Valve Seat, Valve Spring, Pivot Support, Cam Support, Sprocket Support, Guide Support, Tensioner Support, Follower Support, Cam Bearing, Balancer Bearing, Bearing, Engine Mount, Main Bearing and Cylinder. The following define the tags for each set. The tags attached to the base name are dependent on the set name. For example Valve Seat will have Cyl and Val tags whilst Balancer Bearing will have only an ID tag. Cyl This is used to specify the cylinder number tag for the set. Multiple cylinders may be defined either by typing as a comma-separated list, or selected from the Cylinders Panel shown in Figure B-111 that is popped up by pressing Select.

Figure B-111 Cylinders Panel If a set has a cylinder tag then the centre and axis defining the set, defined using the Definition tab (see below), are relative to the cylinder rather than the ENGDYN origin. The cylinder origin is at the top of the cylinder defined by the Cylinder Height using the Cylinder with the y-axis running along the cylinder axis towards the head. If more than one cylinder is defined then a set will be generated per cylinder. This allows for rapid generation of sets defining locations on the cylinder head and in the cylinder. Val This is used to specify the valve VAL tag for those sets defining those locations where the valve train interacts with the structure. This can be any string or integer. Examples might be I1, I2, E1 and E2.

329

HELP Model Generation ID This is used to specify the identification ID tag. This can be any string or integer. 3.5.2.1.2

Clip Definition

Figure B-112 shows the Definition tab selected. This is used to define the boundaries of a shape in which the required set exists within a given tolerance. It has four tabs; Centre, Axis, Extent and Diameter.

Figure B-112 Define Model Panel Clip Definition Tab Shape Different clip shapes may be defined depending on the shape and complexity of the set being clipped. The options are Plane, Cylinder, Frustum and Sphere. The chosen shape will change the input data requirements of the Axis, Extent and Diameter tabs. Centre This is used to define the X, Y and Z centre coordinates of the set. Axis This is used to specify X, Y and Z which define the axis of the shape. Extent This is used to define two lengths A and B, which define the extent of the set along the cylinder axis if the shape is either a Frustum or Cylinder. Extent This is used to define two diameters A and B. 3.5.2.1.2.1 Defining a Plane Annulus

A plane annulus as shown in Figure B-113 is defined by an axis normal to the plane and two diameters DA and DB which are the outer and inner diameters respectively. External element faces which fall within DB – t and DB + t and within ± t about the centre along the

330

HELP Model Generation axis will be included in the set, where t is the linear tolerance. In addition for a face to be included the face normal has to be parallel with the axis ± the angular tolerance. This shape is used for defining a Valve Spring set.

DA DB

C

Figure B-113 Plane Annulus Definition

3.5.2.1.2.2 Defining a Cylinder

A cylinder as shown in Figure B-114 is defined by an axis, a diameter DA and two lengths LA and LB which define the extent of the set from the centre along the axis. The length LA defines the length in the positive direction of the axis whilst defines LB the length in the opposite direction of the axis and will be negative. External element faces which fall within the DA ± t and within LB – t and LA + t will be included in the set, where t is the linear tolerance.

DB LB DA

LA

C

Figure B-114 Cylinder Definition

3.5.2.1.2.3 Defining a Frustum

331


DA

A frustum as shown in Figure B-115 is defined by an axis, two lengths LA and LB which define the extent of the set from the centre along the axis and two corresponding diameters DA and DB at the ends. The length LA defines the length in the positive direction of the axis whilst LB defines the length in the opposite direction of the axis and will be negative. External element faces which fall within the D ± t and within L B – t and LA + t will be included in the set, where t is the linear tolerance and D is the diameter at any given point along the axis. This shape is used for defining a Valve Seat set.

LA

LB C

DB

Figure B-115 Frustum Definition 3.5.2.1.2.4 Defining a Sphere

A sphere as shown in Figure B-116 is defined simply by the diameter DA.. External element faces, which fall within DA ± the linear tolerance, will be included in the set.

DA C

Figure B-116 Sphere Definition

3.5.2.1.3

Clip Tolerance

Figure B-117 shows the Tolerance tab selected. This is used to define the tolerances within which the clip shape is clipped.

332


Figure B-117 Define Model Panel Clip Tolerance Tab Linear This is used to specify the linear tolerance within which the shape is clipped. The default value will be half the distance between any two adjacent nodes of the finite element model. Angular This is used to specify the angular tolerance within which the shape is clipped and is only used for clipping using a plane shape. 3.5.2.2 Reading Sets Existing sets can be read from the .SFE file, to add nodes and elements to the reduced model, using the Read tab. Figure B-118 shows the Define Model Panel with the Read tab selected.

Figure B-118 Define Model Panel Read Tab

333

HELP Model Generation This has three tabs; Node Sets, Face Sets and Constrained Sets. Each tab consists of a list of sets that exist within the finite element model, allowing the user to select those sets that he wishes to include in the reduced model. Only those sets that follow the naming convention will be listed. Selecting Read Set will read the selected sets from the .SFE file and the nodes and elements defining the sets will be displayed. (The viewing parameters for the read sets can be defined using the Model Appearance Panel using the Set tab.) The sets can then be added to the reduced model by selecting Add Set. When face sets are added for main bearings and the bearing model type is elastohydrodynamic, lubrication mesh node sets are checked against corresponding FE model face sets. Any lubrication mesh node lying outside the FE model face set will be marked by a red cross and a warning for unmatched nodes will be printed. Unmatched lubrication mesh nodes are always made inactive (removed) during model assembly. Note that unmatched nodes are detected based on an internal tolerance. The detection of unmatched nodes is also influenced by the FE model element size. Therefore, some lubrication mesh nodes outside (but close to) the corresponding FE model face set may not be determined to be unmatched.

Figure B-119 Unmatched Lubrication Mesh Nodes

334


Figure B-120 Removed Unmatched Nodes Even though unmatched nodes are removed during the model assembly, it is recommended to change the bearing lubrication mesh geometry to match the FE model where possible. In the example shown above, the bearing groove defined on the bearing panel is too narrow leading to unmatched nodes as shown in Figure B-119. After correcting the groove width, no unmatched nodes will be found, as shown in Figure B-120. Removing unmatched nodes can also be used to define lubrication mesh for non-trivial bearing shapes which can not be defined on the bearing panel. 3.5.2.3 Deleting Sets Sets and the nodes and elements defining each set can be deleted from the reduced model, using the Delete tab. Figure B-121Figure B-121 shows the Define Model Panel with the Delete tab selected.

335


Figure B-121 Define Model Panel Delete Tab This has three tabs; Node Sets, Face Sets and Constrained Sets. Each tab consists of a list of sets that exist within the reduced model, allowing the user to select those sets that he wishes to delete. Selecting Delete Set will delete the selected sets and the nodes and elements defining the sets will be deleted from the reduced model.

3.5.3

Editing the Cylinder Block Components

The area on the left of the Cylinder Block Tool Panel shown in Figure B-108 lists the individual components of the cylinder block that can be edited. In this case, the Mechanical Link is shown highlighted; other components from the list can be highlighted by selecting them with the mouse. These may be edited in any order, but they all must be completed before progressing to Assemble Model. Individual components may be either selected or de-selected using the mouse on the relevant component shown in the main drawing. De-selected components and those with an appropriate set name will be shown in green. Having selected the particular components that you wish to define, they may be edited by clicking on the Edit Selected button. This will open up the relevant panel. 3.5.3.1 Engine Mount Definition Engine mounts can either be modelled using a quasi-elastomeric or hydroelastic model. Each model is a 3 degree of freedom model connecting a node representing the engine side bracket to ground. Each mount is defined such that x lies along the axis of the mount.

336

HELP Model Generation 3.5.3.1.1

Elastomeric Engine Mount

The quasi-elastomeric model approximates the damping characteristic of a rubber spring which has a constant dynamic magnifier across the frequency range. Following the method described by Tonar12 the force in the mount in each direction is given by

y F  x  v  v. v

1 2



K P M

[ B-2]

where K

=

Dynamic Stiffness

M

=

Dynamic Magnifier

P

=

Preload

x

=

Displacement of engine side node

v

=

Velocity of engine side node

and where y is a state variable defined by

y  h  y  xt  t where h is a filter frequency. If we choose x(t )  sin t  , the solution of this differential equation is

yt  

h  sin t     sin t  h2   2

which, if h is small compared to  , gives

337


yt  

  cost  v(t )  2 2 

It can be seen that the damping term in equation [ B-2] approximates to K   M  , which is a form commonly used when modelling rubber elements in the frequency domain. The value of h should be chosen to be much smaller than, but not negligible with respect to, the first natural frequency of the system to which the engine mount is attached. Tests using the Ricardo program VALDYN using the QSTIFF element, which has the same formulation as this engine mount model, suggest that setting h to 10-20% of the first natural frequency gives results that correlate well with frequency domain results. In this case it is suggested that the frequency is set to 1 Hz. Figure B-122 shows the Elastomeric Engine Mount Panel which allows the parameters of an elastomeric engine mount to be edited and applied using the Apply button or for the mount to be deleted using the Delete button. The panel has two tabs, Properties and Orientation.

Figure B-122 Elastomeric Engine Mount Panel ID This is used to display the identification ID tag of the set name corresponding to the selected node and cannot be edited. The Properties tab as show in Figure B-122 contains three identical tabs X, Y and Z to define the properties of the elastomer in each of the three directions. Static Stiffness This is used to specify the static stiffness. This is not currently used in the solution.

338

HELP Model Generation Dynamic Stiffness This is used to specify the dynamic stiffness. Typically the value of this stiffness is 1.4 times the static stiffness. Dynamic Magnifier This is used to specify the dynamic magnifier of the elastomer. This is related to proportion of critical damping c co and to the loss factor tan  as follows:

c 1 1  tan   co 2 2M

Filter Frequency This is used to define the filter frequency. A value of 1 Hz is recommended. Figure B-123 shows the panel with the Orientation tab selected.

Figure B-123 Elastomeric Engine Mount Orientation Tab This allows the mount to be orientated with respect to the ENGDYN co-ordinate system described in 2.2. The orientation vector is defined by three rotations about X, Y and Z. The transformation matrix is built up in the following order; rotation about Z, then Y and finally about X. The default vector, as shown in the figure, orientates the mount such that it is vertical. 3.5.3.1.2

Hydroelastic Engine Mount

Hydroelastic mounts operate in a similar manner to a piston forcing a fluid through a restricted orifice between two chambers to provide damping. Typically a hydroelastic mount consists of upper and lower chambers filled with a glycol fluid mixture, together with an upper element connected to the engine made of a rubber material to support the static engine weight. The two fluid chambers are separated by an inertia track which is a

339

HELP Model Generation channel of specific length used to transport fluid between the two chambers as the engine oscillates. In addition, there is a decoupler that provides amplitude dependence and consists of a small flexible diaphragm. This allows fluid to remain in the upper chamber for small amplitude displacements, and by doing so provides less damping because of the lack of fluid flow through the inertia track. A hydroelastic mount can be represented by the mass-elastic system shown in Figure B-124

Cv, Kv Cs, Ks m Cd, Kd

Figure B-124 Hydroelastic Engine Mount where Ks Cs Kv Cv Cd Kd m

= = = = = = =

Support or static stiffness Support damping Volume stiffness Volume damping Inertia track damping Inertia track stiffness Effective fluid mass

Given that some of this data is not always readily available, and the static stiffness, volume stiffness and inertia track damping dominate, this model can be further simplified as shown in Figure B-125. It is these inputs that are required for ENGDYN in it’s hydroelastic engine model.

340


Kv Ks

m Cd

Figure B-125 Simplified Hydroelastic Engine Mount Model The hydroelastic engine mount model in ENGDYN assumes that the mount acts as a hydroelastic element in the local x axis of the mount only. In the other two directions y and z the mount behaves as an elastomer in the same way as for an elastomeric engine mount as described above. Figure B-126 shows the Hydroelastic Engine Mount Panel which allows the parameters of an hydroelastic engine mount to be edited and applied using the Apply button or for the mount to be deleted using the Delete button. The panel has two tabs, Properties and Orientation.

Figure B-126 Hydroelastic Engine Mount Panel ID This is used to display the identification ID tag of the set name corresponding to the selected node and cannot be edited. The Properties tab as show in Figure B-126 contains three tabs X, Y and Z to define the properties of the mount in each of the three directions. In the X direction the mount behaves as a hydroelastic element as shown in Figure B-125 and has the following inputs:

341

HELP Model Generation Static Stiffness This is used to specify the static stiffness, Ks. This is not currently used in the solution. Dynamic Stiffness This is used to specify the dynamic stiffness, Kd.. Typically the value of this stiffness is 1.4 times the static stiffness. The dynamic stiffness is given by K d  K s  K v . Damping This is used to define the damping coefficient c of the viscous damper. Mass This is used to define the effective mass of the fluid, m. This can be derived from knowledge of the natural frequency of the mount and its dynamic stiffness such that

m

Kd 4 2 f

2

In the Y and Z directions the mount behaves as an elastomer with the following inputs: Static Stiffness This is used to specify the static stiffness. This is not currently used in the solution. Dynamic Stiffness This is used to specify the dynamic stiffness. Typically the value of this stiffness is 1.4 times the static stiffness. Dynamic Magnifier This is used to specify the dynamic magnifier of the elastomer. This is related to proportion of critical damping c co and to the loss factor tan  as follows:

c 1 1  tan   co 2 2M

Filter Frequency This is used to define the filter frequency. A value of 1 Hz is recommended. Figure B-127 shows the panel with the Orientation tab selected.

342


Figure B-127 Hydroelastic Engine Mount Orientation Tab This allows the mount to be orientated with respect to the ENGDYN co-ordinate system described in 2.2. The orientation vector is defined by three rotations about X, Y and Z. The transformation matrix is built up in the following order; rotation about Z, then Y and finally about X. The default vector, as shown in the figure, orientates the mount such that it is vertical. This is shown below as viewed from the front of the engine. Y is out of the page for the engine mount Engine X

Z

Y Z Figure B-128 Engine Mount Orientation

3.5.3.2 Mechanical Links The Define Mechanical Link Panel is shown in Figure B-129 and is used to define mechanical links and their associated simulation child at a particular node on the cylinder block.

343


Figure B-129 Mechanical Link Panel Node This is used to display the selected cylinder block node number at which the links are defined and cannot be edited. Children Selecting the LIST button adjacent to Children will display the Simulation Children Panel shown in Figure B-130.

Figure B-130 Simulation Children Panel This lists the currently defined simulation children. Child number zero is always defined as being the parent process. Links Selecting the LIST button adjacent to Links will display the Cylinder Block Mechanical Links Panel shown in Figure B-131

344

HELP Model Generation Figure B-131 Cylinder Block Mechanical Links Panel This lists all the currently defined mechanical links attached to the cylinder block. Directions Selecting the EDIT button adjacent to Directions will display the Edit Direction Panel shown in Figure B-132 if local directions have been defined.

Figure B-132 Edit Direction Panel This allows the direction vector to be edited for each local direction Mechanical links are entered using the table where each row defines a unique link. Rows are added to and deleted using the right hand mouse button with the mouse positioned over the row number. This displays a Row Operations Pop-Up menu with Insert Before, Insert After and Delete options. All the links currently defined at the selected node can be deleted using the Delete All button. 3.5.3.2.1 Defining a Mechanical Link A link is defined or edited using the right hand mouse positioned over the appropriate Link ID cell. This will display a pop-up menu with Edit and Define options. Edit allows existing links at that node to be edited whilst Define allows a new link to be defined. In each case the Define Mechanical Link Panel shown in Figure B-133 is displayed.

345


Figure B-133 Define Mechanical Link Panel Link ID This used to specify the unique identification number of the mechanical link shared with the child process. Type This is used to specify the type of the link. Currently the only option is Force. Dimension This is used to specify the dimension of the link. The possible options are, 1DTranslational, 1D-Rotational, 3D-Translational, 3D-Rotational and 3D-6DOF. Axis This is always set to Fixed for the cylinder block model. Relative To This is used to specify the co-ordinate system of the cylinder block axis in the connecting application pertaining to the link. The options are Node or Origin. If connecting to VALDYN then Origin would be selected. Axis This is used to specify the co-ordinate system of the cylinder block axis in the connecting application pertaining to the link. The options are either Local or Global. Global This must be specified if Axis is set to Global. The options are X, Y or Z. Local This must be specified if Axis is set to Local. An existing local direction can be selected using the Select button, which will display the Link Directions Panel shown in Figure B-134.

346


Figure B-134 Link Directions Panel A new local direction is defined using the Define button, which will display the Define Direction Panel shown in Figure B-135.

Figure B-135 Define Direction Panel This allows a local direction vector to be defined in the co-ordinate system of the connecting Application. 3.5.3.2.2

Defining a Simulation Child

A simulation child is defined or selected using the right hand mouse positioned over the appropriate Simulation Child cell. This will display a pop-up menu with Define and a list of current simulation child numbers. Define allows a new child to be defined, displaying the Define Simulation Child Panel shown in Figure B-136. This has two tabs, Properties and Transformation.

Figure B-136 Define Simulation Child Panel Figure B-136 shows the Properties tab selected.

347

HELP Model Generation Number This is used to specify the number of the Simulation Child number where zero is reserved for the parent process. Application This is used to define the Simulation Child Process. It can be entered either directly or by clicking on the Select button adjacent to the Application box., which will display the Applications Menu shown in Figure B-137.

Figure B-137 Applications Menu Arguments This used to specify the arguments to the Application. For example, if the simulation child is a VALDYN simulation then this would be set to –s .dat. Host This is used to define the host on which the Application is to run on. At present all the processes have to run from the same host and so this is set to local. The Transformation tab is shown selected in Figure B-138.

Figure B-138 Define Simulation Child Panel Transformation Tab This is used to define any necessary transformation from the co-ordinate system used by the Application to that used by ENGDYN. All vectors are in the ENGDYN global axes. The Mirror vector specifies a normal vector to the mirror, the plane of which passes through the origin at the crankshaft co-ordinate system. If zero is entered for the X, Y or Z values of a transformation, then that transformation will not occur. The transformation matrix is built up in the following order; mirror, rotation about Z, then Y and finally about X and then the translation.

348


Matrix Reduction

Selecting the Matrix Reduction button on the Cylinder Block Tool Panel will pop up Matrix Reduction Panel shown in Figure B-139. This panel will only be displayed if all the sets (as described above) defining each cylinder and main bearing have been defined.

Figure B-139 Matrix Reduction Panel This panel is used to perform a finite element solution to derive the reduced mass and stiffness matrices. This is only required if the cylinder block model (as defined using the Model Definitions Panel) is Compliant, Compliant Crankcase or Dynamic. If the model is dynamic then both mass and stiffness matrices are derived otherwise only a stiffness matrix is evaluated. The matrices can either be evaluated using the imbedded Vectorised Sparse Solver (VSS) or using MSC/NASTRAN or ABAQUS (when evaluating the stiffness matrix.) Solver This is used to specify the solver. Options are Fearce (if the Vectorised Sparse Solver is to be used), MSC/Nastran and Abaqus. Matrix Formulation This is used to specify the reduction method. If the model is dynamic then the option menu is set to Component Mode Synthesis (CMS), otherwise it is set to Static Condensation. No. of Modes This is used to specify the number of modes to be included in the CMS analysis.

349

HELP Model Generation Output Name This is used to specify the name of the .FRC command file and any analysis data file that is written. Selecting the Default button will set the output name to be the name of the cylinder block .SFE file name as defined using the Model Definitions Panel. Element Check This toggle is used to switch on element checking. Limited Memory The memory used by the Vectorised Sparse Solver can be limited using this toggle Memory This is used to specific the maximum memory to be used by the Vectorised Sparse Solver (VSS). Selecting Solve will write the .FRC command file that is contains commands to define the sets (defined using the Define Model Panel) and to execute the solution. A prompt is then given allowing the user to proceed with the solution. If Solver is set to Fearce then the solution of the finite element model will be performed and a progress bar will be displayed (as shown in Figure B-140), otherwise a data file will be written that can be executed by the selected FE solver outside ENGDYN.

Figure B-140 Block Reduction Progress Panel

350


Assembling the Model

The cylinder block model is assembled by selecting the Assemble Model button on the Cylinder Block Tool Panel shown in Figure B-107. If there are items that have not been defined or have not been defined correctly an error message will be displayed.

1

I.Mech.E C99/71, 1971 ‘Vibrations of a Crankshaft’. D.Hodgetts, Advanced School of Automobile Engineering, Cranfield.

2

PhD Thesis (University of London) ‘The Vibrations of Crankshafts’. D.Hodgetts, Cranfield Institute of Technology.

3

I.Mech.E C216/76,1976 ‘The Whirl Modes of a Crankshaft’, D.Hodgetts, Cranfield Institute of Technology.

4

FISITA,1986 ‘The Dynamic Response of Crankshafts and Camshafts’, D.Hodgetts, Cranfield Institute of Technology.

5

I.Mech.E AD Autotech Congress 1987 ‘The Measurement and Prediction of Flywheel Whirl’. A.R.Heath, Ricardo.

6

I.Mech.E C14/87, 1987 ‘Computers in Analysis Techniques for Reciprocating Engine Design’. D.J.Lacy, Ricardo.

7

Ricardo Software, Aprill 2009, ‘PISDYN Manual – Documentation / User’s Manual Release 5.1’

8

Ricardo Software Manual RD.03/61301, April 2003, ’PDTOSF Release 3.0 User’s Guide Version 1.1’

9

Ricardo Software Manual RD.06/333901, October 2006, ‘SDFBROWSER – Documentation / User’s Manual Version 2.0’

10

‘Handbook of Lubrication’, Volume II, E. Richard Booser

11

12th European ADAMS Users’ Conference, 18-19 Nov 1997, Marburg Germany, ‘Modelizing Structural Hysteric Damping’, U. Tonar.

12

12th European ADAMS Users’ Conference, 18-19 Nov 1997, Marburg Germany, ‘Modelizing Structural Hysteric Damping’, U. Tonar.

351

HELP Solution

4 Solution 4.1 General Chapters 1 and 3 described how the crank train and cylinder block models are generated. These models define the physical properties of the engine and are considered constants. This chapter describes how the graphical user interface may be used to perform analyses using these models and in particular how to define various parameters such as oil type, engine speed and load and solution type. The user is able to change these parameters for each load case. Three panels can be opened by clicking on buttons from the Main Panel, these are Lubrication, Loading and Evaluate Solution. The Lubrication Definition Panel and Loading Definition Panel are used to define oil fluid properties and loading respectively and can be opened to add, delete and update data between solutions. Data can be deleted or updated provided it has not already been used in one of the previous solutions. The Evaluate Solution Panel is used to define the analyses to be performed and allows the solution to be executed directly or in batch. This panel can be opened any number of times to perform different solutions, and on completion of each solution the results are appended to the .EDSF file. The solution may be executed as many times as required by the user. This enables the user to solve for different loading conditions, configurations of the component sub-models and different solution methods. The results of each solution are appended to the .EDSF standard data file and may be post-processed (see Chapter 5). A summary file with the suffix .EDSUM is created which contains summary results for each load case. An example file is shown in Figure B-141.

Figure B-141 Example Summary File A number of different solution methods are available to the user. Bearing loads and journal orbits may be calculated using statically-determinate or statically-indeterminate

353

HELP Solution solution methods or alternatively bearing loads, journal orbits and the 3-dimensional vibration behaviour of the crankshaft and cylinder block may be calculated using a timemarching dynamic solution method. For each of these solution methods the oil viscosity at each journal bearing may either be user-defined or derived iteratively by ensuring thermal balance of the oil.

4.2 Lubrication The Lubrication Definition Panel is used to read oil fluid properties into the .EDSF file for use by ENGDYN during solution. Any number of oil types can be defined using this panel for selection during solution. The panel is opened by clicking on the Lubrication button from the main panel. A Lubrication Definition Panel in which a single oil types have been defined is shown in Figure B-142.

Figure B-142 Lubrication Definition Panel The oil property data are read from ASCII files in which the data definitions are set out in a similar way to the Command ‘File format in which arguments are set equal to values. Each file, ending with the suffix .MAT, contains oil property data for a single oil. The filename donates the name of the oil. The first line of each file is assumed to define the base properties of the oil; each subsequent line defines the viscosity at a specific temperature. If the file only contains two pairs of viscosity and temperature as shown in the example below the user has the option of fitting the Walther-ASTM equation to the supplied data using the Equation option. The recognised arguments are as follows:

354

HELP Solution Base Properties: DENSITY

[kg/m3]

SPECIFIC_HEAT_CAPACITY

[kg.K]

BULK_MODULUS

[Pa]

CAVITATION_PRESSURE

[Pa]

PIEZO_VISCOSITY_COEFFICIENT

[Pa-1]

a

[K-1]

b

[K-1]

c

[-]

m

[-]

Temperature Depended Properties (entered in pairs): KINEMATIC_VISCOSITY

[m2/s]

DYNAMIC_VISCOSITY

[N.s/m2]

DENSITY

[kg/m3]

TEMPERATURE

[oC]

In the .MAT file at least one of DYNAMIC_VISCOSITY and KINEMATIC_VISCOSITY should be defined. If only DYNAMIC_VISCOSITY is defined then KINEMATIC_VISCOSITY is derived from it using the supplied density. If varying density is not supplied then the base value is used. For multi-grade oils, viscosity becomes thinner as the shear rate increases. This shear thinning effect may be expressed by the Cross equation as

355

HELP Solution

      

 

where  is shear rate in s 1 ,  0 and thinned viscosities Pa.s  ,

halfway between  0 and

 c s

1

0  

1    c m

  are the low-shear-rate and the fully shear-

 is the shear rate at which the lubricaint viscosity is

  and m is a dimensionless parameter.

The ratio

 0  c is a constant. The halfway shear rate  c is normally given by  c  10 a b (T 273)

where a and b( K 1 ) are constants and T is the oil film temperature in K. These Non-Newtonian properties can be defined by including the constants a, b, c and m of the Cross equation as part of the base properties of the lubricant on the first line of the .MAT file. If they are excluded or are all 0 then shear thinning is disabled. For example, a file for 5W30 would have the name 5W30.MAT and would contain DENSITY=862 SPECIFIC_HEAT_CAPACITY=2000 BULK_MODULUS=5E9 \ CAVITATION_PRESSURE=99990 PIEZO_VISCOSITY_COEFFICIENT=0.0 TEMPERATURE=40 KINEMATIC_VISCOSITY=66.76E-6 TEMPERATURE=100 KINEMATIC_VISCOSITY=11.64E-6

The \ at the end of the first line is a continuation character. Item The Add button is used to add a chosen oil to the list. The oil is assigned a reference number (displayed in the Item field), when it is added to the list. Its mass density and specific heat capacity, which are read from the file, are also displayed. A graph of LogLog Kinematic Viscosity versus Log Temperature will be displayed on the right hand side of the Lubrication Definitions panel. For each successive oil, a line is appended to the currently displayed graph, allowing a comparison of viscous properties. The user can move between oils by clicking on the and buttons. Additionally, the currently displayed oil may be deleted from the list by clicking on the Del button or the data updated by clicking on the Update button. Filename This defines the name of the file containing the current oil type in the format described above. Using the browse button will open the Select Panel shown in Figure B-143.

356

HELP Solution

Figure B-143 Select Panel The Filter unless changed by the user will contain the same path name of the database directory containing a library of oil types. In Figure B-143 the file 15W40.MAT has been selected. Clicking on OK will accept the file and return you to the main panel. Equation The data to define the variation of oil viscosity with temperature may be provided in one of there forms. If Walther-ASTM is selected the oil properties file must contain the kinematic viscosity of the oil at two specified temperatures only. ENGDYN uses the Walther-ASTM equation to calculate the oil viscosity at 0.1 C temperature intervals from -100 C to 300 C. The Walther-ASTM equation is given by log log ( + ) = A + B log T where  and T are the kinematic viscosity in cStokes and temperature in K respectively. A and B are constants, specific for each oil, calculated from the two pairs of viscosity and temperature values supplied in the file and  is a constant as defined using Gamma. If Vogel is seleted the oil properties file must contain the kinematic viscosity of the oil at three specified temperatures only. ENGDYN uses the Vogel equation to calculate the oil viscosity at 0.1 C temperature intervals from -100 C to 300 C. The Vogel equation is given by

 1   2  T 

 T    . exp 

where  (cSt),  1 (degC) and  2 (degC) are contants, specific for each oil, calculated from the three pairs of viscosity and temperature values supplied in the file. T (degC) is the temperature at which the dynamic viscosity,  T  is required If User Defined is selected it is expected that the oil properties file contains the kinematic viscosity of the oil in at least ten specific temperatures from which ENGDYN interpolates during solution.

357

HELP Solution Gamma The value of Gamma, used in the Walther-ASTM equation, may be set as either 0.6cstokes or 0.7cstokes. Plot, Viscosity and Temperature are used to control the axes used for the graph on the right hand side of the Lubrication Definitions panel. Plot This specifies whether Kinematic or Dynamic viscosity is plotted. Viscosity This specifies whether the viscosity axis is LogLog, Log or Linear. Temperature This specifies whether the temperature axis is Log or Linear.

358

HELP Solution

4.3 Loading The Loading Definition Panel shown in Figure B-144 is used to define the loading for the analysis and is opened by clicking on the Loading button from the Main Panel. This panel can only be opened only once the crank train and cylinder block models have been assembled as described in Chapter 3. The panel consists of four tabs, Force Profile, Force Equation, Distortion and In-Cylinder.

Figure B-144 Loading Definition Panel Each tab has a Load option menu, which is used to identify the load case to which the loading refers and has the options Full Load, Part Load or No Load. All but the InCylinder tab has a Model option menu, which is used to identify the model to which the loading refers and has the options Crankshaft and Cylinder Block. The Plot button on each tab is used to display the Loading Diagrams Panel shown in Figure B-145 and is used to plot the loading data. The data that can be plotted depends on the tab that is currently selected.

359

HELP Solution

Figure B-145 Loading Diagrams Panel If the data plotted on the x-axis is Crank Angle then 0 degree corresponds to the datum shown in Figure B-12 and described in 2.2. For an in-line engine this datum will also correspond to TDC for Cylinder No. 1. A tracer and an associated Co-ordinates Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu, which has Tracer and Print options. Selecting Print will open a Print Panel that determines how the plot will be printed 4.3.1

In-Cylinder Loads

4.3.1.1 Gas Cylinder Pressure Loads If the In-Cylinder Model is defined as Slider Crank using the Model Definitions Panel then the Loading Definition Panel is displayed as shown in Figure B-144. This is used to create cylinder pressure maps, which are defined as a series of cylinder pressure diagrams against engine speed. The cylinder pressure maps are stored in the .EDSF file as absolute pressure but may be input as gauge or absolute pressure. During solution, ENGDYN evaluates cylinder pressures at a given engine speed by interpolation from the maps defined using this panel. Data is entered using the table where each row defines data at a single engine speed. Rows are added to and deleted using the right hand mouse button with the mouse position over the row number. This displays a Row Operations Pop-Up menu with Insert Before, Insert After and Delete options.

360

HELP Solution The cylinder pressure data are read from ASCII files in which the data is defined in block or column free format. Each file, ending with the suffix .PRES, is supplied either by typing the name of the file in the File Name column of the table or by using the right hand mouse button positioned over the cell which will display a Select Pressure File pop-up menu. This is then used to display the Pressure File Panel similar to that shown in Figure B-143. Each file contains either a single gas cylinder pressure diagram which applies to all cylinders or multiple diagrams where there is a diagram for each cylinder. The pressures do not necessarily have to be at equally spaced crank angle intervals, unless in the special case when only pressures are defined in the files. The data in each file must be in either blocks or columns, not both, with each block labelled using the names ‘CRANK ANGLE’, ‘GAUGE CYLINDER PRESSURE’ or ‘ABSOLUTE CYLINDER PRESSURE’ as appropriate. Each block or column must be defined with an appropriate unit type. If the units of pressure are ‘pounds per square inch’ then the unit type should be set to ‘psi’ or ‘lbf/in^2’. An example of a file in which a single cylinder pressure diagram is defined using block format is as follows: BLOCK

NUMBER=1 NAME='CRANK ANGLE' UNITS='deg'

0.00000 80.0000 160.000 240.000 320.000 400.000 480.000 560.000 640.000 BLOCK

10.0000 90.0000 170.000 250.000 330.000 410.000 490.000 570.000 650.000

20.0000 100.000 180.000 260.000 340.000 420.000 500.000 580.000 660.000

30.0000 110.000 190.000 270.000 350.000 430.000 510.000 590.000 670.000

40.0000 120.000 200.000 280.000 360.000 440.000 520.000 600.000 680.000

50.0000 130.000 210.000 290.000 370.000 450.000 530.000 610.000 690.000

60.0000 140.000 220.000 300.000 380.000 460.000 540.000 620.000 700.000

70.0000 150.000 230.000 310.000 390.000 470.000 550.000 630.000 710.000

NUMBER=2 NAME='GAUGE CYLINDER PRESSURE' UNITS='bar'

15.8583 4.93880 1.49395 0.85086 1.04215 0.45371 0.49962 0.57071 1.38595

24.5757 4.03838 1.20749 0.84090 1.03976 0.45462 0.50857 0.58182 1.73402

26.1159 3.41322 0.99790 0.87401 1.02032 0.48842 0.52708 0.62143 2.20045

20.1354 2.92003 0.88764 0.95214 0.98457 0.49955 0.52381 0.60344 2.89864

14.4201 2.56355 0.83746 0.98992 0.90603 0.49111 0.53186 0.70102 3.97223

10.4879 2.28901 0.82952 1.02439 0.78639 0.49418 0.53775 0.84878 5.55619

7.89986 2.04112 0.82177 1.03059 0.60423 0.49719 0.54718 0.98019 7.68286

6.13410 1.78568 0.85677 1.04399 0.50758 0.51053 0.54746 1.14777 9.98740

The same data in column format would be as follows: COLUMN COLUMN 0.00000 10.0000 20.0000 30.0000 40.0000 50.0000 60.0000 . . 640.000 650.000 660.000 670.000 680.000 690.000

NUMBER=1 NAME='CRANK ANGLE' UNITS='deg' NUMBER=2 NAME='GAUGE CYLINDER PRESSURE' UNITS='bar' 15.8583 24.5757 26.1159 20.1354 14.4201 10.4879 . . . 1.38595 1.73402 2.20045 2.89864 3.97223 5.55619

361

HELP Solution 700.000 710.000

7.68286 9.98740

If a different cylinder pressure diagram is to be defined for each cylinder then each block or column is differentiated using LEGEND = ‘CYLINDER 1’. For example, for a 4-cylinder engine using column format the data would be of the form: COLUMN COLUMN COLUMN COLUMN COLUMN

NUMBER=1 NUMBER=2 NUMBER=3 NUMBER=4 NUMBER=5

0.00000 10.0000 20.0000 30.0000 40.0000 50.0000 60.0000. . 710.000

15.8583 24.5757 26.1159 20.1354 14.4201 10.4879 . . 9.98740

NAME='CRANK NAME='GAUGE NAME='GAUGE NAME='GAUGE NAME='GAUGE 0.6543 0.6783 0.2000 0.2500 0.4500 . . . 0.2000

ANGLE' UNITS='deg' CYLINDER PRESSURE' CYLINDER PRESSURE' CYLINDER PRESSURE' CYLINDER PRESSURE'

0.1000 0.6200 0.5300 0.3456 . . . . 0.0123

UNITS='bar' UNITS='bar' UNITS='bar' UNITS='bar'

LEGEND=’CYLINDER LEGEND=’CYLINDER LEGEND=’CYLINDER LEGEND=’CYLINDER

1’ 2’ 3’ 4’

0.5689 0.4367 0.9876 . . . . . 0.9774

If multiple cylinder pressure diagrams are supplied then it is expected that they are correctly phased and are with respect to the same datum. Absolute cylinder pressures are calculated by ENGDYN and stored in the .EDSF file at regular angular intervals using a cubic spline fit to the supplied data. The interval at which the data is calculated for a given Load type depends on the engine type and on the minimum interval at which the data is supplied for the first pressure diagram. For an inline engine or a vee engine with even firing the data will be stored at 0.1, 0.2, 0.25, 0.5 or 1.0 degree intervals whichever is closest to the minimum interval of the supplied data. Conversely for a vee engine with an odd number vee angle the data will be stored 0.05, 0.1, 0.125, 0.25 or 0.5 degree intervals. The Ambient Pressure is used to calculate the absolute cylinder pressures if the cylinder pressures in the .PRES file are supplied as gauge pressure. The Crankcase Pressure is used during solution to calculate the nett gas force acting on the piston. The cylinder pressure data stored in the .PRES file is expected to be with respect to Top Dead Centre (TDC) for Cylinder No.1. If the data is with respect to a different datum then the Offset Angle can be used to shift the data. The Interval column is used to specify the crank angle interval at which the data is stored in the .PRES file if no crank angle data exists in the file. If crank angle data does exist then the value returned in this column is the interval at which the data is stored in the .EDSF file. The Factor column is used to factor the absolute cylinder pressures. The supplied factor can be a constant as shown in the figure or an expression as a function of T, IMEP, P or PMAX where these variables correspond to the Indicated Torque, Indicated Mean Effective Pressure, Indicated Power or Maximum Cylinder Pressure calculated for the supplied cylinder pressure diagram. The expression must have consistent units. For example if the cylinder pressures are to be factored to give an indicated torque of 210 N.m then the expression would be 210[N.m]/T. This enables cylinder pressures to be factored to give

362

HELP Solution the correct torque, power or pressure characteristic. This can be helpful in instances where the cylinder pressure data may be limited. If the Loading Definition Panel is opened and cylinder pressure data had previously been defined using the Pressure Definition Panel used in earlier versions of ENGDYN the data is updated. For each .PRES file a new file is created with a _UPDATED.PRES suffix. This new file contains the same data but defines whether the pressure is gauge or absolute as described above. This definition was previously defined on the panel and not in the file itself. If the Plot button is selected the Loading Diagrams Panel shown in Figure B-145 is displayed. Data is plotted by selecting at least one item from each of the lists. Only one Data item can be selected and if an item other than Cylinder Pressure is selected the Cylinder and Speed lists are ghosted. Absolute Cylinder Pressure can be plotted against crank angle for any number of cylinders and engine speeds. Crankcase Pressure, Ambient Pressure, Indicated Torque, Indicated Power, Maximum Pressure and I.M.E.P. can be plotted against engine speed for all cylinders. 4.3.1.2 Secondary Dynamic Loads If the In-Cylinder Model is defined as Secondary Dynamic using the Model Definitions Panel then the Loading Definition Panel is displayed as shown in Figure B-146.

Figure B-146 Loading Definition Panel In this case the in-cylinder loads are calculated using the Ricardo program PISDYN1 in which the secondary dynamics of the piston are considered. Filename This is used to specify the name of the Ricardo Standard Data file (sdf) containing the loads calculated by PISDYN. The user may click on the Browse button to open a Select Panel similar to that shown in Figure B-143. If the user clicks on the Default button then the file name defined using the Model Definitions Panel is selected. The contents of the sdf file can be viewed using the Ricardo Program SDFBROWSER2 and should contain the following data: g:cylinder:pressure g:cylinder:crankcasePressure g:conrod:bigEndForce g:conrod:bigEndMoment g:piston:crownForce

363

HELP Solution g:piston:crownMoment g:piston:skirtForce g:piston:skirtMoment Each of these variables is an array except g:cylinder:crankcasePressure. Each array will have a corresponding crank angle ordinate array stored as one of the following g:cylinder:crankAngle g:conrod:crankAngle g:piston:crankAngle Each item of data will be tagged by PISDYN1 to associate the data with a particular cylinder at a particular load and speed. Each tag is of the form :*_*. The tags SP, LD and CYL are used to differentiate between engine speed, engine load and cylinder number respectively. For example the data array g:cylinder:pressure:LD_100:SP_2000:CYL_1 contains absolute cylinder pressures for cylinder 1 at 2000 rev/min 100% load. Each data item as described above and stored in the sdf will contain the speed tag SP. The load tag LD and cylinder tag CYL are optional. If these tags are excluded then ENGDYN assumes this data to apply to all cylinders regardless of load. If the load tag LD is used then it must have the value 100, 50 or 0 dependent on the selected Load type, Full load, Part Load or No Load respectively. If the cylinder tag CYL is used then the numbering convention must be the same as that defined using the Engine Configuration Panel. Data are calculated by ENGDYN and stored in the .EDSF file at regular angular intervals using a cubic spline fit to the supplied cylinder pressure data and using a linear fit to the supplied force and moment data. The interval at which the data is calculated for a given Load type depends on the engine type and on the minimum interval at which the data is supplied. For an inline engine or a vee engine with even firing the data will be stored at 0.1, 0.2, 0.25, 0.5 or 1.0 degree intervals whichever is closest to the minimum interval of the supplied data. Conversely for a vee engine with an odd number vee angle the data will be stored 0.05, 0.1, 0.125, 0.25 or 0.5 degree intervals. If the Plot button is selected the Loading Diagrams Panel shown in Figure B-147 is displayed.

364

HELP Solution

Figure B-147 Loading Diagrams Panel Data is plotted by selecting at least one item from each of the lists. Only one Data item can be selected. If Indicated Torque, Indicated Power, Maximum Pressure or I.M.E.P are selected then the Cylinder and Speed lists are ghosted and the data is plotted against engine speed for all cylinders. Absolute Cylinder Pressure, Big End Force, Crown Force, Crown Moment, Skirt Force and Skirt Moment can be plotted against crank angle for any number of cylinders and engine speeds. The big end force is the force acting on the crank pin with respect to a fixed axes set at the pin centre in which Y and Z are as defined in Figure B-12. The skirt and crown forces are with respect to the piston pin centre and are the forces and moments acting on the cylinder bore. The skirt forces and moments are integrated from the hydrodynamic and asperity pressures calculated at each time step by PISDYN1. 4.3.2

Force Profile

The Force Profile tab shown in Figure B-148 is used to apply loads to the cylinder block and crankshaft which are stored in an .sdf file and originate from VALDYN or in an ASCII data file.

365

HELP Solution

Figure B-148 Loading Definition Force Profile Tab

Applying Loads from VALDYN Applying loads written by VALDYN is described in the appendix Applying Loads Predicted by VALDYN to the Cylinder Block and Crankshaft Models

Applying Loads from an ASCII Data File Appendix ASCII data file format read by Sdf_Read_Ascii() describes the required format of the file. An example file to apply loads can be as follows: COL g:profile1D:crankAngle[deg] 1 . . . 720 TAGS = TYP_crankNode:ID_gear:SP_2000 COL g:profile1D:force:DIR_4[N.m] COL g:profile1D:force:DIR_7[N] COL g:profile1D:force:DIR_8[N.m] -48 1450 94 . . . . . . . . . -96 2150 56

366

HELP Solution TAGS = TYP_crankNode:ID_belt:SP_2000 COL g:profile1D:force:DIR_2[N] COL g:profile1D:force:DIR_3[N] COL g:profile1D:force:DIR_4[N.m] COL g:profile1D:force:DIR_5[N.m] COL g:profile1D:force:DIR_6[N.m] 360 1450 -75 -24 38 . . . . . . . . . . . . . . . 315 630 -35 -34 16 The local directions DIR_1, DIR_2, DIR_3, DIR_4, DIR_5, DIR_6, DIR_7, and DIR_8 then need to be defined by pressing the Edit button adjacent to Directions. It is also possible to apply loads in ENGDYN coordinate system. The same example file above to apply loads can be modified to define loading both in ENGDYN coordinates and local coordinates as follows: COL g:profile1D:crankAngle[deg] 1 . . . 720 TAGS = TYP_crankNode:ID_gear:SP_2000 COL g:profile1D:force:DIR_phiX[N.m] COL g:profile1D:force:DIR_7[N] COL g:profile1D:force:DIR_8[N.m] -48 1450 94 . . . . . . . . . -96 2150 56 TAGS = TYP_crankNode:ID_belt:SP_2000 COL g:profile1D:force:DIR_Y[N] COL g:profile1D:force:DIR_Z[N] COL g:profile1D:force:DIR_phiX[N.m] COL g:profile1D:force:DIR_phiY[N.m] COL g:profile1D:force:DIR_phiZ[N.m] 360 1450 -75 -24 38 . . . . . . . . . . . . . . . 315 630 -35 -34 16

367

HELP Solution Local directions DIR_7 and DIR_8 then need to be defined by pressing the Edit button adjacent to Directions. Each force profile is applied either to a node or elemental face set of the reduced model of the crankshaft or cylinder block. The table displays all those sets that have loads applied and cannot be edited. The table’s contents are changed by selecting an .sdf written by VALDYN or ASCII data file using the Browse button. This will update the File field, whose contents may be deleted by pressing the Delete button. For the crankshaft model additional sets can be added to the model using the file CRANKSHAFT.SETS as described in Section 3.4.8. In the example above, then this would contain the following: crankNode:ID_gear crankNode:ID_belt crankNode:ID_chain

1 1 1

The centre co-ordinate and axis of each set can be edited by pressing the Edit button adjacent to Sets, this pops up the panel shown in Figure B-149.

Figure B-149 Edit Loaded Set Panel The local coordinates can be defined by direction vectors which can be edited by pressing the Edit button adjacent to Directions, this pops up the panel shown in Figure B-150.

Figure B-150 Edit Direction Panel Pressing plot will bring up the Loading Diagrams Panel shown in Figure B-151.

368

HELP Solution

Figure B-151 Loading Diagrams for Force Equation and Force Profile This displays the sum of the forces or moments due to force profile and force equation definitions at a given speed for a given set in a particular direction. If the load is applied to the crankshaft then the plotted loads will be with respect to the rotating axis system. A tracer and an associated Co-ordinates Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu that has Tracer and Print options. Selecting Print will open a Print Panel that determines how the plot will be printed

4.3.3

Force Equation Loads

The Force Equation tab shown in Figure B-152 is used to apply force and moment loads to the crankshaft and cylinder block that can be expressed as an equation. Typical examples where this may be used would be applying balancer shaft bearing loads and applying a constant belt load to the nose of the crankshaft.

369

HELP Solution

Figure B-152 Loading Definition Force Equation Tab The centre co-ordinate and axis of each set can be edited by pressing the Edit button adjacent to Sets, this pops up the panel shown in Figure B-149. The direction vectors are edited by pressing the Edit button adjacent to Directions, this pops up the panel shown in Figure B-150. Pressing plot will bring up the Loading Diagrams Panel shown in Figure B-151. A force equation is entered using the table where each row defines a load applied to a set in a given direction. Rows are added to and deleted using the right hand mouse button with the mouse positioned over the row number. This displays a Row Operations Pop-Up menu with Insert Before, Insert After and Delete options. A Set is defined or deleted using the right hand mouse positioned over an appropriate cell in the set column. This will display a pop-up menu with Define and Delete options. Selecting the latter will delete the selected allows a new set to be defined, whilst selecting Define will display the Define Loaded Set Panel shown in Figure B-153.

Figure B-153 Define Loaded Set Panel Selecting the Select button will display the Select Set Panel shown in Figure B-154.

370

HELP Solution

Figure B-154 Select Set Panel This lists the node and face sets that been defined previously for the given Model. This panel allows a set to be selected. If a node set is selected then the centre and axis fields on the Define Loaded Set Panel will be ghosted since the centre co-ordinate is the coordinate of the node and the axis has no meaning. If a face set is selected then centre defines the centre of the set, whilst axis defines the set axis dependent on the shape that the set is defining. Default values will be displayed which are derived by evaluating a best-fit cylinder, cone or plane. If the set is defining a frustum or cylinder then axis defines the axis of that shape. If the set is defining a plane then the axis defines the vector normal to the plane. A Direction is defined using the right hand mouse positioned over an appropriate cell in the direction column. This will display a pop-up menu containing the global directions X, Y and Z allowing the load to be applied in one of those directions; any previously defined local directions; and Define. Selecting Define will display the Define Load Direction Panel shown in Figure B-155, which allows a new local direction vector to be defined.

Figure B-155 Define Load Direction Panel The Equation can be a constant as shown in the figure or an expression as a function of N, OMEGA, THETA, T and PI where these variables correspond to speed, rotational speed, crank angle time and pi. The expression must have consistent units. The equation can have an equality as shown in the figure where the function can be F or M where these variables correspond to force and moment. Alternatively, the equation can have no equality provided the resultant units of the function are either N or N.m. In the example above 10000[N.m] would be equivalent to F=10000

371

HELP Solution If the specified Model is the Crankshaft then there is an additional column which allows the load to be applied with respect to the rotating axis system or the fixed axis system. Using the right hand mouse will display a pop-up menu containing Fixed and Rotating.

4.3.4

Distortion Loads

The Distortion tab shown in Figure B-156 is used to define the distortion of the journals and bearings for each of the journal bearings under operating conditions. This effectively allows the user to define the hot shape of the journal and bearing.

Figure B-156 Loading Definition Distortion Tab If Model is set to Crankshaft then the journal distortion for each of the bearings is defined. If Cylinder Block is selected then the bearing distortion for each of the main bearings is defined. Similarly if Connecting Rod is selected then the bearing distortion of each pin bearing is defined. The Distortion menu has the options None, Constant and From Distorted Shape. If None is selected, no distortions are applied for the selected Load and Model. If Constant is selected, a fixed distortion value may be entered for each bearing. Rows are added to and deleted by using the right mouse button with the cursor positioned over the row number. This displays a Row Operations Pop-Up menu with Insert Before, Insert After and Delete options. Radial distortion values are entered for each bearing. These values are the total radial distortion of the bearing or journal when the bearing is hot at the given Load and Engine Speed. If From Distorted Shape is selected then the displacements of a finite element solution (loaded to represent the hot running engine) are used to derive the radial distortion of each journal or bearing relative to the axis of a best-fit cylinder. For each bearing the mean radial distortion and the distortion relative to the best-fit cylinder are evaluated. These are then added to the cold manufactured values defined using the Main Bearing

372

HELP Solution Panel as appropriate. The total mean radial displacements are displayed in the table. The .SFE file containing the finite element results is selected using the Browse button. This will update the File field. Selecting Delete will blank the File field. The model can be transformed by selecting the Define button adjacent to Transformation. A Loadcase is defined using the right hand mouse positioned over an appropriate cell in the Loadcase column. This will display a pop-up menu containing the list of loadcases in the .SFE file. Once a Loadcase and an Engine Speed have been selected the distortions are calculated. Selecting View will display the Contour Panel and Distortion Speeds Panel shown in Figure B-157 and Figure B-158 respectively.

Figure B-157 Contour The Contour panel shown in Figure B-157 is used to control the display of the contours on the lubrication mesh on the Main Panel. The option menu specifies the Value that is displayed. The options are Off, Journal Profile, Bearing Profile and Clearance.

Figure B-158 Distortion Speeds Panel The Distortion Speeds Panel is used to select the engine speed to be viewed.

373

HELP Solution

4.4 Evaluate Solution 4.4.1

Solution

The Evaluate Solution Panel shown in Figure B-159 is used to define the analyses to be performed and allows the solution to be executed directly or in batch. This panel consists of a number of tabs which are described below. This figure shows the Solution tab selected. The panel is opened by clicking on the Evaluate Solution button from the Main Panel. If the file was not saved before selecting the panel, the user will be prompted to do so before being allowed to continue. If a solution has already been performed then the parameters used for the last load case stored in the .EDSF file will be used as defaults wherever possible.

Figure B-159 Evaluate Solution Panel

Solution Type The solution type may be set to Determinate, Indeterminate or Dynamic:

374

HELP Solution

Determinate

Main bearing reactions and main and big end journal bearing orbits are calculated assuming a statically-determinate solution.

Indeterminate

Displacements, main bearing reactions and journal orbits are calculated assuming a statically-indeterminate solution allowing for the compliance of the crankshaft and the cylinder block (if included).

Dynamic

Main bearing reactions, journal orbits and vibratory displacements and velocities of the crankshaft and cylinder block (if included) are calculated for a fully three-dimensional dynamic model.

If Indeterminate is selected a Determinate solution will be performed to provide an initial condition to the solution. Likewise if Dynamic is selected the quasi-static Determinate and Indeterminate solutions will also be performed. Cylinder Block Model This parameter allows the cylinder block model type to be selected for the current solution. The model type can be Rigid, Rigid Body, Compliant or Dynamic dependant on the model defined during model generation using the Model Definitions Panel. Connecting Rod Model This defines whether the connecting rod is either Dynamic or Compliant in any small end or big end bearing EHL solutions. This only applies if the solution Type is Determinate as defined on the Solution tab above and the bearing model boundary condition is not defined as Mobility chosen on the Bearing Model tab below. Obviously if no connecting rod models have been defined as Dynamic or Compliant using the Connecting Rod tab of the Model Definitions Panel then this option menu does not apply. The parameters that control the solution are defined as follows: If a Dynamic or Indeterminate solution has been selected, the following parameters must be defined. Abs Tolerance This is used to define the absolute error tolerance at each integration step. Rel Tolerance This is used to define the relative error tolerance at each integration step.

375

HELP Solution Global Tolerance This is used to define the global error tolerance, g. At a given crank angle, n, the solution has converged provided the equation,

F   g

F  Fn180. N  Fn is satisfied for each force and moment at every journal and thrust bearing and where N is the engine cycle, 2 or 4. The moments are divided by the bearing shell length to give an equivalent force. The absolute and relative error tolerances control the error at each integration step. If global errors are to be reduced to satisfy the global error tolerance then the absolute and relative error tolerances need to be reduced. It is recommended that these error tolerances are scaled down uniformly. For statically-indeterminate and dynamic solutions the value F is stored at each crank angle and at each bearing, enabling the data to be plotted using the Plot Results Panel. For each bearing the data is plotted against crank angle, enabling users to determine at which crank angles the solution did not converge (within the Global Tolerance) and at which bearing. End Angle This is used to specify the crank angle at which the solution is to finish. If the solution has converged, satisfying the Global Tolerance at a smaller crank angle, then the solution will finish at that crank angle. If the thermal balance solution is switched on, then this angle refers to the crank angle for each thermal iteration. Initial Step Length This is used to specify the step length to be attempted on the first integration step. Max No of Steps This specifies the maximum number of integration steps that may be attempted per CA. If this limit is exceeded, the solution is terminated with an error. Number of Processors Simulation can be distributed to a specified number of processors. Simulation Children Selecting Edit will display the Edit Simulation Children Panel to allow the child data to be edited if required. The parameters that control the output are defined as follows: Print Interval This is used to specify the crank angle interval at which the results are stored. The default value is either 1.0 or 0.5 degree CA depending on the engine type. A value less

376

HELP Solution than these default values can be specified in order to enable high frequencies (typically 2500 Hz and above) to be resolved when performing subsequent frequency response and acoustic analyses. This can be a particular problem at low engine speeds. For an inline engine or a vee engine with even firing and an even number vee angle the possible crank angle divisions will be in the range (0.1,1.0) with possible values of 0.1, 0.2, 0.25, 0.5 and 1.0. Conversely for a vee engine with an odd number vee angle the possible crank angle divisions will be in the range (0.05,0.5) with possible values of 0.05, 0.1, 0.125, 0.25 and 0.5. The memory requirement and .EDSF file size will increase with decreasing crank angle interval. Reduced Bearing Output This is used to control whether or not the oil film thickness’ and oil film pressures at each crank angle and bearing angle are to be saved to file on completion of each loadcase. The default is not to save these results to file. If these results are saved there is a large overhead in terms of the size of the .EDSF file. Once this has been set for the first loadcase it cannot be subsequently changed. Data Recovery For those models that are compliant or dynamic model and have been derived using the finite element method it is possible to perform data recovery from the reduced model (used by ENGDYN) to the complete finite element model. This uses a data recovery matrix derived during the matrix reduction procedure and which is stored in the SFE file. If this is selected then at the end of each completed loadcase ENGDYN will perform the data recovery for each model from the reduced degrees of freedom to all the nodes of the reduced model. This will only happen for each model if both the SFE exists and the data recovery matrix is in the file. Save Static Results This parameter is used to specify whether the statically-indeterminate solution used as an initial condition for the dynamic solution is to be saved and is only relevant for a Dynamic solution. If this is switched on the statically-indeterminate results are stored as a separate loadcase.

4.4.2

Cases

Figure B-160 shows the Evaluate Solution Panel with the Loadcases tab selected.

377

HELP Solution

Figure B-160 Evaluate Solution Panel Cases Tab This tab is used to define the load cases to be analysed. During one solution any number of load cases may be analysed for a given selected loading. Loading This is used to specify the loading condition to apply at the given engine speed. Full Load, Part Load and No Load refer to the loading found in the .EDSF file and defined using the Loading Definition Panel described in 4.3. Alternatively Inertia Only will only apply the inertia loading to the crankshaft (no cylinder pressure is applied). Loads are derived from the appropriate loading map stored in the .EDSF file. If the InCylinder Model was defined as Slider Crank using the Model Definitions Panel then the loads are derived by interpolation at the specified engine speed. If an engine speed is specified that is beyond the range stored in the appropriate loading map, loads are not extrapolated but the closest available load definition is used. If only one load data point defines the map then that loading is used regardless of engine speed. Conversely, if the In-Cylinder Model was defined as Secondary Dynamic then solutions can only be performed at the speeds at which the loads were derived using PISDYN (see Section 4.3.1.2). Interpolation This option menu is used to specify the interpolation method for interpolating the cylinder pressure at the given engine speeds and load from the user defined cylinder pressure maps. Two methods are available, Spline or Linear. This option menu is only relevant if the In-Cylinder Model was defined as Slider Crank using the Model Definitions Panel.

378

HELP Solution Speeds may be entered either directly using the table or clicking on the Select button adjacent to Speeds. If the table is used then rows are added to and deleted using the right hand mouse button with the mouse position over the row number. This displays a Row Operations Pop-Up menu with Insert Before, Insert After and Delete options. If the user wishes to delete all the speeds currently defined in the table then the Delete All button is used. Speeds The user may define multiple loadcases to perform a speed sweep by clicking on the Select button adjacent to Speeds. If the In-Cylinder Model was defined as Slider Crank using the Model Definitions Panel then this opens the Speed Sweep Panel shown in Figure B-161.

Figure B-161 Speed Sweep Panel The option menu Series is used to select whether the list of speeds are to be generated using an Arithmetic or Geometric series. A loadcase will be generated automatically for each speed and will be added to the table on the left of the Evaluate Solution Panel once OK is clicked. The Start and End engine speeds must be entered regardless of the selected series. If an Arithmetic series has been selected then the desired speed Step between successive loadcases must also be supplied. If a Geometric series has been selected then the list of speeds is defined using the equation:

N  N 0 r i i  0, n where N0 is the first speed defined by Start and r is the Ratio which must be > 1.0 Conversely if the In-Cylinder Model was defined as Secondary Dynamic using the Model Definitions Panel then the Speeds Panel shown in Figure B-162 is displayed using the Select button. This panel is used to select those speeds for which solutions are required.

Figure B-162 Speeds Panel

379

HELP Solution

4.4.3

Model Options

Figure B-163 shows the Evaluate Solution Panel with the Model Options tab selected and is used to define the dynamic parameters for the model.

Figure B-163 Evaluate Solution Panel Model Options Tab Crankshaft Vibration Damper This toggle is used to select and de-select any crankshaft vibration dampers that have been defined. Centrifugal Loads This field is used to select whether the centrifugal loads due to the mass and inertia of the crankshaft are Excluded or Included and is relevant only for a quasi-static solution and if the crankshaft model includes mass. The Excluded option would be used when the results of the solution are to be used as applied loads for a static stress analysis of the crankshaft in which only the influence due to cylinder gas pressure is considered.

380

HELP Solution Cylinder parameters Cylinder Damping This parameter is used to specify the cylinder-damping coefficient. This parameter is not required if all the connecting rod models have been defined as Rigid or Dynamic using the Connecting Rod tab of the Model Definitions Panel, in which case the Pin Constant and Ring Constant are defined instead. For the case when any of the connecting rod models have been defined as either Rigid or Dynamic using the Connecting Rod tab of the Model Definitions Panel then two constants are required for the piston friction model. Piston Constant The piston constant C p is the constant in the equation

 .  f p  Cp  W 

0.5

defining the friction coefficient for the piston where where

W

= Load per unit length acting on the piston



= piston velocity



= kinematic viscosity

The default for the constant C p is 92.4 Ring Constant The ring constant C r is the constant in the equation

 .  f r  Cr   W 

0.5

defining the friction coefficient for each ring where where

W




= piston velocity

381

HELP Solution




The default for the constant C r is 159.1. Coupling In order to ensure that the crankshaft rotates at a constant speed at the flywheel and to constrain the rigid body motion about the crankshaft axis a torsional stiffness and damper is added at the flywheel connected to ground. The value of this stiffness should be such that the natural frequency of the rigid body mode in torsion should be significantly lower than the first flexible torsion mode of the crankshaft. A default value of 50 Hz is usually adequate for most automotive applications. The stiffness is defined either by specifying the Stiffness or the Frequency. The option menu Variable is used to select which variable is entered. If the frequency is specified then the corresponding stiffness is calculated, and vice-versa. The frequency is calculated from the polar inertia of the crankshaft J and the stiffness K such that:

f 

1 2

K J

The damping coefficient of the coupling damper is defined either by specifying the Coefficient or the damping Ratio. The option menu Damping is used to select which variable is entered. If the damping ratio is specified then the corresponding damping coefficient is calculated, and vice versa. The damping coefficient is calculated from the polar inertia of the crankshaft J, the coupling stiffness K and the damping ratio such that

C  2 JK A value of 0.5 for the damping ratio is recommended. The inertia of the crankshaft J includes the inertia due to the sum of the rotating mass and half the reciprocating mass at each cylinder. Cylinder Block Modal Damping The modal damping characteristic is defined by selecting the Define button which displays the Modal Damping Characteristic panel shown in Figure B-164. Modal damping can only be defined if a Dynamic solution has been selected and the cylinder block model type is Dynamic. The modal damping characteristic of the cylinder block is defined as a curve of proportion of critical damping (/c) against frequency. During solution, each mode of the crankshaft is assigned by linear interpolation a proportion of critical damping. For modes

382

HELP Solution with a frequency greater than the maximum defined frequency a value of 1.0 is assigned to /c, whilst for modes with a frequency less than the minimum defined frequency the value of /c at the minimum is assumed. At least two pairs of damping and frequency values are expected to be defined using the panel.

