AM 2036 (10 May 2004) - Normes_fem_global

AM2036

Means and Methods

Finite Element Analysis Manual

PURPOSE: This document shall form the basis for the preparation and exchange of major component FE models within Airbus, and will be applicable to all future analysis structures and projects. The intention is to create a common FE model of a complete aircraft structure by assembling the component parts as supplied, without the need to alter FE interface details, and which will analyse first time without the need to re-work the FE structure. The model obtained will provide the starting point for all future analysis work and for all engineering disciplines, thus avoiding duplication of effort. This specification has been prepared by representatives of each site of Airbus.

SCOPE:

Owner’s Approval: (signed)

Name Function

: JL. de la Gándara Verano : Global FE Analysis (IESSM)

Authorization: (signed) Date

: 10 May 2004

Name Function

: JF. Imbert : CoC Structures - Head of Structural Analysis (ESA)

 Airbus SAS 2004. All rights reserved. Confidential and proprietary document

Issue: E

Page 1 of 68 Printed Copies are not controlled. Confirm this is the latest issue available through the Portal.

AM2036 Finite Element Analysis Manual

Table of contents 1

Objective.................................................................................... 3

2

Finite Element Analysis Code.................................................. 4

3

Finite Element Model Definition............................................... 5

4

General Interface Requirements.............................................. 9

5

Quality Assurance of Components and Documentation ..... 12

6

Complete Model Assembly, Quality Assurance and Documentation........................................................................ 17

7

Mass Data ................................................................................ 18

8

Dynamic Model........................................................................ 19

9

Update Procedure ................................................................... 20

APPENDIX A Forms............................................................................................... 21 APPENDIX B General Explanations and Guidelines .......................................... 28 APPENDIX C........................................................................................................... 64 Reference documents ........................................................................................... 67 Glossary ................................................................................................................. 67 Approval ................................................................................................................. 67 Record of revisions ............................................................................................... 68

Issue: E

 Airbus SAS 2004. All rights reserved. Confidential and proprietary document Printed Copies are not controlled. Confirm this is the latest issue available through the Portal.

Page 2 of 68


1

Objective This specification shall form the basis for the preparation and exchange of major component FE models within AIRBUS, and will be applicable to all future analysis structures and projects. The intention is to create a common FE model of a complete aircraft structure by assembling the component parts as supplied, without the need to alter FE interface details, and which will analyse first time without the need to re-work the FE structure. This model will provide the starting point for all future analysis work and for all engineering disciplines, thus avoiding duplication of effort. The specification also identifies common validation procedures to assure that all the analysis work undertaken by Airbus continues to be of the highest quality consistent with world class design organisations. The correct application of AM2036 will not guarantee that the model results are acceptable to the project, only that the results will reliably reflect the response of the structure and will not be affected by analysis anomalies. This specification will be subject to review on a regular basis.

Issue: E


Page 3 of 68


2

Finite Element Analysis Code The agreed F.E. Analysis Code for all work shall be MSC.Nastran. The specific version of MSC.Nastran to be used shall be determined and agreed by all parties for each major project, or major revision of project.

Issue: E


Page 4 of 68


3

Finite Element Model Definition The component models supplied must assemble into a Complete Aircraft model without further work being necessary, and must have sufficient and well balanced detail to permit combined aeroelastic and strength optimisation and, when assembled as part of the overall aircraft model, must provide the correct load paths, stiffness behaviour and deflection of the complete model. The FEM of the complete aircraft shall comprise the following components: •

A complete full diameter fuselage.

•

Both Left and Right wings, including wing-let and wing-tip devices if applicable.

•

Both Left and Right Horizontal Tailplanes.

•

Vertical Tailplane.

•

All Engine Attachment Pylons.

The following components will be supplied if available: •

All Landing Gears.

•

All moving surfaces required for Dynamic Analysis (including actuators and attachments structures).

•

All Fairings required for Dynamic Analysis.

Each structural component shall be supplied as a separate MSC.Nastran bulk data file, additional data required to enable test analysis of the component should also be supplied as a further computer file. The model supplied must contain the following data: •

The model data shall be supplied in N and mm.

•

The model must include all the requisite attachments for the application of masses for Dynamic Analysis.

•

The constituent components of an A-set for use in creating a condensed model.

•

Sufficient comments on idealisation (Boundaries, Spec., Features, …)

In order to permit the generation of a half-aircraft model from the full aircraft model, no elements must bridge across the aircraft vertical plane-of-symmetry. Also all the DOF on the plane of symmetry must be independent. Continued…

Issue: E


Page 5 of 68


3

Finite Element Model Definition, continued For elements on or in the vertical plane-of-symmetry, they will be represented in full, and it will be required to be divided by 2 when generating the half-model. Doubled elements on the plane-of-symmetry may be used providing all parties involved are in agreement; but property references must be unique to that location to eliminate the problem of multiple addressing of property cards. The NASTRAN "basic co-ordinate system" of the model shall be the basic aircraft coordinate system as defined in the DBD. An alternative master co-ordinate system may be used providing all parties agree. The model bulk data shall be numbered to the following convention: GRID

→

C yyy zzz (*)

CORDxR/FORCE/MOMENT

→

C xxx

SPC/LOAD

→

C xxx

Element/Property/Material (except for the following)

→

C x yyy zzz (*)

PCOMP/MAT8

→

C xx xxx (*)(**)

Where C represents the component identification number. No element number should be duplicated, and the numbers assigned should be meaningful. Where possible the numbering logic should be retained between different versions of model for the same project. Note: (*) The range of numbering for the element, property, material and GRID identifications for the Fuselage will be changed to starting values of 1 (always) after the availability of a new MSC.Nastran validated version (see paragraph 2) with extended PCOMP and MAT8 numbering (8 digits) and not further aeroelastic limitation. (**) From Nastran 2001 the numbering restrictions on PCOMP cards will no longer exist. The following number sequences are proposed as a generic set. Project teams may agree alternate sequences to meet alternate project requirement. Continued…

Issue: E


Page 6 of 68


3

Finite Element Model Definition, continued Component

Element/Property/Material Grid Identification range → 2 9 999 999 1

→ 2 999 999

Fuselage

1

Pylons & Engines

3 0 000 000 → 3 9 999 999 3 000 000

→ 3 999 999

Vertical Tailplane & Rudder

4 0 000 000 → 4 9 999 999 4 000 000

→ 4 999 999

Horizontal Tailplane & Elevator

5 0 000 000 → 5 9 999 999 5 000 000

→ 5 999 999

Wing - (Left) (*)

6 0 000 000 → 6 9 999 999 6 000 000

→ 6 999 999

Wing - (Right) (*)

7 0 000 000 → 9 999 999

7 000 000

→ 7 999 999

Landing Gear (**)

8 0 000 000 → 8 9 999 999 8 000 000

→ 8 999 999

Aeroelastic

9 0 000 000 → 9 9 999 999 9 000 000

→ 9 999 999

Notes: (*) Including Flaps, Slats, Tracks, Ailerons, Winglets, etc. (**) Including Nose, Centre, and Main Landing Gears. Only those cards identified within the following list should be used within the model data deck: PARAM cards which are required to be used/included: PARAM

AUTOSPC NO

(PARAM,AUTOSPC,YES is now the default, and changing the PARAM to NO has no effect on the end result. Where the "Grid Point Singularity Table", listed in the F06 output dataset, shows a value of 0., this singularity should be removed by an appropriate SPC card. Any such SPC's should be grouped separately, with a unique set ID(s), from boundary SPC's, and SPCADD should be used to combine sets for analysis.)

PARAM

BAILOUT

0

(a value of - 1 may be used during model testing)

PARAM

GRDPNT

0

PARAM

WTMASS

1.E-3 (if model is in N and mm and mass in kg)

PARAM

SNORM

(must only be used with system default value) Continued…

Issue: E


Page 7 of 68


3

Finite Element Model Definition, continued PARAM cards which must be used with care and by agreement with all parties: PARAM

MAXRATIO

(this PARAM should only be used if a value different to the MSC.Nastran default value has been mutually agreed between all parties).

PARAM

K6ROT

1.0

(This PARAM must only be used in combination with PARAM,SNORM)

Bulk Data cards (*) a)

GRID

CORDxR

CORDxC

CORDxS

b)

CBAR

CBEAM

CROD

CONROD

CQUAD4

CTRIA3

CSHEAR

PBAR

PBEAM

PROD

PCOMP

PSHELL

PSHEAR

d)

MAT1

MAT2

MAT8

MAT9

(MAT4)

e)

CONM1

CONM2

CELASx (**)

f)

FORCE

MOMENT

PLOAD

PLOADx

LOAD

g)

[CPENTA]

[CHEXA]

[CTETRA]

[PSOLID]

h)

[GENEL]

[DMIG]

c)

The following are Bulk Data cards that should only be used with care and/or after consultation with all other parties: RBE1

RBE2

RBE3

RBAR

RROD

The RSPLINE element should be avoided, and mesh changes incorporated using CQUAD4 & CTRIA3 where necessary. MPC

should not to be used. (Recommended)

SPC

should only be used to suppress rotational DOF’s with no stiffness or to fully suppress DOF of reference nodes. All the translation DOF's have to have stiffness enough (except reference grids). It is forbidden to use field 8 of GRID card to define "permanent single point constraint".

Where continuation cards are required, the NASTRAN default continuation, which does not require continuation labels, should be used at all times. Under NO circumstances should any non-linear elements be included in the Bulk Data. Notes: (*) Any linear element not included in categories a) to h) can be used subject to approval. (**) To be used with care.

Issue: E


Page 8 of 68


4

General Interface Requirements All parties shall agree to a common FEM methodology to be applied at interfaces. This agreed approach should include the following: •

Agreed interface Geometry.

•

Agreed Meshing standard and Idealisation philosophy

•

Agreed Grid and Element data numbering.

•

Compatible Element types at the boundary.

•

Agreed displacement co-ordinate system.

•

Agreed that DOF at both sides of interface must be compatible.

