Design Sensitivity and Optimization in MSC - Nastran

MSC.
Software Corporation
2 MacArthur Place
Santa Ana, CA 92707, USA
Tel: (714) 540-8900
Fax: (714) 784-4056
Web: http://www.mscsoftware.com
Tokyo, Japan United States Munich, Germany

Tel: 81-3-3505-0266 MSC.Patran Support Tel: (+49)-89-43 19 87 0
Fax: 81-3-3505-0914 Tel: 1-800-732-7284 Fax: (+49)-89-43 61 716
Fax: (714) 979-2990
Design Sensitivity and Optimization in

MSC.NASTRAN
NAS107 Exercise Book
MSC.Nastran 2005
May 2005
NA*V2005R2*Z*Z*Z*SM-NAS107-WBK
DISCLAIMER
MSC.Software Corporation reserves the right to make changes in specifications and other information contained in this
document without prior notice.
The concepts, methods, and examples presented in this text are for illustrative and educational purposes only, and are not
intended to be exhaustive or to apply to any particular engineering problem or design. MSC.Software Corporation assumes
no liability or responsibility to any person or company for direct or indirect damages resulting from the use of any
information contained herein.
User Documentation: Copyright© 2005 MSC.Software Corporation. Printed in U.S.A. All Rights Reserved.
This notice shall be marked on any reproduction of this documentation, in whole or in part. Any reproduction or distribution
of this document, in whole or in part, without the prior written consent of MSC.Software Corporation is prohibited.
MSC and MSC. are registered trademarks and service marks of MSC.Software Corporation. NASTRAN is a registered
trademark of the National Aeronautics and Space Administration. MSC.Nastran is an enhanced proprietary version
developed and maintained by MSC.Software Corporation. MSC.Patran is a trademark of MSC.Software Corporation.
All other trademarks are the property of their respective owners.
TABLE OF CONTENTS
PAGE
1. Optimization of a 3-bar truss problem subject to static loadings. 1-1
2. Minimization of surface material required for shipping box. 2-1
3. Frame structure design with different scaling 3-1
4. Calculation of design sensitivity coefficients. 4-1
5. Mode tracking exercise. 5-1
6. Shape optimization of a planar culvert. 6-1
7. Multi-disciplinary optimization problem subject to static, normal 7-1
modes, and frequency response.
8. Design structure with strength and buckling limits satisfying both 8-1
the Euler’s and Johnson’s criteria.
9. Random response optimization 9-1
10. Discrete variable optimization 10-1
11. Enhancement of dynamic response creation 11-1
12. External response (DRESP3) 12-1
13. Avoiding A Frequency Range 13-1
14. Stiffened Plate Optimization With Objective Formed using Crossing-Subcase Responses 14-1
WORKSHOP 1
3- BAR TRUSS OPTIMIZATION SUBJECT TO
STATIC LOADING
Subcase 1 Subcase 2
X: -16,000 lbs X: 16,000 lbs
Y: -12,000 lbs Y: -12,000 lbs
NAS 107, Workshop 1, May 2005

Copyright© 2005 MSC.Software Corporation WS1-1
Design Model Description
X Objective Function: Weight minimization
X Design Variables: Cross-sectional area A1 and A2
X Constraints:
Stress Allowable: 20 ksi tension
15 ksi compression
Displacement at grid 4: X direction ≤ ±0.2 in
Y direction ≤ ±0.2 in

Q Suggested Exercise Steps
1. Open a new database and call it wkshp1.db.
2. Turn on all the labels and select front view.
3. Create new nodes for the model.
4. Create bar elements from the previous nodes
5. Make a new material called alum and sets its properties.
6. Create a 1-D rod with property set named prop_1 and sets its material property to
alum with area of 1.
7. Create another 1-D rod with property set named prop_2 and sets its material
property to alum with area of 2.
8. Under Loads/BCs, create a new nodal displacement called disp_1 and sets its
translation and rotation parameters.
9. Create another nodal displacement called disp_2 and sets its parameters.
10. Create a nodal force called force_1 and sets its force property and application
region.
11. Create another force called force_2 and sets its force properties and application
region.
12. Create a load case called case_1 and sets its properties.
13. Create another load case called case_2 and sets its displacement and force
properties..
14. Create design variables using 1-D rod with prop_1 and prop_2 and associated areas.
15. Create design objective from the design study pre-processing tools.
16. Create design constraints on X displacement, Y displacement, and axial stress.

17. Create a design study called opt_1 by including design variables, objectives and
constraints.
18. Review the design study contents using the Design Study Summary option.
19. Create an input file for optimization – Set necessary parameters.
20. Select the existing design study and global objective.
21. Create the subcases by including desired constraints. Select subcase for the job.
Generate the analysis Bulk Data file: wkshp1.bdf
22. Compare wkshp1.bdf with the sample input file.
23. Compare wkshp1.bdf with the sample input file (cont'd).
24. Compare wkshp1.bdf with the sample intput file (cont'd).
25. Submit the wkshp1.bdf for Nastran analysis and check for errors.
26. Compare the results with the sample output.
27. In Patran, read the result file using Nastran's generated file, wkshp1.op2.
28. Plot the Design Variable History on the XY plot.
29. Plot the Objective Function History on the XY plot.
30. Plot the Maximum Constraint History on the XY plot.
31. Quit MSC.Patran.