Figure B-164 Modal Damping Characteristic Panel The option menu Definition is used to define the modal damping characteristic. The options are User or Default. If User is selected values of frequency and damping ratio can be added to the table on the left of the panel. Rows are added to and deleted from the table using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and Delete Row options. If Default is selected the default equation

  2.443  10 2 e 0.0007f c is used which is based on a curve fit to experimental data. The units of frequency are defined using Freq Units. A tracer and an associated Co-ordinates Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu, which has Show Tracer and Hide Tracer options.

383

HELP Solution 4.4.4

Bearing Model

Figure B-165 shows the Evaluate Solution Panel with the Bearing Model tab selected and is used to define the parameters for the bearing model.

Figure B-165 Evaluate Solution Panel Bearing Model Tab Model (Boundary Conditions) The solution can be Mobility Half-Sommerfeld Reynolds + JFO Half-Sommerfeld, Mass Conserving (Recommended) Reynolds, Mass Conserving For further details of these models please see Section 2.6.3 and Theory Section 4.2.

384

HELP Solution Composite GWT This parameter is used to specify which method is used to calculate the composite Greenwood-Tripp parameters will be used. The options are: v4.0 and Earlier (Deprecated) Hertzian (Recommended) It is recommended to use the Hertzian calculation as the “v4.0 and Earlier (Deprecated)” option is provided for backwards compatibility only. Please see Theory Section 4.4 for more details. GWT This parameter is used to specify which method is used to calculate the GreenwoodTripp parameters will be used. The options are: RMS Surface Height (Deprecated) Summit Height Correction (Recommended) It is recommended to use the Summit Height Correction method as the “RMS Surface Height (Deprecated)” is provided for backwards compatibility only. Please see Theory Section 4.4 for more details. Average Flow Model This parameter is used to specify whether or not the AFM will be used. The options are: Patir and Cheng Disabled The “Patir and Cheng” model accounts for surface roughnesses in the solution of the fluid dynamics. Please see Theory Section 4.2.1for more details. Mobility Map The mobility map can be Short Bearing or Finite Bearing. Please see Theory Section 4.1.1 for more details. Flow Summation This parameter is used to specify which method is used for adding the pressure (QP) and velocity (QT) induced flow when calculating the instantaneous oil film flow rate for each journal bearing. Currently only one option exists and this is Pressure (Qp only). Please see Theory Section 4.1.2.3 for more details.

385

HELP Solution Shear This argument is used to denote whether the film extent assumed for the calculation of the oil film shear power loss for each journal bearing is  (PI) or 2 (2PI). Please see Theory Section 4.1.3 for more details. Lubricant The oil types defined using the Lubrication Definitions panel selected from the main panel are listed. The user is required to select one of the oil types by clicking on the appropriate name. If only one oil is contained in the list it will automatically be selected. Supply Temperature This defines the inlet temperature of the oil to the journal bearings. Supply Pressure (Gauge) This defines the feed pressure of the oil to the main journal bearings. The parameters that control the oil film calculations are defined as follows: Oil Temperature Derivation There are four available options to calculate the oil temperature calculation: Constant; Thermal Balance; Thermal Balance, Axial; Use static Temps. Constant If the Oil Temperature Derivation is set to Constant and the solution type is either Determinate or Indeterminate it is necessary to define an oil temperature at each journal bearing by clicking on the Define button adjacent to Bearing Oil Temps. This will open the Bearing Oil Temperatures Panel as shown in Figure B-166.

Figure B-166 Bearing Oil Temperatures Panel The temperatures are defined for each of the Main and Pin journal bearings using the option menu Bearing Type and editing the values in the table. If a solution has already been performed then the temperature values will default to those used for the last loadcase stored in the .EDSF file. If this is the first solution then the values will default to the value of the Inlet Temperature defined using the Evaluate Solution Panel shown in Figure B-165.

386

HELP Solution Thermal Balance and Thermal Balance, Axial This option allows the user to switch on the thermal balance calculation for each of the journal bearings. This is an iterative procedure that evaluates the temperature rise in the oil from the calculated power loss and oil flow. The solution continues until the temperature of the oil for each bearing is within a user-specified tolerance. For more details please see Section 2.6.4 and Theory Section 4.6. Use Static Temps If the Oil Temperature Derivation is set to Use Static Temps and the solution type is Dynamic the oil temperature at each journal bearing may be defined in one of two ways. The user can define an oil temperature at each journal bearing by clicking on the Define button adjacent to Bearing Oil Temps and enter the temperatures as shown in Figure B-166. Alternatively the temperatures may be evaluated by a thermal balance solution from the quasi-static solution that always precedes the dynamic solution. This is achieved by switching Use Static Temps on. Initial Temp This is used to specify a starting temperature for each of the journal bearings for the oil for the thermal balance solution. By default, this parameter will initially take the value defined by Inlet Temperature. This starting temperature is only used for the first journal bearing of the statically-determinate solution and for subsequent bearings the iterated temperature of the previous bearing is used as the starting temperature. For subsequent solution methods the temperatures of the lower level solution method are used as the starting temperature for each bearing Temp Tolerance This parameter is used to specify the temperature tolerance, T, for the thermal balance iteration. The solution for a given cycle has converged providing the equation,

Tn1  Tn   T is satisfied for all journal bearings. The default value is 0.25 degC. For the ‘Thermal Balance’ and ‘Thermal Balance, Axial’ option it is recommend to use a lower value of between 0.05 and 0.1 degC. Relaxation Factor This parameter is used to control the temperature that is used on the next thermal balance iteration step. The relaxation factor for each journal bearing is defined as:

Tn  Tn - 1 + F .T - Tn - 1 where F

= Relaxation factor

387

HELP Solution Tn

= Journal oil bearing temperature for next step.

Tn-1

= Journal oil bearing temperature at previous step

T

= Journal oil bearing temperature calculated from power loss and oil flow rate

The default value is 0.4.

4.4.5

Performing the solution

Once the Evaluate Solution Panel has been completed, the ENGDYN solution can be performed. Two solution modes are offered; Direct or Batch. If Direct is selected the solution is run directly from the Graphical User Interface environment and the GUI will remain locked until the solution is finished. If on the other hand Batch is selected the Batch Control Panel as shown in Figure B-167 is displayed allowing the user to batch queue the solution which then allows continued use of the GUI.

Figure B-167 Batch Control Panel The Batch Command field controls how the Batch Script is to be executed and by default is set to the character string as defined by the environment variable BATCH_COMMAND. By default the batch command is “isisq submit”. The environment variable BATCH_HOST can also be set and if present, the chosen host will be added to the Batch Command line. The Batch Command field may be edited by the user as required. If it is deleted altogether, the job will be run immediately on the same machine that the graphical user interface is running on. The Batch Script contains the command to run the solver and may be modified as required. Additional lines may also be added to the script.

388

HELP Solution If Cancel is selected either from the Batch Control Panel or from the Query Panel an .EDSO command file will be generated but the solver will not be executed.

4.4.6

Running a solution from the command line

An .EDSO file command will be produced by ENGDYN if a solution is run from the GUI,the Cancel button is pressed from the Batch Control Panel or “File->Generate” or “File->Generate As” are selected.. The content of the .EDSO command file is described in Section Solver Commands. The .EDSO command file can be run from the command line by opening a command prompt in windows or linux and typing: engdyn –solve example.EDSO Other options available are: -V -32 -debug= -np=

Specify a version number to run Run in 32-bit mode on 64-bit processors Set debug message level Set number of processors

For example, typing: engdyn –V 2012.2 –solve -32 –debug=4 –np=2 example.EDSO will run the ENGDYN command file “example.EDSO” using a 32 bit executable of ENGDYN version 2012.2 with the debug message level set to four where ENGDYN is run over 2 processors when possible. Similar information can be accessed by typing engdyn –help in a command prompt. The data within the .EDSO command file can be hand edited in a text editor. For possible inputs please see Section Solver Commands. This allows the user to make changes to the model, such as different oil temperatures and pressures, without accessing the ENGDYN GUI. Furthermore, by utilizing external optimization and design software which would automate the updating of the .EDSO command file, it is possible to optimize the model in relation to some specified goal. The other command files described in Section Solver Commands allow a wider range of model inputs to be updated by hand or by an automated process.

1

Ricardo Software, April 2009, ‘PISDYN Manual – Documentation / User’s Manual Version 5.1’

2

Ricardo Software Manual RD.06/333901, October 2006, ‘SDFBROWSER – Documentation / User’s Manual Version 2.0’

389

HELP Post-Processing

5 Post-Processing 5.1 General The Graphical User Interface allows simulation results to be plotted, animated and exported to an ASCII file. The interface has a built in library of pre-defined plots that enables the user to rapidly view the results of the analyses performed. The plots can either be printed or saved to file. Results can also be exported to allow further data manipulation by the user not currently supported by ENGDYN. In addition, the displacement and velocity results can be animated interactively in both the time and frequency domains.

5.2 Selecting Loadcases Loadcases for which results are to be post-processed are selected using the Select Loadcases Panel which is opened by clicking on the Select Loadcases button from the Main Panel. This is shown in Figure B-168.

Figure B-168 Select Loadcases Panel Load cases are stored by solution type and engine load condition. Solution Type and Loading act as filters enabling load cases of a particular solution type and engine loading to be listed in the lower part of the panel for selection. The list of displayed load cases will be updated depending on the selection. The user can then select from this list, by highlighting with the mouse, those load cases they wish to post-process. The solution parameters used for each load case displayed for information only are grouped under 3 tabs General, Lubricant and Tolerances. The figure shows the General tab selected.

391

HELP Post-Processing Solution Type This lists all those load cases of a given solution type, which can either be StaticallyDeterminate, Statically-Indeterminate or Dynamic. Only the solution types which have been previously simulated and for which results are available will be selectable. Loading This lists all those load cases of a given engine loading condition, which can either be Inertia Only, No Load, Part Load or Full Load. Only the engine load conditions which have been previously simulated and for which results are available will be selectable. Select All This button can be used to select all the displayed load cases. Deselect All This button can be used to deselect all the displayed load cases.

5.3 Selecting Crank Train Modes Free-free mode shapes of the crank train that are to be plotted are selected using the Select Modes Panel shown in Figure B-169 which is opened by clicking on the Select Modes button from the Main Panel. The free-free modes are calculated from the reduced model once the model has been assembled as described in 3.4.

Figure B-169 Select Modes Panel The user can select from the list of displayed modal frequencies, by highlighting with the mouse, those mode shapes they wish to plot. A description can be added in the right hand column, which is then displayed on any plot containing that mode.

392

HELP Post-Processing 5.4 Plotting EHL Results

Figure B-170 Lubrication Mesh Plotting Panel Edit

Edits the created plots to redefine

Delete

Deletes selected plots from the list

Delete All

Deletes all the plots in the list

Add Plot 2D/3D

Adds plots

Page

Allows user to control the page plots are created.

Plot

Allows user to control the plots Loads saved plot template from file Saves created plots as a template Starts RPLOT

Plot Panel produces contour or XYZ type plots and ASCII format data tables of various quantities on the bearing lubrication mesh. Some plots are available at selected crank angles, and some are available as one plot per cycle. Crank angles are selected with an interactive menu. Data is interpolated to the selected crank angles. The plots are superimposed on a scale drawing of the bearing. The bearing drawings and contour plots are presented as unwrapped two-dimensional images.

393

HELP Post-Processing 5.4.1

Plot Definition Panel

Figure B-171 Plot Definition Panel The Plot Definition Panel contains controls for the plots to be created. All the plots, which are either Contour or XYZ Type, are available on Lubrication Mesh. The plot Set is the available lubrication mesh sets, which means Data sets that can be plotted on lubrication mesh. The available Data sets are listed in the table below. Model: Lubrication Mesh Location: mainBearing:ID_<> bigEndBearing:ID_<> smallEndBearing:ID_<> Data: Asperity Pressure Asperity Shear Stress Axial Flow Rate Axial Shear Stress Dynamic Viscosity Hydrodynamic Pressure Pressure (Hydrodynamic Pressure + Asperity Pressure) Oil Film Thickness Radial Deformation Tangential Shear Stress Tangential Flow Rate Temperature Void Fraction

394

HELP Post-Processing Type: Contour (See Figure B-178) XYZ (See Figure B-179) Compute: None Mean RMS Maxima Minima Angle: Define (See Figure B-172) where Mean Value of a Variable: Mean value of a variable is the average value from the start of the engine cycle. T

Var( x, y ) 

Var( x, y, )dt 0

T

where T = the period of one engine cycle. RMS Value of a Variable: 2

 Var( x, y, ) d

VarRMS ( x, y ) 

2

1

 2  1

where 1 = the crank angle at the beginning of the cycle, and the cycle.

 2 = the crank angle at the end of

Maxima of a Variable: Maxima of a variable is the plot of maximum value of each mesh node over the entire engine cycle. Minima of a Variable: Minima of a variable is the plot of minimum value of each mesh node over the entire engine cycle.

395


Crank Angle Panel

Figure B-172 Crank Angle Panel The crank angles that the plots are created can be selected by four possible options Using Interval, Defined List, Angle of Minimum Value and Angle of Maximum Value. In Using Interval option user inputs Start, End and Interval of the crank angles in which plots are created. Defined List allows user to specify the crank angles the plots are desired. It is also possible to plot the crank angle at which Maximum or Minimum value of the any variable appears for the given interval (Start and End).

5.4.3

Page Attributes Panel

Figure B-173 Page Attributes Panel

396


Plot Attributes Panel

Figure B-174 Plot Attributes Panel Misc Tab

Figure B-175 Plot Attributes Panel Axes Tab

397


5.4.5

Automatic Generation of EHL Plots

On completion of a solution for a model with EHL bearings, plotting information will automatically be added to the model EDSF file. This information is held in an SDF file in ..\Ricardo\2013.1\Products\ENGDYN\Config\xengdyn\Templates. Two pages for each EHL bearing are included; one showing the upper and lower bearing lubrication mesh and the second showing the journal lubrication mesh. There are two ways in which to view the results through the GUI or using the “–plot” switch when calling ENGDYN from the command line. 5.4.5.1 Generation using GUI

398

HELP Post-Processing After a model has run, open the model in the GUI and select “Plot EHL Results”. The lubrication mesh plotting panel will already be populated with results to display for the upper and lower beaing and the journal for each EHL bearing. Press the “Start R-Plot” button and the results will be shown in R-Plot. The default results displayed are shown in Figure B-176 and Figure B-177 for one EHL bearing.

Figure B-176 Default EHL plot for upper and lower bearing shells

399


Figure B-177 Default EHL plot for journal

5.4.5.2 Generation using command line After a model has run the plotting information is included in the file and therefore the plot can be automatically ouput using the ENGDYN “plot” switch. The command is:

400

HELP Post-Processing engdyn –plot example.EDSF Other options available are: -V Specify a version number to run -output Specify the name of the output .rp file -loadcase Set loadcase number to output results for if not included results for all loadcase will be output For example, typing: engdyn –plot –output results.rp –loadcase 2 example.EDSF will output the bearing EHL results to a file called results.rp for loadcase 2. The .rp file can then be opened in R-Plot by typing: rplot results.rp 5.4.6

Plot Examples

The following set of plots was made from the EHL Bearing Analysis tutorial model.

Figure B-178 Main Bearing 1 Pressure Angle of Maximum Contour Plot

401


Figure B-179 Main Bearing 1 Pressure Angle of Maximum XYZ Plot

5.5 Plotting Results The plotting of analysis results is controlled by the Plot Results Panel which is be opened by clicking on the Plot Results button on the Main Panel and is shown in Figure B-180. For a given model, plots of a given type can be generated for a subset of the model. The selection of the subset and plot type is controlled with the Subset, Plot, Results and Relative To lists. It is not possible to generate results for every combination, so only those combinations that are valid will be selectable.

402


Figure B-180 Plot Results Panel

5.5.1

Selection

Model This is used to select the ENGDYN model for which the results are applicable. The plot options available in the lower half of the panel will be updated depending on which Model is chosen. The possible model types are Crankshaft, Journal Bearing, Connecting Rod, Piston and Cylinder Block. Subset The list on the left of the Plot Results Panel shown in Figure B-180 lists the individual components of the Model that can be selected for which results are required. In the example shown in Figure B-181 Nodes is selected. Other components from the list can be highlighted by selecting them with the mouse. By clicking on the Select All button, all components of that type will be selected, and will be shown in red in the main drawing area. Alternatively, individual components may be either selected, or de-selected using the mouse on the relevant component shown in the main drawing, as shown in Figure B-181. De-selected components will be shown in green. All selected components may be de-selected by clicking the Deselect All button. Non-selectable items will be shown in grey. The selected items are displayed in the message area of the Main Panel.

403


Figure B-181 Selection for Plotting Plot This lists the available types of plot for a given Model and Subset. The required plot is selected using the mouse. The possible types are listed in Table B-11.

404

HELP Post-Processing Plot Type History Spectra Orders Campbell Mode Orbit Polar Extent

Description Time history plot of dependant variable against crank angle at a particular engine speed. Harmonic content of dependant variable plotted against frequency at a particular engine speed. Harmonic content of dependant variable plotted against engine speed. Harmonic content of dependant variable plotted against engine speed and frequency. Mode shape plotted as a deformed shape in 3 views against its undeformed shape. Journal and bearing load orbits. Polar plots of y and z and y and z data. The extent of the positive oil film pressure distribution plotted against crank angle and bearing angle. Table B-11 Plot Types

Results This lists the results type (or dependant variable) that can be plotted for a given Model, Subset and Plot. The required results type is selected using the mouse. The possible results types are listed in Table B-12. Results Type Displacement Velocity Applied Forces External Forces Internal Forces Bearing Load Oil Film Pressure

Description Displacements. Velocities. Forces and moments applied to the crankshaft or cylinder block. All forces applied to the crankshaft including both applied forces (as described above) and centrifugal and gyroscopic forces. Internal shear forces and bending moments acting at each crankshaft element. Bearing forces and moments. Bearing oil film pressures. Table B-12 Results Types

Relative To This lists the axes against which the results are to be plotted and is only relevant when the selected Model is Crankshaft. This can be selected using the mouse either as Fixed, Rotating, Nearest Bearing or Crankshaft Node. Fixed is relative to ground, whilst Rotating is relative to the crankshaft and rotates with it. Nearest Bearing allows crankshaft displacement results to be plotted relative to the displacement of the nearest main journal bearing. This effectively means the results are relative to a ‘fixed’ axis that moves with the bearing. Crankshaft Node allows crankshaft displacements at selected nodes to be plotted relative to another crankshaft node selected by the user. For example this enables the relative twist between the crank nose and flywheel to be plotted. If this option is selected the Plot Query Panel shown in Figure B-182 is displayed when user clicks on the Apply button. A further node is selected using the mouse which will then be displayed in Magenta. Clicking the Apply button again will display the plot.

405


Figure B-182 Plot Query Panel

5.5.2

Viewing

Figure B-183 Graph Panel

406

HELP Post-Processing Once the required plot has been selected for a particular Model and Subset, as described in 5.5.1, the plots are generated by clicking on the Apply button. They are displayed in RPLOT as shown in Figure B-183. The example shown shows an oil film extent plot for a big end bearing of a single cylinder engine at 2000 rev/min Full Load. One plot is displayed per page. At the top of the panel, the page number and number of pages is displayed. The user can navigate between pages either by using the keyboard with the Page Up, Page Down, Home and End keys or by navigate buttons the bottom of RPlot page. RPlot operations can be found in RPlot help system which can be accessed from Help menu as shown in Figure B-184.

Figure B-184 RPlot Help Menu

5.6 Exporting Results For a given model, results can be exported from the .EDSF file to an ASCII file to allow further data manipulation by the user not currently supported by ENGDYN using the Plot Results Panel. The data is written using the Export Results Panel shown in Figure B-185 and may be written either in Experimental format as used by RPlot, in Block format or in Column format. The Export Results Panel is opened by clicking on the Export Results button on the Main Panel.

407


Figure B-185 Export Results Panel Model This is used to select the ENGDYN model for which the results are applicable. The export options available in the lower half of the panel will be updated depending on which Model is chosen. The possible model types are Crankshaft, Journal Bearing and Cylinder Block. Dataset This list is used to select the name of the dataset to be written. Table B-13, Table B-14 and Table B-15 list those datasets that can currently be written together with their units, the data type (vector, scalar or scalar array) and the independent variable against which the data varies. For the vector arrays in which both translational and rotational data are stored two units are defined. A full description of the selected dataset is displayed in the Dataset Definition field. This table also indicates those datasets on which a Fast Fourier Transform (FFT) can be performed. Dataset Name

Units

Data Type

Independent Variable

FFT

JOURNAL_BEARING_ECCENTRICITIES

m, rad

Vector

Crank angle

Yes

JOURNAL_BEARING_LOADS

N, N.m

Vector

Crank angle

Yes

JOURNAL_BEARING_SPECIFIC_LOADS

N/m2

Scalar

Crank angle

Yes

JOURNAL_BEARING_MAXIMUM_OIL_PRESSURE

N/m2

Scalar

Crank angle

Yes

JOURNAL_BEARING_MAXIMUM_OIL_FILM_PRESSURE_LOCATION

deg

Crank angle

No

JOURNAL_BEARING_MINIMUM_CLEARANCE

m

Crank angle

Yes

JOURNAL_BEARING_MINIMUM_CLEARANCE_LOCATIO

deg

Crank angle

No

408

Scalar

HELP Post-Processing JOURNAL_BEARING_MINIMUM_VOID_FRACTION

-

JOURNAL_BEARING_MINIMUM_VOID_FRACTION_LOCATION

deg

JOURNAL_BEARING_LEAKAGE_FLOW_RATE

m3/s

Scalar

Crank Angle

JOURNAL_BEARING_SUPPLY_OIL_FLOW_RATE

m3/s

Scalar

Crank Angle

JOURNAL_BEARING_OIL_PRESSURES

N/m2

Scalar Array

Crank angle

No

JOURNAL_BEARING_OIL_THICKNESSES

m

Scalar Array

Crank angle

No

JOURNAL_BEARING_OIL_TEMPERATURE

C

Scalar

Engine speed

No

JOURNAL_BEARING_OIL_VISCOSITY

Pa.s

Scalar

Engine speed

No

JOURNAL_BEARING_OIL_FLOWRATE

m3/s

Scalar

Engine speed

No

JOURNAL_BEARING_OIL_POWER_LOSS

W

Scalar

Engine speed

No

JOURNAL_BEARING_PRESS_VELOCITY

N/m.s

Scalar

Engine speed

No

JOURNAL_BEARING_CYCLIC_HYDRODYNAMIC_POWER_LOSS

W

Scalar

Engine Speed

No

JOURNAL_BEARING_CYCLIC_BOUNDARY_LUBRICATION_POWER_LOSS

W

Scalar

Engine Speed

No

JOURNAL_BEARING_MEAN_FRICTION_TORQUE

N.m

Scalar

Engine Speed

No

JOURNAL_BEARING_FMEP

N/m2

Scalar

Engine Speed

No

Scalar

Crank Angle

Yes

Crank Angle

No

Table B-13 ENGDYN Journal Bearing Results Data Dataset Name

Units

Data Type


FFT

CRANKSHAFT_DISPLACEMENTS

m, rad

Vector

Crank angle

Yes

CRANKSHAFT_VELOCITIES

m/s, rad/s

Vector

Crank angle

Yes

CRANKSHAFT_APPLIED_FORCES

N, N.m

Vector

Crank angle

Yes

CRANKSHAFT_EXTERNAL_FORCES

N, N.m

Vector

Crank angle

Yes

CYLINDER_PRESSURE

N/m2

Scalar

Crank angle

Yes

Table B-14 ENGDYN Crankshaft Results Data Dataset Name

Units

Data Type


FFT

CYLINDER_BLOCK_DISPLACEMENTS

m, rad

Vector

Crank angle

Yes

CYLINDER_BLOCK_VELOCITIES

m/s, rad/s

Vector

Crank angle

Yes

409

HELP Post-Processing CYLINDER_BLOCK_APPLIED_FORCES

N, N.m

Vector

Crank angle

Yes

CYLINDER_PRESSURE

N/m2

Scalar

Crank angle

Yes

Table B-15 ENGDYN Cylinder Block Results Data

Dataset Definition This field displays a full description of the selected dataset. Subset The list on the left of the Export Results Panel shown in Figure B-185 lists the components of the Model that can be selected for which results are required. The method of selection is similar to that described in 5.4 for plotting results. In the example shown in Figure B-186 Nodes is selected. Other components from the list can be highlighted by selecting them with the mouse. By clicking on the Select All button, all components of that type will be selected, and will be shown in red in the main drawing area. Alternatively, individual components may be either selected, or de-selected using the mouse on the relevant component shown in the main drawing, as shown in Figure B-181. De-selected components will be shown in green. All selected components may be de-selected by clicking the Deselect All button. Non-selectable items will be shown in grey. The selected items are displayed in the message area of the Main Panel.

410


Figure B-186 Selection for Exporting Filename This defines the name of the output file. A suffix dependant on the independent variable is added to the name. If the independent variable is crank angle then a file is created for each selected load case with a name of the form NAME.L001.EDRES where the load case number is written as part of the name. If the independent variable is engine speed then a single file is created with a name of the form NAME.FL.EDRES where the engine load type is written as FL, PL, NL and IO as part of the name. FL, PL, NL and IO correspond to Full Load, Part Load, No Load and Inertia Only respectively. The user may alternatively select a file by clicking on the browse button to open the Export File Panel. Format This defines the format of the data. The data may either be written in Exp(RPlot) format as used by Rplot, in Block format or in Column format. An example of an Exp(RPlot) formatted file is as follows: EXP: Engine_speed : ONED COUNT=

3

411

HELP Post-Processing 3.33333E+01 5.00000E+01 7.83333E+01 Oil_film_power_loss_@_Main_Journal_Bearing_1 : ONED COUNT= 3 9.55107E+01 1.87223E+02 4.74661E+02

The same data in Block format is as follows: ;Written by ENGDYN Version 2.1.0 on 1-Jan-2000 at 17:21:57 ;Dataset JOURNAL_BEARING_OIL_POWER_LOSS BLOCK NUMBER=001 NAME=“ENGINE SPEED” UNITS=“rev/s” 3.33333E+01 5.00000E+01 7.83333E+01 BLOCK NUMBER=002 NAME=“OIL POWER LOSS” UNITS=“W” LEGEND=“Main Bearing 1” 9.55107E+01 1.87223E+02 4.74661E+02

The same data in Column format is as follows: ‘Engine_speed’ ‘Oil_film_power_loss_@_Main_Journal_Bearing_1’ ‘rev/s’ ‘W’ 3.33333E+01 9.55107E+01 5.00000E+01 1.87223E+02 7.83333E+01 4.74661E+02

If the data is three-dimensional, in the case of a scalar array, the experimental format is the only format that is recognised. Relative To This is used select whether the crankshaft results are written with respect to the Fixed or Rotating axes. Rotating is relative to the crankshaft and rotates with it, whilst Fixed is relative to ground. FFT This is used to perform a Fast Fourier Transform (FFT) on the data of the selected dataset. If set to None then no FFT is performed and the data is written in the time domain. If Real/Imag is selected the real and imaginary coefficients at each engine order frequency up to and including the frequency corresponding to the maximum engine order as specified by Max Order will be written to the file. If Mag/Phase is selected the coefficients are written as magnitude and phase. Interval This is used to specify a crank angle interval at which the data is written for those datasets with crank angle as their independent variable. The interval defaults to the interval at which the data is stored in the .EDSF file and which was defined at solution. Max Order This is used to specify the maximum engine order to which frequency domain data is written when FFT is Real/Imag or Mag/Phase Mag/Phase. Directions These are used to select particular directions of a vector array.

412

HELP Post-Processing 5.7 Backsubstitution For those models that are compliant or dynamic model and have been derived using the finite element method it is possible to perform data recovery from the reduced model (used by ENGDYN) to the complete finite element model. This uses a data recovery matrix derived during the matrix reduction procedure and which is stored in the SFE file.

5.8 Animate Results

Figure B-187 Animation Example The results of the simulation can be animated by clicking on the Animate Results button on the main panel. Two panels will be opened, the Animation Control Panel and the Animation Results Panel, as shown in Figure B-187. The displaced or vibrating shapes of the reduced

413

HELP Post-Processing models of the crank train and cylinder block may be animated either in the time or frequency domain with displacements or velocities shown as a colour contour. Initially, the display area will show in grey the reduced models of the crank train and cylinder block in their undeformed state. Once the animation is initiated using the button on the Animation Control Panel titles and current result attributes will be displayed in the top left corner of the display area together with a contour bar at the bottom. This is shown in Figure B-187.

5.8.1

Animation Control

The Animation Control Panel shown in Figure B-188 is used to control the animation.

Figure B-188 Animation Control Panel The control buttons have the following functions: Start recording Start animation Stop animation Go to Previous Loadcase Step forward by one frame Step backward by one frame Go to next loadcase

The slider bar is used to select a particular angle in the simulation cycle by selecting with the left mouse button and then dragging.

414


Animation Results

The Animation Results Panel shown in Figure B-192 is used to control which results are to be animated. It has four tabs, Contour, Deformation, Domain and Options: 5.8.2.1 Contour Tab This tab consists of seven tabs, for the bearing, conrod, crankshaft, cylinder block, piston, acoustic field and boundary mesh. Each tab has the same layout, but the contents of the Value and Direction menu will vary.

Figure B-189 Animation Results Contour Tab Value This defines the results and their vector component that are to be displayed as a colour contour. Currently None, Cavitation, Pressure, Film Pressure, Contact Pressure, Axial Shear Stress, Tangential Shear Stress, Contact Shear, Axial Flow, Tangential Flow, Radial Deformation and Film Thickness may be presented for the lubrication mesh; None, Displacement and Velocity may be presented for the conrod, crankshaft, cylinder block and piston; None, Pressure and Pressure [dbL] may be presented for the acoustic field and None, Intensity and Intensity [dbL] may be presented for the Boundary Mesh. Direction Depending on the model type, this may contain X, Y, Z, PhiX, PhiY, PhiZ, All or Scalar. If All is selected the vector sum of the x, y and z components of the results vector is calculated and displayed. Min Cont This is used to define the minimum value of the colour contour of the currently displayed results. The value can be adjusted either by using the slider bar or by typing the value directly. By default the minimum value of the data will be displayed and used. Max Cont This is used to define the maximum value of the colour contour of the currently displayed results. The value can be adjusted either by using the slider bar or by typing the value directly. By default the maximum value of the data will be displayed and used.

5.8.2.2 Deformation Tab This tab consists of five tabs, for the bearing, conrod, crankshaft, cylinder block and piston. Each tab has the same layout, but the contents of the Value and Direction menu will vary.

415


Figure B-190 Animation Results Panel Deformation Tab Value This defines the results and their vector component that are to be used to deform the models. Currently, None, Pressure, Film Pressure, Contact Pressure and Radial Deformation may be selected for the bearing and None or Displacement may be selected for the remaining models. Direction Depending on the model type, this may contain X, Y, Z, and All, or Scalar (+ve) and Scalar (ve). If All is selected the deformed shape is the vector sum of the x, y and z components. If Scalar (-ve) is selected then the model will be deformed in the opposite direction to reality. Factor This is used to define the factor by which the deformed shapes are magnified. The factor may either be selected by using the slider bar or by typing a value. 5.8.2.3 Domain Tab

Figure B-191 Animation Results Domain Tab Domain This is used to select whether the results are to be animated in the Time or Frequency domain. Note that the bearing may only be animated in the time domain. Order This is used to select the order of vibration to be animated and may be defined either by typing the harmonic order directly or by using slider bar. Cycling Is used to select whether or not to automatically cycle through each order of vibration. No of Cycles

416

HELP Post-Processing This defines the number of cycles of animation at each loadcase condition when Cycling is selected.

5.8.2.4 Options Tab

Figure B-192 Animation Results Panel Options Tab Relative To This is used to select whether the deformation of the crank train is relative to the rotating or fixed axes. Unwrap Bearing This applies to the lubrication mesh only. If selected, the mesh will be drawn as a flat surface rather than as a cylinder. Fixed Crank This is used to fix the crank train model from rotating. If this is switched on then the results are viewed as if the user were rotating with the crankshaft. Eye Locate If this option is selected the undeformed shapes of the reduced crank train and cylinder block models are displayed in grey. Contour Bar This is used to select whether the contour bar is Horizontal or Vertical on the main panel. Number of Bands This is used to control the number bands of the contour. Animation Speed This slider bar is used to control the animation speed. It is necessary to press the play button to resume animating after changing the speed.

5.8.3

Image Panel

The Image Panel (Plot Bitmap from File Menu) shown in Figure B-193 is used to define how and where individual frames are recorded when the record button is selected.

417


Figure B-193 Image Panel Directory This is used to select a directory in which the files will be saved. Prefix This is used to define a prefix that will be added to each image file name. Not that each image will have a name of the form _n_m, where n is a 3 digit loadcase number and m is a 3 digit crank angle value. Format This is used to define the format of the image. Options are jpeg, bmp, png, rgb and mpeg (movie). Quality This is only applicable to jpeg and mpeg, and has the options High, Medium and Low. It determines how detailed the image is (note that a high quality image takes longer to produce and takes up more disk space than a low quality image). Invert B/W Change all instances of black to white, and all instances of white to black. Frame Rate This only applies to mpeg movies and determines how many frames per second the movie contains. Options are 1, 2, 3, 5, 6, 10, 15 and 30. Size This determines the size (in pixels) of the saved image or movie. Options are pal, ntsc, Screen and Custom. This affects the numbers displayed in the Resolution fields. Resolution These two fields determine the respective width and height (in pixels) of the saved image or movie.

418

HELP Crank Shaft Stress Analysis

6 Crank Shaft Stress Analysis 6.1 Overview Operating loads calculated by ENGDYN can be used to evaluate the stresses and fatigue safety factors at the critical locations of the crankshaft using two methods developed by Ricardo1. The first method which is referred to as the Classical Method considers simple bending theory in which the stress is considered as being onedimensional with lateral stresses being neglected, whilst the second method uses the finite element method in which a quasi-static analysis is performed. The Graphical User Interface provides a single environment to perform both methods in which the majority of the data input is common to both. The user is required to define the geometry of the fillets, the crankshaft webs and the oil hole breakouts together with the base material properties of the crankshaft (although these may be obtained from a database of typical crankshaft steels and Spheroidal Graphite (SG) irons.) The elevated strengths of the material, due to surface treatment processes, may either in the absence of measured values be calculated using ENGDYN as described in Appendix 3 or supplied by the user. Stress concentration factors (SCF) at the crankshaft fillets and oil hole breakouts are calculated based on work by Lowell2, Arai3, Peterson4 and Oldberg5. Fatigue notch factors (Kf) are calculated from the SCF values based on work by Lowell5 and modified according to Ricardo experience. Alternatively, the SCF and Kf values can either be defined by the user or calculated using alternative algorithms supplied by the user using environment variables. The derivation of SCF and Kf values is described in Appendices C and D respectively. Stresses and fatigue safety factors results may be plotted as history plots against engine speed or crank angle or as a Goodman Diagram. The finite element method uses the Ricardo program FEARCE to apply and to combine the unit loads as part of the method. This program is executed by ENGDYN within the interface, but more experienced users may wish to execute these programs outside the interface.

6.1.1

Stress Analysis Using Classical Methods

Stresses and fatigue safety factors are calculated at the oil hole breakout(s) on each journal and at the journal fillets corresponding to sections through the journal and web overlap. Stresses and fatigue safety factors are only calculated at those locations which have geometry defined. In order to calculate the journal fillet stresses on section through the web overlap it is necessary to define the geometry of all the lands and fillets adjacent to the web including the web itself. The calculated stresses are appended to the .EDSF standard data file. The stresses corresponding to the maximum and minimum stress conditions at the lowest calculated

419

HELP Crank Shaft Stress Analysis factor of safety for each location together with the material properties, SCF and FNF values and section moduli are written to a file with the suffix .EDSTRESS. Bending and torsion stresses are calculated for the complete engine cycle for each selected loadcase at the crank angle interval defined during solution to calculate the operating loads. For quasi-static determinate and indeterminate solutions the bending stresses are due to quasi-static bending only. For these solution types a stress due to torsional vibration may be calculated and combined with the quasi-static bending stress. These may be combined either by assuming the torsional behaviour is in phase with the quasi-static bending, that is the mean torque and torque amplitude vary at each crank angle, or by calculating a maximum and minimum torque and applying that at every crank angle. In order to account for the uncertainty concerning the phasing of the vibratory torque with respect to the quasi-static loads the latter method is preferred for this type of solution. For dynamic solutions the bending stresses are to vibratory bending which includes the quasi-static content. For this solution type the stress due to the torsional vibration is always combined with the bending stress. As for the quasi-static solutions the stresses may be combined using one of the two methods described above. However a dynamic solution removes the uncertainty concerning the phasing between the torsional and bending behaviour it is valid to use the former method.

6.1.2

Quasi-Static Analysis Using the Finite Element Method

6.1.2.1 General Description The approach adopted for calculating the stresses and fatigue safety factors using the finite element method is a unit load method, in which loads are applied to the model, displacements and stresses are calculated using an appropriate finite-element package and the loads are factored and combined during post-processing. The method is described by the flowchart of Figure B-194. A torsional load is applied at the nose of the crankshaft together with unit loads at each journal bearing. An optional centrifugal load and unit axial load at the nose of the crankshaft is also applied. The stresses corresponding to the maximum and minimum stress conditions at the lowest calculated factor of safety for each location together with the material properties and SCF and FNF values are written to a file with the suffix .EDSTRESS. 6.1.2.2 Finite Element Model 6.1.2.2.1

Geometry and Restraints

The finite-element model can either be of the complete crankshaft or a subset of it. The mesh defining the fillets and oil holes of interest should be of sufficient mesh density to accurately predict the stress at these locations. Typically, unless the oil holes are radial cross-drillings it is difficult to model the oil holes using hexahedral elements. Using tetrahedral elements overcomes this problem but results in a larger model. An example hexahedral model for a four-cylinder in-line engine is shown in Figure B-195

420


FE data or binary file containing crankshaft model and set definitions e.g. IDEAS Universal File

CRANK MODEL

Unit Load Calculations

FE Data File

FE SOLVER

.FRC command file defining unit loads

.FRC File

.SFE file contains model and unit loads. On completion of the FE solution the stresses and displacements are appended using the appropriate translator

Translator

Fatigue Safety Calculations

.FRC command file defining unit load factors, stress correction factors and fatigue safety calculations

.FRC File

.SFE file contains external surfaces of FE model with operating load stresses and fatigue safety factors

FE Viewing Package

Plotted Output

.EDSTRESS file

.EDSTRESS contains summary of results including minimum fatigue safety factors for each set

Figure B-194 Quasi-Static Analysis Using the Finite-Element Method

421


Figure B-195 An Example Finite Element Model for Crankshaft Stress Analysis If a finite element model of the crankshaft was used during model generation as described in 3.3.1 then the model used for the stress analysis must be in the same coordinate system as that model. In addition the material name of the crankshaft and the base material properties (Young’s Modulus, Poisson’s Ratio and Mass density) should also be the same. If centrifugal loading is to be considered as part of the analysis then it is important that the total mass moments M.y, M.z, M.y.x and M.z.x of the stress quality model are the same as the stiffness quality model using during model generation. If the mass moments are not same then an error may be introduced during post-processing in which there is a nett out-of-balance force acting on the model. For this reason it is often prudent to use the same model for both model generation and for the stress analysis. The model should include at each end of the crankshaft a layer of low-stiffness massless elements to ameliorate the local influence of the boundary conditions and torsional loading at the nose of the crankshaft. These layers should be equivalent to about one quarter of the typical bearing shell length and are drawn in red in Figure B-195 and typically the Young’s Modulus of these elements should by a factor of 10 less than for the crankshaft itself. In addition a layer of shell elements should be overlaid on the plane at the front of the front layer of low-stiffness elements as shown in the figure such that the shell elements share the same nodes as the solid elements. The material properties of these elements should be the same as the crankshaft, except for the mass density, which should be set to zero. The nodes in the plane at the rear of the rear layer of low-stiffness elements are fully restrained as indicated in the figure.

422

HELP Crank Shaft Stress Analysis 6.1.2.2.2

Set Definition

A number of elemental face and node sets are required for applying the unit loads and restraint and for calculating the stresses at particular locations. For loading, face sets are required at each main and pin journal bearing using the names MAIN_BEARING and PIN_BEARING respectively. Each set must extend over the area in which the journal oil film acts such that the axial position and extent of the set should correspond to the bearing centre and bearing shell length defined during model generation within ENGDYN. An example is shown in Figure B-196 with the set displayed in red.

Figure B-196 Face Set Definition for Applying Bearing Loads Typically the oil film doesn’t act over the entire journal between the fillets, which usually necessitates a thin layer of elements between the fillet and the start of the set as shown in the figure. If this is not possible, usually because the element aspect ratio is unacceptable, then it is possible to define the set as extending between the two fillets. The names of the sets must follow the same numbering convention as defined during model generation within ENGDYN in which a number is used to indicate the number of the main bearing or cylinder corresponding to the pin bearing. For example, in the case of the four-cylinder in-line example shown in Figure B-195 the required sets would be as follows: MAIN_BEARING_01 MAIN_BEARING_02 MAIN_BEARING_03 MAIN_BEARING_04 MAIN_BEARING_05

423

PIN_BEARING_01 PIN_BEARING_02 PIN_BEARING_03 PIN_BEARING_04

HELP Crank Shaft Stress Analysis The nodes in the plane at the rear of the rear layer of low-stiffness elements need to be included in a node set called RESTRAINT so that the model can be restrained at this location as indicated in Figure B-195. If other external loads have been applied to the crankshaft using either the Force Equation or Force Profile tabs part of the Loading Definition Panel as described in Sections 4.3.2 and 4.3.3 then it is necessary to have additional face sets corresponding to each loaded set. For example, if the set crankNode:ID_belt has been defined (see Section 3.4.8) and loaded then the crankshaft stress model needs to have a face set defined with the same name suitable for applying unit loads. It’s desirable that the centre of this face set corresponds to the node set in the crankshaft reduced model. If they are not aligned then the program accounts for this during the stress analysis to ensure that static equilibrium is maintained. It should be noted that if a node has been loaded with multiple force profiles, by defining multiple sets for a single node using the CRANKSHAFT.SETS file e.g: crankNode:ID_gear crankNode:ID_belt crankNode:ID_chain

1 1 1

and applying a force to each set, the forces will be combined and applied at the first set representing the node for post-processing purposes. Though this means that the force is not applied in the correct location in the stress analysis, the total load and moment will be correct. In addition to the face sets required for applying the loading a node set is required for each location of the crankshaft for which stresses and fatigue safety factors are required. Each set has its own material properties, stress correction factor to account for notch sensitivity and combination factors to account for vibratory loading at that location. Node sets can be defined for the journal fillets, the journal oil hole breakouts, the journals and the facing lands. By default the following naming convention has been adopted although each set name can be redefined using the interface: REAR_MAIN_FILLET_01 LEAD_MAIN_OILHOLE_01 REAR_MAIN_LAND_01 MAIN_JOURNAL_01 In each of the above names the words MAIN, REAR and LEAD can be replaced with PIN, FRONT and TRAIL respectively. The definition of FRONT and REAR depends on the numbering of the main bearings as defining during model generation. If the bearing numbers ascend with x then FRONT is the fillet or land nearest the crank nose. Conversely if the bearing numbers descend with x then FRONT is the fillet or land nearest the flywheel. The definition of LEAD and TRAIL corresponding to leading and trailing oil hole breakouts respectively is defined in Figure B-197.

424

HELP Crank Shaft Stress Analysis Trailing

Leading

Figure B-197 Definition of Leading and Trailing Oil Holes The numbering convention is similar to the loading sets in that the number relates to the main bearing or cylinder (defined during model generation) depending on whether the set is at a main or pin journal. An example of the node sets for a crank pin journal with a cross-drilling oil hole is shown in Figure B-198. In this example, if the main bearing numbers ascend with x then the rear and front fillets sets are shown as blue and violet respectively and the rear and front facing lands sets are shown as yellow and grey respectively. The leading and trailing oil hole breakout sets are shown in view 2 as green and grey respectively, whilst the journal set is shown in red. Only those sets of interest need to be defined.

Figure B-198 Journal Node Set Definitions for Crankshaft Stressing

425

HELP Crank Shaft Stress Analysis The main bearing fillets at the ends of the crankshaft which are not adjacent to the crankshaft web can also be stressed. The naming convention as described above applies, such that for an in-line 4 cylinder engine with 5 main bearings the sets would by default be called FRONT_MAIN_FILLET_01 and REAR_MAIN_FILLET_05. Although there may be more than one fillet at the cranknose, only one fillet can be stressed currently. If the model is a subset of the crankshaft then the model must include at least two main bearings and must extend from the flywheel end of the crankshaft and include the flywheel flange. For example a rear bay model of the model shown in Figure B-195 is shown in Figure B-199. The front most main bearing should extend as far as the front facing land.

Figure B-199 An Example Rear Bay Finite Element Model for Crankshaft Stress Analysis

6.1.2.3 Application of the Quasi-Static Unit Loads ENGDYN applies unit loads at each main and pin journal bearing, together with a unit torsional load applied at the nose of the crankshaft. If centrifugal loads are to be included in the analysis then a centrifugal load is also applied. If loads due to vibratory bending are to be included a unit axial load is applied at the nose of the crankshaft. The bearing loads are applied to the sets as defined in 6.1.2.2 as distributed loads using nodal forces that give a pressure distribution that varies parabolically along the journal and around its circumference with an extent of 180o. View 1 of Figure B-200 shows the pressure distribution due to a force load in the -Y direction, whilst view 2 shows a moment load in the Y direction. The latter is applied using loading in Z applied in opposite directions on each half of the bearing to give a couple. Viewing the nodal forces as opposed to the face pressures will not necessary give a smooth distribution as indicated in the figure since the nodal forces are dependant on element size and whether the elements are linear or parabolic.

426


Figure B-200 Journal Bearing Unit Loads Unit force loads and moment loads are applied in each of 4 orthogonal directions (+Y -Y +Z and -Z) in the ENGDYN co-ordinate system at each of the main and pin journal bearings. A unit torsional load is applied using nodal force loads applied at two nodes on the shell elements at the nose of the crankshaft. This is shown in Figure B-201 for the example shown in Figure B-195.

Figure B-201 Torsional Load

427

HELP Crank Shaft Stress Analysis A body load equivalent to a constant angular velocity about the crankshaft axis is applied if centrifugal loads are to be applied in the analysis. A unit axial load is applied if vibratory bending loads are to be included in the analysis. This load is applied using face pressures to the faces of the solid elements that have the layer of shell elements. For all other sets where additional external loads have been applied during the ENGDYN solution unit force and moment loads are applied in each of 4 orthogonal directions (+Y Y +Z and -Z) For a conventional four-cylinder in-line engine in which durability of the crankshaft is to be assessed using dynamic loading 75 unit loadcases would be applied to the complete model of the crankshaft. These would include 36 unit force loadcases, 36 unit moment loadcases, 1 unit torsional loadcase, 1 centrifugal loadcase and 1 axial loadcase. ENGDYN writes an analysis data file that contains the geometry and restraints described in 6.1.2.2.1 together with the loading as defined above for solution in an appropriate user defined finite-element package. The loading is also written to a FEARCE command file with an .FRC suffix which is used to write the loading to the Ricardo Standard Finite Element File (SFE) and to export the data to the appropriate format. Part of an example .FRC written by ENGDYN in which a MSC/NASTRAN data file is exported is shown in Figure B-202. The figure shows the commands to apply a unit force bearing load of 1000 N.m in the +Y direction, the unit torsional and axial loads and the unit centrifugal load corresponding to an angular velocity of 1000 rev/min. In addition the commands required to restrain the model and to export the data to MSC/Nastran are also shown.

428

HELP Crank Shaft Stress Analysis ; Written by ENGDYN

Version 2.1.2

Run on 12-Feb-2000 at 12:20:15

OPEN FILE='EXAMPLE' STATUS=OLD EDIT DELETE =’Force Loads’ ‘Body Loads’ ‘Face Pressure Loads’

\ \

BEARING

\ \ \ \ \ \ \ \ \ \

SET LOADCASE NAME DESCRIPTION CENTRE AXIS_VECTOR WIDTH FORCE TORQUE ANGLE TYPE

= = = = = = = = = = =

MAIN_BEARING_01 1 MBUL01#STEP01# 'Force Load +Y -0.172050E+00, 0.100000E+01, 0.225000E-01 0.000000E+00, 0.000000E+00, 0.180000E+03 JOURNAL

Main bearing 01' 0.000000E+00, 0.000000E+00 0.000000E+00, 0.000000E+00 0.100000E+04, 0.000000E+00,

0.000000E+00 0.000000E+00

. . . FORCE LOADCASE NAME DESCRIPTION COORDINATE FORCE

= 57 = TORQUE = 'Unit torque load' = -0.258744E+00, 0.140003E-01, -0.236511E-06 = 0.000000E+00, -0.973213E-02, 0.357143E+05

\ \ \ \ \

FORCE LOADCASE NAME DESCRIPTION COORDINATE FORCE

= 57 = TORQUE = 'Unit torque load' = -0.258744E+00, -0.139997E-01, -0.244141E-06 = 0.000000E+00, 0.973213E-02, -0.357143E+05

\ \ \ \ \

SET

SET = SHELLS TYPE = FACE SHAPE

= SQUARE,TRIANGULAR

SET

SET = FRONT

TYPE = NODE SHAPE

= SQUARE,TRIANGULAR

SET

SET = FRONT

TYPE = FACE CONVERT

= NODE

SET

SET = FRONT

TYPE = FACE SUBTRACT = FRONT,SHELLS

PRESSURE SET LOADCASE NAME DESCRIPTION FORCE

= FRONT = 58 = ALOAD = 'Axial Load' = 0.100000E+04

\ \ \ \

BODY LOADCASE NAME DESCRIPTION CENTRE

= 60 = CLOAD = 'Centrifugal Load @ 1000 rev/min' = -0.172050E+00, 0.000000E+00, 0.000000E+00, 0.000000E+00, 0.000000E+00, 0.000000E+00 OMEGA = 0.104720E+03, 0.000000E+00, 0.000000E+00, 0.000000E+00, 0.000000E+00, 0.000000E+00 ACCELERATION = 0.000000E+00, 0.000000E+00, 0.000000E+00

RESTRAINT SET

\ \ \ \ \ \ \

SET = RESTRAINT DIRECTION = X,Y,Z

SET = EXTERNAL TYPE = ELEMENT

SAVE EXPORT TO ANALYSIS UNIT NAME

= = = =

'MSC/NASTRAN' STRESS SI 'EXAMPLE'

\ \ \

Figure B-202 An Example .FPR File for Applying Unit Loading Before any new loads are defined an EDIT command is defined that deletes any loads that may already exist in the SFE. If any temperature loads exist in the SFE then the unit

429

HELP Crank Shaft Stress Analysis load case numbers are forced to start at n+1 where n is the maximum temperature load case number. A SET command is also defined that generates an element set that contains all the solid elements that have a node on the surface. If MSC/Nastran is used to solve for displacement and stress then the stresses will be calculated for this set only. 6.1.2.4 Requirements if the Power Take-Off is at some location along the Crankshaft Typically the power take off (PTO) is at the rear of crankshaft as shown Figure B-195. Alternatively if the PTO is at some other point along the crankshaft, typically via a gear drive, a layer of low-stiffness elements need to be attached at that location and the node set RESTRAINT defined so as to restrain the model. The axial location of this set needs to correspond to the node in the reduced crankshaft model where the mean torque is applied. In this case, unit axial and torsional loads are applied at both ends of the crankshaft and as such it is necessary to have an additional layer of shell elements overlaid at both the front and back of the crankshaft. There are occasions, typically due to the application of an external axial load at the PTO, when it is necessary to have the axial restraint at a different location to the torsional restraint. In this case 2 node sets are required, one called axialRestraint and the other torqueRestraint. The model is restrained in X at axialRestraint, and in Y and Z at torqueRestraint. Low stiffness elements are used in the same way as before at each restrained location. If the node set torqueRestraint is defined at some location along the crankshaft then unit torsional loads are applied at both ends of the crankshaft. Likewise, if the node set axialRestraint is defined at some location along the crankshaft then unit axial loads are also applied at both ends of the crankshaft.

6.1.2.5 Calculating Stresses and Fatigue Safety Factors Once operating loads have been calculated as described in Chapter 5 and a finiteelement solution has been performed to calculate the displacements and stresses due to the unit loading as described in 6.1.2.3 the stresses at a user-specified crank angle interval are calculated by factoring and combining the unit loads throughout the engine cycle. Fatigue safety factors are then calculated from the resultant stress history. . The displacement results are deleted before the loadcases are combined. This reduces the size of the subsequent file and increases the speed of the post-processing. The stresses due to quasi-static loading due to bending and torque are calculated separately as are the stresses due to vibratory loading such that a number of combined loadcases are created at each crank angle. At each crank angle every location along the crankshaft is factored differently according to the notch sensitivity and the vibratory load at that location. Bending and torsion stress correction factors arising from notch sensitivity (as described in Appendix 4) are applied as appropriate depending on whether the load is due to bending or torsion.

430

HELP Crank Shaft Stress Analysis The fatigue safety factors are calculated at each location along the crankshaft using the material strengths appropriate to that location. Safety factors are calculated for both quasi-static loading only and quasi-static and vibratory loading combined. The loadcase factors and combinations are written by ENGDYN to a FEARCE command file with an .FRC suffix. This is used by ENGDYN to calculate the combined stresses and fatigue safety factors, which are written to the Ricardo Standard Finite Element File (SFE) and which then can be exported to an appropriate finite-element package for viewing. The VARIABLE and COMBINE commands are used to combine and factor stresses. Table B-16 lists the load case prefixes that are used to combine and factor the stresses. If any temperature loads are included in the SFE then these loads are included in the load case QSB. Load cases with the prefixes QSB and QST are deleted from the SFE file in order to minimise the size of the file. Load Case Prefix QSB QST QSL MXTV MNTV TVIB MXBV MNBV BVIB QMXT QMNT QMXV QMNV QSLT QSLV

Description Quasi-static bending Quasi-static torque Quasi-static loading Maximum torsional vibration throughout engine cycle. This value is calculated once and does not vary with crank angle. Minimum torsional vibration throughout engine cycle. This value is calculated once and does not vary with crank angle. Torsional vibration Maximum bending vibration throughout engine cycle. This value is calculated once and does not vary with crank angle. Minimum bending vibration throughout engine cycle. This value is calculated once and does not vary with crank angle Bending Vibration Quasi-static loading + maximum torsional vibration Quasi-static loading + minimum torsional vibration Quasi-static loading + maximum vibration Quasi-static loading + minimum vibration Quasi-static loading + torsional vibration Quasi-static loading + torsional and bending vibration

Table B-16 Naming of Combined Stress Loadcases Part of an example .FRC written by ENGDYN in which stresses and fatigue safety factors are calculated for a dynamic solution for a crankpin oil hole and fillet is described as follows. The combinations at 0.0 degree CA for this example in which some of the prefixes of Table B-16 are used are as follows: ;ENGDYN Loadcase 59 @ 4700.0 rev/min FL (Indeterminate solution) ; Vibratory moments and forces calculated from loadcases 18 (dynamic) and 59(static) ; Combinations for crank angle = 0.00 deg VARIABLE NAME SAVE

= BEARINGLOADFACTOR = YES

\ \

431

HELP Crank Shaft Stress Analysis VALUE =

0.22090E+02, 0.47364E-01, 0.34435E-01, 0.23893E-01, 0.40781E-04, 0.14414E-01, 0.44125E-06, 0.19881E-05

0.19800E+02, 0.32647E+02, 0.70548E+01, 0.42450E-01, 0.65983E+01, 0.44551E+02, 0.94911E+01,

0.15990E+00, 0.29263E+00, 0.11375E+00, 0.48106E-01, 0.29125E-01, 0.00000E+00, 0.99380E-07,

VARIABLE NAME = FACTOR_SET001 VALUE = $BEARINGLOADFACTOR, VARIABLE NAME = FACTOR_SET002 VALUE = $BEARINGLOADFACTOR,

0.36584E-02, 0.13397E-02, 0.62735E-03, 0.19381E-04, 0.13559E-02, 0.10792E+02, 0.11974E+02,

\ \ \ \ \ \ \

0.10368E-06 SAVE = YES 0.10368E-06 SAVE = YES

COMBINE NAME = QSB59#CA0001#0.00 DESCRIPTION = ' 4700 rev/min FL Q-S Bending Only - Indet Soln' OLDLOADCASE = 59 1 3 5 8 9 12 13 15 17 19 22 23 26 27 29 31 34 36 38 40 42 43 46 47 50 51 53 56 57 SET = FRONT_PIN_FILLET_01 LEAD_PIN_OILHOLE_01 FACTOR = $FACTOR_SET001 $FACTOR_SET002 NOTCH = $BENDSCF_SET001 $BENDSCF_SET002

\ \ \ \ \ \ \

VARIABLE NAME = FACTOR_SET001 VALUE = -0.10368E-06 SAVE = YES VARIABLE NAME = FACTOR_SET002 VALUE = -0.10368E-06 SAVE = YES COMBINE NAME DESCRIPTION OLDLOADCASE SET FACTOR NOTCH

= = = = = =

QST59#CA0001#0.00 ' 4700 rev/min FL Q-S Torque Only - Indet Soln' 57 FRONT_PIN_FILLET_01 LEAD_PIN_OILHOLE_01 $FACTOR_SET001 $FACTOR_SET002 $TORSSCF_SET001 $TORSSCF_SET002

\ \ \ \ \

COMBINE NAME DESCRIPTION OLDLOADCASE SET FACTOR NOTCH

= = = = = =

QSL59#CA0001#0.00 ' 4700 rev/min FL Q-S Loading Only - Indet Soln' QSB59#CA0001#0.00 QST59#CA0001#0.00 STRESSED 0.1000E+01 0.1000E+01 0.1000E+01

\ \ \ \ \

VARIABLE NAME = VARIABLE NAME =

TVIB_SET001 VALUE = TVIB_SET002 VALUE =

-0.17605E+00 SAVE = YES -0.17605E+00 SAVE = YES

VARIABLE NAME = BVIB_SET001 VALUE = -0.12131E+01, -0.12315E+01, -0.29455E+00 SAVE = YES VARIABLE NAME = BVIB_SET002 VALUE = -0.12131E+01, -0.12315E+01, -0.29455E+00 SAVE = YES

0.20454E+01, -0.41258E+00, 0.20454E+01, -0.41258E+00,


= = = = = =

TVIB59#CA0001#0.00 ' 4700 rev/min FL TV Only - Indet Soln' 57 FRONT_PIN_FILLET_01 LEAD_PIN_OILHOLE_01 $TVIB_SET001 $TVIB_SET002 $TORSSCF_SET001 $TORSSCF_SET002

\ \ \ \ \


= = = = = =

BVIB59#CA0001#0.00 ' 4700 rev/min FL BV Only - Indet Soln' 58 33 35 37 39 FRONT_PIN_FILLET_01 LEAD_PIN_OILHOLE_01 $BVIB_SET001 $BVIB_SET002 $BENDSCF_SET001 $BENDSCF_SET002

\ \ \ \ \


= = = = = =

QSLT59#CA0001#0.00 ' 4700 rev/min FL Q-S Load + TV - Indet Soln' QSL59#CA0001#0.00 TVIB59#CA0001#0.00 STRESSED 0.1000E+01 0.1000E+01 0.1000E+01

\ \ \ \ \

COMBINE NAME = QSLV59#CA0001#0.00 DESCRIPTION = ' 4700 rev/min FL Q-S Load + Vib - Indet Soln' OLDLOADCASE = QSL59#CA0001#0.00 TVIB59#CA0001#0.00 BVIB59#CA0001#0.00 SET = STRESSED FACTOR = 0.1000E+01 0.1000E+01 0.1000E+01 NOTCH = 0.1000E+01

\ \ \ \ \ \

The SAFETY commands are used to perform the fatigue safety factor calculations. Table B-17 lists the loadcase prefixes that are used for the calculation of fatigue safety

432

HELP Crank Shaft Stress Analysis factors. Table B-18 lists the loadcases that are used for a number of different loadcase options that are available within ENGDYN. Loadcase Prefix QS QST QSV

Description Quasi-static loading Quasi-static loading and torsional vibration Quasi-static loading and bending and torsional vibration

Table B-17 Naming of Fatigue Safety Factor Loadcases

1 2 3 4 5

Loadcase Combination Option Quasi-static loading only Quasi-static loading and Max/Min Torsional Vibration Quasi-static loading and Varying Torsional Vibration Quasi-static loading and Max/Min Bending and Torsional Vibration Quasi-static loading and Varying Bending and Torsional Vibration

Combined Loadcases Used QSB, QST, QSL QSB, QST, QSL, MXTV, MNTV, QMXT, QMNT

Fatigue Used QS QS,QST

Loadcases

QSB, QST, QSL, TVIB, QS,QST QSLT QSB, QST, QSL, MXTV, QS,QST,QSV MNTV, MXBV, MNBV, QMXV, QMNV QSB, QST, QSL, TVIB, QS,QST,QSV BVIB, QSLT,QSLV

Table B-18 Use of Combined and Fatigue Loadcases For the above example the following SAFETY commands are used. SAFETY

NAME DESCRIPTION TYPE LOADCASES SET MATERIAL

= = = = = =

QS59 ' 4700 rev/min FL Q-S Loading - Indet Soln' 'Alternative Goodman' 'QSL59*' FRONT_PIN_FILLET_01 LEAD_PIN_OILHOLE_01 'IRON_PIN_FILLET_D45.0mm' 'IRON_PIN_OILHOLE_D45.0mm'

\ \ \ \ \ \

SAFETY


= = = = = =

QST59 ' 4700 rev/min FL Q-S Loading + TV - Indet Soln' 'Alternative Goodman' 'QSLT59*' FRONT_PIN_FILLET_01 LEAD_PIN_OILHOLE_01 'IRON_PIN_FILLET_D45.0mm' 'IRON_PIN_OILHOLE_D45.0mm'

\ \ \ \ \ \

SAFETY


= = = = = =

QSV59 ' 4700 rev/min FL Q-S Loading + Vib - Indet Soln' 'Alternative Goodman' 'QSLV59*' FRONT_PIN_FILLET_01 LEAD_PIN_OILHOLE_01 'IRON_PIN_FILLET_D45.0mm' 'IRON_PIN_OILHOLE_D45.0mm'

\ \ \ \ \ \

In all cases either quasi-static determinate and indeterminate solutions are required. These are used to calculate stresses due to quasi-static loading only with the option of including torsional vibration only or bending, axial and torsional vibration.