Interface Grid points must be shared using the same co-ordinate systems, and the same element types must be used at the boundary in order to provide continuous stiffness paths across the interface. Each interface definition must be agreed between all parties at the very early phases of each project. Two optional types of interface are shown, the first shares information on a common basis, and in this case all parties must agree on the ownership of the individual data items: see Fig. 4-1. This is the recommended form of interface model. The second (see Fig. 4-2) shows an interface using duplicated data, i.e. the geometry of points 1470220 and 6199020 will be identical, and elements 10470119 and 61000021 will overlay each other. In this case the sum of the property data for these elements must agree with the total stiffness of the actual structure. As with the simpler joint shown above, the ownership of the bridging elements joining the components together, and the type of elements to be used will need to be agreed; but each component is structurally independent, and interface loads may be recovered at the joint. This interface method should only be used in exceptional circumstances. In order to perform the model assembly more easily, it is recommended to include each different interface in a separate file. Continued…

Issue: E


Page 9 of 68


4

General Interface Requirements, continued For a conventional AIRBUS aircraft, the following Interfaces have been identified, and details of agreements reached for each should be included as part of the model documentation: •

Rear Fuselage to Vertical Tailplane

•

Rear Fuselage to Horizontal Tailplane

•

Rear Fuselage to Centre Fuselage

•

Centre Fuselage to Wing

•

Centre Fuselage to Centre Landing Gear(s)

•

Centre Fuselage to Front Fuselage

•

Front Fuselage to Nose Landing Gear

•

Front Fuselage to Cockpit

•

Wing to Main Landing Gear

•

Wing to Engine Pylon

•

Wing to Moving Surfaces e.g. Flaps, Slats, Ailerons, etc.

•

Any other identifiable interface (subject to workshare)

For any other Aircraft, an equivalent list of interfaces should be considered. For Dynamic Analysis, if required, the A-set information for the component should be supplied. This definition must be agreed between all parties.

Fig. 4-1 - Typical Interface Continued…

Issue: E


Page 10 of 68


4

General Interface Requirements, continued

Fig. 4-2 - Alternative Interface

Issue: E


Page 11 of 68


5

Quality Assurance of Components and Documentation Each supplier has the responsibility of verifying their component models before delivery to the other parties. A. The following verification checks are required to be undertaken with the model: A.1. Geometric Checks: A series of basic model data checks shall be performed on the analysis model. These checks shall include the following: a. Free edge/face to confirm there are no voids in the model. b. Direction normals to confirm consistency of element orientations + G1G2 direction. c. Bending planes and or offsets for element consistency. d. Panel shape assessments for Aspect Ratio, Taper, Twist and Skew. e. Identification of co-incident Grids and Elements. f. Identification of all co-ordinate systems. g. Dependant/Independent node conflicts for Rigid elements. A.2. Static Checks: Analysis of Static cases to determine deformations, path loads and local stresses. The cases used should be significant for assessing the response of the component and the BC will be representative of the final assembly. A.3. Rigid Body Checks analysis: Only rigid body motion will be acceptable, there should be no local deformations with the energy matrix defined as part of the N-set. A.4. Normal modes analysis with test masses (using densities in MATx cards): to determine that only six rigid body modes exist for the model in free/free condition and with the effects of the control surface modes added. This check will require the addition of the material density parameter to all of the MATx cards; these modified MATx cards should be used for test purposes only. For a description of the required tests: see table 5-1 Items 1 thru 3. A.5. Thermo-elastic analysis with unique isotropic expansion coefficient in all MATx cards): to determine that null stresses are induced by an uniform ∆T applied to the model.

B. Mass model QA: The mass model should be verified by producing XY ~ XZ plots using scaled mass points. Additionally the normal modes check defined in the attached table should also be carried out using a mass case that includes payload, fuel, and both structural and systems masses (M1 or M2). Notes: → M1 = artificial mass case to test every mass and RBE3 connections.

→ M2 = realistic mass case. Continued…

Issue: E


Page 12 of 68


5

Quality Assurance of Components and Documentation, continued

C. The following verification checks will be required to be undertaken with the model using CONMx cards: C.1. Normal modes analysis with mass model (using CONMx cards): The normal modes should be calculated for M1 or M2 mass conditions: see table 5-1. Items 4 and 5. C.2. Normal modes analysis with mass model (using CONMx cards) in A-set: The normal modes should be calculated for a M2 mass status, using the A-set for the conditions (∗) specified: see table 5-1 Item 6. Normal modes assessment tests are defined as follows: CHECK

Test Masses

Method

Frequencies

1

Densities in MATx cards

Lanczos

[ -1, 1]

2

Densities in MATx cards Densities in MATx cards

Lanczos

[ 0, 1]

Lanczos

[ 1, ∞] 15 first modes

3

Boundary Conditions free/free with actuators 8 (Kact ~ 10 N/m) free/free without actuators free/free with actuators

Target 6 and only 6 rigid body modes (RBM) 6 rigid body modes plus 0 Hz rotations local or unrealistic modes detection

And for the Mass-model CHECK

Test Masses

Method

Frequencies

4

RBE3 and CONMx (M1)

Lanczos

[ -1, 1]

Boundary Conditions free/free with actuators

Target 6 and only 6 rigid body modes (Mass model verification)

And for the Dynamic model CHECK

Test Masses

Method

Frequencies

5

RBE3 and CONMx (M2) RBE3 and CONMx (M2)

Lanczos

[ 0, ∞] 15 first modes [ 0, ∞] 15 first modes

6

MGIVENS or MHOU and A-set

Boundary Conditions free/free

free/free

Target Normal modes for structure (used as reference) Complementary (for Analysis set verification)

Table 5-1 - List of Normal Modes Checks

(∗)

Continued…

Any alternative to A-set definition should be agreed between all parties.

Issue: E


Page 13 of 68


5


Notes: → Where MSC.Nastran DMAP Alters have been used in any of the QA checks, this Alter should be declared and agreed between all parties. → The highest frequencies for Rigid Body Modes, Checks 1 and 2, should be as close to 0.0 as possible, but should not be greater than 1/100 of the magnitude of the lowest predicted flight mechanics frequency for first flexible mode. Values below 0.01 Hz are desirable; any values above this level must be discussed between all parties. D. Information to be supplied with component models (when distributing): D.1. Statement by an authorised person that the component model conforms to the specification AM2036, and that ALL checks required by the specification have been completed. D.2. NASTRAN .F06 (using ECHO=SORTED and PARAM,PRTMAXMIN,YES) listing for SOL 101 giving the following information: •

OLOAD summary table.

•

MAX. DISPLACEMENTS table.

•

SPCFORCES table.

D.3. NASTRAN

.F06

(using

ECHO=SORTED,

PARAM,PRTMAXMIN,YES

and

SPCFORCE=ALL) listing for SOL 103 giving the following information: •

Output of the WEIGHT POINT GENERATOR.

•

The value of the first 9 eigenvalues.

D.4. The following model files for each SOL (101 and 103): •

NASTRAN job submission file (with the "includes")

•

PARAM cards necessary for the execution

•

Data of the interface with other components.

•

ASET definition.

•

Mass RBEs and CONMx data (SOL 103 only).

•

All the files containing the NASTRAN Bulk Data. Continued…

Issue: E


Page 14 of 68


5


E. Complete documentation shall be supplied for each component model. The following items of documentation shall be included: E.1. A statement identifying the engineering standard/design status used in the preparation of the model. E.2. Overall plots of relevant parts, especially for the interface areas. E.3. Descriptions of the model Bulk Data deck to enable users to easily identify the data and definition of the rules to decode the numbering sequence used for the GRID and element numbering. E.4. The interface grid points complete with co-ordinate frame, and the boundary element characteristics for each interface relevant to the component. Where relevant this interface document should also include an actuator installation sketch. E.5. Verification that the model geometry checks identified in A.1. have been performed correctly. E.6. A brief description of the static test load cases used to verify the model. E.7. A diagram or other description of the support conditions and a summary of the generated reaction forces from the static test cases. E.8. Results from the static test load cases in the form of the OLOAD, Maximum Displacements and SPCFORCEs resultant tables. E.9. The maximum Von Mises stresses generated by the "ZERO STRESS" thermoelastic test. E.10. The rigid body modes, frequencies and mode shapes (plots), plus control surface modes generated from the normal modes tests (items 1 and 2 in table 5-1). E.11. Tabulated frequencies for flexible normal modes, plus plots of first three flexible modes (item 3 in table 5-1). These modes should also be identified. E.12. The energy matrix from the rigid body test (items 3, 4, 5 and 6 in the table 5-1). E.13. Tabulated frequencies and flexible modes displacement plots for all DOFs involved in RBE3 mass connections (Item 5 in table 5-1). E.14. A list of all supplied filenames. E.15. EI and GJ curves and other specific information (for pylons, Centre Section, etc.) may be supplied as optional extras. E.16. MSC.Nastran version and computer description, including Operating System level. The distributed masses used as part of the tests should also be supplied for use in the verification of the assembled model only. This data should not be used for any other purpose. Continued…

Issue: E


Page 15 of 68


5


F. The following forms have to be completed and attached to the full documentation of the FEM: •

FORM 1: "REFERENCES": In this form a FEM overall description is included: e.g. configuration, geometry/property/material references, drawings (if exists), document of dimensioning or whichever good reference that identify correctly the model.

•

FORM 2: "FILES RECORD": Describe each one of the datasets that define the model and the different runs used in FORM 4. Because the main use of this form is to verify the right transmission of the full component FEM, this form has to be sent at the same time as the model.

•

FORM 3: "MODEL CHECK LIST": Describe all the geometry checks performed on the model. As well as indicating if the checks have been performed, give details of any problems that may have been found, along with maximum/minimum values where appropriate. Where necessary append plots for clarity (e.g. co-ordinate systems).

•

FORM 4: "ANALYSIS CHECK LIST": Describe mainly the results of static and Rigid Body checks.

•

FORM 5: "ANALYSIS CHECK LIST": NORMAL MODES": Describe all the checks performed with the model using density in the MATx cards.

•

FORM 6: "ANALYSIS CHECK LIST: DYNAMIC NORMAL MODES": Includes the results of all the dynamic checks performed with the model using CONM2 cards.

•

FORM 7: "ANALYSIS CHECK LIST: THERMO-ELASTIC STRESS": Includes the results of the "ZERO STRESS" test.

Issue: E


Page 16 of 68


6

Complete Model Assembly, Quality Assurance and Documentation Component models shall be supplied, along with test results and documentation, as MSC.Nastran Bulk Data files to the team responsible for undertaking the assembly. The model bulk data shall be supplied via a suitable electronic data transfer link. In the event of sub-assemblies of minor components being used, the assembler of the sub-assembly model must perform the same validation checks defined for a component model or as directed by project management, and supply the appropriate documentation to the final model assembly. The team responsible for the final assembly shall assemble all the component models to form the complete full aircraft model and shall carry out the same verification checks as have been identified for the component models in section 5.0 of this specification. Equivalent documentation for the complete model shall also be prepared as per section 5.0. If the verification checks of the complete model are unsatisfactory, then a workshop may be called to solve those problems relevant to the assembly or to the interface data. Where the problem lies within a component model, then the team responsible for that model must rework the component to a satisfactory state. After all the verification tests have been completed and accepted by all parties, the complete model will be distributed to all users, via a suitable electronic data transfer link. The documentation that will accompany the complete model will be as follows: 1. A brief description of the static test load cases used to verify the model shall be included. 2. A diagram or other description of the support conditions and a summary of the generated reaction forces from the static test cases shall be included. 3. Deformations at the same defined locations used for the checks on the component models, and stresses/strains at the same selected elements shall be included, along with the table of maximum values for displacements and stresses. 4. The combined A-set (∗) for the complete model, assembled from the component model A-sets. 5. The basic 6 rigid body modes plus control surface modes, and results for Component rigid body tests should also be included for those moving surfaces where it was not possible to test the control surface modes due to the need for an interface at a control surface boundary. 6. The first "n" normal modes using the defined A-set, frequencies and mode shapes should be supplied. The number of modes to be listed should be agreed between all parties.