1 2 3
1 2 3
Y
Z X 4
Figure 1.1 – Geometry and Finite Elements
Figure 1.2 – Constraints and Applied Forces (Case 1) Figure 1.3 – Constraints and Applied Forces (Case 2)
Table 1.1 – Material Properties
Elastic Modulus 10e6 psi

Poisson's Ratio 0.3
Density 0.1

Step 1. Create a New Database
Open database.
d
a. File / New.
b. Enter wkshp1 for File
Name.
c. Click OK.
d. Under New Model
e
Preferences, select Based
on Model Tolerance.
f
e. Select MSC.NASTRAN for
Analysis Code. b c g
f. Select Structural for
Analysis Type.
g. OK.

Step 2. Show All Labels
a b
Show labels and change the view to

front view.
a. Show all entity labels
b. Front view
Whenever possible, deselect Auto

Execute option.

Step 3. Create New Nodes
Create new nodes

a. Elements b
b. Create/Node/Edit
c. Deselect Associate with
Geometry
d. Enter [-10 0 0] for Node Location
List
e. Apply
Repeat the same steps a – c with
[0 0 0] for Node Location List
and click Apply c
[10 0 0] for Node Location List
and click Apply d z
[0 -10 0] for Node Location List
and click Apply
e

Step 4. Create bars
Create bars
a. Elements
b. Create/Element/Edit
c. Select Bar for Shape b
d. Select Node 1 from the viewport
for Node 1 =
e. Enter Node 4 from the viewport
for Node 2 =
f. Apply
c
Repeat steps a – c with:
Node 2 for Node 1 =
Node 4 for Node 2 =
and click Apply.
Repeat steps a – c with:
Node 3 for Node 1 = d
Node 4 for Node 2 = z
e
and click Apply.
f

Step 5. Create and Define the Material’s properties
Define a material using the specified b

modulus of elasticity and
allowable stresses. e
a. Materials f
b. Create/Isotropic/Manual Input
c. Enter alum for Material Name g
d. Input Properties…
e. Enter 10e6 for Elastic Modulus =
f. Enter 0.3 for Poisson Ratio =
g. Enter 0.1 for Density =
h. OK c
i. Apply
h
i

Step 6. Create a 1-D Rod and Set its Properties
Create a 1D rod with aluminum e

properties. b f
a. Properties
b. Create/1D/Rod
c. Enter prop_1 for Property Set
Name
e. Select alum in the Select
Material databox for Material c
Name.
f. Enter 1 for Area
g. OK
h. For Select Members, click on d
Beam Element icon
h g
i
and select rod elements 1 and 3
from the view port.
i. Add
j. Apply
j

Step 7. Create another Rod and Set its Properties
Create another property with the new b e

input properties. f
a. Properties
b. Create/1D/Rod
c. Enter prop_2 for Property Set
Name
e. Select alum in the Select Material c
databox for Material Name.
f. Enter 2 for the Area
g. OK
d
h. For Select Members, click on
Beam Element icon g
h
i
and select Elm 2 for Select
Members
i. Add
j. Apply j

Step 8. Create a Load/BC
i
Create a nodal displacement called
b
disp_1. e
a. Loads/BCs f
b. Create / Displacement / Nodal.
c. Enter disp_1 for New Set Name j
d. Input Data… k
e. Enter <0,0,0> for Translations
<T1 T2 T3>
f. Enter <0,0,0> for Rotations <R1
R2 R3>
g. OK
h. Select Application Regions… l
c g
i. Select FEM under Geometry
Filter
j. Drag the mouse to select Node 1,
Node 2, and Node 3 from the
viewport for Select Nodes d
k. Add
l. OK h
m. Apply m

Step 9. Create a Load/BC (Cont.)
b i
Create another nodal displacement
called disp_2. e
a. Loads/BCs
f
b. Create / Displacement / Nodal.
c. Enter disp_2 for New Set Name j
d. Input Data…
e. Enter < , ,0> for Translations <T1 k
T2 T3>
f. Enter <0,0,0> for Rotations <R1
R2 R3>
g. OK
h. Select Application Regions…
c
i. Select FEM under Geometry l
Filter
j. Select Node 4 from the viewport g
for Select Nodes
k. Add d
l. OK h
m. Apply
m

Step 10. Apply Forces on X and Y
a
h
b
Create a new nodal force called
e
force_1. i
a. Loads/BCs j
b. Create/Force/Nodal
c. Enter force_1 for New Set Name
d. Input Data…
e. Enter < -16000, -12000, > for
Force <F1 F2 F3> k
f. OK
g. Select Application Region…
h. Select FEM under Geometry
Filter
c f
i. Select Node 4 from the viewport
for Select Nodes
j. Add
k. OK
l. Apply d
g
l