433

HELP Crank Shaft Stress Analysis Torsional vibration is included either by calculating the torsional vibration within ENGDYN or by using the Ricardo programs VALDYN or TVFORCED. The torsional vibration may be combined either by assuming the torsional behaviour is in phase with the quasi-static loading (Option 3 of Table B-18), that is the mean torque and torque amplitude vary at each crank angle, or by calculating a maximum and minimum torque and applying that at every crank angle (Option 2 of Table B-18). In order to account for the uncertainty concerning the phasing of the vibratory torque with respect to the quasistatic loads the latter method (Option 2) is used. Bending and torsional vibration is included by performing a dynamic solution equivalent to the quasi-static indeterminate solution. This is best achieved by using the Save Static Results option on the Evaluate Solution Panel when the dynamic solution is performed. The vibration may be combined either by assuming the vibratory behaviour is in phase with the quasi-static loading (Option 5 of Table B-18), that is the mean vibrations and vibration amplitudes vary at each crank angle, or by calculating maximum and minimum vibration values and applying that at every crank angle (Option 4 of Table B-18). Since a dynamic solution removes the uncertainty concerning the phasing between the torsional and bending behaviour it is preferable to use the former method (Option 5).

6.2 Using the Graphical User Interface The Crankshaft Stress Analysis Panel shown in Figure B-203 is displayed when the Crank Analysis button is selected from the Main Panel.

Figure B-203 Crankshaft Stress Analysis Panel

434

HELP Crank Shaft Stress Analysis Method is used to specify the stress analysis method. If FE Quasi-Static is selected then two Output options are available. Quasi-Static Unit Loads is selected to apply unit loads to the finite element model of the crankshaft as described in 6.1.2.3, before proceeding to Stress and Fatigue Safety Factors which is selected once the displacements and stresses due to the unit loads have been calculated using a suitable finite-element solver. The Select Model button is used to specify the finite element model of the crankshaft, which must be selected before the crankshaft geometry can be edited as described below in 6.2.2. The Define Material button can be selected at any time to define the crankshaft material properties once a model has been selected. The remaining buttons (below Define Material) on the right of the panel are used to define various parameters and to execute the analysis. If Quasi-Static Units Loads is selected it is only necessary to select a model using Select Model and to define and generate the output using Define Output and Generate Output to apply the units loads to the model. If the finite element method is being used the crankshaft model can be displayed in the display area of the Main Panel by selecting the Model toggle. This will display the edges of the model. The Sets toggle will display the mesh of the node set corresponding to each selected entity when editing the crankshaft geometry. The Labels toggle will display the node numbers for each node set.

6.2.1

Crankshaft Finite Element Model

The stress quality finite element model of the crankshaft is selected using the Select Model button on the Crankshaft Stress Analysis Panel shown in Figure B-203, which displays the Crankshaft Model Panel shown in Figure B-204. This figure shows the File tab selected.

Figure B-204 Crankshaft Model Panel InputName This is used to specify the name of the file containing the stress quality finite-element model of the crankshaft. If a finite element model of the crankshaft was used during model generation as described in 3.3.1 then this file name will initially default to that model. Specify SFE file name when translated from another FE package.

435

HELP Crank Shaft Stress Analysis Origin This is used to specify from which finite-element package the crankshaft model is being translated. The options are RICARDO-SFE, MSC/NASTRAN and IDEAS. The model will be translated from the native finite-element package to a Ricardo Standard File (SDF) format known as the Standard Finite Element File (SFE) from which ENGDYN reads and writes data for analysis. Table B-6 summarises the formats from which the crankshaft model can be read Units This is used to specify the units of the finite element model of the crankshaft. The options are SI, MM, METRIC, FT and INCH. Table B-7 summarises the units of length, force, density and mass for each of the options. If the File Origin is specified as IDEAS or RICARDO-SFE this option is not required. The unit system is always assumed to be defined in the IDEAS Universal file and the SFE file is always assumed to be in the SI unit system.

Figure B-205 Crankshaft Model Panel Figure B-205 shows the Crankshaft Model Panel with the Transformation tab selected. It is used to define any necessary transformation from the co-ordinate system used by the FE model of the crankshaft to that used by ENGDYN. If a finite element model of the crankshaft was used during model generation, as described in 3.3.1, then the fields of this tab will default to those values defined using the Model Definitions Panel shown in Figure B-25. All vectors are in the ENGDYN global axes. The Mirror vector specifies a normal vector to the mirror, the plane of which passes through the origin at the ENGDYN co-ordinate system. If zero is entered for the X, Y or Z values of a transformation, then that transformation will not occur. The transformation matrix is built up in the following order; mirror, rotation about Z, then Y and finally about X and then the translation. The FE model must be orthogonal to the ENGDYN co-ordinate system.

6.2.2

Crankshaft Geometry Definition

The area on the left of the Crankshaft Stress Analysis Panel shown in Figure B-203 and Figure B-206 lists the individual components of the crankshaft that can be edited. In this

436

HELP Crank Shaft Stress Analysis case, the Crankshaft Web is shown highlighted. Other components from the list can be highlighted using the mouse. If the Classical method is used Pin Journals and Main Journals will not be displayed in the list. The components may be edited in any order with the exception of Pin Journal Fillets and Main Journal Fillets for which the crankshaft web adjacent to the fillet must be defined before editing the fillet data. By clicking on the Select All button, all components of that type will be selected, and will be shown in red in the main drawing area. Alternatively, individual components may be either selected, or de-selected using the mouse on the relevant component shown in the main drawing, as shown in Figure B-206. De-selected components will be shown in green. All selected components may be de-selected by clicking the Deselect All button.

Figure B-206 Editing Geometry for Crankshaft Stress Analysis If the data are non-uniform for a particular entity then it is recommended that the user start by selecting all to define the dominant dimensions before selecting components individually with the mouse to make any required changes. Having selected the particular components that you wish to define, they may be edited by clicking on the Edit Selected button. This will open up the relevant panel. In each panel a drawing area will display a drawing of the current component, defining the model parameters.

437

HELP Crank Shaft Stress Analysis It is not necessary to define all the crankshaft geometry. If, for example, only pin journal oilholes are of interest then it is only necessary to select and edit Pin Journal Oilholes using the Pin Journal Oilholes Panel shown in Figure B-210. Conversely, if all the geometry is defined, and the finite element method is used the user has the option of subsequently selecting the node sets to be stressed when using the Crankshaft FE Stress Analysis Panel as described below. If stresses and fatigue safety factors are required using the Classical Method at the pin and main journal fillets on the crankshaft web overlap section as shown below in Figure B-208, then the geometry of the fillets and lands of the crankshaft web must be defined as well as the geometry of the web itself. 6.2.2.1 Crankshaft Webs

The Crankshaft Web Panel is shown in Figure B-207 and is used to define the geometry for calculating the adjacent journal fillet stress concentration factor (SCF) values and the bending modulus for each of the selected crankshaft webs.

438


Figure B-207 Crankshaft Web Panel The Full Thickness is the thickness between the facing lands of the web and has previously been derived from geometry defined using the Engine Configuration Panel as shown in Figure B-18 and so its value is displayed, and may not be edited. Reduced Thickness This defines the thickness of the web at the extremity of the web overlap section between the journal centres. This parameter is only used for the calculation of the bending modulus of the overlap section as required for the classical method. Minimum Width This is used to specify the minimum width Bmin of the web between the journal centres as shown in the figure. Maximum Width This is used to specify the maximum width, Bmax of the web between the journal centres as shown in the figure. Equivalent Width This defines the equivalent width, B of the crankshaft web and is calculated from the maximum and minimum widths as defined by

439


1 1 1 1   3  3 3 2  Bmin Bmax B

  

This value is displayed and may not be edited. Offset This is used to specify the offset of the reduced web section at the extremity of the web overlap section and is only used in the rigorous calculation of the bending modulus of the overlap section as required for the classical method and as defined on the Crankshaft Classical Stress Analysis Panel. If the Classical Method is used and the geometry of the fillets and facing lands for a given web are defined as well as the geometry of the web itself the section modulus of the overlap section as defined in Figure B-208 is calculated and stresses and fatigue safety factors are calculated at the pin and main journal fillets. This figure shows a section in which the Offset is zero.

Figure B-208 Definition of Flexural Modulus at Web Overlap Section If there is no journal overlap for a given web, where the overlap accounts for the undercut into each journal, the modulus is calculated assuming a rectangular section using the values of Reduced Thickness and Equivalent Width. 6.2.2.2 Pin Journal Fillets The Pin Journal Fillets Panel is shown in Figure B-209 and is used to edit the geometry and calculate the Stress Concentration Factor (SCF) values for the selected pin journal fillets.

440


Figure B-209 Pin Journal Fillets Panel Radius This is used to specify the radius of the fillet. Web Undercut This is used to specify the undercut of the fillet into the adjacent web. Journal Undercut This is used to specify the undercut of the fillet into the journal. The option menu Calculation is used to select the method for calculating the SCF values which are displayed in the table at the bottom of the panel for each selected pin journal fillet together with the journal diameter and the geometry of the adjacent web. The options are Lowell et al, User or a user-defined method. The default method Lowell et al and the procedure for defining a user-defined method using environment variables is described in Appendix 4. If User is selected the columns Kb and Kt are editable and the SCF values are entered directly. The column Setname, which is only displayed if the finite-element method is used, shows the node set name for each selected pin fillet as defined in 6.1.2.2.2. These set names may be edited, otherwise the default names as shown in Figure B-209 are used.

441

HELP Crank Shaft Stress Analysis The location of each node set is checked against the expected location as defined by the supplied geometry when either the OK or Apply buttons are selected. The mesh of each node set can be viewed by selecting the Sets toggle on the Crankshaft Stress Analysis Panel. The graphs of Kb and Kt versus the variable selected using the Variable option menu are displayed for the current pin journal fillet selected from the table. If no fillet is selected then the first fillet displayed in the table is plotted. The Variable options are Radius, Web Undercut or Journal Undercut. A tracer and an associated Fillet SCF Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu, which has Show Tracer and Hide Tracer options. 6.2.2.3 Main Journal Fillets The geometry and calculation of the Stress Concentration Factor (SCF) values for the main journal fillets are defined using the Main Journal Fillet Panel similar to the Pin Journal Fillets Panel shown in Figure B-209. 6.2.2.4 Pin Journal Oilholes The Pin Journal Oilholes Panel is shown in Figure B-210 and is used to edit the geometry and calculate the Stress Concentration Factor (SCF) values for the selected pin journal oil holes. Type This is used to specify whether the oil hole type is Single (Leading), Single (Trailing), Twin, Cross or None. Changing this option will change the drawing displayed on the panel and the items that may be edited. Diameter, d This defines the diameter of each of each oil hole breakout. include the chamfer of the hole.

This diameter should

Height, h This defines the perpendicular distance of the oil hole breakout from the journal centre as shown in Figure B-210

442


Figure B-210 Pin Journal Oilholes Panel The inclination angle of the oil hole from the normal to the journal is calculated by defining the vector of the oil hole with respect to the global axes at the oil hole breakout. This vector is defined using X, Y and Z. The calculated inclination is displayed in the table at the bottom of the panel and may not be edited. The option menu Calculation is used to select the method for calculating the SCF values which are displayed in the table at the bottom of the panel for each selected pin journal oil hole together with the journal diameters. The options are Peterson et al, User or a user-defined method. The default method Peterson et al and the procedure for defining a user-defined method using environment variables is described in Appendix 4. If User is selected the columns Kb and Kt are editable and the SCF values are entered directly. The column Setname, which is only displayed if the finite-element method is used, shows the node set name for each selected pin oil hole as defined in 6.1.2.2.2. These set names may be edited, otherwise the default names as shown in Figure B-210 are used. The location of each node set is checked against the expected location as defined by the supplied geometry when either the OK or Apply buttons are selected. The mesh

443

HELP Crank Shaft Stress Analysis of each node set can be viewed by selecting the Sets toggle on the Crankshaft Stress Analysis Panel. The graphs of Kb and Kt versus the variable selected using the Variable option menu are displayed for the current pin journal oil hole selected from the table. If no oil hole is selected then the first oil hole displayed in the table is plotted. The Variable options are Diameter, d or Inclination Angle. A tracer and an associated Fillet SCF Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu that has Show Tracer and Hide Tracer options. 6.2.2.5 Main Journal Oilholes The geometry and calculation of the Stress Concentration Factor (SCF) values for the main journal oil holes are defined using the Main Journal Oilholes Panel similar to the Pin Journal Oilholes Panel shown in Figure B-210. 6.2.2.6 Pin Journal Lands The Pin Journal Lands Panel is shown in Figure B-211 and is used to edit the geometry and node set name corresponding to each selected pin journal facing land.

Figure B-211 Pin Journal Lands Panel

444

HELP Crank Shaft Stress Analysis Diameter This defines the diameter of each facing land as shown in Figure B-211. Thickness This defines the thickness of each facing land as shown in Figure B-211. The column Setname, which is only displayed if the finite-element method is used, shows the node set name for each selected pin journal land as defined in 6.1.2.2.2. These set names may be edited, otherwise the default names as shown in Figure B-211 are used. The location of each node set is checked against the expected location as defined by the supplied geometry when either the OK or Apply buttons are selected. The mesh of each node set can be viewed by selecting the Sets toggle on the Crankshaft Stress Analysis Panel. 6.2.2.7 Main Journal Lands The geometry and node set names for the main journal lands are defined using the Main Journal Lands Panel similar to the Pin Journal Lands Panel shown in Figure B-211. 6.2.2.8 Pin Journals

Figure B-212 Pin Journal Panel The Pin Journal Panel is shown in Figure B-212 and is used to edit the node set name corresponding to each selected pin journal as defined in 6.1.2.2.2. The location of each node set is checked against the expected location as defined by the supplied geometry when either the OK or Apply buttons are selected. The mesh of each node set can be viewed by selecting the Sets toggle on the Crankshaft Stress Analysis Panel. 6.2.2.9 Main Journals

445

HELP Crank Shaft Stress Analysis The node set names for the main journals are defined using the Main Journals Panel similar to the Pin Journal Panel shown in Figure B-212

6.2.3

Material Definition

The material properties of the crankshaft are defined by selecting Define Material on the Crankshaft Stress Analysis Panel as shown in Figure B-203, which displays the Crankshaft Material Properties Panel shown in Figure B-213. 6.2.3.1 Base Properties

Figure B-213 Crankshaft Material Properties Panel showing Base Properties If a finite-element model of the crankshaft has been used during model generation to define either the stiffness or mass properties of the ENGDYN model of the crankshaft then the Material Name and values of Young’s Modulus (E), Poisson’s Ratio () and Mass Density () for the crankshaft material will be displayed in the Base Properties tab of the panel. In this case the name and values cannot be edited since they have been used to define the ENGDYN crankshaft mass-elastic system. The Material Type will default to Steel or SG-Iron as appropriate if iron or steel is in the name. Otherwise the type will default to Steel. The Material Type effects the default definition of Size Factor and the treatment of notch sensitivity as described in 6.2.4. If the finite-element method is used to calculate stresses and fatigue safety factors then the material name and the values of E,  and  of the crankshaft material of the stress quality finite element model must be the same as those displayed on the panel. If a material with the correct name is not found in the SFE file then a Material Properties Panel is displayed listing all those materials in the file that have the correct values of E,  and . The user is then able to select the appropriate material. This material is then renamed in the SFE file.

446

HELP Crank Shaft Stress Analysis Conversely, if no finite element model has been used during model generation then the values of E,  and  and the material name are either edited if the Classical Method is used or selected from the stress quality finite element model using the Select button adjacent to Material Name.

6.2.3.2 Base Strengths

Figure B-214 Crankshaft Material Strengths Panel showing Base Strengths The base strengths UTS, ty, cy, fsb, h and fs of the crankshaft material as cast or forged are defined using Ultimate Tensile, Tensile Yield, Compressive Yield, Infinite-Life Fatigue, Hydrostatic Fatigue and Torsional Fatigue in the Base Strengths tab as shown in Figure B-214. Each of these base strength values should have a value greater than 0.0. The torsional fatigue strength fs is only required when using either the Multi-Axial or Dang Van fatigue safety factor calculation options selected using the Safety Factor Calculation Panel as shown in Figure B-226. The hydrostatic fatigue strength h is only required when performing the Dang Van fatigue safety factor calculation. In the absence of any suitable material property data the torsional and hydrostatic fatigue strengths can be derived from the infinite-life tensile fatigue strength such that:

 fs  a fsb  h  b fs

where for Steel crankshafts a and b are 0.58 and 2.5 respectively, and for SG-Iron crankshafts a and b are 0.68 and 0.83.

447

HELP Crank Shaft Stress Analysis Strengths from a database can be selected using the Select button, which displays the Materials Database Panel shown in Figure B-215. This panel displays the strengths for all the materials in the database for the given material type as defined by Material Type. Once a material is selected and OK is selected the values of E,  and  for the selected material are checked against the values on the Materials Database Panel. If any of these values are different a warning message is displayed.

Figure B-215 Materials Database Panel

6.2.3.3 Elevated Strengths

Figure B-216 Crankshaft Material Properties Panel showing Elevated Strengths The base strengths will be modified due to size, surface finish and surface treatment effects. These modified strengths, referred to as Elevated Strengths, are displayed in the table at the bottom of the Crankshaft Material Properties Panel shown in Figure B-216 for each fillet, oil hole, facing land and journal corresponding to each unique journal diameter. In the absence of any measured strength values for the crankshaft the elevated strengths may be calculated using the assumptions described in Appendix 3.

448

HELP Crank Shaft Stress Analysis The elevated values are calculated by setting the Strengths option menu to Calculated, and are displayed in the table when the APPLY button is selected. The surface treatments and corresponding factors for the fillets, oil holes, facing lands and journals for both the pin and main journals are defined using the relevant options available on the panel. The surface treatment options are None, Cold-Rolled, Rolled and Hardened, Tufftrided, Induction Hardened and Nitrided. Cold-Rolled and Rolled and Hardened only apply to the fillets, while Induction Hardened only applies to the fillets and journals. If any one of these options is selected the default factor corresponding to that surface treatment is displayed. This value can be subsequently edited if required. The size factors are defined by selecting the Define button adjacent to Size Factor, which displays the Size Factor Definition Panel shown in Figure B-217.

Figure B-217 Size Factor Definition Panel The option menu Definition is used to define the characteristic of size factor against diameter. The options are Default, User and Equation. If Default is selected the default size factor equation for the given Material Type, as defined on the Crankshaft Material Properties Panel, is used. The default equations for Steel and SG-Iron materials are defined in Appendix 3. If User is selected values of diameter and size factor can be added to the table on the left of the panel. Rows are added to and deleted from the table using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and Delete Row options. If Equation is selected an equation as a function of the diameter D is entered. A tracer and an associated Size Factor Values Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu that has Show Tracer and Hide Tracer options. Conversely if the elevated strengths are known, the Strength option menu can be set to User Defined and the elevated values can be entered as required in the table at the

449

HELP Crank Shaft Stress Analysis bottom of the Crankshaft Material Properties Panel. column is no longer displayed.

In this case, the Size Factor

If the finite element method is used the elevated and base strength material properties are written to the .SFE file containing the crankshaft model, together with a .MAT file for each material definition. These are written when either the OK or APPLY buttons are selected. The name of the material and .MAT file are constructed from the crankshaft material name as defined by Material Name, the location to which the material refers and the diameter of the journal. In the case shown in Figure B-213 in which there are two unique main journal diameters the following materials would be defined: STEEL_PIN_FILLET_D53.0mm STEEL_PIN_OILHOLE_D53.0mm STEEL_PIN_LAND_D53.0mm STEEL_PIN_JOURNAL_D53.0mm STEEL_MAIN_FILLET_D65.0mm STEEL_MAIN_OILHOLE_D65.0mm STEEL_MAIN_LAND_D65.0mm STEEL_MAIN_JOURNAL_D65.0mm STEEL_MAIN_FILLET_D70.0mm STEEL_MAIN_OILHOLE_D70.0mm STEEL_MAIN_LAND_D70.0mm STEEL_MAIN_JOURNAL_D70.0mm

6.2.4

Notch Sensitivity

The treatment of the notch sensitivity of the crankshaft material is described in detail in Appendix 5. Once the material properties and geometry of at least one fillet or oil hole have been defined, the Fatigue Notch Factor values (FNF) of the fillets and oil holes can be calculated by selecting Notch Factors on the Crankshaft Material Properties Panel, which displays the Fatigue Notch Factors Panel shown in Figure B-218.

Figure B-218 Fatigue Notch Factors Panel

450

HELP Crank Shaft Stress Analysis The option menu Characteristic is used to select the method for calculating the FNF values which are displayed in the table on the right of the panel for each oil hole and fillet for which geometry has been defined. The options are Modified Lowell, Equation or a user-defined method. The default method Modified Lowell and the procedure for defining a user-defined method using environment variables is described in Appendix 5. If Equation is selected an equation as a function of the Stress Concentration Factor values SCF is entered. For those materials that are 100% notch sensitive, such as very high strength steels, this option can be used to define an equation in which the FNF value is simply equal to the SCF value. The columns Bending FE SCF and Torsion FE SCF, which are only displayed if the finite-element method is used, display the FE stress correction factors for each oil hole and fillet as defined by Stress Concentration Factor equation in Appendix 5. The Variable option menu is used to select the variable that is plotted against Stress Concentration Factor. The options are FNF or FE SCF. A tracer and an associated FNF Values Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu that has Show Tracer and Hide Tracer options.

6.2.5

Application of the Quasi-Static Unit Loads Using the Finite-Element Method

The quasi-static unit loads, as part of the finite-element method, are defined by selecting the Define Output on the Crankshaft Stress Analysis Panel shown in Figure B-203, which displays the Crankshaft FE Unit Loads Analysis Panel shown in Figure B-219. This panel is used for defining the unit loads that are written. The application of the unit loads is described in detail in 6.1.2.3.

Figure B-219 Crankshaft FE Unit Loads Analysis Panel Solver This is used to specify the FE solver that is used to calculate the unit load displacements and stresses. The options are FeaRCE(VSS), MSC-NASTRAN, ABAQUS and IDEAS .

451

HELP Crank Shaft Stress Analysis Centrifugal Load If this is selected a body load equivalent to a constant angular velocity about the crankshaft axis is applied to the model. If this is not selected it will not be possible to post-process load cases that include centrifugal loading. Unit Moments If this is selected unit moment bearing loads are apply to each main bearing. In addition the unit axial load will also be applied at the nose of the crankshaft. If this is not selected it will only be possible to post-process statically-determinate solution load cases. It is recommended that both the Centrifugal Load and Unit Moments be selected so that all load cases can be post-processed. Output Name This is used to specify the name of the .FRC command file and any analysis data file that is written. Once the calculation parameters have been defined successfully the Generate Output button is selected to perform the calculation. The Query Panel of Figure B-220 will be displayed once the FEARCE .FRC command file is written.

Figure B-220 Query Panel This gives the user the option, by answering Yes of calculating the unit load displacements and stresses. If the user replies with No, no output is generated other than the .FPR command file and the calculation is stopped.

6.2.6

Calculation of Stresses and Fatigue Safety Factors

The Define Output button on the Crankshaft Stress Analysis Panel is used to display a panel for defining parameters for the calculation of stresses and fatigue safety factors. If the Classical Method is used the Crankshaft Classical Stress Analysis Panel shown in Figure B-221 is displayed. Conversely, if the finite-element method is used the Crankshaft FE Stress Analysis Panel shown in Figure B-224 is displayed. These panels can only be displayed once the Fatigue Notch Factor (FNF) values have been calculated (for at least one fillet or oil hole) and the load cases for which stress and fatigue results are required have been selected using the Select Loadcases Panel described in 5.2. Once the calculation parameters have been defined the Generate Output button is selected to perform the calculation. 6.2.6.1 Using the Classical Method

452

HELP Crank Shaft Stress Analysis The Crankshaft Classical Stress Analysis Panel shown in Figure B-221 is used to define parameters for the calculation of stresses and fatigue safety factors using the Classical method.

Figure B-221 Crankshaft Classical Stress Analysis Panel Vibratory Load This is used to specify whether vibratory loads are to be included in the calculation of stresses and fatigue safety factors and if so how. If None is selected no vibration loads are included and only stresses due to quasi-static bending are calculated. If dynamic load cases have been selected using the Select Loadcases Panel then by implication vibratory loads must be included and this option cannot be selected. Vibratory loads are included either by selecting Max/Min or Varying. The combination of stresses due to quasi-static loads and vibration loads is described in detail in 6.1.1. Torsional Vibration This is used to specify how the torsional vibration loading is derived. If quasi-static load cases have been selected using the Select Loadcases Panel then the torsional vibration loading can only currently be derived using the Ricardo programs VALDYN or TVFORCED. Data from VALDYN may be extracted either from a linear frequency domain model using the VALDYN (Freq) option or from a time domain model using the VALDYN (Time) option. The data required from these programs to derive the torsional vibration loading for this analysis is described in Appendix 6.. If dynamic load cases have been selected then torsional vibration loading is included in the ENGDYN solution so the option is set to Engdyn. Torques This is used to specify either the VALDYN sdf output file name or the TVFORCED input data file name dependent on the setting of Torsional Vibration. Safety Factor Algorithm This is used to specify the algorithm for calculating the fatigue safety factors. The Max/Min Stress method derives the minimum safety factor by calculating a safety factor

453

HELP Crank Shaft Stress Analysis from the maximum and minimum Signed Von Mises around the engine cycle. This assumes that the maximum and minimum stresses give rise to the minimum safety factor. On the other hand the Rigorous method derives the minimum safety factor by calculating the safety factor corresponding to each combination of stress values around the engine cycle. Crank Web Modulus This is used to specify how the bending modulus of the crank web overlap section, as shown in Figure B-208, is calculated. The Simplified calculation (which is consistent with previous Ricardo programs) assumes that the reduced web section at the extremity of the web overlap section is symmetrical about the web centre. In this case the Offset parameter defined on the Crankshaft Web Panel is ignored. This assumption is not made if the Rigorous calculation is used. If there is no journal overlap for a given web, where the overlap accounts for the undercut into each journal, the modulus is calculated assuming a rectangular section regardless of whether this option is specified as Simplified or Rigorous. Mean Torque This is used to specify whether the mean torque component of the vibratory torque is derived from the indicated mean torque or a user-defined brake mean torque. If Indicated is selected the indicated mean torque derived from the cylinder pressure diagram used during the solution is used in the stress and fatigue calculation. Conversely, if User is selected a torque characteristic is defined using the Mean Torque Definition Panel shown in Figure B-222 which is displayed when the Define button adjacent to the option menu is selected.

Figure B-222 Mean Torque Definition Panel Initially the table will include the indicated mean torque at each speed derived from the cylinder pressure map corresponding to the load type selected using the Select Loadcases Panel. The speeds and torques are edited to enter a new torque characteristic, which is then plotted against the indicated mean torque curve as shown in the figure above. Rows are added and deleted using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and

454

HELP Crank Shaft Stress Analysis Delete Row options. A tracer and an associated Co-ordinates Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu that has Show Tracer and Hide Tracer options. Overload This is used to specify whether the fatigue safety factors are calculated for Cyclic or General overload as defined in Appendix 7. Bending Moments This is used to specify the method for calculating the bending moment at each fillet and oil hole when deriving bending moments from loads calculated using a determinate solution. For this solution method the bending moments are calculated such that each crank unit or crank throw is considered in isolation, unaffected by the forces acting on the rest of the shaft. Two options, Single Bay and Complete Crank, are available for calculating the bending moments. Single bay assumes that in practice, the flexure of the shaft throws load towards the inner edges of the main bearings such that the effective points of support are assumed to be situated halfway between the bearing mid-length and its edge. This is shown in Figure B-223. If centrifugal loads are included in any of the selected load cases then the bearing reactions RL and RR will react these loads as well as the load W due to the gas pressure and the inertia forces of the piston and connecting rod.

Figure B-223 Effective Main Bearing Supports for Bending Moment Calculation Complete Crank assumes that the loads are reacted at the main bearing centres. It is recommended that Single Bay be used since this gives a more realistic approximation of the bending moment at each fillet and oil hole. If Indeterminate or Dynamic load cases are selected using the Select Loadcases Panel then this option menu is not relevant and will be ghosted.

455

HELP Crank Shaft Stress Analysis Output Name This is used to specify the name of the .EDSTRESS file. This file contains the stresses corresponding to the maximum and minimum stress conditions at the lowest calculated factor of safety for each location together with the material properties, SCF and FNF values and section moduli. 6.2.6.2 Using the Finite-Element Method The Crankshaft FE Stress Analysis Panel shown in Figure B-224 is used to define parameters for the calculation of stresses and fatigue safety factors using the finite element method.

Figure B-224 Crankshaft FE Stress Analysis Panel Results To This is used to specify the output format of the results. The options are RICARDO-SFE, IDEAS and FEMVIEW. If RICARDO-SFE is selected then the results can be viewed using the Ricardo program FEVIEWER. If IDEAS is selected a Universal file is written; if FEMVIEW is selected the results are written directly to a FEMVIEW binary file. Crank Angle Interval This is used to specify the crank angle interval at which the operating stresses are calculated. This value defaults to the interval at which results were stored during the solution as defined by Print on the Evaluate Solution Panel. This default may be prohibitive in terms of the time required to calculate the combined stresses and to evaluate the fatigue safety factors, and also in terms of the disk usage. Intervals of 5 or 10 degrees CA are more typical.

456

HELP Crank Shaft Stress Analysis Crank Angle Offset This is used to specify a crank angle offset and corresponds to the first angle at which the operating stresses are calculated. This is useful, for example, for resolving the crank angle at which the maximum gas cylinder pressure occurs through the engine cycle. Node Sets Stresses and fatigue safety factors can be calculated for all the node sets defined during geometry definition, as described in 6.2.2, or for a subset of those sets using the Select button which displays the Node Set Panel shown in Figure B-225. This panel is used to select those node sets for which results are required.

Figure B-225 Node Set Panel Export When the solution is performed a file is created for each selected ENGDYN loadcase and this file is called _L0001.SFE where L0001 is the loadcase number. On completion of the solution each file will contain an exported copy of the model and the stress and safety factor results for that loadcase. The extent of the model that is exported is defined by this option. If Sets is selected then each exported file will contain a model that consists of the node sets for which results are required. If External is selected then each exported file will contain the external surfaces of the model. If All is selected then each exported file will contain a copy of the whole model. Vibratory Load This is used to specify whether vibratory loads are to be included in the calculation of stresses and fatigue safety factors and if so how. If None is selected then no vibration loads are included and only stresses and fatigue safety factors due to quasi-static bending and torsion are calculated. Vibratory loads are included either by selecting Max/Min or Varying. The contribution due to torsional and bending vibration is defined using Torsional Vibration and Bending Vibration as described below. If bending vibration is included it is recommended that this option be set to Varying. The combination of stresses due to quasi-static loads and vibration loads is described in detail in 6.1.2.5 Torsional Vibration This is used to specify how the torsional vibration loading is derived. If bending vibration is included (as defined below) then the torsional vibration is derived from the dynamic

457

HELP Crank Shaft Stress Analysis load case and so the option is set to Engdyn. Otherwise, the torsional vibration loading can be derived using either the Ricardo programs VALDYN or TVFORCED. Data from VALDYN may be extracted either from a linear frequency domain model using the VALDYN (Freq) or from a time domain model using the VALDYN (Time) option. The data required from this program to derive the torsional vibration loading for this analysis is described in Appendix 6. Torques This is used to specify either the VALDYN sdf output file name or the TVFORCED input data file dependent on the setting of the Torsional Vibration menu. Bending Vibration This is used to specify whether the contribution due to bending and axial vibration is included in the calculation of operating stresses. This option can only be selected if Vibratory Load is set to Varying or Max/Min and if statically-indeterminate load cases have been selected using the Select Loadcases Panel, and for each of these load cases there is a corresponding dynamic load case. To ensure that there are matching pairs of statically-indeterminate and dynamic load cases it is recommended that the Save Static Results option be used on the Evaluate Solution Panel when the dynamic solution is performed. This is described in more detail in 6.1.2.5. Mean Torque This is used to specify whether the mean torque component of the vibratory torque is derived from the indicated mean torque or a user-defined brake mean torque. If Indicated is selected the indicated mean torque derived from the cylinder pressure diagram used during the solution is used in the stress and fatigue calculation. Conversely, if User is selected a torque characteristic is defined using the Mean Torque Definition Panel shown in Figure B-222 which is displayed when the Define button adjacent to the option menu is selected. This panel is described above in 6.2.6.1. Safety Calculation The safety factor calculations to be performed are defined using the Select button, which displays the Safety Factor Calculation Panel shown in Figure B-226.

458


Figure B-226 Safety Factor Calculation Panel Any number of safety factor calculation types can be selected, each of which are described in detail in Appendix 7. Output Name This is used to specify the base name of files written during solution. The following files will be written .FRC .EDSTRESS together with _L001.FRC _L001.SFE _l001.jnl for each loadcase. The .EDSTRESS file contains the stresses corresponding to the maximum and minimum stress conditions at the lowest calculated factors of safety for each location together with the material properties and SCF and FNF values. The file _l001.jnl is a journal file that can be used to create a rotating animation of the crankshaft using the Ricardo Program FEVIEWER.

Once the calculation parameters have been defined successfully the Solution button is selected to perform the calculation. The quasi-static unit loads are read from the .SFE file containing the crankshaft model and are displayed in the Crankshaft Unit Loads Panel shown in Figure B-227.

459


Figure B-227 Crankshaft Unit Loads Panel This lists the load case number and force and moment vectors for the unit loads as described in 6.1.2.3. The Query Panel of Figure B-228 is also displayed once the FEARCE .FRC command files are written.

Figure B-228 Query Panel

460

HELP Crank Shaft Stress Analysis This gives the user the option, by answering Yes of calculating the stresses and fatigue safety factors. If the user replies with No, no output is generated and the calculation is stopped.

6.2.7

Writing the Summary .EDSTRESS file

On completion of the solution using FEARCE the .EDSTRESS summary file is writtem. The Write Summary button enables this file to be rewritten. The results are read from the SFE for each loadcase and summary data is written to the file .EDSTRESS.

6.2.8

Plotting Results

Results are plotted by selecting Plot Results on the Crankshaft Stress Analysis Panel shown in Figure B-203, which displays the Plot Results Panel shown in Figure B-229.

Figure B-229 Plot Results Panel

6.2.8.1 Selection Four types of plot can be displayed; Stress History, Safety Factor Summary and a Goodman diagram (style A or B). The selection of each plot type is controlled with the Plot and Results lists. The minimum fatigue safety factor and corresponding stress is plotted for each selected item corresponding to the safety calculation type selected from the Safety Calculation list. If the Classical method is used only Goodman-StandardCyclic and Goodman-Standard-General will ever be displayed in this list. The Crankshaft Stress Analysis Panel is used a given component of the crankshaft, such as pin journal fillets, for which plots are required. Items are then selected and deselected using the mouse and/or the Select All and Deselect All buttons in a similar manner to that described in 6.2.2 when defining the crankshaft geometry. Load cases are selected using the Select Loadcases Panel shown in Figure B-168

461

HELP Crank Shaft Stress Analysis Design Factor This is used to specify the design safety factor displayed on the Goodman and safety summary plots. If the Material Type is defined as Steel on the Crankshaft Material Properties Panel then a default value of 1.5 is displayed, otherwise a value of 1.7 is used as the default.

6.2.8.2 Viewing and Printing Once the required plot has been selected as described in 6.2.8.1, the plots are generated by clicking on the Apply button. They are displayed in a separate Graph Panel as shown in Figure B-183 and described in 5.5.2.

1

ASME, ICE-Vol.9, Book No.100295, 1989, ‘Crankshaft Stress Analysis – The Combination of Finite Element and Classical Analysis Techniques’, A.R.Heath and P.M.McNamara, Ricardo.

2

Published by the Author, 1975, ‘The Crankshaft : Designing for Structural Strength and Fatigue Resistance’, C.M.Lowell.

3

Bulletin of the JSME, 1965, ‘The Bending Stress Concentration Factor of Solid Crankshaft’, J.Arai.

4

Wiley-Interscience, 1974, ‘Stress Concentration Factors’, R.E.Peterson.

5

Experimental Stress Analysis, Vol.2, No.2, 1945, ‘Structural Evolution of a Crankshaft’, S.Oldberg and C.Lipson, Chrysler Corporation.

462

HELP Cylinder Block Analysis

7 Cylinder Block Analysis 7.1 Overview Operating loads calculated by ENGDYN can be applied to the cylinder block either in the frequency domain to perform modal frequency response and acoustic analyses or in the time domain to perform quasi-static analyses. The analyses may be performed on the model used to derive the reduced cylinder block model as described in 2.5, 3.3.4 and 3.5 or can be performed on a different model, but which must be in the same co-ordinate system as the reduced model. If a quasi-static analysis is being performed then the model can consist of just a slice of the cylinder block. The FEARCE suite of programs including FEPRE and FEPOS1 are used to apply and combine loads and to perform analyses. These programs are executed by ENGDYN within the interface, but more experienced users may wish to execute these programs outside the interface.

7.2 Frequency Response and Acoustic Analysis 7.2.1

General Description

A Fast Fourier Transform (FFT) can be performed on the operating loads calculated by ENGDYN to derive frequency response loading that can either be exported to a finiteelement package such as NASTRAN or be used to perform frequency response and acoustic analyses within ENGDYN. A modal frequency response analysis method is used by ENGDYN to predict the response of the cylinder block. This method, as described in Appendix 9, uses the mode shapes of the cylinder block to minimise the size of the solution and uses modal damping to uncouple the equations of motion; this makes the numerical solution more efficient. The predicted response of the cylinder block calculated by ENGDYN can either be exported as a boundary condition to an acoustic calculation program such as SYSNOISE or alternatively an acoustic analysis can be performed in ENGDYN to calculate the radiated noise from the surface vibrations using the Rayleigh equation. Although this equation is applicable only to flat plates, in practice it is possible to obtain reasonable results from surfaces that are generally flat such as the bare sides of a cylinder block. Two solutions to the Rayleigh equation are implemented. The Rayleigh method is a simplified method in which each node is treated as a piston in a baffle assuming that the vibration is the same over the surface of the piston. This is normally acceptable for models where the mesh density is sufficient to give a good representation of the mode shapes and the wavelength is large compared to the element size. The Helmholtz method solves the same Rayleigh equation but to a greater accuracy (and therefore takes longer). The two methods are described in detail in Appendix 8.

463

HELP Cylinder Block Analysis The procedure for deriving the frequency response loading and predicting the nodal vibrations and radiated noise is described in Figure B-230

FE data or binary file containing cylinder block model and modal results if required e.g. NASTRAN OUTPUT2 file

BLOCK MODEL

.FRL File

.FPR File

.FPO file contains commands to execute analyses. .FRL file contains frequency response loading

.SFE file contains external surfaces of FE model, modal results and frequency response loading

FE Data File

Forced Response Analysis

Acoustic Analysis

.SFE file contains nodal vibration and radiated noise results as well as model and modal results

SYSNOISE Data File

VIBPLOT

FE Viewing Package

Figure B-230 Frequency Response and Acoustic Analysis

464

HELP Cylinder Block Analysis 7.2.2


7.2.2.1 Geometry and Restraints The finite-element model of the cylinder block must be in the same co-ordinate system as the model used during model generation as described in 3.3.4 and have nodes that are in the same position as the reduced cylinder block model. Typically this model will be of the engine power train and will not necessarily be constrained such that the modal behaviour is ‘free-free’. If the model is restrained then the engine mount stiffness’ should be included using spring elements that are attached to ground on the vehicle side. This ensures that the rigid-body frequencies and modes are correctly modelled. 7.2.2.2 Modal Characteristics If a frequency response or acoustic analysis is to be performed the modal frequencies and mode shapes of the cylinder block model are required. These can be calculated using a finite-element solver such as MSC/NASTRAN. 7.2.2.3 Set Definition No node sets are required to define the nodes at which the loads are applied. ENGDYN determines the nodes of the model to be loaded by finding those that are coincident with the nodes of the reduced cylinder block model. A single node set called REDUCED_NODES is created which contains these nodes and is written to the .SFE file containing the cylinder block model when the loading is written. Nodal vibrations can be calculated for the complete cylinder block model, or for a subset of selected node sets that are included in the model. A single node set called VIBRATION is created which is the union of the selected sets If an acoustic analysis is performed an elemental face set is required to define each radiating surface. Since the Rayleigh integration is only applicable to flat plates the selected face sets should define radiating surfaces that are generally flat, for example, each side of the cylinder block. It is preferable wherever possible to define sets which are the complete sides of the cylinder block rather than dividing each side up into multiple sets. The nodes of each face set are included with any vibration node sets in a set called VIBRATION. The VIBRATION node set is written to the .SFE file containing the cylinder block model when the calculation is performed.

7.2.3

Frequency Response Loading

The frequency response loading is written to a file with .FRL suffix. This file contains loading in the frequency domain, the degrees of freedom at which the loads act and also modal damping data if modes are present in the .SFE file containing the cylinder block model. Each load case has the same number as the ENGDYN load case number and has a FRL prefix. At the end of this file for each selected loadcase the total forces and moments (with respect to the model origin) acting on the cylinder block model are

465

HELP Cylinder Block Analysis tabulated in the frequency domain. Figure B-231 shows an example of the total tabulated loads for a single loadcase at four frequencies.

Figure B-231 Example Tabulated Total Frequency Response Loading Cylinder pressure data is also written to general variable arrays to the SFE file together with the frequency response loading. These data are written to allow structural attenuation to be calculated using the Ricardo Program VIBPLOT2. For example for a 4cylinder engine with loads at 2000 rev/min Full Load the following arrays will be written: g:cylinder:pressure:sp_2000:ld_100:cyl_1 g:cylinder:pressure:sp_2000:ld_100:cyl_2 g:cylinder:pressure:sp_2000:ld_100:cyl_3 g:cylinder:pressure:sp_2000:ld_100:cyl_4 g:cylinder:crankangle:sp_2000:ld_100

7.2.4

Calculating Nodal Vibration and Radiated Noise

ENGDYN writes a FEARCE command file with an .FRC suffix that contains commands to calculate nodal vibration and radiated noise and is executed using the interface. An example file is shown in Figure B-232. This file creates the VIBRATION node set that includes a node set called MOUNTS and the nodes of a face set called BLOCK_LEFT that defines a radiating surface of the cylinder block. Node and face sets called SUBSET are also created. The node set contains the nodes of the VIBRATION node set together with the nodes at which the loads are applied. The face contains the external faces of the model together with the faces of the VIBRATION set. The sets called SUBSET are then exported to a second .SFE file called EXAMPLE_EXTERNAL.SFE that will contain the modal data and results once the calculation has finished.

466

HELP Cylinder Block Analysis ; Written by ENGDYN OPEN

Version 2.1.2

Run on 01-Jun-2000 at 10:47:31

FILE = 'EXAMPLE' STATUS = OLD

SET TYPE DELETE

= NODE = BLOCK_LEFT

\

SET TYPE = NODE CONVERT = FACE SET = BLOCK_LEFT

\ \

SET DELETE TYPE

= VIBRATION = NODE

\

SET TYPE SET UNION

= NODE = VIBRATION = MOUNTS

\ \

SET DELETE TYPE

= SUBSET = NODE

\

SET SET TYPE UNION

= SUBSET = NODE = VIBRATION

\ \

SET DELETE TYPE

= EXTERNAL = FACE

SET SET

= EXTERNAL

SET DELETE TYPE

= SUBSET = FACE

\

SET SET TYPE UNION

= SUBSET = FACE = VIBRATION

\ \

EXPORT TO FACE_SET NODE_SET STATUS NAME OPEN

= = = = =

BLOCK_LEFT

REDUCED_NODES \

EXTERNAL

SFE SUBSET SUBSET FRESH 'EXAMPLE_EXTERNAL'

\ \

\ \

FILE = 'EXAMPLE_EXTERNAL' STATUS = OLD

SINUSOIDAL FILE = 'EXAMPLE.FRL' SOLVE ;

ANALYSIS = MODAL

2000 rev/min Full Load

Dynamic Solution

VIBRATION

SET = VIBRATION LOADCASES =

NOISE

METHOD LOAD SET DENSITY SPEED_OF_SOUND FILE

;

3000 rev/min Full Load

= = = = = =

8 FRESH = YES

RAYLEIGH 8 BLOCK_LEFT 1.290 344.000 'EXAMPLE.L008'

\ \ \ \ \

Dynamic Solution

VIBRATION

SET = VIBRATION LOADCASES =

NOISE

METHOD LOAD SET DENSITY SPEED_OF_SOUND

12 FRESH = YES

= RAYLEIGH = 12 = BLOCK_LEFT = 1.290 = 344.000

\ \ \ \ \

467

HELP Cylinder Block Analysis FILE

= 'EXAMPLE.L012'

Figure B-232 An Example .FRC File for Frequency Response and Acoustic Analysis The SINUSOIDAL command reads the frequency response loading for two load cases at 2000 and 3000 rev/min contained in the .FRL file called EXAMPLE.FRL. The SOLVE command calculates for each of these load cases the modal contribution factors as described in Appendix 9. These factors are written to the .SFE file and to files called EXAMPLE.L008.ALPHA and EXAMPLE.L012.ALPHA for each load case. Nodal vibrations and radiated noise for the face called BLOCK_LEFT are then calculated for each of the two load cases with the nodal vibrations for load case 8 deleted from the file. Radiated sound power and radiation efficiency for the set BLOCK_LEFT are printed in tabular form to files called EXAMPLE.L008.RES and EXAMPLE.L012.RES for each load case. Sound intensities for the set BLOCK_LEFT are written to the .SFE file. The vibration and noise results can then be plotted using the Ricardo program VIBPLOT2, which reads the sinusoidal loading and vibration and sound intensity results stored in the .SFE file.

7.3 Quasi-Static Analysis 7.3.1

General Description

Two methods are available for performing a quasi-static analysis of a cylinder block to predict quasi-static displacements, stresses and fatigue safety factors using the finiteelement method. The first method is a unit-load approach in which unit loads are applied to the model, displacements and stresses are calculated using an appropriate finiteelement package and the loads are factored and combined during post-processing. The second method, referred to as the Direct Method, is to calculate the loads and to apply these loads to the model so that there is a load case corresponding to each selected crank angle. A solution is then performed on these load cases using an appropriate finite-element package. The main advantage of this method over the Unit Load Method is that the piston side loads are applied to the cylinder bore using a single distributed load at each crank angle, whereas the latter combines unit loads applied at the connecting rod small end reversal positions. However each method has advantages and disadvantages and the chosen method depends on the objective of the analysis and the configuration and operating conditions of the engine. For example, the Direct Method is preferred when analysing a large single speed engine such as a V16 since the number of unit loads will be greater than the number of load cases to analyse the engine cycle at every 5 degrees CA.

7.3.2


7.3.2.1 Geometry and Constraints The finite-element model can either be of the complete cylinder block including the cylinder head or a subset of it and must be in the same co-ordinate system as the model

468

HELP Cylinder Block Analysis used during model generation as described in 3.3.4. If the model consists of a slice of the cylinder block then the model must include at least two main bearings. The cutplanes of the slice can either be at the main bearing centres or cylinder centres. In this case half the load is applied to the cylinder or main bearing at each cut-plane. The model should be restrained using a 3-point constraint. If the model consists of a slice then it is the responsibility of the user to apply appropriate constraints to the cutplanes. These constraints can be applied using the program FEPRE1. 7.3.2.2 Set Definition A number of elemental face sets are required to apply the loading. Sets are required for each modelled main bearing, cylinder bore and cylinder head gas face using the names MAIN_JOURNAL_BEARING, CYLINDER_BORE and HEAD_GAS_FACE respectively. If axial loads are to be included in the analysis then sets are required for each bearing surface of the thrust bearing using the names REAR_THRUST_BEARING and FRONT_THRUST_BEARING where the definition of FRONT and REAR depends on the numbering of the main bearings as defined during model generation. If a given face set does not exist but a node set of the same name does the node set will be converted to create the face set. Each main bearing face set must extend over the area in which the journal oil film acts such that the axial position and extent of the set should correspond to the bearing centre and bearing shell length defined during model generation using the Engine Configuration Panel and Main Bearing . Figure B-233, which shows a slice through a four-cylinder inline engine, shows the main bearing sets displayed in red. Each cylinder bore set should include the complete axial extent of the cylinder bore as shown in Figure B-233. The axial position and modelled bore diameter should correspond to the cylinder centre and cylinder bore defined during model generation using the Engine Configuration Panel. The axial extent of all the cylinder bore sets should be the same. The figure shows the cylinder bore set in yellow with the cylinder head gas face set shown in green. This set need not necessarily include the valves as shown in the figure, since the pressure load is factored to apply the correct load to account for the difference between the actual bore area and the modelled area. If axial loads are to be included in the analysis then the sets defining each bearing surface of the thrust bearing should have axial positions such that they are equidistant from the centre of the appropriate main bearing. Figure B-233 shows the front thrust bearing set in blue. The rear thrust bearing set is out of view. The names of the sets must follow the same numbering convention as defined during model generation as described in 3.2. For example, in the case of the four-cylinder inline engine the required sets would be as follows: MAIN_JOURNAL_BEARING_01 MAIN_JOURNAL_BEARING_02 MAIN_JOURNAL_BEARING_03 MAIN_JOURNAL_BEARING_04 MAIN_JOURNAL_BEARING_05

HEAD_GAS_FACE_01 HEAD_GAS_FACE_02 HEAD_GAS_FACE_03 HEAD_GAS_FACE_04

469

CYLINDER_BORE_01 CYLINDER_BORE_02 CYLINDER_BORE_03 CYLINDER_BORE_04

HELP Cylinder Block Analysis If axial loads were to be included in the analysis then two additional sets would be required as follows: REAR_THRUST_BEARING_01

FRONT_THRUST_BEARING_01

Figure B-233 Loading Sets

7.3.3

Unit Load Method

In this method, as described by the flowchart of Figure B-234, unit loads are applied to the model, displacements and stresses are calculated using an appropriate finiteelement package and the loads are factored and combined during post-processing.

470


FE data or binary file containing cylinder block model and set definitions e.g. IDEAS Universal File

BLOCK MODEL

Unit Load Calculations

FE Data File

FE SOLVER

.FPR command file defining unit loads

.FPR File

.SFE file contains model and unit loads. On completion of the FE solution the stresses and displacements are appended using the appropriate translator

Translator

Fatigue Safety Calculations

.FPO command file defining unit load factors, any stress correction factors and fatigue safety calculations

.FPO File

.SFE file contains external surfaces of FE model with operating load stresses and fatigue safety factors

FE Viewing Package

Figure B-234 Quasi-Static Cylinder Block Analysis Using the Unit Load Method

471

HELP Cylinder Block Analysis 7.3.3.1 Application of the Quasi-Static Unit Loads ENGDYN applies unit loads at each main journal bearing, cylinder bore and cylinder head gas face. If axial loads are to be included in the analysis then thrust bearing unit loads are also applied. The bearing loads are applied to the sets as defined in 7.3.2 as distributed loads using nodal forces that give a pressure distribution that varies as a parabolic function along the journal and around its circumference with an extent of 180o. View 1 of Figure B-235 shows the pressure distribution due to a force load in the -Y direction, whilst view 2 shows a moment load in the -Y direction. The latter is applied using loading in Z applied in opposite directions on each half of the bearing to give a couple. Viewing the nodal forces as opposed to the face pressures will not necessarily give a smooth distribution as indicated in the figure since the nodal forces are dependant on element size and whether the elements are linear or parabolic.

Figure B-235 Journal Bearing Unit Loads Unit force loads are applied in each of 4 orthogonal directions (+Y -Y +Z and -Z) in the ENGDYN co-ordinate system at each of the main and pin journal bearings. In addition, for quasi-static indeterminate or dynamic solutions, unit moment loads are applied in each of 4 orthogonal directions (+Y -Y +Z and -Z) at each of the main journal bearings. The cylinder bore loads are applied to the sets as defined in 7.3.2 as shown in Figure B-236 in a similar manner to the bearing loads. View 1 shows the pressure distribution due to the force load at the small end lower reversal position acting on the thrust side, whilst view 2 shows the force load at the small end upper reversal position. Four unit force loads are applied on the cylinder bore in the ENGDYN co-ordinate system at each cylinder. These are applied at small end lower and upper reversal positions acting on the thrust side as shown in Figure B-236 together with the similar

472

HELP Cylinder Block Analysis loads acting on the anti-thrust side. A fifth unit load is applied as a pressure load to the cylinder head gas face.

Figure B-236 Cylinder Bore Unit Loads If axial loads are to be included in the analysis a unit load is applied as a pressure load to each bearing surface of the thrust bearing. For a four-cylinder in-line engine in which durability of the cylinder block is to be assessed using dynamic loading 62 unit load cases would be applied to the complete model of the cylinder block. These would include 20 unit force and 20 unit moment load cases at the main bearings, 20 unit load cases at the cylinders and 2 unit load cases at the thrust bearing. ENGDYN writes an analysis data file that contains the geometry and restraints defined by the user together with the loading as defined above for solution in an appropriate user defined finite-element package. The loading is also written to a FEARCE command file with an .FRC suffix which is used to write the loading to the Ricardo Standard Finite Element File (SFE) and to export the data to the appropriate format. Part of an example .FRC written by ENGDYN in which a MSC/NASTRAN data file is exported is shown in Figure B-237. The figure shows the commands to apply a unit-force bearing load of 1000 Nm in the +Y direction and to export the data to MSC/NASTRAN. The user may define additional loads by editing the command file. If any of these loads have a resultant vector that is the same as any unit load then it must have a loadcase number higher than the unit load.

473

HELP Cylinder Block Analysis ; Written by ENGDYN

Version 2.1.2

Run on 12-Feb-2000 at 12:20:15

OPEN FILE='EXAMPLE' STATUS=OLD BEARING

SET LOADCASE NAME DESCRIPTION CENTRE AXIS_VECTOR WIDTH FORCE TORQUE ANGLE TYPE

= = = = = = = = = = =

MAIN_BEARING_01 1 MBUL01#STEP01# 'Force Load +Y -0.172050E+00, 0.100000E+01, 0.225000E-01 0.000000E+00, 0.000000E+00, 0.180000E+03 SHELL

Main bearing 01' 0.000000E+00, 0.000000E+00 0.000000E+00, 0.000000E+00 0.100000E+04, 0.000000E+00,

0.000000E+00 0.000000E+00

\ \ \ \ \ \ \ \ \ \

. . . SAVE EXPORT TO ANALYSIS UNIT NAME

= = = =

'MSC/NASTRAN' STRESS SI 'EXAMPLE'

\ \ \

Figure B-237 An Example .FPR File for Applying Unit Loading

7.3.3.2 Calculating Quasi-Static Displacements, Stresses and Fatigue Safety Factors Once operating loads have been calculated as described in Chapter 4 and a finiteelement solution has been performed to calculate the displacements and stresses due to the unit loading as described in 7.3.3.1 the displacements and stresses at a userspecified crank angle interval are calculated by factoring and combining the unit loads throughout the engine cycle. Thermal and assembly load cases can also be included with the operating loads if required. Fatigue safety factors can then calculated from the resultant stress history. Stress correction factors arising from notch sensitivity can be applied to node sets specified by the user. The loadcase factors and combinations are written by ENGDYN to a FEARCE command file with an .FRC suffix. This is used by ENGDYN to calculate the combined stresses and fatigue safety factors that are written to the Ricardo Standard Finite Element File (SFE) and which then can be exported to an appropriate finite-element package for viewing.

7.3.4

Direct Method

In this method, as described by the flowchart of Figure B-238, operating loads which have previously been calculated as described in Chapter 4 are applied to the model during pre-processing. A load case is defined for each selected crank angle of each selected ENGDYN load case. Loads are applied at each main journal bearing, at each cylinder and if axial loads are included in the analysis at each thrust bearing also.

474

HELP Cylinder Block Analysis The bearing loads are applied to the sets as defined in 7.3.2.2 as distributed loads using nodal forces that give a pressure distribution that varies as a parabolic function along the journal and around its circumference. Similarly the piston side loads are applied as distributed load to each cylinder bore at the appropriate location as defined by the extent of the piston skirt defined using the Cylinder shown in Figure B-59.

FE data or binary file containing cylinder block model, any loading other than operating loads and set definitions e.g. IDEAS Universal File

BLOCK MODEL

.FPR command file defining operating loads

.FPR File

.SFE file contains model and loads. On completion of the FE solution the stress and displacement results are appended using the appropriate translator FE analysis data file contains model and loads.

FE Data File

FE SOLVER Translator .SFE file contains model, loads and results. Translator FE Viewing Package

Figure B-238 Quasi-Static Cylinder Block Analysis Using the Direct Method

475

HELP Cylinder Block Analysis The load acting on the cylinder head gas face due to the gas cylinder pressure is also applied. If the piston top ring position has been defined using the Cylinder then the gas cylinder pressure is applied to each cylinder bore above the top ring. ENGDYN writes an analysis data file that containing the geometry, the user-defined restraints, the operating loads as defined above together with any other loading, for solution in an appropriate user defined finite-element package. The loading is also written to a FEARCE command file with an .FRC suffix that is used to write the loading to the Ricardo Standard Finite Element File (SFE) and to export the data to the appropriate format. A solution is then performed using an appropriate finite-element package.

7.4 Using the Graphical User Interface The Cylinder Block FEA Panel shown in Figure B-239 is displayed when the Block Analysis button is selected from the Main Panel.

Figure B-239 Cylinder Block FEA Panel The panel consists of four buttons that are selected sequentially. The required analysis and output is selected using the Select Output button, which displays the Cylinder Block FEA Output Panel shown in Figure B-240.

Figure B-240 Cylinder Block FEA Output Panel If the unit-load method is used to calculate quasi-static displacements and stresses, as described in 7.3.3, Quasi-Static Unit Loads is selected to apply unit loads to the finite element model of the cylinder block as described in 7.3.3.1 before proceeding to QuasiStatic Unit Load Factors or Quasi-Static Combined Results. These are selected once the displacements and stresses due to the unit loads have been calculated using a

476

HELP Cylinder Block Analysis suitable finite-element solver such as MSC/NASTRAN. Quasi-Static Unit load Factors is selected to write the load case factors and combinations to the FEPOS command file as described in 7.3.3.2. In addition, if Quasi-Static Combined Results is selected the command file is used to calculate the quasi-static displacements and stresses and to calculate if required the fatigue safety factors from the resultant stress history. Conversely, if Quasi-Static Loading is selected loads are applied using the direct method as described in 7.3.2. If Frequency Response Loading is selected a Fast Fourier Transform (FFT) is performed on the time domain cylinder block forcing data and the frequency response loading is written in an appropriate format. If Nodal Vibrations is also selected a modal forced response analysis, as described in 7.2 and Appendix 9, is performed using modes calculated using a suitable finite element solver to predict frequency domain displacements, velocities or accelerations at selected cylinder block nodes. Radiated Noise is selected to calculate radiated noise from vibrating surfaces using the Rayleigh or Helmholtz method as described in 7.2 and Appendix 8. The Select Model button on the Cylinder Block FEA Panel shown in Figure B-239 is used to select, as described below in 7.4.1, the finite element model of the cylinder block for the analysis. The Define Output button is used to display the relevant panel for defining parameters for the chosen output. Once the calculation parameters have been defined, and if required, load cases have been selected using the Select Loadcases Panel described in 5.2, the Generate Output button is selected to perform the appropriate calculation.

7.4.1

Cylinder Block Finite Element Model

The finite element model of the cylinder block is selected using the Select Model button on the Cylinder Block FEA Output Panel shown in Figure B-240, which displays the Cylinder Block Model Panel shown in Figure B-241. This figure shows the Model tab selected.

Figure B-241 Cylinder Block Model Panel

477

HELP Cylinder Block Analysis File Name This is used to specify the name of the file containing the finite-element model of the cylinder block. If a finite element model of the cylinder block was used during model generation as described in 3.3.4 then this file name will initially default to that model. File Origin This is used to specify from which finite-element package the cylinder block model is being translated. The options are RICARDO-SFE, MSC/NASTRAN and IDEAS. The model will be translated from the native finite-element package to a Ricardo Standard File (SDF) format known as the Standard Finite Element File (SFE) from which ENGDYN reads and writes data for analysis. Table B-6 summarises the formats from which the cylinder block model can be read. File Units This is used to specify the units of the finite element model of the crankshaft. The options are SI, MM, METRIC, FT and INCH. Table B-7 summarises the units of length, force, density and mass for each of the options. If the File Origin is specified as IDEAS or RICARDO-SFE this option is not required. The unit system is always assumed to be defined in the IDEAS Universal file and the SFE file is always assumed to be in the SI unit system.