(∗)

Any alternative to A-set definition should be agreed between all parties.

Issue: E


Page 17 of 68


7

Mass Data All parties shall provide the locations and the connection system for mass points for their component(s), for masses subdivided into separate distributions, e.g. Structures, Systems, etc. Based on this information, the responsible team shall supply mass data for appropriate configurations, and shall distribute this data to all other parties. Mass data will be supplied for the complete aircraft, in kg, m and s using CONM2 cards. Prior to delivery mass quality assurance checks shall be undertaken including density plots for the XY~XZ plane. If the model is to be used for optimisation, it may be necessary for the component supplier to provide mass distributions for their components equivalent to the mass of the FEM generated by the density parameter on the material cards. This distribution will be required to be subtracted from the Fixed Structure mass distribution.

Issue: E


Page 18 of 68


8

Dynamic Model The team responsible for the assembly of the complete aircraft model shall transform the model from N and mm to N and m, without any other modification, maintaining the Bulk Data deck in its original sequence and also maintaining the same data formats: see table 8-1. The aeroelastic calculation shall use both the half-aircraft and the full-aircraft models to properly represent the effects of the non-symmetrical considerations. The responsible team will then prepare a symmetric half-aircraft model, left side, for the specific needs of the Aeroelastics and Loads departments. These aircraft models shall be distributed via a suitable electronic data transfer link. In addition to the model bulk data, files containing the correct boundary constraint data for the half model, for both symmetric and anti-symmetric boundary conditions, shall also be supplied. These models shall form the basis of any dynamic investigations that may be required. For the half-aircraft model, mass connection elements shall be taken from those supplied for the complete aircraft model, with suitable modifications to allow for the splitting of the model. Mass linkage to structure: Mass should be assigned to single "non structural points" (GRIDs), linked to the structure using RBE3 elements connected to a set of "local hard structural points". Where A-set (*) is to be used there are two possibilities: a) The A-set should include part of the "local hard points". Therefore, the "non structural mass points" should be the DEPENDENT nodes on the RBE3. b) The A-set should include part of the "non structural mass points". Therefore, these points should be the INDEPENDENT nodes on the RBE3 and unique for all loading conditions. Complete model N and mm

Stress

N and m

Aeroelastic

Half model Loads and Aeroelastic

Table 8-1 - Usage of FEMs

(*)

Any alternative to A-set definition should be agreed between all parties

Issue: E


Page 19 of 68


9

Update Procedure For each complete model, a record of each component model shall be maintained by the party responsible for the complete model assembly. This record to include the issue number and the preparation date of the component. Where relevant changes have been made to any component, a new complete model will be assembled, verified, documented and distributed via a suitable electronic data transfer link to all parties as defined elsewhere in this specification. The following numbering system (in capital letters) will be used for identification of the complete assembled models and for formal referencing. a/c model/engine version/design standard/model issue/model type/model size Where: •

•

Model type: •

S = Static Model

•

D = Dynamic or Aeroelastic Model

Model size: •

C = Complete Model

•

H = Half Model

For example: A3XX100/01/08/01/D/C Represents the first issue of a complete aircraft model of the A3XX-100 at Status 8, with engine option 1. The model is for Dynamics usage and will therefore be in N and m. Where configuration or other studies are undertaken using parts or all of one of these models, these will be considered as independent activities, and will not be recorded as part of this numbering.

Issue: E


Page 20 of 68


APPENDIX A Forms QUALITY ASSURANCE OF FEM

PROJECT:

REFERENCES PART COMPONENT:

A/C COMPONENT:

ANALYSIS IDENTIFICATION:

PAGE:

VERSION:

Finite Element Model general description:

Reference Coordinate System (FEM Component related to DBD)

Geometry

Meshing

Properties

Materials

Name (Dept.)

Signature

Date

Compiled by: Approved by: Continued…

Issue: E


Page 21 of 68


APPENDIX A Forms, continued QUALITY ASSURANCE OF FEM

PROJECT:

FILES RECORD PART COMPONENT

A/C COMPONENT:

ANALYSIS IDENTIFICATION: COMPUTER DATASET/FILE NAME

DATE OF CREATION

Name (Dept.)

PAGE:

VERSION:

COMMENTS/DESCRIPTION

Signature

Date

Compiled by: Approved by:

Continued…

Issue: E


Page 22 of 68



PROJECT:

MODEL CHECK LIST PART COMPONENT:

A/C COMPONENT:


GRIDs

PAGE: VERSION:

1-D

MODEL SIZE 3-D

2-D

SPRINGs

DEPARTMENT RESPONSIBLE & PRINCIPAL DATA VALUE (If it is relevant)

CHECK LIST

RIGID BODYs

SIGNATURE

Coincident Nodes Coincident Elements Free Edges/Free Faces Element Warping Element Twist Interior Angles Distortion of Elements Connectivity Bending Planes (BAR&BEAMs) Offsets (all elements) Positive normal (QUAD4/TRIA3) Element Co-ordinate System Material Co-ordinate System Co-ordinate Systems Rigid Elements, MPCs Computer

ANALYSIS DATA Operating System

Max. Diagonal Factor If YES

MSC.Nastran Version

Is Grid Point Singularity. Table Empty? If NO Attached as page No.:

COMMENTS:

Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 23 of 68


APPENDIX A Forms, continued PROJECT:

QUALITY ASSURANCE OF FEM

ANALYSIS CHECK LIST PART COMPONENT:

A/C COMPONENT:


PAGE:

VERSION:

BOUNDARY CONDITIONS DESCRIPTION: (text, figure or references)

LOAD CASE DESCRIPTION:

Load Balance Checks Applied Reaction (SPCs) Difference (%) Maximum Displa GRID Value Epsilon COMMENTS:

FX

FY

FZ

MX

MY

MZ

TX

TY

TZ

RX

RY

RZ

External Work

Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 24 of 68


APPENDIX A Forms, continued PROJECT:

QUALITY ASSURANCE OF FEM

ANALYSIS CHECK LIST: NORMAL MODES PART COMPONENT: ANALYSIS IDENTIFICATION:

A/C COMPONENT: VERSION:

BOUNDARY CONDITIONS DESCRIPTION: (text, figure or references) Check 1 Mass Reference Mass Standard Value Mass Output from GPWG Mode Plots & Frequency List

Mode No: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Comments:

Check 1 (Densities in MATx) [-1,1] free/free with actuators

Check 2

Frequency [Hz] Check 2 (Densities in MATx) [0,1] free/free without actuators

Name (Dept.)

Page:

Check 3

Check 3 (Densities in MATx) [1,∞] free/free with actuators

Signature

Date

Compiled by: Approved by: Dynamic Acceptance:

Continued…

Issue: E


Page 25 of 68



PROJECT:



BOUNDARY CONDITIONS DESCRIPTION: (text, figure or references) Check 4 Mass Reference Mass Standard Value Mass Output from GPWG Mode Plots & Frequency List

Mode No: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Comments:

Check 4 (M1) (Masses in CONMx) [-1,1] free/free with actuators

Check 5

Frequency [Hz] Check 5 (M2) (Masses in CONMx) [0,∞] free/free with actuators

Name (Dept.)

Page:

Check 6

Check 6 (M2) (Masses in CONMx) [0,∞] free/free with actuators

Signature

Date


Continued…

Issue: E


Page 26 of 68



PROJECT:

THERMO-ELASTIC ANALYSIS CHECK LIST PART COMPONENT:

A/C COMPONENT:


PAGE:

VERSION:

BOUNDARY CONDITIONS DESCRIPTION: (text, figure or references) Statically determined constrain

LOAD CASE DESCRIPTION: Defined Temperature:

Load Balance Checks Applied Reaction (SPCs) Difference (%) Maximum Stress ELEMENT Value Epsilon COMMENTS:

FX

FY

FZ

MX

MY

MZ

TX

TY

TZ

RX

RY

RZ

External Work

Name (Dept.)

Signature

Date


Issue: E


Page 27 of 68


APPENDIX B General Explanations and Guidelines 1

Introduction The objective of this appendix is to clarify the correct application of the recommendations written in AM2036, and identify areas where it will be necessary to expect definitions of local project variations. Although driven by the needs of Loads and Aeroelastics analyses, AM2036 outlines basic principles of good practice that should be applied to all forms of analysis including Linear Elastic. Users of FE Analysis should also consider objectively whether their FEM might possibly be used as part of an analysis investigation involving dynamic analysis, and if this possibility exists, should then apply AM2036 rigorously to their model from the beginning. AM2036 has been directed towards models for linear static and dynamic applications. Where additional analysis requirements exist it will be necessary for analysts to consider additional validation checks based on those extra requirements. The same physical checks are required to be performed for the assembly of component models as have been applied to the individual components models. Similarly the data supply and documentation is also identical. This is in line with Section 6 of the main Document. For each of the geometric and analysis checks, this appendix will be provide a best practice method for performing them. These are only recommendations, the analysts may use any method that is considered suitable to achieve the required results and satisfactorily completes the checks. The results from the assembled model will reflect the behaviour of the whole aircraft (instead of only the separate components) and should be used as the basis for any decisions regarding model acceptability or real aircraft characteristics.

2

Finite Element Analysis Code The specific version of MSC.Nastran to be used on any given project needs to be agreed by all affected parties. Continued…

Issue: E


Page 28 of 68


APPENDIX B General Explanations and Guidelines, continued 3

Finite Element Model Definition In order to properly define the Finite Element Model, it is necessary to determine how the model is going to be organised. •

How many components will the main assembly be divided into?

•

Where will the boundaries between these various components be?

•

Who will be responsible for: •

Each sub-component?

•

Each sub-assembly of these sub-components?

•

The final assembly of the overall model?

It is the responsibility of the Chief Engineer’s organisation to ensure that these questions are addressed and suitable actions agreed. This would be best undertaken by a formal meeting between all parties at which these and other related general issues should be addressed. In principle, most of those definitions, where variations are permitted by AM2036, should be agreed at this meeting. It must be recognised that regular meetings between analysts involved in this modelling and analysis activity are essential to ensure correct compliance with these rules. Therefore this initial meeting should be considered as the launch meeting for this task.

