Jon Fernandez de Antona Beng Project Excerpt3



Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0

Jon Fernandez de Antona Beng Project Excerpt3



Uploaded by Ramadan Duraku



Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News



Documents



Sheet Music

2005 Paper IJVSMT JPombo

1

of 112

 

Design

Inerter Nonlinearities

 Search document



DEVELOPMENT OF A FULL VEHICLE DYNAMIC DYNAMIC MODEL MODEL OF A PASSENGER CAR USING ADAMS/CAR Excerpt from my BEng Project at Oxford Brookes University. Jon Fernandez de Antona

Master your semester Thiswith workScribd is licensed under the Creative Read Free Foron 30this Days Sign up to vote title Commons Attribution-NonCommercial& The New York Times  Useful  Not useful Special offer for students: Only $4.99/month.

Cancel anytime.

NoDerivs 3.0 Unported License. To view

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music

Master your semester with Scribd & The New York Times Special offer for students: Only $4.99/month.



Upload

Sign In

Read Free For 30 Days Cancel anytime.

Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document



Development Development of a full vehicle vehicle dynamic dynamic model of a passeng passenger er car using Adams/Car Adams/Car



Documents



Sheet Music

Jon Fernandez de Antona

Abstract Multibody simulation (MBS) software is a key dynamic simulation tool for engineers working in the automotive industry, in which the use of predictive methods is vital to increase i ncrease the efficiency of concurrent engineering processes. The aim of this project was to create a multibody model of a mid-size passenger car, comprising of a Macpherson front suspension and a multilink rear suspension. This model was to be valid for simulating the primary ride and handling responses of the vehicle. In order to achieve this, component-level parameters parameters were experimentally measured and template-based MSC ADAMS® products such as ADAMS/Car were used to build the multibody model. Subsequently, system- and vehicle-level tests, focusing on suspension compliances and full vehicle dynamic responses, were carried out in a MTS 329® road simulator. These tests were used to compare and validate the MBS results with reality. The correlation study showed that, although the model followed trends accurately, the magnitudes of the computational and experimental results were not always always equivalent. The effect of the model implementation implementation and validation methods on these magnitudes was discussed.


Read Free Foron 30this Days Sign up to vote title



 Not useful Cancel anytime.

Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Abstract Multibody simulation (MBS) software is a key dynamic simulation tool for engineers working in the automotive industry, in which the use of predictive methods is vital to increase i ncrease the efficiency of concurrent engineering processes. The aim of this project was to create a multibody model of a mid-size passenger car, comprising of a Macpherson front suspension and a multilink rear suspension. This model was to be valid for simulating the primary ride and handling responses of the vehicle. In order to achieve this, component-level parameters parameters were experimentally measured and template-based MSC ADAMS® products such as ADAMS/Car were used to build the multibody model. Subsequently, system- and vehicle-level tests, focusing on suspension compliances and full vehicle dynamic responses, were carried out in a MTS 329® road simulator. These tests were used to compare and validate the MBS results with reality. The correlation study showed that, although the model followed trends accurately, the magnitudes of the computational and experimental results were not always always equivalent. The effect of the model implementation implementation and validation methods on these magnitudes was discussed.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print


1

Download



Magazines



News

of 112

 

Design


 Search document






Documents



Sheet Music


Table of Contents 1.

Introduct Introduction ion .................................................................. ..................................................................................................... ...................................................... ................... 1.1.

Background Background and motivation motivation ...................................................................... ......................................................................................... ...................

1.2.

Vehicle Dynamics Dynamics and Multibody Multibody Simulation Simulation ................................................................ ................................................................

1.3.

Project Project Sponsor Sponsor .................................... ....................................................................... ....................................................................... ....................................

1.4.

Research Research goals and approach approach .................................................................... ....................................................................................... ...................

1.4.1. Project Project aim ................................................................... ...................................................................................................... ................................................ .............

1.4.2. Objectives Objectives .................................................................... ....................................................................................................... ................................................ .............

1.4.3. Methodolog Methodology y .................................................................... ........................................................................................................ ........................................... .......

1.4.4. Report outline outline ................................................................... ....................................................................................................... ........................................... ....... 2.

Literature Literature Review ..................................................................... ......................................................................................................... ........................................... ....... 2.1.

Overview...................... Overview......................................................... ...................................................................... ............................................................. ..........................

2.2.

Vehicle to be modelled................... modelled...................................................... ....................................................................... ........................................... .......

2.3.

Typical components of an automotive multibody model .............. ....... ............... ............... .............. .............. .......

2.4.

Tyre models ...................................................................... .......................................................................................................... ........................................... .......

2.5.

Suspension Suspension configuratio configurations.................... ns...................................................... ...................................................................... ....................................

2.6.

Influence of suspension isolator models on MBS .............. ....... ............... ............... .............. .............. .............. ............... ........

2.7.

Target specificatio specifications ns of a multibody multibody model .................................... .................................................................. ..............................

2.8.

Typical methodologies for implementing and validating multibody models .............. ....... ............ .....

2.8.1. Techniques for measuring component-level parameters .............. ....... ............... ............... .............. .............. ....... 2.8.2. Techniques for measuring system-level parameters .............. ....... .............. ............... ............... .............. .............. ....... 2.8.2.1.

Kinematics Kinematics and Compliance Compliance testing testing ............................................................... ...............................................................

Full vehicle vehicle dynamic testing testing ................................................................... .......................................................................... ....... Master your2.8.2.2. semester with Scribd  Read Free Foron 30this Days Sign up to vote title  2.8.3. Correlation Correlation................................................................... ...................................................................................................... ................................................ ............. & The New York Times  Useful  Not useful Cancel anytime.

2.9. Only Use $4.99/month. of MBS within the automotive automotive industry industry ................................................................. ................................................................. Special offer for students: 2.10.

Summary Summary





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document



Development of a full vehicle dynamic model of a passenger car using Adams/Car



Documents



Sheet Music


3.7.

Spring stiffness measurement ......................................................................................

3.8.

Anti Roll Bar stiffness measurement.............................................................................

3.9.

Bushing stiffness measurement....................................................................................

3.10. 4.

5.

Characterisation of Bumpstops.................................................................................

Multibody model implementation ....................................................................................... 4.1.

Model characteristics ...................................................................................................

4.2.

Templates and model topology ....................................................................................

4.3.

Modelling force elements.............................................................................................

4.4.

Static suspension alignment .........................................................................................

4.5.

Modelling full vehicle mass properties .........................................................................

4.6.

Additional requests ......................................................................................................

Road simulator testing ........................................................................................................ 5.1.

Characteristics of the road simulator ............................................................................

5.2.

Using the road simulator to validate the multibody model ...........................................

5.2.1. Vertical excitation tests ................................................................................................ 5.2.1.1.

Experimental setup ....................................................................................... You're Reading a Preview

5.2.1.2.

Procedure ..................................................................................................... Unlock full access with a free trial.

5.2.2. Compliance tests ......................................................................................................... 6.

Download With Free Trial

Multibody model validation ................................................................................................. 6.1.

Suspension parameter analyses ...................................................................................

6.1.1. Correlation of suspension compliance simulations ................................................... 6.1.1.1.

Front suspension compliances .....................................................................

6.1.1.2.

Rear suspension compliances.......................................................................

Master your6.1.1.3. semester with Scribd Conclusion on the compliance tests ..............................................................  Read Free Foron 30this Days Sign up to vote title  6.1.2. Simulated front suspension kinematics and rates .......................................................... & The New York Times  Useful  Not useful Cancel anytime.

Special offer for students: Only $4.99/month. 6.1.2.1. Quasi-static front PWT simulations ...............................................................





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print


1

Download



Magazines



News

of 112

 

Design


 Search document






Documents



Sheet Music


6.2.3.3.

A study on validating suspension kinematics and rates during vertical excitat

tests.

99

6.2.3.4.

Conclusion on the vertical excitation tests ...................................................

7.

Conclusion and future work .................................................................................................

8.

Original contribution............................................................................................................

9.

Bibliography ........................................................................................................................

10. Appendix ............................................................................................................................. 11.1.

Appendix A: Miscellaneous information .................................................................

11.2.

Appendix B: Suspension hardpoint coordinates ........................................................

11.3.

Appendix C: Component mass properties .................................................................

11.4.

Appendix D: Exploded view of the suspension assemblies ....................................

11.5.

Appendix E:Technical drawings for bushing sleeves (next page).................................

11.6.

Appendix F: Bushing stiffness measurements ..........................................................

11.6.1. Trailing arm bushing .....................................................................................................

11.6.2. Rear track rod outboard bushing ..................................................................................

11.6.3. Rear track rod inboard bushing ....................................................................................

You're Reading a Preview

11.6.4. Lateral link bushing ...................................................................................................... Unlock full access with a free trial.

11.6.5. Lower control arm rear bushing ................................................................................... 11.7.

FreeofTrial Appendix G: Normalized Download inertias for With a variety vehicles ..........................................

11.8.

Appendix H: Road Simulator testing .........................................................................

11.9.

Appendix I: Signal post-processing using MATLAB ...................................................

11.10.

Appendix J: Full vehicle excitation test results ..........................................................






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


List of Figures

Figure 2.2.1: Front suspension (MacPherson) ............................................................................................. Figure 2.2.2: Rear suspension (Multilink) .................................................................................................. Figure 2.6.1: Vehicle performance classifications arranged by primary frequency range ......... .................

Figure 2.6.2: Effect of FD bushings on vertical body acceleration. ............................................................... Figure 2.10.1: The process of validating an ADAMS model, synthesized (45). .............................................

Figure 3.1.1: Front RVDT installation ........................................................................................................

Figure 3.1.2: Rear RVDT Installation ......................................................................................................... Figure 3.1.3: Installation of the handwheel torque and angle transducer ...............................................

Figure 3.1.4: Installation of the capacitive uniaxial accelerometers............................................................. Figure 3.1.5: The ADMA Gyro/accelerometer, the Data acquisition system and the display .......................

Figure 3.2.1: Coordinate Measuring Machine (CMM) .................................................................................

Figure 3.3.1: Weighting of the front track rod ............................................................................................

Figure 3.3.2: Full vehicle CoG position measurement by axle lift method.....................................................

Figure 3.4.1: Wheel alignment gauge.......................................................................................................

Figure 3.4.2: Steering ratio analysis ......................................................................................................... Figure 3.4.3: Changes in front RVDT readings due to steer.......................................................................

Figure 3.5.1: Front and rear damper force versus velocity ...........................................................................

Figure 3.6.1: Experimental set-up for tyre characterisation.........................................................................

Figure 3.6.2: Force-displacement curves for tyres at different pressures. .................................................... Figure 3.7.1: Experimental set-up for the characterisation of springs .....................................................

Reading .......................................................................... a Preview Figure 3.7.2: Front and rear spring forceYou're vs. displacement

Figure 3.8.1: Experimental set-up for the ARB measurements ..................................................................... Unlock full access with a free trial.

Figure 3.8.2: Mechanism comprised by the ARB, the swivel rod end and the actuator. .............................

Figure 3.8.3: ADAMS/View model of the experimental set-up .....................................................................


Figure 3.8.4: ADAMS/View results for ARB displacement as a function of actuator force .................. ............

Figure 3.8.5: FARB torque vs. angular displacement (fixed end) ..................................................................

Figure 3.8.6: FARB torque vs. angular displacement (free end) ...................................................................

Figure 3.8.7: RARB torque vs. angular displacement (fixed end) ..................................................................

Figure 3.8.8: RARB torque vs. angular displacement (free end) ...................................................................

Figure 3.9.1: Housing and sleeves used to fix the bushings .........................................................................

Figure 3.9.2: Applying axial preload to the bushings ...................................................................................

Figure 3.9.3: Bushing test configuration 1 - axial rotation ........................................................................... Master your semester with Scribd  Read Free Foron 30this Days Sign up to vote title Figure 3.9.4: Bushing test configuration 2 - radial rotation .........................................................................  & The New York Times Useful Not useful   Figure 3.9.5: Bushing test configuration 3 - radial translation ..................................................................... Cancel anytime.

Special offer forFigure students: Only $4.99/month. 3.9.6: Bushing test configuration 4 - axial translation ......................................................................

Figure 3.9.7: Averaging hysteresis loops ...................................................................................................





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document




 

Documents


Figure 5.1.1: MTS 329 Road Simulator layout (one corner) (53) .................................................................. Sheet Music

Figure 5.2.1: Experimental setup of the road simulator (front and rear axles shown in operation, respectivel

Figure 5.2.2: Input signal for PWT vertical excitation tests .......................................................................... Figure 6.1.1: Comparison of lateral force compliance test results for the front axle ................................... Figure 6.1.2: Comparison of overturning moment compliance test results for the front axle ....................... Figure 6.1.3 Comparison of aligning moment compliance test results for the front axle .......... .................

Figure 6.1.4: First scenario - effect on maximum simulated camber angle and change required to achieve th

target response for different bushing stiffnesses. .......................................................................................

Figure 6.1.5: Second scenario - effect on maximum simulated camber angle for different bushing stiffnesses Figure 6.1.6: Comparison of lateral force compliance test results for the rear axle ...................................

Figure 6.1.7: Comparison of overturning moment compliance test results for the rear axle ......... ................. Figure 6.1.8: Comparison of aligning moment compliance test results for the rear axle ......... .................

Figure 6.1.9: A summary of the measured and simulated compliances for front and rear suspensions ...........

Figure 6.1.10: Quasi-static PWT and roll angle tests for the front suspension..............................................

Figure 6.1.11: Simulated front axle camber angle vs. wheel position for kinematic and compliant configurat

.............................................................................................................................................................. Figure 6.1.12: Simulated front axle toe angle vs. wheel position for kinematic and compliant configurations

Figure 6.1.13: Simulated front axle wheel rate vs. wheel position for kinematic and compliant configurations

Figure 6.1.14: Simulated front axle camber angle (Y axis) vs. roll angle (X axis) for kinematic and compliant

configurations ......................................................................................................................................... Figure 6.1.15: Simulated front axle toe angle (Y axis) vs. roll angle (X axis) for kinematic and compliant

configurations .........................................................................................................................................

Figure 6.1.16: Simulated front axle roll rate vs. rollReading angle forakinematic You're Previewand compliant configurations ..........

Figure 6.1.17: Simulated rear axle wheel rate, camber and toe angle vs. wheel position .............................

Unlock full access with a free(Ytrial. Figure 6.1.18: Simulated rear axle roll rate, camber and toe angle axes) versus roll angle (X axis) ............

Figure 6.2.1: Graphical representation of the computational setup.............................................................

Download With Free Trial Figure 6.2.2: Data post-processing flowchart .............................................................................................

Figure 6.2.3: Simulated and experimental responses to FPWT excitations - time-domain analysis of damper

displacements and vertical spindle forces. .................................................................................................

Figure 6.2.4: Simulated and experimental responses to FOWT excitations - time-domain analysis of damper

displacements and vertical spindle forces. ................................................................................................. Figure 6.2.5: Simulated and experimental responses to FPWT excitations -............................................... Figure 6.2.6: Simulated and experimental responses to FOWT excitations - .............................................

Figure 6.2.7: Simulated and experimental responses to FPWT excitations - ............................................... Master your semester with Scribd  Figure 6.2.8: Simulated and experimental responses to FOWT excitations - to .............................................. Read Free Foron 30this Days Sign up vote title  Figure 6.2.9: Camber and toe angle versus suspension displacement obtained from experimental and & The New York Times Useful useful plots,  Not Cancel anytime.

vertical excitations tests. ......................................................................................................... Special offer forsimulated students: Only $4.99/month. Figure 10.1.1: Typical applications for different tyre models [4] ..................................................................





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document




 


Documents

Figure 10.6.7: Trailing arm bushing - Rotation around Z ............................................................................. Sheet Music

Figure 10.6.8: Rear track rod outboard bushing - Reference axis................................................................. Figure 10.6.9: Rtrod outbd bushing - Translation along X and Y (axisymmetric) .........................................

Figure 10.6.10: Rtrod outbd bushing - Translation along Z .......................................................................... Figure 10.6.11: Rtrod outbd bushing - Rotation around X and Y (axisymmetric) .......... .............................

Figure 10.6.12: Rtrod outbd bushing - Rotation around Z ...........................................................................

Figure 10.6.13: Rear track rod inboard bushing - Reference axis ................................................................. Figure 10.6.14: Rtrod inbd bushing - Translation along X and Y (axisymmetric) .........................................

Figure 10.6.15: Rtrod inbd bushing - Translation along Z ............................................................................ Figure 10.6.16: Rtrod inbd bushing - Rotation around X and Y (axisymmetric)........... .............................

Figure 10.6.17: Rtrod inbd bushing - Rotation around Z ..............................................................................

Figure 10.6.18: Lateral link bushing - Reference axis...................................................................................

Figure 10.6.19: Lateral link bushing - Translation along X ...........................................................................

Figure 10.6.20: Lateral link bushing - Translation along Y ...........................................................................

Figure 10.6.21: Lateral link bushing - Translation along Z ...........................................................................

Figure 10.6.22: Lateral link bushing - Rotation around X .............................................................................

Figure 10.6.23: Lateral link bushing - Rotation around Y .............................................................................

Figure 10.6.24: Lateral link bushing - Rotation around Z .............................................................................

Figure 10.6.25: Lower control arm rear bushing - Reference axis ................................................................

Figure 10.6.26: LCA rear bushing: Translation along X ................................................................................

Figure 10.6.27: LCA rear bushing: Translation along Y ................................................................................

Figure 10.6.28: LCA rear bushing: Translation along Z ................................................................................

Figure 10.6.29: LCA rear bushing: Rotation around X .................................................................................. You're Reading a Preview

Figure 10.6.30: LCA rear bushing: Rotation around Y ..................................................................................

Unlock full access with a free trial. Figure 10.6.31: LCA rear bushing: Rotation around Z ..................................................................................