Step 11. Apply Forces on X and Y (Cont.)
a
h
b e
Create another nodal force called i
force_2.
j
a. Loads/BCs
b. Create/Force/Nodal
c. Enter force_2 for New Set Name
d. Input Data…
e. Enter < 16000, -12000, > for k
Force <F1 F2 F3>
f. OK
g. Select Application Region…
f
h. Select FEM under Geometry
Filter
i. Select Node 4 from the viewport c
for Select Nodes
j. Add
k. OK
l. Apply d
g
l

Step 12. Create Load Cases
e
b
Create a new Load Case called
case_1.
a. Load Cases.
b. Create
c. Enter case_1 as Load Case
Name
d. Input Data…
e. Under Select Individual
Loads/BCs databox, select
Displ_disp_1
Displ_disp_2
Force_force_1
c
f. OK
g. Apply f
g
Step 13. Create Load Cases (Cont.)
b
Create another Load Case called e
case_2.
a. Load Cases.
b. Create
c. Enter case_2 as Load Case
Name
d. Input Data…
e. Under Select Individual
Loads/BCs databox, select
Displ_disp_1
Displ_disp_2
Force_force_2 c
f. OK
g. Apply
d
f
g Note: The viewport stays the same.

Step 14. Create Design Variable from Tools
Use Tools to create the Design

Variables for the model.
a. Tools/Design Study/Pre-
Process…
b. Create/Design
Variable/Property
c. Select 1D for Dimensions
d. Select Rod for Type
c
e. Select prop_1 from Select
Property Set databox d
f. Select Area from Select Property e
Name databox
g. Apply
g
Step 14a. Create Design Variable from Tools (Cont.)
b
a
Use Tools to create the Design

Variables for the model.
Process…
b. Create/Design
Variable/Property
c. Select 1D for Dimensions
d. Select Rod for Type c
e. Select prop_2 from Select d
Property Set databox
f. Select Area from Select Property e
Name databox
g. Apply
h. Close
f
g h
Step 15. Create Design Objective from Tools
b
Create Objective for the Design a
Study.
a. Tools/Design Study/Pre- c
Process d
b. Create / Objective
c. Select Global as the Solution.
d. Select Weight as the Response.
e. Enter Total_Weight as the
Objective Name
e
f. Select minimize under Min/Max
selection box.
g. Apply
NAS 107, Workshop 1, May 2005 g

Step 16. Create Design Constraints from Tools (Cont.)
b
Create Design Constraints for the a
Design Study
Process
b. Create /Constraint
c. DISP_1 for Constraint Name
d. Select Node 4 for Select Node
e. Select TX option under
Displacement Component c i
f. Enter –0.2 for Lower Bound
g. Enter 0.2 for Upper Bound
h. Apply
i. DISP_2 for Constraint Name
j. Select Node 4 for Select Node d j
k. Select TY option under
Displacement Component e
l. Enter –0.2 for Lower Bound k
m. Enter 0.2 for Upper Bound
n. Apply
f g l m
h n
Step 16a. Create Design Constraints from Tools (Cont.)
Create Stress Constraints for the

Design Study b
a. Tools/Design Study/Pre- a
Process
b. Create/Constraint
c. Select Stress for the
c
Response.
d. STRESS_1 for Constraint
Name
e. Select FEM under Constraint
Region d
f. Select 1D
g. Select Rod
h. Under Select Finite Element, e
drag your mouse to select
Element 1, Element 2, and f g
Element 3 from the viewport
i. For Select Component, select h
Axial
j. Enter –15000 for Lower Bound
input box. i
k. Enter 20000 for Upper Bound
input box
l. Apply k
j
l
Step 17. Create Design Study from Tools
Create Design Study and set its

properties.
b
a. Tools/Design Study/Pre- a
Process
b. Create / Design Study
c. Enter opt_1 for Design Study
Name
i
d. Select Design Variables… l
e. For prop_1_Area, enter 0.1 under
Lower Bound and press Enter, c
and 100 under Upper Bound and
press Enter for
f. For prop_2_Area, enter 0.1 under j m
Lower Bound and press Enter,
and 100 under Upper Bound and
press Enter for d
g. OK
h. Select Objective… e h
i. Select Total_Weight for the f k
study
j. Close
k. Select Constraints…
l. Select desired constraints (all of
them for this study). n
m. Close
n. Apply
Step 18. Design Study Summary From Tools
Review contents in a Design Study.

Process b
a
b. Summary / Design Study
c. Select opt_1 from Design Study
Listbox
d. Review various contents of the
design study. c
e. Close
d
e
Step 19a. Create an Input File for Analysis – Translation Parameters
e
b
Generate an input file and sets its
parameters for analysis.
a. Analysis
b. Optimize/Entire Model/Analysis
Deck
c. Enter wkshp1 for Job Name
d. Translation Parameters…
e. For Data Output, select OP2 and
Print
f. For MSC.Nastran Version, enter c
2005
g. OK
f