3.1

FE Model Content The list included in the body of AM2036 is a generic list, the contents can be modified by specific requirements of each project and must be agreed at the launch meeting. Also the desired model forms should be agreed i.e., full aircraft, half aircraft etc. Only model configurations actually required should be prepared.

3.2

Data Numbering Convention The numbering convention to be used must be fixed at the introductory meeting according to the component breakdown and work share of the project. The convention given in the AM2036 is intended to show the form such a convention should take and should be used as default in absence of any other requirements. In addition the "programme" should manage and document the complete number spectrum. Continued…

Issue: E


Page 29 of 68


APPENDIX B General Explanations and Guidelines, continued 3.3

Acceptable NASTRAN Card Types The PARAM options quoted show best advice at the time of writing and in general should be adhered to. As analysis techniques change, the desired PARAM options may also change, and therefore it is important that the analysis PARAM configurations should be agreed at the launch meeting to ensure analysis compatibility. The "ECHO=SORT(PARAM)" card in the case control is recommended in order to check the PARAMs used (with this option the PARAMs used are written in F06). The list of the structural cards given includes all those standard elements that might be expected to be used. The use of any special element types outside the list should be discouraged. However where it is agreed as essential to use unlisted special elements, their use must be agreed at the launch meeting.

4

General Interface Requirements The list of modelling interfaces must be agreed at the launch meeting as discussed in section 3.0. For each of these interfaces, AM2036 has defined basic criteria that need to be considered. It is essential that the configuration at each interface is resolved by both affected parties and the assembler, within the AM2036 guidelines. This is best undertaken by meetings between these parties and mutual exchange of interface data until both parties are satisfied as to the quality of the modelling interface they have constructed. Documentation of the agreed interface is required within the documentation rules of AM2036 and must be included by both parties. Continued…

Issue: E


Page 30 of 68



Quality Assurance of Components and Documentation Each of the following checks must be performed in order to ensure the model being assessed will be suitable for use and will provide an easy assembly. A method for performing each of the checks is presented. These are proposals for each check, but the analyst may use any suitable method providing they satisfactorily perform the required tests.

5.1

Geometry Checks In order to fully validate model elemental geometric relationships it is "necessary" to use both MSC.Patran and MSC.Nastran in two distinct phases. These two phases are not independent, but complement each other, and it is recommended that the checks using MSC.Patran are performed first followed by checks using MSC.Nastran. When all the checks in this section have been performed, Form 3 (for geometry checks) must be completed. Before any specific data checks are performed, the MSC.Nastran bulk data deck should be read into MSC.Patran in order to identify any errors in element connectivity, numbering and entity duplication. At this stage plots for the complete structure can be prepared for later use in the documentation. These plots also form part of the data checks, as they will illustrate fundamental misalignments between groups of nodes and elements.

5.1.1

Validation Checks Using MSC.Patran These checks should be performed using the "Finite Element" or "Utilities" menus Within MSC.Patran set the "graphic utilities" to either "shrink", "hide line" or "shaded" mode and verify element connectivity and type. The "Utilities Display" menu can be used to check Element Type, Property data, Material data and Element co-ordinate Systems by using "Show Element Info". These checks are the following: Continued…

Issue: E


Page 31 of 68


APPENDIX B General Explanations and Guidelines, continued 5.1.1.1

Coincident Nodes To be made using the command "Equivalence". This command eliminates, for one list, group, or the whole model, the nodes located in the same position, with a user defined tolerance, erasing the undesired free edges. The coincident nodes linked by ELAS, GAP, MPC, and rigid elements are not eliminated by MSC.Patran. The erased nodes are highlighted and the command may be reversed with "UNDO". To display the coincident nodes linked by ELAS and/or GAP elements, you may use the menu "Utilities/Display/Mark 1D Elements". By putting the elements with length smaller than a user defined tolerance value in a list, the user may create a list of nodes related to these elements using the menu "List" command.

5.1.1.2

Free Edges The command "Verify/Element/Boundaries" is used to check the free edges of surfaces and identify disconnects between 2D elements. This command shows the free edges or free faces of the elements included in the group or groups shown in the screen, which are not connected to other elements of the model. The analyst must confirm that any free edges identified are valid. A plot should be obtained illustrating the free edge boundary, and It is recommended that this is drawn using 2D (QUAD4 and TRIA3) elements only as 1D elements will always show as free edges.

5.1.1.3

Coincident Elements The command "Verify/Element/Duplicates" is used to check for coincident elements. This command shows and lists all the elements with the same nodal connectivity. The command will erase those duplicate elements with the highest identification or viceversa. The analyst must confirm that only planned duplicate elements exist.

5.1.1.4

Normals Verification The command "Verify/Element/Normals" is used to verify element normals. The option "draw normal vector" displays the direction of element normals for 2D elements included in the current group. The command allows the user to reverse the normal of some elements to match the normal of the correctly aligned elements. Continued…

Issue: E


Page 32 of 68


APPENDIX B General Explanations and Guidelines, continued The QUAD4 element co-ordinate system used by MSC.Patran to obtain the element normal is shown in figure App B.5.1.1.4.1.

Fig. App B.5.1.1.4.1 - QUAD4 element co-ordinate system in MSC.Patran 5.1.1.5

Element Distortions The command "Finite Element/Verify" is used to check for element distortions. This command measures the deviation between the actual element and the ideal one. The worst element distortions, according to the MSC.Patran criteria, must be located outside the zones with large stress concentration or high stress gradient. Two parameters are checked separately: 1) Aspect Ratio (TRIA3 & QUAD4) The Aspect Ratio is calculated according to figure App B.5.1.1.5.1 for TRIA3 (ideal element is the equilateral triangle) and figure App B.5.1.1.5.2 for QUAD4 (ideal element is the square or the diamond). Aspect ratios smaller than 5 are recommended and always must be smaller than 10. The use of any elements with aspect ratio higher than 10 must be justified. Continued…

Issue: E


Page 33 of 68


APPENDIX B General Explanations and Guidelines, continued

Fig. App B.5.1.1.5.1 - Aspect Ratio for TRIA3 element in MSC.Patran

Fig. App B.5.1.1.5.2 - Aspect Ratio for QUAD4 element in MSC.Patran 2) Taper (QUAD4) The taper is calculated according to figure App B.5.1.1.5.3 (ideal element is the rectangle, taper = 1). Taper recommended values are between 0.8 and 1. Continued…

Issue: E


Page 34 of 68



Fig. App B.5.1.1.5.3 - Taper for QUAD4 element in MSC.Patran 5.1.1.6

Element Warping The command "Finite Element/Verify/QUAD/Warp" is used to check the element warping. This command measures the angle of deviation with respect to the element plane. Warping angle is calculated according to figure App B.5.1.1.6.1. Theta = arcsin (h/l) Warping angles smaller than 10 degrees are recommended.

Fig. App B.5.1.1.6.1 - Warp angle for QUAD4 element in MSC.Patran Continued…

Issue: E


Page 35 of 68



Internal Angles (Skew) and Skew Factor The command "Verify/QUAD or TRIA/Skew" is used to check element internal angles. This command presents the largest element internal angle, as shown in figures App B.5.1.1.7.1 & App B.5.1.1.7.2. The command shows elements with an internal angle greater than 30 degrees (default). The skew factor, is the ratio between the skew angle and 90 degrees (see figures App B.5.1.1.7.1 & App B.5.1.1.7.2), this factor must be in the range between 0 and 0.33. Elements with internal angles higher than recommended values must be located outside the zones with high stress concentration or high stress gradient.

Fig. App B.5.1.1.7.1 - TRIA3 element internal angles in MSC.Patran

Fig. App B.5.1.1.7.2 - QUAD4 element internal angles in MSC.Patran Continued…

Issue: E


Page 36 of 68



ELAS Elements It must be confirmed that all ELAS elements have zero length and have coincident nodes. This operation may be combined with the operations described in the section Coincident Nodes (see 5.1.1.1). The menu "Utilities/Display/Mark 1D Elements" is used to put elements with null length, i.e. smaller than a user defined tolerance value, into a group. Using the menu "Utilities/Display/Show/Element info", you may check the stiffness (K), the co-ordinates of the nodes, the co-ordinate system for displacements of those nodes, and the degrees of freedom affected for all elements within the group.

5.1.1.9

Rigid Elements and MPCs MPCs, and Rigid Elements, are equations that define relationships between degrees of freedom of nodes. It must be confirmed that the relationship between the degrees of freedom used are correct, and do not produce changes to local load paths due to erroneous or undesired stiffness. To verify the relationship between degrees of freedom you may to use the menu "Finite Element/Show/MPC" and "Utilities/Display/Show MPCs".

5.1.1.10 Co-Ordinate Systems The orientation and location of all general and local co-ordinate systems must be confirmed. Plots must be made showing all model co-ordinate systems used. 5.1.1.11 Element Connectivities Plots showing element connectivities must be obtained to verify they are correct. This verification is very important to ensure the correct interpretation of MSC.Nastran element forces and the correct application of pressure loads. Correct connectivity is also required to guarantee the correct operation of any software used for automated strength analysis. The command to check connectivities for 2D (TRIA3 & QUAD4) elements is "Utilities/FEM Elements/Modify 2D Connectivities (Display)". This instruction displays a vector between the first and second node for each element. The command to display connectivities for 3D (Solid elements) is "Verify/Element/Connectivity". This instruction will mark those elements with negative volume. Continued…

Issue: E


Page 37 of 68


APPENDIX B General Explanations and Guidelines, continued 5.1.1.12 Element Offsets Figures showing the element offsets in BAR/BEAM and QUAD4/TRIA3 must be obtained to verify the correct application in the model. To display the BAR/BEAM with offset, the function "Display/LBC/Element Properties/Beam Display 1D Line + Offset" option must be used. To display the offset in nodes as a vector, the function "Properties/Show/Offset @ Node 1-2" is used. To display the offset in plate elements, the function "Utilities/ Display/Display Plate Offset" is used. The offsets values of concentrated masses (CONM2 cards) are obtained with the function "Properties/Mass Offset". The CONM2 cards in MSC.Patran are shown as a triangle located in the assigned node. There is, however, a reported error in MSC.Patran where the position of mass is defined in absolute co-ordinates (option CID= -1 in CONM2 card). In this situation MSC.Patran takes these co-ordinates as the offset values. 5.1.1.13 Element Co-Ordinate Systems For the correct interpretation of element forces, stresses, and strains it is necessary to know the direction of the element co-ordinate system axes. With the function "Utilities/Display/Finite Element Coordinate Frames" you may be able to display the element co-ordinate systems for selected elements (select MSC.Nastran option). Plots showing the element co-ordinate system for areas where consistent element orientations are necessary should be obtained, checking the axes in 1D (BAR/BEAM) and 2D (QUAD4/TRIA3) elements as necessary. 5.1.1.14 Material Co-Ordinate Systems When a model uses composite materials, the material co-ordinate system, and the relative angle between the material and the element co-ordinate systems must be verified and plotted. It is recommended that the material axes are defined using an auxiliary global coordinate system in order to avoid problems due to irregular elements or future mesh modifications. Do not define Xelement by specifying the angle between G1G2 and the material axis in CQUAD4/CTRIA3 cards. The projection of the material co-ordinate system over the elements can be obtained with the command "Utilities/Display/Plot Material Orientation". Continued…