Figure 10.7.1: Normalized roll inertias for different vehicles (52) ................................................................

Download With Free Figure 10.7.2: Normalized pitch inertias for different vehicles (52)Trial ..............................................................

Figure 10.7.3: Normalized yaw inertias for different vehicles (52) ...............................................................

Figure 10.8.1: Mounting of a MTS SWIFT wheel force transducer on a MTS 329 road simulator (55) ............ Figure 10.10.1: Simulated and experimental responses to FPWT excitations - ......................................... Figure 10.10.2: Simulated and experimental responses to FPWT excitations - ......................................... Figure 10.10.3: Simulated and experimental responses to FPWT excitations - ......................................... Figure 10.10.4: Simulated and experimental responses to FOWT excitations - .........................................

Figure 10.10.5: Simulated and experimental responses to FOWT excitations - ........... ............................. Master your semester with Scribd  Figure 10.10.6: Simulated and experimental responses to FOWT excitations ......................................... Read Free Foron 30this Days Sign up to- vote title  & The New York Times  Useful  Not useful

Special offer for students: Only $4.99/month.

Cancel anytime.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


List of Tables

Table 2.2.1: Technical specifications of the vehicle to be modelled..............................................................

Table 2.3.1: Degrees of freedom removed by joints ....................................................................................

Table 3.3.1: Full vehicle centre of gravity coordinates .................................................................................

Table 4.2.1: Modifications to original templates.........................................................................................

Table 4.2.2: Legend showing the abbreviation of each joint type. ...............................................................

Table 4.2.3: Legend showing the force elements. .......................................................................................

Table 4.3.1: Methods followed to model different bushings ........................................................................

Table 4.3.2: Approximate characteristics of non-tested bushings ................................................................ Table 4.4.1: Static camber and toe values at nominal operating conditions ............................................ Table 4.5.1: Normalized and absolute vehicle inertias from bibliography [52] .........................................

Table 5.2.1: Actuator control modes for vertical excitation tests .................................................................

Table 5.2.2: Description of the input signal for PWT vertical excitation tests ...............................................

Table 5.2.3: Load ranges for front compliance tests....................................................................................

Table 5.2.4: Actuator control modes for the front Fy compliance test .......................................................... Table 5.2.5: Actuator control modes for the front Mx compliance test .....................................................

Table 5.2.6: Actuator control modes for the front Mz compliance test ........................................................

Table 5.2.7: Load ranges for rear compliance tests ..................................................................................... Table 6.1.1: First scenario - change required to achieve the target response for different bushing stiffnesses

Table 6.2.1: RMSD of the simulated and experimental results for damper displacements and wheel forces ...

Table 10.2.1: Front suspension hardpoint coordinates ................................................................................

Reading a Preview Table 10.2.2: Rear suspension hardpointYou're coordinates .................................................................................

Table 10.3.1: Front suspension component mass properties ....................................................................... Unlock full access with a free trial.

Table 10.3.2: Rear suspension component mass properties ........................................................................ Table 10.8.1: Model 329 6DOF Spindle-coupled road simulator specifications [53]...................................







Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Acronyms and abbreviations MBS: FEA: DOF: NVH: CAD: CAE: CoG: CMM: DOE: K&C: LVDT: RVDT: ABS: FWD: WFT: LCA: f-: r-: LF: RF: LR: RR: -trod: -urt: -ARB: PWT: OWT: RMSD: FFT:

Multibody simulation/multibody system Finite Element Analysis Degree of Freedom Noise, Vibration and Harshness Computer Aided Design Computer Aided Engineering Centre of Gravity Coordinate Measuring Machine Design of Experiment Kinematics and Compliance Linear Variable Differential Transformer Rotary Variable Differential Transformer Anti-lock braking system Front Wheel Drive Wheel Force Transducer Lower control arm Front... Rear... Left Front Right Front You're Reading a Preview Left Rear Unlock full access with a free trial. Right Rear Track rod UprightDownload With Free Trial Anti Roll Bar Parallel Wheel Travel Opposite Wheel Travel Root Mean Square Deviation Fast Fourier Transform






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


1. Introduction 1.1. Background and motivation

Computer Aided Engineering (CAE) in general, and Multibody Simulation (MBS particular, have gradually transformed the former unidirectional, sequential pro design process into a concurrent process in which it is not necessary for preceding tasks to have ended before engineers can start working on the next ta downstream.

MBS enables engineers to numerically solve complex dynamic problems w would have taken a lot of effort by simply using analytical methods. Virtual dyna models of products which do not physically exist can be created, hence provi valuable information early in the design stage. In addition, even when the product has already been produced, MBS can reduce the amount of costly physical testing required to optimise it.

1.2. Vehicle Dynamics and Multibody Simulation You're Reading a Preview

In order to be competitive, automotive manufacturers are forced to reduce Unlock full access with a free trial. duration and cost of the design process of their new products, while meeting ever-increasing customer expectations in terms of quality, comfort, efficiency Download With Free Trial performance.

MBS software is a key dynamic simulation tool for engineers working in automotive industry, in which the use of predictive methods is vital to increase efficiency of the engineering process.

Within the automotive industry, MBS software has been traditionally used by veh dynamicists for a variety of tasks, including, but not limited to: 

Master your semester with Scribd Read Free Foron 30this Days Sign up to vote title  & The New York Times Useful Not useful   Ride comfort analysis (isolation from external and internal disturbances) 

Special offer for students: Only $4.99/month. Analysis of suspension 

Cancel anytime.

kinematics and elastokinematics.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


1.3. Project Sponsor

This project was carried out in collaboration with the Automotive Technolog Innovation Centre of Navarre (CITEAN) in Pamplona, Spain. The Departmen Kinematics and Dynamics was in charge of supervising the project and supplying required funds, equipment and knowledge.

CITEAN is a government-backed entity which provides consultancy service companies in a variety of sectors, including automotive, rail and wind po generation. Its main lines of activity are structural design and analysis, kinema and dynamics, mechanical and environmental tests and NVH. Their facilities include: 

 

A range of seismic platforms on which modular rigs and hydraulic actua can be mounted to carry out component characterisation tests. A workshop with a lathe and a mill. A single-axle, spindle-coupled MTS 329 road simulator.

This project will be the stepping stone to a larger-scale investment on veh dynamics-related MBS by CITEAN. You're Reading a Preview

In the medium-term, the company is seeking to implement a Multibody model w Unlock both full access a free trial.high-frequency loads which can be used to accurately simulate thewith low and transmitted to the suspension components. The dynamic loads obtained from M With Free Trial can be processed to generate Download time-compressed pseudo-white noise signals w would then be used to reduce the duration of fatigue tests on suspen components without losing accuracy. Future evolutions of this model would also valuable for studies on ride comfort, NVH and active dampers.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


1.4. Research goals and approach 1.4.1.

Project aim

The aim of this project is to implement and correlate a multibody dynamic model mid-size passenger car, comprising a Macpherson front suspension and a mult rear suspension, in order to simulate its handling and primary ride behaviour. ADAMS/Car software will be used. 1.4.2.

Objectives

A. Identify and measure the unknown variables required to implement the ADAMS/Car model, such as: a. Mass properties of individual components and of the complete vehic b. Topology of the vehicle c. Steering ratio d. Damper curves e. Tyre characteristics f. Component stiffness: i. Springs ii. Anti Roll bars iii. Bushings iv. BumpstopsYou're and droopstops Reading a Preview Unlock full access withmeasured a free trial. B. Build the ADAMS/Car model using the data.

C. Design appropriate testsDownload to measure the systemWith Free Trial and vehicle-level characteristics of the vehicle, both in ADAMS/Car and in real life. D. Run the selected simulations in ADAMS/Car. E. Carry out the experimental tests. F. Process the results and produce a correlation study which can be used to

Master your draw semester with Scribd conclusions and suggest further work Read on the subject. Free Foron 30this Days Sign up to vote title & The New York Times  Useful  Not useful Special offer for students: Only $4.99/month.

Cancel anytime.

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


1.4.3.

Methodology

A. All the characterisation tests were carried in-house within CITEAN’s prem and existing resources were used where possible.

a. The masses of all the suspension components were measured usin precision scale. The components were drawn in CAD in orde estimate the positions of their centres of gravity and inertias. In orde measure the mass and the CoG position of the full vehicle an axl test was carried out as specified in ISO 10392 [1]. No equipment available to measure the full vehicle inertias so these were der from bibliography.

b. Suspension hardpoint coordinates were acquired by lifting the ve and using a CMM. Wheel plane orientations were measured u wheel alignment gauges. c. The handwheel to road wheel steering ratio was measured us handwheel angle sensor and a wheel alignment gauge.

d. Damper force-velocity curves were acquired using a hydraulic dam dynamometer. You're Reading a Preview

e. The static radial Unlock stiffness of the was measured using a hy full access with atyre free trial. actuator coupled to a strain gauge. Measurements were carried ou different inflation pressures. No Free additional Download With Trial tyre data was available.

f. A variety of component stiffness characterisation tests were carried using one or two hydraulic actuators coupled to strain gau generating a range of quasi-static axial forces and moments. The e of displacement frequency on the stiffness was ignored in all tests.

B. Subsystems and assemblies which were specific to the vehicle's topology Master your semester with Scribd  Free Foron 30default Days characteristics were implemented based onRead ADAMS/Car Sign up to vote this title templates  & The New York Times  Useful  Not useful

Cancel anytime. side, a series of dynamic excitation and suspen compliance tests were planned using the MTS 329 road simulator. Th

C. On experimental Special offer for students: Onlythe $4.99/month.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


F. The logged data from the experimental tests was filtered and resampled ease of analysis. Subsequently, the quality of curve fitting was numeric quantified where possible. 1.4.4.

Report outline

The project report consists of 8 main sections plus an appendix. Section 1 expl the project background and states the project goals, outlining the approach follo to achieve them.

Section 2 contains a detailed review of literature on automotive MBS applicat and dynamic model correlation.

The process of measuring system- and component-level parameters of the veh along with the results of these measurements, are described in section 3.

Section 4 shows how these measurements are then used to build the multib model.

The use of road simulator tests as a tool to correlate the multibody mode explained in section 5 and the correlation between experimental data and multib simulations is discussed in section 6. Reading a Preview You're

Unlock full access a free trial. Finally the conclusion and suggestions forwith future work are presented in sectio followed by a statement of the original contribution of this project, in section 8.







Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


2. Literature Review 2.1. Overview

At the time of this investigation, extensive work is still being done in optimis adding new functionalities and finding new applications to the existing MBS softw However, the application of MBS to vehicle dynamics has been a standard prac in industry since the early nineties, which means that a vast amount of literatur available in this field, a portion of which is shown here.

2.2. Vehicle to be modelled

A Volkswagen Passat B6 (2006) was the vehicle chosen for this project. The pri reason for this choice was the fact that the sponsoring company had previo worked with it and some data was readily available.

Those specifications which are relevant to the purposes of this project can be s in Table 2.2.1 [2][3]: Brand/Model You're Volkswagen Passat B6 (2006) Reading a Preview Body style 4 door sedan Unlock full access with a free trial. Layout Front engine, front wheel drive Engine 2.0 litre, inline 4, TDi With Free Trial Transmission Download 5 speed manual Wheelbase 2709mm Track (front, rear) 1552mm, 1551mm Front suspension MacPherson Rear suspension Multilink Tyres Pirelli P7 215/55 R16 97 W Table 2.2.1: Technical specifications of the vehicle to be modelled

Master your semester with Scribd  2.2.1 Phantom views of the front and rear suspensionsRead can be seen in title Figure Free For 30this Days Sign up to vote on  FigureYork 2.2.2, Times respectively[2]. & The New  Useful  Not useful Special offer for students: Only $4.99/month.

Cancel anytime.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 2.2.1: Front suspension (MacPherson) You're Reading a Preview Unlock full access with a free trial.







Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design




 Search document




Documents



Sheet Music


2.3. Typical components of an automotive multibody model

A multibody system, as the name implies, can be understood as an assembl multiple bodies, connected to each other or to the "ground" by constraints an forces.

The constraints limit the number of independent kinematical possibilities, or deg of freedom (DOF), the bodies have to move. Therefore they can be us understand what movement the bodies (members of a mechanical system) will fo i.e. their kinematic behaviour.

Typical constraint elements, or joints, are shown in Table 2.3.1 [4]. It can be s that some joints in this table remove "half-constraints". This is because these jo relate translational and rotational motion, generating a single, coupled DOF for b motions: Joint

Translational DOF removed

Rotational DOF removed

Total DOF removed

Constant velocity

3

1

4

Cylindrical

2

2

4

Fixed

3

3

6

Hooke

You're Reading a Preview 3 1

4

Planar

2 Unlock full1access with a free trial.

3

Rack and pinion Revolute

0.5

0.5

Download 3 With Free Trial 2

1 5

Screw

0.5

0.5

1

Spherical

3

0

3

Translational

2

3

5

Universal

3

1

4

Table 2.3.1: Degrees of freedom removed by joints

Master your semester with Scribd & The New York Times


 

 Useful  Not useful Cancel anytime. On the other hand, the bodies in a MBS, which can either be perfectly rigid or flex Special offer for students: Only $4.99/month. have a mass and a specific inertia tensor. If the forces acting on the bodies





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


The basic forces typically involved in an automotive MBS application are following [6]: 

Internal (action-reaction) forces, involving two bodies: o

o

o

o

o

Springs: A uniaxial force is applied along a line joining two points in bodies. This force is a function of the relative displacement of the t points. Dampers: Same as springs, but the force is a function of the relative velocities of the points. Bumpstops and rebound stops (droopstops): Same as springs, but force only acts when the distance between the two points is smaller than a given clearance distance. Anti roll bars: A moment is applied around an axis joining two points two bodies. This moment is a function of the relative angular displacement of the two bodies around the aforementioned axis. Bushings: The forces and moments created by a bushing along the DOF of one of the bodies can be represented by six uncoupled equations of motion of the form: You're Reading a Preview Unlock full accessalong/around with a free trial. where F is the force/moment a given axis, K is the stiffness of the bushing for that axis, c is the damping, and f is the With Free Trial preforce/pretorqueDownload (preload).

It must be noted that K and c may not necessarily be constants, i.e. they be defined respectively by functions of displacement and velocity, therefor producing a nonlinear bushing. This also applies to the stiffness paramete the rest of force elements. 

Master your semester with Scribd & The New York Times Special offer for students: Only $4.99/month. 





Useful

External (action only) forces, involving a single body, including:







Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


2.4. Tyre models

Tyres are the only contact point between a vehicle and the road. Thus, the fo and moments generated by the tyres are of crucial importance for the dyna behaviour of the road vehicle.

In order to predict these forces and moments under a variety of operating conditi different tyre models have been developed during the last few decades. Th models are usually based on a large amount of experimental data, which can obtained via different types of tests [7][8].

Different models have different strengths and weaknesses and therefore eac them is best suited for a given application. Figure 10.1.1 in appendix A shows typ applications for different tyre models. The 2002 version of the Pacejka tyre mode and the FTire flexible ring tyre model [9] are arguably the most complete polyvalent models offered in ADAMS. In general terms, the Pacejka model is u for all applications except for those involving high excitation frequencies (over 15 or those requiring modelling of non-linear tyre enveloping effects, such as dura studies[4]. You're Reading a Preview

2.5. Suspension configurations

Unlock full access with a free trial.

The main functions of automotive suspensions are[10]: Download With Free Trial





To keep the tyres at optimal angles with respect to the road in orde generate the desired forces on the tyre contact patches.

To transmit these forces into the sprung mass while maximizing the isola of occupants from road disturbances

vice versa,with to reactScribd the forces caused by the motion of the sprung m Master your And semester  Read Free Foron 30this Days Signvariations up to vote title while reducing/optimising the vertical load on the tyre con  & The New York Times  Useful  Not useful patches. 


Cancel anytime.

Ever since the automobile was invented a wide range of suspension systems





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


This suspension configuration is very popular on the front axles of mod transverse-engined, FWD cars, as its inherent geometry provides enough space a wide engine bay and unrivalled design simplicity [10].

However, the fact that the strut acts effectively as an infinitely long upper control generally means that negative camber is lost with bump[11]. This reduces the lat force generation capabilities of the tyres. In a cornering situation, an increas negative camber is desired on the outside wheel [11]. This can be explained phenomena such as "camber thrust" [7].

Multilink suspensions normally consist of four or five links. If non-compliant (a "kinematic") joints were used between all the links, the fifth link, when present, w effectively over-constrain the mechanism, leaving no remaining DOFs for it to mo

However, due to the fact that bushings are widely used instead of kinematic join real life, the compliance of the joints has to be taken into account. In this case fifth link provides more accurate control of the compliant toe angles caused by cornering forces and moments [10].

Increased control on the system compliances affecting the wheel plane orienta (a.k.a "elastokinematics") is beneficial in many ways: optimised suspen elastokinematics improve the handling of the vehicle, help reduce tyre wear, You're Reading a Preview provide a better ride quality due to the possibility of achieving greater longitud Unlock full access with a free trial. compliances without detriment to suspension kinematics[12]. Download With Free Trial

2.6. Influence of suspension isolator models on MBS

It has been mentioned in section 2.3 that bushings can be modelled as set nonlinear forces and moments. The nonlinearities are caused primarily by reasons:

Master your semester with Scribd  Read Free Foron 30this Days SignIn up to vote title Coupling of different modes of deformation: automotive suspensions  & The New York Times Useful Not useful   bushing is mostly deformed in such a way that multiaxial rotations 


Cancel anytime.

translations occur simultaneously. However, the torsional stiffness of bushing is significantly affected by the amount of axial displacement.[13]





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


The effects of these phenomena on automotive suspensions are best appreciate relatively high road disturbance frequencies, well within the "secondary ride" area shown in Figure 2.6.1 [16]

Figure 2.6.1: Vehicle performance classifications arranged by primary frequency ran

The effects of a frequency dependant (FD) bushing model on vertical cha You're Reading a Preview acceleration for a given road disturbance and vehicle speed are shown in Fig Unlock full access with a free trial. 2.6.2 [16]. Baseline data (in blue) represents a quasi-static bushing model in w stiffness is purely a function of displacement. Download With Free Trial






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Computational effort is a major factor to be considered when modelling bushings MBS. FD and hysteretic bushing models require more time to compute than qu static models. For this reason, quasi-static models are still widely used for prim ride simulations. In these cases, the damping ratio is generally defined as a d proportion of the instantaneous stiffness. In most cases this value is considere be in the region of 1% [17][4].