Step 19b. Create an Input File for Analysis – Optimization Parameters
Generate an input file and set its

parameters for analysis (Cont.)
a. Optimization Parameters…
b. Enter 10 for Maximum Number of
Standard Design Cycles b
(DESMAX) =
c. Enter 1 for Print Design Data (P1)
every n-th cycle where n=
d. Enter 1 for Print Analysis
Results(NASPRT) every n-th
cycle where n =
e. OK
c
a d
e
Step 20a. Create an Input File for Analysis – Design Study Select
Select the Design Study b

a. Design Study Select…
b. Select opt_1

Step 20b. Create an Input File for Analysis – Global Objective Select
Select a Global Objective

b
a. Global Objective Select…
b. Select Total_Weight
a
c

Step 21a. Create an Input File for Analysis – Subcase Create
Select constraints
a. Subcases… b
b. Select case_1 from the h
Available Subcases
c. Select
d
Constraints/Objective…
d. Select Constraints
e
e. Select all of the existing
constraints.
f. OK
g. Apply
c
h. Select case_2 from the i
Available Subcases
i. Repeat steps c.-g. for case_2.
j. Cancel
f
g j

Step 21b. Create an Input File for Analysis – Subcase Select
Generate an input file and sets its

parameters for analysis
a. Subcase Select… b
b. Select 101 LINEAR STATIC for
Solution Type c
c. Under Subcases Available, select
case_1 and case_2
d. OK
e. Apply
An MSC.Nastran input file called
wkshp1.bdf will be generated.
This process of translating your
model into an input file is called
the Forward Translation. The e
Forward Translation is complete
when the Heartbeat turns green.
MSC.Nastran users should
proceed to the next step.
a
d
e
Step 22. Generated Input File
Look for the generated input file named wkshp1.bdf. It should be similar to the
output below.
SOL 200
TIME 600
$ Direct Text Input for Executive Control
CEND
TITLE = MSC.Nastran job created on 25-Apr-05 at 13:06:35
ECHO = NONE
MAXLINES = 999999999
DESOBJ(MIN) = 1
ANALYSIS = STATICS
$ Direct Text Input for Global Case Control Data
SUBCASE 1
$ Subcase name : case_1
SUBTITLE=case_1
SPC = 2
LOAD = 2
DISPLACEMENT(SORT1,REAL)=ALL
SPCFORCES(SORT1,REAL)=ALL
STRESS(SORT1,REAL,VONMISES,BILIN)=ALL
DESSUB = 21
$ Direct Text Input for this Subcase
SUBCASE 2
$ Subcase name : case_2
SUBTITLE=case_2
SPC = 2
LOAD = 4
DISPLACEMENT(SORT1,REAL)=ALL
SPCFORCES(SORT1,REAL)=ALL
STRESS(SORT1,REAL,VONMISES,BILIN)=ALL
DESSUB = 22
$ Direct Text Input for this Subcase
BEGIN BULK
PARAM POST -1
PARAM PRTMAXIM YES
PARAM NASPRT 1
$ Direct Text Input for Bulk Data
$ Elements and Element Properties for region : prop_1
PROD 1 1 1.

Step 23. Generated Input File (Cont.)
$ Pset: "prop_1" will be imported as: "prod.1"

CROD 1 1 1 4
CROD 3 1 3 4
$ Elements and Element Properties for region : prop_2
PROD 2 1 2.
$ Pset: "prop_2" will be imported as: "prod.2"
CROD 2 2 2 4
$ Referenced Material Records
$ Material Record : alum
$ Description of Material : Date: 25-Apr-05 Time: 08:32:04
MAT1 1 1.+7 .3 .101
$ Nodes of the Entire Model
GRID 1 -10. 0. 0.
GRID 2 0. 0. 0.
GRID 3 10. 0. 0.
GRID 4 0. -10. 0.
$ Loads for Load Case : case_1
SPCADD 2 4 6
LOAD 2 1. 1. 1
$ Displacement Constraints of Load Set : disp_1
SPC1 4 123456 1 2 3
$ Displacement Constraints of Load Set : disp_2
SPC1 6 3456 4
$ Loads for Load Case : case_2
LOAD 4 1. 1. 3
$ Nodal Forces of Load Set : force_1
FORCE 1 4 0 20000. -.8 -.6 0.
$ Nodal Forces of Load Set : force_2
FORCE 3 4 0 20000. .8 -.6 0.
$ Referenced Coordinate Frames
$ ...DESIGN VARIABLE DEFINITION
$ prop_1_Area
DESVAR 1 prop_1:11. .1 100. 1.
$ prop_2_Area
DESVAR 2 prop_2:22. .1 100. 1.
$ ...DEFINITION OF DESIGN VARIABLE TO ANALYSIS MODEL PARAMETER RELATIONS
DVPREL1 1 PROD 1 A
1 1.
DVPREL1 2 PROD 2 A
2 1.