Figure B-242 Cylinder Block Model Panel Figure B-242 shows the Cylinder Block Model Panel with the Transformation tab selected. It is used to define any necessary transformation from the co-ordinate system used by the FE model of the cylinder block to that used by ENGDYN. If a finite element model of the cylinder block was used during model generation, as described in 3.3.4, then the fields of this tab will default to those values defined using the Model Transformation and Translation Panel shown in Figure B-26. All vectors are in the ENGDYN global axes. The Mirror vector specifies a normal vector to the mirror, the plane of which passes through the origin at the ENGDYN co-ordinate system. If zero is entered for the X, Y or Z values of a transformation, then that transformation will not occur. The transformation matrix is built up in the following order; mirror, rotation about Z, then Y and finally about X and then the translation. The FE model must be orthogonal to the ENGDYN co-ordinate system unless a quasi-static analysis using the direct method, as described 7.3.4, is being performed, in which case the transformation matrix can be non-orthogonal.

478

HELP Cylinder Block Analysis 7.4.2

Frequency Response and Acoustic Analysis

The Define Output button on the Cylinder Block FEA Panel is used to display a panel for defining calculation parameters. If Frequency Response Loading, Nodal Vibrations or Radiated Noise have been selected on the Cylinder Block FEA Output Panel the Frequency Response and Acoustic Analysis Panel shown in Figure B-243 is displayed. This panel has three tabs and can only be displayed once the output and model have been selected. The figure shows the Frequency Response tab selected. Once the calculation parameters have been defined and the load cases for which results are required have been selected using the Select Loadcases Panel described in 5.2 the Generate Output button is selected to perform the calculation.

Figure B-243 Frequency Response and Acoustic Analysis Panel

7.4.2.1 Frequency Response Analysis The Frequency Response tab of the Frequency Response and Acoustic Analysis Panel shown in Figure B-243 is used to define the frequency response loading for the frequency response analysis and subsequent acoustic analysis if Radiated Noise has been selected on the Cylinder Block FEA Output Panel. Vibration Set Nodal vibrations can be calculated for the complete cylinder block model, or for a subset of selected node sets included in the model using the Select button, which displays the Node Set Panel shown in Figure B-244. This panel is used to select those node sets for which results are required. A single node set called VIBRATION is created which is the union of the selected sets. This set is written with the loading to the .SFE file containing the cylinder block model.

479

HELP Cylinder Block Analysis Figure B-244 Node Set Panel Modal damping Modal damping, as described in Appendix 9, is defined using the Define button, which displays the Modal Damping Characteristic Panel shown in Figure B-245. Modal damping values are only written with the loading if the .SFE file containing the cylinder block model (selected using the Cylinder Block Model Panel) contains modal frequencies. If a dynamic cylinder block model has been defined during model generation using the Model Definitions Panel then a line will be plotted for each selected loadcase. These will be the modal damping characteristics used during solution and defined before executing the solution using Modal Damping Characteristic Panel shown in Figure B-245. Conversely, if the cylinder block was not defined as dynamic then the default characteristic will be plotted as shown in the below figure. The option menu Definition is used to define the modal damping characteristic. The options are As Solution, Equation, User or Default. If As Solution is selected then the modal damping characteristic used during solution will be used for each selected load case. This option is only available if the cylinder block model was defined as dynamic during model generation. If Equation is selected an equation as a function of frequency F is entered. If User is selected values of frequency and damping ratio can be added to the table on the left of the panel. Rows are added to and deleted from the table using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and Delete Row options. If Default is selected the default equation

  2.443  10 2 e 0.0007f c

is used which is based on a curve fit to experimental data. The units of frequency are defined using Freq Units.

480


Figure B-245 Modal Damping Characteristic Panel A tracer and an associated Co-ordinates Panel can be displayed and hidden by positioning the mouse on the graph and using the right mouse to display a pop-up menu, which has Show Tracer and Hide Tracer options. The frequencies at which the response is calculated may be defined in one of three ways: 1. A forced response load case is defined corresponding to each selected load case with the response calculated at the harmonic frequencies between the minimum and maximum frequencies defined by Min Frequency and Max Frequency. To obtain the loading in this format the Order Tracking toggle should be de-activated as shown in Figure B-243. 2. A forced response load case is defined corresponding to each selected load case with the response calculated at the engine order frequencies defined using the Order Tracking Panel shown in Figure B-246 and setting the Loadcase option menu to for each Speed.

481


Figure B-246 Order Tracking Panel This panel is displayed by selecting the Order Tracking toggle and selecting the Define button adjacent to Orders. Engine orders are then entered using the table. Rows are added to and deleted from the table using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and Delete Row options. If the engine is a 4-stroke engine then the orders must be divisible by 0.5, otherwise the orders must be integer values. 3. A forced response loadcase is defined corresponding to each selected engine order defined using the Order Tracking Panel shown in Figure B-246 and setting the Loadcase option menu to for each Order. The response is calculated at the engine order frequencies. The Order Tracking Panel is displayed and the engine orders are defined as described above for option 2. 7.4.2.2 Acoustic Analysis The Acoustic tab of the Frequency Response and Acoustic Analysis Panel shown in Figure B-247 is used to define parameters for the radiated noise calculation as described in Appendix 8. This tab can only be edited if Radiated Noise has been selected on the Cylinder Block FEA Output Panel.

Figure B-247 Frequency Response and Acoustic Analysis Panel Method

482

HELP Cylinder Block Analysis This is used to specify the method for calculating sound intensity and radiated sound power. The available methods, which are described in detail in Appendix 8 are Rayleigh, Helmholtz and Indirect BEM. The acoustic tab contains four child tabs: Boundary, Acoustic Mesh, Fluid and Options. 7.4.2.2.1

Boundary

Figure B-248 Acoustic Analysis Boundary Tab Radiating Set The face sets defining each radiating surface are selected by using the Select button, which displays the Face Sets Panel shown in Figure B-249.

Figure B-249 Face Sets Panel Since the Rayleigh integration is only applicable to flat plates the selected face sets should define radiating surfaces that are generally flat, for example, each side of the cylinder block. It is preferable wherever possible to define sets which are the complete sides of the cylinder block rather than dividing each side up into multiple sets. The nodes of each face set are added to the node set called VIBRATION and is to the .SFE containing the cylinder block when the calculation is performed. The radiated sound power is calculated for each set. The total radiated sound power contribution from a number of radiating surfaces can be calculated and plotted as described below in 7.4.2.4

483

HELP Cylinder Block Analysis 7.4.2.2.2

Acoustic Mesh

Figure B-250 Acoustic Mesh Tab Mesh This is used to define the acoustic mesh. It may be None, Imported or Defined. If set to imported then the mesh is read from an FE model, which is selected by pressing the Browse button which in turn pops up the Model Transformation and Translation Panel shown in Figure B-26. If set to Defined then it’s possible to define a mesh manually, as described in section 7.4.2.2.2.1. 7.4.2.2.2.1 Acoustic Mesh Definition

Figure B-251 Acoustic Mesh Definition File Name This is used to define the name of the SFE file that will contain the acoustic mesh. The name may be typed in directly or selected via a standard file browser accessed by pressing the Browse button. Shape This is used to define the shape of the defined mesh. It may be one of None, Sphere, Hemisphere, Cylinder, Pane or Box. Depending on the shape chosen, different elements in the child tabs will require values to be entered. When the details have been entered, pressing the Generate button generates the SFE file. (a)

Centre

484


Figure B-252 Acoustic Mesh Centre This is used to define the centre coordinate of the Sphere, Hemisphere, Cylinder, Pane or Box. (b)

Radius

Figure B-253 Acoustic Mesh Radius This is used to define the radius of the Sphere, Hemisphere or Cylinder. (c)

Normal

Figure B-254 Acoustic Mesh Normal This is used to define the normal to the Pane. (d)

Vector

Figure B-255 Acoustic Mesh Vector This is used to define a normal to the cut-plane of the Hemisphere (pointing into the mesh side), the axis of the Cylinder, or the orientation of the first edge of a Pane.

485

HELP Cylinder Block Analysis (e)

Length

Figure B-256 Acoustic Mesh Length This is used to define the length of the Cylinder, the length and width of the Pane, or the length, width and height of the Box. 7.4.2.2.3

Fluid

Figure B-257 Acoustic Mesh Fluid Tab Fluid Density This is used to specify the density of the fluid (normally air) in which the noise is being calculated. The default value is 1.29 kg/m3. Speed of Sound This is used to specify the speed of sound (normally air) in which the noise is being calculated. The default value is 344 m/s2. 7.4.2.2.4

Options

Figure B-258 Acoustic Mesh Options Tab Integration Increment This is used to specify an integer number defining by how much the integration order is to be increased or decreased and applies to the Helmholtz method only. Increasing the order will improve the accuracy of the results but will also increase the computation effort. Increasing the integration order by 1 is likely to double the execution time. The

486

HELP Cylinder Block Analysis default integration orders have been chosen to give an accuracy of approximately 3 significant figures in the linear pressure and intensity results. Delete Vibrations If this is selected the nodal vibrations are calculated prior to the acoustic analysis for each load case and are then deleted from the .SFE file once the analysis has been completed. This option is recommended, especially when the calculation is performed on a large number of load cases, since the size of the .SFE file can grow very large if the vibration results remain in the file. If this option is not selected the nodal vibrations are first calculated for all the selected load cases before the acoustic analyses are performed. 7.4.2.3 Output The Output tab of the Frequency Response and Acoustic Analysis Panel shown in Figure B-259 is used to define the output file names and the format of the output for the frequency response and acoustic analysis.

Figure B-259 Frequency Response and Acoustic Analysis Panel Output Name This is used to specify the name of the .FPO command file and any analysis data file that is written. Loading To This is used to specify the output format of the frequency response loading and can only be selected when Frequency Response Loading has been selected on the Cylinder Block FEA Output Panel. The options are FeaRCE, RICARDO-SFE, MSC-NASTRAN and IDEAS. If FeaRCE is selected the loads will only be written to an .FRL file. If RICARDO-SFE is selected the loads will be written to the .SFE file containing the finite element model of the cylinder block, but no loads will be written to an analysis data file. If MSC-NASTRAN is selected two analysis data files will be written. The first file will have a _modes.dat suffix and will contain the model geometry and data to perform a modal solution and which will create a database file when NASTRAN is executed. The second file will have a _forcing.dat suffix and will contain the loading and parameters to

487

HELP Cylinder Block Analysis perform an analysis restart using the database file to allow a frequency response analysis to be performed. Results to This is used to specify the output format of the results and can only be selected when Nodal Vibrations or Radiated Noise have been selected on the Cylinder Block FEA Output Panel. The options are FeaRCE, RICARDO-SFE, IDEAS, FEMVIEW and SYSNOISE. If FeaRCE is selected only the .FPO command file together with the .FRL file containing the loading will be written and no vibrations or radiated noise will be calculated. If any of the other options are selected the vibrations and radiated noise (if selected) are calculated and written to the .SFE file. If IDEAS or FEMVIEW are selected the results are also written to data files so that nodal vibrations and sound intensities on the finite-element mesh can be viewed using either of these two finite element packages. If IDEAS is selected a Universal file is written; if FEMVIEW is selected the results are written directly to a FEMVIEW binary file. The option SYSNOISE can only be selected if Nodal Vibrations have been selected as the chosen output. If this option is selected two SYSNOISE free format files will be written. The first file will have a .sysnoise suffix and will contain the mesh of the vibration set, whilst the second file will have a .sysnoise.disps suffix and will contain the nodal vibration results for use as a boundary condition within SYSNOISE. Export This is used to specify whether the nodal vibration and radiated noise results are written to an exported .SFE file. The options are None, Sets and External. If None is selected no export is performed and the results are written to the .SFE file selected using the Cylinder Block Model Panel. In this case it is recommended that the original file be copied before performing the analysis so that this file can be used again. If Sets is selected the vibration and radiating sets for which results are required together with the loaded nodes will be exported to a file called basename_SUBSET.SFE, where basename is the name of the .SFE file selected using the Cylinder Block Model Panel.. If External is selected the external surfaces of the model together with the loaded nodes are exported to a file a called basename_EXTERNAL.SFE. The exported file will contain the modal frequencies and mode shapes as well as the nodal vibration and radiated noise results, and will be significantly smaller than the original file. For this reason it is strongly recommended that either of the export options are used. 7.4.2.4 Generating and Plotting Results Once the calculation parameters have been defined using the Frequency Response and Acoustic Analysis Panel the Generate Output button is selected on the Cylinder Block FEA Panel to perform the calculation. Vibration and noise results can then be plotted using the Ricardo program VIBPLOT 2

7.4.3

Quasi-Static Analysis

The Define Output button on the Cylinder Block FEA Panel is used to display a panel for defining calculation parameters. If Quasi-Static Unit Loads, Quasi-Static Unit Load Factors, Quasi-Static Combined Results, or Quasi-Static Loading have been selected on

488

HELP Cylinder Block Analysis the Cylinder Block FEA Output Panel the Cylinder Block Quasi-Static Analysis Panel shown in Figure B-260 is displayed. This panel can only be displayed once the output and model have been selected. Once the calculation parameters have been defined the Generate Output button is selected to perform the calculation.

Figure B-260 Cylinder Block Quasi-Static Analysis Panel

7.4.3.1 Applying Quasi-Static Unit Loads The quasi-static unit loads, as part of the Unit Load Method described in 7.3.3, are defined using the Cylinder Block Quasi-Static Analysis Panel as shown in Figure B-260. Q-S Loading To This is used to specify the output format of the loading. The options are FeaRCE, RICARDO-SFE, MSC-NASTRAN and IDEAS. If FeaRCE is selected the unit loads will only written to the .FPR Command File as described in 7.3.3.1 No loads will be written to the .SFE file containing the finite element model of the cylinder block or to an analysis data file. If RICARDO-SFE is selected the loads will be written to the .SFE file, but no loads will be written to an analysis data file. Start Loadcase This is used to define the load case number at which the loads are to start and defaults to n  1 where n is the maximum load case number already stored in the .SFE file. Selecting the Default button will restore the default value. Moments Included If this is selected unit moment bearing loads are apply to each main bearing. If this is not selected it will only be possible to post-process statically-determinate solution load cases. It is recommended that this be selected so that all load cases can subsequently be post-processed. Output Name

489

HELP Cylinder Block Analysis This is used to specify the name of the .FPR command file and any analysis data file that is written. The other options and data fields that are shown Figure B-260 are ghosted and are only relevant when defining parameters for the calculation of stresses and fatigue safety factors as described below or when writing quasi-static loading using the Direct Method as described in 7.4.3.3. 7.4.3.2 Calculation of Displacements, Stresses and Fatigue Safety Factors

Figure B-261 Cylinder Block Quasi-Static Analysis Panel The Cylinder Block Quasi-Static Analysis Panel as shown in Figure B-261 is used to define parameters for calculating quasi-static displacements, stresses and fatigue safety factors using the Unit Load Method described in 7.3.3. This panel can only be displayed once the load cases for which results are required have been selected using the Select Loadcases Panel described in 5.2. Crank Angle The crank angles at which quasi-static displacements and stresses are calculated are defined using the Select button, which displays the Crank Angle Panel shown in Figure B-262.

490


Figure B-262 Crank Angle Panel Definition This is used to define how the crank angles are selected. The options are Using Interval or Defined List. If Using Interval the crank angles are defined using Interval and Offset as described below. If Defined List is selected crank angles are entered using the table. Rows are added and deleted using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and Delete Row options. Interval This is used to specify the crank angle interval at which the quasi-static displacements and stresses are calculated. This value defaults to the interval at which results were stored during the solution as defined by Print on the Evaluate Solution Panel. This default may be prohibitive in terms of the time required to calculate the combined displacements and stresses and to evaluate the fatigue safety factors, and also in terms of disk usage. Intervals of 5 or 10 degrees CA are more typical. Offset This is used to specify a crank angle offset and corresponds to the first angle at which the quasi-static displacements and stresses are calculated. This is useful, for example, for resolving the crank angle at which the maximum gas cylinder pressure occurs through the engine cycle. Q-S Results To This is used to specify the output format of the results and can only be selected if QuasiStatic Combined Results has been selected on the Cylinder Block FEA Output Panel. The options are RICARDO-SFE, IDEAS and FEMVIEW. If RICARDO-SFE is selected the results are written to the .SFE file only. If IDEAS or FEMVIEW are selected the results are also written to data files so that nodal displacement, stress and fatigue safety factor distributions on the finite-element mesh can be viewed using either of these two finite element packages. If IDEAS is selected a Universal file is written; if FEMVIEW is selected the results are written directly to a FEMVIEW binary file.

491

HELP Cylinder Block Analysis Export This is used to specify the extent of the model to which the results are written. The options are Sets, External and All. If Sets is selected the elemental face sets for which results are required will be exported to a file called _SUBSET.SFE where is defined by Output Name. The sets are selected using the Select button, which displays the Face Sets Panel shown in Figure B-249. Any sets at which stress correction factors have been defined using the Notch Sensitivity Panel, as described below, will also be included. If External is selected the external surfaces of the model are exported to a file a called _EXTERNAL.SFE. If All is selected the whole model is exported to a file a called .SFE. If this name conflicts with the name of the .SFE file selected using the Cylinder Block Model Panel then the model is exported to a file called _COPY.SFE. The exported file will contain the unit loads, displacements and stresses as well as the quasi-static displacement, stress and fatigue safety factor results. It is strongly recommended that either Sets or External be selected since the file will be significantly smaller than the original file. Thermal Load This is used to specify whether a thermal load case is to be combined with the operating load cases. The supplied load case number must correspond to a load case with temperature loads. Assembly Loads This is used to specify whether any assembly load cases, defining for example bolt and bearing crush loads, are to be combined with the operating load cases. The assembly loads are defined by selecting the Define button, which displays the Assembly Loads Panel shown in Figure B-263.

Figure B-263 Assembly Loads Panel The assembly load case numbers and their associated factors are added to the table. Rows are added to and deleted from the table using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and Delete Row options. Notch Sensitivity This is used to specify whether any locations of the model are notch sensitive. These locations are defined by selecting the Define button, which displays the Notch Sensitivity Panel shown in Figure B-264. The treatment of notch sensitivity is described in

492

HELP Cylinder Block Analysis Appendix 5. Although this appendix considers the treatment in crankshaft analysis the principles are also relevant for other analyses.

Figure B-264 Notch Sensitivity Panel Each notch sensitive area of the model is defined by a node set which has a stress correction factor and material associated with it. The stress correction factor is defined by Stress Correction Factor equation as described in Appendix 5. The node set names, factors and material names are entered in the table columns as shown. Rows are added to and deleted from the table using the right mouse button with the mouse positioned over the row number. This displays a pop-up menu with Insert and Delete options. Selecting the right mouse button whilst editing the Set Name column without the cell highlighted will display a pop-up menu with a Select Node Set option. Selecting this option will display the Node Set Panel shown in Figure B-265 from which a set can then be selected from the displayed list of node sets in the .SFE file containing the cylinder block model.

Figure B-265 Node Set Panel Selecting the right mouse button whilst editing the Material column without the cell highlighted will display a pop-up menu with a Select Material File option. Selecting this option will display the Material File Panel shown in Figure B-266.

493


Figure B-266 Material File Panel This is used to select a material file as defined in 3.4.1.13.1 containing the material properties corresponding to the selected node set. The Filter unless changed by the user will contain the same path name of the database directory containing properties of a library of materials. Safety Calculation This is used to specify whether fatigue safety factors are to be calculated. The safety factor calculations to be performed are defined using the Select button, which displays the Safety Calculation Panel as shown in Figure B-267.

Figure B-267 Safety Calculation Panel Any number of safety factor calculation types can be selected, each of which are described in detail in Appendix 7. Output Name This is used to specify the name of the .FPO command file and any other file as defined by Q-S Results To that is written. The other options and data fields shown in Figure B-261 are ghosted and are only relevant when defining parameters for applying quasi-static unit loads as described above or when writing quasi-static loading using the Direct Method as described in 7.4.3.3.

494

HELP Cylinder Block Analysis Once the calculation parameters have been defined successfully the Generate Output button is selected to perform the calculation. The quasi-static unit loads are read from the .SFE file containing the cylinder block model and are displayed in the Cylinder Block Unit Loads Panel as shown in Figure B-268. If there are any additional loads other than the unit loads defined by the user that have the same resultant vector then these will be ignored provided they have a loadcase number higher than the unit load.

Figure B-268 Cylinder Block Unit Loads Panel This lists the load case number and force and moment vectors for the unit loads as described in 7.3.3.1 at each modelled main journal bearing, cylinder and thrust bearing. The Query Panel of Figure B-269 is also displayed.

495


Figure B-269 Query Panel This gives the user the option, by answering Yes of calculating the unit load factors, and if Quasi-Static Combined Results has been selected the quasi-static displacements and stresses and fatigue safety factors. If the user replies with No, no output is generated and the calculation is stopped. 7.4.3.3 Quasi-Static Loading The quasi-static loads, as part of the Direct Method described in 7.3.4, are defined using the Cylinder Block Quasi-Static Analysis Panel as shown in Figure B-270. This panel can only be displayed once the load cases for which loads are required have been selected using the Select Loadcases Panel described in 5.2.

Figure B-270 Cylinder Block Quasi-Static Analysis Panel Crank Angle The crank angles at which loading are written are defined using the Select button which displays the Crank Angle Panel shown in Figure B-271.

496


Figure B-271 Crank Angle Panel Definition This is used to define how the crank angles are selected. The options are Using Interval or Defined List. If Using Interval the crank angles are defined using Interval and Offset as described below. If Defined List is selected crank angles are entered using the table. Rows are added and deleted using the right mouse button with the mouse positioned over the row. This displays a pop-up menu with Insert Row and Delete Row options. Interval This is used to specify the crank angle interval at which quasi-static loads are applied to the model. This value defaults to the interval at which results were stored during the solution as defined by Print on the Evaluate Solution Panel. This default may be prohibitive in terms of the time required to calculate the combined displacements and stresses and to evaluate the fatigue safety factors, and also in terms of disk usage. Intervals of 5 or 10 degrees CA are more typical. Offset This is used to specify a crank angle offset and corresponds to the first angle at which the quasi-static loads are applied to the model. This is useful, for example, for resolving the crank angle at which the maximum gas cylinder pressure occurs through the engine cycle. Q-S Loading To This is used to specify the output format of the loading. The options are FeaRCE, RICARDO-SFE, MSC-NASTRAN and IDEAS. If FeaRCE is selected the unit loads will only written to the .FPR Command File as described in 7.4.3.3. No loads will be written to the .SFE file containing the finite element model of the cylinder block or to an analysis data file. If RICARDO-SFE is selected the loads will be written to the .SFE file, but no loads will be written to an analysis data file. Start Loadcase This is used to define the load case number at which the loads are to start and defaults to n  1 where n is the maximum load case number stored in the .SFE file already. Selecting the Default button will restore the default value. Output Name

497

HELP Cylinder Block Analysis This is used to specify the name of the .FPR command file and any analysis data file that is written. The other options and data fields that are shown are ghosted and are only relevant when defining parameters for applying unit loads and calculating stresses and fatigue safety factors using the Unit Load Method as described in 7.4.3.1 and 7.4.3.2. 1

Ricardo Report DP01/2245 FEPRE and FEPOS - Revision 5.5 – Finite Element Pre- and PostProcessor’.

2

Ricardo Report DP01/2330 ‘VIBPLOT Revision 2.0.4 Noise and Vibration Plotting Utility Program’.

498

HELP Additional Post-Processing

8 Additional Post-Processing 8.1 Overview Quasi-static loads predicted by ENGDYN can be applied to finite element models of the piston and connecting rod to perform stress analysis of those components. These postprocessing options are not currently available using the Graphical User Interface, but are available using an .EDPO command file that is described in this section. ENGDYN is executed by entering the following command on the command line: engdyn -post .EDPO

8.2 Quasi-Static Analysis of a Piston 8.2.1


8.2.1.1 Geometry and Restraints The finite-element model can be of the complete piston or half the piston with a plane of symmetry in the thrust plane. The model will include the piston pin which is restrained as appropriate to prevent rigid body motion and which is connected to the piston with gap elements. At present the restraint and gap definitions are not written by ENGDYN. The mass of the model should match the mass of the piston defined using the Cylinder. 8.2.1.2 Set Definition Two elemental face sets are required for applying the loads, these are pistonGasFace and pistonSkirt. The set pistonGasFace defines the area over which the cylinder gas pressure acts and will include the circumferential portion of the piston above the top ring. The set pistonSkirt defines the piston skirt. The extent of this set should agree with the values of Top of Skirt and Bottom of Skirt specified using the Cylinder. An additional set called pistonPinJournal is required that defines the outer surface of the piston pin and is used to test the location of the model.

8.2.2

Applying the Quasi-Static Loads

Loads can be applied at any number of crank angles and load cases. Three loads are applied at each crank angle. These are the cylinder gas pressure load, the skirt load and a body load based on the nominal acceleration of the piston. A parabolic distributed load is applied to the piston skirt unless the In-Cylinder Model, defined using the Model Definitions Panel, is set to Secondary Dynamic, in which case the pressures from the PISDYN solution are mapped on to the FE model.

499

HELP Additional Post-Processing The loads applied to the piston model can be plotted using the Plot Results Panel and selecting Piston as the Model type. Figure B-272 shows an .EDPO command file to write a FEARCE .FPR file, which will then apply loads to the finite element model. A copy of this .EDPO file is included as part of the installation in ../Ricardo/engdyn/3.0/examples/piston_analysis. OPEN

DATA=.EDSF

;Name of EDSF file containing results

PISTON

LOAD=1 CYLINDER=1 OUTPUT= FILE= FROM=ABAQUS UNIT=MM SOLVER=ABAQUS ROTATE=0.0,0.0,0.0 TRANSLATE=0.0,0.0,0.0 MIRROR=0.0,0.0,0.0 INTERVAL=10.0 OFFSET=5.0

; ; ;

The INTERVAL and OFFSET arguments may be replaced with a CRANK_ANGLE argument that requires a list of angles w.r.t. TDC for any given cylinder. CRANK_ANGLE=0.0,180.0,270.0,570.0

\ \ \ \ \ \ \ \ \ \ \

;List of ENGDYN loadcases ;List of cylinder numbers ;Name of output files ;FE model file name ;FE model format (SFE,ABAQUS,ANSYS,IDEAS,FEMGEN,PAFEC,MSC) ;Units of FE model (Not required if FROM=SFE or IDEAS) ;FE solver (FEARCE(VSS),MSC,ABAQUS,IDEAS,ANSYS) ;Rotation, translation and mirror vectors to transform FE ;model to ENGDYN local coordinate system w.r.t. cylinder ;where the origin is at the centre of the small end ;Crank angle interval at which loading is required ;First angle w.r.t. TDC at which loading is required

Figure B-272 EDPO Command File for Piston Analysis Two commands are required in the .EDPO file; an OPEN to open the .EDSF file and a PISTON Command to apply the loads. The .FRC file written by ENGDYN will either write an analysis data file for a 3rd party FE solver such as MSC/NASTRAN or will run a solution using the FEARCE VSS solver.

8.3 Applying Loads to a Finite Element Model of a Connecting Rod 8.3.1


8.3.1.1 Geometry and Constraints The finite-element model can be of the complete connecting rod or half the connecting rod with a plane of symmetry in the thrust plane. The model may also include the piston pin. If the pin is included then density of the piston pin should be negligible so that the mass of the pin is excluded as part of the mass of the connecting rod. There are two different options dependent on whether EHL derived loads are being applied to the model. Option 1: Applying non-EHL derived loads In this case bearing loads are applied to the piston bearing journal surfaces of the piston pin, with the big end bearing restrained to ground. The pin is restrained as appropriate to prevent rigid body motion of the connecting rod about the big end centre and is

500

HELP Additional Post-Processing connected to the connecting rod with gap elements. The big end of the connecting rod is gapped to ground. Option 2(a): Applying EHL derived loads at the Big End Bearing In this case the calculated EHL oil film and asperity pressures are mapped from the lubrication mesh onto the face set bigEndBearing defining the bearing surface. If the model includes the piston pin then the pin is restrained at the piston bearing journal surfaces of the pin and the pin is connected to the connecting rod with gap elements. If the model excludes the piston pin, that is as the model is supplied for the EHL analysis for the ENGYDN solution, then the model is restrained in all directions at the node set smallEndBearing. Option 2(b): Applying EHL derived loads at the Small End Bearing In this case the calculated EHL oil film and asperity pressures are mapped from the lubrication mesh onto the face set smallEndBearing defining the bearing surface. In this case the piston pin is not included. The model is restrained in all directions at the node set bigEndBearing. At present the restraint and gap definitions are not written by ENGDYN. The mass of the model should match the mass and inertia of the connecting defined using the Cylinder Panel. 8.3.1.2 Set Definitions If the piston pin is included in the model then two face sets are required for applying the loads, these are pistonPinJournal:ID_front and pistonPinJournal:ID_rear. These sets represent the piston-bearing surface of the piston pin. Additional sets called bigEndBearing, smallEndBearing and pistonPinJournal are required to check the location of the model. The sets bigEndBearing and smallEndBearing define the bearing surface of the big and small end bearing respectively, whilst the set pistonPinJournal defines the outer surface of the piston pin.

8.3.2

Applying the Quasi-Static Loads

Loads can be applied at any number of crank angles and load cases. Two loads are applied at each crank angle. These are bearing loads and a body load based on the nominal acceleration of the connecting rod. If non-EHL loads are being applied to the connecting rod then a parabolic distributed load is applied to the piston pin journal to represent the bearing load. The loads applied to the connecting rod model can be plotted using the Plot Results Panel and selecting Connecting Rod as the Model type.

501

HELP Additional Post-Processing Figure B-273 shows an .EDPO command file to write a FEARCE .FRC file, which will then apply loads to the finite element model. A copy of this .EDPO file is included as part of the installation in ../Ricardo/engdyn/4.0/examples/conrod_analysis. OPEN

DATA=.EDSF

;Name of EDSF file containing results

CONROD

LOAD=1 CYLINDER=1 OUTPUT= FILE= FROM=ABAQUS UNIT=MM SOLVER=ABAQUS ROTATE=0.0,0.0,0.0 TRANSLATE=0.0,0.0,0.0 MIRROR=0.0,0.0,0.0 INTERVAL=10.0 OFFSET=5.0

; ; ;

The INTERVAL and OFFSET arguments may be replaced with a CRANK_ANGLE argument that requires a list of angles w.r.t. TDC for any given cylinder. CRANK_ANGLE=0.0,180.0,270.0,570.0

\ \ \ \ \ \ \ \ \ \ \

;List of ENGDYN loadcases ;List of cylinder numbers ;Name of output files ;FE model file name ;FE model format (SFE,ABAQUS,ANSYS,IDEAS,FEMGEN,PAFEC,MSC) ;Units of FE model (Not required if FROM=SFE or IDEAS) ;FE solver (FEARCE(VSS),MSC,ABAQUS,IDEAS,ANSYS) ;Rotation, translation and mirror vectors to transform FE ;model to ENGDYN local coordinate system w.r.t. cylinder ;where the origin is at the centre of the big end ;Crank angle interval at which loading is required ;First angle w.r.t. TDC at which loading is required

Figure B-273 EDPO Command File for Connecting Rod Analysis Two commands are required in the .EDPO file; an OPEN to open the .EDSF file and a CONROD Command to apply the loads. The .FRC file written by ENGDYN will either write an analysis data file for a 3rd party FE solver such as MSC/NASTRAN or will run a solution using the FEARCE VSS solver. The latter is only appropriate if the analysis is linear. The arguments for the CONROD Command are: LOADCASE CYLINDER OUTPUT FILE FROM

SOLVER UNIT INTERVAL OFFSET

This is the list of ENGDYN load cases that are to be applied to the model. This is the cylinder number of the connecting rod to be loaded. This is the name of the output file. This is the file name of the FE model of the connecting rod to be loaded. This is format of the FE model file. The options are SFE, ABAQUS, ANSYS, IDEAS, FEMGEN, PAFEC and NASTRAN. Default = SFE This is the output format of the loading data. The options are FEARCE(VSS), NASTRAN, ABAQUS, IDEAS or ANSYS. The default is FEARCE(VSS) This is the unit definition of the file containing the FE model. The options are SI, MM, METRIC, FT and INCH. Default = SI This is the crank angular interval at which the data is written. This is a crank angle offset and corresponds to the first angle at which the quasi-static loads are applied. This is useful, for example, for resolving the crank angle at which the maximum gas cylinder pressure occurs through the

502

HELP Additional Post-Processing

CRANK_ANGLES ROTATE TRANSLATE MIRROR STARTING_LOADCASE ASSEMBLY LOADCASE

HALF_MODEL NUMBER_OF_CYCLES

engine cycle. Default = 0.0 This is a list of crank angles at which loads are required and is an alternative to INTERVAL and OFFSET This is a rotation vector to define the transformation of the SFE to the ENGDYN coordinate system. Default=0.0,0.0,0.0 This is a translation vector to define the transformation of the SFE to the ENGDYN coordinate system. Default=0.0,0.0,0.0 This is a mirror vector to define the transformation of the SFE to the ENGDYN coordinate system. Default=0.0,0.0,0.0 This is the starting load case number at which the loads are written. Default = 0 This is an assembly load case number in the SFE name. If the analysis is linear and this is specified then FEARCE COMBINE commands are written to combine the assembly loads with the applied mechanical loads. Default = 0 Switch to denote whether the connecting rod model to be loaded is a half model. Options are YES and NO. Default = NO This is the number of cycles of loading data that are written. More than one cycle is required for a non-linear analysis where there may be slippage between any defined contact surfaces. A number of cycles is required in this case for the non-linear solution to converge. Default =1

503

COMMAND FILES Additional Post-Processing

C. COMMAND FILES The ASCII command files can be used to build, run and post-process ENGDYN models via the command line. The command files can be generated from model built in the ENGDYN GUI by selecting “File->Generate” or “File->Generate As”. This will produce an .EDPR file, an .EDSO file and, if crankshaft stress analysis postprocessing has been setup, an .EDPO file. The files can also be generated from the command line by running: engdyn –gen model.EDSF where model should be substituted with the appropriate model name. The EDPR file is processed by ENGDYN by running: engdyn –pre model.EDPR The EDSO file is processed by ENGDYN by running: engdyn –solve model.EDSO The EDPO file is processed by ENGDYN by running: engdyn –post model.EDPO

505

COMMAND FILES Pre-Processor Commands

1

Pre-Processor Commands

This program is used to define the component sub-models for the SOLVER. The program requires a command file with the suffix .EDPR and creates a standard data file with the suffix .EDSF. Those commands that are dependent on the previous execution of other commands are clearly stated.

506


1.1 AXES The AXES command is used for the rotation of the local coordinate system. This command broadens the capabilities of the LINK command. Pre-Processor Commands has the following arguments: AXES

NUMBER

=

Axes rotation identifier

ROTATIONS

=

Rotations along axes

where: NUMBER This argument is used to identify the local coordinate system. Non-defaultable Integer ROTATIONS This argument is used to define the rotation of axes of the local coordinate system. Default=0,0,0 Real deg

507


1.2 BEARING The command BEARING is used to define a journal or thrust bearing. This command should be used for each of the main bearings, crankpin bearings and thrust bearing of the engine and should appear after the NODE command. BEARING has the following arguments: BEARING

TYPE NUMBER MODEL FEED_FROM JOURNAL_NODE BEARING_NODE LENGTH DIAMETER RADIAL_CLEARANCE JOURNAL_OILHOLE_DIAMETER JOURNAL_OILHOLE_POSITION JOURNAL_OILHOLE_OFFSET BEARING_OILHOLE_DIAMETER BEARING_OILHOLE_POSITION BEARING_OILHOLE_OFFSET JOURNAL_GROOVE_WIDTH JOURNAL_GROOVE_EXTENT JOURNAL_GROOVE_OFFSET BEARING_GROOVE_ANGLE BEARING_GROOVE_WIDTH BEARING_GROOVE_EXTENT BEARING_GROOVE_OFFSET BEARING_GROOVE_ANGLE MESH

= = = = = = = = = = = = = = = = = = = = = = = =

MATERIAL

=

FRICTION_COEFFICIENT WEAR_COEFFICIENT JOURNAL_PROFILE BEARING_PROFILE SHAFT_SPEED_RATIO BEARING_STIFFNESS

= = = = = =

AXIAL_CLEARANCE

=

THRUST_FACE_DIAMETER STIFFNESS

= =

DAMPING

=

AXIAL_SYMMETRY BEARING_EXTENT

= =

where:

508

Type of bearing Bearing number Oil model type Bearing feeding the oil Crankshaft journal node Bearing node Bearing length Bearing diameter Radial clearance Journal oilhole diameter Journal oilhole position Journal oilhole offset Bearing oilhole diameter Bearing oilhole position Bearing oilhole offset Journal groove width Journal groove extent Journal groove offset Journal groove angle Bearing groove width Bearing groove extent Bearing groove offset Bearing groove angle Lubrication mesh size Bearing and journal material name names Friction coefficient Wear coefficient Journal profile Bearing profile Shaft speed ratio Bearing stiffness Axial clearance of thrust bearing Thrust bearing face diameters Stiffness of the thrust bearing Damping applied to the thrust bearing Axial symmetry flag Circumferential extent of the bearing


TYPE This argument is used to specify the type of bearing. The available options are: MAIN BIG_END SMALL_END BALANCER_SHIFT CAMSHAFT OTHER Non-defaultable Text NUMBER This argument is used to specify the bearing number or name. A number should be supplied if the bearing type has been specified as MAIN, BIG_END or SMALL_END using the TYPE command. Non-defaultable Text MODEL This argument is used to specify the oil model type. The options available are: MOBILITY HYDRODYNAMIC ELASTOHYDRODYNAMIC Default=MOBILITY Text FEED_FROM This argument specifies where the oil feed is located. The options available are: NONE BEARING MAINS JOURNAL Non-Defaultable Text JOURNAL_NODE This argument is used to specify the crankshaft node of the journal corresponding to this bearing. See Note 1. Non-defaultable Integer BEARING_NODE This argument is used to specify the crankshaft node of the bearing. See Note 1. Non-defaultable Integer LENGTH This argument is used to specify the bearing length. This input can be either a single number or a pair of values specifying upper and lower bearing length. The latter is applicable only for small end bearings (TYPE is set to SMALL_END) and enables the definition of keystone-shaped small end bearings. If a pair is specified for other bearing types, the first value is used as the bearing length and the second value ignored. Non-defaultable Real m 509


DIAMETER This argument is used to specify the bearing diameter. Only one diameter per bearing can be set. Non-defaultable Real m RADIAL_CLEARANCE This argument specifies the radial clearance of the bearing. Non-defaultable Real microns JOURNAL_OILHOLE_DIAMETER This argument specifies the diameter of the journal oil holes. The number of journal oilhole diameters, journal oilhole positions and journal oilhole offsets should be identical. Non-defaultable Real mm JOURNAL_OILHOLE_POSITION This argument specifies the radial position of the journal oilholes. The position must be between 0 and 360 degrees. The number of journal oilhole diameters, journal oilhole positions and journal oilhole offsets should be identical. Non-defaultable Real Degrees JOURNAL_OILHOLE_OFFSET This argument specifies the axial offset of the journal oilholes from the axial centre of the journal. The number of journal oilhole diameters, journal oilhole positions and journal oilhole offsets should be identical. Non-defaultable Real mm BEARING_OILHOLE_DIAMETER This argument specifies the diameter of the bearing oilholes. The number of bearing oilhole diameters, bearing oilhole positions and bearing oilhole offsets should be identical. Non-defaultable Real mm BEARING_OILHOLE_POSITION This argument specifies the radial position of the journal oilholes. The position must be between 0 and 360 degrees. The number of bearing oilhole diameters, bearing oilhole positions and bearing oilhole offsets should be identical. Non-defaultable Real Degrees BEARING_OILHOLE_OFFSET This argument specifies the axial offset of the bearing oilholes from the axial centre of the bearing. The number of bearing oilhole diameters, bearing oilhole positions and bearing oilhole offsets should be identical. Non-defaultable Real mm

510


JOURNAL_GROOVE_WIDTH This argument specifies the width of the journal grooves. The number of journal groove widths, journal groove extent pairs, journal groove angles and journal groove offsets should be identical. Non-defaultable Real mm JOURNAL_GROOVE_EXTENT This argument specifies the extent of the journal grooves. The extent of each groove is represented by a pair of angular positions, each of which must be between 0 and 360 degrees. The number of journal groove widths, journal groove extent pairs, journal groove angles and journal groove offsets should be identical. Non-defaultable Real degrees JOURNAL_GROOVE_OFFSET This argument specifies the axial offset of the journal grooves from the axial centre of the journal. The number of journal groove widths, journal groove extent pairs, journal groove angles and journal groove offsets should be identical. Non-defaultable Real mm JOURNAL_GROOVE_ANGLE This is used to define the grooves with an angle. The number of journal groove widths, journal groove extent pairs, journal groove angles and journal groove offsets should be identical. Non-defaultable Real degrees BEARING_GROOVE_WIDTH This argument specifies the width of the bearing grooves. The number of bearing groove widths, bearing groove extent pairs, bearing groove angles and bearing groove offsets should be identical. Non-defaultable Real mm BEARING_GROOVE_EXTENT This argument specifies the extent of the bearing grooves. The extent of each groove is represented by a pair of angular positions, each of which must be between 0 and 360 degrees. The number of bearing groove widths, bearing groove extent pairs, bearing groove angles and bearing groove offsets should be identical. Non-defaultable Real degrees BEARING_GROOVE_OFFSET This argument specifies the axial offset of the bearing grooves from the axial centre of the bearing. The number of bearing groove widths, bearing groove extent pairs, bearing groove angles and bearing groove offsets should be identical. Non-defaultable Real mm

511

COMMAND FILES Pre-Processor Commands BEARING_GROOVE_ANGLE This is used to define the grooves with an angle. The number of bearing groove widths, bearing groove extent pairs, bearing groove angles and bearing groove offsets should be identical. Non-defaultable Real degrees MESH This argument defines the size of the lubrication mesh. The argument’s first value is the number of nodes in the axial direction and the second value is the number of nodes in the circumferential direction. Non-defaultable Integer MATERIAL This argument specifies the name of the bearing and journal materials. The bearing material name is entered first followed by the journal material name. The materials should already have been specified using the MATERIAL command. Non-defaultable Text FRICTION_COEFFICIENT This argument is used to specify the friction coefficient for the contacting surfaces. Non-defaultable Real WEAR_COEFFICIENT This argument is used to specify the wear coefficient. See Note 2. Default= Text JOURNAL_PROFILE This argument is used to define the cold manufactured profiles of the journal when the MODEL argument is set to HYDRODYNAMIC or ELASTOHYDRODYNAMIC. See Note 3. Default=CYLINDRICAL Text BEARING_PROFILE This argument is used to define the cold manufactured profiles of the bearing when the MODEL argument is set to HYDRODYNAMIC or ELASTOHYDRODYNAMIC. See Note 3. Default=CYLINDRICAL Text SHAFT_SPEED_RATIO The shaft speed ratio can be specified for a bearing of TYPE MAIN and must be specified for a bearing of TYPE BALANCER_SHIFT, CAMSHAFT or OTHER. The argument must be set to 1.0 for a bearing of TYPE MAIN. Default=see above Real

512

COMMAND FILES Pre-Processor Commands BEARING_STIFFNESS This argument represents the bearing stiffness. It takes six values which describe the three translational and three rotational stiffnesses. This argument should only be set for a bearing of TYPE MAIN. Default=(0.0,0.0,0.0,0.0,0.0,0.0) Real N.m-1,N.m.rad-1 AXIAL_CLEARANCE This argument specifies the axial clearance of the bearing. See Note 4. Non-defaultable Real Microns THRUST_FACE_DIAMETER This argument specifies the thrust face diameter of the thrust bearing. diameters should be supplied. See Note 4. Non-defautable Real M

Two

STIFFNESS This argument specifies the six contributions to the thrust stiffness. These values should be set if the bearing is a thrust bearing or the MODEL is set to MOBILITY. Non-defautable Real N.m-1, N.m.rad-1 DAMPING This argument specifies the six contributions to the damping. These values should be set if the bearing is a thrust bearing or the MODEL is set to MOBILITY. Non-defautable Real AXIAL_SYMMETRY This argument defines whether the model should have axial symmetry. Default=No Text BEARING_EXTENT This argument defines the circumferential extent of the bearing. The extent is represented by a pair of angular positions, each of which must be between 0 and 360 degrees. Note that this argument is applied only if the feed is from the bearing .(FEED_FROM is set to BEARING). See also note 5. Default=0,360 Real degrees

Notes: 1

One bearing or journal node must be set. No more than one journal and one bearing node can be set.

2

The wear coefficient value is not easily obtained and is a function of lubrication conditions, lubricity and other factors. However, users need not be concerned about the validity of the wear coefficient value, unless that is the focus of their study. Wear coefficient values do not impact the simulation and are used only for post-processing.

513

COMMAND FILES Pre-Processor Commands 3

The profile types can be either: CYLINDRICAL CIRCULAR NON-CIRCULAR If the journal is CYLINDRICAL then is assumed to be a cylinder with a diameter equal to the DIAMETER. If the bearing is CYLINDRICAL then it is assumed to be a cylinder with a diameter equal to the Nominal Diameter plus twice the RADIAL_CLEARANCE. For a CIRCULAR profile, the axial location in mm and the mean diameter in mm at all known locations should be defined. For a NON-CIRCULAR profile the axial location in mm and the profile number in mm at all known locations should be defined.

4

If a bearing of TYPE MAIN has either AXIAL_CLEARANCE or THRUST_FACE_DIAMETER defined it is presumed to be a thrust bearing.

5

BEARING_EXTENT can be used to define partial bearings. Note that the GUI does not provide the option to define bearing extent.

514


1.3 BLOCK The BLOCK command is used to define the cylinder block sub-model. This command should appear after the BEARING, CYLINDER and CRANK commands. BLOCK has the following arguments: BLOCK

TYPE FILE FROM UNITS

= = = =

MATRIX_FORMULATION

=

DATA

=

ROTATE

=

TRANSLATE MIRROR NODE_SET FACE_SET

= = = =

CONSTRAINED_NODE_SET

=

SOLVER

=

NUMBER_OF_MODES

=

ELEMENT_CHECK OUTPUT

= =

MEMORY

=

INCLUDE DATA_RECOVERY

= =

Type of block model File name File format Units of cylinder block model The method for solving the derived solution matrix Type of data which represents the cylinder block Vector rotations about global x,y,z axes Translation in global x,y,z axes Normal vector for mirroring Set of nodes in FE model Set of faces in FE model Set of constrained nodes in FE model FE model solver Number of modes that the FE solver should calculate Element check option Output filename Maximum amount of memory to use of solving FE model The name of SFE file to include Data recovery option

where: TYPE This argument specifies the type of model to use to describe the cylinder block. The options are: DYNAMIC COMPLIANT COMPLIANT_BEARING COMPLIANT_CRANKCASE RIGID Default=DYNAMIC Text

515

COMMAND FILES Pre-Processor Commands FILE This argument is used to specify the name of the file containing the finite element geometry, and reduced matrices of the cylinder block model (also see Note 1 below). The name does not need to contain the suffix which is best omitted. This is only needed if the DATA command specifies FE. Non-defaultable Text FROM This argument specifies which format of the data within the FE model file. The options are: SFE ABAQUS ANSYS IDEAS FEMGEN PAFEC MSC/NASTRAN Default=SFE Text UNITS This argument specifies the units of the data within the FE model file. The options are: SI MM METRIC FT INCH The model will then be translated into SI units Default=SI Text MATRIX_FORMULATION This argument describes the method to solve the solution matrix. The options are: STATIC_CONDENSATION COMPONENT_MODE_SYNTHESIS This option should only be used if a dynamic model has been selected using the TYPE command. Default=COMPONENT_MODE_SYNTHESIS Text DATA This argument specifies from where the data describing the cylinder block geometry will be provided if the TYPE has been set to RIGID. The options are: NONE USER_DEFINED FE If FE is selected an FE model file must be provided. Default=None Text

516

COMMAND FILES Pre-Processor Commands ROTATE This argument is used to specify the vector rotations about the x,y and z crankshaft global axes. For further explanation see Note 2. Default=(0,0,0) Real degrees TRANSLATE This argument is used to specify the translation vector in crankshaft global axes. For further explanation see Note 2. Default=(0,0,0) Real m MIRROR This argument is used to specify the normal vector to the mirror (the plane of which passes through the origin at the crankshaft coordinate system). Default is to not perform any mirroring. For further explanation see Note 2. Default=No mirroring Real NODE_SET This argument can be used to specify a set of nodes. The node numbers should be specified a list corresponding to the node numbers in the FE model. Default=None Integer FACE_SET This argument can be used to specify a set of faces. The face numbers should be specified a list corresponding to the face numbers in the FE model. Default=None Integer CONSTRAINED_NODE_SET This argument can be used to specify a set of constrained nodes. The node numbers should be specified a list corresponding to the node numbers in the FE model. Default=INPUT_2_default_value Type of input_2 Units_of_input_2 SOLVER This argument specifies the solver for the FE model. The options are: FEARCE MSC/NASTRAN ABAQUS If FEARCE is chosen then the FE model will be solved for automatically. Choosing the remaining options will result in an input file for the chosen solver being written to file. Default=FEARCE Text NUMBER_OF_MODES If a dynamic model is chosen using the TYPE, the number of modes for which the FE model should be solved must be given. For all other models this argument should not be set. Non-defaultable Integer

517

COMMAND FILES Pre-Processor Commands ELEMENT_CHECK This option specifies whether an element check should be performed when exporting the results. Default=Yes Text OUTPUT This argument specifies the name and position of the file to export the results of the FE model transformation and calculation to. Non-defaultable Text MEMORY This argument specifies maximum amount of RAM which the calculations should consume. Default=Unlimited Integer Mbytes INCLUDE The name of an SFE file to include. Default=None Text DATA_RECOVERY This argument specifies whether data recovery matrices are to be saved to the output file. Default=No Text

Notes: 1

This command reads the coordinates, node numbers and reduced matrices and freedoms of the cylinder block model. In an SFE file these are stored in the following data sets : NODAL_COORDINATES NODE_NUMBERS REDUCED_MASS_MATRIX REDUCED_STIFFNESS_MATRIX DOF_SET_REDUCED_FREEDOMS

2

The arguments ROTATE, TRANSLATE and MIRROR are used to define a transformation matrix that transforms the cylinder block model to the crankshaft coordinate system. It is assumed that the cylinder block and crankshaft models are orthogonal to one another. The transformation matrix is built up in the following order : Mirror Translate Rotate about x, then y and finally z crankshaft global axes

518


1.4 CHILD The CHILD command is used to define simulation children. Child number zero is always defined as being the parent process. The CHILD command is used in association with the LINK command. CHILD has the following arguments: CHILD

NUMBER

=

Simulation child number

APPLICATION

=

Define child process

HOST

=

Select host to run

ARGUMENTS

=

APPLICATION arguments

FILE

=

Header file for child

ROTATE

=

Rotate child

TRANSLATE

=

Translate child

MIRROR

=

Mirror child

where: NUMBER This is used to specify the number of the Simulation Child number where zero is reserved for the parent process. Non-defaultable Integer APPLICATION This argument is used to define the Simulation Child Process. Non-defaultable Text HOST This argument is used to define the host on which the Application is to run on. At present all the processes have to run from the same host and so this is set to local. Default=local Text ARGUMENTS This argument is used to specify the arguments to the child APPLICATION. For example, if the simulation child is a VALDYN simulation then this would be set to –s .dat. Non-defaultable Text

519

COMMAND FILES Pre-Processor Commands FILE This argument specifies the header file for the child. Non-defaultable Text ROTATE This argument is used to define any necessary rotation from the co-ordinate system used by the APPLICATION to that used by ENGDYN. All vectors are in the ENGDYN global axes. See Note 1 Default=0,0,0 Real deg TRANSLATE This argument is used to define any necessary translation from the co-ordinate system used by the APPLICATION to that used by ENGDYN. All vectors are in the ENGDYN global axes. See Note 1 Default=0,0,0 Real mm MIRROR This argument is used to specify a mirror vector as a normal vector to the mirror, the plane of which passes through the origin at the crankshaft co-ordinate system. If zero is entered for the X, Y or Z values of a transformation, then that transformation will not occur. See Note 1 Default=0,0,0 Real

Notes: 1

The transformation matrix is built up in the following order: mirror, rotation about Z, then Y and finally about X and then the translation.

520


1.5 CONROD The Pre-Processor CommandsCONROD command is used to define the specification of the connecting-rods. Pre-Processor Commands has the following arguments: CONROD

TYPE

=

FILE FROM

= =

UNITS

=

ROTATE

=

TRANSLATE

=

MIRROR SOLVER NUMBER_OF_MODES ELEMENT_CHECK OUTPUT NUMBER

= = = = = =

MEMORY

=

DATA_RECOVERY TOLERANCE

= =

Type of model to use for the connecting rod Filename of FE model FE code which the model Units of the data within the FE model file Rotation vector to apply to the FE model Translation vector to apply to the FE model Plane to reflect the model. FE model solver Number of modes to solve for Element check option Output filename Cylinder identification numbers Maximum amount of memory to use of solving FE model Data recovery option Linear and angular clipping tolerance

where: TYPE This argument specifies the type of model to use to describe the connecting rods. The options are: NONE RIGID COMPLIANT_BIG_END COMPLIANT_SMALL_END COMPLIANT DYNAMIC COMPLIANT_BIG_END_SYMMETRIC COMPLIANT_SMALL_END_SYMMETRIC COMPLIANT_SYMMETRIC RIGID and DYNAMIC are to be used for dynamic calculations only and the remaining options for static calculations. Non-defaultable Text FILE This argument specifies the file containing the FE model of the connecting rod. The full path and filename should be specified. There should be a file specified for each rod. Non-defaultable Text

521


FROM This argument specifies which format of the data within the FE model file. The options are: SFE ABAQUS ANSYS IDEAS FEMGEN PAFEC MSC/NASTRAN Default=SFE Text UNITS This argument specifies the units of the data within the FE model file. The options are: SI MM METRIC FT INCH The model will then be translated into SI units Default=SI Text ROTATE This argument specifies the rotation vector for each connecting rod. There should be a vector for each rod. See Note 1. Default=(0,0,0) Real, Vector degrees TRANSLATE This argument specifies the translation vector for each connecting rod. There should be a vector for each rod. See Note 1. Default=(0,0,0) Real, Vector m MIRROR This argument specifies the mirror vector for each connecting rod. There should be a vector for each rod. See Note 1. Default=No mirroring Real, Vector SOLVER This argument specifies the solver for the FE model. The options are: FEARCE MSC/NASTRAN ABAQUS If FEARCE is chosen then the FE model will be solved for automatically. Choosing the remaining options will result in an input file for the chosen solver being written to file. See Note 2. Default=FEARCE Text 522


NUMBER_OF_MODES If a dynamic model is chosen, the number of modes for which the FE model should be solved must be given. For all other models this argument should not be set. See Note 2. Non-defaultable Integer ELEMENT_CHECK This option specifies whether an element check should be performed when exporting the results. See Note 2. Default=Yes Text OUTPUT This argument specifies the name and position of the file to export the results of the FE model transformation and calculation to. See Note 2. Non-defaultable Text NUMBER This option specifies the identification numbers of the cylinders for which connecting rods are specified. Non-defaultable Integer MEMORY This argument specifies maximum amount of RAM which the calculations should consume. See Note 2. Default=Unlimited Integer Mbytes DATA_RECOVERY This argument specifies whether data recovery matrices are to be saved to the output file. See Note 2. Default=No Text TOLERANCE This argument specifies linear and angular tolerance used for set clipping. It is possible to specify a single value for linear tolerance or a pair of linear and angular tolerances. If the tolerances are not set or set to zero, system tolerances will be applied. Default=(0,0) Text mm, degrees

Notes: 1

The arguments ROTATE, TRANSLATE and MIRROR are used to define a transformation matrix that transforms the cylinder block model to the crankshaft coordinate system. It is assumed that the cylinder block and crankshaft models are orthogonal to one another. The transformation matrix is built up in the following order :

523

COMMAND FILES Pre-Processor Commands Mirror Translate Rotate about x, then y and finally z crankshaft global axes 2

SOLVER, NUMBER_OF_MODES, ELEMENT_CHECK, OUTPUT, MEMORY and DATA_RECOVERY arguments can have either a single value common for each rod or the number of values should equal the number of rods defined by the command. The latter is useful to define models with more than one unique connecting rod.

524


1.6 CRANK The CRANK command is used to define the type of crankshaft model and to extract data from a file containing a finite element model of the crankshaft assembly. This command should appear after the MATERIAL commands. Pre-Processor Commands has the following arguments: CRANK

TYPE

= =

The type of model to represent the crankshaft The name and position of the crankshaft FE model

FILE CALCULATE

=

FROM

=

UNITS

=

OVERWRITE

=

MATRIX_FORMULATION = MATERIAL

=

OUTPUT

=

EXCLUDE_ELEMENTS

=

SOLVER MY MZ MYX MZX

= = = = =

WEB_NUMBERS

=

X

=

RADIUS

=

MASS

=

where:

525

Specifies the properties of the model to be solved for The format of the crankshaft FE model The units of the data describing the FE model Denotes whether to overwrite the SFE output files The method for deriving the reduced mass and stiffness model The name of the crankshaft and crank counterweight material Output filename Elements which are excluded from the FE model Applied matrix solver Required M.y mass moment Required M.z mass moment Required M.y.z mass moment Required M.z.x mass moment The webs to be drilled in the primary balance calculation The axial position of each counterweight The radial position of the counterweights Mass of the balance weights

COMMAND FILES Pre-Processor Commands TYPE This argument specifies the model to calculate the stiffness of the crankshaft. The available options are: DYNAMIC COMPLIANT COMPLIANT_WITH_NO_MASS RIGID RIGID_WITH_NO_MASS Default=DYNAMIC Text FILE This argument specifies the file containing the FE model of the crankshaft. The full path and filename should be specified. Non-defaultable Text CALCULATE This argument defines which properties of the crankshaft should be calculated. The options are: NONE MASS_PROPERTIES STIFFNESSES ALL The ALL argument will indicate that both mass properties and stiffnesses of the crankshaft should be calculated. Default=NONE Text FROM This argument specifies which format of the data within the FE model file. The options are: SFE-RICARDO ABAQUS ANSYS IDEAS FEMGEN PAFEC MSC/NASTRAN PATRAN Default=SFE-RICARDO Text UNITS This argument specifies the units of the data within the FE model file. The options are: SI MM METRIC FT INCH The model will then be translated into SI units Default=SI Text OVERWRITE Denotes whether to overwrite the SFE output files if they exist 526

COMMAND FILES Pre-Processor Commands Default=No

Text

MATRIX_FORMULATION This argument is used to specify the method for deriving the reduced mass and stiffness model of the crankshaft. .At present only one method is available. Default= HODGETTS Text MATERIAL This argument specifies the name of the crankshaft and crank counterweight materials. One or two materials can be specified. If only one is specified then the crankshaft and crank counterweight material will be set to that material name. If two are specified the first and second names will be taken as the crankshaft and the crank counterweight materials respectively. Non-defaultable Text OUTPUT This argument specifies the name and position of the file to export the results of the FE model transformation and calculation to. Non-defaultable Text EXCLUDE_ELEMENTS This arguments defines the beam elements not included in the FE model. Non-defaultable Integer SOLVER This argument specifies the matrix solver to be applied to the stiffness matrix. The options are: FULL_CONJUGENT_GRADIENT SYMMETRIC_CONJUGENT_GRADIENT OUT_OF_CORE_SKYLINE_CHOELESKI SKYLINE_CHOELESKI_DECOMPOSITION VECTORIZED_SPARSE_SOLVER Default=VECTORIZED_SPARSE_SOLVER Text MY The required value of the M.y total mass moment for the primary balance calculation. See Note 2. Default=0.0 Real kg.mm MZ The required value of the M.z total mass moment for the primary balance calculation. See Note 2. Default=0.0 Real kg.mm MYX The required value of the M.y.x total mass moment for the primary balance calculation. See Note 2. 527

COMMAND FILES Pre-Processor Commands Default=0.0

kg.mm2

Real

MZX The required value of the M.z.x total mass moment for the primary balance calculation. See Note 2. Default=0.0 Real kg.mm2 WEB_NUMBERS The web numbers to be drilled to obtain the required mass moments in the primary balance calculation. See Note 2. Non-defaultable Integer X The axial positions of the two counterweights used to obtain the required mass moments in the primary balance calculation. See Note 2. Non-defaultable Real m RADIUS The radial positions of the two counterweights used to obtain the required mass moments in the primary balance calculation. See Note 2. Non-defaultable Real m MASS The masses of the two counterweights used to obtain the required mass moments in the primary balance calculation. See Note 2. Non-defaultable Real kg

Notes: 1

The Hodgetts model refers to a model that idealises the crank to rigid, massive nodes interconnected by flexible massless elements. The elemental properties of this model may either be derived from stiffnesses obtained using the finite element method or using empirical formulae. The mass properties may either be derived from a finite element model or from a CAD solid model.