Another factor having an influence on computational effort is the value of the bus stiffness. High stiffness bushing models require small step-sizes in the nume integration[18]. This can be explained by the fact that bushings generate hig frequency responses than vehicle dynamic responses [19]. One way to work aro this is the subsystem synthesis method, in which separate equations of motion (E are generated for the chassis and suspension subsystem [18].

2.7. Target specifications of a multibody model

Back in 1992, Lotus obtained good correlation of yaw rate between their ADA model and real tests for an 80km/h single lane change manoeuvre [20]. In orde achieve this, the multibody model they used had in excess of 200 DOF and use complex Pacejka tyre model. You're Reading a Preview Unlock full access with a free trial.

Ever-increasing computing power can be a temptation to keep increasing complexity and amount of DOF of multibody models. However, excessive comple Download With Free Trial can lead to a “paralysis of analysis”[21], in which the large number of parame required to build the model and the time required to gather them actually reduce usefulness of the model.

It is recommended that the level of complexity of the model to be impleme should just match the complexity of the problem to be solved [22]. The ideal mo would be the least complex one which provides a solution of an “accepta accuracy/correctness, where the task t o define “acceptable” to thejudgem Read Free Foris 30left Days Sign up to vote on this title  of the engineer.  Useful  Not useful


Cancel anytime.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


2.8. Typical methodologies for implementing and validating multibody models

The methodology for this project, defined in section 1.2.3., is based largely general industry practice.

This types of project generally involves three major phases. First, component-l parameters are measured and the model is formulated. Then, system-level da collected (this can also be done before implementing the model). Finally, simulation predictions and the experimental data are compared using the sa driver control inputs in both cases[23]. 2.8.1.

Techniques for measuring component-level parameters

Suspension component parameters are generally measured using similar techniq to those defined in section 1.2.3. Once the parameters have been measured, n of the results should be adjusted to improve correlation unless the quality of measurements has been improved; otherwise the correlation process me becomes a process of iterative curve fitting[23].

In addition to measuring the stiffness of suspension isolators, dampers, etc, effect of linkage and support You're stiffness is also considered in some cases w Reading a Preview increased accuracy is required. Unlock full access with a free trial.

According to [24], in order to understand the extent to which this affects veh Withbelonging Free Trial to a multilink suspension, response, FEA was carried outDownload on a linkage boundary conditions being those corresponding to the vehicle at static equilibri The computed deflection of the linkage turned out to be less than half of precision of the CMM used to measure the link. However, considerable deflect do occur in other components when they are subjected to greater dynamic and loadings[12].

Master your with Scribd  Lack ofsemester access to vehicle design data is also a common issue. Even when de Read Free Foron 30this Days Sign up to vote title  data is available from the manufacturer, determining the actual values & The New York Times useful of parame  Useful  Not asOnly vehicle and Special offer forsuch students: $4.99/month.

Cancel anytime.

component inertias or exact hardpoint locations, can challenging[24]. This issue is further aggravated when no design data is availabl





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print


1

Download



Magazines



News

of 112

 

Design


 Search document






Documents



Sheet Music


In order to measure full vehicle inertias, devices such as the Vehicle In Measurement Facility (VIMF) [26], have been widely used [19][27].

As far as suspension hardpoints are concerned, the difficulty of measuring coordinates accurately can be overcome by creating an approximate multib model, carrying out a statistical design of experiments (DOE), comparing simulated results to actual test results, and thus adjusting the coordinates in model to achieve correlation. DOE can be carried out using commercial simula software such as ADAMS/Insight [4], although this method usually involves computational costs [24].

2.8.2.

Techniques for measuring system-level parameters

Full-vehicle behaviour depends on system level parameters (weight distribution, stiffness, etc.), rather than on individual component parameters [28]. In orde acquire data at system level, two main methods are generally used: K&C testing full vehicle dynamic testing. Aerodynamic testing would also be required if hig speed manoeuvres are to be modelled[23]. Regardless of which tests are to carried out, the design process of the system-level tests should be comple independent from the component parameter phase to avoid bias on You're Reading identification a Preview selected tests[23]. Unlock full access with a free trial.


2.8.2.1.

Kinematics and Compliance testing

There is a range of commercial and in-house designed K&C rigs in the market, different operating principles can be found. Some rigs fix the vehicle body to ground and use actuators to apply vertical and horizontal forces on the tyre con patch via high-friction wheel platforms. Another approach is to move the entire b of the vehicle in heave, pitch and roll while the wheel platforms are allowed  to m Read Free Foron 30this Days Sign up to vote title within the road plane to apply lateral and longitudinal forces. In all cases, the lin Not useful  Useful under Cancel anytime. and angular displacements of the wheel reference planes, different loa Special offer for students: Only $4.99/month. conditions, are measured[29]. When carrying out a K&C test, the vehicle body ha






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


A K&C test can be used to obtain valuable data for the multibody model [28][31]: 











Vertical suspension rates and hysteresis: Mainly the stiffness of springs bumpstops, and the clearance distance of the latter ones are measured. contribution of the bushings can be measured by removing the springs.[32 Front/rear roll stiffness distribution: same as before, but ARB stiffness is taken into account1. Change of toe and camber angles with bump/roll. Instant centre locations : geometrical parameters such as roll centre and ratios are measured.

Longitudinal compliance: the compliances of bushings, subframes and w hub assemblies2 under braking forces are measured. Their effect on toe a is assessed.

Lateral compliance: when forces are opposing, the compliances of bushings and the wheel hub assembly are measured. When forces parallel, the compliances of the subframe 3 and the steering system are ad to these. The effect of all this on toe angle and camber is assessed. You're Reading a Preview



Aligning moment compliances: when Unlock full access withmoments a free trial. are opposing, mainly w hub and bushing compliance is measured. When parallel, steering sys behaviour is measured. The steering Download Withratio Free and Trial feedback can be measured installing a handwheel angle and moment sensor. The effect of po steering can be evaluated by switching the engine on and off.

Due to the fact that K&C rigs are very specialized machinery, attempts have b made to use other devices to achieve similar results in a more economical way. instance, there have been reports of 90% correlation between traditional K&C t and measurements taken using a combination of a MTS 329® road simulator, M  Read Free Foron 30this Days Sign up to vote title SWIFT® wheel force transducers and wheel vector sensors[33].  Not useful  Useful  Cancel anytime.






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


2.8.2.2.

Full vehicle dynamic testing

Track acquisitions are the most common method of obtaining information on vehicle dynamic behaviour. In these tests, vehicles are equipped with sensor acquire data on the parameters of interest, and standard manoeuvres are perfor in accordance with ISO 15037-1:2006 [34] The most frequent standard manoeuvres are listed below [35]: 

Steady-state cornering: This can be achieved by either driving at a cons radius while increasing the speed, by maintaining a fixed handwheel a while increasing the speed, or by maintaining a constant speed w increasing the handwheel angle. The increments can either be discret continuous[36]. o



Swept steer: This is a particularly popular version of the steady-s cornering manoeuvres. The handwheel is turned at a constant ang velocity at an order of magnitude of 10deg/s, while the vehicl travelling at a constant speed[23].

(Pseudo) step steer: Used to evaluate the lateral frequency respo functions of the vehicle to Reading transient inputs; a pseudo-discrete step You're a Preview approximately 40° is applied to the handwheel angle, at an approxim Unlock full access with a free trial. angular velocity of 500 deg/s[23].

Withand Free a Trial Pulse steer: A linear Download increment linear decrement are app consecutively on the steering wheel angle, producing a "fishhook" or "J-t type manoeuvre which excites both steady-state an transient responses o vehicle. This manoeuvre is progressively replacing the traditional swept handwheel angle inputs. If suitably short step sizes are chosen, step s inputs are able to excite the full range of vehicle directional frequencies similar way to swept sine inputs, while being quicker and simple  perform[35]. It must be noted that frequency responses are Read Free Foron 30this Days Sign up to vote title dependen  vehicle speed, and therefore can only  beUseful derived from constant sp Not useful  Cancel anytime. Special offer for students:maneouvres[23]. Only $4.99/month. 






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


Whenever it is possible to have repeatable control inputs, manoeuvres should repeated until an estimation of the random error level can be obtained by statis methods. This can be achieved after approximately 10 repetitions. The calcula random error level can be used as a reference when validating the simulations.[2

2.8.3.

Correlation

"A model will be considered to be valid if, within some specified operating range and for some given inputs, the simulation’s predictions agree wi th the physical system’s responses within some specified level of accuracy" [23]

Simulation predictions can only be correct within a portion of the operating rang the vehicle. For instance, out of a given input frequency range predictions become progressively worse. In order to understand these phenomena, simulat should be validated within both time and frequency domains[23].

As mentioned in section 2.8.1, when a simulation is found to be poorly correlat the only parameters allowed to be changed are those for which no accurate da available. Parameters could either be iterated manually (by changing one at a ti You're Reading a Preview or by using more efficient software-enabled statistical DOE. However, if this me Unlock full access withcorrelation a free trial. is used there is a risk of obtaining good with a model that is representative of reality, due to the fact that the solutions to the MBS may no Download With Free Trial be identified as accurately unique[24]. In order to avoid this, parameters should possible and sanity checks carried out on the model regularly to ensure that results are logical[35].

2.9. Use of MBS within the automotive industry

Master your semester with Scribd  Read Free Foron 30this Days Sign up to vote Multibody models of varying complexity have been implemented intitle recentyear & The New Times manufacturers, includingFord[42], Useful Volkswagen Not useful most York major automotive [43], B Special offer for[44], students: Only $4.99/month. Jaguar [45] and Nissan

Cancel anytime.

[46]. In general, the process of implementing the in multibody models is strongly dependant on the traditional product developm





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


2.10. Summary

Modelling the behaviour of automotive suspensions is a complex task. In orde implement a multibody model of a vehicle, all relevant component-level parame must be identified and measured by experimental tests.

Once the component-level parameters are measured, they are incorporated into multibody model, and simulations are run.

Subsequently, a correlation study is carried out to see if the simulated behaviou the vehicle at system level is comparable to the actual behaviour.

In order to certify this, two main methods are used: The loading-induced change wheel plane orientations and positions are measured, and the response of the w vehicle to different dynamic events is analysed (Figure 2.10.1). The first me generally involves using a Kinematics and Compliance (K&C) rig, and the sec generally consists of driving the vehicle on a test track.

You're Reading a Preview Unlock full access with a free trial.







Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


3. Measurement of system and component-lev parameters 3.1. Instrumentation

In order to acquire the required data, a variety of sensors were installed on vehicle.

Suspension displacements were measured by installing one string potentiom (RVDT) in each corner. The body of the front potentiometer was fixed to the sp perch, while the string was attached to one of the bolts of the upper strut m (Figure 3.1.1)



Figure 3.1.1: Front RVDT installation

Similarly, rear string potentiometers were fixed to the damper body by an in-ho built clamp. The strings were attached to the top damper mount (Figure 3.1.2).






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 3.1.2: Rear RVDT Installation You're Reading a Preview

A handwheel torque and angle Unlock sensor was with alsoa free installed, making sure that the pu full access trial. of the angle transducer remained in tension throughout a full turn of the handwh (the outer perimeter of the handwheel isWith not Free concentric Download Trial with its turning axis) (Fig 3.1.3).






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


In order to measure the angular orientation of the vehicle and all six acceleration Genesys® ADMA (Automotive Dynamic Motion Analyser) accelerometer gyroscope system was installed on the central console, between the driver passenger seats (see Figure 3.1.5). This was the closest position available from estimated CoG of the vehicle. The exact position of the CoG was to be estima later, experimentally.

Its orientation was adjusted with a digital scale so that its top face was comple horizontal with the car standing at its nominal operating ride height. The nom operating conditions meant that the fuel tank was completely full (in orde minimize fluid oscillations), all the sensors and acquisition systems were insta and a driver was sitting in the vehicle.

The fact that the ADMA system was supported by the plastic of the central con meant that the acceleration measurements might be distorted. To offset these r and validate the measurements, three extra uniaxial capacitive accelerometers w installed on the two top strut mounts at the front, and on the rear subframe, above the inboard hardpoint of the right lateral link (see Figure 3.1.4). This loca was chosen for ease of access and because the coordinates of the hardpoint w be measured when creating the vehicle topology. The accelerometer was pla exactly 60mm above the hardpoint. You're Reading a Preview Unlock full access with a free trial.


Figure 3.1.4: Installation of the capacitive uniaxial accelerometers






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


3.2. Suspension topology measurement

The topology of the suspension was measured using a CMM (see Figure 3.2.1). centres of cylindrical bushings were estimated by measuring three points on e boundary of the cylinder to create two circumferences, and finding the midp between their centres. All measurements were taken with the vehicle at the nom operating ride height. Bushing centre displacements caused by preloads at nom ride height were not taken into account, and could therefore be a source of erro the measurements.

It must be noted that the arm of the CMM could not reach both axles without mo the base. A reference axis was created at a point underneath the front subframe the translation of the CMM base was measured with respect to it, in order to ob the relative positions of the rear axle hardpoints. This might be a source of errone offsets between the front and rear axle measurements.



Master your semester with Scribd Read Free Foron 30this Days Sign up to vote title & The New York Times Useful  Not useful Machine Figure 3.2.1: Coordinate Measuring (CMM) Special offer for students: Only $4.99/month.

Cancel anytime.

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


3.3. Estimation of mass properties Suspension components were weighted using an electronic scale, as seen in 3.3.1.

Figure 3.3.1: Weighting of the front track rod

The components were measured and drawn in CAD in order to estimate positions of their CoG and inertias. estimated component masses and ine You'reThe Reading a Preview can be found in Appendix C. The full vehicle inertia estimations were left for a Unlock full access with a free trial. stage in the project.

Download With Freean TrialIntercomp® SW500 E-Z w Vehicle corner weights were acquired placing scale system on one of the levelled seismic platforms and weighting the veh under the aforementioned nominal operating conditions. The measured mass 1518kg.

From the corner weights, the position of the CoG within the horizontal plane calculated knowing that:


where





Useful

 

is the distance from the front axle centreline to the CoG (pos





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


where

is the distance from the longitudinal vehicle centreline to the CoG (pos

to the left),

's are the front and rear tracks, and the

's are the masses

individual tyre contact patches (FL= front left, and so on).

The vertical (Z) position of the CoG was measured using the axle lift met specified in ISO 10392 [1].

The set-up of the experiment can be seen in Figure 3.3.2. The front axle was pla on an elevating platform while the rear axle rested on scales on the workshop floo


Download With Free Trial Figure 3.3.2: Full vehicle CoG position measurement by axle lift method

The ISO standard specifies that the suspension should be blocked so that deflection occurs. Arguably, the most straightforward way to do this was to weld dampers so that they became solid rods. However, the price of second-h dampers for such a contemporary vehicle wasRead too and this  option Free Foron 30 Days Sign up tohigh, vote this title  discarded at the expense of accuracy of measurements.  Useful  Not useful


Cancel anytime.

The axle loads for different front axle heights were measured. Then, the CoG he





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


XCG 1070mm YCG -3mm ZCG 460mm Table 3.3.1: Full vehicle centre of gravity coordinates

3.4. Steering ratio and wheel alignment

The suspension alignment was measured using a FACOM® GTR 300 suspen geometry set-up device. The vehicle body was lifted using the elevating platform measurements were carried out at different suspension displacements. measured parameters included camber, toe, kingpin inclination and castor. In addition, the suspension set-up device was used for measuring the steering (Figure 3.4.1).



Figure 3.4.1: Wheel alignment gauge Master your semester with Scribd Read Free Foron 30this Days Sign up to vote title & The New York Times  Useful  Not useful Special offer forIn students: $4.99/month. orderOnly to achieve this,

 

Cancel anytime.

the steering angle was increased at regular intervals and readings of the handwheel angle sensor and the suspension set up device w





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 3.4.2: Steering ratio analysis

The fact that the string potentiometers had been attached to the spring perc meant that their reading would change with steer, as the whole strut would ro around the upper mount. The suspension set-up device was used to measure phenomenon. Potentiometer readings were recorded for various roadwheel an and a polynomial regression was applied to the resulting values. Initially You're Reading a Preview polynomial was to be used to weight the suspension displacement readings du Unlock accessacquisitions, with a free trial. these measurements (Fig track acquisitions. In the absence offull track 3.4.3) were kept for future reference. The curves for the left and right wheels Download Freeare Trialnot exactly equal on both si different as the mounting positions of the With RVDTs






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


3.5. Measurement of the damper force-velocity curves

Damper curves were obtained using a hydraulic dynamometer. The results for f and rear dampers can be seen in Figure 3.5.1.

Figure 3.5.1: Front and rear damper force versus velocity You're Reading a Preview Unlock full access with a free trial.

3.6. Tyre static radial stiffness measurement Download With Free Trial

The radial stiffness of the tyres was measured using an actuator attached to an a force transducer, as seen in Figure 3.6.1.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 3.6.2: Force-displacement curves for tyres at different pressures. You're Reading a Preview Unlock full access with a free trial.