Issue: E


Page 38 of 68


APPENDIX B General Explanations and Guidelines, continued 5.1.2

Validation with MSC.Nastran These checks are complementary to those performed using MSC.Patran. Having validated the topology and properties, it is necessary to run an analysis of the model using MSC.Nastran to obtain one additional check for model consistency, and thus prevent a MSC.Nastran fatal error. It is not necessary to complete a full analysis run for this check, the analysis may be terminated after the data-checking phase of the analysis. In MSC.Nastran version 70.5, the model must be executed with SOL 101 to verify the model geometry, and to calculate the load resultant for each of the validation static load cases. MSC.Nastran will perform some element geometrical and connectivity checks. These are similar to the checks performed by MSC.Patran, but some of the assessment criteria used by MSC.Nastran differ from those used by MSC.Patran. It is still considered necessary to perform both sets of tests as the initial MSC.Patran checks will detect the most severe errors, leaving the MSC.Nastran checks to identify any residual problems. Verifications for TRIA3 & QUAD4 "Skew Angle" and for QUAD4 "Taper" are performed. The elements that do not pass the test are listed in the F06 file as either "User Information Message (UIM)" or "User Warning Message (UWM)". The "Skew Angle" is measured as shown in figure App B.5.1.2.1, and will produce a warning message when the value is smaller than 30.0 degrees for QUAD4 elements, and 10 degrees for TRIA3 elements. "Taper" is calculated as shown in figure App B.5.1.2.2 and a warning message is printed when the taper is higher than 0.5. Both "Taper" and "Skew" are very important for the efficiency of the MSC.Nastran QUAD4 element, while large warping angles will create artificial local internal forces.

Fig. App B.5.1.2.1 QUAD4 element Skew Angle in MSC.Nastran Continued…

Issue: E


Page 39 of 68



Fig. App B.5.1.2.2 QUAD4 element Taper Ratio in MSC.Nastran In order to perform the analysis check with MSC.Nastran V70.5 it is necessary to select the following PARAM values: a) PARAM,CHECKOUT,YES → To check the model geometry and the matrix consistency, and to produce the load resultant for the applied loads, in the origin of basic co-ordinate system (default). If you need the load resultant relative to a different point use the following parameter. b) PARAM,GRPNT,G0 → This defines the node about which to calculate the load resultant, the reactions, the mass and inertia, etc. If you do not use this parameter the default value is the origin of basic co-ordinate system. c) PARAM,PRTMAXIM,YES → To show the maximum displacements, maximum spcforces, maximum applied load and maximum mpcforces for each load case in basic co-ordinate system. This printing is automatic in V68 and V69 when you ask for displacement, spcforces in the "Case Control" but not in V70.5. The MSC.Nastran execution shows also if there are RBEs not correctly defined or are quasi singular. Continued…

Issue: E


Page 40 of 68



Properties and Material Cards Element properties must correspond to the relevant drawings and calculations must conform to defined modelling rules for the item concerned. The material cards must contain data which agrees with values specified either by the customer or with validated published reference documentation, e.g. MIL-HDBK, etc. Figures showing model element property data should be produced for illustration purposes. These may be generated using MSC.Patran "Properties/Show".

5.3

Constrains and Boundary Conditions Constraint data and Boundary Conditions must be applied to the correct nodes/degreesof-freedom, in the correct directions, and with the correct stiffness where relevant. Evidence must be produced to demonstrate that these criteria have been met. Cases representing each of the maximum predicted loading cases should be used to evaluate the reaction distribution through each of the constraint systems. e.g. in a HTP FEM, the most critical bending and torsion load case should be used, and also the elevator Maximum Torque load case if possible, and the results of these checks must be included in Form 4. The use of PARAM, AUTOSPC, YES is only permitted in the preparation and model validation model phase, and not in the final model. Any additional SPC’s identified by PARAM, AUTOSPC must produce negligible reactions within the final model. If these reactions are not negligible, then corrective action to the model is required. SPC cards which MSC.Nastran creates automatically when the model is run using the PARAM AUTOSPC YES may be created automatically by executing the analysis using PARAM, SPCGEN, 1. This produces a PUNCH file of the additional SPC cards identified by AUTOSPC that the model needs to analyse without stiffness problems. These cards can then be included in the Bulk Data of the FEM for the production runs using PARAM AUTOSPC NO. NB: Except when part of the main support of the model, it is only acceptable to SPC rotational DOFs, never translation DOFs, translation constraints are in general forbidden even when they generate zero reactions. Continued…

Issue: E


Page 41 of 68


APPENDIX B General Explanations and Guidelines, continued This restriction also applies for dynamic validations, as it is the intention to use the same model for both static and dynamic calculations. This will therefore be the normal configuration. A general sketch/plot of each set of boundary conditions used in the model must be created for the purposes of validation and documentation. The reaction resultant, i.e. spcforce resultant, at the reference point must be equal and opposite in sign to the applied load resultant, i.e. oload resultant.

5.4

Loads It must be confirmed that all loads applied to the model are applied in the correct directions, have the correct magnitudes, and are at the correct nodal locations. The following specific items must be taken into account: •

The total applied load in all 6 DOFs with respect to the reference point. This point is by default the origin of the basic co-ordinate system.

•

The loading distributions along the component main dimension for shear load, bending moment, and torque moment derived using MSC.Nastran internal loads.

•

The comparison between applied and theoretical load distributions from MSC.Nastran, at each load monitoring station, e.g. ribs or frame stations.

•

The reaction resultant is equal and opposite of the applied load resultant. This output is printed automatically in solutions such as SOL 101.

These checks will be included in a technical report and referenced on Form 4. When it is necessary to introduce new nodes for loading introduction, it is recommended that they are included by using an RBE3. This RBE3 is defined as follows: The new node must be the "REFERENCE GRID" for the RBE3 with either a) 6 dependant degrees-of-freedom, or b) 6 independent degrees-of-freedom as shown on page 19. In the case of b), the independent DOFs must be taken from structural GRIDs and must be selected to define completely the rigid body motion of reference point. Continued…

Issue: E


Page 42 of 68



Dynamic Checks Using Density Properties on MAT Cards Performing a normal modes calculation allows confirmation that all the nodes are adequately connected to the structure, and there are no mechanisms or low stiffness areas. This check is made using SOL 103 of MSC.Nastran using the LANCZOS method. The check is made in the following three steps: 1) Calculate the rigid body modes within the range of - 1 to + 1 Hz with the model free in space and with moving surface actuators connected if present. There must be 6, and only 6, modes i.e. three for translations and three for rotations or linear combinations of these. The frequencies will be very small, almost zero. The maximum frequency must be fRBM < 0,01 Hz. 2) Calculate the rigid body modes within the range 0 to + 1 Hz plus the rotations at 0 Hz for models including movable surfaces with the model free in space and the actuators disconnected, or if attached with RODs, these should be removed. There must be 6 rigid body modes and one rotation at 0 Hz for each movable surface. 3) Calculate the first 15 flexible modes with the model free and the actuators connected; the standard frequency range is from 1 to ≈ 40 Hz. This run is in order to detect local and spurious modes, and the flexible modes of the structure modelled. In the three executions the following rules must be fulfilled: a) This validation will be made using the material density on material card; there must be no exceptions. The object of the validation is to confirm the model not to obtain the correct flexible modes. b) All analyses must be made with "PARAM, AUTOSPC, NO". This is normal practice. c) Translation singularities must be eliminated with the correct use of RBE elements. Any rotational singularities not constrained by using PARAM, K6ROT, 1. and PARAM, SNORM, 20. must be eliminated with SPC's or RBE's, as normal practice. d) Under NO circumstances must users use field 8 of the GRID and GRIDSET cards to include permanent single point constraints. This is normal practice as the use of that field can cause the inadvertent grounding of masses during dynamic analyses or loads during static analyses if the node is used as a mass or loading point.. Singularities should be eliminated via SPC1 or SPC cards. e) The deflected shape of the model should be assessed, with drawings of the deformed structure, using the basic co-ordinate system. Areas where the displacement co-ordinate system is not the basic one should be specifically identified and plotted where possible. Continued…

Issue: E


Page 43 of 68


APPENDIX B General Explanations and Guidelines, continued Tables App B.5.5.1 to App B.5.5.3 show examples of the specific cards used for each of the above analyses for Executive, Case Control, and Bulk Data decks. ASSIGN OUTPUT2='~/neptuno/juanluis/a3xx/din/d1.op2',UNIT=12,delete ID =A3XXHTP,STATUS10C SOL 103 CEND $-------------------------------------------------$.. CASE CONTROL DECK TITLE = A3XX STATUS10C: HTP FEM LABEL = DYNAMIC VALIDATION SOL 103 LANCZOS: NORMAL MODES SUBTITLE= RUN 1: EIGENVALUES CALCULATIONS RIGID BODY MODES [-1,1] Hz. LINE=84 MAXLINES=4000000 $ Only SPCs in order to avoid singularities in the model (rotations only). ECHO=NONE SPC=510 MPC=510 $ Necessary in order to check the boundary conditions. SPCFORC=ALL MPCFORC=ALL $ Necessary in order to create the normal modes. DISP=ALL $ Method for eigenvalues calculations. METHOD = 10 $-------------------------------------------------$.. BULK DATA CARDS BEGIN BULK $.. ANALYSIS MECA PARAM POST -1 $ Method for eigenvalues calculations: Search modes from -1 to 1 Hz. $ (Rigid Body Modes) EIGRL 10 -1.0 1.0 $ UNITS USED IN THE MODEL: $ Length unit: mm ! $ stiffness unit (E in MATs card): N/mm2 ----> Force unit = N. ! $ The Time unit = s $ In order to use the coherent units is necessary to put mass in Ton. $ but normally the CONM2 card are in kg (and the RHO of MATs in kg/mm**3) $ ====> Therefore must be used PARAM,WTMASS,0.001. If the model mass is not the actual one it $ must to adjust the value in order to obtain the correct order of magnitude of the $ eigenvalues (this value do not produces changes in the output of model mass). $ 1 .. 2 .. 3 .. 4 .. 5 .. 6 .. 7 .. 8 .. 9 .. 10 . PARAM WTMASS 1.7-3 $ In order to put off PARAM AUTOSPC PARAM,AUTOSPC,NO $ Put on the "GRID POINT WEIGHT GENERATOR" PARAM GRDPNT 0 $------------------------------------------------------------------------------$ All the SPC for the boundary conditions of the model must be removed, only the necessary in $ order to avoid the rotations singularities must be presents in the model. $ The GRID card must be not contain any permanent constrain in field 8. $----------------------------------------------------------------------------$ Materials with RHO in Kg/mm**3 $ Note: It is possible to use a multiplier in RHO in order to obtain the actual mass of the $ structure $ $ Aluminium Alloy $ 1 .. 2 .. 3 .. 4 .. 5 .. 6 .. 7 .. 8 .. 9 .. 10 . $ - RHO$MAT1 1 72000. 0.33 2.7-6 $..