2

The primary balance calculation will be undertaken if any web numbers are chosen, the mass and axial position of the two counterweights are chosen or the radial and axial position of the two counterweights are chosen. If web numbers are specified there should be no information regarding the counter weight present in the command. Conversely, if counter weight information is present then should be no web numbers included in the command file. If the counterweight method of balancing is chosen, the mass of the counterweights will be calculated if the radial position of the weights is specified. Conversely, if the mass of the counterweights are specified their radial positions will be calculated.

528


529


1.7 CYLINDER The CYLINDER command is used to describe physical properties of each engine cylinder. Pre-Processor Commands has the following arguments: CYLINDER

NUMBER FIRING_ANGLE JOURNAL_NODE CYLINDER_NODES BORE CRANK_THROW CONROD_MASS CONROD_LENGTH CONROD_INERTIA CONROD_CENTRE_OF_GRAVITY

= = = = = = = = = =

PISTON_MASS TOP_RING_POSITION TOP_OF_SKIRT PIN_OFFSET CYLINDER_OFFSET HEIGHT INCLINATION_ANGLE RING_RADIAL_THICKNESS RING_AXIAL_THICKNESS RING_TIP_LOAD RECIPROCATING_MASS_RATIO BIG_END_BEARING_OFFSET

= = = = = = = = = = = =

Cylinder number Firing angle Crank pin node Cylinder nodal definition Cylinder bore Crank throw Connecting rod mass Connecting rod length Connecting rod inertia Connecting rod centre of gravity Piston mass Top ring position Top of skirt Piston pin offset Cylinder offset Cylinder height Angle of inclination Radial ring thickness Axial ring thickness Tip load Reciprocating mass ratio Bearing offset

where: NUMBER This defines the cylinder number and must be unique. Non-defaultable Integer FIRING_ANGLE This argument is used to define the crankshaft angle at which firing is ignited. Non-defaultable Real deg JOURNAL_NODE This argument defines the crank pin crankshaft node. Non-defaultable Integer

530

COMMAND FILES Pre-Processor Commands CYLINDER_NODES This argument is used to define four node sets to define the upper and lower reversal points of the small end of the connecting rod on the thrust and anti-thrust sides of the cylinder bore. Each node set will contain a single structural node. Non-defaultable Integer BORE This argument is used to define the cylinder bore. Non-defaultable Real

mm

CRANK_THROW This argument is used to define the crank throw radius. Non-defaultable Real

mm

CONROD_MASS This argument is used to define the total mass of the connecting rod including bolts and bearing shells. The value for g:conrod:mass is used if the data is read from the sdf file. Non-defaultable Real kg CONROD_LENGTH This argument is used to define the distance between the bearing centres of the connecting rod. Non-defaultable Real mm CONROD_INERTIA This argument is the connecting rod polar moment of inertia about its centre of gravity and will be the inertia of the complete rod assembly including bolts and shells. The value for g:conrod:inertia is used if the data is read from the sdf file. Non-defaultable Real kg.mm2 CONROD_CENTRE_OF_GRAVITY This argument is used to define the distance of the centre of gravity from the big-end centre. The z-co-ordinate of g:conrod:cgOffset is used if the data is read from the sdf file. Non-defaultable Real mm PISTON_MASS This argument is used to define the mass of the piston assembly. The value for g:piston:mass is used if the data is read from the sdf file. Non-defaultable Real kg TOP_RING_POSITION This argument is used to specify the distance of the top face of the top ring from the piston pin centre and must be greater than the Top of Skirt value. The value for g:piston:topRingHeight is used if the data is read from the sdf file. See Note 1. Non-defaultable Real mm 531


TOP_OF_SKIRT This argument is used to specify the distance of the top of the piston skirt from the piston pin centre and must be a positive value. The value for g:piston:skirtTop is used if the data is read from the sdf file. See Note 1. Non-defaultable Real mm BOTTOM_OF_SKIRT This argument is used to specify the distance of the bottom of the piston skirt from the piston pin centre and must be positive value. The value for g:piston:skirtBottom is used if the data is read from the sdf file. See Note 1. Non-defaultable Real mm PIN_OFFSET This argument is used to define the piston pin offset and is positive if the pin centre is on the anti-thrust side. The value for g:piston:pinOffset is used if the data is read from the sdf file. Non-defaultable Real mm CYLINDER_OFFSET This argument is used to define the distance between the leading and the trailing banks for a Vee engine. It must have a positive value. Non-defaultable Real mm HEIGHT This argument is used to define the distance from the crankshaft centre to the centre of the cylinder head. If the cylinder head has a flat roof chamber then this height will be the same as Cylinder Height. If the head has a pent-roof chamber then this height will be greater than Cylinder Height. This is only required if the cylinder block data is USER_DEFINED. Otherwise if data is loaded from an FE file the height is calculated from the FE model when the cylinder block model is assembled. Non-defaultable Real mm INCLINATION_ANGLE This argument is used to edit the angle of inclination for Vee engine of each cylinder about the x-axis from y. Non-defaultable Real deg RING_RADIAL_THICKNESS This argument is used to define radial thickness of the ring (the difference between the outer and inner radius). Non-defaultable Real mm RING_AXIAL_THICKNESS This argument is used to define axial (height) dimension of the ring. Non-defaultable Real mm

532

COMMAND FILES Pre-Processor Commands RING_TIP_LOAD This argument is used to define load applied to the on ring. Non-defaultable Real N RECIPROCATING_MASS_RATIO For all dynamic solutions, where the loads due to the connecting rod and piston are applied to the crank-pin, it is necessary to add the effective mass due to the rotating and reciprocating masses of the connecting rod and piston to the crankpin node. This is only required as an input if the crankshaft model is dynamic. Non-defaultable Real mm BIG_END_BEARING_OFFSET This argument is used to define the axial offset of the big-end bearing centre from the cylinder centre and is typically equal to zero. The offset is measured from the cylinder centre axis. A positive value denotes that the bearing is to the rear of the cylinder, whilst a negative value denotes that the bearing is in front of the cylinder. If an offset is specified then there will be moment with respect to the big end bearing centre due to the offset cylinder forces. For reasons of packaging Vee-configured engines often have a small offset. For one bank the big end bearing offset will be to the rear of the cylinder centre (+ve), whilst for the other bank the offset will be in front of the cylinder (-ve). Non-defaultable Real mm

Notes: 1

The parameters Top of Skirt, Bottom of Skirt and Top Ring are only used to apply loads to the cylinder bore when performing a quasi-static cylinder block analysis using the direct method. If the user is not intending to perform this type of analysis then these parameters can all be set to zero. If any of these parameters are non-zero then they should all be positive. Top of Skirt and Bottom of Skirt are used to define the extent of the pressure distribution when applying the piston side load. Top Ring is used to define the position above which the cylinder gas pressure is applied to the cylinder bore.

533


1.8 DAMPER The DAMPER command is used to specify the characteristics of a vibration damper. If the crankshaft has no vibration dampers then this command can be excluded from the command file. All referenced files should be in the standard Ricardo ASCII format. DAMPER has the following arguments: DAMPER

NAME NODE MODEL

= = =

TYPE

=

MASS

=

INERTIA

=

STIFFNESS_TYPE STIFFNESS

= =

BASE_STIFFNESS

=

DAMPING_TYPE

=

DAMPING

=

BASE_DAMPING HUB_OUTER_DIAMETER RING_INNER_DIAMETER AXIAL_LENGTH OIL_GAP SHEAR_AREA VOLUME

= = = = = =

TEMPERATURE

=

INITIAL_TEMPERATURE

=

Vibration damper name Attachment point of damper TV or general Damper Type of damper (viscous or elastomeric) Mass of seismic mass Inertia properties of seismic mass Dependency of the stiffnesses Stiffnesses of the damper Stiffnesses for a time domain analysis Dependency of the damping coefficients Damping co-efficents or a file containing the co-efficients Damping co-efficients for a time domain analysis Damper hub outer diameter Damper ring inner diameter Damper axial length Damper oil gap Damper shear area Damper volume Temperature of the damper which can vary with speed Initial temperature

where: NAME This argument is used to specify a name for the vibration damper. The name is used to identify and select dampers to be used in solution. The name is limited to 8 characters. Non-defaultable Text NODE The node at which the damper is attached. Non-defaultable Integer

534

COMMAND FILES Pre-Processor Commands MODEL This option defines the model type of the damper. The options are TORISIONAL or GENERAL. TORSIONAL represents torsional damping about the x-axis only. GENERAL represents damping in all 6 degrees of freedom. Therefore, this option affects the number of inputs in many of the other DAMPER arguments. See Note 1. Default=TORSIONAL Text TYPE This option defines the type of damper. The options are VISCOUS or ELASTOMERIC. Non-defaultable Text MASS This argument is used to specify the mass and inertia of the seismic mass of the damper. This argument is only needed if MODEL is set to GENERAL. Non-defaultable Real kg and kg.m-2 INERTIA This argument is used to specify the inertia about the x-axis of the seismic mass of the damper with respect to the global crankshaft axis. This argument is only needed if MODEL is set to TORSIONAL. Non-defaultable Real kg.m2 STIFFNESS_TYPE This argument specifies the dependency of the stiffnesses. The options are: CONSTANT The stiffnesses are constant TEMPERATURE_DEPENDENT The stiffnesses vary with temperature FREQUENCY_DEPENDENT The stiffnesses vary with frequency VARYING The stiffnesses vary with temperature and frequency If CONSTANT is specified then the code will expect inputs of type Real for the DAMPING argument. The number of inputs depends on the MODEL argument. If any other option is specified then the code will expect an input of type Text for the DAMPING argument. Non-defaultable Text STIFFNESS This argument is used to specify the stiffnesses or a file containing the varying stiffnesses. Non-defaultable Real or Text N.m-1 and N.m.rad-1 BASE_STIFFNESS Base stiffnesses to be used in a time-domain analysis if only frequency dependent data are provided. Non-defaultable Text N.m-1 and N.m.rad-1

535

COMMAND FILES Pre-Processor Commands DAMPING_TYPE This argument specifies the dependency of the damping co-efficients. The options are: CONSTANT The damping coefficients are constant TEMPERATURE_DEPENDENT The damping coefficients vary with temperature FREQUENCY_DEPENDENT The damping coefficients vary with frequency VARYING The damping coefficients vary with temperature and frequency If CONSTANT is specified then the code will expect inputs of type Real for the DAMPING argument. The number of inputs depends on the MODEL argument. If any other option is specified then the code will expect an input of type Text for the DAMPING argument. Non-defaultable Text DAMPING This argument specifies the damping co-efficients or the file in which the varying damping coefficients exist. This input is needed if one direction is specified as TYPE VISCOUS. Non-defaultable Real or Text N.s.m-1 and N.m.s.rad-1 BASE_DAMPING Base damping co-efficients to be used in a time-domain analysis if only frequencydependent data are provided. This input is needed if one direction is specified as TYPE VISCOUS. Non-defaultable Real N.s.m-1 and N.m.s.rad-1 DYNAMIC_MAGNIFIER Dynamic magnifier values or the file in which the varying dynamic magnifiers exist. Dynamic magnifiers can only vary with temperature. This input is needed if one direction is specified as TYPE ELASTOMERIC. Non-defaultable Real or Text FILTER_FREQUENCY Filter frequency values for each direction. This input is needed if one direction is specified as TYPE ELASTOMERIC. Non-defaultable Text Hz HUB_OUTER_DIAMETER This argument is used to specify the hub outer diameter of the vibration damper. See Note 2. Non-defaultable Text m RING_INNER_DIAMETER This argument is used to specify the ring inner diameter of the vibration damper. See Note 2. Non-defaultable Text m

536

COMMAND FILES Pre-Processor Commands AXIAL_LENGTH This argument is used to specify the axial length of the vibration damper. See Note 2. Non-defaultable Text m OIL_GAP This argument is used to specify the oil gap of the vibration damper. This input is only applicable to a viscous damper. See Note 2. Non-defaultable Text m SHEAR_AREA This is used to specify the effective shear area in the selected direction so as to calculate the Maximum Shear Stress. See Note 2. Non-defaultable Text m2 VOLUME This is use to specify the volume of the elastomer or the oil so as to calculate the power density. See Note 2. Non-defaultable Text m3 TEMPERATURE This argument can be either CONSTANT or the name of a file containing the speed dependent temperature data. Default=CONSTANT Text INITIAL_TEMPERATURE This is the initial temperature of the vibration damper for a thermal balance calculation of the damper. o Default=20 oC Text C

Notes: 1

If MODEL is set to GENERAL then the following arguments will take 6 inputs:  TYPE  MASS  STIFFNESS_TYPE  STIFFNESS  BASE_STIFFNESS  DAMPING_TYPE  DAMPING (if one direction has TYPE VISCOUS)  BASE_DAMPING (if one direction has TYPE VISCOUS)  DYNAMIC_MAGNIFIER (if one direction has TYPE ELASTOMERIC)  FILTER_FREQUENCY (if one direction has TYPE ELASTOMERIC)  HUB_OUTER_DIAMETER (Optional)  RING_INNER_DIAMETER (Optional)  AXIAL_LENGTH (Optional)  OIL_GAP (Optional) 537

COMMAND FILES Pre-Processor Commands  SHEAR_AREA (Optional)  VOLUME (Optional) If MODEL is set to TORSIONAL then the following arguments will take 1 input:  TYPE  INERTIA  STIFFNESS_TYPE  STIFFNESS  BASE_STIFFNESS  DAMPING_TYPE  DAMPING (if TYPE is VISCOUS)  BASE_DAMPING (if TYPE is VISCOUS)  DYNAMIC_MAGNIFIER (if TYPE is ELASTOMERIC)  FILTER_FREQUENCY (if TYPE is ELASTOMERIC)  HUB_OUTER_DIAMETER (Optional)  RING_INNER_DIAMETER (Optional)  AXIAL_LENGTH (Optional)  OIL_GAP (Optional)  SHEAR_AREA (Optional)  VOLUME (Optional) 2

SHEAR_AREA and VOLUME inputs are used to calculate the Maximum Shear Stress, Power Loss and Power Density for the vibration damper. If these inputs are not supplied and the geometry of the damper is supplied by the HUB_OUTER_DIAMETER, RING_INNER_DIAMETER, AXIAL_LENGTH and OIL_GAP arguments then the calculate the Maximum Shear Stress, Power Loss and Power Density for the vibration damper will be calculated using these values. If none of the inputs described in this note are provided then the Maximum Shear Stress, Power Loss and Power Density for the vibration damper will not be calculated.

3

An example of a torsional, viscous damper DAMPER NAME = MODEL = NODE = TYPE = INERTIA = STIFFNESS_TYPE = STIFFNESS = BASE_STIFFNESS = DAMPING_TYPE = DAMPING = BASE_DAMPING = SHEAR_AREA = VOLUME = TEMPERATURE = INITIAL_TEMPERATURE =

538

damper:ID_viscous TORSIONAL 1 VISCOUS 0.005550 [kg.m^2] CONSTANT 0.016982 [MN.m/rad] 0.016982 [MN.m/rad] CONSTANT 2.912475 [N.m.s/rad] 2.912475 [N.m.s/rad] 1000.000000 [mm^2] 2000.000000 [mm^3] 20.000000 [degC] 0.000000 [degC]

\ \ \ \ \ \ \ \ \ \ \ \ \ \


1.9 DIRECTION The DIRECTION command is used to set a direction vector for each local direction. This command broadens the LINK command capabilities. Pre-Processor Commands has the following arguments: DIRECTION

NUMBER

=

Local direction identifier

VECTOR

=

Direction vector

where: NUMBER This argument is used to identify the local direction vector. Non-defaultable Integer VECTOR This argument is used to define a local direction. Non-defaultable Real

539


1.10 DRIVE The DRIVE command is used to define the crankshaft node at which the mean torque of the engine is reacted. DRIVE has the following arguments: DRIVE

NODE

=

Crankshaft node at which mean torque is reacted

where: NODE This argument is used to specify the crankshaft node at which the mean torque of the engine is reacted. Non-defaultable Integer

540


1.11 ELEMENT The ELEMENT command is used to define a crankshaft element. The program calculates the flexibility coefficients based on the data specified under this command. This command should be used for each element of the crankshaft and should appear after the MASS commands. ELEMENT has the following arguments: ELEMENT

NUMBER TOPOLOGY TYPE

= Element number = Element topology = Element type

CYLINDRICAL PRISM

RIGID_LENGTHS

= Rigid lengths of element = Outside diameter of element = Inside diameter of element

OUTSIDE_DIAMETERINSIDE_ DIAMETER INSIDE_DIAMETER

CRANKWEB FLEXIBILITIES

EQUIVALENT_WIDTH FULL_THICKNESS IN_PLANE_STIFFNESSIN_PL ANE_STIFFNESSIN_PLANE_S TIFFNESSIN_PLANE_STIFFN ESSIN_PLANE_STIFFNESS OUT_PLANE_STIFFNESS

WEB_NUMBER

= Out-of-plane stiffness of web = Torsional stiffness of web = Radius of counterweight = Start of counterweight = End of counterweight = Number of web

UFXY UFXZ UMXY UMXZ VFXY VFXZ VMXY VMXZ CFLEX TFLEX

= = = = = = = = = =

TORSIONAL_STIFFNESS RADIUS_OF_COUNTERWEIG HT START_OF_COUNTERWEIGH T END_OF_COUNTERWEIGHT

ELEMENT FLEXIBILITY COEFFICIENTS

= Width of web in global YZ-plane = Width of web in global XY-plane = In-plane stiffness of web

541


SHAFT STIFFNESSES

BENDING_STIFFNESS TORSIONAL_STIFFNESS AXIAL_STIFFNESS

= Bending stiffness of element = Torsional stiffness of element = Axial stiffness of element

SPRING STIFFNESS AND DAMPING

STIFFNESS DAMPING PRELOAD PRELOAD_DISTANCE

= = = =

FRICTION AND DAMPING

DAMPING FRICTION_COEFFICIENT

= Element damping Element coefficient of friction

Stiffness of spring Spring damping Spring preload Distance between nodes at which preload is applied

where: NUMBER This argument is used to set the element number. Non-defaultable Integer TOPOLOGY This argument is used to set the element number. Non-defaultable Integer TYPE This argument is used to specify the type of geometric section. It takes the following values : CYLINDRICAL, MAIN, PIN Cylindrical prism WEB Crankweb properties PROPERTIES Element flexibilities SHAFT Shaft flexibilities SPRING_AND_DAMPER Spring flexibilities and damping RIGID_LINK Rigid link FRICTION Friction and damping element The type of element determines which information should be included in the PreProcessor Commands command TYPE=CYLINDRICAL If TYPE=CYLINDRICAL, MAIN or PIN then the element flexibilities are evaluated assuming a cylindrical prism, and the following arguments apply. This option is used

542

COMMAND FILES Pre-Processor Commands to determine the flexibilities at the crankpins and main bearing journals of the crankshaft.. Default=PROPERTIES Text RIGID_LENGTHS This argument is used to specify the rigid lengths of the cylindrical prism. Two values are expected and these are the rigid lengths at each end of the element where the first value is the rigid length at the end nearest to x=0. If TYPE=MAIN or PIN then this argument is ignored and the FULL_THICKNESS value for the adjacent crankweb is used to calculate the current rigid length. Default=0.0,0.0 Real m OUTSIDE_DIAMETER This argument is used to specify the outside diameter of the cylinder. If this argument is not specified then it defaults to the outside diameter at the node with the same number as the element specified using the MASS command. If the diameter at that node has not been specified using the MASS command then the default is 0.0. Default=(see above) Real m INSIDE_DIAMETER This argument is used to specify the inside diameter of the cylinder. If this argument is not specified then it defaults to the inside diameter at the node with the same number as the element specified using the `MASS' command. If the diameter at that node has not been specified using the `MASS' command then the default is 0.0. Default=(see above) Real m TYPE=WEB If TYPE=WEB then the element flexibilities are evaluated assuming stiffness values obtained from a finite-element analysis of a crankweb, and the following arguments apply. Figures 2.6 and 2.7 define the dimensions and stiffnesses for this argument. Appendix 1 describes how the in-plane bending, out-of-plane bending and torsional stiffnesses are obtained using the finite element method. EQUIVALENT_WIDTH This argument is used to specify the width of the crankweb in the global YZ-plane. This dimension is derived using the formula for the equivalent length of a crankweb given in Figure 2.6 and described in Reference [10]. If this is ignored a preferred method of evaluating the flexibilities is employed using FULL_THICKNESS only. Default=(see above) Real m FULL_THICKNESS This argument is used to specify the width of the crankweb in the global XY-plane. Non-defaultable Real m IN_PLANE_STIFFNESS This argument is used to specify the in-plane stiffness of the crankshaft web. Non-defaultable Real N.m/rad

543

COMMAND FILES Pre-Processor Commands OUT_PLANE_STIFFNESS This argument is used to specify the out-of-plane stiffness of the crankshaft web. Non-defaultable Real N.m/rad TORSIONAL_STIFFNESS This argument is used to define torsional stiffness. It is evaluated by applying a moment as a displacement load to the model, which has an anti-symmetric with expansion constraint at the crankpin cut plane. In addition, three nodes are restrained to prevent rigid body motion. Non-defaultable Real N.m/rad

The following arguments are used to define the extent of the counterweight at a web and are used to determine the location of balance drillings for primary crankshaft balance. Only those webs referred to in the BALANCE command require these arguments. RADIUS_OF_COUNTERWEIGHT This argument is used to specify the radius from the crankshaft axis to the outer edge of the counterweight. Default=0.0 Real m START_OF_COUNTERWEIGHT This argument is used to specify the angular position corresponding to the start of the counterweight measured clockwise relative to the Y-axis of the ENGDYN Co-ordinate System. This angle should lie between 0 and 360°. Default=0.0 Real deg END_OF_COUNTERWEIGHT This argument is used to specify the angular position corresponding to the end of the counterweight measured clockwise relative to the Y-axis of the ENGDYN Co-ordinate System. This angle should lie between 0 and 360°. Default=0.0 Real deg WEB_NUMBER This argument is used to define a web number. Non-defaultable Integer TYPE=PROPERTIES If TYPE=PROPERTIES then the element flexibilities as defined in NEED FIGURE are entered directly by the user, and the following arguments apply. It is anticipated that all element flexibilities should be specified using TYPE=CYLINDRICAL and WEB. This option is included for completeness.

UFXY 544

COMMAND FILES Pre-Processor Commands This argument is used to specify the flexibility coefficient UFXY. Non-defaultable Real m/N UFXZ This argument is used to specify the flexibility coefficient UFXY. Non-defaultable Real m/N UMXY This argument is used to specify the flexibility coefficient UFXY. Non-defaultable Real rad/(N.m) UMXZ This argument is used to specify the flexibility coefficient UFXY. Non-defaultable Real rad/(N.m) VFXY This argument is used to specify the flexibility coefficient UFXY. Non-defaultable Real m/N VFXZ This argument is used to specify the flexibility coefficient UFXY. Non-defaultable Real m/N VMXY This argument is used to specify the flexibility coefficient UFXY. Non-defaultable Real rad/(N.m) VMXZ This argument is used to specify the flexibility coefficient UFXY. Non-defaultable Real rad/(N.m) CFLEX This argument is used to specify the flexibility coefficient CFLEX. Non-defaultable Real m/N TFLEX This argument is used to specify the flexibility coefficient TFLEX. Non-defaultable Real rad/(N.m) TYPE=SHAFT If TYPE=SHAFT then the element flexibilities are evaluated assuming stiffness values obtained from a finite-element analysis of the flywheel or cranknose respectively, and the following arguments apply. The user must ensure that the element length as defined by the nodes using the NODE command corresponds to 545

COMMAND FILES Pre-Processor Commands the distance between the constraint face and applied force in the finite element model. Appendix 1 describes how the bending, axial and torsional stiffnesses are obtained using the finite-element method. BENDING_STIFFNESS This argument is used to specify the bending stiffness of the crankweb. Non-defaultable Real N.m/rad TORSIONAL_STIFFNESS This argument is used to specify the torsional stiffness of the crankweb. Non-defaultable Real N.m/rad AXIAL_STIFFNESS This argument is used to specify the axial stiffness of the crankweb. Non-defaultable Real N.m TYPE= SPRING_AND_DAMPER STIFFNESS This argument is used to specify the stiffness of the spring. Non-defaultable Real N/m DAMPING This argument is used to define damping factor of the damper. Default=0.0 Real Ns PRELOAD This argument is used to define the preload of the spring. Default=0.0 Real

N

PRELOAD_DISTANCE is the description Default=0.0

m

Real

TYPE=FRICTION DAMPING This argument is used to define damping factor of the element. Non-defaultable Real Nsm-1 FRICTION_COEFFICIENT This argument is used to set value of the friction coefficient for the contacting surfaces. Non-defaultable Real

546


Notes: 1

The first and last elements are assumed to have zero flexibility. The properties for these elements, which correspond to the elements at the front and rear of the crankshaft, should not be input by the user.

2

It is assumed that a given element n joins the node n to node n+1.

547


1.12 ENGINE The ENGINE command is used to specify the type of engine considered. ENGINE has the following arguments: ENGINE

CYCLE ROTATION

= =

The type of cycle the engine performs The direction of rotation of the

where: CYCLE The cycle describes the stroke of the engine and can be 2 or 4 stroke. Default=4 Integer ROTATION The rotation describes the direction of the crankshaft and can be either CLOCKWISE or ANTICLOCKWISE. Default=CLOCKWISE Text

548


1.13 IN_CYLINDER The IN_CYLINDER command is used to describe in cylinder model of motion. IN_CYLINDER has the following arguments: IN_CYLINDER

MODEL METHOD

= =

Model type Method of simulation

FILE

=

PISDYN sdf file name

EQUATION

=

Equation defining user model

where: MODEL This is used to select the piston motion model. The available options are: SLIDER_CRANK, SECONDARY_DYNAMIC, USER_EQUATION. Default=SLIDER_CRANK Text METHOD This argument is used to select a numerical method for simulation of piston motion. Only SIMPLIFIED method is available so far. Default=SIMPLIFIED Text FILE This argument is used to select a data file from PISDYN analysis if SECONDARY_DYNAMIC is chosen. Non-defaultable SDF file EQUATION This argument is used to define USER_EQUATION. Non-defaultable String

549


1.14 LINK The LINK command is used to define mechanical links and their associated simulation children at a particular node on the crankshaft. The LINK command has the following arguments: LINK

NUMBER MODEL

= Identification number = Engine part

NODE

= Selected node

SET DIRECTION TYPE DIMENSION

= = = =

CHILD

= Simulation children

RELATIVE_TO

= Crankshaft axis definition

AXIS

= Crankshaft axis definition

Coordinate system definition Coordinate system definition Link type Link dimension

where: NUMBER This argument is used to specify the unique identification number of the mechanical link shared with the child process. Non-defaultable Integer MODEL This argument is used to select a part of assembly. It is possible to select CRANKSHAFT or CYLINDER_BLOCK. Non-defaultable Text NODE This argument is used to select a node where mechanical link is to be applied. Non-defaultable Integer SET This argument is used to set coordinate system to LOCAL or GLOBAL. Non-defaultable Text DIRECTION This argument must be specified if SET is GLOBAL. The options are X, Y or Z. If SET is LOCAL, default value is (0,0,0) deg. Non-defaultable Text or Real

550


TYPE This is used to specify the type of the link either FORCE or MOTION. Default=FORCE Text DIMENSION This argument is used to specify the dimension of the link. The possible options are, 1D_TRANSLATIONAL, 1D_ROTATION, 3D_TRANSLATIONAL, 3D_ROTATION and 3D_6DOF. Non-defaultable Text CHILD This lists the currently defined simulation children. Child number zero is always defined as being the parent process. Non-defaultable RELATIVE_TO This is used to specify the co-ordinate system of the crankshaft axis in the connecting application pertaining to the link. The options are NODE or ORIGIN. If connecting to VALDYN then Origin would be selected. Non-defaultable Text AXIS This is used to specify the co-ordinate system of the crankshaft axis in the connecting application pertaining to the link. The options are either FIXED or ROTATING. If connecting to VALDYN then Fixed would be selected. Default=0, 0, 0 Real mm

551


1.15 LOADING The LOADING command is used to apply pressures and loads to the model LOADING has the following arguments: LOADING

TYPE NAME SPEED CYLINDER_PRESSURE

= = = =

AMBIENT_PRESSURE CRANKCASE_PRESSURE FACTOR INTERVAL OFFSET FILE FROM UNIT ROTATE TRANSLATE MIRROR EQUATION SET DIRECTION DISTORTION

= = = = = = = = = = = = = = =

CASE MODEL RELATIVE_TO

= = =

Type of loading Loading application Revolutions of shaft Specify pressure inside cylinder Reference pressure Pressure at bearing edges Scaling of pressure Angle between solution steps Offset of piston pressure data Input date file Data format Units specification Rotate model Translate model Mirror model Force equation Applied force set Force direction Distortion of bearings and journals Loadcase number Model loading Axis that the force equation is applied to

where: TYPE This argument used to select type of loading and which component is to be under loading. The possible types are: IN_CYLINDER FORCE_EQUATION FORCE_PROFILE OIL_TEMPERATURE OIL_FEED_TEMPERATURE OIL_FEED_PRESSURE JOURNAL_TEMPERATURE BEARING_TEMPERATURE DISTORTION PRESSURE Non-defaultable

Text

552

COMMAND FILES Pre-Processor Commands NAME This argument is used to select how the loading should be applied. Possible values are: FULL_LOAD, PART_LOAD and NO_LOAD. Default=FULL_LOAD Text SPEED This argument is used to specify the revolutions per minute of the engine. See Note 1. Non-defaultable Real rpm CYLINDER_PRESSURE This argument is used to load a file containing pressure data for given speed and cylinder. See Note 1. Non-defaultable *.PRES file AMBIENT_PRESSURE This argument is used to specify the ambient pressure. See Note 1. Default=1 Real bar CRANKCASE_PRESSURE This argument is used for hydrodynamic solutions to define the conditions at the edge of the bearing. See Note 1. Default=1 Real bar FACTOR This argument is used to scale the data in the PRES file to a different rating. The FACTOR argument stands for either a constant factor or an expression (in terms of T, IMEP, P, PMAX). See Note 1. Default=1 Real INTERVAL This argument is used to work with data in the form of a pressure column at equal intervals and defines the angle between each pressure value. This is used in conjunction with the OFFSET argument. The interval angle is only used if there are no crank angles in the .PRES file. Default=1.0 Real deg OFFSET This argument is used to offset supplied in cylinder pressures by a defined angle relative to the shaft angle. Default=0.0 Real deg

553

COMMAND FILES Pre-Processor Commands FILE This argument is used to specify a file containing loading data. The type of file depends on the TYPE of loading specified. IN_CYLINDER SDF file containing in-cylinder pressures FORCE_PROFILE ASCII or SDF file containing force profiles DISTORTION SFE containing distortions of journals and bearings Non-defaultable

Text

FROM This argument is used to select the source format of the FILE if the TYPE of loading is DISTORTION. The possibilities are: SFE ABAQUS ANSYS IDEAS FEMGEN PAFEC MSC NASTRAN Default=SFE

Text

UNIT This argument is used to specify the units for the FE model in the file specified by FROM. The options are SI, MM, METRIC, FT and INCH. It is not required when data are loaded from and SFE or IDEAS file using the FROM command Default=SI Text ROTATE This argument is used to rotate the FE model in the file specified by FROM. Default=0,0,0 Real TRANSLATE This argument is used to translate the FE model in the file specified by FROM. Default=0,0,0 Real MIRROR This argument is used to specify the vector in which to mirror the FE model in the file specified by FROM. Default=0,0,0 Real CASE This argument is used to identify loadcase number for the data in the file specified by FROM. Non-defaultable Integer

554

COMMAND FILES Pre-Processor Commands EQUATION This argument is used to define a force equation which acts on the sets defined by SET. It is needed when TYPE is specified as FORCE_EQUATION. See Note 2. Non-defaultable SET This argument specifies the set to apply the force equation to. See Note 2. Default=0,0,0 Real mm DIRECTION This argument is used to define if the load is acting in the local or global direction. See Note 2. Non-defaultable Integer RELATIVE_TO This argument is used to specify the axis to apply the force equation to. The options are FIXED and ROTATING. This option is only relevant is the MODEL is CRANKSHAFT Non-defaultable Text DISTORTION This argument is used to define the distortion of the journals and bearings for each of the journal bearings under operating conditions. This effectively allows the user to define the hot shape of the journal and bearing. The options are: NONE CONSTANT FROM_DISTORTED_SHAPE Default=NONE

Text

MODEL This argument is used to choose which model part the loading is applied to. The possibilities are: CRANKSHAFT CYLINDER_BLOCK CONNECTING_ROD Non-defaultable

Text

RELATIVE_TO This argument is used to specify the axis which the CRANKSHAFT MODEL is relative to. Non-defaultable

Text

555


Notes: 1

The number of provided inputs for SPEED, AMBIENT_PRESSURE, CRANKCASE_PRESSURE, CYLINDER_PRESSURE, FACTOR should be consistent.

2

The number of provided inputs for EQUATION, SET and DIRECTION should be consistent.

3

An example of the LOADING command is given below: LOADING TYPE=IN_CYLINDER NAME=FULL_LOAD SPEED=1000 CYLINDER_PRESSURE='p10_so.PRES' AMBIENT_PRESSURE=0 CRANKCASE_PRESSURE=0 FACTOR=1.0 INTERVAL=1

556


1.16 LUBRICANT The LUBRICANT command is used to specify the type of lubricant used and its fluid properties. LUBRICANT has the following arguments: LUBRICANT

NAME FILE

= =

EQUATION GAMMA

= =

PRESSURE_VISCOSITY

=

The name of the lubricant The name of the file containing the fluid properties of the lubricant The type of lubrication equation The value of Gamma in the Walther-ASTM equation The type of equation used for pressure-viscosity dependence

where: NAME The name of the lubricant can be specified. The lubricant information is taken to be included in a file denoted by the NAME identifier with an extension of .MAT in the same directory as the input file. If the file is not found in the current directory, the ENGDYN database directory is searched for the lubricant file denoted by the NAME argument. NAME cannot be used with FILE. See Note 1 for further information. Non-defaultable Text FILE This argument specifies the explicit position of the file which contains the lubricant properties. FILE cannot be used with NAME. See Note 1 for further information. Non-defaultable Text EQUATION This argument specifies the type of equation to be used which relates the viscosity of the lubricant with its temperature. The choices are: 1. USER_DEFINED 2. WALTHER_ASTM 3. VOGEL If USER_DEFINED is chosen see Note 2 for more information Default= WALTHER_ASTM Text GAMMA This argument specifies the value of GAMMA needed for the WALTHER_ASTM equation. The values must be either 0.6 or 0.7 cstokes. Default=0.6 Real cstokes PRESSURE_VISCOSITY This argument specifies the type of equation to be used which relates the viscosity of the lubricant with its pressure. The choices are:

557

COMMAND FILES Pre-Processor Commands NONE BARUS ROELANDS For more information please see note 3. Default=NONE Text

Notes: 1

The fluid property file should contain data for the specific heat capacity, density, cavitation pressure, bulk modulus and piezo viscosity coefficient. If the user defined lubrication equation is chosen the file should also contain data relating the temperature to the kinematic viscosity. An example is shown below: SPECIFIC_HEAT_CAPACITY=2000 DENSITY=879 \ CAVITATION_PRESSURE=99990 BULK_MODULUS=5.0E8 \ PIEZO_VISCOSITY_COEFFICIENT=0.0 TEMPERTURE=40 KINEMATIC_VISCOSITY=90.48E-6 TEMPERTURE=100 KINEMATIC_VISCOSITY=14.52E-6

2

If the user defined model is chosen for the viscosity temperature equation the lubricant fluid properties file should contain data relating the viscosity and temperature (see Note 1).

3

The fluid property file can include PRESSURE_VISCOSITY_INDEX property used when PRESSURE_VISCOSITY=ROELANDS. If the PRESSURE_VISCOSITY_INDEX property is not defined, default value 0.67 is used. See Engdyn Theory for details on Roelands equation.

558


1.17 MASS The command MASS is used to define a section at a given nodal position along the crankshaft. The program calculates the mass properties at that node based on the data specified under this command. This command should be used for each section of the crankshaft. This command should appear after the NODE and MATERIAL commands. MASS has the following arguments: MASS

NODE AXIS TYPE

= = =

Node number Axis type Section type

TYPE=CYLINDRICAL, JOURNAL or PIN LENGTH INSIDE OUTSIDE X_OFFSET MATERIAL

= = = = =

Axial length of section Inside diameter of cylindrical section Outside diameter of cylindrical section Distance of centroid from node in XX Material properties file name

TYPE=PROPERTIES or WEB

MASS

=

Mass at node

GX GY GZ

= = =

OFFSET

=

X_OFFSET

=

RADIUS

=

THETA IXX IYY IZZ IXY IYZ IZX

= = = = = = =

INERTIA

=

Node centre of gravity x-coordinate Node centre of gravity y-coordinate Node centre of gravity z-coordinate Node centre of gravity x-, y- and zcoordinates Axial distance of centroid from defined axis Radial distance of centroid from defined axis Angular distance of centroid from defined axis Nodal inertia about XX Nodal inertia about YY Nodal inertia about ZZ Nodal inertia product about XY Nodal inertia product about YZ Nodal inertia product about ZX Nodal inertia about all rotational directions

where: NODE This argument is used to specify the node at which the properties of the section are to be concentrated. Non-defaultable Integer

559

COMMAND FILES Pre-Processor Commands AXIS This argument is used to define the axis (LOCAL, ORIGIN or CENTROID) which the properties are relative to if TYPE=PROPERTIES. Default=LOCAL Text TYPE This argument is used to specify the type of geometric section defined on the MASS command line. It takes the following values : CYLINDRICAL, JOURNAL, PIN Cylindrical prism PROPERTIES, WEB Crankweb properties The required additional inputs are dependent on the type of geometric section denoted by the TYPE argument. Default=PROPERTIES Text TYPE=CYLINDRICAL, JOURNAL or PIN If TYPE=CYLINDRICAL, JOURNAL or PIN then the mass properties for the given section are evaluated assuming a cylindrical prism, and the following arguments apply. As the values of the argument imply, this option is used to determine the properties at the crankpins and main bearing journals of the crankshaft. LENGTH This argument is used to specify the length of the cylindrical prism. Non-defaultable Real m INSIDE This argument is used to specify the inner diameter of the cylindrical prism. Non-defaultable Real m OUTSIDE This argument is used to specify the outer diameter of the cylindrical prism. Non-defaultable Real m X_OFFSET This argument is used to specify the axial offset of the prism relative to the nodal position. Default=0.0 Real m MATERIAL This argument is used to specify the name of the file that contains the required material properties (see Note 1). This file should already have been read into the buffer via the command MATERIAL. Non-defaultable Text TYPE=PROPERTIES OR WEB If TYPE=PROPERTIES or WEB then the mass properties for the given section are entered directly by the user, and the following arguments apply. As the values of the argument imply, this option is used to specify the mass properties at each of the 560

COMMAND FILES Pre-Processor Commands crankwebs obtained from finite-element stiffness models. Appendix 1 describes how the mass properties of the finite-element model are evaluated. The centre of gravity relative to the defined axis can be specified in three ways. These methods are mutually exclusive and therefore only one method should be used. The centre of gravity can be specified by using the GX, GY and GZ arguments, the OFFSET argument or the X_OFFSET, RADIUS and THETA arguments. The GX, GY and GZ arguments and the OFFSET argument specify the centre of gravity using a Cartesian coordinate system. The X_OFFSET, RADIUS and THETA arguments specify the centre of gravity using a Cylindrical coordinate system. The inertia of the section can be specified in one of two ways. Again, these methods are mutually exclusive and therefore only one method should be used. The inertia can be specified by using the INERTIA, or IXX, IXY, IYY, IYZ, IZX and IZZ arguments MASS This argument is used to specify the mass of the section. Non-defaultable Real

kg

GX This argument is used to specify the x-coordinate of the centre of gravity relative to the defined axis. Non-defaultable Real m GY This argument is used to specify the y-coordinate of the centre of gravity relative to the defined axis. Non-defaultable Real m GZ This argument is used to specify the z-coordinate of the centre of gravity relative to the defined axis. Non-defaultable Real m OFFSET This argument specifies the offset of the centre of gravity relative to the defined axis by specifying the three directional offsets using a Cartesian coordinate system. Non-defaultable Real m X_OFFSET This argument is used to specify the axial offset of the element relative to the defined axis. Non-defaultable Real m RADIUS This argument is used to specify the radial offset of the element relative to the defined axis. Non-defaultable Real m 561

COMMAND FILES Pre-Processor Commands THETA This argument is used to specify the angular offset of the element relative to the defined axis. Non-defaultable Real m IXX This argument is used to specify the inertia about XX of the defined axis. Non-defaultable Real kg.m2 IYY This argument is used to specify the inertia about YY of the defined axis. Non-defaultable Real kg.m2 IZZ This argument is used to specify the inertia about ZZ of the defined axis. Non-defaultable Real kg.m2 IXY This argument is used to specify the inertia about XY of the defined axis. Non-defaultable Real kg.m2 IYZ This argument is used to specify the inertia about YZ of the defined axis. Non-defaultable Real kg.m2 IZX This argument is used to specify the inertia about ZX of the defined axis. Non-defaultable Real kg.m2 INERTIA This argument is used to specify the inertia about all six rotational directions of the defined axis. Non-defaultable Real kg.m2

562


Notes: 1

The materials file holds quoted strengths for a particular material. These files always have the suffix .MAT and are kept either in the current working directory or stored together in a directory called materialsdb (usually /usr/local/materialsdb) to form an ever increasing library of materials data. If this location is unsatisfactory for your system it can be put anywhere provided that the environment variable RICARDO_FE_DIRECTORY points to the new location (this environment variable is normally set in the script file that runs ENGDYN). For example, if the materialsdb directory is put in /usr/databases, the environment variable must be set up with the command: setenv RICARDO_FE_DIRECTORY /usr/databases A version of a file in the working directory takes precedence over the library version. The data definitions are set out in a similar way to a Ricardo command file with names being set equal to values. Long lines can be continued with an '&' character. The recognised arguments for a material property definitions are as follows : UTS N.m-2 TENSILE_YIELD N.m-2 COMPRESSIVE_YIELD N.m-2 FATIGUE_STRENGTH N.m-2 DENSITY kg.m-3 Only sufficient characters to avoid ambiguity need to be typed. The first letter must be typed but any subsequent letters may be omitted (e.g TEY represents TENSILE_YIELD).

2

More than one section can be specified for each node. The mass properties of a given node are calculated based on all the sections specified at that node.

563


1.18 MATERIAL The MATERIAL command is used to specify material properties for a component. MATERIAL has the following arguments: MATERIAL

NAME

=

Name of material properties file

where: NAME This argument is used to specify the name of the file that contains the required material properties. Non-defaultable Text

564


1.19 MOUNT The MOUNT command models engine mounts and it can either be modelled using a quasi-elastomeric or hydroelastic model. Each model is a 3 degree of freedom model connecting a node representing the engine side bracket to ground. Each mount is defined such that x lies along the axis of the mount. MOUNT has the following arguments: MOUNT

TYPE

=

Type of mount

NAME

=

Given name of the mount

ROTATIONS

=

Mount orientation

DYNAMIC_STIFFNESS

=

Dynamic stiffness

DYNAMIC_MAGNIFIER

=

Dynamic magnifier

FILTER_FREQUENCY

=

Frequency filter

STATIC_STIFFNESS

=

Static stiffness

DAMPING_COEFFICIENT

=

Damper coefficient

MASS

=

Mass of mount

where: TYPE This argument is used to define the type of mounting. Possibilities to select are: ELASTOMERIC HYDROELASTIC ELASTIC Non-defaultable

Text

NAME This argument is used to specify the given name of the mount. Non-defaultable Integer ROTATIONS This argument allows the mount to be orientated with respect to the ENGDYN coordinate system. The orientation vector is defined by three rotations about X, Y and Z. The transformation matrix is built up in the following order; rotation about Z, then Y and finally about X. Non-defaultable Real deg

565

COMMAND FILES Pre-Processor Commands DYNAMIC_STIFFNESS This argument is used to specify the dynamic stiffness. Typically the value of this stiffness is 1.4 times the static stiffness. Default=0 Real MN/m DYNAMIC_MAGNIFIER This argument is used to specify the dynamic magnifier of the elastomer. Default=0 Real FILTER_FREQUENCY This argumentis used to define the filter frequency. A value of 1 Hz is recommended. Default=1 Real Hz STATIC_STIFFNESS This is used to specify the static stiffness. This is not currently used in the solution. Default=0 Real MN/m DAMPING_COEFFICIENT This argument is used to define the damping coefficient c of the viscous damper. Default=INPUT_2_default_value Real N.s/m MASS This argument is used to specify mass of a hydroelastic engine mount. Default=0 Real kg

566


1.20 NODE The NODE command is used to define the nodal coordinates of a node belonging to a given model. NODE has the following arguments: NODE

NUMBER COORDINATE MODEL

= = =

Node number x, y and z-coordinates of node Model type

where: NUMBER This argument is used to specify the node number to which the coordinates apply. Non-defaultable Integer COORDINATE This argument is used to specify the x,y and z-coordinates for this node. Non-defaultable Real m MODEL This argument denotes the reduced model type of this node. At present only the crankshaft model is available Default=CRANKSHAFT Integer

Notes: 1

For the crankshaft the nodes must proceed along the length of the crankshaft in sequential order. If the .EDPR file is written from the Graphical User Interface then the first node will be at the crank nose end. If RIGID_LINK elements are used then the crankshaft model can be branched such that the nodes are no longer sequential, but the nodes defining the flexibility of the crankshaft must be in sequential order.

567


1.21 OPEN The OPEN command is used to open the ENGDYN standard data base file. This must be the first command in the file. If this file already exists then it will be overwritten. OPEN

DATABASE

=

Name of the ENGDYN standard data file

where: DATABASE This argument is used to specify the name of the standard data file. The name does not need to contain the .EDSF suffix which is best omitted. Non-defaultable Text

568


1.22 PROFILE The PROFILE is used to is used to define the cold manufactured profiles of the journal and bearing when the Model Type is set Hydrodynamic or Elastohydrodynamic. The hydrodynamic and boundary lubrication models are dependent on clearance values. The clearance profile between the bearing and journal surfaces determines the oil film and contact pressures, which are generated in this region. This profile is based on the nodal clearance values. PROFILE has the following arguments: PROFILE

NUMBER

=

Profile number

TYPE

=

Profile type

RADIAL_DEVIATION

=

Maximal radial deviation

ANGULAR_OFFSET

=

Angular offset

NUMBER_OF_LOBES =

Number of profile lobes

EQUATION

=

Profile description

FILE

=

Profile description

DIAMETER

=

Diameter of profile

OFFSET

=

Profile centre offset

where: NUMBER This argument is used to set the profile number. Non-defaultable Integer TYPE This argument is used to specify the profile type. The options are CIRCULAR, ELLIPTICAL, SINUSOIDAL, EQUATION and FILE. The availability of certain further arguments depends on the TYPE. Non-defaultable Text RADIAL_DEVIATION This argument is used to define the maximum radial deviation from the DIAMETER. Default=0 Real mm ANGULAR_OFFSET This argument is used to define the angular offset (i.e. rotation) of the profile about the centre. Default=0 Real deg

569

COMMAND FILES Pre-Processor Commands NUMBER_OF_LOBES This argument is used to define how many lobes the profile has and can only be used if the profile TYPE is sinusoidal. Default=0 Integer EQUATION This argument is used to define an equation as a function of D, R, N, θ and φ where these variables correspond to Diameter, Radial Deviation, Number of Lobes, Angle and Angular Offset. The expression must have consistent units where the resultant has a unit of length. Non-defaultable Real FILE This argument is used to define the name of the ASCII file from which the profile is read. Each file contains either a single profile or multiple profiles of radii or radial deviations varying with angle, in which the data is defined in block or column free format. The data do not necessarily have to be at equally spaced angle intervals. The data in the file must be either in blocks or columns, not both, with each block labelled using the names ‘ANGLE’ and ‘RADIUS’ or ‘RADIAL DEVIATION’ as appropriate. Each block or column must be defined with an appropriate unit type. See Note 1 Non-defaultable Text DIAMETER This argument is used to define the diameter of the profile. Non-defaultable Real mm OFFSET This argument is used to define an offset of the profile centre in the Y and Z direction. Default=0,0 Real mm

Notes: 1

An example of a file in which a single profile is defined using block format is as follows: BLOCK NUMBER=1 NAME='ANGLE' UNITS='deg' 0.00000 10.0000 20.0000 30.0000 40.0000 50.0000 60.0000 70.0000 80.0000 90.0000 100.000 110.000 120.000 130.000 140.000 150.000 160.000 170.000 180.000 190.000 200.000 210.000 220.000 230.000 240.000 250.000 260.000 270.000 280.000 290.000 300.000 310.000 320.000 330.000 340.000 350.000 BLOCK NUMBER=2 NAME='RADIAL DEVIATION' UNITS='micron' 1.01086 1.00000 0.99790 0.88764 0.83746 0.82952 0.82177 0.85677 0.85086 0.84090 0.87401 0.95214 0.98992 1.02439 1.03059 1.04399 1.04215 1.03976 1.02032 0.98457 0.90603 0.78639 0.60423 0.50758 0.45371 0.45462 0.48842 0.49955 0.49111 0.49418 0.49719 0.51053 0.49962 0.50857 0.52708 0.62381

The same data in column format would be as follows: COLUMN NUMBER=1 NAME='CRANK ANGLE' UNITS='deg'

570

COMMAND FILES Pre-Processor Commands COLUMN NUMBER=2 NAME='RADIAL DEVIATION' UNITS='micron' 0.00000 1.01086 10.0000 1.00000 20.0000 0.99790 30.0000 0.88764 40.0000 0.83746 50.0000 0.82952 60.0000 . .. .. 340.000 0.52708 350.000 0.62381

If multiple profiles are defined in the same file then each block or column is differentiated using LEGEND = ‘PROFILE 1’. For example, using column format the data would be of the form: COLUMN NUMBER=1 NAME='ANGLE' UNITS='deg' COLUMN NUMBER=2 NAME='RADIAL DEVIATION' UNITS='micron' LEGEND=’PROFILE 1’ COLUMN NUMBER=3 NAME='RADIAL DEVIATION' UNITS='micron' LEGEND=’PROFILE 2’ COLUMN NUMBER=4 NAME='RADIAL DEVIATION' UNITS='micron' LEGEND=’PROFILE 3’ COLUMN NUMBER=5 NAME='RADIAL DEVIATION' UNITS='micron' LEGEND=’PROFILE 4’ 0.00000 1.8583 0.6543 0.1000 0.5689 10.0000 2.5757 0.6783 0.6200 0.4367 20.0000 2.1159 0.2000 0.5300 0.9876 30.0000 2.1354 0.2500 0.3456 . 40.0000 1.4201 0.4500 . . 50.0000 1.4879 . . . 60.0000. . . . .

571


1.23 TITLE The TITLEPre-Processor Commands command is used to define a title and has a single text value. TITLEPreProcessor Commands

VALUE

=

Set Title

where: VALUE This argument is used to set a title. The command `TITLE' may be used three times to define up to three main titles. The text value may be any text (enclosed in quotes) up a maximum length of 30 characters. Default=’ ‘ Text

572

COMMAND FILES Solver Commands

2 Solver Commands This program is used to perform the solution on the model for a given loadcase(s) and may be executed as many times as required by the user. This enables the user to solve for different loading conditions and configurations of the component sub-models. The program requires a command file with the suffix .EDSO. This section defines the commands for this file.

573


2.1 BEARING The command BEARING is used to define the oil temperature at a journal or thrust bearing. BEARING has the following arguments: BEARING

TYPE NUMBER OIL_TEMPERATURE

= = =

Type of bearing Bearing number Oil temperature at bearing

where: TYPE This argument is used to specify the type of bearing. The available options are: MAIN BIG_END SMALL_END The oil temperature of multiple bearings of the same type can be specified using a single BEARING command. Non-defaultable Text NUMBER The number denoting the bearing at which the oil temperature will be applied. Non-defaultable Integer OIL_TEMPERATURE User defined oil temperature at bearing. Non-defaultable Real

574


2.2 BLOCK The BLOCK command is used to define the cylinder block model type. The cylinder block may be rigid, compliant or dynamic BLOCK has the following arguments: BLOCK

TYPE

=

Type of cylinder block model

where: TYPE This argument is used to select the type of cylinder block model to be used in the solution. The cylinder block model options are: DYNAMIC COMPLIANT RIGID_BODY RIGID The models are defined using the PRE-PROCESSOR and the block model specified by the solution is restricted according to the model defined using the PRE-PROCESSOR. If the block has been modelled as compliant using the PREPROCESSOR, a rigid body or dynamic block cannot be solved for. If the block has been modelled as rigid using the PRE-PROCESSOR, a compliant block or a dynamic block cannot be solved for. If the block has been models as ancillary only using the PRE-PROCESSOR, a compliant block or a dynamic block cannot be solved for. Default=DYNAMIC Text

575


2.3 COUPLING The COUPLING command is used to specify the coupling stiffness and damper. In order to ensure that the crankshaft rotates at a constant speed at the flywheel and to constrain the rigid body motion about the crankshaft axis a torsional stiffness and damper is added at the flywheel connected to ground. The value of this stiffness should be such that the natural frequency of the rigid body mode in torsion should be significantly lower than the first flexible torsion mode of the crankshaft. A default value of 50 Hz is usually adequate for most automotive applications COUPLING has the following arguments: COUPLING

STIFFNESS FREQUENCY DAMPING_RATIO

= = =

DAMPING_COEFFICIENT

=

The coupling stiffness value The coupling frequency value The coupling damping ratio value The coupling damping coefficient value

where: STIFFNESS Torsional stiffness applied at the flywheel. See Note 1. Non-defaultable Real

N.m.rad-

FREQUENCY Frequency of the rigid body mode in torsion at the flywheel. See Note 1. Non-defaultable Real Hz DAMPING_RATIO The damping ratio of the coupling damper. See Note 2. Non-defaultable Text

Ratio

DAMPING_COEFFICIENT The damping coefficient of the coupling damper. See Note 2. Non-defaultable Text N.m.s.rad-

576


Notes: 1

The stiffness is defined either by specifying the Stiffness or the Frequency. If the frequency is specified then the corresponding stiffness is calculated, and vice-versa. The frequency is calculated from the polar inertia of the crankshaft J and the stiffness K such that:

f 

1 2

K J

The inertia of the crankshaft J includes the inertia due to the sum of the rotating mass and half the reciprocating mass at each cylinder. 2

The damping coefficient of the coupling damper is defined either by specifying the Coefficient or the damping Ratio If the damping ratio is specified then the corresponding damping coefficient is calculated, and vice versa. The damping coefficient is calculated from the polar inertia of the crankshaft J, the coupling stiffness K and the damping ratio such that

C  2 JK A value of 0.5 for the damping ratio is recommended. The inertia of the crankshaft J includes the inertia due to the sum of the rotating mass and half the reciprocating mass at each cylinder.