3.7. Spring stiffness measurement


Front and rear springs were tested in a similar way to the tyres. Actuator speed 1mm/s were used and hysteresis was negligible. The experimental set up can seen in Figure 3.7.1 and the results can be seen in Figure 3.7.2.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 3.7.2: Front and rear spring force vs. displacement

3.8. Anti Roll Bar stiffness measurement

The front and rear ARBs were fixed through its bushing mounts to a universal jig order to apply a torque to the ARBs, one of their ends was attached to the actu via an in-house fabricated attachment fork. The opposite end was first left fre Reading order to measure the bushingYou're stiffness, anda Preview then fixed to the jig to measure stiffness of the whole assembly. The experimental set-up can is shown in Fig Unlock full access with a free trial. 3.8.1. Download With Free Trial






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


An inherent problem of this setup was that the swivel rod end of the actuator and fabricated attachment fork created a four-bar mechanism, which made it laboriou identify the moment applied on the ARB and its rotation purely using the actu force and displacement data. This difficulty is illustrated in Figure 3.8.2.

Figure 3.8.2: Mechanism comprised by the ARB, the swivel rod end and the actuato

Although the problem of calculating the moment and angular displacement at from the actuator force and displacement data could have been solved analytic You're Reading a Preview ADAMS/View was used to facilitate the process. The positions of points "A", "B" Unlock full access with a free trial. "C" were measured and an equivalent multibody model was created, the ARB tor being simulated by a torsional spring of unit stiffness (see Figure 3.8.3). The i Download With Free Trial variables were the force and displacement of the actuator, and the output varia were the angular displacement and the reaction torque at point "C".






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


The results showed that the relationship between the actuator force and the ang displacement of the ARB was largely linear (see Figure 3.8.4). Therefore, the e of the mechanism geometry was ignored in subsequent calculations, and torque point "C" were simply calculated by multiplying the actuator force by the ver component of the distance between points "B" and "C". Similarly, the ang displacements at point "C" were calculated by simple trigonometry of right-a triangles.


Figure 3.8.4: ADAMS/View results for ARBWith displacement Download Free Trial as a function of actuator for

The processed results of the experimental tests for the front and rear anti roll b for both fixed and free ends, are shown in the four figures below.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 3.8.6: FARB torque vs. angular displacement (free end)



Figure 3.8.7: RARB torque vs. angular displacement (fixed end)






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


In the free end tests, a considerable hysteresis can be seen between compression and expansion cycles, despite the relatively low actuator sp (1mm/s). This behaviour is typical of the polymers employed in the bushings support the ARBs.

In addition, in Figure 3.8.8 a peak can be appreciated due to the clash between free end of the ARB and the jig supporting it, which effectively makes the ARB s working in torsion. All results beyond the contact point were ignored for bus stiffness determination. However, the linear slope generated provides a good ide the linear contribution that the torsion of purely elastic steel brings to the overall A stiffness.

In fixed end tests, hysteresis is hidden by the fact that the measured stiffnes considerably higher and it is produced mainly by the torsion of highly elastic s The "steps" seen in Figure 3.8.7 could be due to sticktion or stick/slip behaviou the steel/rubber interface in the ARB bushings.

3.9. Bushing stiffness measurement

The stiffness of the bushings around and along their three axes had to be measu You're Reading a Preview for the multibody model. Unlock full access with a free trial.

Given the large number of bushings present in the suspension assemblies, it Download With Free Trial decided that the most representative bushing models should be selected experimentation, and the same measurements would be used for other sim bushing constructions.

Bushings were chosen according to their accessibility and their character features. Some bushings were found to be unique in shape due to their spe distribution of voids and reinforcements, features which are commonly used to ta their stiffness independently along each axis. However, those bushings  which w Read Free Foron 30this Days Sign up to vote title  solid cylinders of rubber were only characterised once. Not useful  Useful  Cancel anytime. Special offer forThe students: Only $4.99/month. selected bushings were the rear track rod outboard (component 2, Fig 10.4.1), the rear track rod inboard (component 3, Figure 10.4.1), the trailing






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 3.9.1: Housing and sleeves used to fix the bushings

In order to simulate reality as accurately as possible, the bushings were preloa axially by two nuts as they would be in the car (see Figure 3.9.2).



Figure 3.9.2: Applying axial preload to the bushings

With the bushings held in place, different experiment set-ups were designed to a Master your semester with Scribd  the required translations and rotations to the bushings. Read Free Foron 30this Days Sign up to vote title  & The New York Times  Useful  Not useful Special offer for students: Only $4.99/month.

Cancel anytime.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 3.9.3: Bushing test configuration 1 - axial rotation You're Reading a Preview Unlock full access with a free trial.







Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 3.9.5: Bushing test configuration 3 - radial translation

Master your semester with Scribd Read Free Foron 30this Days Sign up to vote title & The New York Times Not useful Figure 3.9.6: Bushing test configuration - axial  translation  4Useful Special offer for students: Only $4.99/month.

 

Cancel anytime.

For test configurations 2 (Figure 3.9.4) and 3 (Figure 3.9.5), bushings were teste





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


The full results can be seen in Appendix F. When importing the curves into multibody model, the hysteretic properties were ignored by averaging the results the loading and unloading cycles, as shown shown by the red lines in Figure 3.9.7.

Figure 3.9.7: Averaging hysteresis loops

3.10. Characterisation of Bumpstops

The front and rear bumpstop stiffness curves were obtained in a similar way to springs. However, due to their propensity to buckling, a guide rod had to be inse though the specimens, as seen in Figure 3.10.1.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Large nonlinearities can be found both on the front and rear bumpstops as deflections start to become relatively large, as seen in Figure 3.10.2 and Fig 3.10.3.

Figure 3.10.2: front bumpstop force vs. displacement






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design




 Search document




Documents



Sheet Music


4. Multibody model implementation Once the component and system level parameters were measured, implementation of the ADAMS/Car model began.

4.1. Model characteristics

As mentioned in section 1.4.1, the main purpose of this model is to simulate primary ride and handling behaviour of the car. Bibliography suggests that complexity of the model should be adequate for the complexity of the problem to solved (see section 2.7).

Although initially limited track testing was scheduled in order to obtain experime data for the model correlation, eventually it was cancelled due to lack of resour Instead, a series of tests were planned in the road simulator, in which forces displacements would be applied to the suspension of the static vehicle. This wil further explained in section 5.

Therefore, having accurate powertrain and brake models was no longer immediate priority, other than to account for the effects of their mass propertie the dynamic behaviour of the vehicle. Similarly, default values were used for Reading Preview aerodynamic properties of the You're vehicle body aand the importance of the tyre mo was limited to simulating their radial deflection normal loads. Unlock full access with under a free trial.

In addition, a series of generalDownload assumptions and simplifications were made w With Free Trial implementing the model: 



All linkages and bodies were considered to be perfectly rigid (including chassis).

The hysteresis and frequency dependence of all suspension isolator mo was ignored by averaging the experimental results as shown in section 3.9

Master your semester with Scribd The preloads on the bushings & The New York Times operating conditions. 


Default

properties

(provided




were considered to be zero at nom  Useful  Not useful Cancel anytime. by

ADAMS/Car)

were

used

whene





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


4.2. Templates and model topology

ADAMS/Car could be considered as a "template-based" version of ADAMS/View database of pre-built templates is available to the user, which includes all subsystems required to build the model of a full car. Depending on the vehicle to modelled, some of these templates can be used with very little modifications. T 4.2.1 shows the pre-defined templates which were used in this model and modifications made to them. Original subsystem Rigid chassis Powertrain

Modifications Mass, CoG position, Inertias Location (front-mounted), position of differential tripots. Tyres (PAC2002 model for 205_55R16) Dimensions, static radial stiffness Brakes Dimensions Steering Topology (rack location, etc.), steering ratio Anti roll bar Duplication (one at the front, one at the rear), topologies Table 4.2.1: Modifications to original templates You're Reading a Preview

As far as this particular model is concerned, whenever enough data was availabl Unlock full access with a free trial. the topology and characteristics of a given subsystem, it was decided to build subsystem from scratch. Download With Free Trial

Based on the suspension hardpoints measured in section 3.2, the parts which w considered to be geometrically intricate were modelled in Catia V5 and subseque imported into ADAMS. The geometry for simple parts was created directly in ADA

In addition to ease of visualization, the primary reason for producing representa geometry was the fact that component inertias were unknown, and hence neede  be numerically calculated from 3D models. Read Free Foron 30this Days Sign up to vote title

Master your semester with Scribd & The New York Times This way, the topologies of front suspension Special offer for students: Only $4.99/month.

(Figure 4.2.2) were modelled.



useful  Useful  Not Cancel anytime. (Figure 4.2.1) and rear suspen





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music



Figure 4.2.1: Graphical representation of the front suspension model (MacPherson Unlock full access with a free trial. Download With Free Trial






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


It must be noted that, as mentioned in section 2.8.1, ADAMS calculates component inertias with respect to the absolute origin of a given model. This is reason why geometry was imported from CATIA V5 to ADAMS and why ine were calculated within ADAMS. The results of this process can be seen in appe C.

Once the parts and their geometries were produced, the next step was to join parts together appropriately. ADAMS/Car enables the user to build kinematic ( compliant) joints in such a way that they can be switched by bushings at a l stage by using a simple toggle button. This is particularly useful when comparing kinematic and compliant behaviour of a subsystem.

Figure 4.2.3 shows how the parts were joined together. If the diagram is conside to represent the plan view of the vehicle, the upper half (in blue) would represent front subsystems in the car and the bottom half (in orange) would represent the subsystems. Only the driver's side (left) is shown, along with single or non-symm parts (those which are inside the coloured rectangles).

Rigid bodies and embedded subsystems (such as the powertrain) are shown in ovals and the links between them are represented by lines: a single line for a sin joint, multiple lines for multiple joints. Each joint has a unique number. This num You're Reading a Preview is sometimes followed by a letter, representing the type of joint. A legend for th letters is shown in Table 4.2.2. The numbers which Unlock full access with a free trial. are not followed by a le correspond to force elements such as springs, dampers, bumpstops or rebo Download TrialIn addition, some of the stops; Table 4.2.3 contains the legendWith for Free these. numbers are followed by an asterisk: switchable kinematic and comp characteristics will be assigned to these, as mentioned above.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music




Figure 4.2.3: A schematic representation of the full vehicle topology C F

Constant velocity joint Fixed joint

H

Hooke joint

Master your semester with Scribd R Revolute joint Read Free Foron 30this Days Sign up to vote title S Spherical joint & The New York Times  Useful  Not useful T Translational joint Special offer for students: Only $4.99/month.

Cancel anytime.

*

Switchable by bushings

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print


1

Download



Magazines



News

of 112

 

Design


 Search document






Documents



Sheet Music


4.3. Modelling force elements

Whereas the nature of kinematic joints shown in Figure 4.2.3 is self explanatory ( Table 2.3.1 for more info on the DOF removed by each of them), the compliant jo or bushings, need additional parameters to define them.

As far as the front suspension is concerned, Figure 4.2.3 shows that the f subframe is rigidly fixed to the chassis at all times, effectively forming an integral of it. However, whenever the subsystem is simulated in compliant mode, all components attached to the subframe are linked by bushings.

When in compliant mode, the rear suspension has a different layout: the whole subframe is linked to the chassis through bushings. In addition, due to the natur the multilink suspension (see section 2.5 for more information), all the linkages w are connected to the subframe use bushings too.

One of the main issues encountered when creating the multibody model of the was that some bushings could not be characterised, as it would have b necessary to destroy some suspension components in order to extract them f their housings. Table 4.3.1 shows how the different bushings have been mode and which assumptions/simplifications have been made in the process. You're Reading a Preview ID

Name

Method


4 fsubframe to powertrain Default ADAMS/Car values 9a fsubframe to LCA (front) Estimated Download With Free Trial 9b fsubframe to LCA (rear) Experimental measurements (pp. 137 to 140) 10 fsubframe to steering rack Default ADAMS/Car values 17 fsubframe to FARB Experimental measurements (Figure 3.8.6) 19 chassis to strut Estimated 22 chassis to rsubframe Estimated 23 rurt to rdamper Default ADAMS/Car values 27 lateral link to rurt Copied from "rsubframe to rtrod" 28 rsubrame to lateral link Experimental measurements (pp. 133 to 136)  Read Free Foron 30this Days Sign up to vote title  29 trailing arm to chassis Experimental measurements (pp. 123 to 126) Useful  Not useful  Cancel to anytime. 32 upper link to rurt Copied from "rsubframe rtrod" Special offer for students: Only $4.99/month. 33 rsubframe to upper link Copied from "rsubframe to rtrod"






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Whenever the material properties, shape and dimensions of a bushing were kno existing mathematical models were used to estimate its stiffness in every directio

Bushing models of increasing complexity have been developed over the years; f simple quasi-static approximations such as[47], [48],[49] and[50], to frequency amplitude-dependant models [51].

In this case, the load-deflection relationships proposed by Adkins et al [47] Horton et al. [48] [50] [49] were deemed sufficient to estimate the static stiffness cylindrical solid rubber bushing. Two of the assumptions made in these relations are that the bushing does not contain any asymmetric features such as voids, that the rubber is perfectly isotropic. Figure 4.3.1 shows the nomenclature used the models.


Figure 4.3.1: Geometrical parameters of the bushing model Download With Free Trial

Let there be any "radial" axis which lies on the xy plane (Figure 4.3.1) and cros through the origin. If it is assumed that the bushing is axisymetric with respect to and is clamped on its outer face, the moment required to rotate the inner face (small) angle β around any radial axis (a.k.a. "conically") is named and be approximated by:






Useful

 

where "G" is the shear modulus of the rubber, "l" is the length of the bushin





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


where the value of constant " " is dependent on whether "l" is very large ( very small ( (

). In order to use this formula on bushings whose aspec

) was roughly 1, the calculated stiffness for the two limiting cases was aver

On the other hand, the axial moment which produces a (torsional) deflection of degrees around "z" is:

Finally, the axial force required to deflect the inner face by "ω" millimetres along is given by:

where:


and



Table 4.3.2 shows the approximate dimensions and material properties of the th bushings whose stiffness was estimated by this method. ID 9a 19 22

Name fsubframe to LCA (front) chassis to strut chassis to rsubframe

l (mm) 50 25 70

a (mm) 20 70 30

b (mm) 8 50 10

Material rubber polyurethane rubber

Master your semester with Scribd ReadupFree Foron 30this Days to vote title Table 4.3.2: Approximate characteristicsSign of non-tested bushings & The New York Times  Useful  Not useful must be that these Special offer forItstudents: Onlynoted $4.99/month.

Cancel anytime.

G (MPa) 1.5 100 1.5  

values were calculated assuming that the bushings w completely uniform in shape. However, the "chassis to strut" bush/mount had a s





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


angle compliances (Figure 4.3.2). The effect of these stiffnesses on suspen compliances will be further explained in section 6.1.1.

Figure 4.3.2: Top view of the LCA in a front MacPherson suspension, showing the effec the front (element D) and rear (el. 4) LCA bushing stiffnesses on compliant toe angle caused by braking and rolling resistance loads [12].

As far as the "chassis to rsubframe" bushings is concerned, it was noticed that t You're Reading a Preview construction was not axisimmetrical, as they were reinforced by a lump of rub full accessno with more a free trial.data was available on along one radial direction. Unlock However, construction and it was decided to model them as axisimmetric bushings. Download With Free Trial

On the other hand Table 4.3.1 shows that some experimental measurements w reused for various bushings. This resource was used whenever more than bushing shared similar characteristics, such as shape, material, and construction

Bushing damping values were set to 1% of the corresponding stiffness in all ca (see section 2.6 for more information).

Master your semester with Scribd  Subsequently the remaining force elements were Read included in model Free Forthe 30this Days Sign up to vote on title by assig  the stiffness measured in the experimentaltests to springs, dampers, ant & The New York values Times Useful  Not useful andOnly bumpstops. Special offer forbars students: $4.99/month.

Cancel anytime.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


4.4. Static suspension alignment

Once all force elements were modelled, the next step was to assign static cam and toe values to the front and rear suspension, using the measurements carried in section 3.4. The alignment of both axles was considered to be symmetrical, he the toe and camber values for the left and right wheels were assumed to be eq Table 4.4.1 shows the input values. Front suspension Rear suspension

Camber (deg) -1 0

Toe (deg) 0 0

Table 4.4.1: Static camber and toe values at nominal operating conditions

The reason for the lack of decimal figures in the aforementioned values is significant error of the experimental measurements. Repeatability betw measurements was poor, and it was noticed that the camber and toe readings on rear right corner were consistently offset, even when measurements were carri on different vehicles. This suggested that the FACOM® GTR 300 suspen geometry set-up device was either poorly calibrated or faulty. In order to overco this problem, approximate figures were used at the expense of accuracy You're Reading a Preview simulation. Unlock full access with a free trial.

4.5. Modelling full vehicle mass properties Download With Free Trial

Once all the suspension components were modelled and their individual m properties were known, the next step was to define full vehicle inertias and position of the vehicle centre of gravity.

The main difficulty when setting full vehicle mass properties can be illustrated Figure 4.5.1; the full car model consists of many subsystems, each having its mass properties and contributing in a different way to the full car. 

Master your semester with Scribd Read Free Foron 30this Days Sign up to vote title  & The New YorkanTimes Useful Not useful experimen To mention example, the full car CoG location  had been  measured Special offer for(section students: Only $4.99/month. 3.3), and therefore

Cancel anytime.

became the target to be achieved for the model. In o to adjust the full vehicle CoG location, the mass of subassemblies would have ha





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 4.5.1: Subsystems and full vehicle mass properties

Another issue found during this process was the fact that the full vehicle ine were unknown. As mentioned in the literature review, specific equipment is requ to measure this, which the sponsoring company did not have access to.