Table App B.5.5.1 - Example of execution: test #1

Issue: E


Continued…

Page 44 of 68


APPENDIX B General Explanations and Guidelines, continued ASSIGN OUTPUT2='~/neptuno/juanluis/a3xx/din/d2.op2',UNIT=12,delete ID =A3XXHTP,STATUS10C SOL 103 CEND $-------------------------------------------------$.. CASE CONTROL DECK $.. TITLE = A3XX STATUS10C: HTP FEM LABEL = DYNAMIC VALIDATION SOL 103 LANCZOS: NORMAL MODES. SUBTITLE= RUN 2: EIGENVALUES CALCULATIONS RIGID BODY MODES + MOBILE SURFACES ROTATIONS [0,1] Hz $ Actuators without connection LINE=84 MAXLINES=4000000 ECHO=NONE $ Only SPCs in order to avoid singularities in the model (rotations only). SPC=510 MPC=510 $ Necessary in order to check the boundary conditions. SPCFORC=ALL MPCFORC=ALL $ Necessary in order to create the normal modes. DISP=ALL $ Method for eigenvalues calculations. METHOD = 11 $-------------------------------------------------$.. BULK DATA CARDS $.. BEGIN BULK $.. ANALYSIS MECA PARAM POST -1 $ Method for eigenvalues calculations: Search modes from 0 to 1 Hz. $ (Rigid Body Modes + mobile surface rotations) $ 1 .. 2 .. 3 .. 4 .. 5 .. 6 .. 7 .. 8 .. 9 .. 10 . EIGRL 11 0.0 1.0 $ UNITS USED IN THE MODEL: $ Length unit: mm ! $ stiffness unit (E in MATs card): N/mm2 ----> Force unit = N. ! $ The Time unit = s $ In order to use the coherent units is necessary to put mass in Ton. $ but normally the CONM2 card are in kg (and the RHO of MATs in kg/mm**3) $ ====> Therefore must be used PARAM,WTMASS,0.001. If the model mass is not the actual one it $ must to adjust the value in order to obtain the correct order of magnitude of the $ eigenvalues (this value do not produces changes in the output of model mass). $ 1 .. 2 .. 3 .. 4 .. 5 .. 6 .. 7 .. 8 .. 9 .. 10 . PARAM WTMASS 1.7-3 $In order to put off PARAM AUTOSPC PARAM,AUTOSPC,NO $ Put on the "GRID POINT WEIGHT GENERATOR" PARAM GRDPNT 0 $------------------------------------------------------------------------------$ All the SPC for the boundary conditions of the model must be removed, only the necessary in $ $ order to avoid the rotations singularities must be presents in the model. $ The GRID card must be not contain any permanent constrain in field 8. $----------------------------------------------------------------------------$ Materials with RHO in Kg/mm**3 $ Note: It is possible to use a multiplier in RHO in order to obtain the actual mass of the $ structure $ $ Aluminium alloy $ $ 1 .. 2 .. 3 .. 4 .. 5 .. 6 .. 7 .. 8 .. 9 .. 10 . $ - RHO$MAT1 1 72000. 0.33 2.7-6 $..


Issue: E


Continued…

Page 45 of 68


APPENDIX B General Explanations and Guidelines, continued ASSIGN OUTPUT2='~/neptuno/juanluis/a3xx/din/d3.op2',UNIT=12,delete ID =A3XXHTP,STATUS10C SOL 103 CEND $-------------------------------------------------$.. CASE CONTROL DECK $.. TITLE = A3XX STATUS10C: HTP FEM LABEL = DYNAMIC VALIDATION SOL 103 LANCZOS: NORMAL MODES. SUBTITLE= RUN 3: EIGENVALUES CALCULATIONS FLEXIBLE MODES [0, ∞] Hz (FIRST 15TH MODES) LINE=84 MAXLINES=4000000 $ Only SPCs in order to avoid singularities in the model (rotations only). ECHO=NONE SPC=510 MPC=510 $ Necessary in order to check the boundary conditions. SPCFORC=ALL MPCFORC=ALL $ Necessary in order to create the normal modes. DISP=ALL $ Method for eigenvalues calculations. METHOD = 12 $-------------------------------------------------$.. BULK DATA CARDS $.. BEGIN BULK $.. ANALYSIS MECA PARAM POST -1 $ Method for eigenvalues calculations: Search FIRST 15TH modes from 0 to ∞ Hz. $ (Flexible modes) $ 1 .. 2 .. 3 .. 4 .. 5 .. 6 .. 7 .. 8 .. 9 .. 10 . EIGRL 12 15 $ UNITS USED IN THE MODEL: $ Length unit: mm ! $ stiffness unit (E in MATs card): N/mm2 ----> Force unit = N. ! $ The Time unit = s $ In order to use the coherent units is necessary to put mass in Ton. $ but normally the CONM2 card are in kg (and the RHO of MATs in kg/mm**3) $ ====> Therefore must be used PARAM,WTMASS,0.001. If the model mass is not the actual one it $ must to adjust the value in order to obtain the correct order of magnitude of the $ eigenvalues (this value do not produces changes in the output of model mass). $ 1 .. 2 .. 3 .. 4 .. 5 .. 6 .. 7 .. 8 .. 9 .. 10 . PARAM WTMASS 1.7-3 $In order to put off PARAM AUTOSPC PARAM,AUTOSPC,NO $ Put on the "GRID POINT WEIGHT GENERATOR" PARAM GRDPNT 0 $------------------------------------------------------------------------------$ All the SPC for the boundary conditions of the model must be removed, only the necessary in $ order to avoid the rotations singularities must be presents in the model. $ The GRID card must be not contain any permanent constrain in field 8. $----------------------------------------------------------------------------$ Materials with RHO in Kg/mm**3 $ Note: It is possible to use a multiplier in RHO in order to obtain the actual mass of the $ structure $ $ Aluminium alloy $ $ 1 .. 2 .. 3 .. 4 .. 5 .. 6 .. 7 .. 8 .. 9 .. 10 . $ - RHO$MAT1 1 72000. 0.33 2.7-6 $.. $..


Issue: E


Continued…

Page 46 of 68


APPENDIX B General Explanations and Guidelines, continued Verifications: 1. In Test #1, the following items must be taken into account: 1.1. Check using output from the GRID point weight generator that the total mass and C.G. are consistent. The following table shows an example: DIRECTION MASS AXIS SYSTEM (S)

MASS

X

3.97216E+0.3

0.000000E+00

6.120751E+03

2.540000E+02

Y

3.97216E+03

2.683207E-15

0.000000E+00

2.540000E+02

Z

3.972164E+03

2.683207E-15

6.120751E+03

0.000000E+00

X-C.G.

Y-C.G.

Z.C.G.

When the mass or the C.G. is inconsistent, the following tasks must be performed: •

Verify the mass units.

•

Check the density values in MATi cards.

•

Adjust the mass of model to the actual mass with the parameter PARAM, WTMASS, FACT in order to obtain the correct order of magnitude in the eigenvalue calculation. See example in tables App B.5.1 - 3.

1.2. Verify the lists of eigenvalues, checking there are exactly 6 rigid body modes. The frequency must be lower than 0.01 Hz. If not, perform the following tasks: •

When there are less than 6 modes, or the frequencies are higher than 0.01 Hz, it indicates that the structure is over-constrained. The "SPCFORCES" listing must be checked in detail in a static run, SOL 101, to verify which degrees of freedom are the reason of the problem and eliminate the incorrect constraints. The problem may also be due to ELAS elements with no coincident nodes.

•

Modes from 0.01 to 0.1 Hz may be acceptable in very complicated structures and when only rotational modes are affected. This must be agreed by all affected parties.

1.3. Verify the "Grid Point Singularity Table" and confirm that the translation constraints TX; TY, TZ are at the correct nodes and DOFs. Continued…

Issue: E


Page 47 of 68


APPENDIX B General Explanations and Guidelines, continued 2. In Test #2, the following items must be taken into account: 2.1. The modes obtained must be 6 rigid body modes similar to those obtained in the previous test #1, plus one more per each modelled movable surface. These modes should be plotted with MSC.Patran to show that they are real modes and not spurious mode. The frequency range is from 0.0 to 0.01 Hz. 2.2. This test is used to detect and to correct the undesired connections between movable surfaces and the main structure. It will also identify situations where the hinge line points are not perfectly aligned. 2.3. This test checks that the moving surfaces have been modelled correctly, and will detect problems with DOFs not detected in test #1. 3. In Test #3, the following items must be taken into account: 3.1. The eigenvalues of flexible modes must be consistent with the expectations for the structure under consideration. First wing bending mode should be from 4 to 10 Hz, and torque from 8 to 25 Hz, etc. 3.2. No local or spurious modes must appear in the first 15 values. Again this is best demonstrated by plotting mode shapes using MSC.Patran. 3.3. Any flexible modes outside the physical natural range may indicate an error in mass units. 3.4. Local or spurious modes in skins may indicate a singularity normal to the skin in nodes without rib or spar elements (see figure App B.5.10). In this case if the singularities appear at small frequencies then the relevant nodes must be supported, e.g. with RBE1 elements but not if the model is used for thermal analysis. 3.5. Other local modes may be due to projections of stiffness in one specific direction, e.g. BAR elements, and are not detected by PARAM, AUTOSPC, YES in the previous NASTRAN tests. In this case the definition of a local coordinate frame allowing the creation of a SPC card in the correct stiffness direction will solve the problem. The completion of these tests #1 to #3 will satisfy the requirements of Form 5. Continued…

Issue: E


Page 48 of 68



Figure App B.5.10 - Local or spurious Modes in skin

5.6

Dynamic Checks Using Concentrated Mass (Conm2 Cards) The normal modes calculation using concentrated masses allows verification of realistic modes in the structure from an actual mass distribution. Test #4 will check the rigid body modes using a real mass distribution. In this test all density values in the MAT card must be zero or blank or the structural mass will be included twice. Test #5 must also be made with masses applied via CONM2 cards and all densities on material cards must be equal to zero or blank. The object of this test is to obtain correct flexible modes for the structure. The values obtained in Test #3 were for model checking only, and are not "real" values. Test #6 is a complementary check for the A-set. The completion of these tests #4 to #6 will satisfy the requirements of Form 6. Continued…

Issue: E


Page 49 of 68



Complete Model Assembly, Quality Assurance and Documentation In addition to those items identified in chapter 6 of the main body of the document concerning the documentation that will accompany the complete model, a complete copy of the validation forms 1 to 6 must also be included.