577


2.4 DAMPING The DAMPING command is used to define the modal damping characteristics of the crankshaft and cylinder block and the cylinder damping coefficient. If modal damping characteristics for the cylinder block and crankshaft are not defined using this command then default characteristics are used as defined in Appendix 9. If a cylinder damping coefficient is not defined then a default value of 2000 N.s/m is used. This command is repeated as many times as there are components to be assigned userdefined damping values and characteristics. DAMPING has the following arguments: DAMPING

COMPONENT DAMPING_RATIO FREQUENCIES VALUE

= = = =

Component name List of damping values List of frequencies Damping value

where: COMPONENT This argument is used to select the component to which the damping values are to apply. If COMPONENT=CYLINDER_BLOCK or CRANKSHAFT then damping characteristics are defined using the FREQUENCIES and DAMPING_RATIO arguments whilst for COMPONENT=CYLINDER a damping coefficient is defined using the VALUE argument. Non-defaultable Real The type of component determines which information should be included in the command COMPONENT=CYLINDER_BLOCK or CRANKSHAFT If either the cylinder block or crankshaft components are selected a modal damping characteristic is defined as proportions of critical damping (./.crit) against frequency and is specified using the DAMPING_RATIOS and FREQUENCIES arguments. Proportions of critical damping are assigned to each mode of the component by linear interpolation. For modes with a frequency greater than the maximum defined in the FREQUENCIES option a value of 1.0 is assigned to ./.crit whilst for modes with a frequency less than the minimum defined in the FREQUENCIES argument the value of ./.crit at the minimum is assumed. At least two pairs of damping and frequency values are expected. DAMPING_RATIO This argument is used to define the modal damping characteristic and is entered as a list of critical damping values corresponding to frequency values specified using the FREQUENCIES argument. Non-defaultable Real Ratio

578

COMMAND FILES Solver Commands FREQUENCIES This argument is used to specify a list of frequency values corresponding to the damping ratio values specified using the DAMPING_RATIOS argument. Non-defaultable Real Hz COMPONENT=CYLINDER If the cylinder component is selected then a damping coefficient is specified to represent cylinder damping using the VALUE argument. VALUE This argument is used to specify a damping coefficient corresponding to the component selected using the COMPONENT argument. Non-defaultable Real N.s/m

Notes: 1

579


2.5 LUBRICANT The LUBRICANT command is used to specify the type of lubricant used and its fluid properties. LUBRICANT has the following arguments: LUBRICANT

NAME TEMPERATURE PRESSURE

= = =

The name of the lubricant Oil inlet temperature Oil feed pressure

where: NAME This input specifies the name of the lubricant to be used. This lubricant must have been added to the model via the Lubrication Definitions panel or the LUBRICANT pre-processor command. Non-defaultable Text TEMPERATURE This input defines the inlet temperature of the oil to the main journal bearings. o Non-defaultable Real C PRESSURE This input defines the feed pressure of the oil to the journal bearings. Non-defaultable Real bar

Notes: 1

580


2.6 OPEN The OPEN command is used to open the ENGDYN standard data base file. This must be the first command in the file. If this file already exists then it will be overwritten. OPEN

DATABASE

=

Name of the ENGDYN standard data file

where: DATABASE This argument is used to specify the name of the standard data file. The name does not need to contain the .EDSF suffix which is best omitted. Non-defaultable Text

581


2.7 SOLUTION The SOLUTION command is used to specify the type of lubricant used and its fluid properties. SOLUTION has the following arguments: SOLUTION

TYPE

CYLINDER_BLOCK CONNECTING_ROD PRINT_INTERVAL END ABSOLUTE_TOLERANCE RELATIVE_TOLERANCE GLOBAL_TOLERANCE INITIAL_STEP_LENGTH MAXIMUM_NUMBER_OF_STEPS NUMBER_OF_PROCESSORS LOAD SPEED INTERPOLATION CENTRIFUGAL_LOADS DAMPER CP CR OILFILM_SOLUTION BEARING_TEMPERATURE

582

= Solution type: dynamic, indeterminate or determinate Type of cylinder block included in model = Type of connecting rod if included in model = The crank angle interval at which the results are stored = The crank angle at which to end the simulation = The absolute error tolerance = The relative error tolerance = Global error tolerance = The first integration step length = Maximum number of integration steps per CA = Number of processors to run the simulation over = Full, part of no load = RPM of the crankshaft = Interpolation method for calculating the cylinder pressure = Inclusion or exclusion of centrifugal loads = Option to include or exclude vibrations dampers = Piston constant to define the piston friction coefficient = Ring constant to define the ring friction = Method to evaluate the pressure within the bearing oilfilm = Thermal balance calculation option

COMMAND FILES Solver Commands INITIAL_TEMPERATURE RELAXATION_FACTOR TEMPERATURE_TOLERANCE

SAVE_STATIC_RESULT

REDUCED_OUTPUT MAP

SHEAR FLOW_SUMMATION

= Initial temperature of oilfilm for each bearing = Relaxation factor for bearing temperature calculation = Temperature tolerance for bearing temperature calculation = Option to indicate saving of indeterminate results for a dynamic solution = Option to specify reduced output information = This options defines the mapping type to use with the Mobility method = Option to define the film extent = Option to define the calculation of instantaneous oil film flow rate

GREENWOOD_WILLIAMSON_TRIPP = GreenwoodWilliamson-Trip Asperity model option AVERAGE_FLOW_MODEL = Average Flow Model option BACKSUBSTITUTE

= Backsubstitution of results of reduced model onto FE model

where: TYPE This argument is used to specify the solution type. The possible solution types are as follows:

583

COMMAND FILES Solver Commands DETERMINATE

Main bearing reactions are calculated assuming a statically-determinate solution. INDETERMINATE Main bearing reactions are calculated assuming a statically-indeterminate solution but neglect the mass of the crankshaft and cylinder block. DYNAMIC Bearing reactions and vibratory displacements and velocities of the crankshaft and cylinder block (if included) are calculated for a fully threedimensional dynamic model. If TYPE=DYNAMIC the statically-determinate and statically-indeterminate solutions are also performed. Default=DYNAMIC Text CYLINDER_BLOCK This argument is used to select the type of cylinder block model to be used in the solution. The cylinder block model options are: DYNAMIC COMPLIANT RIGID_BODY RIGID The models are defined using the PRE-PROCESSOR (GUI or EDPR file) and the block model specified by the solution is restricted according to the model defined using the PRE-PROCESSOR. If the block has been modelled as compliant using the PREPROCESSOR, a rigid body or dynamic block cannot be solved for. If the block has been modelled as rigid using the PRE-PROCESSOR, a compliant block or a dynamic block cannot be solved for. If the block has been modelled as ancillary only using the PRE-PROCESSOR, a compliant block or a dynamic block cannot be solved for. Default=DYNAMIC Text CONNECTING_ROD This option defines whether the connecting rod is rigid, compliant or dynamic in any small end or big end bearing EHL solutions. This option only applies if the solution TYPE is DETERMINATE and the bearing oilfilm solution is not MOBILITY as specified by the OILFILM_SOLUTION option. For INDETERMINATE and DYNAMIC solutions the connecting rod type is automatically set to rigid and dynamic respectively. If no connecting rod models have been defined as DYNAMIC or COMPLIANT using the CONROD command, this option does not apply. The options are: RIGID COMPLIANT DYNAMIC Default=RIGID

Text

END This argument is used to specify the crank angle (CA) at which the solution is to finish. If the solution has converged satisfying the GLOBAL_TOLERANCE at a smaller crank angle then the solution will finish at that crank angle. This argument is ignored for all solution types other than TYPE=DYNAMIC Default=2880.0 Real Degrees

584

COMMAND FILES Solver Commands PRINT_INTERVAL This argument is used to specify the print interval in ECA for storing of results. Default=1.0 Real ECA The next five inputs are only needed if the TYPE is INDETERMINATE or DYNAMIC ABSOLUTE_TOLERANCE This argument is used to specify the absolute error tolerance. Default=1.0x10-9 Real m RELATIVE_TOLERANCE This argument is used to specify the relative error tolerance. Default=0.001 Text GLOBAL_TOLERANCE This argument is used to specify the global error tolerance,  g . The solution at a given crank angle n has converged provided the equation,

| Fn-720 - Fn |   g | Fn |

Default=0.001

Real

INITIAL_STEP_LENGTH This argument is used to specify the step length to be attempted on the first integration step. This argument is ignored for all solution types other than TYPE=DYNAMIC. Default=1.0x10-6 Real s MAXIMUM_NUMBER_OF_STEPS This argument is used to specify the maximum number of integration steps that may be attempted per ECA. If this number is exceeded the solution is terminated with an error. This argument is ignored for all solution types other than TYPE=DYNAMIC. Default=100 Integer NUMBER_OF_PROCESSORS This option specifies the number of cores or processors over which to run the ENGDYN solution. Default=1 Integer

585

COMMAND FILES Solver Commands LOAD This is used to specify the loading condition to apply at the given engine speed. Full Load, Part Load and No Load refer to the loading found in the .EDSF file and defined by the GUI or by LOADING command. The options are: FULL_LOAD PART_LOAD NO_LOAD INERTIA_ONLY Inertia_Only will only apply the inertia loading to the crankshaft (no cylinder pressure is applied). Default=FULL_LOAD Text SPEED This option is used to specify the speed at which the crankshaft is rotating. The user may define multiple loadcases to perform a speed sweep by specifying multiple speeds. In this case, each speed will be a separate loadcase. Non-defaultable Real RPM INTERPOLATION This option menu is used to specify the interpolation method for interpolating the cylinder pressure at the given engine speeds and load from the user defined cylinder pressure maps. Two methods are available, SPLINE or LINEAR. This option is only relevant if the In-Cylinder MODEL was defined as SLIDER_CRANK using the GUI or the IN_CYLINDER command. Default=LINEAR Text CENTRIFUGAL_LOADS This option is used to select whether the centrifugal loads due to the mass and inertia of the crankshaft are EXCLUDED or INCLUDED and is relevant only for a quasistatic solution and if the crankshaft model includes mass. See Note 1. Default=See Note 1 Text DAMPER This options defines whether the crankshaft vibration dampers defined using the DAMPER will be included or excluded in the ENGDYN solution. The options are INCLUDED or EXCLUDED. Default=INCLUDED Text CP This input defines the piston constant which defines the piston friction coefficient. See Note 2. Default=92.4 Real CR This input defines the ring constant which defines the ring friction. See Note 3. Default=159 Real

586

COMMAND FILES Solver Commands OILFILM_SOLUTION This option defines the method which is used to model the bearing oilfilm. The options are: MOBILITY FINITE_DIFFERENCE FINITE_VOLUME HALF_SOMMERFELD HALF_SOMMERFELD, MASS_CONSERVING REYNOLDS, DEPRECATED REYNOLDS, MASS_CONSERVING REYNOLDS, JAKOBSSON_FLOBERG_OLSSON The FINITE_DIFFERENCE and FINITE_VOLUME options are deprecated and are included for backwards compatibility. They are equivalent to the HALF_SOMMERFELD and REYNOLDS, DEPRECATED options respectively. Default= HALF_SOMMERFELD, Text MASS_CONSERVING BEARING_TEMPERATURE This argument specifies the method to calculate the bearing oilfilm temperatures. The options are: USER_DEFINED The bearing temperatures are defined by the user THERMAL_BALANCE Temperature is averaged over the bearing THERMAL_BALANCE, AXIAL Temperatures vary axially RADIAL_HEAT_TRANSPORT Temperatures vary axially and circumferentially. Heat is only transported radially on the metal layer. THERMAL_BALANCE, NODAL Temperatures vary axially and circumferentially. Heat is transported radially and circumferentially on the metal layer. USE_STATIC_TEMPERATURES The temperatures are taken from the static analysis preceeding the dynamic analysis Default=USER_DEFINED

Text

The following three inputs are only need if BEARING_TEMPERATURE is not defined as USER_DEFINED. INITIAL_TEMPERATURE This option defines the initial temperature of the bearing oilfilm. The input can either be ‘SUPPLY’ or a user defined temperature. If ‘SUPPLY’ is chosen the initial temperature is set to the oil inlet temperature. o Default=SUPPLY Text or Real C RELAXATION_FACTOR This parameter is used to control the temperature that is used on the next thermal balance iteration step. See Note 44. Default=0.4 Real Non-Dimensional

587

COMMAND FILES Solver Commands TEMPERATURE_TOLERANCE This parameter is used to specify the temperature tolerance for the thermal balance iteration. See Note 5. o Non-defaultable Text C MAP This option specifies the mobility map to be used. The options are: SHORT_BEARING FINITE_BEARING When the ratio of bearing length to bearing diameter ranges between 0.75 and 4, a bearing is assumed to be a finite length bearing. The SHORT_BEARING option must be used for indeterminate and dynamic solutions. Default=SHORT_BEARING Text SHEAR This argument is used to denote whether the film extent assumed for the calculation of the oil film shear power loss for each journal bearing is  (PI) or 2 (2PI) when the mobility method is applied. Default=PI Text FLOW_SUMMATION This parameter is used to specify which method is used for adding the pressure and velocity induced flow when calculating the instantaneous oil film flow rate using the mobility method for each journal bearing. The options are: SUMMATION Induced oil rate due to pressure and velocity is used PRESSURE_ONLY Induced oil rate due to pressure is used SELECTIVE_SUMMATION Large of the two incuded flow rates is used At present only oil flow due to feed pressure is available. Default=PRESSURE_ONLY Text The next two inputs are required only for solutions which contain oilfilms not solved using the MOBILITY OILFILM_SOLUTION option. GREENWOOD_WILLIAMSON_TRIPP This input defines the method to calculate the composite asperity parameters and whether the GWT formulae contains the summit height correction. The options are: v4_DEPRECATED, RMS_SURFACE_HEIGHT v4_DEPRECATED, SUMMIT_HEIGHT_CORRECTION HERTZIAN, RMS_SURFACE_HEIGHT HERTZIAN, SUMMIT_HEIGHT_CORRECTION For further details please see the Boundary Lubrication Model section of the Theory Manual. Default= HERTZIAN, Text SUMMIT_HEIGHT_CORRECTION

588

COMMAND FILES Solver Commands AVERAGE_FLOW_MODEL This option defines whether the Average Flow Model of the Reynolds equation is enabled or disabled. The options are: ENABLED DISABLED For further details please see the Average Flow Model section of the Theory Manual. Default=DISABLED Text SAVE_STATIC_RESULT This option is used to specify whether the statically-indeterminate solution used as an initial condition for the dynamic solution is to be saved and is only relevant for a dynamic solution. If this is set to YES the statically-indeterminate results are stored as a separate loadcase. Default=NO Text REDUCED_OUTPUT This option is used to control whether or not the oil film thickness’ and oil film pressures at each crank angle and bearing angle are to be saved to file on completion of each loadcase. If these results are saved there is a large overhead in terms of the size of the .EDSF file. Once this has been set for the first loadcase it cannot be subsequently changed. Default=YES Text BACKSUBSTITUTE For those models that are compliant or dynamic model and have been derived using the finite element method it is possible to perform data recovery from the reduced model (used by ENGDYN) to the complete finite element model. This uses a data recovery matrix derived during the matrix reduction procedure and which is stored in the SFE file. If this is selected then at the end of each completed loadcase ENGDYN will perform the data recovery for each model from the reduced degrees of freedom to all the nodes of the reduced model. This will only happen for each model if both the SFE exists and the data recovery matrix is in the file Default=NO Text

589


Notes: 1

The EXCLUDED option would be used when the results of the solution are to be used as applied loads for a static stress analysis of the crankshaft in which case only the influence due to cylinder gas pressure is considered. When the crank has mass the default is INCLUDED, when the crank does not have mass the default is EXCLUDED.

2

The piston constant C p is the constant in the equation

 .  f p  Cp  W 

0.5

defining the friction coefficient for the piston where where

3

W




= piston velocity




The ring constant C r is the constant in the equation

 .  f r  Cr   W 

0.5

defining the friction coefficient for each ring where where

4

W




= piston velocity




The relaxation factor for each journal bearing is defined as:

Tn  Tn - 1 + F .T - Tn - 1 where

590


5

F

= Relaxation factor

Tn

= Journal oil bearing temperature for next step.

Tn-1

= Journal oil bearing temperature at previous step

T

= Journal oil bearing temperature calculated from power loss and oil flow rate

The solution for a given cycle has converged providing the equation,

Tn1  Tn   T is satisfied for all journal bearings where T is the temperature tolerance.

591

COMMAND FILES Post-Processor Commands

3 Post-Processor Commands This section defines the commands that may be used in the POST-PROCESSOR. This program is used to post-process the results of the analyses performed by the SOLVER. The program requires a command file with the suffix .EDPO. This provides a means of calculating displacements and stresses for complete finite element models of the crankshaft and cylinder block by a number of methods using loads calculated using the `reduced' ENGDYN models.

3.1 CRANK_ANALYSIS The CRANK_ANALYSIS command is used to perform a crankshaft stress analysis using the operating loads calculated by ENGDYN. For more information on crank shaft stress analyses in ENGDYN please see Section Crank Shaft Stress Analysis. CRANK_ANALYSIS has the following arguments: CRANK_ANALYSIS

LOADCASES METHOD OUTPUT OUTPUT_NAME FILE FROM TO UNIT CENTRIFUGAL_LOAD UNIT_MOMENTS BENDING_MOMENTS INTERVAL OFFSET SAFETY_CALCULATION OVERLOAD SCF GEOMETRY AFTER_TREATMENTS STRENGTHS MATERIAL MATERIAL_TYPE 592

= List of ENGDYN loadcases = Stress analysis method = Requested output = Output file name = List of crankshaft model names = File format of input data = File format of output data = Units of input data = Argument to include or exclude centrifugal loads = Argument to include or exclude unit moments = Type of bending moments = Crank angle interval of output = Crank angle offset of output = Method to calculate safety factors = Overload option = Name of the file containing user-defined SCF values = Name of the file containing geometry = Name of file containing after treatments = Name of the file containing elevated material strengths = Crankshaft material name = Type of material

COMMAND FILES Post-Processor Commands MATERIAL_FILE SIZE_FACTOR SIZE_FACTOR_EQUATION SIZE_FACTOR_UNITS FNF_CALCULATION FNF_EQUATION VIBRATORY_LOAD TOURQES SETS DYNAMIC_LOADCASES SAFETY_ALGORITHMS

= Name of file containing material properties = Name of file containing table of size factors = Size factor equation = Units for size factor diameter = Type of Fatigue Notch Factor calculation = Fatigue Notch Factor equation = Type of vibratory load to apply = Base name of vibratory load output = List of node sets to be stressed. List of ENGDYN dynamic loadcases Algorithm to calculate safety factors

where: LOADCASES This argument specifies the list of ENGDYN loadcases to use in the crankshaft analysis. Non-defaultable Integer METHOD This argument specifies the type of stress analysis to be undertaken on the crankshaft. The options are: CLASSICAL FE QUASI-STATIC Non-defaultable Text OUTPUT This argument specifies the type of output required. The argument only applies if METHOD is set to FE QUASI-STATIC; CLASSICAL will always output STRESSES AND SAFETY FACTORS. The options are: STRESSES AND SAFETY FACTORS UNIT LOADS Default=STRESSES AND Text SAFETY FACTORS

OUTPUT_NAME This argument specifies the name of the output file. The output must be specified for the CLASSICAL METHOD. For the FE QUASI-STATIC METHOD, the output file 593

COMMAND FILES Post-Processor Commands name will default to the name of the first FEA model name which contains the unit loading. See above Text FILE This argument specifies a list of crankshaft model names where the first model name contains the unit loading. This argument is only needed where METHOD has been defined as FE QUASI-STATIC. Non-defaultable Text FROM This argument specifies the format of the data within the FE model file. This argument is only needed where METHOD has been defined as FE QUASI-STATIC. The options are: SFE_RICARDO IDEAS MSC_NASTRAN Default=SFE_RICARDO Text TO This argument specifies the format in which to output the data. This argument is only needed where METHOD has been defined as FE QUASI-STATIC. The options for this argument are dependent on the type of OUTPUT specified. The options, where OUTPUT is STRESSES AND SAFETY FACTORS, are: RICARDO-SFE IDEAS FEMVIEW Default=RICARDO-SFE Text Where the OUTPUT is UNIT LOADS the options are: FEARCE ABAQUS ANSYS IDEAS NASTRAN Default=FEARCE Text UNIT This argument specifies the units of the data within the FE model file. This argument is only needed where METHOD has been defined as FE QUASI-STATIC. The options are: SI MM METRIC FT INCH The model will then be translated into SI units. Default=SI Text CENTRIFUGAL_LOAD 594

COMMAND FILES Post-Processor Commands This argument specifies if centrifugal loads will be included in the stress analysis. . This argument is only needed where METHOD has been defined as FE QUASI-STATIC. The options are: EXCLUDED INCLUDED Default=INCLUDED Text UNIT_MOMENTS This argument specifies if unit moments will be included in the stress analysis. This argument is only needed where METHOD has been defined as FE QUASI-STATIC. The options are: EXCLUDED INCLUDED Default=INCLUDED

Text

BENDING_MOMENTS This argument specifies the bending moment calculation type. The argument specifies whether the bending moment calculation will be calculated from a single bay or the complete crankshaft. This argument is only needed where METHOD has been defined as CLASSICAL. The options are: SINGLE BAY COMPLETE CRANKSHAFT Default=SINGLE BAY Text INTERVAL This argument sets the crank angle interval of the output. If this argument is not set the crank angle interval at which the ENGDYN results are stored is used. This argument is only needed where METHOD has been defined as FE QUASI-STATIC. Default=see above Real Degrees OFFSET This argument sets the crank angle offset of the output. This argument is only needed where METHOD has been defined as FE QUASI-STATIC. Default=0.0 Real Degrees SAFETY_CALCULATION This argument specifies the calculation to evaluate the safety factors. The options depend on the type of METHOD chosen. Multiple safety calculation methods can be specified. For FE QUASI-STATIC the options are: GOODMAN ALTERNATIVE GOODMAN MULTIAXIAL FATIGUE PRINCIPAL GOODMAN DANG VAN LINEAR SWT P1 ALTERNATIVE GOODMAN

For CLASSICAL the options are: 595

COMMAND FILES Post-Processor Commands GOODMAN Non-defaultable

Real

OVERLOAD This argument specifies the type of overload calculation for Goodman based safety factor calculations. The options are: CYCLIC GENERAL Default=CYCLIC Text

SCF This argument specifies the name of the file containing user-defined Stress Concentration Factors. The file must have a .SCF suffix. The format of this file is described in Section 3.1.1.1. This argument is optional and if it is not supplied the SCF values will be calculated from the geometry. Default=see above Text GEOMETRY This argument specifies the name of the file containing geometry. The file must have a .GEOM suffix. The format of this file is described in Section 3.1.1.2. Non-defaultable Text AFTER_TREATMENTS This argument specifies the name of the file containing after treatments. The file must have a .AT suffix. The format of this file is described in Section 3.1.1.3. Non-defaultable Text STRENGTHS This argument specifies the name of the file containing elevated material strengths. The file must have a .STR suffix. If the file is not supplied then the strengths are calculated from the after treatments and the size factors. If the argument is not present, the AFTER_TREATMENTS argument must be specified. The format of this file is described in Section 3.1.1.4. Default=see above Text MATERIAL This argument specifies the name of the crankshaft material. This argument is optional. Default=NULL Text MATERIAL_TYPE This argument specifies the type of crankshaft material. The options are STEEL SG_IRON Non-defaultable Text

596

COMMAND FILES Post-Processor Commands MATERIAL_FILE This argument specifies the name of the file containing the material properties. The file must gave a .MAT suffix. Non-defautable Text SIZE_FACTOR This argument specifies the name of the file containing the table of size factor values. The file must have a .SIZE suffix. The SIZE_FACTOR and SIZE_FACTOR_EQUATION arguments are mutually exclusive. See Note 1. Non-defautable Text SIZE_FACTOR_EQUATION This argument specifies the size factor equation. The SIZE_FACTOR and SIZE_FACTOR_EQUATION arguments are mutually exclusive. The SIZE_FACTOR_UNITS argument must be present if this argument is specified. See Note 1. Non-defautable Text SIZE_FACTOR_UNITS This argument specifies the units corresponding to the diameter to be used in the size factor equation set by the SIZE_FACTOR_EQUATION argument. This argument must be present if the SIZE_FACTOR_EQUATION is specified. See Note 1. Non-defautable Text FNF_CALCULATION This argument specifies the Fatigue Notch Factor (FNF) calculation method. The options are: MODIFIED LOWELL EQUATION User Defined If EQUATION is specified then the argument FNF_EQUATION must be present. If the argument is specified as User Defined, where User Defined denotes the name of an FNF calculation, an environment variable should be set called FNF_CALCULATION_METHOD which corresponds to the name of the User Defined FNF calculation. Furthermore, another environment variable FNF_STEEL_EQUATION or FNF_SG_IRON_EQUATION, depending on the specification of the MATERIAL_TYPE argument, should be set specifying the form of FNF equation. Default=MODIFIED LOWELL Text FNF_EQUATION This argument specifies the equation to be used to calculate the Fatigue Notch Factors and must be present if the FNF_CALCULATION is specified as EQUATION. Non-defaultable Text VIBRATORY_LOAD This argument specifies the vibration load to be applied to the crankshaft stress analysis. The options are: NONE 597

COMMAND FILES Post-Processor Commands MAX/MIN VARYING If the argument is set to MAX/MIN or VARYING then the TOURQES argument must be set. Default=NONE Text

TOURQES This argument specifies the base name of the file containing the vibratory loads which must be set if the VIBRATORY_LOAD argument is present. If there is a file with basename.sdf, the vibratory loads are considered to be from VALDYN, if not the loads taken to be Torsional Vibration (TVFORCED) loads. Non-defaultable Text SETS This argument specifies the list of node sets to be stressed and is only needed when the METHOD argument is set to FE QUASI-STATIC. Default=No sets specified Text DYNAMIC_LOADCASES This argument specifies the list of ENGDYN dynamic loadcases to use in the crankshaft analysis. The number of dynamic loadcases should be consistent with the number of loadcases declared using the LOADCASES argument. This argument should only be used when the METHOD argument is set to FE QUASI-STATIC. Default=No sets specified Text SAFETY_ALGORITHMS This argument specifies the type of algorithm to use when calculating the safety factors where the METHOD has been specified as CLASSICAL. The options are: MAX/MIN STRESS RIGOROUS Default=MAX/MIN STRESS Text WEB_MODULUS This argument specifies the type of algorithm to use when calculating the web modulus where the METHOD has been specified as CLASSICAL. The options are: SIMPLIFIED RIGOROUS Default=SIMPLIFIED Text EXPORT This argument specifies what to output when the output type STRESSES AND SAFETY FACTORS. The options are: NONE SETS 598

COMMAND FILES Post-Processor Commands EXTERNAL ALL Default=SETS

Text

MEAN_TORQUES This argument specifies the file from which to read the user-defined mean torques. The file should have a .TORQUES suffix. If this argument is not present the mean torques are taken as indicated. Default=Indicated Text

3.1.1

Additional Files

Four files should be supplied by the user using arguments in the CRANK_ANALYSIS command. These are used to supply data relating to crankshaft geometry, Stress Concentration Factor (SCF) values, after treatments and fatigue strengths. 3.1.1.1 Stress Concentration Factor File The stress concentration file is optional and is specified using the SCF argument of the CRANK_ANALYSIS command. The file has a .SCF suffix. The file is used either to supply user-specified SCF values or to define alternative methods for calculating the SCF values of fillets and oilholes. If this file is not supplied, the SCF values are calculated using the Lowell et al and Peterson et al methods for fillets and oilholes respectively. The file can be used to specify SCF values for particular fillets or oilholes whilst the remainder are calculated using the default methods. There are 2 keywords which are used in the SCF File: FILLET OILHOLE 3.1.1.1.1

Keyword FILLET

FILLET JTYPE NUMB LOCATION METHOD SCFBending SCFTorsion JTYPE NUMB LOCATION METHOD SCFBending SCFTorsion

Journal Type (‘MAIN’ or ‘PIN’) Main bearing or cylinder number (dependent on JTYPE) Fillet location Method for calculating SCF values SCF in Bending for the leading oilhole SCF in Torsion for the leading oilhole

Example: FILLET PIN, 1, REAR, User, 3.45, 2.98 3.1.1.1.2

Keyword OILHOLE

OILHOLE JTYPE NUMB METHOD TYPE SCFBending1 SCFTorsion1 SCFBending1 SCFTorsion1 JTYPE

Journal Type (‘MAIN’ or ‘PIN’)

599

COMMAND FILES Post-Processor Commands NUMB METHOD TYPE SCFBending1 SCFTorsion1 SCFBending2 SCFTorsion2

Main bearing or cylinder number (dependent on JTYPE) Method for calculating SCF values. This can be equal to ‘Peterson et al’, the method defined using the environment variable OILHOLE_SCF_CALCULATION_METHOD or ‘User’ Oilhole type. This can be equal to ‘Single (Leading)’, ‘Single (Trailing)’, Twin or Cross. SCF in Bending for the leading oilhole. This is supplied if METHOD=User and TYPE=‘Single (Leading)’, Twin or Cross, otherwise the value should be blank. SCF in Torsion for the leading oilhole. This is supplied if METHOD=User and TYPE=‘Single (Leading)’, Twin or Cross, otherwise the value should be blank. SCF in Bending for the trailing oilhole. This is supplied if METHOD=User and TYPE=‘Single (Trailing)’ or Twin, otherwise the value should be blank. SCF in Torsion for the trailing oilhole. This is supplied if METHOD=User and TYPE=‘Single (Trailing)’ or Twin, otherwise the value should be blank.

Example: OILHOLE PIN, 1, User, ‘Single (Trailing)’, , , 3.45, 2.98 3.1.1.2 Geometry File The geometry file is specified using the GEOMETRY argument of the CRANK_ANALYSIS command and has a .GEOM suffix. The geometry file contains geometry data of the journal fillets, journal oilholes, journal lands, journals and crankwebs. The data contained in this file defines which areas of the crankshaft are to be stressed. For example, if the file only contains data relating to the rearmost crank pin oilhole then that is the only are to be stressed. There are 5 keywords which are used in the Geometry File: FILLET OILHOLE JOURNAL LAND CRANKWEB 3.1.1.2.1 FILLET

Keyword FILLET JTYPE

NUMB

JTYPE NUMB LOCATION SETNAME RADIUS WEB_UNDERCUT JOURNAL_UNDERCUT

LOCATION

SET

RADIUS

WEB UNDERCUT

JOURNAL UNDERCUT

Journal Type (‘MAIN’ or ‘PIN’) Main bearing or cylinder number (dependent on JTYPE) Fillet location (‘REAR’ or ‘FRONT’) Node set name (If blank then default name is used) Fillet radius [m] Fillet web undercut [m] Fillet journal undercut [m]

600

COMMAND FILES Post-Processor Commands Example: FILLET MAIN, 1, FRONT, MAIN_FILLET_REAR_01, 0.003, 0.000 3.1.1.2.2

Keyword OILHOLE

SHOLUD HAVE OFFSET AS WELL OILHOLE JTYPE NUMB TYPE SET1 d1 h1 o1 x1 y1 z1 SET2 d2 h2 o2 x2 y2 z2

JTYPE NUMB TYPE

Journal Type (‘MAIN’ or ‘PIN’) Main bearing or cylinder number (dependent on JTYPE) Oilhole type. This can be equal to ‘Single (Leading)’, ‘Single (Trailing)’, Twin or Cross. Node set name (If blank then default name is used). Oilhole diameter [m]. Height of the oilhole breakout with respect to journal centre [m]. Oilhole axial offset [m] Vector defining orientation of oilhole with respect to breakout [m].

SET d

h o x, y, z Example:

OILHOLE PIN, 1, ‘Single (Leading)’, , 0.005, 0.010, 0.0, 0.010, 0.002, 0.0 3.1.1.2.3

Keyword JOURNAL

JOURNAL JTYPE NUMB SET JTYPE NUMB SET

3.1.1.2.4

Journal Type (‘MAIN’ or ‘PIN’) Main bearing or cylinder number (dependent on JTYPE) Node set name (If blank then default name is used).

Keyword LAND

LAND JTYPE NUMB LOCATION SETNAME DIAMETER THICKNESS JTYPE NUMB LOCATION SETNAME DIAMETER THICKNESS

Journal Type (‘MAIN’ or ‘PIN’) Main bearing or cylinder number (dependent on JTYPE) Land location (‘REAR’ or ‘FRONT’) Node set name (If blank then default name is used). Facing land diameter [m] Facing land thickness [m]

601


3.1.1.2.5

Keyword CRANKWEB

CRANKWEB NUMB ThicknessReduced WidthMin WidthMax CRANKWEB

NUMB ThicknessReduced WidthMin WidthMax

= = = =

Crankshaft web number Reduced thickness of web [m] Minimum width [m] Maximum width [m]

3.1.1.3 After Treatment File The after treatment file is optional and is specified using the AFTER_TREATMENTS argument of the CRANK_ANALYSIS command and has a .AT suffix. If this file is not supplied it is assumed that there is no after treatment of the crankshaft. This file contains the after treatment process and fatigue lift factor due to that treatment for each area (fillets, oilholes, lands and journals) of the main and pin journals. The possible after treatment processes for each area of a journal are summarised in Table 19and the default factors for each of these processes are summarised in Table 20.

Fillets Oilholes Journals Lands

None

ColdRolled

   

   

Rolled and Nitrided    

Tuffrided

Induction Hardened

Nitrided

   

   

   

Tuffrided

Induction Hardened

Nitrided

1.15

1.5

1.5

1.15 1.15 1.15

1.5 -

1.5 1.5 1.5

Table 19: After Treatment Processes

None

ColdRolled

Fillets

1.0

Oilholes Journals Lands

1.0 1.0 1.0

1.7 (SGIron) 1.4 (Steel) -

Rolled and Nitrided 2.5 (SGIron) 2.1 (Steel) -

Table 20: Default After Treatment Factors

There are 2 keywords which are used in the After Treatments File: PIN MAIN

602

COMMAND FILES Post-Processor Commands 3.1.1.3.1

Keyword PIN

PIN LOCATION TREATMENT FACTOR LOCATION TREATMENT FACTOR

3.1.1.3.2

Location (‘FILLETS’, ‘OILHOLES’, ‘LANDS’ or ‘JOURNALS’) One of the processes listed in Table 19 Fatigue lift factor. (Defaults to value in Table 20 if nonsupplied)

Keyword MAIN

MAIN LOCATION TREATMENT FACTOR LOCATION TREATMENT FACTOR

Location (‘FILLETS’, ‘OILHOLES’, ‘LANDS’ or ‘JOURNALS’) One of the processes listed in Table 19 Fatigue lift factor. (Defaults to value in Table 20 if nonsupplied)

3.1.1.4 Strengths File The strengths file is specified using the optional STRENGTHS argument of the CRANK_ANALYSIS command and has a .STR suffix. If a strengths file is not supplied the elevated strengths are calculated from the supplied size factors and after treatment factors. The strengths file defines the elevated values of Ultimate Tensile Strength (UTS), Tensile and Compressive Yield Strengths and Infinite-Life Fatigue strength for journal fillets, journal oilholes, journal lands and journals. For those areas which have no strength defined it is assumed that the base strengths apply. There are 4 keywords which are used in the Strengths File: FILLET OILHOLE JOURNAL LAND

3.1.1.4.1

Keyword FILLET

FILLET JTYPE DIAMETER σUTS σTYS σCYS σFS JTYPE DIAMETER σUTS

Journal type (‘MAIN’ or ‘PIN’) Journal diameter Ultimate Tensile Strength [Nm-2]

603

COMMAND FILES Post-Processor Commands Tensile Yield Strength [Nm-2] Compressive Yield Strength [Nm-2] Infinite-Life Fatigue Strength [Nm-2]

σTYS σCYS σFS

3.1.1.4.2

Keyword OILHOLE

OILHOLE JTYPE DIAMETER σUTS σTYS σCYS σFS JTYPE DIAMETER σUTS σTYS σCYS σFS

3.1.1.4.3

Journal type (‘MAIN’ or ‘PIN’) Journal diameter Ultimate Tensile Strength [Nm-2] Tensile Yield Strength [Nm-2] Compressive Yield Strength [Nm-2] Infinite-Life Fatigue Strength [Nm-2]

Keyword JOURNAL

JOURNAL JTYPE DIAMETER σUTS σTYS σCYS σFS JTYPE DIAMETER σUTS σTYS σCYS σFS

3.1.1.4.4

Journal type (‘MAIN’ or ‘PIN’) Journal diameter Ultimate Tensile Strength [Nm-2] Tensile Yield Strength [Nm-2] Compressive Yield Strength [Nm-2] Infinite-Life Fatigue Strength [Nm-2]

Keyword LAND

LAND JTYPE DIAMETER σUTS σTYS σCYS σFS JTYPE

Journal type (‘MAIN’ or ‘PIN’) 604

COMMAND FILES Post-Processor Commands DIAMETER σUTS σTYS σCYS σFS

Journal diameter Ultimate Tensile Strength [Nm-2] Tensile Yield Strength [Nm-2] Compressive Yield Strength [Nm-2] Infinite-Life Fatigue Strength [Nm-2]

Notes: 1

The size factors can be specified using three methods: User-defined size factors User-defined size factor equation and units Default equations depending on the material type To choose the first method set the SIZE_FACTOR argument to a file containing the user defined size factors. To choose the second method set the SIZE_FACTOR_EQUATION and SIZE_FACTOR_UNITS arguments. If neither SIZE_FACTOR or SIZE_FACTOR_EQUATION are present the third method will be used.

605


3.2 BLOCK_ANALYSIS The BLOCK_ANALYSIS command is used to perform a crankshaft stress analysis using the operating loads calculated by ENGDYN. For more information on block analyses in ENGDYN please see Section Cylinder Block Analysis. BLOCK_ANALYSIS has the following arguments: BLOCK_ANALYSIS

OUTPUT LOADCASES OUTPUT_NAME FILE FROM UNIT ROTATE TRANSLATE MIRROR TO CRANK_ANGLES

INTERVAL OFFSET THERMAL_LOADCASE ASSEMBLY_LOADS NOTCH_SENSITIVITY SAFETY_CALCULATION OVERLOAD

606

= Requested output = List of ENGDYN loadcases = Output file name = List of crankshaft model names = File format of input data = Units of input data = Vector rotations about global x,y,z axes = Translation in global x,y,z axes = Normal vector for mirroring = File format of output data = Crank angles at which to perform calculations = Crank angle interval = Crank angle offset = Include thermal load = Include assembly loads = Notch sensitive locations = Calculate fatigue safety factor = Overload option for Goodman safety calculation

COMMAND FILES Post-Processor Commands STARTING_LOADCASE UNIT_MOMENTS

EXPORT FREQUENCY_RANGE

ORDERS LOADCASE_OPTION DAMPING DAMPING_EQUATION DAMPING_EQUATION_UNITS VIBRATION_SETS NOISE_SETS NOISE_METHOD

INTEGRATION_ORDER_INCREMENT

DELETE_VIBRATIONS DENSITY SPEED_OF_SOUND

where:

607

= Starting loadcase in SFE file = Argument to include or exclude unit moments = SFE export option = Minimum and maximum frequencies for forced response loading = Engine order frequencies = Option for specifying engine orders = Type of applied damping = Specification of damping equation = Units of damping equation = Vibration sets = Noise radiating sets = Method to calculate sound intensity and power = Integration order for Helmholtz calculation = Delete nodal vibration data = Fluid density = Fluid speed of sound

COMMAND FILES Post-Processor Commands OUTPUT This argument specifies the type of output produced. The available options are: ‘QS Unit Loads’ ‘QS Unit Load Factors’ ‘QS Combined Results’ ‘QS Loading’ ‘Frequency Response Loading’ ‘Nodal Vibrations’ ‘Radiated Noise’ Some options will also automatically run other analyses. QS Combined Results’ will run ‘QS Unit Load Factors’, ‘Nodal Vibrations’ will run ‘Frequency Response Loading’ and ‘Radiated Noise’ will run ‘Nodal Vibrations’. The requirement for all other arguments is dependent on the OUTPUT option. A tick in the table below indicates that an argument is relevant to the stated OUTPUT option in the top column. QS Unit Loads LOADCASES OUTPUT_NAME FILE FROM UNIT ROTATE TRANSLATE MIRROR TO

       

INTERVAL OFFSET THERMAL_LOADCASE ASSEMBLY_LOADS NOTCH_SENSITIVITY SAFETY_CALCULATION OVERLOAD UNIT_MOMENTS EXPORT

               

CRANK_ANGLES

STARTING_LOADCASE

QS Unit Load Factors

  

FREQUENCY_RANGE ORDERS LOADCASE_OPTION DAMPING DAMPING_EQUATION DAMPING_EQUATION_UNITS VIBRATION_SETS NOISE_SETS NOISE_METHOD INTEGRATION_ORDER_INCREMENT DELETE_VIBRATIONS DENSITY SPEED_OF_SOUND

Non-defaultable

Integer 608

QS Combined Results

                 

QS Loading

           

Frequency Response Loading

Nodal Vibrations

Radiated Noise

        

        

        

      

       

             



COMMAND FILES Post-Processor Commands LOADCASES This argument specifies the list of ENGDYN loadcases to use in the block analysis. Non-defaultable Integer OUTPUT_NAME Name of FEARCE .FRC produced by ENGDYN containing instructions for the block analyses. See Note 1 Text FILE This argument specifies name of the file containing the FE model of the block. Non-defaultable Text FROM This argument identifies the format of the file specified by the FILE argument. The options are: SFE IDEAS MSC/NASTRAN Default=SFE Text UNIT This argument specifies the units of the data within the FE model file. The options are: SI MM METRIC FT INCH The model will then be translated into SI units. Default=SI Text ROTATE This argument is used to specify the vector rotations about the x,y and z crankshaft global axes. For further explanation see Note 2. Default=(0,0,0) Real degrees TRANSLATE This argument is used to specify the translation vector in crankshaft global axes. For further explanation see Note 2. Default=(0,0,0) Real m MIRROR This argument is used to specify the normal vector to the mirror (the plane of which passes through the origin at the crankshaft coordinate system). Default is to not perform any mirroring. For further explanation see Note 2. Default=No mirroring Real

609

COMMAND FILES Post-Processor Commands TO This argument specifies the format of the file to which the analysis results will be exported to. The options are: SFE IDEAS MSC/NASTRAN Default=SFE Text CRANK_ANGLES This argument specifies the list of crank angles at which to undertake the chosen analysis. See Note 3. Non-defaultable Real INTERVAL This argument specifies the crank angle interval, which together with the value defined by the OFFSET argument, is used to generate the crank angles at which to undertake the chosen analysis. See Note 3. Non-defaultable Real OFFSET This argument specifies the crank angle interval, which together with the value defined by the INTERVAL argument, is used to generate the crank angles at which to undertake the chosen analysis. See Note 3. Default=0.0 Real THERMAL_LOADCASE This argument identifies the loadcase containing thermal results in the FE model file specified by the FILE argument. Default=No thermal loads Integer ASSEMBLY_LOADS This argument identifies the loadcase containing thermal results in the FE model file specified by the FILE argument. Default=No assembly loads Integer NOTCH_SENSITIVITY This argument specifies the file containing notch sensitivities to be included in the safety calculation. Default=No notch sensitivities Text

610

COMMAND FILES Post-Processor Commands SAFETY_CALCULATION This argument specifies the fatigue safety factor calculations to be performed. Multiple safety factors can be specified using this argument. The options are: Goodman Alternative Goodman Multiaxial Fatigue Principal Goodman P1 Standard Goodman Dang Van Linear SWT P1 Alternative Goodman Default=No safety calculation Text performed OVERLOAD This argument specifies the overload calculation type for Goodman based safety factor calculations. The options are: CYCLIC GENERAL Default=CYCLIC Text STARTING_LOADCASE This argument is used to define the load case number at which the loads are to start. Default=see Note 4 Integer UNIT_MOMENTS This argument specifies whether the unit moment bearing loads are included in the analysis. The options are: EXCLUDED INCLUDED If INCLUDED is specified, unit moment bearing loads are applied to each main bearing. If EXCLUDED is specified, it will only be possible to post-process staticallydeterminate solution load cases. It is recommended that this be selected so that all load cases can subsequently be post-processed. Default=INCLUDED Text

EXPORT This argument is used to specify whether the analysis results are written to an exported .SFE file. The options are: None Sets Only External See Note 5. Default=Sets Only Text

611

COMMAND FILES Post-Processor Commands FREQUENCY_RANGE This argument specifies the range of frequencies at which to perform the chosen analyses. Two values are expected; the first the minimum frequency and the second the maximum frequency. Non-defaultable Real ORDERS This argument specifies the list of engine orders at which to perform the analyses. Non-defaultable Real LOADCASE_OPTION This argument states whether the orders specified by the ORDERS argument are with respect to the order or speed of the engine. Non-defaultable Text DAMPING This argument specifies the file containing model damping values. See Note 6. Default=See Note 7 Text DAMPING_EQUATION This argument specifies the damping equation to be used for the block model damping. See Note 6. Default= See Note 7 Text DAMPING_EQUATION_UNITS This argument specifies the list of ENGDYN loadcases to use in the crankshaft analysis. This argument is only needed if the DAMPING_EQUATION argument is used. Non-defaultable Text VIBRATION_SETS This argument specifies the sets for which nodal vibrations are calculated. Default=All sets Text NOISE_SETS This argument specifies the list of face sets defining each radiating surface. Non-defaultable Text NOISE_METHOD This argument specifies the method used to calculate the radiating noise. The options are: RAYLEIGH HELMHOLTZ Default=RAYLEIGH Text

612

COMMAND FILES Post-Processor Commands INTEGRATION_ORDER_INCREMENT This argument is used to specify an integer number defining by how much the integration order is to be increased or decreased and applies to the Helmholtz method only. Default=0 Integer DELETE_VIBRATIONS This argument is used to specify whether the nodal vibrations are calculated prior to the acoustic analysis for each load case and are then deleted from the .SFE file once the analysis has been completed. The options are: YES NO Default=YES Text DENSITY This argument is used to specify the density of the fluid (normally air) in which the noise is being calculated. Default=1.29 kg/m3 Real kg/m3 SPEED_OF_SOUND This argument is used to specify the speed of sound in the fluid (normally air) in which the noise is being calculated .Default=344 m/s Real m/s

Notes: 1

If OUTPUT_NAME name is not specified then the name of the .FRC file is taken from the FILE argument.

2

The arguments ROTATE, TRANSLATE and MIRROR are used to define a transformation matrix that transforms the cylinder block model to the crankshaft coordinate system. It is assumed that the cylinder block and crankshaft models are orthogonal to one another. The transformation matrix is built up in the following order : Mirror Translate Rotate about x, then y and finally z crankshaft global axes

3

The arguments CRANK_ANGLES and INTERVAL are mutually exclusive and an error is reported if they are both included in the .EDPO file.

4

The starting loadcase defaults to n 1 where n is the maximum load case number stored in the .SFE file already.

613

COMMAND FILES Post-Processor Commands 5

If None is selected no export is performed and the results are written to the .SFE file selected using the Cylinder Block Model Panel. In this case it is recommended that the original file be copied before performing the analysis so that this file can be used again. If Sets is selected the vibration and radiating sets for which results are required together with the loaded nodes will be exported to a file called basename_SUBSET.SFE, where basename is the name of the .SFE file selected using the Cylinder Block Model Panel.. If External is selected the external surfaces of the model together with the loaded nodes are exported to a file a called basename_EXTERNAL.SFE. The exported file will contain the modal frequencies and mode shapes as well as the nodal vibration and radiated noise results, and will be significantly smaller than the original file. For this reason it is strongly recommended that either of the export options is used.

6

The arguments DAMPING and DAMPING_EQUATION are mutually exclusive and an error is reported if they are both included in the .EDPO file.

7

If no damping information is supplied the modal damping data used will be dependent on the type of block modelled in the ENGDYN solution. For a dynamic block, the modal damping characteristic used during solution will be used for each selected load case. For other types of blocks the default damping equation will be used:

  2.443  10 2 e 0.0007f . c

The equation is based on a curve fit to experimental data.

614


615

THEORY Introduction

D. THEORY 1 Introduction This section provides the user with an understanding of the relevant academic theory which is implemented within ENGDYN in order to simulate the forces applied to a rotating crankshaft.

616

THEORY Static Solutions

2 Static Solutions There are two types of static solutions; determinate and indeterminate. 2.1 Determinate Solution For a determinate solution, the main bearing reactions and main and big end journal bearing orbits are calculated assuming a statically-determinate solution. 2.2 Indeterminate Solution For an indeterminate solution, the displacements, main bearing reactions and journal orbits are calculated assuming a statically-indeterminate solution allowing for the compliance of the crankshaft and the cylinder block (if included).

617

THEORY Dynamic Solutions

3 Dynamic Solutions For a dynamic solution, the main bearing reactions, journal orbits and vibratory displacements and velocities of the crankshaft and cylinder block (if included) are calculated for a fully three-dimensional dynamic model.

618

THEORY Bearing Model

4 Bearing Model The bearing model in ENGDYN is able to perform three solutions of increasing fidelity: 1) Mobility based solution. 2) Hydrodynamic (HD) analyses (rigid bearing model). 3) Elasto-hydrodynamic (EHD) analyses (bearing deformation included). Mobility based solution (MB) assumes a rigid, circular bearing where the presence of holes and grooves is included for purposes of the the oil flow rate calculation only. This method is described in the Section 4.1. The Hydrodynamic model (HD), which is based on the Reynolds equation solution on a computational mesh (Section 4.2), fully accounts for the presence of holes and grooves. The HD model assumes that the bearing is rigid. An example journal bearing configuration capable of being solved by the HD model is shown in Figure D-1. There is an oil hole on journal and the bearing is partially grooved. NOTE: At present the ENGDYN GUI enables the definition of holes or grooves on the journal or bearing only. The combination of both on a journal or bearing is not possible.

Figure D-1 Example of the main bearing geometry A schematic representation of the oil flows in the journal bearing is shown in Figure D-2.

619


Figure D-2 Schema of oil flows in the bearing The Elasto-Hydrodynamic model (EHD) is an extension of HD model where the bearing deformation is included in the prediction of the oil flow. The deformation is usually very significant and therefore should generally be included when predicting wear loads. Figure D-3 shows and example of the radial deformation of the bearing where the deformation is exaggerated for illustration purposes.

620


Figure D-3 Example of Main Bearing Deformation (High Scale Factor) All the bearing models can be solved with or without thermal balancing. By the inclusion of thermal balancing the correct temperature and therefore oil viscosity values are applied in the bearing model. This is important for the accurate prediction of bearing friction loss, power loss and wear rate. Thermal balance models are described at Section 4.6. 4.1 Mobility Method 4.1.1

Theory

The mobility method (References 1 and 2) is a technique used extensively in the automotive industry since its inception approximately 35 years ago. It provides a solution extremely rapidly, typically taking only seconds or a few minutes of CPU time for a typical model, since it does not involve solving the Reynolds Equation. The limitation is that it may be used only in cases of ideal circular bearings without oil feed holes or grooves. However, it does provide a quick initial estimate of operating clearances for any design project related to journal bearings. In the mobility method, the effect of rotation is converted to an equivalent squeeze motion under steady load. The mobility vector (M), which is the dimensionless squeeze velocity with which a non-rotating shaft moves to support the load, is then calculated. The squeeze velocity (é) for a rotating shaft can then be expressed as

621


| F | [C r /R] é= LD/C r

2

M + e

(1)

where: L R,D Cr |F| μ e ω

bearing length, bearing radius and diameter, radial clearance, bearing load, lubricant viscosity, eccentricity, and bearing rotational speed.

The components of the mobility vector (M) may be implemented in the dynamic analysis in two ways: (a) components are calculated instantaneously based on “mobility maps” generated through a pre-processor or (b) components are calculated instantaneously using analytical curve fits (Reference 3), as is done in this code. In general, the mobility vector may be written as: M = f(ε,γ)

(2)

where: ε γ

Eccentricity ratio (e/Cr), and Bearing length to diameter ratio (L/D)

The analytical expressions for the curve fits of mobility components using a finite bearing analysis are:





f( ,  )  2 4  2  3  0.24 e  Mx = 1  0.4 1     2       2  1   3 

(3)

 5   2  3   2     1     0.3 1   8 4 7 2 3 -    My = f( ,  )  2 1  0.034  1    2    151     0.0161   2  1.03        

(4)

and

where:

f  ,   

1   2 1  2

2

 

2

  21    

(5)

and α

Eccentricity ratio in the x direction, and 622

THEORY Bearing Model β

Eccentricity ratio in the y direction.

Using these mobility components the orbit (or locus of instantaneous locations of the shaft center) is obtained by marching the solution forward in time using Equation (1). This provides the user with the minimum film thickness value over a typical load cycle. The usefulness of the mobility method has been further increased by extending the analytical curve fits (Reference 3) for maximum oil film pressure value, location and pressure curve extents. With this information, the oil film pressure profile and hydrodynamic power loss can be estimated. Within ENGDYN, the curve fits for the mobility components have been further refined using the finite difference solution, this is described in the proceeding section. Additionally, a loose coupling is also provided with the boundary lubrication model such that when the bearing clearances are extremely low, contact pressures will be predicted. However, the user should note that the contact pressures are solely based on the bearing clearance criteria and not on an explicit force balance and hence only trends in contact forces should be reported rather than the quantitative values. As a result, in situations where a rapid solution is required, using basic bearing geometry parameters (bearing length, diameter and radial clearance) and bearing loads, the user can obtain an initial estimate of minimum film thickness, peak oil film pressure, friction torque and journal orbit for a load cycle. 4.1.2

Determination of Bearing Oil Leakage Flow Rate

4.1.2.1 Pressure Induced Flow (a) Plain Bearings with Small Circular Hole Feed A relationship for the pressure induced flow for a circular feed hole to a plain bearing has been developed using geometric flux plotting techniques (Reference 3) and is as follows:

h 3pf  Dh  Q p  0.675  0.4   L 

1.75

(6)

For a radial clearance Cr small compared with the bearing radius, the clearance h at the angle  relative to the minimum oil film thickness is given accurately by h = Cr (1 -  cos)

(7)

Therefore equation (6) becomes:

C3 (1 -  cos )3 p f  D h  Q p  0.675 r  0 . 4  L  

623

1.75

(8)


0 h

dh

hmax hh e



hmin

min 

F Figure D-4 Bearing and Oil Hole Geometry for a Fixed Bearing

624

THEORY Bearing Model The angle  for a bearing that is fixed as shown in Figure D-4 is given by  = ( + ) - h or

(9)

 = min - h



l

F  hmin

hh 

 h

e

CA

CA+ 

0

Figure D-5 Bearing and Oil Hole Geometry for Rotating Bearing

625

THEORY Bearing Model For a bearing that is rotating the angle  in Figure D-5 is given by  = ( +) - (h + (CA + ))

(10)

 = min - (h + (CA + ))

or where

 r sin  CA   l  

  sin -1 

(11)

For a cross-drilled circular oil feed the program ENGDYN considers each oilhole feed individually and sums the two flows. (b)

Partially Grooved Bearing

A general curve-fitted formula, for practical use has been developed from data derived from an accurate finite difference model (Reference 3). This considers any rectangular grooving at any position in the bearing and any journal eccentricity and angular position of the journal centre (allowing for varying film thickness). The general comprehensive curve-fitted equation considering any groove extent (up to 270°), any groove aspect ratio, groove position and journal position is as follows

 a D  C3rp f 1.25  0.25 L   L   Qp  . f . f  L  1 2   6(  1) 0.3333   6(1 - a )   a L    

(12)

where



f 1  1   cos b

  1   3

cos e



3

 sin 3 θ b 3   f 2  θ b  3ε sinθ b  ε 2 1.5 θ b  sin2θ b  + ε 3  sinθ b  4 3     sin 3 θ e 3   θ e  3ε sinθ e  ε 2 1.5 θ e  sin2θ e  + ε 3  sinθ e  4 3   

   

   

and b, e are the angles defining the beginning and end of the groove relative to the maximum film thickness position (Figure D-6). Since the groove moves with the bearing, this equation is true for both fixed and rotating bearings where 1 and 2 are given by (13) 1 = 2 - (min - ) 626

THEORY Bearing Model 2 = 2 - (min - )

(14)

0 e

b

hmax hb e 

 min

he

hmin F Figure D-6 Bearing and Groove Geometry (c)

Fully Grooved Bearing

This case is dealt with in a number of references including Reference 3. The pressure induced flow for a fully grooved bearing is given by:

Qp 

4.1.2.2

16  0.327  DC3rpf 1  1.5 2 . L





(15)

Velocity Induced Flow

The equation for the oil flow rate from a short bearing, due to shift translational velocities is detailed in Reference 6 and given by .  .   Q T  DLC r       cos  1   sin  1     

.

(16)

Where  represents rate of change of attitude angle and θ1 is the film reformation value.

627

THEORY Bearing Model 4.1.2.3 Summation Technique ENGDYN has a number of methods for adding the instantaneous oil flow rates induced by pressure and velocity before numerically integrating to determine the cyclic oil flow rate. The methods are (i)

Summation of Qp and QT

(ii)

Use Qp only

(iii)

Selective summation of Qp and QT

The selective summation technique takes the greater of the two instantaneous oil flow rates. This technique was introduced in Reference 6.

4.1.3

Determination of Bearing Oil Power Loss

Expressions for the instantaneous power loss due to shear and squeeze effects are detailed in References 7, 8 and 9, and results in the following scalar expressions:

D 3 L( j   b ) 2 oo PSH  J1 2C r   j  b  F C r sin + FV cos 2 PSQ   2  

x

b j

e

V

y

where

F

2

J 1oo  

1



 (1  cos )

2

V = e  e 2 2

2

  e  2    tan     e    1

Figure D-7 Notation for Basic Power Loss Relations and the angles 2 and  are defined in Figure D-7. The ‘shear’ (or ‘Couette’) constituent arises from differences in journal and bearing rotation (relative to any observer) whilst the ‘squeeze’ (or ‘Poiseville’) constituent arises 628

THEORY Bearing Model from journal translation (relative to an observer rotating at the mean ‘entrainment’ velocity of journal and bearing). The instantaneous total power dissipated in the lubricant film arises chiefly from these two components, with the shear power loss being the dominant quantity. The shear power loss (PSH) has two possible solutions depending on the limits of integration of the journal bearing intergral J. These are detailed in Reference 10 and given below: For a 2 film,

J 

oo 1  2 1 1



2

(17)

1   

1 2 2

For a  film,

If sin 1  0

J 

f sin 1  0

J 

oo 1  1 1

oo 1  1 1





cos 1 - A   cos 1 - B

1   

1 2 2

cos 1 A   cos 1 B

1   

1 2 2

(18)

where   + cos 1     cos 1   and B =   A =   1  cos 1   1  cos 1  ENGDYN allows the shear power loss to be calculated using a  or 2 film assumption; although for the short bearing theory only  is strictly valid. The overall cyclic power loss P is found by the numerical integration of the instantaneous power losses.

4.2 Solution of Reynolds Equation on a Computational Mesh The second group of bearing models are based on a direct numerical solution of the Reynolds equation. The general form of the Reynolds equation used for incompressible fluids is shown below:

629


  h 3 p    h 3 p  v h h      x  12 x  y  12 y  2 y t

(19)

where: μ x y h p v

Oil viscosity Axial coordinate Circumferential coordinate Clearance distribution Hydrodynamic pressure distribution Circumferential velocity

h represents here the clearance distribution between the journal and the bearing. It consists of three components:

h  ho  h   s where ho hε δs

(20)

Nominal clearance coming from journal and shaft input profiles, Clearance due to shaft eccentricity and tilt Bearing radial deformation

In all the direct solutions of Reynolds equation the Average Flow Model form of the equation is used. This is shown in Section 4.2.1; Equation (23). 4.2.1

Average Flow Model (by Patir & Cheng)

The local film thickness is assumed to be consisting of nominal film thickness and roughness amplitudes on the two surfaces. It is assumed that roughness heights have a Gaussian frequency density with a standard deviation of σ. The average film thickness h can then be defined as:

H  h , z  H 3

(21)

h  h 3 h  35  z  140  z 2   70  z 2  28  5 z 2 256











H 3 H 3

(22)

The directional properties of roughness are described by a surface parameter  . It is defined as the ratio of x and y correlation lengths (i.e. the length to width ratio of the asperities). The  values for typical contact areas can be seen in Figure D-8. Pressure flow and shear flow factors were defined such that the average Reynolds equation for a rough bearing is given by:

630


 h 3 p    h 3 p    x     y  x  12 x  y  12 y 

v   h  s   2  y y

 h   t 

(23)

(24)

where x is the axial coordinate and y is the circumferential coordinate. . Coefficients C, r, A1, A2, 1, 2 and 3 depend on  and are interpolated from tables given in Reference 13.

Figure D-8 Typical contact areas for transversely oriented, isotropic and longitudinally oriented surfaces Reynold’s equation (23) is discretized using the Finite Volume Method (FVM). FVM balances the oil flows over each finite volume of the computational mesh, so it is advantageous to express it in the form which balances the oil flows:

m x m y h   x y t

(25)

631


Variables m x and m y represent the mass fluxes in the axial, x, and circumferential, y, directions, respectively. The mass flux in each direction comprises of a Couette flow component and a Poisseuille flow component such that in the circumferential direction

m   m  y c

 my

y p

(26)

where



(27)

h 3  p   y   12  x 

(28)

m 

y c

m 

y p





v h  S 2

Axial x direction is analogous. At the axial boundaries of the domain the pressure is set to a reference (crankcase) pressure pcrc:

p(0, y)  p( L, y)  pcrc

(29)

where L is the bearing length. In circumferential boundaries the identity between 0 and 2π nodes is set:

p( x,0)  p( x,2 )

(30)

In the positions of oil holes or oil grooves the immediate oil pressure is set:

p( x, y)  phole

(31)

For bearing oil holes and grooves overlapping with the bearing axial boundary, the pressure is set to reference (crankcase) pressure Pcrc instead of oil supply pressure Phole,

P( x, y)  Pcrc

4.2.2

(32)

Half-Sommerfeld Boundary Conditions

The set of linear equations (25), (29), (30) and (31) is solved to predict the hydrodynamic pressure distribution. The PDE is of an elliptical type. After the discretization of the equation, the system of equations can be solved for on a rectangular mesh:

L

i 1, j

   h  s Pi 1, j  Li 1, j Pi 1, j  Li , j 1 Pi , j 1  Li , j 1 Pi , j 1  Li , j Pi , j      y  2  y

632

 h      t 

(33)

THEORY Bearing Model The solution may contain areas where the hydrodynamic pressure is below the cavitation pressure or is even negative. Therefore the cavitation pressure limit (Half-Sommerfeld boundary conditions) is applied as a post-processing step:

p( x, y)  pcavitation Then

If

p( x, y)  pcavitation

(34)

By using the Half-Sommerfeld boundary conditions the flooded and cavitated area can be estimated directly. The estimate is not precise, but it provides a fast and approximate solution. At this stage of the calculation, the oil flows do not satisfy the oil mass conservation requirement. As the requirement for oil mass conservation is crucial when calculating the thermal balance of the journal bearing the oil flows are re-evaluated during a post-processing step. The basic premise of the applied approach is to correct the obtained flows and oil accumulation such that the continuity equation with oil flow corrections  is valid:

  m x     m y  (  h)   x y t

(35)

The values of the oil flow corrections  and oil mass flows are strictly controlled by upwind stabilization. At the edges of any holes and grooves it is assumed that there is only oil (due to oil feed pressure). Therefore the value of the oil flow corrections is set to 1.0.

 hole  1

(36)

The set of linear equations represented in (35) and (36) is then solved to obtain the oil flows that satisfy the mass conservation. After the discretization the following system of equations are solved on rectangular mesh:

R

i 1, j

 i 1, j  Ri 1, j i 1, j  Ri , j 1 i , j 1  Ri , j 1 i , j 1  Ri , j i , j  

hi , j i , j t

(37)

Notes:

 The Half-Sommerfeld boundary conditions provide a fast, non-iterative solution of the Reynolds equation.  The estimate of hydrodynamic pressure distribution is not precise as a consequence the pressure gradient at the flooded/cavitated area may be very steep.  In the post-processing step the system of corrected flows is calculated where oil mass conservation is satisfied.

633

THEORY Bearing Model 4.2.3

JFO (Deprecated) Boundary Conditions

As stated previously, the Reynolds equation for incompressible fluids enforces both the momentum and mass balance but does not take cavitation into account In order to satisfy both the momentum and mass balances in the entire fluid, the solution is split into two parts, one representing the flooded and the other the cavitated region. In the flooded region the Reynolds equation is solved and the hydrodynamic pressures calculated, in the cavitated region the oil mass balance, which is derived from the Reynolds equation, is preserved and the mass fractions calculated. The fluxes in the oil mass balance under cavitation pressure conditions are equal to:

m 

y c





v h  S 2



(38)

where  is the mass fraction. On the edges of any holes and grooves present it is assumed that there is only oil (due to oil feed pressure). Therefore the values of mass fractions at these points are set to 1.0.

 hole  1

(39)

Therefore equations (25), (29), (30) and (31) are solved to calculate the hydrodynamic pressure distribution in the flooded area and equations (25) with (38) and (39) are solved in the cavitated area to calculate the mass fraction distribution. The solution reached by this method is fully mass conserving over the entire bearing area. The resulting system of linear equations is solved iteratively. The need for iterations comes from the determination of which nodes are flooded and which are cavitated:

L

i 1, j

R

i 1, j

Pi 1, j  Li 1, j Pi 1, j  Li , j 1 Pi , j 1  Li , j 1 Pi , j 1  Li , j Pi , j  

 i 1, j  Ri 1, j  i 1, j  Ri , j 1 i , j 1  Ri , j 1 i , j 1  Ri , j i , j  

If

node cavitated

If

node flooded

 hi , j  i , j    s  t 2 y

Pi , j  p cavitation and  i , j  solved ;  i , j  1.0

 i , j  1.0 and

Pi , j  solved ; Pi , j  p cavitation

Notes:  JFO (Deprecated) boundary conditions provide a slower, iterative solution of Reynolds equation. The system can suffer from convergence problem.  JFO (Deprecated) boundary conditions conserve momentum and mass in the flooded areas and conserve mass in cavitated areas.

634

(40)

THEORY Bearing Model 4.2.4

Half-Sommerfeld, Mass Conserving (Recommended)

This option is an extension to the Half-Sommerfeld Boundary Conditions described in Section 4.2.2. The maindifference between the algorithms is in the definition of the feed holes and groovesm that is the definition of internal boundaries. The approach is based on the linear combination of the Reynolds equation with the boundary condition equation.

  h 3 P    h 3 P    h  s   x     y    x  12 x  y  12 y  2  y y P  Phole

 h    t

(41)

After discretization on the rectangular mesh the following system is obtained:

Li 1, j Pi 1, j  Li 1, j Pi 1, j  Li , j 1 Pi , j 1  Li , j 1 Pi , j 1  Li , j Pi , j 

  h

 s   2  y y

 h    t

(42)

Pij  Phole This linear combination of equations is weighted using the definition of the hole coverage ratio . It is the ratio of the immediate hole area (Ahole) in the finite volume area (A).



Ahole A

(43)

BY applying the weighting, the numerical strength of the Reynolds terms is weakened close to the hole and groove boundaries and conversely, the strength of boundary condition is increased. To ensure the correct weighting between the Reynolds equation and Dirichlet boundary condition (feed pressure in the hole/groove) - all the terms in the equation must have the same dimension:

(1   )Li 1, j Pi 1, j  Li 1, j Pi 1, j  Li , j 1 Pi , j 1  Li , j 1 Pi , j 1  Li , j Pi , j   Li , j Pi , j    h  s  (1   )   y  2  y

 h      Li , j Phole  t 

(44)

The solution may contain areas where the hydrodynamic pressure is below the cavitation pressure or is even negative. Therefore, the cavitation pressure limit (Half-Sommerfeld boundary conditions) is applied in the post-processing step:

If

p( x, y)  pcavitation Then

p( x, y)  pcavitation

(45)

By using the Half-Sommerfeld boundary conditions the flooded and cavitated area can be estimated directly. The estimate is not precise, but it provides a fast and approximate solution. At this stage of the calculation, the oil flows do not satisfy the oil mass conservation requirement.