Step 24. Generated Input File (Cont.)
$ ...STRUCTURAL RESPONSE IDENTIFICATION

DRESP1 1 W WEIGHT
$ DCONADD21
DCONADD 21 1 2 3
$ DCONADD22
DCONADD 22 1 2 3
$ DISP_1
DRESP1 2 DIS2 DISP 1 4
$ DISP_2
DRESP1 3 DIS3 DISP 2 4
$ STRESS_1
DRESP1 4 STR4 STRESS ELEM 2 1
2 3
$ ...CONSTRAINTS
DCONSTR 1 2 -.2 .2
DCONSTR 2 3 -.2 .2
DCONSTR 3 4 -15000. 20000.
$ ...OPTIMIZATION CONTROL
DOPTPRM DESMAX 10 FSDMAX 0 P1 1 P2 1
CONV1 .001 CONV2 1.-20 CONVDV .001 CONVPR .01
DELP .2 DELX 1. DPMIN .01 DXMIN .05
ENDDATA 3951b41d

Step 25. Submit the Model to Nastran for Analysis
a
If you have MSC.NASTRAN on your
Network, you can submit the
wkshp1.bdf for analysis.
Open MSC.NASTRAN.
a. Find and Open wkshp1.bdf .
b. Open. b
c. Run.
d. Check for fatal errors by
opening up wkshp1.f06 file as
a text document and
searching for the word
FATAL. If no fatal errors
exist, then the analysis
completed successfully.
If no matches exist, search
for the word WARNING.
Determine whether existing
WARNING messages
indicate modeling errors.
c

Step 26. View Results in the f06 File (Cont.)
Compare the results obtained in the .f06 file with the following:
***************************************************************
S U M M A R Y O F D E S I G N C Y C L E H I S T O R Y
***************************************************************
(HARD CONVERGENCE ACHIEVED)
NUMBER OF FINITE ELEMENT ANALYSES COMPLETED 7

NUMBER OF OPTIMIZATIONS W.R.T. APPROXIMATE MODELS 6
OBJECTIVE AND MAXIMUM CONSTRAINT HISTORY

---------------------------------------------------------------------------------------------------------------
OBJECTIVE FROM OBJECTIVE FROM FRACTIONAL ERROR MAXIMUM VALUE
CYCLE APPROXIMATE EXACT OF OF
NUMBER OPTIMIZATION ANALYSIS APPROXIMATION CONSTRAINT
---------------------------------------------------------------------------------------------------------------
INITIAL 4.828427E+00 -3.234952E-01
1 2.887955E+00 2.888277E+00 -1.114384E-04 -2.479180E-02
2 2.754460E+00 2.754398E+00 2.250540E-05 -4.326953E-03
3 2.717907E+00 2.717877E+00 1.105302E-05 6.591797E-05
4 2.705185E+00 2.705179E+00 2.203353E-06 1.682617E-04
5 2.697357E+00 2.697340E+00 6.098927E-06 1.448145E-03
6 2.695266E+00 2.695282E+00 -6.015127E-06 1.754883E-03

---------------------------------------------------------------------------------------------------------------
1 MSC.NASTRAN JOB CREATED ON 27-APR-05 AT 09:58:58 APRIL 27, 2005 MSC.NASTRAN 9/23/04 PAGE 137
0 SUBCASE 1
DESIGN VARIABLE HISTORY
----------------------------------------------------------------------------------------------------------------------------------
INTERNAL | EXTERNAL | |
DV. ID. | DV. ID. | LABEL | INITIAL : 1 : 2 : 3 : 4 : 5 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | PROP_1:1 | 1.0000E+00 : 7.6733E-01 : 7.8602E-01 : 8.0348E-01 : 8.2228E-01 : 8.3472E-01 :
2 | 2 | PROP_2:2 | 2.0000E+00 : 7.1794E-01 : 5.3120E-01 : 4.4529E-01 : 3.7943E-01 : 3.3640E-01 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 8 : 9 : 10 : 11 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | PROP_1:1 | 8.4580E-01 :
2 | 2 | PROP_2:2 | 3.0301E-01 :

Step 27. View Results using Patran
b
Continue to view results in PATRAN.
a. Analysis.
b. Access Results/Read
Output2/Result Entities.
c. Select Results File…
d
d. Select wkshp1.op2.
e. OK.
f. Apply.
e
f
Step 28. Post Design Variable History XYWINDOW
b
Post DesignVariableHistory Plot
a. XY Plot
c
b. Post / XYWindow
c. For Select Current XY Window,
select DesignVariableHistory
d. For Post/Unpost XY Windows,
select DesignVariableHistory
e. Apply
e
Step 29. Post Objective Function XYWINDOW
b
Post ObjectiveFunction Plot
a. XY Plot
b. Post / XYWindow
c
select ObjectiveFunction
select ObjectiveFunction
e. Apply
e
Step 30. Post Maximum Constraint Curve
b
Post Maximum Constraint History Plot
a. XY Plot
b. Post / XYWindow c
select MaximumConstraintHistory
select MaximumConstraintHistory
e. Apply

e
Step 31. Quit MSC.Patran
Quit MSC.PATRAN.
a. File / Quit.
This ends this exercise.

WORKSHOP 2
MINIMIZATION OF SURFACE MATERIAL FOR
SHIPPING BOX

1. Generate an Input File for MSC.NASTRAN users.
2. Compare the Input File
3. Submit Input File to MSC.Nastran For Analysis
4. View the Results.
5. XY Plots of Design Results.