7

Mass Data The masses of the F.E. model will be checked using PARAM, GRDPNT, GID where GID is the reference node identification. This card activates the GRID POINT WEIGHT GENERATOR module, which will calculate the masses and the moment of inertia with respect to the defined reference point. To obtain these results it is obviously necessary to include the density parameter in the MATi material cards or alternatively the actual mass distribution with concentrate mass cards, CONM2. When the actual mass distribution is supplied using CONM2 cards, all density parameters on material cards must be zero or blank without exception. NB: If GID = 0, the reference point is the origin of the basic co-ordinate system.

8

Dynamic Model Only one official dynamic model must exist, the only difference between the dynamic model and the stress model must be the units, m for dynamic and mm for stress. There must be no other differences.

9

Update Procedure Any revision to component models which must then be included in revised assemblies, or revisions to existing assemblies, will require the new assembly to be reissued under the update procedure defined in the main document, Section 9.0. Continued…

Issue: E


Page 50 of 68



Comments to Appendix A The following pages contain explanatory guidelines to completing the Forms required for conformance to this AM procedure. Two sets of guidelines are given; the first contains explanatory text defining the contents of sections of the forms to assist the user in completing them. The second set contains a completed example based on an Airbus HTP example. For the convenience of the user, the two sets have been interleaved so that the explanation guidelines for each form are immediately followed by the example for that form. NB: As Forms 5 & 6 and Forms 4 & 7 are identical in layout and manner of completion, only Form 4 and 5 has been included in the example. Continued…

Issue: E


Page 51 of 68


APPENDIX B General Explanations and Guidelines, continued QUALITY ASSURANCE OF FEM

PROJECT:

REFERENCES PART COMPONENT

A/C COMPONENT:


PAGE:

VERSION:

Finite Element Model general description: < Analyst should provide a simple description of the FE model, if necessary outlining the function to which it is being put >

Reference Coordinate System (FEM Component related to DBD) < A definition of the geometric axis system used for the model data, and the relationship of this system to that given within the relevant DBD. Include a reference for the source of data used. >

Geometry < Reference to a report or other form of data source, e.g. a CADs database or list of drawing references, used to define the geometry data used within the model. If necessary append a separate list to the end of this report and refer to the list here. > Meshing < Reference to a report or other form of data source used to define the standard of meshing density used within the model. >

Properties < Reference to a report or other form of data source used to define the standard of idealisation used within the model. If necessary append a separate list to the end of this report and refer to the list here. >

Materials < Reference to a report or other form of data source used to define the materials used within the model, and the data values assigned to them. If necessary append a separate list to the end of this report and refer to the list here. > Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 52 of 68



PROJECT:

REFERENCES PART COMPONENT: A/C COMPONENT: HTP HTP ANALYSIS IDENTIFICATION: VERSION: A320EHTPG5 02 Finite Element Model general description:

PAGE: 1

General model of half A320 HTP, including torque box, elevator and leading edge. The used elements are ROD, TRIA3, QUAD4, ELAS2, RBE1 and RBE3 (for EI and GJ calculations). DOF = 5586 (symmetric cases)/5563 (antisymmetric cases) (Kll) More information in NT-D-ADS-01003, chapter 1.

Reference Coordinate System (FEM Component related to DBD) System 0, global aircraft axes. All the grids, except for the elevator, are referred to this system. Other auxiliar systems are used to provide the correct directions for constrains, loads or material properties. More information in NT-D-ADS-01003, chapter 3. Geometry No changes affect geometry. Meshing MSC/PATRAN V9.0

Properties Only torsion box properties have been modified, according to ATL sizing. More information in NT-D-ADS-01003, chapter 6. Materials GR V Z19.775 HERCULES 8552/AS4 TH0 0.18 mm Tape has been used for the torsion box, since ATL manufacturing process is going to be used. Properties calculation has been made using ANCOLAB V3.0. Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 53 of 68



PROJECT:

FILES RECORD PART COMPONENT

A/C COMPONENT:

ANALYSIS IDENTIFICATION: COMPUTER DATASET/FILE NAME

DATE OF CREATION

PAGE:

VERSION:


Enter list of ALL associated computer files for this model and the validation activity. Include BULK DATA, submission macros, test loading and results. In particular those files that form the deliverable to Airbus should be clearly marked and perhaps grouped together for convenience. Where files are stored in a common directory, give the directory path name at the top of the table.

Name (Dept.)

Signature

Date

Compiled by: Approved by: Continued…

Issue: E


Page 54 of 68



PROJECT:

FILES RECORD PART COMPONENT: HTP

A/C COMPONENT: HTP

ANALYSIS IDENTIFICATION: A320EHTPG5

PAGE: 2

VERSION: 02

COMPUTER DATASET/FILE DATE OF NAME CREATION


a320new_hbo_updated_e.dat

19-03-01

Torsion Box Bulk DATA (right hand side)

a320new_ele.dat

16-01-01

Right Elevator Bulk DATA

a320new_hle.dat

16-01-01

Right Leading Edge Bulk DATA

a320new_fit.dat

16-01-01

Fittings Bulk DATA

a320new_act.dat

16-01-01

Actuators Bulk DATA

a320new_spc.dat

16-01-01

SPC’s to avoid singularities

a320gr_dup_new.dat

16-01-01

Duplicated grids in appropriated axes for a-set

fem0_6_e_2.dat

01-02-01

Half HTP Static Model with load test case.

a320m1.dat

13-02-01

Half HTP Static Model: (Run 1).

a320m2.dat

13-02-01


a320m3.dat

13-02-01


Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 55 of 68



PROJECT:

MODEL CHECK LIST PART COMPONENT: ANALYSIS IDENTIFICATION:

A/C COMPONENT:

PAGE: VERSION:

MODEL SIZE GRIDs

1-D CHECK LIST

2-D

3-D

SPRINGs

RIGID BODYs

DEPARTMENT RESPONSIBLE & PRINCIPAL DATA VALUE (If it is relevant)

SIGNATURE

Coincident Nodes Coincident Elements Free Edges/Free Faces Element Warping Element Twist Interior Angles Distortion of Elements Connectivity Bending Planes (BAR&BEAMs) Offsets (all elements) Positive normal (QUAD4/TRIA3) Element Co-ordinate System Material Co-ordinate System Co-ordinate Systems Rigid Elements, MPCs Computer

ANALYSIS DATA Operating System

Max. Diagonal Factor

MSC.Nastran Version

Is Grid Point Singularity. Table Empty? If YES

If NO

Attached as page No.

COMMENTS: < Against each entry in the table above, enter Name and Department/Group reference or Siglum of the person responsible for completing the checklist item. Also, where relevant, give the most appropriate value, e.g. for Element Twist quote the worst twist measured. If required, hard copy output can be appended and a reference made to the relevant page numbers in this report. The person named should then sign the end column. > < For the rest of the form, complete the entries and sign below as Compiler. > Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 56 of 68



PROJECT:

MODEL CHECK LIST PART COMPONENT: HTP ANALYSIS IDENTIFICATION: A320EHTPG5 GRIDs 1187

1-D 734 CHECK LIST

Coincident Nodes Coincident Elements Free Edges/Free Faces Element Warping Element Twist Interior Angles Distorsion of Elements Connectivity Bending Planes (BAR&BEAMs) Offsets (all elements) Positive normal (QUAD4/TRIA3) Element Co-ordinate System Material Co-ordinate System Co-ordinate Systems Rigid Elements, MPCs Computer SUN UltraSPARC/336 Max. Diagonal Factor 1.2E+06

A/C COMPONENT: HTP

PAGE: 3 VERSION: 02

MODEL SIZE 2-D 3-D SPRINGs RIGID BODYs 1338 0 105 17+14* DEPARTMENT RESPONSIBLE & SIGNATURE PRINCIPAL DATA VALUE (If it is relevant) < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > < reference to attached plot/diagramn given > ANALYSIS DATA Operating System MSC/NASTRAN Version SunOS 5.8 70.5.2 r4 (Oct 1998) Is Grid Point Singularity. Table Empty? If YES

X

If NO

Attached as page No.

COMMENTS: *The 14 RBE3 are used only for EI and GJ calculations

Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 57 of 68



PROJECT:

ANALYSIS CHECK LIST PART COMPONENT: ANALYSIS IDENTIFICATION:

A/C COMPONENT:

PAGE: VERSION:

BOUNDARY CONDITIONS DESCRIPTION: (text, figure or references) < Describe the method of constraining the model. Where necessary add appropriate diagrams to illustrate

LOAD CASE DESCRIPTION: < Describe the load case being used for model testing. >


FX

FY

FZ

MX

MY

MZ

TX

TY

TZ

RX

RY

RZ

External Work

< From the analysis of the test load case described above, complete the above table. >

< Complete one of these pages for EACH static test load case used. >

Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 58 of 68



PROJECT:

ANALYSIS CHECK LIST PART COMPONENT: HTP ANALYSIS IDENTIFICATION: A320EHTPG5 BOUNDARY CONDITIONS DESCRIPTION: (text, figure or references) See NT-D-ADS-01003, chapter 7

A/C COMPONENT: HTP

LOAD CASE DESCRIPTION:

OTW19E733

PAGE: 4 VERSION: 02

NASTRAN ID.: 306033


FX FY ≈0.0 ≈0.0 ≈0.0 ≈0.0 --TX TY 12106 14010 -6.05E+00 3.19E+01 -3.383E-10

FZ MX -1.01E+05 -1.13E+08 1.01E+05 1.13E+08 0.0 0.0 TZ RX 30085 21058 -2.77E+02 -2.28E-01 External Work

MY MZ -5.67E+07 -5.05E+05 5.67E+07 5.05E+05 0.0 0.0 RY RZ 1038 1038 2.96E-01 -2.48E+00 5.218E+06

OLOAD RES and SPCFOR RES values for Fx and Fy are roughly zero, so percentage difference between them is not significant. In rotations there are some great values, because NASTRAN plate elements do not have stiffness in the normal direction of the plate, but these singularities do not have any influence in the model performances.

Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 59 of 68



PROJECT:



Page:

BOUNDARY CONDITIONS DESCRIPTION: < Describe the model constraints. If necessary refer to (text, figure or references) an appended diagram/plot. > Check 1 Check 2 Check 3 Mass Reference Mass Standard Value Mass Output from GPWG Mode Plots & Frequency List Frequency [Hz] Check 1 Check 2 Check 3 (Densities in MATx) (Densities in MATx) (Densities in MATx) [-1,1] [0,1] [1,∞] free/free with actuators free/free without actuators Mode No: free/free with actuators 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Comments: < From the normal modes analysis, complete the table. Only entries in the UNSHADED areas will be used for the assessment. > < NB. This sheet requires a DYNAMIC acceptance signature in addition to the normal approval. >

Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 60 of 68



PROJECT:

ANALYSIS CHECK LIST: NORMAL MODES PART COMPONENT: A/C COMPONENT: Page: HTP HTP 5 ANALYSIS IDENTIFICATION: VERSION: A320EHTPG5 02 BOUNDARY CONDITIONS DESCRIPTION: Free Model (text, figure or references) NT-D-ADS-01003, section 9.2 Check 1 Check 2 Check 3 Mass Reference Mass Standard Value 335 335 335 Mass Output from GPWG 296 295 296 < reference to plot < reference to plot < reference to plot Mode Plots & Frequency List given > given > given > Frequency [Hz] Check 1 Check 2 Check 3 (Densities in MATx) (Densities in MATx) (Densities in MATx) [-1,1] [0,1] [1,∞] free/free with actuators free/free without actuators Mode No: free/free with actuators 1 2.293611E-04 3.438683E-04 2 1.701064E-04 2.326994E-04 3 9.762136E-05 1.406094E-04 4 5.137773E-05 6.064041E-06 5 2.509644E-06 1.500369E-04 6 1.662269E-04 2.142871E-04 7 1.189490E-02 1.737560E+01 8 1.956964E+01 9 2.400370E+01 10 2.467126E+01 11 3.586686E+01 12 4.168311E+01 13 4.190392E+01 14 4.250319E+01 15 4.794977E+01 Comments:

Name (Dept.)

Signature

Date

Compiled by: Approved by: Dynamic Aceptance:

Continued…

Issue: E


Page 61 of 68



PROJECT:



Page:

BOUNDARY CONDITIONS DESCRIPTION: < Describe the model constraints. If necessary refer to (text, figure or references) an appended diagram/plot. > Check 4 Check 5 Check 6 Mass Reference Mass Standard Value Mass Output from GPWG Mode Plots & Frequency List Frequency [Hz] Check 4 (M1) Check 5 (M2) Check 6 (M2) (Masses in CONMx) (Masses in CONMx) (Masses in CONMx) [-1,1] [0,∞] [0,∞] free/free with actuators Mode No: free/free with actuators free/free with actuators 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Comments: < From the normal modes analysis, complete the table. Only entries in the UNSHADED areas will be used for the assessment. > < NB. This sheet requires a DYNAMIC acceptance signature in addition to the normal approval. >

Name (Dept.)

Signature

Date


Continued…

Issue: E


Page 62 of 68



PROJECT:

THERMO-ELASTIC ANALYSIS CHECK LIST PART COMPONENT:

A/C COMPONENT:


PAGE:

VERSION:

BOUNDARY CONDITIONS DESCRIPTION: (text, figure or references) Statically determined constrain

LOAD CASE DESCRIPTION: Defined Temperature:

Load Balance Checks Applied Reaction (SPCs) Difference (%) Maximum Stress ELEMENT Value Epsilon COMMENTS:

FX

FY

FZ

MX

MY

MZ

TX

TY

TZ

RX

RY

RZ

External Work

Name (Dept.)

Signature

Date


Issue: E


Page 63 of 68


APPENDIX C 1

Introduction A Finite Element Model, used to analyse the mechanical forces and constraints coming from thermal expansion, must satisfy the usual validity criteria for mechanical FEM (See the previous pages of this manual). The most important aspects are: zero frequency rigid vibration modes, loads-constraints balance , limited displacements verifying the linear hypothesis, etc. In addition to these criteria, a FEM has to satisfy a supplementary zero stress check, which then validates the use of thermo-elastic loads. The zero stress check requires zero mechanical stresses over the entire structure with uniform expansion of the structure, uniform CTE (coefficient of thermal expansion) and arbitrary temperature loading

2

Test of Zero Stress To carry out this check, it is necessary to modify the expansion coefficients by applying identical and uniform CTE to all MAT cards of the structure. The user must input a constant value for all of the

αi,i

expansion coefficients and ensure that the

αi,j

(i#j)

coupling coefficients are set to zero. To satisfy this "Zero Stress" check, a mechanically validated model must be modified, according to the following rules: Elements: No Restriction

•

ROD:

•

BAR - BEAM: •

The offsets of shear axis must not modify the distance between the 2 nodes of the beam (vectors W1 and W2)

• •

The offsets of neutral axis (BEAM) have no influence.

TRIA-QUAD: •

If the offsets ZOFFS or Z0 are used, the parameter SNORM must be set to 0.5 instead of 20. (Due to Nastran limitations)

•

A warped QUAD results in a non-zero stress (Nastran limitation) but is generally acceptable (under the accuracy level)

•

TETRA-HEXA:

No Restriction. Continued…

Issue: E


Page 64 of 68


APPENDIX C, continued •

GENEL-DMIG: •

Generally non valid Elements for a thermo-elastic analysis (except under circumstances where their inclusion does not modify the displacements in another part of the model)

•

CELAS: •

CELAS between two coincident nodes: No Restriction

•

CELAS of non zero length not valid: see GENEL-DMIG above.

D.O.F RELATIONS: •

•

RBE2: •

RBE2 between two coincident nodes for 6dof: No Restriction

•

Otherwise: see GENEL

RBE3: •

If the slave node has no associated rigidity (Mass or Load introduction node): No Restriction

• •

Otherwise: see GENEL-DMIG

MPC: •

If the slave node is a barycenter of two master nodes with coefficients in the same ratio as the distances (Menu Airbus in Patran): No Restriction.

•

Otherwise: see GENEL-DMIG

Within the Nastran input deck, the reference temperature is set to a constant value for all materials. The parameter TEMP (INIT) = 20 and TEMP (LOAD) = 2 are inserted in the Case Control Section and similarly TEMPD, 20, 20 and TEMPD, 2, 120 into the Bulk Data Section. (The indicated values are set to perform a uniform ∆T = 100°C). Upon completion of the modifications, the whole model is run using Nastran. Applying the TEMP(INIT) parameter is preferred over the use of TREF in the MATi and the PCOMP, because it reduces the modifications to the Nastran input deck. (Within Nastran, TEMP (INIT) prevails over TREF) Setting in the Case Control parameter ESE=ALL allows the user to view the element stresses and identify those which fail to satisfy the Zero Stress criteria (values exceeding 1 Mpa). Continued…

Issue: E


Page 65 of 68


APPENDIX C, continued 3

Temperature Loading The temperature map is generated from a preliminary calculation on a Finite Element Thermal Model. The temperature map results are transferred from the thermal FEM to the mechanical FEM by the thermal team (ESAN). The team then applies temperatures directly at the nodes of the mechanical FEM by TEMP cards. It may be necessary to refine the mechanical meshing in zones of strong thermal gradients, to obtain a better appreciation of the associated mechanical constraints. If undertaken, the work must be done in agreement with the thermal team (ESAN) to ensure coherent thermal and mechanical models in these zones. For two parts of structure in contact but with different temperatures, it is necessary to have 2 coincident nodes connected by an RBE2 element. This allows the user to apply the 2 corresponding temperatures. In instances of meshing refinement of the local model, e.g. when a partial model is extracted from a global model for local studies, the temperature map must be interpolated to all nodes of the refined model.

Issue: E


Page 66 of 68


Reference documents AP2000

Procedural Documentation Handling and Writing

AP2080

Airbus Reference Language Main

MSC.Nastran

Documentation (Definitions for BAR, ROD etc.)

Glossary See AP2080 - Airbus Reference Language Main Assembly A number of parts, subassemblies or any combination thereof joined together to perform a specific function and which can be disassembled without destruction of designed use. Component Any self-contained part, combination of parts, subassemblies or units, which perform a distinctive function necessary to the operation of a system. Computer See ISO 2382/1-1984 Data See ISO 2382/1-1984 File See ISO 2382/1-1984 Input(Data)

See ISO 2382/1-1984

Quality Assurance

All those planned or systematic actions necessary to provide adequate confidence that a product or service will satisfy given needs.

Validation

The act of confirming that a specific requirement has been meet.

Approval This document has been approved by: (signatures or proof of agreement are archived together with the master document)

Function/Organization

Issue: E

Approval/Name

Structure Coc - Global F.E. Analysis - ESAE

D. Fiztsimmons

Flight Physics Coc-Loads & Aeroelasticity - EGL

S. Dirmeier


Page 67 of 68


Record of revisions Issue

Date

Summary and reasons for changes

A

Mar 1998

Initial issue.

B

Jan 1999

Including forms.

C

Feb 2001

Revision.

D

Dec 2002

Revised to reflect new Airbus organisation. Inclusion of appendix B.

E

May 2004

Revised to reflect the thermal checks (Chapter 5, appendix A, appendix B and appendix C).

If you have a query concerning the implementation or updating of this document, please contact the Owner on page 1 or the following:

Team members

Company/Department

Telephone

JC. Sourisseau

ESAET - Global FE Analysis

+33561939649

V. Rollo

ESAEU - Global FE Analysis

+441179364074

JL. de la Gándara

IESSM - Global FE Analysis

+34916241269

KD. Hamann

ESAEG - Global FE Analysis

+494074375504

For general queries or information contact: Airbus Documentation Office, Address: Airbus - 31707 Blagnac CEDEX - France e-mail: [email protected] This document and all information contained herein is the sole property of AIRBUS S.A.S. No intellectual property rights are granted by the delivery of this document or the disclosure of its content. This document shall not be reproduced or disclosed to a third party without the express written consent of AIRBUS S.A.S. This document and its content shall not be used for any purpose other than that for which it is supplied. The statements made herein do not constitute an offer. They are based on the mentioned assumptions and are expressed in good faith. Where the supporting grounds for these statements are not shown, AIRBUS S.A.S. will be pleased to explain the basis thereof.

Issue: E


Page 68 of 68

AM 2036 (10 May 2004) - Normes_fem_global

Recommend Documents