It was decided that the most straightforward way to overcome this problem wa refer to bibliography. The National Highway Traffic Safety Administration's (NHT Light Vehicle Inertial Parameter Database [52] was used to obtain data on the typ moments of inertia of passenger cars. Appendix G shows the normalized roll, p and yaw inertias of a variety of road vehicles of different masses. The normaliza is achieved by dividing the measured inertiaa Preview by (for Ixx) or by You're Reading

(for Iyy and Izz), where "M" is the vehicle mass, "T" is the track and "L" is Unlock full access with a free trial. wheelbase. Download With Free Trial

Knowing that for the Passat, under nominal operating conditions, M=151 T≈1551,5mm and L=2709mm, the values in Table 4.5.1 could be estimated. Normalized Absolute

Ixx 0.7 600 kgm2

Iyy 0.9 3000 kgm2

Izz 1 3000 kgm2

Table 4.5.1: Normalized and absolute vehicle inertias from bibliography [52]

Master your semester with Scribd & The New York Times 4.6. Special offer for students: OnlyAdditional $4.99/month.

requests





Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


5. Road simulator testing 5.1. Characteristics of the road simulator

MTS 329® 6DOF road simulators are generally used to recreate the multia stresses that the suspension assemblies encounter during their operating life. R life events such as cornering, braking, road disturbances or any combination of th can be recreated in a controlled environment, which improves the repeatability accuracy of tests as opposed to road testing. Due to these characteristics, th kinds of rigs are specially suited for component fatigue testing [53].

The layout of one corner of the model 329 road simulator can be seen in Figure 5 The road simulator in CITEAN is comprised of one axle (two corners), the remai axle being supported in a static platform.



Master your semester with Scribd Figure 5.1.1: MTS 329 Road Simulator layout (one corner) [53]  Read Free For 30this Days Sign up to vote on title  & The New York Times Useful  Not useful  The vehicle can be set either in a "floating" configuration, in which the

veh chassis is free to move, or in a "clamped" configuration, in which the chassis is f


Cancel anytime.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


All moments and forces are measured by a MTS SWIFT® wheel force transd (WFT) bolted to the wheel hub (see appendix H). These measurements can be u to control the actuators in a closed loop. Each of the six DOF of each corner can controlled in force or displacement mode. In displacement control mode, the u input is the position of the spindle along a particular DOF and the actuator movem matches the input. Whenever a DOF is controlled in force mode, the user inputs spindle forces and moments, and the relevant actuators constantly adjust position to match the WFT readings to the target values.

As far as the force control mode is concerned, it must be noted that the co system will tend to induce an oscillatory motion on the actuators in its quest to m the target force readings. Although this might not be an issue when the vehicle is to a clamped configuration, it could produce large, uncontrollable oscillations if vehicle was in a floating configuration and large asymmetrical forces were app laterally.

When the vehicle is in a floating configuration, the risk of uncontrollable oscillat can be reduced by setting one corner of the car to force control while the o corner is kept in displacement control. You're Reading a Preview

5.2. Using the road simulator to validate the multibody mode Unlock full access with a free trial.

Although originally conceived for fatigue testing, MTS 329 road simulators h Download With Free Trial previously been used for non-standard applications such as K&C testing, mentioned in section 2.8.2.1. In that instance, in order to achieve an accept accuracy, the vehicle body was clamped to the floor and wheel vector sensors w used to measure the position and orientation of the wheel at all times [33].

As far as this project is concerned, the lack of access to a four-post rig and the of time to carry out track testing meant that interest was not only limited to the  testing capabilities of the road simulator, but aRead way measuring Free For 30this Days Sign up toofvote on title the dyna  responses of the car was also required. Therefore new  ways of using the Not useful  Useful Cancel anytime. model validation were implemented, producing a series of ver Special offer forsimulator students: Onlyfor $4.99/month. excitation tests and compliance tests.






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print


1

Download



Magazines



News

of 112

 

Design


 Search document






Documents



Sheet Music


5.2.1.

Vertical excitation tests

The relatively high-frequency, high-amplitude operating capabilities of the r simulator along the vertical axis meant that, if the vehicle was set to a floa configuration, the rig could be used effectively as a 4-post shaker. Two main iss were brought into play when considering this application.

Firstly, and as mentioned before, CITEAN's road simulator is comprised of only axle, which meant that only a limited number of excitation modes could be applie the car.

Secondly, the road simulator is not conceived to be a precision instrument. actuators are joined to the rockers and pushrods through compliant joints, w means that input displacements are not as accurate as in a 4 post-rig (however forces measured by the WFT are very accurate).

Once these issues were understood, it was decided that the resulting inaccura would still be acceptable for the purposes of validating the model behaviour at h amplitude, low-frequency road disturbances. Inaccuracies due to the constructio the road simulator were expected to become larger as input frequencies increa However, it must be noted that the multibody model was not adapted to simulate vehicle response at high input frequencies accurately. You're Reading a Preview In addition, if track testing been carried out, the roughness of the tarmac would have added a high freque Unlock full access with a free trial. component anyway, so the road simulator option was assumed to be the best op available given the absence of aDownload 4 post rig. With Free Trial 5.2.1.1.

Experimental setup

The general layout of the experiment can be seen in Figure 5.2.1. The car had t raised into position by a bridge crane. In order to avoid damaging the c underbody and bodywork with the lifting straps a structure was built out of stand UPN channel section beams and machined nylon blocks. The structure  dimensioned so that the jacking points in the underbody of on30 the car Read Free For Days Sign up to vote this title rested on  nylon blocks, while having enough overhang on both sides of car to keep Useful Notthe useful  Cancel anytime. straps away from the bodywork. Special offer forlifting students: Only $4.99/month.






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


The underbody structure was lowered but not removed, so that the vehicle would fall to the floor in case the road simulator failed. Subsequently the position of road simulator actuators was fine-tuned to reduce residual forces and mom being measured by the WFT. At the same time the attitude of the car chassis monitored by the accelerometer/gyro system in order to maintain it close to s operating conditions. This procedure was used for both axles.



Figure 5.2.1: Experimental of the road simulator (front and rear axles shown i Master your semester withsetup Scribd  Read Free Foron 30this Days Sign up to vote title operation, respectively)  & The New York Times  Useful  Not useful

setting up the experiments wasCancel theanytime. fact that all the inbo sensors (shown in section 3.1) were connected to one data acquisition syst

issue when Special offer forOne students: Onlyfound $4.99/month.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


5.2.1.2.

Procedure

Four different vertical excitation tests were carried out: 







A “Front Parallel Wheel Travel” (FPWT) test, in which both wheels of the f axle were moved in phase.

A “Front Opposite Wheel Travel” (FOWT) test, in which both wheels of front axle were moved in inverse phase. A “Rear Parallel Wheel Travel” (RPWT) test, in which both wheels of the axle were moved in phase.

A “Rear Opposite Wheel Travel” (ROWT) test, in which both wheels of rear axle were moved in inverse phase.

The same actuator control modes were used for both corners in all four tests. Th are shown in Table 5.2.1. DOF Control mode (target value) Dx (longitudinal position) Displacement (zero) Dy (lateral position) Displacement (zero) Dz (vertical position) Displacement You're Reading a Preview (input) Rx (camber angle) Force (zero) Unlock full access with a free trial. (zero) Ry (wheel rotation ) Displacement Rz (steer angle) Force (zero)

Download With Free Table 5.2.1: Actuator control modes forTrial vertical excitation tests

From Table 5.2.1 it can be seen that the only input for the vertical excitation tes the vertical position of the wheel. When deciding which input signals to use, t main possibilities were studied: using pre-recorded road input signals, usin random white noise signal, or using a sinusoidal input. The last option turned o be the easiest one to generate and post-process.

Master your semester with Scribd  Free Foron 30this Days Sign to vote title low-freque It was decided that the input signal had to haveRead a up high-amplitude,  & The New York useful  Useful  Notas content that Times could be used for time-domain analysis, as well a high-freque Special offer forcontent students: Only that$4.99/month. could be used

Cancel anytime.

to analyse frequency-dependant effects. In order to k the actuator velocities within a controllable range, the amplitude had to be redu





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


and

where "A" is the peak (initial) amplitude, "n" is the number of samples, " starting frequency and "

" is

" is the finishing frequency.

A 5 cycle-long standard sine wave of amplitude "A" and frequency "

" was

before the start of the swept sine, so that redundant low-frequency data would available for repeatability checks during time-domain analysis. Finally the signal tapered at the beginning and at the end.

As far as the parallel wheel travel (PWT) tests were concerned, a (relatively la peak amplitude of 50mm and a frequency range of 0.5Hz to 10Hz were deem appropriate. It must be noted that, in PWT tests, the vehicle was expected to ro in "pitch" around the fixed axle, so large displacements did not suppose a struct risk for the vehicle. However, for opposite wheel travel (OWT) tests, the p amplitude was reduced to 20mm in order to avoid applying excessive stresses You're Reading a Preview the anti roll bars and the chassis, caused by the "warp" motion. Unlock full access with a free trial.

The resulting signal for PWT tests is shown in Figure 5.2.2. Download With Free Trial






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


Step 1 2 3 4

Time (s) 3.5 - 12 12 - 22 22 - 45

Description Amplitude is increased progressively from 0 to 50mm at 0.5Hz. Amplitude (50mm) and frequency (0.5Hz) are kept constant for 5 cycle Frequency is increased linearly from 0.5 to 10Hz during 23s. Amplitud decreased from 50 to 0.5mm, its rate of decrement being defined by A 45 - 46 Amplitude is tapered to zero. Table 5.2.2: Description of the input signal for PWT vertical excitation tests

5.2.2.

Compliance tests

As far as measuring the compliance of the suspensions is concerned, aforementioned K&C test rig “conversion” of the road simulator [33] was used a example. However, due to the limited time and resources available, no jigs could fabricated to clamp the vehicle chassis to the floor. In addition, no resources w available to purchase a wheel vector sensor. A crude way to work around th issues was to leave the vehicle in a floating configuration, apply a series of disc static loads to the wheels and to measure the wheel plane displacements by different methods: 

Using a CMM to measure the wheel plane position with respect to a refere You're Reading a Preview part in the vehicle chassis (e.g. the front subframe). Unlock full access with a free trial.



Using the absolute wheel plane positions calculated by the road simul Download With Free Trial control system.

In order to carry out these tests, the experiment setup shown in section 5.2.1.1 reused.

Table 5.2.3 shows the forces and moments used in the front axle tests. Th separate tests were carried out. Firstly a lateral force was applied at the centre single wheel, while the other corner was kept fixed at zero displacement. It  Read Free Foron 30 Days Sign up to vote title corners w assumed that, due to Newton's third law, the reaction forces inthis both  Useful Not useful   be roughly the same. When the displacements were being measured, it had to Cancel anytime. Special offer fortaken students: Onlyaccount $4.99/month. into that they were caused by the deflections of both corners of axle, as the vehicle body was not fixed. The displacement of each corner






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Load Range Lateral force (Fy) -3kN (outboard) to 4kN (inboard) Overturning moment (Mx) -1kNm (increasing camber angle) to 2kNm (decreasing cambe Aligning moment (Mz) 0Nm to -300Nm (increasing toe in) Table 5.2.3: Load ranges for front compliance tests

The load ranges had to be broad enough to acquire good quality data, w discarding any risks of component failures. A series of assumptions were m when estimating the appropriate ranges:

For the initial lateral force and camber moment estimations, a perfectly ne handling bicycle model [10] was used, operating at a (very optimistic) 1G ste state cornering situation. Static vertical wheel loads of 4kN and a wheel radiu 0.5m were assumed, and the camber moment (2kNm) was obtained by multipl the estimated lateral force at the contact patch (4kN) by the wheel radius.

Subsequently, it was assumed that the inboard wheel in a corner would encou smaller loads, so slightly smaller figures (-3kN and -1kNm) were used in this case

As far as the steer moment calculations are concerned, it was known that the f trackrod had withstood peak axial loads of 2kN in fatigue tests which were previo carried out for another project within CITEAN. The distance from the outb Reading a Preview trackrod balljoint to the centreYou're of the wheel spindle was approximately 150 therefore a steer moment of 300Nm was deemed Unlock full access with a freeappropriate. trial.

The modified actuator controls for Fy, Mx and Mz tests are shown in Table 5 Download With Free Trial Table 5.2.5 and Table 5.2.6, respectively. It must be noted that the co nomenclature in the tables corresponds to when the front axle was attached to road simulator. When the car was turned around for rear compliance tests, the front corner became the right rear and right front became the left rear.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


DOF

LF Control mode (target value) Dx (longitudinal position) Displacement (zero) Dy (lateral position) Force (input) Dz (vertical position) Displacement (zero) Rx (camber angle) Force (zero) Ry (wheel rotation ) Displacement (zero) Rz (steer angle) Force (zero)

RF Control mode (target value) Displacement (zero) Displacement (zero) Displacement (zero) Force (zero) Displacement (zero) Force (zero)

Table 5.2.4: Actuator control modes for the front Fy compliance test

DOF

LF Control mode (target value) Dx (longitudinal position) Displacement (zero) Dy (lateral position) Displacement (zero) Dz (vertical position) Displacement (zero) Rx (camber angle) Force (input) Ry (wheel rotation ) Displacement (zero) Rz (steer angle) Force (zero)

RF Control mode (target value) Displacement (zero) Displacement (zero) Displacement (zero) Force (input) Displacement (zero) Force (zero)

Table 5.2.5: Actuator control modes for the front Mx compliance test

DOF

LF Control mode (target value) full access with a (zero) free trial. Dx (longitudinal position)Unlock Displacement Dy (lateral position) Displacement (zero) Dz (vertical position) Displacement (zero) Download With Free Trial Rx (camber angle) Force (zero) Ry (wheel rotation ) Displacement (zero) Rz (steer angle) Force (input) You're Reading a Preview

RF Control mode (target value) Displacement (zero) Displacement (zero) Displacement (zero) Force (zero) Displacement (zero) Displacement (zero)

Table 5.2.6: Actuator control modes for the front Mz compliance test

As mentioned before, once the front axle tests were finished, the vehicle was tur around and the same tests were carried out on the rear axle. However, having s  accu that the method for CMM measurements was very time-consuming for the Read Free Foron 30this Days Sign up to vote title  it provided, it was decided that the wheel positions beuseful acquired from Useful only Not  would  Cancel anytime. simulator control system. In this case, using discrete, static loads was no lo Special offer forroad students: Only $4.99/month. required as there was no need to stop for CMM measurements to be ta






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print


1

Download



Magazines



News

of 112

 

Design


 Search document






Documents



Sheet Music


6. Multibody model validation

It has been previously stated that, whenever possible, the experimental data wa be compared with simulated data in order to validate the multibody model. simulated data was based on a range of quasi-static, single axle suspen parameter analyses and full vehicle dynamic analyses, to be described in section.

6.1. Suspension parameter analyses

In order to extract system-level parameters from the multibody model, front and suspensions were analysed individually by carrying out a series of singleanalyses. The data obtained from the simulations was useful to understand general characteristics of the suspension model. However, not all suspen parameter simulations had an equivalent experimental test, and therefore could be directly correlated. 6.1.1.

Correlation of suspension compliance simulations

6.1.1.1.

Front suspension compliances

The front suspension compliance tests (section You're Reading a Preview 5.2.2) provided a variety experimental data which was to be compared with the multibody simulations. W Unlock full access with a free trial. carrying out multibody simulations, the loads shown in Table 5.2.3 were applie the wheels in the form of quasi-static sweeps. Download With Free Trial

However, the exact boundary conditions of the experimental tests could not replicated using the default simulation options provided by ADAMS/Car. experimental tests were carried out with the chassis set to a floating configura whereas the multibody simulations clamped the body to the ground. The de boundary conditions of the simulations could not be edited, as this would req advanced programming skills and the time and resources available were limited.  Read Free Forbeen 30this Days movement of the chassis in the experimental tests measured, w Signhad up tonot vote on title  Useful  Not useful made it impossible to apply any corrections to the multibody simulations. Cancel anytime.


In order to compare different sources of data, the simulated results were plo





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 6.1.1: Comparison of lateral force compliance test results for the front axle

For lateral force compliance tests, the lateral stiffness of the wheel plane measu on the road simulator control system turned out to be roughly three times sm than the simulated stiffness (Figure 6.1.1). However, if the "noise" due to the p precision of the method is ignored, it can be seen that the CMM measurem Reading a Preview correlate remarkably well withYou're the simulated data. This would make sense if compliance of the road simulator mechanisms was being considered: the contr Unlock full access with a free trial. calculates the wheel plane position based on the displacement of the individ actuators, but ignores the deflection of the joints which connect the actuators to Download With Free Trial rockers, hence recording a larger-than-real wheel plane displacement.

However, the results for overturning moment compliance (Figure 6.1.2) suggest opposite: the results for both experimental methods are coherent, while difference between experiments and simulations becomes very large.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


While the camber angle at zero load is similar for all the measurements, experimental rates of change in camber angle are two orders of magnitude la than the simulated results. This could be related to the aforementioned difference boundary conditions between the experimental and the computational results. fact that the experimental measurements also take into account the movement o chassis would explain the larger compliances.

As far as the aligning moment compliance is concerned (Figure 6.1.3), experimental compliances are again approximately four times larger than simulated compliances. Noticeably, the compliances for the two experime methods are identical. However, considerable offset is observed on the steer an at zero load. Compared to the simulated offset (-0.12°), CMM results show exaggerated static toe out (-1.22°), while the road simulator control system indicating a toe in attitude (0.47°).



Figure 6.1.3 Comparison of aligning moment compliance test results for the front ax

Master your semester with Scribd  For 30this Days Sign to vote on title control sys An explanation for this offset could be the fact thatRead theupFree road simulator  & The New York Times Useful Not useful   measures the absolute position of the wheel plane while the CMM measures Special offer forrelative students: Only $4.99/month. position with respect

Cancel anytime.

to the vehicle body. The method by which the veh was positioned on the road simulator was not very accurate, and it is likely that





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


In order to check whether this lack of correlation was due to differences in boun conditions, or if there was anything fundamentally wrong in the multibody mod DOE study was carried out.