Model Description
X Objective Function: Minimize the material surface area.
S = 2 * ( W * H + L * H + 2.0 * W * L )
X Design Variables: W (Width), H (Height), L (Length)
X Constraints:
Volume: V = H * W * L ≥ 2.0
X Initial Design Optimum Design

H = 1.5 H = 2.0
W = 1.5 W = 1.0
L = 1.5 L = 1.0
S = 18.0 S = 12.0
V = 3.375 V = 2.0

1. Generating an Input File for MSC.Nastran Users
Q Generate an input file using the data from pages 2-1 through 2-3. Use the following input file as a
starting point:
$ $ DUMMY STRUCTURAL OPTIMIZATION STATEMENTS
$ wkshp2.dat $
$ DESVAR 4 DUM 1.0 0.1 10.0
DVPREL1 10 PROD 1 4
ID MSC, D200X8 $ 4 1.0
TIME 5 DRESP1 10 DISPL DISP 1
SOL 200 2
a.CEND
Generate an input file using DCONSTR 1 10 1.0 1.0
$
the data from
TITLE = BOXpage 2-3. UseMINIMIZATION WITH A VOLUME
MATERIAL $ ---------------------------------------------------------
CONSTRAINT D200X8
the following input file as a --
DESOBJ = point.
starting $
DESSUB = 1 $ BOX DESIGN PROBLEM DEFINITION
$
SUBTITLE = INITIAL DESIGN: W=1.5, L=1.5, AND H=1.5 $ Define three design variables
$DISPLACEMENT = ALL $
STRESS = NONE DESVAR 1 W 1.5 0.1 10.0
SPC = 1 DESVAR 2 L 1.5 0.1 10.0
DESVAR 3 H 1.5 0.1 10.0
ANALYSIS = STATICS $
LOAD = 1 $ Define the Objective function using DRESP2 and DEQATN
BEGIN BULK $ the DRESP2 should be referenced by DESOBJ in Subcase
param,post,-1 level.
$
$ DRESP2...
$ ---------------------------------------------------- DEQATN...
------- $
$ $ Define a Volume response function using DRESP2 and DEQATN
$ The box must be able to hold a minimum volume of 2.0
$ DUMMY STRUCTURAL MODEL (NOT USED) $
FORCE 1 2 1.0E7 1.0 0.0 DEQATN...
0.0 DRESP2...
GRID 1 0.0 0.0 0.0 DCONSTR...
$
GRID 2 1.0 0.0 0.0
$ define the screen entry for equation
CROD 1 1 1 2 DSCREEN EQUA -100.0 1
MAT1 1 1.0E7 0.3 0.1 doptprm p1 1 p2 15
PROD 1 1 1.0 ENDDATA
SPC1 1 23456 2
SPC1 1 123456 1
$

2. Compare the MSC.NASTRAN Input File
Q The completed MSC.NASTRAN input file is shown below:
$
$ wkshp2.dat
$
ID MSC, D200X8 $
TIME 5
SOL 200
CEND
TITLE = BOX MATERIAL MINIMIZATION WITH A VOLUME CONSTRAINT D200X8 $--------------------------------------------------------------------
DESOBJ = 2 $
DESSUB = 1 $ BOX DESIGN PROBLEM DEFINITION
SUBTITLE = INITIAL DESIGN: W=1.5, L=1.5, AND H=1.5 $
$DISPLACEMENT = ALL $ Define three design variables
STRESS = NONE $
SPC =1 DESVAR 1 W 1.5 0.1 10.0
ANALYSIS = STATICS DESVAR 2 L 1.5 0.1 10.0
LOAD =1 DESVAR 3 H 1.5 0.1 10.0
BEGIN BULK $
$ $ Define the Objective function
param,post,-1 $
$ DRESP2 2 SURFACE 2
$--------------------------------------------------------------------- DESVAR 1 2 3
$ DEQATN 2 F(W,L,H)=2.0*(W*H+L*H+2.0*W*L)
$ DUMMY STRUCTURAL MODEL (NOT USED) $
FORCE 1 2 1.0E7 1.0 0.0 0.0 $ Define a Volume constraint function
GRID 1 0.0 0.0 0.0 $
GRID 2 1.0 0.0 0.0 DCONSTR 1 1 2.0 1.0E35
CROD 1 1 1 2 DEQATN 1 F(W,L,H)=W*L*H
MAT1 1 1.0E7 0.3 0.1 DRESP2 1 VOLUME 1
PROD 1 1 1.0 DESVAR 1 2 3
SPC1 1 23456 2 $ define the screen entry for equation
SPC1 1 123456 1 DSCREEN EQUA -100.0 1
$ doptprm p1 1 p2 15
$ DUMMY STRUCTURAL OPTIMIZATION STATEMENTS ENDDATA
$
DESVAR 4 DUM 1.0 0.1 10.0
DVPREL1 10 PROD 1 4
4 1.0
DRESP1 10 DISPL DISP 1 2
DCONSTR 1 10 1.0 1.0
$

Q To submit the MSC.Nastran .dat file,

Q find an available UNIX shell window and at the command prompt enter nastran
wkshp2 scr=yes. Monitor the run using the UNIX ps command:

4. View Results
When the run is completed, edit the wkshp2.f06 file and search for word FATAL. If no matches exist,
search for word WARNING. Determine whether existing WARNING messages indicate modeling errors.
While still editing wkshp2.f06, search for word “H I S T O R Y”. Compare the results obtained in the .f06
file with the following:
***************************************************************
***************************************************************
(SOFT CONVERGENCE ACHIEVED)