635

THEORY Bearing Model As the requirement for oil mass conservation is crucial when calculating the thermal balance of the journal bearing the oil flows are re-evaluated during a post-processing step.

  m x     m y   (  h)   x y t    hole

(46)

The values of the oil flow corrections  and oil mass flows are strictly controlled by upwind stabilization. After the discretization the following system of equations are solved on rectangular mesh:

Ri 1, j  i 1, j  Ri 1, j  i 1, j  Ri , j 1 i , j 1  Ri , j 1 i , j 1  Ri , j  i , j  i , j 

hi , j  i , j t

(47)

 i , j   hole To ensure the correct weighting between the Reynolds equation and Dirichlet boundary condition (feed pressure in the hole/groove) - all the terms in the equation must have the same dimension:

1   Ri 1, j i 1, j  Ri 1, j i 1, j  Ri , j 1 i , j 1  Ri , j 1 i , j 1  Ri , j i , j   Ri , j i , j  1   

hi , j  i , j t

(48)

 Ri , j  hole

At the edges of the holes and grooves, it is assumed that there is only oil (due to oil feed pressure). Therefore the value of the oil flow corrections is set to 1.0.

 hole  1

(49)

Notes:  The Half-Sommerfeld boundary conditions provide fast, non-iterative solution of Reynolds equation.  The estimate of the hydrodynamic pressure distribution is not precise. The pressure gradient at the flooded/cavitated areas may be very steep.

 During the post-processing step, the system of corrected flows is calculated where the oil mass conservation is always satisfied regardless of the oil hole or groove motion. 4.2.5

Reynolds, Mass Conserving

636

THEORY Bearing Model The discretized system (44) derived in Section 4.2.4 is solved by the Successive Over Relaxation (SOR) method. At each iteration of the SOR method, the hydrodynamic pressure of the central node is updated. The update is based on the hydrodynamic pressure of the surrounding nodes whose pressure is limited by the cavitation pressure. Thus, the cavitated area is iteratively searched.

(1   )Li 1, j Pi 1, j  Li 1, j Pi 1, j  Li , j 1 Pi , j 1  Li , j 1 Pi , j 1  Li , j Pi , j   Li , j Pi , j    h  s  h      Li , j Phole  (1   )   2  y  y   t   If node cavitated Pi , j  pcavitation If

node flooded

(50)

Pi , j  solved ; Pi , j  pcavitation

The difference when compared to the solution with Half-Sommerfeld boundary condition is in the application of the cavitation pressure limit. In the solution with Half-Sommerfeld boundary conditions the limit is applied in a post-processing step but in the solution with Reynolds boundary conditions the limit is applied during the iteration process. The estimate of the flooded area when using the Reynolds boundary condition is therefore more realistic than when using the HalfSommerfeld boundary conditions. The pressure gradient at the flooded/cavitated boundary is more realistic than and not as steep as the result obtained using the Half-Sommerfeld boundary conditions. The oil flows do not satisfy the oil mass conservation requirement. As the requirement for oil mass conservation is crucial when calculating the thermal balance of the journal bearing the oil flows are re-evaluated during a post-processing step. The linear system (48) is solved to correct the oil flows to satisfy oil mass conservation. Notes:  Reynolds, Mass Conserving boundary conditions provide a relatively fast, iterative solution to the Reynolds equation.  The estimate of hydrodynamic pressure distribution is generally better compared to the Half-Sommerfeld boundary conditions.  During the post-processing step, the system of corrected flows is calculated where the oil mass conservation is always satisfied regardless of the oil hole or groove motion.  The ability to model holes and grooves that follow the shaft motion will be a future development.

4.2.6

Reynolds and JFO

In the flooded region, the Reynolds equation is solved and the hydrodynamic pressures calculated, in the cavitating region the oil mass balance is preserved (which is derived from the Reynolds equation) and the mass fractions calculated. For this boundary condition, the Reynolds equation for the flooded area is combined the with the continuity equation for the cavitated area.

637

THEORY Bearing Model The approach here is analogous to the method described in Section 4.2.4. Allthough, the Reynolds and JFO boundary conditions are based on the linear combination of the Reynolds and continuity equation with the boundary condition equations. In the flooded area Reynolds equation is solved whilst in the cavitated area the continuity equation is solved. There are two boundary conditions: the hydrodynamic pressure in the flooded area edge and the mass fraction in the cavitation area edge.

  h 3 P    h 3 P    h  s   x     y    x  12 x  y  12 y  2  y y P  Phole

 h    t

(51)

   hole These boundary conditions are applied to the continuity equation. Here the value of the mass fraction  is the ratio between the immediate oil level and the immediate clearance:

h h  y t P  Phole

U

(52)

   hole The continuity equation is updated to be compatible with the Reynolds equation, such that:

  h  s    2  y y P  Phole

 h   t 

(53)

   hole After discretization on a rectangular mesh of equations (51) and (53) the system of equations to be solved are:

Li 1, j Pi 1, j  Li 1, j Pi 1, j  Li , j 1 Pi , j 1  Li , j 1 Pi , j 1  Li , j Pi , j  Pi , j  Phole

  h

 s   2  y y

 h    t

(54)

 i , j   hole Ri 1, j  i 1, j  Ri 1, j  i 1, j  Ri , j 1 i , j 1  Ri , j 1 i , j 1  Ri , j  i , j 

hi , j  i , j t



  s

(55)

2 y

638

THEORY Bearing Model Equations (54) and (55) are linearly combined to form a combined equation which must be solved iteratively to determine the flooded and cavitated area. Note the weighting between the Continuity/Reynolds equation and Dirichlet boundary conditions (mass fraction/hydrodynamic pressure in the hole/groove) where all the terms in the equation must have the same dimension:

(1   )Li 1, j Pi 1, j  Li 1, j Pi 1, j  Li , j 1 Pi , j 1  Li , j 1 Pi , j 1  Li , j Pi , j  

1   Ri 1, j i 1, j  Ri 1, j i 1, j  Ri , j 1 i , j 1  Ri , j 1 i , j 1  Ri , j i , j   Li , j Pi , j  Ri , j  i , j  hi , j  i , j   s  1     2 y  t If

node cavitated

If

node flooded

   Li , j Phole  Ri , j  hole 

(56)

Pi , j  p cavitation and  i , j  solved ;  i , j  1.0

 i , j  1.0 and

Pi , j  solved ; Pi , j  p cavitation

At the edges of holes and grooves, it is assumed that there is only oil (due to the oil feed pressure). Therefore the value of the mass fraction is set to 1.0.

 hole  1

(57)

Notes:  The Reynolds and JFO boundary condition provide a slower, iterative solution of the Reynolds equation when compared to the previouslt described boundary conditions.  The Reynolds and JFO boundary condition ensures the conservation of momentum and mass in flooded areas and the conservation of mass in cavitated areas.

4.2.7

Determination of the Bearing Hydrodynamic Friction Force and Power Loss

Using the instantaneous oil film pressure distributions the bearing hydrodynamic friction force can be calculated as:



h  p  vdA    dA h 2  y  A A

F f hyd  

(58)

If the Average Flow Model is used then Equation (58) is extended to:

F f hyd  

A

 h

v  f   fs dA    fp A

h  p   dA 2  y 

639

(5 9)










f 

3 35  1  z  1  z 2 ln(300( z  1))   55  z 132  z 345  z  160  z  405  z 60  147 z  H 32  60 

f 

35  z 1  z 2 32 

3

ln





z 1 z   66  z 2 30 z 2  80  z  1 15 



H 3

 fs  A3 H  4 e  5  6 H  fp  1  De  sH H h ;



zH

3

Constants in equations for the Φfs and Φfp are tabulated in Reference 13. And similarly, the hydrodynamic power loss is given by:

Phyd



2

2

h 3  p  h 3  p    dA   v dA     dA   h 12  x  12  y  A A A 2

640

(60)


4.3 Oil Viscosity 4.3.1

Temperature Dependence

The Walther-ASTM equation is used to calculate oil viscosity dependence on temperature:

loglog     A  B logT 

(61)

where A and B are constants, specific for each oil. Reference 12 demonstrated that where the viscosity is in centistokes,  is 0.6. ENGDYN uses this equation but allows the user to specify,  to be 0.7 or 0.6 as a value of 0.7 is referred to in Reference 11. ENGDYN also allows for a viscosity-temperature curve to be input for those oils that do not satisfy the Walther-ASTM equation. The value of viscosity at a given temperature T is then obtained by linear interpolation. 4.3.2

Shear Thinning

For multi-grade oils, viscosity becomes thinner as the shear rate increases. This shear thinning effect may be expressed by the Cross Equation as

       Where

0  

(62)

1    c m

 is shear rate in ( s 1 ),  0 and   are the low-shear-rate and the fully shear-

 c ( s 1 ) is the shear rate at which the lubricant viscosity is halfway between  0 and,   and m is a dimensionless parameter. thinned viscosities (Pa-s),

The ratio

 0  c is a constant. The halfway shear rate  c is normally given by

 c  10 a bT 273

(63)

Where a and b (K-1) are constants and T is the oil film temperature in (K). For a multi-grade oil, in addition to the low-shear-rate viscosity and temperature relation, the shear thinning constants a , b , c and m are normally provided in the Cross Equation. The user specified input option can be used to provide low-shear-rate viscosity

 0 and

temperature relation and the values for the constants a , b , c , m . In ENGDYN, oil films in the bearings may experience very high shear rate, where the shear thinning effect of the lubricant can be significant for multi-grade lubricants. The

641

THEORY Bearing Model shear rate of the oil film is assumed to be constant across the film thickness and is normally approximated by:



.R

(64)

h

 is rotational velocity of the journal, R journal radius and h is the oil film thickness. According to the definition, the shear rate of the oil varies along the circumference of the bearing. As a result, oil viscosity also varies along the circumference of the bearing due to the shear thinning effect. Using Equation (62), the viscosity  can be calculated. The calculated viscosity is then fed into the hydrodynamic lubrication solver for the bearings. 4.3.3

Piezo-Viscosity Effect

Many lubricants undergo a considerable increase in viscosity when subject to high pressures. These lubricants are called piezo-viscous. Piezo-viscous lubricant behaviour is modelled by either the Barus equation or the Roelands’s equation both of which are exponential equations. They are described in the proceeding sections. 4.3.3.1

Barus Equation

   0 ep

(65)

This relation is known as the Barus equation.  0 is the base viscosity, that is, the viscosity at low pressure, and the constant  is called the piezo-viscosity coefficient or pressure viscosity index. 10 GPa 4.3.3.2

1

 is measured in Pa 1 and has typical values in the order of

.

Roeland’s Equation 

  0  e

ln(0 )9.67 1 1 p p 



o

  

Z

   

(66)

This relation is known as the Roeland’s equation.  0 is the base viscosity, that is, the viscosity at low pressure, constant z is called the viscosity-pressure index (the default value used in ENGDYN is 0.67) and constant p 0 is 1.96x108 Pa.

642


4.4 Boundary Lubrication Model During the operation of a journal bearing, the minimum film thickness may decrease to such low values that there is contact between the two operating surfaces. In these situations, the load is supported not by just the oil film but also by the asperities. As a result, a boundary lubrication model is coupled with the hydrodynamic lubrication model. This model takes into account surface roughness features such as asperity heights (), radius of curvature of asperities (β), asperity density () standard deviation of summit height (s), and mean summit height (Zs) and composite elastic modulus (E) for the contacting materials. If the option ‘RMS Surface Height (Deprecated)’ is selected then the local asperity contact pressure is given by

pasp 

16 2   hi , j     2 E f 15    

(67)

If the option ‘Summit Height Correction (Recommended)’ is selected then the local asperity contact pressure is given by

pasp 

 s  hi , j  Z s 16 2   s  2 E f 15    s

  

(68)

and 

f x    s  x  e 5

2

s 2

2

ds

(69)

0

ENGDYN offers two methods for deriving the composite asperity data s and Zs used in the equation for the asperity contact pressure (67) or (68). The first method is referred to as ‘v4.0 and Earlier (Deprecated)’ and is implemented for backwards compatibility. It defines the composite data as

   12   2 2   0.5(1   2 )   0.5( 1   2 )

(70)

 s   s1 2   s 2 2

Z s  0.5Z s1  Z s 2 

643

THEORY Bearing Model whereas the second recommended method called ‘Hertzian (Recommended)’ uses

   12   2 2   1 2  1 1      21 2 2

  

1

(71)

 s   s1 2   s 2 2 Z s  Z s1  Z s 2 The asperity data for a single material can be calculated from surface profile measurements using the MATUTIL program supplied with the Ricardo Software installation which is described in Appendix 12. By obtaining the contact pressures and the surface velocities, wear load profile on the bearing shell can be generated using the definition T

Wear load 

1 pasp v dt T 0

(72)

where (v) is the velocity of the journal and (T) is the period of the engine cycle. This wear load calculation can be further extended to extract a wear rate value based on the following definition

Wear rate 

Wear coefficien t   Wear load  dA A

Hardness number  9.81  10

(73)

6

The wear coefficient value, which is supplied by the user, is not easily obtained and is a function of several variables including lubrication conditions, lubricity, and other factors. Users can use their own databases or can use representative values from typical Data Handbooks. However, users need not be concerned about the validity of their input wear coefficient value, unless that is the focus of their study as the input wear coefficient values do not impact the simulation. They are used solely for calculating the wear rate as a post-processing operation. Conversely, ENGDYN could conceivably be used to identify the wear coefficient value itself by the user performing a parametric study of this parameter until the predicted wear rate results match the experimental test results.

4.4.1

Determination of Bearing Boundary Power Loss

Using the instantaneous contact pressure distributions the bearing boundary friction force can be calculated as:

F fbdy   c f paspdA

(74)

A

644

THEORY Bearing Model Where c f is the friction coefficient for the bearing-shaft interface. The boundary power loss is given by:

Pbdy   c f pasp v dA

(75)

A

4.5 Integration of Bearing Forces and Moments By integrating the generated hydrodynamic and asperity pressures and shear stresses the force components (76)(76)(76) and friction force components (78), moment components (77)(77) and friction moment components (79) can be obtained.

Fhyd , x   p cos  dA Fasp, x   pasp cos  dA A

Fhyd , y   p sin  dA A

A

Fasp, y   pasp sin  dA

(76)

A

M hyd , x   x  p cos  dA Fasp, x   x  pasp cos  dA A

A

A

A

M hyd , y   y  p sin  dA Fasp, y   y  pasp sin  dA

(77)

 h  p   F fhyd , x        sin  dA F fasp, x   c f pasp sin  dA 2  x     A A  h  p   F fhyd , y     v     cos  dA F fasp, y    c f pasp cos  dA h 2  y   A  A

(78)

 h  p   M fhyd , x   x       sin  dA M fasp, x   x  c f pasp sin  dA 2  x     A A  h  p   M fhyd , y    x   v     cos  dA M fasp, y    x  c f pasp cos  dA 2  y   A A h

(79)

4.6 Bearing Thermal Balance Solution 4.6.1

Models used with the Mobility Method

Both thermal models in this section calculate the mean oil temperature in the bearing. The heat transport into any metal parts is neglected. The model described in Section 2.1.1.1 updates the oil temperature at the end of each engine cycle. The model described in 2.1.1.2 updates the oil temperature in each evaluation of the Mobility method.

645

THEORY Bearing Model 2.1.1.1 Cycle Averaged Equilibrium Temperature Using the mobility method it is possible to establish the instantaneous oil flow rate and power loss from the bearing, and to numerically integrate to estimate the overall cyclic oil flow rate and power loss. The temperature rise of the oil in the bearing may then be estimated. At the end of simulation performed with a mean oil temperature. The thermal balance is calculated. It is a simple model based on equation:

T  Toil _ feed  P /(   c p  Q)

(80)

Where P is the hydrodynamic power loss, ρ, the oil density, cp, the specific heat capacity and Q is mean oil flow rate. At each iterative step (engine cycle) the dynamic viscosity  is evaluated corresponding to the temperature T. 2.1.1.2 T0D Oil Model – Averaged Temperature over Whole Bearing In this model, the oil mass balance is solved at each time step. This model is able to capture the oil temperature changes within the engine cycle. The heat balance over the whole bearing is used:

c pToil _ feed  Q  P  c pTQ  c pV

dT dt

(81)

Where P is the hydrodynamic power loss, ρ the oil density, cp the specific heat capacity, V is the oil volume present in the bearing and Q is mean oil flow rate. 4.6.2

Models used with the Hydrodynamic and Elastohydrodyanmic Solutions

The solution of Reynolds equation on the lubrication mesh enables the selection of the area for over which to balance the heat transport. The simplest approach is the T0D model described in Section 4.6.2.1 where the area is the whole bearing. The T1D model, described in Section 4.6.2.2, balances heat over each radial slice and therefore the axial distribution of oil temperatures is calculated. This model can predict the oil temperature increase as oil travels from the bearing centre to the bearing axial edges. Bpthe the T1D and T0D models, described in Sections 4.6.2.1 and 4.6.2.2 respectively, neglect the heat transport into the metal parts of the bearing. For a more detailed thermal analysis, that is the determination of the current temperature per node, the assumption of neglecting the heat transport into the metal layers around the oil layer cannot be applied. The metal layers work here as a damping mechanism to the of oil temperature increase. Therefore, the T2D model, described in Section 4.6.2.3, was created which includes the heat transfer from the oil to the metal parts of the bearing. 4.6.2.1 T0D Oil Model - Averaged Temperature over Whole Bearing The average temperature in the oil over the whole bearing is calculated and the heat transport into the metal is neglected. 646


Figure D-9 Schema of 0D Model The heat balance over the whole bearing is used:

 c T

p oil _ feed

OilFeed

 Qoil _ feed  P 

 c T  Q p

Outflow

Outflow

 c pV

dT dt

(82)

Where P is the hydrodynamic power loss, ρ the oil density, cp the specific heat capacity, V is the oil volume present in the bearing and Q is the particular oil flow rate. Equation (82) is integrated together with the journal bearing equations. 4.6.2.2 T1D Oil Model - Temperatures Averaged Circumferentially The averaged axial temperature distribution is obtained with this model by allowing no variation in temperature circumferentially. The heat transport into the metal is neglected.

Figure D-10 Schema of 1D Model

647

THEORY Bearing Model The set of heat balances for each radial slice is used:

c pTleft  Qleft  D  

T y

 c pTright  Qright  D  

 Pslice  c pToil _ feed  Qslice _ oilfeed left

T y

(83)  c pVslice right

dT dt

where P is the hydrodynamic power loss per radial slice, ρ the oil density, cp the specific heat capacity, V is the oil volume present in particular slice and Q is the particular oil flow rate. The equation above is stabilised using the upwind method. Equation (83) is integrated together with the journal bearing equations. 4.6.2.3 T2D Oil Model + 1D Metal Model – Oil / Metal Heat Exchange – Flash Temperatures For this model, the heat transport is calculated for each node of the lubrication mesh and the surrounding metal. The heat transport in the oil and metal is solved separately due to computational memory limitations. In the metal, only radial heat transport is assumed; each node of the lubrication mesh is therefore solved on a separate 1D radial mesh. It is not heat conserving but is able to predict the flash temperature effect.

648


Figure D-11 Schema of 2D Lubrication Mesh

Figure D-12 Extension of Finite Volume into Radial Dimension (1D Mesh for Metal Layer) Figure D-12 depicts the heat flow balance over one finite volume of the lubrication mesh. The heat balance for each finite volume of oil is used is calculated

c pTleft Qleft  c pTrightQright  h  y  

T x

c pTbottomQbottom  c pTtopQtopt  h  x  

 h  y   left

T y

T x

 right

 h  x   bottomt

T y

 top

(84)

P  c pToil _ feed  Q finite _ volume_ oilfeed  c pV

Tcurrent T T  2  x  y  K Metal current t h/2

The heat transfer coefficient (K) is:





metal

K  h / 2 z / 2    metal h / 2 z / 2

(85)

649

THEORY Bearing Model The heat balance for each finite volume of the metal layer (except the radial node neighbouring oil film) is calculated as:

xy   metal

Tmetal z

 xy  metal bellow

Tmetal z

  metal c p _ metal xyz above

Tmetal t

(86)

For the radial node neighboring the oil film the heat transport equation is:

x  y  K

Tmetal  T T  xy  metal metal h/2 z

  metal c p _ metal xyz above

Tmetal t

(87)

The equations above are stabilised by the upwind method. There are two systems of equations solved: 1) Metal layer solution. For each node of the lubrication mesh a 1D system of equations (86), (87) and (85) arises. The unknowns are the metal temperature distribution. 2) Oil layer solution. For all the nodes of the lubrication mesh, the system of heat transport equations (84) is solved. Metal temperatures of the layer next to the oil are provided by the metal layer solution. The unknowns are the oil temperatures. Generally the new oil temperatures are quantitatively similar to the oil temperatures obtained from the first system. Iterations are therefore not used here to equal the oil temperatures from both systems due to computation efficiency considerations. The difference in the obtained oil temperatures between the systems results in heat not being conserved. 4.7 Compliant Model for EHD Bearing It is possible to solve the compliant bearing model using the determinate solution in ENGDYN. 4.7.1

Bearing Shell Deformation

The elasticity relation for a compliant shell is given:

FS  K S δ SFE

(88)

where K S is the shell stiffness matrix with respect to the reduced Finite Element mesh that is the lubrication mesh, FS is the shell forcing vector which is composed of hydrodynamic and boundary contact pressures. The bearing shell stiffness matrix or the stiffness matrix of the entire system of shells (the main bearings of crankcase) used in the EHL calculation is derived from FE models using the FEARCE finite element analysis software. The stiffness matrix of the component is reduced to the lubricated surface which decreases the size of the matrix. The inertia (mass) of the structure is used to resist the applied loadings, that is, an

650

THEORY Bearing Model assumption is made that the structure is in a state of static equilibrium even though it is not constrained. A reference point P is chosen in space at which the displacement is zero. The motion of all other points will be relative to that point. A simple way to think of this is that the solution represents the deformation of the structure one would see if one was standing at that point. For the shell model, six unit acceleration loads are applied. Three linear accelerations are applied together with three rotational accelerations about the chosen reference point P. For each case a load vector is computed during reduction at the reduced degrees of freedom. During the solution, the following set of equations is solved to ensure a state of equilibrium whereby the average weighted motion of the system is zero about the reference point P.

F   K  0   F    r

Fr     0   s

(89)

where F : vector of resulting inertial loads due to the applied forces, FA Fr : load vectors due to the unit accelerations

K : reduced stiffness matrix δ : required displacement vector s : vector of unknown forces to produce equilibrium and should be computational

zeros The vector of applied forces, FA is composed of forces due to hydrodynamic and asperity contact pressures. Load vector is calculated during the reduction process using FEARCE.

4.7.2

Force, Moment and Pressure Equations

The generated forces are put against the bearing load and thus the force balances arise.

Fapp, x  Fhyd , x  Fasp, x  F fhyd , x  F fasp, x

(90)

Fapp, y  Fhyd , y  Fasp, y  F fhyd , y  F fasp, y

The generated moments are put against the bearing load and thus the moment balances arise.

M app, x  M hyd , x  M asp, x  M fhyd , x  M fasp, x M app, y  M hyd , y  M asp, y  M fhyd , y  M fasp, y

(91)

Next in case of EHD bearing analyses the equilibrium between generated pressure and applied pressure used for shell deformation must be reached in each node of lubrication mesh.

651


papp  p  pasp

(92)

The system of non-linear differential-algebraic equations (90), (91) and (92) is solved using Newton method.

652

THEORY Bearing Model 4.8 Nomenclature Notation

Description

Unit

a

Axial length of bearing groove

m

cf

Friction coefficient

cp

Specific heat capacity of oil

J/kg/K

Cr

Bearing radial clearance

m

Dh

Oil hole diameter

m

D

Bearing diameter

m

e

Eccentricity

m

F

Force

N

h

Film thickness

m

h

Average oil film thickness

m

hb

Film thickness at θb

m

he

Film thickness at θe

m

hmin

Minimum film thickness

m

l

Connecting rod length

m

K

Heat transfer coefficient

J/m2/K

L

Bearing length

m

mx

Mass oil flow rate, axial direction

kg/s

M

Moment

Nm

p

Hydrodynamic pressure

Pa

pasp

Asperity pressure

Pa

pf

Oil feed pressure

Pa

P

Overall cyclic power loss

W

PSH

Shear power loss constituent

W

PSQ

Squeeze power loss constituent

W

Q

Volumetric oil flow rate

m3/s

Qp

Total instantaneous feed pressure flow

m3/s

QT

Total instantaneous translational velocity flow

m3/s

r

Crank throw

m

R

Radius

m

T Toil_feed

Temperature or Period of engine cycle Inlet temperature of oil

K s K

v

Circumferential velocity

rad/s

V

Oil volume

m3

x

Axial coordinate

m

653

THEORY Bearing Model y

Circumferential coordinate

m

z

Radial coordinate or Pressure-viscosity index Mean summit height

m m

Angle of oil hole relative to minimum film thickness position or Piezo-viscosity coefficient Load angle relative to bearing datum or Radius of curvature of asperities Angle of journal centre velocity relative to bearing load Angle of connecting rod axis relative to cylinder axis or Walther-ASTM constant

rad

δ

Displacement

m

ε

Eccentricity ratio

-

η

Asperity density

1/m2

θ

Film coordinate

rad

θ1

Film reformation

rad

θ2

Film rupture

rad

 min

Angle of minimum film thickness relative to bearing datum Angle defining beginning of groove relative to bearing datum Angle defining end of groove relative to bearing datum Crank angle

rad

Zs

  2 γ

θb θe θCA θh

Pa-1 rad m rad rad cStokes

rad rad rad

λ

Angle defining oil hole position relative to bearing rad datum Thermal conductivity of oil J/m/K

μ

Dynamic viscosity

Ns/m2



Kinematic viscosity

m2/s

ρ

Oil density

kg/m3

σ

Asperity RMS height

m

σs

Standard deviation of summit height

m

φ

Flow correction

-

Ф

Flow factor (average flow model)

-

ψ

Attitude angle

rad

ω

Angular velocity

rad/s

ωj

Journal angular velocity

rad/s

ωb

Bearing angular velocity

rad/s

ω2

Load angular velocity

rad/s

654


4.9 References 1. OILFILM Suite User Manual, Ricardo DP 93/2072 2. 4.2 ENGDYN User Manual. Engine Dynamics Simulation Program, Ricardo DP 95/1769 3. ‘Feed-Pressure Flow in Plain Journal Bearings’, F.A. Martin, C.S. Lee, ASLE Transactions, Volume 26, July 1983 pp 381-392 4. ‘Calculation Methods for Steadily Loaded Pressure Fed Hydrodynamic Journal Bearings’, Engineering Sciences Data Unit, item 66023, 1966. 5. ‘Design Data for the Steadily Loaded Full Central Circumferential Grooved Plan Journal Bearing’, G.A. Clayton, C.M. Taylor. 6. ‘Engine Bearing Analysis and Design’, G.A. Clayton, PhD Thesis, Department of Mechanical Engineering, University of Leeds, February 1990. 7. ‘Dynamic Analysis of Rocking Journal Bearings with Multiple Offset Segments’, J.F. Booker, P.K. Groenka, H.D. van Leeuwen, ASLE-ASME Lub. Conf., New Orleans, Paper 81-Lub 34, Oct. 81 8. ‘Developments in Engine Bearings’,F.A. Martin, Tribology of Reciprocating Engines Conference, University of Leeds, Sept. 1982. 9. ‘Power Loss in Connecting Rod Bearings’, F.A. Martin, J.F. Booker, I.Mech.E., AD Autotech Congress, 1987. 10. ‘A Table of the Journal Bearing Integral’, J.F. Booker, Trans. ASME Journal of Basic Engineering, June 1965, Vol. 87 pp 533-535. 11. ‘SAE Handbook’, Published yearly by SAE. 12. ‘Viscosity - Temperature Pressure Reduction of Petroleum Products’, L. Grunberg, A.H. Nissan, Science of Petrol, part 3, pp 276-94, 1955. 13. ‘Application of Average Flow Model to Lubrication Between Rough Sliding Surfaces’, Nadir Patir, H. S. Cheng, Trans. ASME Technology, April 1979, Vol. 101 pp 220-230.

655

Journal of Lubrication

APPENDICES Calculation of Crankshaft Stiffness’ using the finite element method

E. APPENDICES 1 Calculation of Crankshaft Stiffness’ using the finite element

method 1.1 Overview The crank train model used by ENGDYN is simplified using a discreet mass-elastic system, idealised to rigid, massive nodes interconnected by flexible, massless elements. The stiffness matrix used for both the dynamic and compliant models is derived using static stiffnesses calculated using the finite element method. This appendix describes how ENGDYN evaluates the stiffnesses for each of the crankshaft webs and for the elements at the ends of the crank train. Three stiffness values, torsion, in-plane bending and out-of-plane bending are calculated for each crankshaft web. The elements at the ends of the crank train are assumed to be axisymmetric, and for each of these elements three stiffness values, torsion, bending and axial are calculated. ENGDYN sub-divides the finite element model supplied by the user into separate files that each contains the finite element model of a crank web or an end element.

1.2 Calculation of Crankshaft Web Stiffnesses ENGDYN orientates each crankshaft web such that the main bearing centre is at the origin and the crank pin lies in the xy-plane as shown in Figure E-1.

Figure E-1 Example Finite Element Model of a Crankshaft Web

657

APPENDICES Calculation of Crankshaft Stiffness’ using the finite element method Three stiffness values are calculated for each crankshaft web, which are derived from two linear-elastic finite element analyses. In each case, the crank pin cut plane is constrained with displacement loads applied at the main bearing cut plane. In the case of a flying web for a Vee-configured engine, the big end bearing cut plane containing the origin is treated as if it were the main bearing cut plane.

1.2.1

Bending Stiffnesses

The in-plane and out-of-plane bending stiffnesses are evaluated by applying bending moments as a displacement load to the model which has a symmetric with expansion constraint at the crank pin cut plane such that the plane remains in plane but can move axially. In addition, two nodes are restrained to prevent rigid body motion. Figure E-2 shows the crank pin cut plane mesh of the model shown in Figure E-1. ENGDYN constrains the model so that:  all the nodes on the crank pin cut plane have the same x displacement,  the node n1 at the centre of the crank pin is restrained in the y and z directions, and  the node n2 is restrained in the z direction.

Figure E-2 Crank Pin Journal Cut-Plane

ENGDYN considers two loadcases, one for in-plane stiffness the other for out-of-plane stiffness. For each loadcase displacement loads are applied to all the nodes on the main bearing cut plane. Displacements in the x direction corresponding to a small rotation about the z and y-axes are applied for in-plane bending and out-of-plane bending stiffness respectively. Figure E-3 and Figure E-4 show the displacement load for each case respectively together with the resulting displaced shape. For each loadcase ENGDYN calculates the reactive torque due to the displacement using the nodal reactions at the main bearing cut-plane. The bending stiffnesses are then evaluated by dividing the reaction by the known rotation.

658


Figure E-3 In-Plane Bending Stiffness Displacement Loads and Results

Figure E-4 Out-of-Plane Bending Stiffness Displacement Loads and Results

1.2.2

Torsional Stiffness

The torsional stiffness is evaluated by applying a moment as a displacement load to the model, which has an anti-symmetric with expansion constraint at the crankpin cut plane. In addition, three nodes are restrained to prevent rigid body motion. ENGDYN constrains the model so that:  all the y and z degrees of freedom on the crank pin cut plane move together, and  the nodes n1, n2 and n3, as shown in Figure E-2, are restrained in the x direction. ENGDYN considers a single loadcase with a displacement load is applied to all the nodes on the main bearing cut plane. Displacements in the y and z directions are applied corresponding to a small rotation about the x-axis, such that at each node the vector of the displacement is in the radial direction only. Figure E-5 shows the displacement load together with the resulting displaced shape.

659


Figure E-5 Torsional Stiffness Displacement Loads and Results ENGDYN calculates the reactive torque due to the displacement using the nodal reactions at the main bearing cut-plane. The torsional stiffness is then evaluated by dividing the reaction by the known rotation.

1.3 Calculation of Crankshaft Element Stiffnesses ENGDYN orientates each model of the ends of the crank train such that the centre of one cut-plane is at the origin and the axis of the model lies along the x-axis. Three stiffness values are calculated for model, which are derived from three linearelastic finite element analyses. The models of each end are assumed to be axisymmetric. In each case, one cut plane is constrained with displacement loads applied at the other.

1.3.1

Bending Stiffness

The bending stiffness is evaluated in an identical manner to that described in A.2.1 for the crankshaft web. Since the model is assumed to be axisymmetric only one bending stiffness is calculated.

1.3.2

Torsional Stiffness

The torsional stiffness is evaluated in an identical manner to that described in A.2.2 for the crankshaft web

660

APPENDICES Calculation of Crankshaft Stiffness’ using the finite element method 1.3.3

Axial Stiffness

The axial stiffness is evaluated by applying an axial force as a displacement load to one cut plane of the model and restraining the other cut plane axially. In addition, a single node at the centre of the axial restraint cut-plane is restrained in y and z to prevent rigid body motion. ENGDYN calculates the reaction due to the displacement using the nodal reactions at the loaded cut plane. The axial stiffness is then evaluated by dividing the reaction by the known displacement.

661

APPENDICES A Method of Evaluating the Mass and Stiffness Properties of a Flywheel

2 A Method of Evaluating the Mass and Stiffness Properties of a

Flywheel 2.1 Overview The flywheel can be included as part of the finite-element model of the crank train as described in Section 2. In this case the stiffness’ of the element connecting the adjacent journal node and the flywheel node, and the mass properties at each of these nodes are calculated by ENGDYN. An alternative, more thorough method is to consider the flywheel in isolation so as to derive equivalent stiffness and mass and inertia values from a modal analysis of the flywheel. This Appendix describes such a method in which either FEARCE or MSC/NASTRAN is used to perform the modal solution.

2.2 Modelling Approach The flywheel is modelled using two nodes connected by a spring and damper element and is excluded from the finite-element model of the crank train used by ENGDYN. The first node n is positioned at the end of the crankshaft at the flange where the flywheel is bolted to the crankshaft. This is shown in Figure E-6.

Figure E-6 Finite Element Model of Crank Train The second node n  1 which represents the outer rim of the flywheel mass may be positioned at any point beyond the first node n.

2.3 Calculation of Mass and Stiffness Properties

663

APPENDICES A Method of Evaluating the Mass and Stiffness Properties of a Flywheel An example finite element model of a conventional flywheel is shown in view 1 of Figure E-7. The model is restrained in all directions at the joint face where the flywheel is bolted. It is important that the model is aligned such that the origin is at the centre of restraint face. This is shown in red of view 2 of the same figure.

Figure E-7 Finite Element Model of Flywheel and Model Restraints An eigenvalue solution is performed to predict the first 20 modes of the restrained flywheel. The 1st axial, bending and torsion modes are identified by inspection using the Ricardo Software program R-Desk or an equivalent FE viewing package. Table E-1 summaries the mode numbers and natural frequencies for the example shown above, whilst. Figure E-8 shows the corresponding mode shapes.

664


Figure E-8 Predicted Restrained Flywheel Modes Mode Number 1,2 3 6 9,10

Frequency [Hz] 780 1080 1809 4016

Description Bending (tilt) Axial Torsion Bending (radial)

Table E-1 Predicted Restrained Flywheel Modes In addition to the modal frequencies, the modal effective mass in each direction is also required. This data is calculated by default when the eigenvalue solution is performed using either FEARCE or MSC/Nastran. How this data is extracted and subsequently used is described below.

2.3.1

Using FEARCE

If an eigenvalue solution is performed using the FEARCE Solver then the modal effective mass is written to the SFE to an array called FE:MODAL_EFFECTIVE_MASS_SUM. This is an array containing 6 values, firstly 3 masses for each global direction (x, y, z) and secondly the inertias in x, y and z.

2.3.2

Using MSC/Nastran

665

APPENDICES A Method of Evaluating the Mass and Stiffness Properties of a Flywheel If the MSC/NASTRAN analysis data deck is written using the FEARCE translator sftona then the In order to derive the mass properties at the second node n  1 it is necessary to include in the Executive Control Section a dmap alter supplied by MSC/NASTRAN and contained in a file called checka.v705. This file, included using the INCLUDE statement, contains the dmap alter KEEFW which calculates the total modal effective mass of the assembly. This is written to a matrix called MERTMERW, with the output from the grid point weight generator to the .f06 file. The matrices for the above example are given in Table E-2 and Table E-3. The values that are of interest for deriving the equivalent system are highlighted.

0 MATRIX MERTMERW (GINO NAME 101 ) IS A DB PREC 6 COLUMN X 6 ROW SQUARE MATRIX. 0COLUMN 1 ROWS 1 THRU 6 -----------------------------------------ROW 1) 1.4373D+01 4.1208D-08 1.3183D-08 1.5375D-10 -2.9117D-08 -4.8004D-08 0COLUMN 2 ROWS 1 THRU 6 -----------------------------------------ROW 1) 4.1208D-08 1.4359D+01 -1.4492D-03 3.2851D-08 -2.8971D-04 8.2488D-02 0COLUMN 3 ROWS 1 THRU 6 -----------------------------------------ROW 1) 1.3183D-08 -1.4492D-03 1.4363D+01 3.7984D-08 -8.1811D-02 2.8971D-04 0COLUMN 4 ROWS 1 THRU 6 -----------------------------------------ROW 1) 1.5375D-10 3.2851D-08 3.7984D-08 1.7991D-01 5.6656D-10 7.1883D-10 0COLUMN 5 ROWS 1 THRU 6 -----------------------------------------ROW 1) -2.9117D-08 -2.8971D-04 -8.1811D-02 5.6656D-10 9.2374D-02 5.7913D-05 0COLUMN 6 ROWS 1 THRU 6 -----------------------------------------ROW 1) -4.8004D-08 8.2488D-02 2.8971D-04 7.1883D-10 5.7913D-05 9.2238D-02

Table E-2 Modal Effective Mass and Inertia in MSC/NASTRAN .f06 file

666


O U T P U T

* * * * * *

F R O M

G R I D P O I N T W E I G H T G E N E R A T O R REFERENCE POINT = 0 M O 1.479350E+01 0.000000E+00 0.000000E+00 0.000000E+00 -3.345901E-08 -3.688397E-08 0.000000E+00 1.479350E+01 0.000000E+00 3.345901E-08 0.000000E+00 8.207294E-02 0.000000E+00 0.000000E+00 1.479350E+01 3.688397E-08 -8.207294E-02 0.000000E+00 0.000000E+00 3.345901E-08 3.688397E-08 1.807075E-01 5.603888E-10 7.154716E-10 -3.345901E-08 0.000000E+00 -8.207294E-02 5.603888E-10 9.278250E-02 -3.578823E-10 -3.688397E-08 8.207294E-02 0.000000E+00 7.154716E-10 -3.578823E-10 9.278250E-02 S * 1.000000E+00 0.000000E+00 0.000000E+00 * * 0.000000E+00 1.000000E+00 0.000000E+00 * * 0.000000E+00 0.000000E+00 1.000000E+00 * DIRECTION MASS AXIS SYSTEM (S) MASS X-C.G. Y-C.G. Z-C.G. X 1.479350E+01 0.000000E+00 2.493255E-09 -2.261737E-09 Y 1.479350E+01 5.547904E-03 0.000000E+00 -2.261737E-09 Z 1.479350E+01 5.547904E-03 2.493255E-09 0.000000E+00 I(S) * 1.807075E-01 -7.650175E-10 -5.298442E-10 * * -7.650175E-10 9.232717E-02 3.578824E-10 * * -5.298442E-10 3.578824E-10 9.232717E-02 * I(Q) * 1.807075E-01 * * 9.232717E-02 * * 9.232717E-02 * Q * 1.000000E+00 0.000000E+00 0.000000E+00 * * 0.000000E+00 1.000000E+00 0.000000E+00 * * 0.000000E+00 0.000000E+00 1.000000E+00 *

* * * * * *

Table E-3 Total Mass and Inertia in MSC/NASTRAN .f06 file

2.3.3

Deriving the Mass and Stiffness Data for the Flywheel Panel

For the flywheel it is necessary to define the mass properties of 2 nodes connected by stiffness’ in each global direction. The second node n  1 represents the outer rim of the flywheel mass and may be positioned at any point beyond the first node n. The mass at the second node is assumed to be the modal effective mass, as described above. ENGDYN does not currently allow the mass to vary with direction. Since the axial mode usually has a lower frequency compared with the radial bending modes the effective mass in the x direction is taken as the mass at node n  1 such that

mn 1  m xeffective

mn  m x  mn 1 The stiffness’ k x , k y and k z are calculated using

f 

1 2

k mn 1

such that

k  4 2 f 2 mn1

667

APPENDICES A Method of Evaluating the Mass and Stiffness Properties of a Flywheel where f is the frequency either of the axial or radial bending mode and k is the corresponding stiffness. Similarly the effective inertia at node n  1 and rotational stiffness’ are calculated as follows:

I n 1  I effective

I n  I  I n 1 k  4 2 f 2 I n 1

It is assumed that the centre of mass of the lumped mass at node n  1 is at the same location as the centre of mass for the complete flywheel. For the above example the mass, inertia and stiffness values are given by,

mn 1 mn I xxn 1 I xxn I yyn 1 I yyn I zzn 1 I zzn

kx ky kx ky kz

 14.37 kg  14.79  14.37  0.42 kg  0.1799 kgm2  0.1807  0.1799  0.0008 kgm2  0.092374 kgm2  0.092783  0.092374  0.000409 kgm2  0.092238 kgm2  0.092783  0.092238  0.000545 kgm2

 4 2 1080  14.37  661.7  10 6 N / m 2  k z  4 2 4016  14.37  9.150  10 9 N / m 2  4 2 1809  0.1799  23.241  10 6 Nm / rad 2  4 2 780  0.092374  2.219  10 6 Nm / rad 2  4 2 780  0.092238  2.215  10 6 Nm / rad 2

If it is assumed that the node n  1 is at the centre of mass of the complete assembly then the mass offset for this mass is 0.0 mm whilst the mass offset for the mass at node n is 5.548 mm.

668

APPENDICES Calculation of Material Strength Properties for Crankshaft Stress Analysis

Calculation of Material Strength Properties for Crankshaft Stress Analysis 3

3.1 Overview Four material properties are required in order to construct the Goodman diagram as described in Appendix 7:

 ut  yt ,  yc  fl

is the ultimate tensile strength are the yield strengths in tension and compression is the uniaxial fatigue limit stress

In addition the torsional fatigue strength fs is required when using either the Multi-Axial or Dang Van fatigue safety factor calculation method and the hydrostatic fatigue strength h is required when performing the Dang Van fatigue safety factor calculation. In the absence of any suitable material property data the torsional and hydrostatic fatigue strengths can be derived from the infinite-life tensile fatigue strength such that:

 fs  a fsb  h  b fs where for Steel crankshafts a and b are 0.58 and 2.5 respectively, and for SG-Iron crankshafts a and b are 0.68 and 0.83.

3.2 Base Strengths The base strengths are defined as those for the as cast or forged material without any consideration of size, surface finish or surface treatment effects.

3.3 Elevated Strengths The elevated strengths are arrived at after consideration of size, surface finish and surface treatment effects. In the absence of any measured strength values these elevated strengths may be calculated as described below. 3.3.1

Infinite-Life Fatigue Strengths

The quoted tensile, torsional and hydrostatic fatigue strengths must be modified to account for size, surface finish and surface treatment effects.

669

APPENDICES Calculation of Material Strength Properties for Crankshaft Stress Analysis The size factor is required because the quoted strength is derived from tests on a 10 mm diameter specimen and the fatigue strength of steel, in general, reduces as specimen size increases. The surface treatment factor is applied to the fillet fatigue strengths to account for the effects of cold-rolling, induction-hardening, etc.

3.3.2

Yield Strengths

The yield strengths are assumed unaffected by the size and surface finish effects and affected by the surface treatment factor only if the surface is cold-rolled up to a maximum of 90% of the ultimate tensile strength value.

3.3.3

Ultimate Tensile Strength

The ultimate tensile strength is assumed unaffected by size, surface finish or surface treatment effects.

670

APPENDICES Calculation of Stress Concentration Factors for Crankshaft Stress Analysis

4 Calculation of Stress Concentration Factors for Crankshaft

Stress Analysis 4.1 Overview Stress concentration factor (SCF) values at each fillet and oil hole breakout are calculated by default from geometry supplied by the user using methods based on work by Lowell1, Arai2, Peterson3 and Oldberg4. Alternatively the SCF values can either be defined by the user or calculated using alternative algorithms supplied by the user using environment variables. At each fillet and oil hole breakout a SCF value is calculated for torsion and for bending.

4.2 Nomenclature

 b d D Djo Dji R T Uj Uw U

          

Inclination of oil hole drilling to normal at breakout Equivalent width of the adjacent web Oil hole diameter Journal diameter at fillet Outer journal diameter Inner journal diameter Journal fillet radius Adjacent web thickness Undercut into journal Undercut into adjacent web Web overlap

deg m m m m m m m m m m

4.3 Journal Fillets The method for calculating the SCF values at each journal fillet is by default based on work by Lowell1 and Arai2 . The following relations are used to define the SCF values for bending Kb and for torsion, Kt. The journal diameter at the fillet, D accounting for any undercut into the journal is given by

D  Djo  2.Uj For bending the stress concentration factor, Kb is given by

Kb  Ab.Vb. K where

Ab and Vb are from Lowell1 and are defined as

671


t Ab  12 .    r

0.455

4

3

2

b b b b Vb  0.0615   0.544   1.873   2.434   1.962  D  D  D  D and K is a correction based on the work by Arai2 to account for any undercut into the adjacent web and is defined as

 Uw   r  K  10 .  81     r   D

2

2  u    0.769   0.407     D  

For torsion the stress concentration factor, Kt is given by

 D Kt    r

0 .1015   u  0.2205        D  

Alternatively the SCF values can either be defined by the user or calculated using alternative algorithms supplied by the user using environment variables that are defined in Table E-4. The equations should be defined as functions of the variable names listed in Table E-5. These are set by the user either using the csh command setenv before executing the program or are included in the engdyn script that resides in the installation directory under Ricardo/bin. Environment Variable FILLET_SCF_CALCULATION_METHOD FILLET_SCF_BENDING_EQUATION FILLET_SCF_TORSION_EQUATION

Definition Name of the method. Equation defining the stress concentration factor for bending, Kb for the journal fillets. Equation defining the stress concentration factor for torsion, Kt for the journal fillets.

Table E-4 Environment Variables Used to Define SCF Values for Journal Fillets Variable Name RADIUS WEB_CUT J_CUT ODJ OVERLAP WIDTH THICK

Description Journal fillet radius, r Undercut into adjacent web, Uw Undercut into journal, Uj Outer journal diameter, Djo Web overlap, u Equivalent width of the adjacent web, b Adjacent web thickness, t

Table E-5 Variable Names to Define SCF Equations for Journal Fillets For example using the setenv command to define equations based on Lowell but ignoring the effect due to undercuts, it would be necessary before executing engdyn to type the following: 672

APPENDICES Calculation of Stress Concentration Factors for Crankshaft Stress Analysis setenv FILLET_SCF_CALCULATION_METHOD ‘Lowell’ > setenv FILLET_SCF_BENDING_EQUATION \ > ‘1.2*(0.0615*(WIDTH/ODJ)**4+0.544*(WIDTH/ODJ)**3+1.873*(WIDTH/ODJ)**2+2.434*WIDTH/ODJ+1.962)*(THICK /RADIUS)**0.455’ setenv FILLET_SCF_TORSION_EQUATION ‘(ODJ/RADIUS)**(0.2205+(OVERLAP/ODJ)**0.1015)’

Alternatively these could be included in the engdyn script as follows: FILLET_SCF_CALCULATION_METHOD = ‘Lowell’ > FILLET_SCF_BENDING_EQUATION > ‘1.2*(0.0615*(WIDTH/ODJ)**4+0.544*(WIDTH/ODJ)**3+1.873*(WIDTH/ODJ)**2+2.434*WIDTH/ODJ+1.962)*(THICK /RADIUS)**0.455’ FILLET_SCF_TORSION_EQUATION = ‘(ODJ/RADIUS)**(0.2205+(OVERLAP/ODJ)**0.1015)’ export FILLET_SCF_CALCULATION_METHOD > export FILLET_SCF_BENDING_EQUATION export FILLET_SCF_TORSION_EQUATION

4.4 Oil Hole Breakouts The method for calculating the SCF values at each oil hole breakout is by default based on Peterson3 and work by Oldberg4. The following relations are used to define the SCF value for bending Kb and for torsion Kt.. For bending the stress concentration factor, Kb is given by 2

d d Kb  A   B   C  Dj   Dj  where A, B and C are coefficients are obtained from quadratic curve-fits and by interpolation from the curves shown in Figure E-9.

STRESS CONCENTRATION FACTOR, Kb

3.5

3 Dji/Djo = 0.0 Dji/Djo = 0.6 Dji/Djo = 0.9 2.5

2 0

0.05

0.1

0.15

d/Djo

Figure E-9 Oil Hole Breakout Stress Concentration Factor for Bending

673

= \

APPENDICES Calculation of Stress Concentration Factors for Crankshaft Stress Analysis Each curve is for a different ratio of (Dji/Djo) where Dji/Djo=0 is for a solid journal. The curve-fits are valid for values of d/ Djo up to 0.15. For torsion the stress concentration factor, Kt is given by

  d 2   d  Kb  F  A   B   C  Djo    Djo   where F is a correction factor based on the work by Oldberg 4 to account for a non-radial oil hole drilling. If the inclination of the oil hole drilling relative to the normal at the breakout is less than 22.5o then the factor F is 1.0 otherwise F is defined by

F  8.79024  10 4   22.5  10411 . 2

A, B and C are coefficients are obtained from quadratic curve-fits and by interpolation from the curves (which are for a radial drilling) shown in Figure E-10. Each curve is for a different ratio of (Dji/Djo) where Dji/Djo=0 is for a solid journal. The curve-fits are valid for values of d/ Djo up to 0.15.

STRESS CONCENTRATION FACTOR, Kt

3.5

3 Dji/Djo = 0.0 Dji/Djo = 0.6 Dji/Djo = 0.9 2.5

2 0

0.05

0.1

0.15

d/Djo

Figure E-10 Radial Oil Hole Breakout Stress Concentration Factor for Torsion Alternatively the SCF values can either be defined by the user or calculated using alternative algorithms supplied by the user using environment variables that are defined in Table E-6. The equations should be defined as functions of the variable names listed in Table E-7. These are set by the user either using the csh command setenv before executing the program or are included in the engdyn script that resides in the installation directory under Ricardo/bin.

674

APPENDICES Calculation of Stress Concentration Factors for Crankshaft Stress Analysis Environment Variable

Definition Name of the method. Equation defining the stress factor for bending, Kb for breakouts. Equation defining the stress factor for torsion, Kt for breakouts.

OILHOLE_SCF_CALCULATION_METHOD OILHOLE_SCF_BENDING_EQUATION

OILHOLE_SCF_TORSION_EQUATION

concentration the oil hole concentration the oil hole

Table E-6 Environment Variables Used to Define SCF Values for Oil Hole Breakouts Variable Name OD ODJ IDJ INCL

Description Oil hole diameter, d Outer journal diameter, Djo Outer journal diameter, Dji Inclination of oil hole drilling to normal at breakout, 

Table E-7 Variable Names to Define SCF Equations for Oil Hole Breakouts For example using the setenv command to define equations based on Peterson but assuming that the journal is always solid and the oil hole drilling is always radial, it would be necessary before executing engdyn to type the following: setenv OILHOLE_SCF_CALCULATION_METHOD ‘Solid Shaft with Radial Drilling’

setenv OILHOLE_SCF_BENDING_EQUATION 6.544*OD/ODJ+2.999)’

‘(37.780*(OD/ODJ)**2-

setenv OILHOLE_SCF_TORSION_EQUATION 4.3302*OD/ODJ+2.3934)’

‘(22.998*(OD/ODJ)**2-

Alternatively these could be included in the engdyn script as follows: OILHOLE_SCF_CALCULATION_METHOD =‘Solid shaft with Radial Drilling’ OILHOLE_SCF_BENDING_EQUATION=(37.780*(OD/ODJ)**2-6.544*OD/ODJ+2.999)’ OILHOLE_SCF_TORSION_EQUATION= ‘(22.998*(OD/ODJ)**2-4.3302*OD/ODJ+2.3934)’ export export export

OILHOLE_SCF_CALCULATION_METHOD OILHOLE_SCF_BENDING_EQUATION OILHOLE_SCF_TORSION_EQUATION

1

Published by the Author, 1975, ‘The Crankshaft : Designing for Structural Strength and Fatigue Resistance’, C.M.Lowell

2

Bulletin of the JSME, 1965, ‘The Bending Stress Concentration Factor of Solid Crankshaft’, J.Arai

3

Wiley-Interscience, 1974, ‘Stress Concentration Factors’, R.E.Peterson.

675


4

Experimental Stress Analysis, Vol.2, No.2, 1945, ‘Structural Evolution of a Crankshaft’, S.Oldberg and C.Lipson, Chrysler Corporation.

676

APPENDICES Treatment of Notch Sensitivity in Crankshaft Stress Analysis

5 Treatment of Notch Sensitivity in Crankshaft Stress Analysis 5.1 Overview In general a component whose geometry gives rise to a stress concentration will show a reduction in fatigue strength. However, it is usual to find that the reduction in fatigue strength is not equal to the stress concentration factor (SCF). In some materials no reduction may be seen and such a material would be said to have no notch sensitivity. Conversely, a material whose fatigue strength reduction is equal to the SCF is said to be 100% notch sensitive. The usual method by which the notch sensitivity of a material is allowed for in conventional stress analysis of crankshafts is to derive a fatigue notch factor (Kf) from the SCF. The fatigue notch factor is then used in place of the SCF in all calculations relating to fatigue. By default in ENGDYN the relationship between Kf and SCF is based on work by Lowell1 and modified according to Ricardo experience. The curves for steel and cast iron are shown in Figure E-11. As can be seen from this figure, cast iron has approximately half the notch sensitivity of steel. For ductile materials such as steel the Kf is only applied to the alternating stress component not the mean stress component.

FATIGUE NOTCH FACTOR

F

3.5 3 2.5 2 1.5 STEELS WITH HARDNESS > 200 BHN 1 SPHEROIDAL GRAPHITE CAST IRONS

0.5 0 1

1.5

2

2.5

3

3.5

4

4.5

STRESS CONCENTRATION FACTOR

Figure E-11 Variation of Fatigue Notch Factor with Stress Concentration Factor Alternatively, the relationship between Kf and SCF can be defined by the user using environment variables that are defined in Table E-8. These are set by the user either using the csh command setenv before executing the program or are included in the engdyn script that resides in the installation directory under Ricardo/bin.

677

APPENDICES Treatment of Notch Sensitivity in Crankshaft Stress Analysis Environment Variable FNF_CALCULATION_METHOD FNF_STEEL_EQUATION FNF_SG_IRON_EQUATION

Definition Name of the method. Equation in terms of a variable SCF defining the relationship between Kf and SCF for steel crankshafts. Equation in terms of a variable SCF defining the relationship between Kf and SCF for SG iron crankshafts.

Table E-8 Environment Variables Used to Define a Relationship Between Kf and SCF For example, to define relationships which result in steel being 100% notch sensitive and SG iron being 50% notch sensitive using the setenv command it would be necessary before executing engdyn to type the following: setenv setenv setenv

FNF_CALCULATION_METHOD FNF_STEEL_EQUATION FNF_SG_IRON_EQUATION

‘Linear’ ‘SCF’ ‘SCF/2.0’

Alternatively these could be included in the engdyn script as follows : FNF_CALCULATION_METHOD FNF_STEEL_EQUATION FNF_SG_IRON_EQUATION export export export

=‘Linear’ =‘SCF’ =‘SCF/2.0’

FNF_CALCULATION_METHOD FNF_STEEL_EQUATION FNF_SG_IRON_EQUATION

5.2 Treatment in Finite Element Analysis In order to allow for notch sensitivity when interpreting the results of a finite element analysis it is necessary to calculate the SCF at critical sections. This is required since notch sensitivity is a function of both material and SCF, though it should be noted that it is a relatively weak function of SCF as shown in Figure E-11. There are two approaches that can be used to calculate the crankshaft SCF values. The first is to apply a simple loadcase to the FE model and derive a SCF on the basis of peak and nominal stresses calculated by the FE analysis. The second approach is to use a SCF that has been calculated using traditional methods. It is Ricardo experience that the second approach is preferable because the calculation is less dependent on judgement and it has the advantage of bringing the overall calculation procedure closer to the traditional approach with which the final result will be compared. A description of the calculation of a SCF using conventional analysis is given inAppendix 4. The calculated SCFs are used to obtain a fatigue notch factor from the curves shown in Figure E-11. Stress values allowing for notch sensitivity are then calculated by the following relationship:

678

APPENDICES Treatment of Notch Sensitivity in Crankshaft Stress Analysis

NotchSensi tivityModi fiedStress 

FatigueNot chFactor  FE Pr edictedStr ess [E-1] StressConcentrationF actor

In practice, this means that a stress correction factor is used given by :

StressCorrectionFact or 

FatigueNot chFactor StressConcentrationF actor

[E-2]

The stress values obtained do not, therefore, represent the real stress state of the crankshaft but can be used for calculating a fatigue safety factor. 1

Published by the Author, 1975, ‘The Crankshaft : Designing for Structural Strength and Fatigue Resistance’, C.M.Lowell

679

APPENDICES Using Output from the Ricardo Program VALDYN and TVFORCED in Crankshaft Stress Calculations

6 Using Output from the Ricardo Program VALDYN and

TVFORCED in Crankshaft Stress Calculations 6.1 Using VALDYN for Crankshaft Stress Calculations The Ricardo program VALDYN1 can be used to predict the torsional vibration of a crankshaft using either the Linear Frequency Domain (LFD) or time domain solver. A typical frequency domain model for a 6-cylinder engine is shown in Figure E-12

Figure E-12 Typical Frequency Domain VALDYN Model for a 6-Cylinder Engine Data is read by ENGDYN from the sdf file written by VALDYN. This is the only file required by ENGDYN.

6.1.1

Defining a Linear Frequency Domain Model

There are a number of modelling considerations when defining a LFD VALDYN model for use by ENGDYN. These are as follows:

681

APPENDICES Using Output from the Ricardo Program VALDYN and TVFORCED in Crankshaft Stress Calculations 1. The LFD Torque Sense for each CRANK element must be set to 1 for an engine rotating clockwise as viewed from the crank nose and –1 for an engine rotating anti-clockwise. 2. The i (white) node of the SSTIFF elements defining the crankshaft stiffness must be the node nearest to the crank nose, as shown in Figure E-12. This is then consistent with the definition of the torque sense as defined above. 3. The number of each CRANK element has to be the same as the cylinder number, as defined in ENGDYN by the ENGDYN will test that the firing angle as defined by the Phase Angle of the CRANK element are consistent with that defined in ENGDYN.

6.1.2

Defining a Time Domain Model

An equivalent time domain model of the frequency domain model is shown in

Figure E-13 Typical Time Domain VALDYN Model for a 6-Cylinder Engine In order to restrain the model an additional MASS element is added (with zero mass) to which is connected each cylinder damper, a coupling stiffness (SSTIFF), and an ANGLE element defining the angle of rotation. It is the responsibility of the user to ensure that 682

APPENDICES Using Output from the Ricardo Program VALDYN and TVFORCED in Crankshaft Stress Calculations the stiffness, damping and preload parameters of the coupling stiffness are defined to prevent any rigid body drift. This may be monitored by an optional LSTIFF as shown. The modelling considerations of the frequency domain model also apply to the time domain model with one exception. The LFD Torque Sense of the CRANK element has no meaning when running in the time-domain and in this case when reading a timedomain ENGDYN reverses the sign as appropriate. It is the responsibility of the user to ensure that the ratio of the ANGLE element should always be set to 1.0 regardless of the rotational direction of the crankshaft. A value of 1.0 with respect to ENGDYN is equivalent to an engine rotating clockwise as viewed from the crank nose. For a counter-rotating crankshaft ENGDYN will reverse the sign of the torques read from the sdf file. For the time domain model ENGDYN will adjust the torques read from VALDYN such that the mean torque at each cylinder is equal to either the ENGDYN calculated indicated torque or the user-defined brake mean torque.

6.1.3

Defining the Output Required by ENGDYN

It is necessary to add the force (torque) of each SSTIFF that connects to a CRANK element as an SDF OUTPUT. Each output is named Crankshaft Stiffness and tagged using CYL. The value supplied for the CYL tag is the number of the CRANK element (i.e. cylinder) that the i (white) node of the STTIFF connects to. Figure E-14 shows the definition of the SDF OUTPUT for the SSTIFF connecting cylinders 2 and 3 of the above model.

Figure E-14 Defining SDF OUTPUT for an SSTIFF element

683

APPENDICES Using Output from the Ricardo Program VALDYN and TVFORCED in Crankshaft Stress Calculations For the SSTIFF defining the stiffness of the crank nose the value of the CYL tag should be set to 0. Figure E-15 shows the SDF OUTPUTs for each of the SSTIFF elements the above model.