As mentioned in section 2.8.1, carrying out statistical design of experiments (DOE a common method when monitoring the effects and interactions of various in factors on some output responses. As opposed to the traditional method of swee the input factors one at a time, experimental design helps to reduce the numbe required tests by changing several factors simultaneously and using statistical t to determine their effect on the outputs.

In this particular case, the input factors consisted in a range of different bus stiffnesses and the output was the maximum camber angle obtained during simulations.

It was expected that a DOE study would demonstrate if correlation could achieved by simply adjusting the unknown parameters of the mutibody model. If turned out to be impossible, it would suggest that the lack of correlation was lar caused by the differences in boundary conditions.

The study was run in ADAMS/Insight. The aim of the study was to quantify the e of increasing and decreasing individual You're Reading abushing Preview stiffnesses on the cam compliance. In addition, the interactions between different stiffnesses and their e Unlock full access with a free trial. on the camber compliance had to be understood.

With Free Trial Before choosing the stiffnessesDownload to be iterated, preliminary "screening" analyses to be run in order to only select the factors (stiffnesses) which had the grea effects on the response, as this helped reduce the processing time.

Two scenarios were considered for the DOEs: in the first scenario the overtur moment was applied to a single wheel, while the other was left free; in sec scenario the overturning moments were applied symmetrically to both wheels (  similar way to what was done in the experimental This Read Free Foron 30this Days Sign up totests). vote title was done  understand the role of the steering rack on the camber compliances Not useful under diffe  Useful  Cancel anytime. conditions. Special offer forboundary students: Only $4.99/month.Again, it must be noted that none of these boundary condit corresponded with the experimental tests, as the chassis was kept fixed in






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Subsequently, the stiffnesses estimated in section 4.3 were used as nominal va for the chosen input factors. These stiffnesses were then scaled down and up u a "linear" model, in order to obtain a "lower" and a "higher" response for each in factor. In order to interpolate the values between the lower and higher respon the investigation strategy was set to "DOE Response Surface". In addition, e possible combination of factors had to be run so that the interactions between th could be understood. Therefore, the DOE design type was set to "full factorial".

If "v" is the number of values per factor and "f" is the number of factors, the num of simulation runs required for a full factorial DOE for each scenario is given by:

Although this number of runs is practical (it took a processing time of 7 minutes scenario using 5 simulation steps per run4), it can be seen that if quadratic or c models were used instead of the linear one, 2187 and 16384 runs would have b required per scenario, respectively. At the same time, the processing time w have increased exponentially as the order of the model was increased.

The analyses made it possible to quantify the effect of each individual stiffness the maximum simulated camber angles. Figure 6.1.4 and Figure 6.1.5 show You're Reading a Preview the value of the factors most effects are negative, meaning that increasing decrease the value of the response simulated camber value Unlock full(i.e. accessthe withmaximum a free trial. the change in camber, will be smaller if the bushing stiffnesses are increased). W the sign convention seems logical, it should beTrial noted that what the effects Download With Free showing is the change in response between the "higher" and "lower" values of given factor. Therefore the magnitude of this number will depend on the higher lower values used in each scenario.

When the overturning moment was being applied to a single wheel (1st scena the lower values of all the factors were set to 1/10th nominal, while the upper va were set to 10 times nominal. This range was expected to give a goodidea ab Read Free For 30this Days Sign up to vote on title whether the orders of magnitude of the bushing stiffnesses were reasonable.  Not useful  Useful  Cancel anytime. results for effects in this scenario (Figure 6.1.4, in blue) showed that the m Special offer forThe students: Only $4.99/month. factor determining the camber compliance was the axial stiffness of the steering






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 6.1.4: First scenario - effect on maximum simulated camber angle and chang required to achieve the target response for different bushing stiffnesses.

This can be understood by the fact that the outboard trackrod balljoint is not at same height as the wheel centre. Therefore, when the camber angle in one co changes, the steering rack moves laterally. Interestingly, Table 4.3.1 shows that steering rack bushings had You're not been at all, and that de Readingcharacterised a Preview ADAMS/Car values had been used. Unlock full access with a free trial.

If time and resources had been available, the correlation study would have b With Free Trial halted at this point and those Download stiffnesses identified by the DOE study as the m relevant would have been measured experimentally. Then, the multibody m would have been changed, and DOEs re-run, completing the loop. However, was not the case, and the default stiffnesses were left unchanged for the time be under full knowledge of the weakness of the multibody model in this field.

Once the different effects for the first scenario were understood, an optimisa

Master your with Scribd  analysissemester was carried out to see which factors, andRead by up how much, had Free Foron 30this Days Sign to vote title to be chan  in order to obtain the target response, which according experimental t & The New York Times Not useful  Useful to  the was in the Special offer forFigure students:6.1.2) Only $4.99/month.

Cancel anytime.

region of 1° of positive camber. The changes propose the optimisation analysis are shown in red within Figure 6.1.4. These changes





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print


1

Download



Magazines



News

of 112

 

Design




 Search document




Documents



Sheet Music


Design variables

Factor

Min

Max

Value

Nominal

Change

lca_front_radial

0.1

10

0.809

1

-19.1%

lca_front_conical

0.1

10

0.986

1

-1.4%

strut_radial

0.1

10

0.973

1

-2.7%

strut_axial

0.1

10

1.000

1

0.0%

strut_conical

0.1

10

0.385

1

-61.5%

rack_axial

0.1

10

0.385

1

-61.5%

lca_front_torsional

0.1

10

0.999

1

-0.1%

Design objectives

Response maximum camber

Min -1.263

Max

Value

1.086

1.012

Table 6.1.1: First scenario - change required to achieve the target response for differe bushing stiffnesses

When all the scaling factors are set to minimum (0.1), the DOE analysis predicts the maximum camber angle obtained during the simulation will be 1.086°. Simil when all scaling factors are set to maximum (10), the maximum simulated cam You're Reading a Preview angle is predicted to be -1.263° . If the scaling factors shown in the "value" col where to be used, the DOE predictions that Unlock full accessshow with a free trial.the maximum simulated a would be 1.012°, which would correlate closely with the experimental results, ev Download With Free Trial the boundary conditions were dissimilar.

The next logical step was to modify the bushing stiffnesses to reflect the chan proposed by the optimization analysis and to re-run the camber complia simulation. Disappointingly, scaling down the stiffnesses only increased maximum simulated camber angles from -0.82° to -0.58°.

The most likely causes for the lack of correspondence between the optimiza Master your semester with Scribd  Read Free For 30this Days Signhigher up to vote on title levels of analysis and MBS is a poor curve fit between the and lower  & The New York Times Not usefulAlthough u  Useful  model. factors, caused by the use of an inappropriate interpolating Special offer forhigher students:order Only $4.99/month. interpolation

Cancel anytime.

models, such as quadratic or cubic, might have impro the fit, the potential improvements were not deemed worthy of the dramatic incre





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


As far as the second scenario is concerned, the screening studies revealed that axial stiffness of the steering rack mountings was no longer an important factor the overturning moment was being applied simultaneously to both wheels, as lateral loads applied on the rack were cancelled out.

On the other hand, the addition of the LCA rear bushing stiffnesses to the grou candidates meant that the lca_front_torsional and strut_axial stiffnesses w discarded, due to their minimal effects. The scaling of the factors for the sec scenario was changed between 0.01 and 100. The reason for varying the factors two orders of magnitude was that the effect of the factors on the response considerably smaller, and the results of single-order-of-magnitude changes w have barely been perceptible.


Download With Free Trial Figure 6.1.5: Second scenario - effect on maximum simulated camber angle for differ bushing stiffnesses.

The results for the linear, full factorial, response surface study for the sec scenario can be seen in Figure 6.1.5. This time the most important factor turned to be the conical stiffness of the strut mount bushing. In a similar way to w happened with the steering rack bushings in the first scenario, it was revealed  Read Free Foron 30this Days Signcharacterised up to vote title the strut mount bushing had not been experimentally (Table 4.3.1).  Not useful  Useful  Cancel anytime. the other hand, the small effects meant that, when the optimization analysis Special offer forOn students: Only $4.99/month. carried out, the target camber compliance could not be reached, even if a sca






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


6.1.1.2.

Rear suspension compliances

The road simulator control system was the only information source for the compliance acquisitions, as mentioned in section 5.2.2. The decision not to use CCM for rear compliance tests was taken before correlation checks were carried and it was based purely on the fact that CMM measurements were very t consuming. This decision meant that no additional sources of data were availabl check whether the road simulator acquisitions were representative of displacements of the wheel with respect to the chassis.

As was done with the front suspension compliances, the load ranges used in experimental tests (Table 5.2.7) were used again for the multibody simulations. O more, the boundary conditions of the simulations did not match those of experimental tests, and this was taken into account when comparing the results.



Figure 6.1.6: Comparison of lateral force compliance test results for the rear axle

Master your semester with Scribd  Read Free Foron 30this Days Sign up to vote title  & The New York Times  Useful  Not useful The correlation for the rear axle lateral force compliances (Figure 6.1.6) turned o Special offer for students: Only $4.99/month.

Cancel anytime.

be even worse than in the front axle. The experimental displacements turned ou





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 6.1.7: Comparison of overturning moment compliance test results for the rear a

As far as the rear axle overturning moment compliances are concerned (Fig 6.1.7), the correlation between the simulated and experimental rates was m better than for the front axle tests, although the simulated rates were still twice large as the experimental ones. Again, this could have been related to You're Reading a Preview aforementioned differences in boundary conditions. Unlock full access with a free trial.

On the other hand, at zero load the road simulator control system measured -1 of camber, while the simulationDownload predictedWith -0.18°. Free The Trial differences in zero offset c have been related to the fact that static camber values were not accurately set in multibody model.

Finally, the same behaviour described above was found in the aligning mom compliance results, both in terms of rates and offsets (Figure 6.1.8).






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


6.1.1.3.

Conclusion on the compliance tests

Having spent a considerable amount of time and effort on characterising suspension components and building the multibody model, the suspen compliance simulations were the first point of contact with the real world. The re of the simulations and the experimental tests correlated very poorly. In addition, coherence between different tests was poor (Figure 6.1.9).


Unlock full access with a free trial. compliances for front and re Figure 6.1.9: A summary of the measured and simulated suspensions Download With Free Trial

These issues made it evident that:

a) The experimental method for measuring suspension compliances was approximation of an already approximate method [33]. The experimental t were not equivalent to the MBS tests, so the correlation study itself was reliable.

Master your semester Scribd  was b) The procedure with followed when implementing the multibody model Read Free Foron 30this Days Sign up to vote title  adequate when it came to modelling suspension compliances. & The New York Times  Useful  Not useful Special offer for students: Only $4.99/month.

Cancel anytime.

Expanding on the first point, it should be noted that the only sources of experime





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


stiffnesses should have been known, without exception. The fact that each bus has 6 stiffnesses (one per each DOF of its inner ring) meant that rough estim were being used for 54 parameters which had a direct effect on suspen compliances.

In addition, all bodies in the model were assumed to be perfectly rigid, when effect of component deflections on suspension compliances cannot be neglected any means [12].

With the time and resources available for this project, it was impossible to res these issues. However, when analysing the suspension compliance results, overall aim of this project must be taken into account. As stated in section 1.4.1 multibody model being implemented in this project must simulate the primary and handling responses of the actual vehicle. In addition, it is assumed that the l of detail of the model should simply match the complexity of the aim to be achie without exceeding it (section 2.7).

The next logical step was to question the effect of having poorly correl suspension compliances on the aims of this project.

It is generally understood that, when actual road disturbances are being conside the effect of suspension compliances on thea vehicle You're Reading Preview ride is paramount [12]. Ac road inputs apply all force and moment components to the wheel, and therefore Unlock full access with a free trial. compliances measured herein would come into play in a real life situation.

Download With Free Trial However, if only vertical force components were considered (e.g. in a ver excitation test), the effect of suspension compliances on vehicle response expected to be reduced.

The fact that track acquisitions had been discarded and the remaining experime data was to be acquired via vertical excitation tests meant that poorly correl suspension compliances were not expected to greatly affect the remai  correlation studies. Read Free Foron 30this Days Sign up to vote title

Master your semester with Scribd & The New York Times Therefore, the multibody model was left



Not useful  Useful the Cancel anytime. unchanged and validation proc Special offer for students: Only $4.99/month. continued, under full knowledge of the weaknesses of the model in this area.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


6.1.2.

Simulated front suspension kinematics and rates

In addition to the quasi-static force tests which were used to compute suspen compliances, ADAMS/Car offers a selection built-in vertical wheel travel tests which the wheels are moved up and down.

Although this type of single-axle simulations provide information on many syst level parameters, it was decided that only three would be considered in orde reduce the post-processing effort. These were:   

The change in camber angle with bump and roll . The change in toe angle with bump and roll. Ride and roll rates (stiffness).

These three parameters provide basic information on suspension kinematics vertical rates, and are usually the ones which have the greatest influence on dynamic behaviour of the car. It should be noted that an in-depth analysi suspension elastokinematics falls out of the scope of this project.

In order to obtain data on these parameters, the vehicle chassis (see Figure 4 was fixed to the ground, and quasi-static vertical displacements were applied to You're Reading a Preview wheels. Figure 6.1.10 shows the two tests which were carried out: a single-axle P Unlock full access with a free trial. test, in which both wheel positions were swept simultaneously, and a "roll+no load" test, in which the displacements and forces applied to the wheels w Download With Free Trial conceived to simulate a pure roll event.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


These tests were carried out with the suspension models set to both kinematic compliant configurations (see Figure 4.2.3), and the results were compared. 6.1.2.1.

Quasi-static front PWT simulations

For the front PWT tests, the suspension displacement range was set between mm (full droop) and 120mm (full bump), the zero being the nominal operating height. These numbers were chosen to ensure that substantial contact was m with the bumptops and rebound stops, so that their effect on suspension rates c be measured.

Figure 6.1.11, Figure 6.1.12 and Figure 6.1.13 compare the camber, toe and w rates of the kinematic and compliant configurations, respectively.


Download With Free Trial Figure 6.1.11: Simulated front axle camber angle vs. wheel position for kinematic an compliant configurations






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 6.1.13: Simulated front axle wheel rate vs. wheel position for kinematic and compliant configurations

It can be seen in Figure 6.1.11 that, as the wheel is displaced towards full dro camber angle increases towards positive. As the wheel moves towards full bu camber is initially decreased towards negative (good for cornering performance the outer strut is compressed in a turn) and subsequently increased from 30 onwards. This behaviour is typical in Macpherson struts, and is related to the that the strut acts as an infinitely long upper control arm (see section 2.5). You're Reading a Preview

Figure 6.1.12 shows that the toe angle increases positive (toe-in) with b Unlock full access with a freetowards trial. (positive displacement). The fact that this increase is relatively linear is beneficia the handling predictability of theDownload vehicle With [12]. Free In addition, this increment means t Trial as the body rolls in a corner, the slip angle of the outer wheel will increase, he generating more lateral force, as long as the tyre is working in the linear range.

Although the kinematic and compliant results for camber angle variation are sim there is a substantial difference in the results for toe angle. For the kinem configuration, the toe angle (0°) at nominal ride height corresponds to what specified in section 4.4. However the compliant configuration produces  a static Read Free Foron 30this Days Sign up to vote title angle of -0.1°.  Not useful  Useful  Cancel anytime. Special offer forGenerally students: OnlyFWD $4.99/month. road vehicle suspensions are designed to have some static to (positive toe) in the front suspension, as it provides straight line stability when






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Although the factors that produce differences between kinematic and compliant angles in the model are complex, they are mainly related to bushing behaviour. fact that the front and rear bushings in the LCA are not perfectly coaxial with rotation axis of the LCA means that radial displacements occur whenever the w is moved up and down. These radial displacements are not of the same magnit for the front and rear bushing, and therefore the whole LCA is translated and rota along the Y and Z directions, respectively. In addition the upright is also constra by a trackrod. Consequently, the compliant toe angle results differ from the kinem results.

As far as the wheel rates are concerned, a considerable offset can be seen betw the kinematic and compliant results: the compliant wheel rate at nominal ride he is 74N/mm, whereas the kinematic wheel rate is just 29N/mm. A straight-forw explanation for this is the fact that both LCA bushings and the FARB bushings h a torsional stiffness which effectively adds to the spring stiffness. When the mod run in kinematic mode, LCA and the FARB are constrained by frictionless revo joints, which do not increase the wheel rate.

Leaving this offset aside, both configurations produce similar curves with the s characteristic peaks. In order to explain these peaks, it must be mentioned ADAMS calculates the wheelYou're rateReading using athe compliance matrix method. Preview compliance matrix is a 12x12 matrix which describes the load vs. displacem Unlock full access with a free trial.

relations for the 6 DOFs of each of the two wheels in an axle, and it is computed each solution point in a simulation [4][6]. Download With Free Trial

Therefore it can be defined as:

In order to calculate the wheel rates, ADAMS uses the elements which correspon and in . In fact, the wheel rate is obtained by taking

Master your semester with Scribd inverse of these elements. & The New York Times






 Useful  Not useful Cancel anytime. Special offer forTherefore, students: Onlywhenever $4.99/month.a sudden change of vertical force occurs between two solu points, the result for fluctuates and it is shown in the wheel rate curve





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


6.1.2.2.

Quasi-static front roll angle simulations

For roll angle tests, ADAMS/Car fixes the chassis to the ground and app opposing displacement to the wheels. The magnitude of these displacements function of the roll angle range and the total axle load selected by the user. As suspension position at zero roll is a direct function of the normal load, it is vital this load is representative of real-world operating conditions of the vehicle. For f axle tests, a (relatively wide) roll angle range of -10°to 10° and a normal load of were deemed appropriate. The results for such input parameters can be seen in the next three figures:


Figure 6.1.14: Simulated front axle camber angle (Y axis) vs. roll angle (X axis) for Download With Free Trial kinematic and compliant configurations






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Figure 6.1.16: Simulated front axle roll rate vs. ro ll angle for kinematic and complian configurations

In Figure 6.1.14 it can be seen that initially the camber angle on the front left w decreases as the vehicle rolls towards that side (negative roll). As mentione section 6.1.1.1, this is beneficial for the cornering performance, as it impro camber thrust. However, and again, due to the inherent characteristics of Macpehrson strut, the camber You're startsReading to increase towards positive as the roll a a Preview further decreases below -3°. Unlock full access with a free trial.