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 1.800000E+01 0.000000E+00
1 1.199275E+01 1.199275E+01 0.000000E+00 9.142756E-04
2 1.199275E+01 1.199275E+01 0.000000E+00 9.142756E-04

---------------------------------------------------------------------------------------------------------------
1 BOX MATERIAL MINIMIZATION WITH A VOLUME CONSTRAINT D200X8 APRIL 24, 2005 MSC.NASTRAN 4/17/05 PAGE 30
INITIAL DESIGN: W=1.5, L=1.5, AND H=1.5
0
--------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | INITIAL : 1 : 2 : 3 : 4 : 5
:
--------------------------------------------------------------------------------------------------------------------------------
1 | 1 | W | 1.5000E+00 : 9.9728E-01 : 9.9728E-01 :
2 | 2 | L | 1.5000E+00 : 9.9728E-01 : 9.9728E-01 :
3 | 3 | H | 1.5000E+00 : 2.0091E+00 : 2.0091E+00 :
4 | 4 | DUM | 1.0000E+00 : 1.0000E+00 : 1.0000E+00 :

5. XY Plots of Design Results
Fig. 2.1 – Objective Function Fig. 2.2 – Maximum Constraint

5. XY Plots of Design Results (Cont.)
Fig. 2.3 – Design Variable

WORKSHOP 3
FRAME STRUCTURE DESIGN WITH
DIFFERENT SCALINGS
10,000 lbs 5 1
A2, I2
A1, I1 A3, I3

Objectives
X Design the following frame to match the prescribed displacements at grid
points 1 and 5.
X This workshop also illustrates the importance of scaling in optimization.
10,000 lbs 5 1
A2, I2
A1, I1 A3, I3

1. Generate an input file for Part A for MSC.Nastran Users.
2. Compare the MSC.Nastran Input File for Part A.
3. Submit Input File to MSC.Nastran for analysis.
4. View Results
5. XY Plots of Objective Function and Design Variable Function for Part A.
6. Generate an input file for Part B for MSC.Nastran Users.
8. View Results
9. XY Plots of Objective Function and Design Variable Function for Part B.
10. Generate an input file for Part C for MSC.Nastran Users.
12. View Results
13. XY Plots of Objective Function and Design Variable Function for Part C.

Model Description
X Part A
X Objective Function: Match the displacements at grid points 1 and 5 to the

values shown in the table using the following formulation:
6
F ( A1, A2, A3, I1, I 2, I 3) = ∑ (ui − uit arg et )2
i =1
Grid Point Tx (in) Ty (in) Rz (in)

1 0.91449 -0.0956148 -0.00097538
5 1.068138 0.09561418 -0.0010012
X Design Variables: A1, A2, A3, I1, I2, I3
X Part B
X Redo the problem by normalizing the displacement using the foolowing

formulation for the objective function:
6 ui
F ( A1, A2, A3, I1, I 2, I 3) = ∑ (1 − t arg et
)2
i =1 ui

Model Description (Cont.)
X Part C
X Redo the problem by normalizing the design variables and the displacements
using the following formulation for the objective function:
6 ui
F (α 1,α 2,α 3, β 1, β 2, β 3) = ∑ (1 − t arg et
)2
i =1 ui
Ai Ii
where α i = and βi =
Aiinit Iiinit

1. Generate an Input File for MSC.NASTRAN Users
Generate an input file using the data from pages 3-5 and 3-6 Part A. Use the following input file as a
starting point.
$wkshp3a.dat CBAR 1 1 1 2 0.000 1.00000 0.000