Figure E-15 SSTIFF SDF OUTPUTS This data is then written by VALDYN to the sdf file for each speed as: g:profile1D:force:TYP_crankStiffness:CYL_0:SP_2000 g:profile1D:force:TYP_crankStiffness:CYL_1:SP_2000 g:profile1D:force:TYP_crankStiffness:CYL_2:SP_2000 g:profile1D:force:TYP_crankStiffness:CYL_3:SP_2000 g:profile1D:force:TYP_crankStiffness:CYL_4:SP_2000 g:profile1D:force:TYP_crankStiffness:CYL_5:SP_2000 g:profile1D:force:TYP_crankStiffness:CYL_6:SP_2000 with corresponding crank angle data written to g:profile1D:crankAngle. ENGDYN will read the last cycle of data and it is necessary that at least one cycle of data is written (as defined by the Analyse Panel in VALDYN)

6.2 Using TVFORCED for Crankshaft Stress Calculations 6.2.1

The TVFORCED Program

The Ricardo program TVFORCED2 calculates the forced and damped torsional vibration of a crankshaft, either of a single harmonic order or at each of the first 24 orders of vibration. The crankshaft is modelled as a lumped parameter mass/spring/damper system, with gas and inertia torque excitation of the cylinder masses. Damping can be both to ground and in the shafts between masses, and damping in shafts can be specified either as a

684

APPENDICES Using Output from the Ricardo Program VALDYN and TVFORCED in Crankshaft Stress Calculations viscous damping coefficient or as a fraction of critical damping. This allows both cylinder damping and most types of torsional vibration damper to be modelled.

6.2.2

Using TVFORCED Output for Crankshaft Stress Calculations

The data file name must always have a .dat suffix. In addition when using TVFORCED output for ENGDYN it is necessary to have an additional suffix to denote the engine load type. For example filename.dat might become basename.fl.dat where fl denotes that this file corresponds to Full Load. Similarly pl and nl denote Part Load and No Load respectively. An example input data file for an in-line 4 engine with a torsional vibration damper is shown in Figure E-16. AN EXAMPLE TVFORCED DATA FILE FOR USE WITH ENGDYN PLOT: DISP:SPEED 2 PLOT: TORQ:SPEED ALL PLOT: POWER 1 PRINT: MAXMINTORQUE ALL PLOT: TORQUE:ANGLE ALL EVERY 1 78.0 85.0 130.0 0.765 INLINE 4 0 0 540 180 360 7 0 2.500e-3 0 15000 -0.125 0 1.496e-3 0 0.08755E6 0 1 5.9601E-3 1.54 0.33663E6 0 1 5.9601E-3 1.54 0.35806E6 0 1 5.9057E-3 1.54 0.35165E6 0 1 5.9038E-3 1.54 0.53842E6 0 0 120.4e-3 0 0 0 0 USER_SPECIFIED 2000 2000.tvf 3000 3000.tvf 4000 4000.tvf

Figure E-16 Example TVFORCED Data File The relative firing order between cylinders must be consistent with that defined in ENGDYN using the Engine Configuration Panel. ENGDYN determines the phase angle between the TVFORCED data and 0 deg CA as defined by ENGDYN. This angle is displayed when the TVFORCED data is read. If a Vee engine is being modelled then the engine should be defined as INLINE not VEE. The data file must contain the line PLOT:TORQUE:ANGLE ALL EVERY 1 as highlighted in the figure and this must be included only once. Inclusion of this plot option will ensure that for each speed analysed output data files will be generated that will contain the 24 order synthesised torque in all the shafts versus crank angle at 1 degree intervals. The file names will be of the form:

685

APPENDICES Using Output from the Ricardo Program VALDYN and TVFORCED in Crankshaft Stress Calculations basename.fl.torq.speed.{plt,RPL} ENGDYN requires the .plt file only and requires a file corresponding to each speed for which stresses are required. For the example given in Figure E-16 the following files will be generated by TVFORCED. basename.fl.torq.2000.plt basename.fl.torq.3000.plt basename.fl.torq.4000.plt The Crankshaft FE Unit Loads Analysis and Crankshaft Classical Stress Analysis panels are used to define the name of TVFORCED data file to be used for the stress and fatigue analysis. The .plt files should reside in the same directory as the .dat file and are opened when the stresses and fatigue safety factors are calculated. 1

Ricardo Document RD06/312401.1 ‘VALDYN Documentation / User Manual Version 4.1’

2

Ricardo Report DP93/229 ‘TVFORCED 5.0’

686

APPENDICES Safety Factor Calculations

7 Safety Factor Calculations 7.1 The Goodman Diagram

Figure E-17 The Goodman Diagram The Goodman diagram as shown in Figure E-17 is used to calculate safety factors, given the minimum and maximum stress at a point, and a set of material properties. It is the most commonly used criterion, as it requires the minimum number of material properties. It is based on data from uniaxial fatigue tests so the single 'stress' value used has to be

687

APPENDICES Safety Factor Calculations derived in some way from a complete 3 dimensional stress field. This is usually done by taking the Von-Mises stress but there are various other methods as described below.

7.2 Goodman Diagram Construction Four material properties are required in order to construct the Goodman diagram:

 ut  yt ,  yc  fl

is the ultimate tensile strength are the yield strengths in tension and compression is the uniaxial fatigue limit stress

The Goodman diagram is drawn with the mean stress on the horizontal axis and the alternating stress on the vertical axis. Six points are plotted on the axes: a

 yc on the negative mean stress axis (compressive yield failure)

b

 yt on the positive mean stress axis (tensile yield failure)

c e

 ut on the positive mean stress axis (tensile failure)  yc on the alternating stress axis (single cycle compressive yield failure)  yt on the alternating stress axis (single cycle tensile yield failure)

f

 fl on the alternating stress axis (eventual fatigue failure)

d

These are the failure points for pure static or pure alternating stresses. It is now necessary to join these points to define the safe area for combined mean and alternating stress situations. Whether these lines are straight or not depends upon the material in question, some need concave lines some convex. For simplicity the Goodman criterion uses straight lines, which in the absence of further information is the best that can be done. The line a-d defines the limit beyond which compressive yield will occur. The line b-e defines the limit beyond which tensile yield will occur. The line c-f defines the limit beyond which a fatigue failure will occur. The three lines bounding the safe working area represent the maximum alternating stress  a  that can be tolerated at any given mean stress  m  . Any stress condition within this envelope will be safe. The safety factor defines how far inside the envelope the stresses are. Two options are offered for the way in which the safety factor is calculated - Cyclic Overload and General Overload.

The safety factor for cyclic overload is defined as the ratio of the maximum allowable alternating stress at  m , to the actual alternating stress  a . This factor will be greater than one for points within the safe working area, less than one but greater than zero for points outside the safe working area but with  m within the bounds  yc   m   yt , and zero elsewhere. If  a is zero, the safety factor will be infinity for  yc   m   yt , and 688

APPENDICES Safety Factor Calculations zero otherwise. The safety factor for general overload is defined as the ratio of the length of the line from (0,0) through  m ,  a  to the point where it intersects the boundary of the safe

working area, to the length of the line from (0,0) to  m ,  a  . This factor will be greater than one for points within the safe working area, and less than one everywhere else. It will only be zero if  m or  a is infinite, and will only be infinite if both  m and  a are zero. The combined overload option uses the General Overload calculation for positive mean stresses and Cyclic Overload for negative mean stress. For situations where the yield line limits are not critical there is an additional option No Yield where both yield strengths are replaced by the UTS. This option is of use when the yielding occurs on the initial loading cycle only. The result must always be interpreted with caution.

7.3 Equivalent Stress Options The stress points on the Goodman diagram are derived from uniaxial tests (i.e. stresses in one direction only) but it is necessary to use the diagram in real problems with multiaxial stress data. The stress tensors therefore have to be reduced to two equivalent stresses  m ,  s  , which can be treated in the same way as the uniaxial stresses. The most commonly used equivalent stress is the Von-Mises stress (originally developed to establish the yield point due to a multi-axial stress field):

2 yt2   1   2    2   3    3   1  2

2

2

[E-3]

where  1 ,  2 and  3 are the three principal stresses. However this equation will always produce a positive value for  yt which will tend to be pessimistic for compressive stress cases or vastly optimistic when the alternating stress varies from compressive to tensile, as both values will give positive equivalent stresses implying that the alternating stress component is small when it should be large. To overcome this limitation it is possible to give a sign to the Von-Mises stress to signify whether the stress is largely tensile or compressive. It can be shown that the Von-Mises stress is positive if the sum of the principal stresses is positive:

 yt 

 1   2 2   2   3 2   3   1 2

+ve if

1   2   3  0

[E-4]

This is the Standard Goodman option equivalent stress used by ENGDYN. This method also has its faults as now a reasonably high stress state may have  1   2   3  0 in which case there can be a sudden switch from a high positive equivalent stress to a high negative equivalent stress due to a very small change in the

689

APPENDICES Safety Factor Calculations complete stress tensor. To overcome this the Alternative Goodman option was developed. Here each stress tensor is broken down into its separate components and two new stress tensors are built up one using the maximum and one the minimum separate values. This produces two new peak stress values from which the two required equivalent stresses  m ,  s  can be calculated. This avoids some problem of sudden switches in the sign of the equivalent stress but does mean that the peak stresses being considered are more extreme than those actually applied, giving a more pessimistic (safer) prediction of failure. To completely overcome problems with sign flipping due to small changes in the mean stress the P1 Alternating Goodman option uses the P1 principal stress of the mean stress instead of the Von-Mises stress. This is more pessimistic again as it tends to predict stresses as being tensile when they are largely compressive. There are two other Goodman stress options currently implemented, the Maximum Principal stress option where the two peak equivalent stresses are taken to be the maximum and minimum of all the principal stresses  1 ,  2 ,  3  and the P1 Principal Stress option where the two peak equivalent stresses are taken to be the maximum and minimum of the P1 principal stresses  1  . For example, the following two stress conditions represent the peak stresses at a node:

 a1  100,  a 2  20,  a 3  100  b1  10,  b 2  2,  b3  10 The standard Goodman method will produce equivalent stresses:

 max  174,  min  17.4



 mean  78.3,  alt  95.7

The alternative Goodman method will produce equivalent stresses:

 max  100,20,10 ,  min  10,2,100  mean  55,9,55 ,  alt  45,11,45



 mean  95.7,  alt  34

The P1 alternative Goodman method will produce equivalent stresses:

 max  100,20,10 ,  min  10,2,100  mean  55,9,55 ,  alt  45,11,45



 mean  55, alt  34

The maximum principal Goodman method will produce equivalent stresses:

 max  100,  min  100



 mean  0,  alt  100

The P1 Principal Goodman method will produce equivalent stresses:

690


 max  100,  min  10

 mean  55,  alt  45



All of these options are approximations, there is no correct method, and each has its uses in different situations.

7.4 Multi-axial Fatigue Safety Factor Multi-axial fatigue safety factor calculations are derived from work by McDiarmid12, which gives an equation for calculating the maximum allowable alternating shear stress on the plane with the maximum mean shear stress. 1.5   fs 0.225 alt  2 mean   allow     1  0.5  uts   fs  3  

0.5

1.5 2  allow  C1  C2 alt  C3 mean

(1973)

[E-5]

(1989)

[E-6]

where

 fs

is the uniaxial reversed fatigue strength limit

 fs  fsb

is the shear stress fatigue limit is the fatigue stress limit under bending

and

 fsb    C1   fs ; C 2   fs  2  

1.5

  fsb        ; C3   fsb   uts   2   2   2 

2

The method is straightforward enough apart from finding the plane of maximum shear stress. There is no analytical method for finding this, as it will almost certainly be in a different direction for each stress state under consideration. An array of normal vectors are considered, each spread over the surface of a unit sphere separated by 15 (Giving a total of 146 normal directions) each of which is used to calculate the in-plane and normal stress for that direction for each stress state (i.e. load case) under consideration. Having shear normal stresses for each load case in each of the 146 directions it is now possible to calculate the allowable alternating shear stress for each direction. This is converted to a safety factor by dividing the allowable alternating shear stress by the actual alternating shear stress in each direction. The stored safety factor is the lowest value calculated. The usual lack of suitable material property data (three separate fatigue stress limits) means that the values of  fs and  fsb have to be calculated from the uniaxial reversed stress fatigue limit where

691


 fsb   fs

 fs 

and

 fs 3

[E-7]

7.5 Dang Van Fatigue Safety Factor The Dang Van fatigue safety factor calculations are based on work by Dang Van et al3. It is applicable to high cycle fatigue analysis of ductile metals and is therefore particularly appropriate for applying to crankshaft fatigue analysis. An example of its application is by Henry et al4 The Dang Van principal is similar to the multi-axial method except it requires the calculation of Hydrostatic and Principal stresses. The stresses are split into Hydrostatic and Deviatoric parts. The Hydrostatic stress is calculated from

h 

1  x   y   z  3

[E-8]

and the deviatoric stress is

 xx   h   d    xy   xz 

 xy  yy   h  yz

 xz    yz   zz   h 

[E-9]

The assumption of this method is that as the component is subjected to the stress cycle, at the microscopic level small amounts of plastic deformation occur within individual grains and at the same time the grains undergo strain hardening. This leaves a residual stress in the component. To calculate this residual stress an iterative procedure is performed as shown in Figure E-18.

692

APPENDICES Safety Factor Calculations Elastic stress range for stress state i+1

Ri+1

Stress Trajectory

Elastic stress range for stress state i

Ri

d

i

i

D

 i 1

d

i1

y

x Figure E-18 Iterative Procedure to Compute the Stabilised Residual Stress The elastic range of the material is initially set to zero (Ro=0). The initial residual stress (  0 ) is guessed as being the mean deviatoric stress throughout the cycle. The isotropic hardening coefficient (  ) is assumed to be 0.05. For each stress in the cycle a line is drawn from  i to the deviatoric stress. If this goes outside the current elastic range a line is drawn from  i to  di 1 . The length of the line is calculated as



D  J 2  di 1   i



[E-10]

The second invariant of deviatoric stress is given by

J 2   



1  1   2 2   2   3 2   3   1 2 6



[E-11]

The elastic range is increased to

Ri 1  Ri   D  Ri 

[E-12]

and the residual stress is moved to

 i 1   i 

D  Ri 1  di 1   i D



693



[E-13]

APPENDICES Safety Factor Calculations This procedure is repeated until R and  have converged. The deviatoric part of the localised stress at each condition is  ld   d   n and the alternating shear stress component is calculated as  

1 VonMises  ld  2

The safety factor is derived from the point on the cycle that is closest to the Dang Van line as shown in Figure E-19 and is given by

SafetyFact or 

 fs  hfs   h   hfs 

[E-14]

where

 fs

is the uniaxial tensile push-pull fatigue strength limit

 fs

is the shear stress fatigue limit

 hfs

is the hydrostatic fatigue stress limit stress.

 

fs

SafetyFact or 

A B A

B

A

 hfs

h

Figure E-19 Dang Van Diagram The hydrostatic fatigue stress limit cannot be measured but must be derived from the Dang Van plot using bi-axial test results. Typically for steel  hfs  2.5 fs and for cast iron  hfs  0.83 fs .

694

APPENDICES Safety Factor Calculations 7.6 LinearSWT Safety Factor

Figure E-20 Linear SWT Modified Goodman Diagram SWT safety calculations are based on work by Smith et al5. From extensive experimental work it has been found that equivalent stress for fatigue life calculations is proportional to

 SWT  where

 mean   alt  alt E

[E-15]

 mean is the mean stress  alt is the alternating stress  alt is the alternating strain amplitude E

is the Young’s Modulus

For the linear case when stress is proportional to strain the relation can be re-written as

 SWT 

 mean   alt  alt

[E-16]

equating this to the fatigue strength gives a curved line on the Goodman diagram where

695


 alt

alloe

where



2   mean   mean  4 2fs

2

[E-17]

 alt

is the allowable alternating stress for a given mean stress  mean

 fs

is the fatigue stress limit for the material

allow

Since the LinearSWT results are to be used in conjunction with the standard Goodman results the tensile yield line is removed entirely. The compressive yield line is left in for the standard results but is replaced with the UTS for the No Yield results.

7.7 Cycles to Failure Calculation For each of the cyclic safety factors calculated above it is possible to calculate an estimate of the number of cycles to failure from SN-curve data. The safety factor is converted back to a uniaxial reversed stress by multiplying it by the fatigue limit stress. This is then multiplied by the design safety factor (defaulted to 1.5) and looked up on the SN-curve to give the number of cycles to failure. For ease of plotting this is stored as a log10 value (i.e. 4 = 10000cycles).

Figure E-21 Calculation of the Number of Cycles to Failure Fatigue data is not generally available at raised temperatures so the fatigue limit stress 696

APPENDICES Safety Factor Calculations at room temperature is used which will give a reversed stress at an equivalent temperature to match the SN-curve data. This assumes that the SN-curve can be scaled using the temperature variation of the fatigue limit stress.

1

‘The Effect of Mean Stress on Biaxial Fatigue where the Stresses are Out-of-Phase and at Different Frequencies’, D.L.McDiarmid, 1989.

2

‘A General Criterion of Fatigue Failure under Multi-Axial Stress’, D.L.McDiarmid, 1973.

3

‘On a New Multi-Axial Fatigue Limit Criterion: Theory and Application", K.Dang Van, B.Griveau, O.Message, 1989.

4

SAE 920087, 1992, ‘Crankshaft Durability Prediction – A new 3-D Approach’, J-P.Henry, J.Toplosky, and M. Abramczuk.

5

‘A Stress-Strain Function for the Fatigue of Metals’, K.N.Smith, P.Watson and T.H.Topper, 1970.

697

APPENDICES Radiated Noise Calculations

8

Radiated Noise Calculations

8.1 Overview Radiated noise is calculated from surface vibrations using the Rayleigh equation. Radiated sound power and radiation efficiency are calculated and printed in tabular form. Sound intensities on the radiating surface are calculated and written to the .SFE file. Radiated sound power is calculated and plotted from the intensities stored in the SFE file using the Ricardo program VIBPLOT1. Sound pressures are calculated for individual points in free space or for grids of points (Spheres, Boxes, Planes etc.) defined in a separate SFE file. If a grid of points is defined then the sound pressures are written to the separate SFE file. The Rayleigh equation is applicable only to flat plates. In practice it is possible to obtain reasonable results from surfaces that are generally flat such as the bare sides of an engine block. The Sigma1 method calculates the noise produced by each node acting as a piston with no interaction with any other node. It is a very quick but inaccurate method. This calculation is automatically performed by both Rayleigh methods in order to calculate the radiation efficiency. Two solutions to the Rayleigh equation are implemented. The Rayleigh method is a simplified method in which each node is treated as a piston in a baffle assuming that the vibration is the same over the surface of the piston. This is normally acceptable for models where the mesh density is sufficient to give a good representation of the mode shapes and the wavelength is large compared to the element size. The Helmholtz method solves the same Rayleigh equation but to a greater accuracy (and therefore takes longer). The face set defining the radiating surface is triangulated and the pressure waves are integrated across each triangle with the integration order automatically adjusted according to the size of the element, the distance from the element and the wavelength of sound at the forcing frequency. Using its minimum integration order this method takes approximately three times longer to run than the Rayleigh method.

8.2 Acoustic Equations and the Rayleigh Integral The acoustic wave equation

  ( p, t )  2

1  2 ( p, t ) c2 t 2

699

[E-18]

APPENDICES Radiated Noise Calculations defines the velocity potential  ( p, t ) which is related to the time dependent particle velocity u ( p, t ) by:

u( p, t )   ( p, t )

[E-19]

and c is the propagation velocity (p & t are the spatial & time variables). The timedependent sound pressure Q( p, t ) is given in terms of the velocity potential by

Q ( p, t )   

 ( p, t ) t

[E-20]

where  is the density of the acoustic medium. The time-dependent velocity potential  ( p, t ) can be reduced to a sum of harmonic components, each of the form:

 ( p, t )   ( p)e it

[E-21]

where  is the angular frequency and  ( p) is the frequency-dependent velocity potential. Substituting [E-21] into [E-18] reduces it to the Helmholtz (reduced wave) equation:

  ( p)  k  ( p)  0 2

where k 2 

2

Fourier form:

c2

2

[E-22]

and k is the wave number. The plate vibration can also be reduced to

u ( p, t )  ( p)  n p  ( p, t )  e it n p n p

[E-23]

where n p is the outward unit normal to the plate at p.

If the Sommerfeld radiation

condition is considered in which the waves in the far field are forced to propagate radially then:

   it lim     r   ik   0    r

[E-24]

where r is the distance from the origin. Sound pressure can be calculated from:

P( p)  i ( p)

[E-25]

and the acoustic intensity on the plate:

I ( p) 





1 Re P * ( p)v( p) 2 700

[E-26]

APPENDICES Radiated Noise Calculations where * denotes the complex conjugate. frequency is:

W   I (q)dS  

The Rayleigh integral equation velocity of the plate:

2

P

The sound power radiated at any given

*

(q) P(q)dS

Sf

[E-27]

relates the velocity potential at a point p to the normal

1 e ikr  ( p)  v(q)dS 2  r

[E-28]

The Rayleigh equation is valid for a flat plate, the plate is set in an infinite baffle and the acoustic medium extends in the positive normal direction. The point p can be either in the acoustic medium, on the plate or on the baffle. The points q are points on the vibrating surface which when integrated represent the whole surface.

8.2.1

The Sigma1 Method

The noise radiated (ignoring the interference across the vibrating surface) is:

W  1 2 c  v * (q)v(q)dS 

[E-29]

which is simply calculated by summing the absolute value of the normal velocities multiplied by the local area over which they act.

8.2.2

The Rayleigh Method

Each node on the surface is treated as a noise source. This simplifies the Rayleigh equation to

 ( p) 

1 2

 area( j ) j

ikr

e j vj rj

[E-30]

The velocity potentials are then used to calculate the surface intensities using equations [E-25] and [E-26]and thence the total radiated power from equation [E-27]. Sound pressures at points in the acoustic medium are calculated in a similar way by evaluating the velocity potentials relative to that point and then using equation [E-25].

701

APPENDICES Radiated Noise Calculations 8.2.3

The Helmholtz Method

The normal velocity is integrated numerically over the surface of each triangular element making up the surface: n

v(q )   v( p j )  j (q)

[E-31]

j 1

where  are the shape function terms for numerical integration. Substituting [E-30] into [E-28] gives:

e ikr  v( p j )  j (q)dSq  j v j hi 2r j 

 ( p)  

[E-32]

Where

e ikr  j (q)dS q , v j  v( p j ) 2r 

hj  

[E-33]

h j is evaluated by Gaussian quadrature, however the integral is singular when q  p  r  0 . To overcome this: 1

 2r ( p , q)  (q)dS i



i

q

1 1 (  i (q)  1)dS q   dS q 2  r ( p , q ) 2  r ( p , q ) i i  



[E-34]

The first part of this equation is now regular. The second part can be integrated analytically. The following example integrates over the surface of a triangular element with vertices a, b and c and angles at those vertices of A, B and C. The integral is evaluated for vertex a: A R ( )

1 1  2r dS  2 0 c sin( B) 1  2r dS  2

A B

 B

 0

A

1 r dr d   R( ) d r 0 A B

c sin( B)   1  cos( )  d   ln  sin( ) 2   1  cos( )  A

[E-35]

The velocity potentials are calculated by simply summing the v j h j components. Surface intensities are then calculated from equations [E-25] and [E-26] and thence the total radiated power from equation [E-28]. Sound pressures at points in the acoustic medium are calculated in a similar way by evaluating the velocity potentials relative to that point and then using equation [E-25]. The accuracy of this numerical solution depends upon the order of integration chosen. The integration order is varied automatically depending upon:

702

APPENDICES Radiated Noise Calculations i.

the size of the element over which the integration is being performed,

ii. the wavelength of sound at the forcing frequency relative to the element size and iii. the distance of the point at which the integral is being performed from the element centroid. The number of gauss points on each triangular element can be 1, 3, 7, 12 or 25. Obviously the more Gauss points used the slower the solution will be. Generally the need to use high Gauss point numbers probably means that the original mesh density was too low. If accuracy is less/more important than speed of solution the integration order can be lowered/raised at the users request. The default values should normally give a good balance between accuracy and speed.

8.3 Radiation Efficiency The radiation efficiency is an index that expresses how effectively a radiating surface converts mechanical vibration energy into sound energy. The radiation efficiency is defined as the ratio of the output power to the input power such that



ReWo  Wi

[E-36]

where conventionally the input power Wi is defined as that calculated using the Sigma1 Method given by equation [E-29]. It is important to note that this convention can result in efficiencies that are larger than one.

8.4 Multiple Face Sets The Rayleigh integration is only valid for flat plates. In order to use the method on real components some simplification must be made. Face sets should be defined for each generally flat radiating surface under consideration, e.g. each side of a cylinder block. The area-weighted normal to each face set is calculated and the nodes of each element face are projected onto the normal plane for that face. The velocities are also calculated in the direction normal to each face. A separate Rayleigh calculation can now proceed for each face. The total radiated sound power contribution from a number of radiating surfaces can be calculated and plotted using the Ricardo program VIBPLOT1 Sound pressures at external points are the complex sum of the pressures contributed by each of the forward facing surfaces. Surface intensities, at nodes that are common to two (or more) faces, are summed after weighting them according to the true (not projected) surface area associated with that node for each face.

703

APPENDICES Radiated Noise Calculations

1

Ricardo Report DP00/1129 ‘VIBPLOT Revision 2.0 Noise and Vibration Plotting Utility Program’.

2

A D Pierce, McGraw-Hill, 1981, ‘Acoustics: An Introduction to its Physical Principles and Applications’

704

APPENDICES Modal Frequency Response Calculations

9 Modal Frequency Response Calculations 9.1 Overview Frequency response analysis is a method used to compute the response of a structure due to steady-state oscillatory excitation. In direct frequency response analysis, the structural response is computed at discrete excitation frequencies by solving a set of coupled matrix equations using complex algebra. Modal frequency response analysis on the other hand, is an alternative method that uses the mode shapes of the structure to reduce the size, uncouple the equations of motion (when modal or no damping is used) and make the numerical solution more efficient.

9.2 Calculation of Modal Contributions and Vibration Response The basic equation for dynamics is:

Mx  Cx  Kx  F

[E-37]

which can be transformed into the frequency domain as:

    M    i 2 m

T

m

 T C     T K    f    T F  f 

[E-38]

where M  , C  and K  are the mass, damping and stiffness matrices and  m and  f

are the mode and forcing frequencies. Since   are the Eigen vectors, by definition

 T M  

 T K   are diagonal matrices. The mode shapes T T normalised to make   M   the unit matrix and   K   is simply   . 9.2.1

and

are mass

Modal Damping

Damping can be defined empirically from experiments on similar structures and can be T defined in terms of modal damping, that is, the diagonal terms of   C   . This makes the solution of the equation [E-38] trivial such that  m , the contribution of each mode to the total vibration at a given forcing frequency, is given by

705

APPENDICES Modal Frequency Response Calculations

a   m2   2force b  2  m f c  a2  b2 n a b  mod e    ,   i  Fi i 1  b c  where

 i Fi

[E-39]

is the damping coefficient for the mode or frequency in question is the complex eigen vector of the mode in question at a forced node is the complex force at that node for that frequency

The alpha values are saved to the SFE file or to a .ALPHA file. The complex displacement vector  f at each node of the structure for each forcing frequency is calculated from the modal contributions using [E-39] and the mode shapes such that

f 



mod e



mod es

[E-40]

where  are the real or complex Eigen vectors.

9.2.2

Discrete Dampers

If discrete damper elements are defined in the model it is no longer possible to T diagonalise the damping matrix and   C   must be calculated explicitly. The

number of dampers will generally be small meaning that C  is a very sparse matrix.

However since the   C   matrix is a fully populated square matrix of order number of modes the trivial solution used for modal damping is no longer possible. Instead a set of simultaneous equations has to be solved for equation [E-38] at each forcing frequency. This calculation requires all the mode shapes to be held in memory at once such that for larger models a substantial amount of RAM may be required. T

706

APPENDICES Applying Loads Predicted by VALDYN to the Cylinder Block and Crankshaft Models

Applying Loads Predicted by VALDYN to the Cylinder Block and Crankshaft Models 10

10.1

Overview

The Ricardo Program VALDYN is used for analysing the dynamic motion of valve train systems. Individual parts of the valve train are represented by modelling elements that can be linked together to form the whole valve train model. The flexibility of the program enables the user to generate models varying from simple single valve-line models to highly complex complete system models. Although VALDYN is intended to model valve train systems a model can be extended to include ancillary drives using gear and chain models and also a representation of the crank train to include its torsional behaviour. VALDYN has the option of writing predicted force histories or profiles to a Ricardo standard data file (sdf.) This appendix describes how these force histories can be read by ENGDYN and applied to either the cylinder block or crank train as excitation forces which are then included as part of the simulation in ENGDYN. This provides a loose coupling between the VALDYN and ENGDYN and is intended primarily as a means of applying valve train loads to the finite element model of the cylinder block.

10.2

Writing Force Profile Data From VALDYN

To apply excitation forces to an ENGDYN model, a VALDYN model is required of the complete valve train of the engine, which may or may not include the front-end drive. To apply the complete set of excitation forces predicted by VALDYN, the following force excitations need to be considered; valve seat, valve spring, rocker pivot, swinging follower pivot, cam support, sprocket and guide supports and tensioner force. These forces are represented in a VALDYN model as either LSTIFF or SSTIFF elements. The properties panels for each of these elements and the properties panel for the GENSPRING element (which generates LSTIFF and SSTIFF elements) have a SDFOUTPUT button that displays the SDFOUTPUT Properties Panel shown in Figure E-22.

707


Figure E-22 SDFOUTPUT Properties Panel This panel allows force profile data to be written to the sdf file for a given selected element. Each force profile is written to an array called g:profile:force with a corresponding ordinate array called g:profile:crankAngle. In order to distinguish between different force profile arrays tags are added to the name such that the name is of the form g:profile1D:force:TYP_valveSeat:CYL_1:VAL_1:DIR_1:SP_1000 Each tag is of the form :*_*. The default tags are defined in the file valdyn.tags, shown in Figure J-2. This is found in $HOME/valdyn/*.*/Config and is supplied as part of the VALDYN installation. # # # # # # # # # # #

This file contains the default tags for Valdyn The format is: The must be 'menu', 'menuitem' or 'field' (a 'menuitem' must follow a 'menu') The should be known by the solver The can be edited to provide language support The must be 'number' or 'string' (defaults to 'string')

menu menuitem menuitem menuitem

TYP none valveSeat valveSpring

"Type" "None" "Valve Seat" "Valve Spring"

708

APPENDICES Applying Loads Predicted by VALDYN to the Cylinder Block and Crankshaft Models menuitem menuitem menuitem menuitem menuitem menuitem field field field field

rockerSupport followerSupport camSupport sprocketSupport guideSupport tensionerSupport CYL VAL DIR ID

"Rocker Pivot" "Follower Pivot" "Cam Support" "Sprocket Support" "Guide Support" "Tensioner" "Cylinder" "Valve" "Direction" "ID"

number string number string

Figure E-23 Default ‘valdyn.tags’ File The tags CYL, VAL, DIR are used to differentiate between cylinders, valves and direction respectively. The ID tag is a generic tag used to differentiate items of the same type such as guide supports. The valdyn.tags file allows the user to add additional menu items to the type list and to change the tag name or gui label. This allows the sdf output to be used generally as a means of transferring data to other programs. For the specific interface with ENGDYN, ENGDYN expects the default tag names to be used and will ignore any additional names that are added. Also the following must be satisfied: 

Each force profile must have a type and direction tag defined.



The direction tag must be a number



The cylinder tag, if defined, must be a number



Each force profile corresponding to a ‘Valve Spring’ or ‘Valve Seat’ must also have a cylinder and valve tag.



Each force profile corresponding to a ‘Rocker Pivot’, ‘Follower Pivot’ or ‘Cam Support’ must also have a cylinder tag

If the first two criteria are not satisfied for any given force profile then that force profile will be ignored when read by ENGDYN. If any of the other criteria are not meet then warnings will be displayed when the data is read. The speed tag SP is added by VALDYN when the force profile is written on completion of the analysis. VALDYN is a one-dimensional simulation program in that it has no concept of threedimensional space. It therefore has no concept of the actual direction of the force when apply to the engine. Hence in order to apply the forces correctly to the ENGDYN model it is necessary to have an obligatory direction tag for each force profile. The actual direction vector for each direction is defined in ENGDYN. To minimise the number of direction vectors defined using ENGDYN it is recommended that each unique direction has a number rather than a unique direction number for each force profile. For example, all the inlet valves of a 2-valve engine are likely to be inclined at the same angle and therefore the force profiles for each inlet valve seat would use the same direction number.

709


10.3

Applying Force Profile Data to the ENGDYN Model

Each force profile is applied either to a node or elemental face set of the reduced model of the crankshaft or cylinder block. Any number of force profiles can be applied to any one set. Typically this allows forces that have been calculated in VALDYN in two orthogonal directions at a single node to be combined so as to apply the vector sum of the forces to the ENGDYN model. A naming convention for the sets has been adopted to allow the force profile data to be automatically matched to sets defined by the user. This convention uses the value of the TYP tag as the base name of the set with the CYL, VAL and ID tags added to the end of the name. The set name is case sensitive. In the example above, the force profile g:profile1D:force:TYP_valveSeat:CYL_1:VAL_1:DIR_1:SP_2000 would be matched with a set called valveSeat:CYL_1:VAL_1 This set name would be stored with either a NODE_SET_ or FACE_SET_ prefix depending on whether the set was a node or face set. The order of the tags is not important so a set called valveSeat:VAL_1:CYL_1 is also valid. Firstly ENGDYN attempts to match each force profile with a node set. If no match is found the program then attempts to match with a face set. If this is unsuccessful a warning is displayed. It is expected that any node set will only contain a single node.

10.4

Applying Force Profile Data to the Cylinder Block Model

Typically force profile data predicted by VALDYN will be forces due to the valve train and these will be applied to sets on the model of the cylinder head included as part of the cylinder block model. However, force profile data may also be written by VALDYN for a chain drive and applied to the front of the cylinder block for example. ENGDYN applies the force profile data to either node or face sets where the expected geometry of the latter is dependant on the TYP tag and therefore the name of the set. The expected geometry for each face set is summarised in Table B-4.

710

APPENDICES Applying Loads Predicted by VALDYN to the Cylinder Block and Crankshaft Models Type Set geometry valveSeat Frustum of a cone valveSpring Plane annulus rockerSupport Cylinder followerSupport Cylinder camSupport Cylinder sprocketSupport Cylinder guideSupport Cylinder tensionerSupport Cylinder Table E-9 Expected Geometry of Each Face Set

The loads are applied to the face sets for the valve seat and spring as uniform pressure loads, whilst loads applied to the cylinders for each of the supports are applied as a distributed pressure load that varies as a parabolic function along the journal and around its circumference with an extent of 180o. For those supports that are not a cylindrical surface a node set should be used. If a node set is used to define a valve seat a face set must also be defined. This set is used to calculate the valve area so that the correct load is applied to the rest of the gas face of the cylinder head. If node sets are used then the Ricardo program FEPRE can be used to constrain a face set so that the displacement of a node known as the control node is the average displacement of the nodes of that set. This control node can be as the loaded node. The advantage of this method is that stiffness at this node is an average stiffness rather than the local stiffness of an element. At present only the torsional behaviour of the camshaft can be modelled in VALDYN. The reactions at each bearing support are not calculated. A cam support represents the stiffness at the position of the cam lobe. The force profile calculated at a cam support is therefore the force applied to the camshaft at that point on the camshaft. It is therefore necessary to include a three-dimensional model of the camshaft in the finite element model of the cylinder block so that the loads can be applied along the camshaft due to each valve train. It is suggested that the camshaft model is constrained to the camshaft bearings using sliding joins. For a 3 cylinder engine with 2 valves per cylinder and a finger follower the sdf output defined by the user and written by VALDYN for a single speed 2000 rev/min may be as follows: g:profile1D:force:TYP_valveSeat:CYL_1:VAL_I1:DIR_1:SP_2000 g:profile1D:force:TYP_valveSpring:CYL_1:VAL_I1:DIR_2:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_1:VAL_I1:DIR_3:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_1:VAL_I1:DIR_4:SP_2000 g:profile1D:force:TYP_camSupport:CYL_1:VAL_I1:DIR_5:SP_2000 g:profile1D:force:TYP_camSupport:CYL_1:VAL_I1:DIR_6:SP_2000

711

APPENDICES Applying Loads Predicted by VALDYN to the Cylinder Block and Crankshaft Models g:profile1D:force:TYP_valveSeat:CYL_1:VAL_E1:DIR_7:SP_2000 g:profile1D:force:TYP_valveSpring:CYL_1:VAL_E1:DIR_8:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_1:VAL_E1:DIR_9:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_1:VAL_E1:DIR_10:SP_2000 g:profile1D:force:TYP_camSupport:CYL_1:VAL_E1:DIR_11:SP_2000 g:profile1D:force:TYP_camSupport:CYL_1:VAL_E1:DIR_12:SP_2000 g:profile1D:force:TYP_valveSeat:CYL_2:VAL_I1:DIR_1:SP_2000 g:profile1D:force:TYP_valveSpring:CYL_2:VAL_I1:DIR_2:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_2:VAL_I1:DIR_3:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_2:VAL_I1:DIR_4:SP_2000 g:profile1D:force:TYP_camSupport:CYL_2:VAL_I1:DIR_5:SP_2000 g:profile1D:force:TYP_camSupport:CYL_2:VAL_I1:DIR_6:SP_2000 g:profile1D:force:TYP_valveSeat:CYL_2:VAL_E1:DIR_7:SP_2000 g:profile1D:force:TYP_valveSpring:CYL_2:VAL_E1:DIR_8:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_2:VAL_E1:DIR_9:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_2:VAL_E1:DIR_10:SP_2000 g:profile1D:force:TYP_camSupport:CYL_2:VAL_E1:DIR_11:SP_2000 g:profile1D:force:TYP_camSupport:CYL_2:VAL_E1:DIR_12:SP_2000 g:profile1D:force:TYP_valveSeat:CYL_3:VAL_I1:DIR_1:SP_2000 g:profile1D:force:TYP_valveSpring:CYL_3:VAL_I1:DIR_2:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_3:VAL_I1:DIR_3:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_3:VAL_I1:DIR_4:SP_2000 g:profile1D:force:TYP_camSupport:CYL_3:VAL_I1:DIR_5:SP_2000 g:profile1D:force:TYP_camSupport:CYL_3:VAL_I1:DIR_6:SP_2000 g:profile1D:force:TYP_valveSeat:CYL_3:VAL_E1:DIR_7:SP_2000 g:profile1D:force:TYP_valveSpring:CYL_3:VAL_E1:DIR_8:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_3:VAL_E1:DIR_9:SP_2000 g:profile1D:force:TYP_followerSupport:CYL_3:VAL_E1:DIR_10:SP_2000 g:profile1D:force:TYP_camSupport:CYL_3:VAL_E1:DIR_11:SP_2000 g:profile1D:force:TYP_camSupport:CYL_3:VAL_E1:DIR_12:SP_2000 These forces would then be written to sets called FACE_SET_valveSeat:CYL_1:VAL_I1 FACE_SET_valveSpring:CYL_1:VAL_I1 NODE_SET_followerSupport:CYL_1:VAL_I1 FACE_SET_camSupport:CYL_1:VAL_I1 FACE_SET_valveSeat:CYL_1:VAL_E1 FACE_SET_valveSpring:CYL_1:VAL_E1 NODE_SET_followerSupport:CYL_1:VAL_E1 FACE_SET_camSupport:CYL_1:VAL_E1 FACE_SET_valveSeat:CYL_2:VAL_I1 FACE_SET_valveSpring:CYL_2:VAL_I1 NODE_SET_followerSupport:CYL_2:VAL_I1 712

APPENDICES Applying Loads Predicted by VALDYN to the Cylinder Block and Crankshaft Models FACE_SET_camSupport:CYL_2:VAL_I1 FACE_SET_valveSeat:CYL_2:VAL_E1 FACE_SET_valveSpring:CYL_2:VAL_E1 NODE_SET_followerSupport:CYL_2:VAL_E1 FACE_SET_camSupport:CYL_2:VAL_E1 FACE_SET_valveSeat:CYL_3:VAL_I1 FACE_SET_valveSpring:CYL_3:VAL_I1 NODE_SET_followerSupport:CYL_3:VAL_I1 FACE_SET_camSupport:CYL_3:VAL_I1 FACE_SET_valveSeat:CYL_3:VAL_E1 FACE_SET_valveSpring:CYL_3:VAL_E1 NODE_SET_followerSupport:CYL_3:VAL_E1 FACE_SET_camSupport:CYL_3:VAL_E1

713

APPENDICES ASCII data file format read by Sdf_Read_Ascii()

11 ASCII data file format read by Sdf_Read_Ascii() 11.1 Introduction The following notes describe the "General" ASCII data file format used by a number of Ricardo programs. The format allows the data contained in a file to be fully described within the file. 11.2 Format 11.2.1 General 1) Blank lines are ignored. 2) Comments are introduced by "#" or "!" and extend to the end of the line. Comments are ignored. A line containing only a comment is treated as a blank line. 3) Data can be in rows, columns or blocks. A set of rows or columns or a block of data is terminated by a non-blank, non-comment line which does NOT consist solely of numbers, or by the end of the file. 4) Individual data are separated by white space (space or tab) and/or a single comma. Two consecutive commas are taken to enclose a zero. 5) Except for their function as data terminators, lines which do not consist solely of numbers have no effect unless they begin with "COLUMN", "ROW", "BLOCK" or "TAGS". 6) Keywords (COLUMN, ROW etc.) are case-insensitive and can be shortened to three characters. 7) COLUMN, ROW and BLOCK lines are referred to as "definition" lines and have the format: COLUMN[=n] [name][[units]] [DESC=desc] defines data in a column. Multiple columns are defined by consecutive COLUMN definitions. [name][[units]] [DESC=desc] defines data in a single row. BLOCK[=n] [name][[units]] [DESC=desc] defines data spread over a number of rows. Where [...] indicates optional input. If no name is given, a "dummy" name of the form "_$#" is generated, where # is the number of dummy names generated so far (including this one). Data units must come immediately after the name and must be enclosed in brackets – e.g. [MPa]. DESC is an oprional description string which will be passed to the program. ROW[=n]

715

APPENDICES ASCII data file format read by Sdf_Read_Ascii() Definition lines must come before the data to which they refer, and when data are encountered they are expected to satisfy all previous unsatisfied definitions. The order of definitions must correspond to the order of the data. The number n is ignored and is provided simply as a means of annotation to make the file easier to read (in fact n does not have to be a number). 11.2.2 Data as columns A COLUMN consists of data in the same ordinal position on consecutive lines. If a line has more items than the first line, additional items will be ignored. If a line has fewer items than the first line, missing items will be interpreted as zeros. In either of these cases a warning will be issued. Examples: a) 2D data as two columns: COL=1 Time[s] required COL=2 Temperature[degC] 0 80.0 10 91.5 20 99.0 30 105.0 50 112.5 70 117.0 100 120.0 b) 3D data as three columns column Frequency[kHz] column Temperature[degC] column Stiffness[N/mm] 1 50 12 1 70 10 1 90 8 1 120 6 3 50 14 3 70 12 3 90 10 3 120 8 5 50 16 5 70 14 5 90 12 5 120 10 7 50 18 7 70 16 7 90 14 7 120 12

716

#Column numbers are not #but add clarity.

APPENDICES ASCII data file format read by Sdf_Read_Ascii() 11.2.3 Data as rows A ROW consists of a single line of data, unless it is the only definition which remains to be satisfied, in which case it can extend over multiple lines and is equivalent to a BLOCK. Example: c) 2D data as two rows: row Time[s] row Temperature[degC] 0 10 20 30 50 70 100 80.0 91.5 99.0 105.0 112.5 117.0 120.0 11.2.4 Data as blocks A BLOCK is 1D data extending over more than one line. Blocks must be separated by a non-blank line containing something other than numbers. Normally this would be another block definition. Example: d) 2D data as two blocks block Time[s] 0 10 20 30 50 70 100 block Temperature[degC] 80.0 91.5 99.0 105.0 112.5 117.0 120.0 or block Time[s] block Temperature[degC] 0 10 20 30 50 70 100 --80.0 91.5 99.0 105.0 112.5 117.0 120.0 11.2.5 3D data as a matrix In the special case where there are more items on the line than there are column definitions to be satisfied, the last column definition is taken to indicate a 2D array with the first dimension corresponding to columns and the second to rows. This is shown in example e) below. In this case the description is appended with "dims(row_name,col_name)" where row_name is the last named row and col_name is the last named column. Thus in example e), the description of "Stiffness" will be "dims(Frequency,Temperature)", indicating that the first dimension corresponds to frequency and the second to temperature.

717

APPENDICES ASCII data file format read by Sdf_Read_Ascii() Example: e) 3D data as matrix (compare with example b)): row Frequency[kHz] column=1 Temperature[degC] ! Column numbers are ignored but column=2> Stiffness[N/m] ! make file more readable. 1 3 5 7 50 12 14 16 18 70 10 12 14 16 90 8 10 12 14 120 6 8 10 12 11.2.6 Anonymous data If data are encountered before any definition lines, it is assumed that the data are in columns and that the number of columns is the number of data on the first line. In this case the columns are given dummy names of the form "_$COLUMN#" where # is the number of dummy column names generated so far (including this one). Example: f)

2D data as two columns (the program sees these as "_$COLUMN1" and "_$COLUMN2"): 0 80.0 10 91.5 20 99.0 30 105.0 50 112.5 70 117.0 100 120.0

11.2.7 Use of tags TAGS lines have the format: TAGS = tagstring Where tagstring contains one or more tags. A tag is part of a name has the form :tagname_value i.e. a tag name must be preceded by a colon and followed by an underscore, which separates it from the tag value. The tag value consists of characters between the underscore an the next colon or the end of the name. The tag name and value can contain any valid variable name characters except colon and, in the case of the name, underscore. Multiple tags should be separated by colons - e.g. "CYL_1:DIR_X". The tags will be appended to each name given. The TAGS definition which is effective for given data is the last one before the start of the data. This means that tags can be changed between sets of data. This mechanism can be more convenient than attaching tags directly to names. Examples:

718

APPENDICES ASCII data file format read by Sdf_Read_Ascii() g) 2D data as two columns using TAGS: TAGS = CYL_1 COL Temperature[degC] COL Time[s] 0 80.0 10 91.5 20 99.0 30 105.0 50 112.5 70 117.0 100 120.0 TAGS = CYL_2 COL Temperature[degC] COL Time[s] 0 80.0 10 92.0 20 100.0 30 107.0 50 115.0 70 120.0 100 130.0 The names seen by the program are Temperature:CYL_1, Time:CYL_1, Temperature:CYL_2, Time:CYL_2. Compare this with: h) 2D data as two columns without TAGS: COL Temperature:CYL_1[degC] COL Time:CYL_1[s] 0 80.0 10 91.5 20 99.0 30 105.0 50 112.5 70 117.0 100 120.0 COL Temperature:CYL_2[degC] COL Time:CYL_2[s] 0 80.0 10 92.0 20 100.0 30 107.0 50 115.0 70 120.0 100 130.0

719

APPENDICES MATUTIL

12 MATUTIL 12.1 Overview Ricardo use the xml schema called MatML published at https://www.matml.org to store material property data of solids used by FEARCE and the mechanical suite of products, ENGDYN, PISDYN, RINGPAK and VALDYN. This includes all necessary properties for performing structural and mechanical analysis using the Ricardo products. In future, this will be extended to support of fluid properties, including lubricants and fuels. The data using this schema is stored in a Ricardo material file with a .rmt suffix, in which each file contains the material properties of a single material (even though the xml schema supports multiple materials in a single file). Currently .rmt files can be created using the FEARCE GUI or from a .MAT or .SFE file. The .MAT file is the native material file used by FEARCE and is simple ASCII file, whilst a .SFE file is the binary standard data file for storing all FE data used by the Ricardo software products. A .rmt file can be created from a .MAT using the FEARCE utility matoxml, whilst data from an .SFE file can be written to a .rmt file using matutil. Matutil is a general command line utility program for editing, printing and calculating material property data. This tool is supplied primarily to allow users to calculate the Greenwood-Williamson-Tripp (GWT) parameters from surface roughness data which are used by the mechanical suite of products for lubricated contact surfaces in each of those products. This functionality is covered in the following sections. Matutil allows the user to print the contents of an .rmt or .SFE file; import data and to calculate properties from other data in the .rmt file or from an ASCII file. Data can be imported from another .rmt file, an ASCII file or .SFE file. It is also possible to convert the units of a given property.

12.2 Using matutil The basic structure of the command is: matutil [-] filename where the available commands are:  print  import  convert  calculate Detailed help for each command can be invoked by typing: matutil -help or, to show help for all commands, type: matutil -help all To list all supported property names and their dimensions, type: matutil -help properties 720

APPENDICES MATUTIL

12.2.1 Printing values Values stored in .rmt or .SFE files can be printed typing: matutil print [-property ] [-material ] . If no property is defined then names of all properties stored in the file are listed. Using the -property option with the desired property name will print the values of that property. In .SFE files more than one material can be stored. A particular material can be selected by using the -material option.

12.2.2 Importing data into an .rmt file Properties can be imported into an .rmt file from different sources another .rmt file, a .SFE file or from an ASCII file. To list all the currently supported properties that can be imported from a text file simply type matutil –help import -property The general command for importing property data is matutil import -file [-ascii ] [-property ][-column=icol|value][-row=irow|value][-replace] [material ] outputname.rmt where the outputname.rmt will be created if it doesn’t exist already, otherwise the properties will be added to the existing file of that name. If no property is defined then all the contents of the file are imported. For example matutil import -file myMaterial1.rmt mymaterial2.rmt would simply copy all property data from myMaterial1.rmt to myMaterial2.rmt The -replace option will overwrite any existing property already in the .rmt file. For example matutil import -file myMaterial1.rmt -property Density -replace mymaterial2.rmt would import the property Density from myMaterial1.rmt and write it to myMaterial2.rmt whilst overwriting any existing property called Density. .SFE files can store more than one material so the -material option must be used to select an individual material to be imported into the .rmt file. For example matutil import -file myFEModel.SFE –material SGIRON600 iron.rmt would import the material SGIRON600 into iron.rmt from the SFE file myFEModel.SFE 12.2.3 Importing Data from an ASCII File Data imported from ASCII file can be in one of two formats, column or matrix selected using the -ascii option. The data in the file can be either comma or space delimited. If the data are in matrix format then the first line must start with comma so as to fit the y

721

APPENDICES MATUTIL values to the correct columns. If no units are defined in the file then the imported data is assumed to be in SI units for that property. 12.2.3.1 Column format By default it is assumed that the columns are ordered as x, y, h(x,y)

When data exists only for one y coordinate value the second column can be left out.

Data can be stored in a different order by labelling each column at the top of the file. This same mechanism can be used to define the units of the data if it is not in SI units. column=1 height [um] column=2 y[mm] column=3 x[mm]

The –row option can be used to import the property value at an individual row by supplying the row index or the x-value of that row. For example -row=3 or -row=0.5 are equally valid. For example matutil import -file profile.dat -ascii column –property “Surface profile height” myMaterial.rmt

722

APPENDICES MATUTIL will import a surface profile stored in column format with 2 columns in a file called profile.dat, into a file called myMaterial.rmt. 12.2.3.2 Matrix format It is assumed that the matrix is in the following form

For the case when the data is not in SI units file as follows: row=1 column=1 column=2>

then the units are specified at the top of the y[mm] x[mm] height[um]

The –row or –column flags can be used to import an individual row or column by supplying the index or value. For example matutil import -file profile.dat -ascii matrix –row=5 –property “Surface profile height” myMaterial.rmt will import the 5th row of a surface profile stored in matrix format in a file called profile.dat into a file called myMaterial.rmt.

12.2.4 Converting units The data of a given property in an .rmt file can be converted to different units in the file using the -unit option. matutil convert -property -unit filename.rmt For example matutil convert -property Density –unit “kg/mm^3” mymaterial2.rmt

723

APPENDICES MATUTIL will change the value of Density to have units of kg/mm^3 in myMaterial2.rmt

12.2.5 Calculating property values Matutil supports the derivation of various property data including parameters used for the boundary lubrication models used in ENGDYN, PISDYN, RINGPAK and VALDYN, together with fatigue property data such as the Smith-Watson-Topper(SWT) and SN curves used by FEARCE. A number of data can be derived from supplied measured surface profile data stored in an ASCII file as defined in Section 12.2.3. These include  Asperity parameters used for the Greenwood-Williamson-Tripp (GWT) boundary lubrication model  Asperity ratio (γ) used for the Patir and Cheng Average Flow Model, and  Honing parameters used by RINGPAK The general command for importing property data is matutil calculate -file input_filename.rmt -property propertyID [method method] [-ascii format][-material material][-replace][-fft] [-density_factor=value] [-column=icol|value] [-row=irow|value] where propertyID can be one of GWT, SWT, SN, Asperity_Ratio, Honing or Brown_Miller. If no output file name is supplied then the result is simply printed to the screen and not written to the any file 12.2.5.1 Deriving Greenwood-Williamson-Tripp (GWT) parameters For the GWT property method is required to define how these parameters are derived and can be one of Abbott_curve, Abbott_parameters or Profile. Each of these methods are described in the Theory Section 12.3.1. As well as the asperities a measured surface profile may also be composed of the macro-shape of the component, such as the profile of a piston skirt. For the calculation of the GWT parameters this macro-shape if it exists must be removed using the -fft option. This option performing a Fast Fourier Transform on the profile, removing the 1 st order and converting the remaining harmonics back to the physical domain. The multiplier factor of asperity density α expresses the ratio of real peaks count and peaks count calculated from 1D profilometer data as defined in Section 12.3.1.1. This factor typically lies within the interval 1.0 to 1.8 and a default value of 1.2 will be used unless -density_factor is used 12.2.5.2

GWT from Abbott Curve

724

APPENDICES MATUTIL In this case the parameters are calculated from an Abbott curve as shown in Figure E-28 and described in Section 12.2.5.3.

Figure E-24 Abbott Curve For example matutil calculate -file material.rmt -property GWT -method Abbott_curve newfile.rmt will calculate the GWT parameters from an Abbott curve stored in material.rmt and write the parameters to a file called newfile.rmt. 12.2.5.3 GWT from Abbott Curve Parameters (ISO 13.565) In this case the parameters are calculated from the Abbott curve parameters Reduced peak height (Rpk), Core height (Rk)at Mr1 and Reduced valley height (Rvk) at Mr2 as shown in Figure E-28 and described in Section 12.2.5.3.

For example matutil calculate -file material.rmt -property GWT –method Abbott_parameters will calculate the GWT parameters from the Abbott curve parameters stored in material.rmt and write the data to that file. 12.2.5.4 GWT from Surface Profile In this case the parameters are calculated from a surface profile that can be one or two dimensional using the method described in Section 12.3.1.1.1. By default, the calculation of GWT parameters is performed on the whole set of data, line by line, and in the case of a piston this would be in the axial direction along the y-axis. Alternatively the calculation can be performed on a given row or column using the -column or -row flags. If the data is loaded from an ASCII file then position in the matrix can be specified by the row or column number or by its coordinate value. If the surface profile data is

725

APPENDICES MATUTIL already in the .rmt file then the coordinate value must be specified to select the given row or column If an outpur names is supplied then two additional files are written so that the probability distribution function can be plotted using RPlot. The data is stored in _distribution.asc and is plotted using _distribution.rp. For example matutil calculate -file profile.dat -ascii column -property GWT method Profile -column 1 will import a surface profile from a file called profile.dat in columnar format with 2 columns and write this data to a file called myMaterial.rmt whilst calculating the GWT parameters using the first column and print these to the screen.

726

APPENDICES MATUTIL

12.3 Theory 12.3.1 Derivation of Surface Property Data 12.3.1.1

Greenwood & Tripp (GWT) Parameters

Figure E-25 Surface Roughness Surface roughness can be defined by the parameters shown in Figure E-25 where Mean Summit Height, Zs The mean asperity summit height of all the counted summits is given by

Zs 

1 ns  zi ns i 1 (1)

where ns = Total number of counted summits Standard Deviation Summit Height, σs The standard deviation of asperity summit height of all counted summits about the mean asperity summit height given by

s 

1 ns ( zi  Zs) 2  ns i 1 (2)

727

APPENDICES MATUTIL The paper defining the Greenwood & Tripp model [1] considers dry asperity contact only and conveniently defines the mean height as zero. When used with a lubrication model this means an offset is added to the contact friction integrals so the oil film can be defined correctly relative to the mean summit height and the asperity contact is defined correctly from the mean summit height and the summit height standard deviation. The contact integral of the Greenwood & Tripp, verified by [2] and [3], is given by

1 Fn (h)  2

n



s2

h  Zs   2  s    e ds    ( h  Zs ) /   (3)

Density, η The density of the asperity peaks is calculated by calculating the asperity summits lying above thethe surface mean line divided by the summit length, squared 2

 Number of peaks   ns           L  Sample length 

2

(4) where α is the ratio of the peaks count and peaks count calculated from 1D profilometer data and typically lies within the interval 1.0 to 1.8. Authors have different recommendations for α. Tomanik [3] recommends 1.0, Nayak [5] 1.2 and [6] 1.8. The default value assumed by matutil is 1.2. Radius of Curvature, β The radius of curvature can be calculated from the average value of that radius calculated at each summit height using a finite difference approximation 1

 d 2z  z  2.zi  zi 1    2   ( xi )  i 1 dx 2  dx  (5) Each of the parameters Zs, σs, η and β can be derived by matutil using one of three methods. 12.3.1.1.1 Surface profile

728

APPENDICES MATUTIL In this case, each of the asperity parameters are obtained using the following approach, proposed by Tomanik [3]. The Mean Summit Height of all the points

is found by first finding all the local peaks above the mean

as shown by the red points of Figure E-25 given by

This then gives the summit height distribution from which can be derived. The Standard Deviation Summit Height σs and Density η are calculated using equations (2) and (4) respectively. The Radius of Curvature, β is found by using the method proposed by Tomanik [3]. Using 2 parabola as shown by Figure E-26

Figure E-26 Parabolic Peak

12.3.1.1.2 Abbott curve

729

APPENDICES MATUTIL The Abbott or Bearing Area curve represents a surface material distribution and gives a quick, qualitative overview of the asperity properties of the surface. The Abbott curve is defined as a fraction of nominal area lying within the surface contour at an elevation y and total surface. This can be seen in the Figure above. On the left side is the surface profile and on the right side is the derived Abbott curve which can be considered as a table of elevations and their fractions.

Figure E-27 Abbot Curve The R. M. S. Summit Height  can be obtained directly from the Abbott curve distribution using the whole curve such that

σ=

1  N 2    yi  Ny 2  N  1  i=1  (6)

The Mean Summit height

is then estimated from the R. M. S. Summit Height such that

Zs  0.4 .

(7)

The Abbott curve does not provide all the information to reconstruct the surface and as such the Asperity Density η is estimated to lie between

10 6 5.10 4 < η < which is σ2 σ2

derived from the following empirical equations:

0.01 < σβη < 0.05 0.01 <

σ < 0.1 β (8)

To improve the estimation Asperity Density a skewness parameter is introduced given by

S sk =

1 N.σ 3

N

  yx   y 

3

i

i=1

(9)

730

APPENDICES MATUTIL and is calculated from the whole curve. For the selected η interval the asperity density is then obtained as follows:

2   η =  2,55.10 4 + atan S sk 2,45.10 4  / σ 2 π   (10) The Radius of Curvature β is estimated from the value of the average asperity peak radius which is calculated from the asperity density and the portion of Abbott curve with the lowest slope. This portion of the curve represents the plateau of the surface, and the elevation at 40% has been chosen. 12.3.1.1.3 ISO 13.565-2 The Abbott curve parameters Reduced peak height (Rpk), Core height (Rk)at Mr1 and Reduced valley height (Rvk) at Mr2 as shown in Figure E-28. The Abbott curve is derived from these parameters so that a table of pairs of elevations and their fractions are obtained. This is then used as described as described in Section 12.3.1.1.2

Figure E-28 Abbott Curve ISO 13.565-2

12.3.1.2

Honing Parameters

731

APPENDICES MATUTIL The honing process of a cylinder bore of an engine creates a characteristic crosshatch with grooves on the surface at a given angle as shown by Figure E-29 . The grooves are considered to have symmetrical distribution in both directions. If the honing angle is too steep then this will lead to rapid oil flow and may cause cylinder and ring scuffing. On the other hand, if the honing angle is too flat then there is less oil flow resulting in a larger oil film layer which may possibly lead to higher oil consumption.

Figure E-29 A Honed Cylinder The model of oil flow through the honing grooves as implemented in RINGPAK is realized by set of triangular shaped channels where the groove cross-section is simplified to a triangle, as shown in Figure E-30. The required inputs to this model are the honing angle, the depth and width of the idealised groove and the density of the grooves.

Figure E-30 Simplification of Honed Groove

732

APPENDICES MATUTIL The honing parameters, groove depth [m], groove width [m] and groove density [1/m] can be derived directly from a surface profile measurement. A typical honed surface profile is shown in Figure E-31.

Figure E-31 Typical Honed Surface Profile The grooves are considered as valleys under the contact plane, which is derived as part of the GWT derivation. The honing triangle is then constructed from an average of the valleys where each valley is defined by its deepest point and the first two points on each side up from the valley above the contact plane. 12.3.1.3 Asperity ratio, γ Asperity ratio or Asperity Correlation Length γ, is applied in the Average Reynolds equation for isotropic surfaces and surfaces with directional patterns. The coefficient represents surface conductivity to oil flow in a particular direction. The ratio is expressed with respect to the axial direction, as shown in Figure E-32, and can be calculated from two given surface profiles measured in perpendicular direction.

Figure E-32 Different Asperity Ratios

1

The Contact of two Nominally Flat Rough Surfaces, J. Greenwood, J. Tripp, IMechE, 1970-71

733

APPENDICES MATUTIL

2

On the Modelling of Elastic Contact between Rough Surfaces, R. Jackson, I. Green, Tribology Transactions 54, 2011, STLE

3

Modelling the hydrodynamic support of cylinder bore and piston rings with laser textured surfaces, E. Tomanik, Tribology International March 2013

734

APPENDICES MATUTIL

735

Engdyn 2014.1 Manual

Recommend Documents