It must be noted that ADAMS/Car measures camber angles with respect to With Free Trial vehicle body. Therefore, for Download independent suspensions such as this one, magnitude of the body roll angle has to be added to the camber angle in orde obtain the “inclination angle” (the “camber angle” of the wheel with respect to ground). If this factor was taken into account, the wheel inclination angles would positive for all body roll angle magnitudes greater than roughly 1°. This is detrime for the cornering performance and therefore it can be said that the cam compensation provided by this Macpherson suspension is not sufficient for large  angles. Read Free Foron 30this Days Sign up to vote title

Master your semester with Scribd  & The New York Times  Useful  Not useful For the rest of the parameters, the trend is similar to what was seen in the PWT Special offer for students: Only $4.99/month.

Cancel anytime.

(section 6.1.1.1). In general, toe-in on the left wheel increases with bump (nega





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


configuration, the vertical tyre deflection is larger, which means that, for the same angle, the suspension displacement is smaller. 6.1.3.

Rear suspension kinematics and rates

The same simulations shown in section 6.1.1 were repeated for the rear axle. In case it must be noted that this kind of multilink suspensions are inherently o constrained by their design and depend on bushing compliances in order to m (see section 2.5). The reason why the kinematic configuration of the rear suspen had been fully modelled (see section 4.2) was to demonstrate that indeed the suspension would not have any remaining DOFs in kinematic mode. This confirmed as soon as the first kinematic simulations for the rear axle were attemp

Figure 6.1.17 shows that, for small displacements, rear PWT results are simila those of the front axle. The toe angle increases towards toe-in with bump, impro straight line stability. Meanwhile, camber angle decreases towards negative va with bump, which is beneficial for lateral force generation. However, displacements become larger towards full bump, the inversion in camber gain characteristic of Macpherson suspensions, is no longer encountered and the cam angle keeps decreasing progressively until it reaches nearly -2.3°.

As far as the wheel rate is concerned, a relatively constant stiffness of approxima You're Reading a Preview 28N/mm is obtained for small displacements. Full droop is reached at -77mm, Unlock full access with a free trial. contact with the bumpstops is made at 29mm of suspension displacement, w can be appreciated as a small step in Figure Download With6.1.17. Free Trial






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


toe angle curve when the car is rolling to the right. These phenomena occu roughly ±6° of body roll.

This might be explained by the fact that the rebound stop forces are transmitte the wheel through the damper (Figure 4.2.2). The top mounts of the rear damp are tilted inboards and forwards with respect to the vehicle chassis, and there any forces transmitted through them have lateral components. In addition, the e rear subframe is mounted to the chassis through compliant bushings. After obser the subframe bushing displacements, it was discovered that the whole subframe tilting and moving laterally by as much as 0.5mm throughout the roll angle simula When one of the rebound stops was reached, the rate of displacement (velocity the subframe suddenly increased (the rebound stops are relatively stiff). The fact the subframe couples the displacements of the left and right corners of the means that this is the origin of the kinks seen on the camber and toe angle curve the opposite corner.

On the other hand, left wheel toe angle is increased towards positive (toe-in) as chassis rolls to the left. This is typically done by automakers to ensure that the axle induces understeer when the car rolls into a corner. Leaving optimum corne performance aside, an understeering car is desirable for the average driver a behaves more predictably [12]. You're Reading a Preview

As far as roll stiffness is concerned, the rear rolltrial. rate is in the region of 1.06 Unlock full access with a free Nmm/deg for small roll angles. The steps at ±3.6° of roll correspond to the bump With Free Trial at ±6°, as mentioned abo contact points, contact with the Download rebound stops occurring






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


6.2. Full-vehicle excitation analyses

Once the system-level suspension parameters were studied, the next step wa understand the vehicle-level behaviour. In order to achieve this, the experime method for full vehicle analyses (section 5.2.1) was compared with MBS. 6.2.1.

Full-vehicle simulation characteristics

As it had happened in section 6.1.1, the default ADAMS/Car analyses did not a for the experimental boundary conditions to be replicated exactly. In experimental set-up for vertical excitation tests (section 5.2.1.1), the front axle rigidly attached to the road simulator through the wheel hubs, while the rear wh were strapped to static platforms.

ADAMS/Car allows the user to choose between applying the forces at the w contact patches or at the hubs, but this option has to be changed simultaneously both the front and rear axles. Modifying this configuration would have requ considerable time and effort which could have not been afforded.

In the experimental tests, both the WFT and the road simulator actuators w mounted on the same axle (Figure 5.2.1). This meant that the forces that had b measured belonged to the axle which wasa Preview constrained through the spindles You're Reading addition, the platforms to which the remaining axle was attached were not Unlock full access with a free trial. horizontal plates, but “U”-shaped fabricated structures. If, in the simulations, forces had been applied at the contact patches, ADAMS/Car would have assum Download With Free Trial that the tyres were resting on a flat surface. Therefore it was decided that the l detrimental solution would be to apply the forces at the wheel hubs (spindles) fo four wheels.

The input vertical displacement signals for the simulations were the same as th used in the experimental tests (Figure 5.2.2). The setup of the simulations can seen in Figure 6.2.1. It should be noted that, although circular plates are sh underneath the tyres, the forces were applied at the wheel hubs, astitle said  before. Read Free Foron 30this Days Sign up to vote  yellow sphere behind the handwheel represents the body. Useful  Not useful  vehicle


Cancel anytime.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


Figure 6.2.1: Graphical representation of the computational setup You're Reading a Preview

6.2.2.

Unlock full access with a free trial. Data post-processing

Once both the experimental and computational Download With Freetests Trial were run, the next step wa convert all the raw data into usable data (Figure 6.2.2). This was carried out MATLAB environment, and the code generated for such purposes can be see appendix I.

Master your semester with Scribd Read Free Foron 30this Days Sign up to vote title & The New York TimesFigure 6.2.2: Data post-processing  Useful  Not useful flowchart Special offer for students: Only $4.99/month.

 

Cancel anytime.

The experimental and computational data had been saved in the form of





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print


1

Download



Magazines



News

of 112

 

Design


 Search document






Documents



Sheet Music


In order to achieve this, the data was filtered first in order to reduce the h frequency noise and hence minimise aliasing. Knowing that the largest frequenc the input signal was 10Hz, the “idealfilter” function [54] was used to remove al power outside the 0-20Hz frequency range.

Once the data was resampled, all units homogenised and the offsets removed, next step was to divide the acquisitions into sections which corresponded to different parts of the input signal (Figure 5.2.2). This step was especially impor as the acquisition starting times for the different data acquisition systems w different, as mentioned in section 5.2.1.1. The time vector of each individual divi was restarted from zero to cancel offsets in the time domain. 6.2.3.

Discussion of full vehicle excitation data

Understandably, analysing the results took considerably longer than carrying out tests themselves. Although both front and rear axles had been excited in the test simultaneously, as the road simulator did not allow it), only the front axle excita results were analysed due to the limited time available. Among the front axle res only four sets of parameters were considered. These were: suspen displacements, spindle positions, chassis accelerations and vertical loads at spindles. You're Reading a Preview

These parameters, although they represented a small fraction of all the Unlock full access with a free trial. available, provided enough information to understand the dynamic response of chassis to the excitations. Download With Free Trial

As the input signal consisted of two main parts (Figure 5.2.2), the acquisitions w divided accordingly into a "sine" part, which was to be analysed in the time dom and a "sweep" part, which was to be analysed in the frequency domain. The res for the different parameters are discussed below. 6.2.3.1.

Damper displacements and vertical wheel forces

Master your semester with Scribd  Read Free Foron 30this Days Sign up to vote title The full experimental and computational acquisitions for the damper displacem  & The New York Times Useful  Not useful  and vertical wheel forces are shown in appendix J. In general, the correla Special offer for students: Only $4.99/month.

Cancel anytime.

between the simulated and experimental results was better for the forces than fo





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music









Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music









Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


In order to quantify the fit between the simulated and experimental results for sinusoidal parts of the tests, the root mean square deviation (RMSD) between two signals was calculated:

where

is the experimental signal,

experimental sample,

is the simulated signal,

is the

is the i -th simulated sample, and n is the total numbe

samples. LF_Fz

RF_Fz

LF_damperpot

RF_damperpot

LR_damperpot

RR_damp

(kN)

(kN)

(mm)

(mm)

(mm)

(mm)

FPWT

0.15

0.21

1.97

1.92

0.21

0.28

FOWT

0.16

0.21

3.26

3.03

4.62

4.47

Table 6.2.1: RMSD of the simulated and experimental results for damper displaceme and wheel forces

As far as the front axle parallel wheel travel (FPWT) excitations are conce (Figure 6.2.3), rear damper acquisitions (first plot from the top) show very s You're Reading a Preview displacements, in the region of ±0.4mm. Therefore a relatively large proportion o experimental results is noise. Unlock full access with a free trial.

It can be observed that the rear dampers operate mainly in the nega Download With Free Trial displacement region, which is these tests represents bump (as opposed to results in section 6.1.2., for which bump was positive). The fact that the experime and computational curves correlate well in the bump region but not in the dr region suggests that the rear tyre, instead of the suspension, might be deflec under compressive loads. In fact, it was observed during all experimental tests ( FPWT and FOWT) that the rear axle was visibly moving around within the s  platforms, due to the compliance of the tyres.Read Obviously this deflection is Free Foron 30this Days Sign up to vote title  accounted for in the simulation, as the forces are applied to the wheels through Not useful  Useful  Cancel anytime. hubs. Special offer for students: Only $4.99/month.






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


the front suspension bushings were not experimentally tested. For instance, a po characterised torsional stiffness on the LCA bushings would have a considera effect on the resulting wheel rate.

Moving on to Figure 6.2.4, it can be seen that the trend for the FOWT test resu similar; the correlation between the spindle forces being better than the correla between the damper displacements.

For the rear damper displacements (first plot), the amplitudes of experime displacements are approximately 25% less than the simulated ones (a m deviation of less than 5mm). The most plausible explanation for this is that the stiffnesses of the model differ from reality, probably due to the aforementioned of experimental characterisation in some of the suspension bushings. In additio this, a phase shift of about 20° is found between experimental and computati results, suggesting that rear tyre deflection has again an important effec suspension displacements.

As far as the front damper displacements are concerned (second plot) the signals are in phase, but again the amplitudes are approximately half of each ot with a mean deviation of less than 4mm.

It should be noted that, both in You're the simulated experimental results, the ampli Reading aand Preview of damper displacements is considerably larger for the rear axle than for the f Unlock full access with a free trial. axle. This is a clear indication of the fact that the rear roll stiffness is smaller than front, which is a typical characteristic in cars forward weight distribution an Download Withwith Free aTrial understeering tendency. This corresponds with what was said in sections 6.1.2 6.1.3.

Interestingly, the vertical force plot shows that the behaviour of the left and corners in the front suspension is asymmetrical, when compared to the simul results. The reason for this is not well understood, but it is suspected that it coul related again to differences in the static weight distribution between the model  Read Free Foron 30this Days Sign up to vote title reality.  Not useful  Useful  Cancel anytime. to analysing the responses to the swept sine input sig Special offer forWhen students: it Onlycame $4.99/month. periodograms (essentially the Fast Fourier Transforms, or FFTs) of the ver






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


purposes of comparing the simulated and experimental data, the de periodogram plotting functionality for time series in MATLAB was deemed sufficie

You're Reading a Preview Figure 6.2.5: Simulated and experimental responses to FPWT excitations Unlock full access with a free trial. Periodograms of vertical forces for front left and right spindles Download With Free Trial






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


In general terms, the periodograms showed that the frequency spectra for the and right spindle loads were almost identical for both the FPWT and FOWT tests far as comparing the experimental and simulated data is concerned, the peak amplitude were located at similar frequencies, but the magnitude of these pe differed significantly between the experimental and simulated results.

For the FPWT tests (Figure 6.2.5) a single and clear resonant frequency was fo at 1.6Hz, and the amplitude progressively reduced thereafter as the freque increased. This peak amplitude corresponds to the mode at which the vehicle b started to resonate in a combined pitch and heave motion, rotating around so point close to the fixed rear axle.

On the other hand, the magnitude of the simulated peak response was roughly th times larger than the experimental. This effect was also observed in the time-dom results shown in Figure 10.10.3 (appendix J), in which the amplitude of the simul signal at around t=25s was considerably larger than the experimental.

Again, the most probable cause for these differences in magnitude is aforementioned lack of correlation in the experimental and simulated wheel rates

When it came to the FOWT tests, an extra peak was found in addition to expected body mode, both in the experimental and simulated results. The first, la You're Reading a Preview peak occurred at 0.8Hz, followed by a smaller one at around 3Hz. Again, Unlock full access with a free trial. experimental peaks had a considerably lower magnitude than the computatio ones, with the second peak being barely With noticeable in the experimental results. Download Free Trial

At this stage, additional sources of data were required in order to interpret s results. In this situation, chassis accelerations were the next point of call.

6.2.3.2.

Chassis accelerations

Master your semester with Scribd  Free Foron 30this Days In section 3.1, mention was made of the ADMA Read accelerometer/gyroscope sys Sign up to vote title  & The New Timesthat additional accelerometers and itYork was explained were placed at known point  Useful  Not useful Special offer forthe students: Only $4.99/month. vehicle, in order to

Cancel anytime.

back up the acquisitions of the first system. Subseque requests were made to measure the chassis accelerations of the multibody mod





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


The locations of additional accelerometers were described in section 3.1; two w placed in the top mounts of the front suspension struts, the remaining one be placed at the rear subframe. The fact that the rear subframe was allowed to m with respect to the chassis did not represent a problem, as the only purpose of th accelerometers was to compare the experimental and computational results.

In order to gain further understanding of the unexpected frequency spectrum sh in Figure 6.2.6, periodograms of the front right strut mount accelerations w generated for both FPWT and FOWT tests. Only the vertical component of s accelerations was considered at this stage. The results for the remaining accelerometers are not shown here, as they did not add any further information.



Figure 6.2.7: Simulated and experimental responses to FPWT excitations Periodograms of vertical accelerations at the front right strut mount.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


Looking at Figure 6.2.7, it can be seen that the frequency spectrum is very simila the one shown in Figure 6.2.5. This makes sense as the vertical loads at spindles should be directly proportional to the acceleration of the chassis, accor to Newton's second law.

A small increase in amplitude can be seen at the higher end of the frequency ra for the experimental results, which is not replicated by the simulations. This m have been caused by the fact that the accelerometers were stuck to the cha using a relatively thick layer of blue-tac, which allowed the sensor to move relativ the base.

As far as the FOWT results are concerned (Figure 6.2.8), the differences with Fig 6.2.6 are much more important. To start with, the peak at 0.8Hz has disappe and only the 3Hz mode is present in the simulated results. This would suggest the initial peak in Figure 6.2.6 might be related to the fact that the rate of chang frequency during the sweep was too large, meaning that not enough cycles w available for every discretized value of frequency, and hence producing a variet issues during the internal computation of the FFT.

Therefore, it could be said that the body mode when the vehicle is undergoin combined roll/warp motion is in the region of 3Hz, for both the experimental You're Reading a Preview simulated results. Unlock full access with a free trial.

On the other hand, the experimental results show an additional peak at roughly 6 with a valley at around 5Hz. Interestingly, a smaller version of this valley can also Download With Free Trial appreciated in the experimental results within Figure 6.2.6. The cause of th unknown at this stage.

6.2.3.3. A study on validating suspension kinematics and rates during vertical excitation tests.

Master your semester with Scribd  Read Free Foron 30this Days Sign up to vote title One of the issues of not being able to fix the vehicle chassis to the ground  during & The New York Times Useful  Not useful  experimental tests was the fact that no experimental data was available on Special offer for students: Only $4.99/month.

Cancel anytime.

suspension kinematics and vertical rates; it was not possible to apply quasi-s





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


However, if such acquisitions were compared with the simulated results un identical (dynamic) conditions, it was expected that valuable information could h been gathered on general trends and quality of fit.

For this purpose, efforts were focused only on the FPWT tests and on the kinema of the right front wheel. This way the amount of data processing required minimized and attention was focused on the most interesting parameters (dyna wheel rate estimations can also be achieved in a 4 post rig, whereas the dyna analyses of suspension kinematics are specific to the capabilities of a r simulator).








Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


The result of this is that, for the simulated results ("adams", in red), the camber toe angles are almost purely a direct function of suspension displacem represented by the relatively thin, linear curves. The fact that these are not per curves is related to residual force components rather than to history dependence

However, when it comes to the road simulator acquisitions, a consider hysteresis can be appreciated for both the camber and toe angle results. In orde interpret this, it should be noted that all DOF were operated in force control target being zero force), except the vertical translation, which was controlled by input displacement (see Table 5.2.1).

This meant that the actuators were required to adjust their position constant order to compensate for the changes in wheel plane orientation as the suspen moved up and down. Therefore, part of the hysteresis seen in Figure 6.2.9 could attributed to delays in high frequency actuator adjustments. A way of measuring quantifying this effect would be to carry out a similar test in a four-post shaker and to measure the wheel plane orientations with a wheel vector sensor. difference in the results would provide a good picture of the effect of actuator con in the road simulator.

On the other hand, it can be seen that the displacement range for the experime You're Readingrange. a Preview results is roughly one third of the simulated This is coherent with what it shown in Figure 6.2.5 and Figure 10.10.1, anda free is therefore not an inherent prob Unlock full access with trial. of the method described herein. Download With Free Trial

As far as the magnitudes of the camber and toe angle curves are concerned, it be seen that the correlation is poor, with large offsets between simulation experiment. However, the rates of change for these parameters are reasonably c

Overall, it could be said that this method is valid for predicting trends (e.g. to k whether the camber increases or decreases with bump). However, it is of very use if the magnitudes of the kinematic parameters are being measured. In such c  Read Free Foron 30this Days Sign up to vote title the use of a conventional K&C rig would be required.  Not useful  Useful  Cancel anytime. 6.2.3.4. Conclusion on the vertical excitation tests Special offer for students: Only $4.99/month.






Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


by fabricating a rigid, spindle-coupled support. However, and once more, time resources were scarce and this was not feasible.

Having said this, the purpose of the excitation tests was not to measure the abso magnitudes of vehicle-level parameters, but to obtain experimental data which c have been used as a reference to compare with the simulated results. When was considered, the excitation tests turned out to be a reasonable way to ach this aim in the absence of a four-post rig, using available resources.

As far as the quality of correlation between the experimental and simulated resul concerned, the results were generally better for the vertical forces at the spin than for the damper displacements.

It was assumed that, for both FPWT and FOWT tests, the most likely cause of issue was the fact that many of the front suspension bushings had not b experimentally characterised. However, this point was difficult to prove as experimental wheel rates could not have been measured accurately without fi the vehicle chassis to the ground.

Something similar happened in the frequency-domain analyses; the reso frequencies correlated very well in most cases, but the amplitudes did not. reasons for this were assumed to be related to inaccurate wheel rate model You're Reading a Preview Again, and for the aforementioned reasons, there was no way of demonstrating th Unlock full access with a free trial.

Finally, the method of using the excitation tests to measure suspension kinema With Free Trial only proved to be good enoughDownload for indentifying trends, but the correlation was poor.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


7. Conclusion and future work

The complexity of implementing a multibody model of a passenger car is determi by the complexity of its required function. In this case, the function of the model to simulate the handling and primary ride responses of the vehicle.

In order to implement a multibody model whose aim is to replicate reality, the m building phase must be followed by a correlation phase. During the build phas this model the constituent components were characterised, while in the correla phase the system- and vehicle-level responses of the model were compared experimental data.

As the function of the model was limited to simulating the low-frequency respon of the vehicle, history and frequency dependence was ignored in all the compon stiffness characterisation tests. Similarly, the characterization of tyres was limite measuring their static vertical stiffness. In addition, several subsystems, such as powertrain or the brakes, were not characterised at all, as the model was required to perform any road manoeuvres at this stage. It should also be noted all bodies in the model were considered to be perfectly rigid.

While some of the model limitations were attributable to conscious decisions limiting the model complexity, others were not. In this respect, the shortag available time and resources played an important role in the model implementa process.

Firstly, several bushings (most of them located at the front suspension) could no experimentally characterised, and mathematical models or plain estimations w used instead. Secondly, the only way to measure the mass properties at compon and system level was to either reverse engineer the parts in CAD and use nume integration methods, or to resort to existing literature. Thirdly, the static nom wheel plane orientations could not be accurately measured, due to equipm limitations. 

Master your semester with Scribd Read Free Foron 30this Days Sign up to vote title  & The New York Times Useful Not useful   Once the initial model had been built, two sets of experiments were carried ou Special offer for students: Only $4.99/month.

Cancel anytime.

first set in which quasi static forces and moments were applied to the wheels an





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


The resulting correlation in both sets of tests was not always good, partly due to fact that the boundary conditions for the experimental and simulated results were equivalent. The reason for this was that the jigs required to apply the diffe constraints could not be built.

On the other hand, the static bushing stiffnesses had not been thorou characterised in the first place, which resulted in the magnitudes of the m compliances and responses differing from those of the actual vehicle.

Although the correlation was not always good, these two sets of tests turned ou be very valuable as they provided evidence on the weaknesses of both the mo implementation and testing methods. Based on the aforementioned res important information was gathered on the behaviour of both the vehicle and model, and basic knowledge was established on how to optimize the valida method.

The results showed that the use of K&C and 4-post 4 -post shaker rigs is completely just when it comes to validating multibody models. Nevertheless, it was also conclu that, in the absence of these devices, road simulator testing would have been ab provide reasonably accurate data if the correct boundary conditions could have b applied. This would have included fixing the vehicle body to the ground for compliance tests and cancelling the effect of tyre stiffness for the excitation tests.

Overall, the main lesson learnt from this project was that it is not possible to shortcuts when implementing and validating a multibody model. If one or m parameters are unknown in the model, it ends up being difficult to ascertain whe the lack of correlation in the validation tests is caused by errors in mo characterization or by inaccuracies in the methodology of the experiment itself.

Although creating a fully detailed multibody model of a vehicle has its obv advantages, using simpler dynamic models should be considered if the b requirements of implementing a multibody model cannot be met. Furthermor  Read Free Fornecessarily 30this Days Signwill up tonot vote on title should be noted that increasing model complexity provide  Not useful  Useful  improved correlation with reality. Cancel anytime.


If these implications are understood and the decision is made to improve this mo





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


8. Original Original contribution contribution

Although the use of multibody systems to model the dynamic behaviou passenger vehicles is nothing new, the way in which the validation study was car out in this project is innovative. i nnovative.

Some precedents [33] exist on the use of MTS 329 road simulators to measure kinematics and compliances of suspensions. However, no published work has b found on the method of using the road simulator to apply vertical excitations to vehicle when it is set to a floating configuration, and on using the measu responses to validate a model.

The main findings of this project, especially those related to the use of the r simulator as a model validation tool, have been summarized in the follow technical paper:

Use of a spindle-coupled multiaxial road simulator to validate a multibody vehicle dynamic model of a passenger car. Olazagoitia, JL., Biera, J. and Fernandez de Antona, J.

2011, an ECCOM The paper is due to be presented in Multibody Dynamics 2011, thematic conference which will be held in Brussels (Belgium), 4th-7th July, 2011.






Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


9. Bibliography

[1]. International Organization for Standardization. Road vehicles with two axles - Determinatio centre of gravity. Geneva, Switzerland : s.n., 1992. ISO 10392:1992.

[2]. Volkswagen of America. Volkswagen Passat 2006, 2007, 2008, 2009 Repair Manual on DVD ROM. [DVD-ROM] Cambridge, MA, Massachusetts, USA : Bentley Publishers, 2009. ISBN: 978-08376-1361-1. [3]. Wikipedia. Volkswagen Passat. Wikipedia. [Online] [Cited: 09 11 2010.]

http://en.wikipedia.org/wiki/Volkswagen_Passat#Passat_B6_and_B7_.282005.E2.80.93present 29. [4]. MSC Software Corporation. MD ADAMS 2008 documentation. 2008.

[5]. García de Jalón, J. and Bayo, E. Kinematic and Dynamic Simulation of Multibody Systems. Th Real-Time Challenge. New York : Springer-Verlag, 1994. ISBN: 0-387-94096-0. [6]. Blundell, M. and Harty, D. The multibody systems approach to vehicle dynamics. Oxford : Butterworth-Heinemann Elsevier, 2004. ISBN: 978-0-7506-5112-7.

[7]. Pacejka, Hans B. Tire and Vehicle Dynamics. s.l. : Butterworth-Heinemann, 2005. ISBN 978-0 7506-6918-4.


Unlock-full access with aand freeData trial. Handling. Kasprzak, Edward M [8]. The Formula SAE Tire Test Consortium Tire Testing

and Gentz, David. Dearborn, USA : SAE International, 2006. Motorsports Engineering Conferenc Exposition. SAE 2006-01-3606.


[9]. Software, Cosin Scientific. FTire Product Brief. Cosin Scientific Software. [Online] [Cited: 17 November 2010.] http://www.cosin.eu/.

[10]. Gillespie, Thomas D. Fundamentals of Vehicle Dynamics. Warrendale, PA : SAE Internationa 1992. ISBN: 978-1-56091-199-9.

Master your semester with Scribd  PA : S [11]. Milliken, William F. and Milliken, Douglas L. Race car Read vehicle dynamics. Warrendale, Free Foron 30this Days Sign up to vote title  International, ISBN: 1-56091-526-9. & The New York1995. Times  Useful  Not useful Cancel anytime.

Special offer for students: Only $4.99/month. [12]. Reimpell, Jörnsen, Stoll, Helmut and Betzler, Jürgen W. The automotive chassis: Engineerin





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


[16]. General modeling of nonlinear isolators for vehicle ride studies. Koppenaal, Jacco, et al. SAE International, 2010, SAE International Journal of Materials and Manufacturing, Vol. 3. (MSC Software Corp.). SAE 2010-01-0950. [17]. Sensitivity of a vehicle ride to the suspension bushing characteristics. Ambrósio, Jorge and

Verissimo, Paulo. s.l. : The Korean Society of Mechanical Engineers, 2009, Journal of Mechanical Science and Technology, Vol. 23, pp. 1075-1082. 1738-494X.

[18]. Real-time multibody vehicle model with bush compliance effect using quasi -static analysis f

HILS. Sung-Soo, Kim and Wan Hee, Jeong. s.l. : Springer, 2009, Multibody System Dynamics, Vol pp. 367-382.

[19]. Parameter Determination and Vehicle Dynamics Modeling for the National Advanced Drivin

Simulator of the 2006 BMW 330i. Salaani, Mohamed Kamel et al. Detroit : SAE International, 20 SAE World Congress & Exhibition. SAE 2007-01-0818.

[20]. Design and prediction of vehicle handling and stability. Hogg, A. S. et al. London : Institutio Mechanical Engineers, 1992. FISITA congress. IMechE paper C389/122.

[21]. The myth of accuracy. Harty, D. January 1999, The Journal of Engineering Integrity Society. [22]. Computer codes for road vehicle dynamic models. Sharp, R. S. Birmingham : Institution of Mechanical Engineers, 1991. Autotech. IMechE paper a427/16/064. You're Reading Preview [23]. Methodology for validating the Unlock National Advanced full access with aDriving free trial.Simulator's vehicle dynamics (NADSdyna). Garrott, W. R. et al. Detroit : SAE International, 1997. International Congress & Exposition. 970562.


[24]. Geometric identification of an elastokinematic model i n a car suspension. Meissonnier, J. e 9, s.l. : Professional Engineering Publishing, 2006, Proceedings of the I nstitution of Mechanical

Engineers, Part D: Journal of Automobile Engineering, Vol. 220, pp. 1209-1220. IMechE paper 09 4070.

[25]. Developing an ADAMS Model of an Automobile Using Test Data. Rao, P. S. et al. s.l. : SAE Master your semester with Scribd  International, 2002. Proceedings of the 2002 SAE Automotive Dynamics and Stability Read Free For 30this Days Sign up to vote on title Conference  Exhibition. p. 337. SAE 2002-01-1567. & The New York Times  Useful  Not useful Cancel anytime.

Special offer for students: Only $4.99/month. [26]. The design of a Vehicle Inertia Measurement Facility. Heydinger, Gary J. et al. Detroit : SAE





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


[29]. Design and operation of a new vehicle suspension kinematics and compliance facility. Best,

Tony, et al. Detroit : SAE International, 1997. SAE International Congress and Exposition. SAE 97

[30]. Suspension kinematics and compliance - measuring and simulation. Holdmann, Peter, et al Detroit : SAE International, 1998. SAE International Congress and Exposition. SAE 980897. [31]. A Model Validation Methodology for Evaluating Rollover Resistance Performance of a Fo

Commercial Vehicle. Ozan, Berzah, et al. (Ford Otosan A.S.). Detroit : SAE International, 2010. SA World Congress & Exhibition. SAE 2010-01-0107.

[32]. Elastokinematic modeling and study of five -rod suspension with subframe. Knapczyk, Józef Maniowski, Michał. 9, s.l. : Elsevier, September 2006, Mechanism and Machine Theory, Vol. 41,

1031 –1047. 0094-114X.

[33]. Somashekhar, Mathad G., Bhaskar, Chaturvedi and Subramaniam, L. An integrated appro to extract basic suspension data through integration of tri-axial spindle coupled road simulator,

wheel force transducer and a wheel vector sensor. s.l. : SAE International, 2009. SAE 2009-28-00

[34]. International Organization for Standardization. Road vehicles - Vehicle dynamics test meth

- Part1: General conditions for passenger cars. Geneva, Switzerland : s.n., 2006. ISO 15037-1:200 [35]. Model validation of the 2006 BMW 330i for the National Advanced Driving Simulator.

International, SA E World Congress & Exhibition. SAE Heydinger, Gary J. et al. Detroit : SAEYou're Reading2007. a Preview 2007-01-0817.


[36]. International Organization for Standardization. Passenger cars - Steady-state circular drivi

With Free:Trial behaviour - Open-loop test methods. Download Geneva, Switzerland s.n., 2004. ISO 4138:2004.

[37]. —. Passenger cars - Test track for a severe line-change maneouvre - Part 1: Double-lane ch Geneva, Switzerland : s.n., 1999. ISO 3888-1:1999.

[38]. —. Passenger cars - Test track for a severe line-change maneouvre - Part 2: Obstacle avoida Geneva, Switzerland : s.n., 2002. ISO 3888−2:2002.

Master your semester with Scribd [39]. —. Road vehicles - Test method for the quantification of on-centre handling - Part1:  Weave Read Free Foron 30this Days Sign up to vote title  Geneva, Switzerland : s.n., 2003. ISO 3674−1:2003. & The New York Times  Useful  Not useful Cancel anytime.

Special offer for[40]. students: Only $4.99/month. vehicles - Test method for the quantification of on-centre handling - Part2: Transit —. Road





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Upload

Sign In

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

1

Download



Magazines



News


of 112

 

Design


 Search document






Documents



Sheet Music


[44]. ADAMS/Car-AT in the chassis development at BMW. Fischer, E. (BMW). Berchtesgaden : M Software Corporation, 2001. European ADAMS User Conference.

[45]. Vehicle dynamics CAE within a development programme. Fussey, N.W. et al. (Jaguar Cars L Berchtesgaden : MSC Software Corporation, 2001. Eur opean ADAMS User Conference. [46]. An application of full vehicle ADAMS modelling with detailed force elements: collaborative activity among design, experiment, & CAE in the field of vehicle dynamics. Sakai, Y. et al (Nissan

Motor Co.). London : MSC Software Corporation, 2002. European ADAMS User Conference.

[47]. Load-deflexion relations of rubber bush mountings. Adkins, J. E. and Gent, A. N. 10, London

The Institute of Physics, October 1954, British Journal of Applied Physics, Vol. 5. DOI: 10.1088/05 3443/5/10/305. [48]. Approximate radial stiffness of rubber bush mountings. Horton, J. M. and Tupholme, G. E.

s.l. : Elsevier, 2006, Materials and Design, Vol. 27, pp. 226-229. DOI: 10.1016/j.matdes.2004.10.0

[49]. Stiffness of rubber bush mountings subjected to radial loading. Horton, J. M., Gover, M. J. C

and Tupholme, G. E. s.l. : American Chemical Society, 2000, Rubber chemistry and technology, V 73, pp. 253-264. [50]. Stiffness of rubber bush mountings subjected to tilting deflection. Horton, J. M., Gover, M.

Society, a2000, Rubber Chemistry and Technology, V and Tupholme, G. E. s.l. : American Chemical You're Reading Preview 73, pp. 619-633. ISSN: 0035-9475 .


[51]. Torsion stiffness of a rubber bushing: a simple engin eering design formula including amplitu

With Free Trial 2007, The journal of strain analy dependence. García, Maria José, et al.Download 1, s.l. : Sage Publications, for engineering design, Vol. 42, pp. 13-21. DOI: 10.1243/03093247JSA246. [52]. Measured vehicle inertial parameters - NHTSA's data through November 1998. Heydinger,

J., et al. Detroit : SAE International, 1999. International Congress & Exposition. SAE 1999-01-133

[53]. MTS Systems Corporation. Model 329 multiaxial spindle-coupled road si mulators. [Online] [Cited: 10 November 2010.] http://www.mts.com/en/vehicles/durability/ssLINK/CM3_002014. Master your semester with Scribd  Read Free Foron 30this Days Sign up to vote title  [54]. The Mathworks, Inc. MATLAB Help. & The New York Times  Useful  Not useful Cancel anytime.

Special offer for[55]. students: $4.99/month. MTSOnly SWIFT. [Online] MTS Systems Corporation. [Cited: 28 February 2011.]

http://www.mts.com/downloads/100-023-513c_SWIFT05.pdf.





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join





Search



Home



Saved

0

50 views



Sign In

Upload

Join

RELATED TITLES

0







Top Charts



Books



Audiobooks









Save

Embed

Share

Print

Download



Magazines



News


1

of 112

 

Design


 Search document






Documents



Sheet Music


[59]. —. Passenger cars - Power-off reaction of a vehicle in a turn - Open-loop test method. Gene Switzerland : s.n., 2006. ISO 9816:2006. [60]. —. Passenger cars - Free-steer behaviour - Part2: Steering-pulse open-loop test method. Geneva, Switzerland : s.n., 2004. ISO 17288-2:2004. [61]. —. Passenger cars - Free steer behaviour - Part1: Steering-release open-loop test method. Geneva, Switzerland : s.n., 2002. ISO 17288-1:2002.

[62]. Integration of kinematics and compliance measurement with vehicle dynamics validation fo

shared platform. Jen, Ming Une and Lu, Ming-Hung. Detroit : SAE International, 2007. SAE Worl Congress. SAE 2007-01-0832.

[63]. Typical vehicle parameters for dynamics studies revised for the 1980's. Riede, Peter M., et a Detroit : SAE International, 1984. SAE International Congress and Exposition. SAE 840561.








Useful

 





Home



Saved



Top Charts



Books



Audiobooks



Magazines



News



Documents



Sheet Music




Upload

Sign In


Join



Jon Fernandez de Antona Beng Project Excerpt3

Recommend Documents