$
CBAR 2 1 2 3 0.000 1.00000 0.000
ID MSC, D200X10
TIME 5 CBAR 3 1 3 4 0.000 1.00000 0.000
SOL 200 CBAR 4 1 4 5 0.000 1.00000 0.000
CEND CBAR 5 2 5 6 1.00000 0.000 0.000
TITLE = DISPLACEMENT MATCHING EXERCISE CBAR 6 2 6 7 1.00000 0.000 0.000
SUBTITLE = PORTAL FRAME SUBJECT TO HORIZONTAL LOAD NO SCALING CBAR 7 2 7 8 1.00000 0.000 0.000
ECHO = SORT CBAR 8 2 8 9 1.00000 0.000 0.000
SPC = 10 CBAR 9 3 1 10 1.00000 0.000 0.000
SET 10 = 1,5 CBAR 10 3 10 11 1.00000 0.000 0.000
DISP = 10
$ CBAR 11 3 11 12 1.00000 0.000 0.000
$ CASE CONTROL FOR ANALYSIS TYPE AND OBJECTIVE FUNCTION CBAR 12 3 12 13 1.00000 0.000 0.000
$ PBAR 1 1 100. 1.+7 1.+7 1.+7
ANALYSIS = ... PBAR 2 1 50. 1.+6 1.+6 1.+7
DESOBJ = ... PBAR 3 1 150. 0.3+7 3.+6 1.+7
$ $ -----------------------------------------------------------
$ $ Design Model Description
SUBCASE 1 $ --- Design Variables (A, I) ---
LABEL = LOAD CONDITION 1
DESVAR...
LOAD = 1
BEGIN BULK $ --- Design Variable to Properties Relation ---
PARAM,POST,-1 DVPREL1...
PARAM AUTOSPC YES $ --- Design Responses ---
PARAM GRDPNT 0 DRESP1...
$ $ --- Equation to describe Objective function ---
$ Analysis Model Description $ The goal is to match the prescribed displacements.
$ $ Define a table containing the prescribed displacements.
MAT1 1 69000.0 0.30000 2.798-6 DRESP2...
SPC1 10 123456 9 13
$
FORCE 1 5 10000.0 1.00000 0.00000
0.00000 DEQATN...
$ $
GRID 1 500.0 1000.0 0.00000 345 DTABLE...
GRID 2 250.0 1000.0 0.00000 345 $
GRID 3 0.000 1000.0 0.00000 345 $ --- Screening and Optimization Control Data ---
GRID 4 -250.0 1000.0 0.00000 345 DSCREEN,DISP,-0.5,1
GRID 5 -500.0 1000.0 0.00000 345 $ 2 3 4 5
GRID 6 -500.0 750.00 0.00000 345 6 7 8 9
GRID 7 -500.0 500.00 0.00000 345
GRID 8 -500.0 250.00 0.00000 345v 10
GRID 9 -500.0 0.0000 0.00000 DOPTPRM DESMAX 20 DELP .8
GRID 10 500.0 750.00 0.00000 345 DPMIN .1
GRID 11 500.0 500.00 0.00000 345 ENDDATA
GRID 12 500.0 250.00 0.00000 345
GRID 13 500.00 0.00000 0.00000

The completed MSC.NASTRAN input file for PART A is shown below:
$ soln3a.dat
ID MSC, D200X10
TIME 5 CBAR 1 1 1 2 0.000 1.00000 0.000
SOL 200 CBAR 2 1 2 3 0.000 1.00000 0.000
CEND CBAR 3 1 3 4 0.000 1.00000 0.000
TITLE = DISPLACEMENT MATCHING EXERCISE CBAR 4 1 4 5 0.000 1.00000 0.000
CBAR 5 2 5 6 1.00000 0.000 0.000
SUBTITLE = PORTAL FRAME SUBJECT TO HORIZONTAL LOAD NO SCALING
CBAR 6 2 6 7 1.00000 0.000 0.000
ECHO = SORT
CBAR 7 2 7 8 1.00000 0.000 0.000
SPC = 10
CBAR 8 2 8 9 1.00000 0.000 0.000
SET 10 = 1,5 CBAR 9 3 1 10 1.00000 0.000 0.000
DISP = 10 CBAR 10 3 10 11 1.00000 0.000 0.000
ANALYSIS = STATICS CBAR 11 3 11 12 1.00000 0.000 0.000
DESOBJ = 700 CBAR 12 3 12 13 1.00000 0.000 0.000
SUBCASE 1 PBAR 1 1 100. 1.+7 1.+7 1.+7
LABEL = LOAD CONDITION 1 PBAR 2 1 50. 1.+6 1.+6 1.+7
LOAD = 1 PBAR 3 1 150. 0.3+7 3.+6 1.+7
BEGIN BULK $ Design Model Description
$ --- Design Variables ---
PARAM,POST,-1
DESVAR,10,AREAHB,100.0,0.05,5000.0
PARAM AUTOSPC YES
DESVAR,20,AREAC1,50.0,0.05,5000.0
PARAM GRDPNT 0 DESVAR,30,AREAC2,150.0,0.05,5000.0
$ Analysis Model Description DESVAR,40,IHB,10.E6,0.1,1.0E9
MAT1 1 69000.0 0.30000 2.798-6 DESVAR,50,IC1,1.E6,0.1,1.0E9
SPC1 10 123456 9 13 DESVAR,60,IC2,3.E6,0.1,1.0E9
FORCE 1 5 10000.0 1.00000 0.00000 0.00000 $ --- Design Variable to Properties Relation ---
$ DVPREL1,100,PBAR,1,4,5.0,,0.0,,+DVPA
GRID 1 500.0 1000.0 0.00000 345 +DVPA,10,1.0
GRID 2 250.0 1000.0 0.00000 345 DVPREL1,200,PBAR,2,4,5.0,,0.0,,+DVPB
+DVPB,20,1.0
GRID 3 0.000 1000.0 0.00000 345
DVPREL1,300,PBAR,3,4,5.0,,0.0,,+DVPC
GRID 4 -250.0 1000.0 0.00000 345
+DVPC,30,1.0
GRID 5 -500.0 1000.0 0.00000 345
DVPREL1,400,PBAR,1,5,1.0,,0.0,,+DVPD
GRID 6 -500.0 750.00 0.00000 345 +DVPD,40,1.0
GRID 7 -500.0 500.00 0.00000 345 DVPREL1,500,PBAR,2,5,1.0,,0.0,,+DVPE
GRID 8 -500.0 250.00 0.00000 345 +DVPE,50,1.0
GRID 9 -500.0 0.0000 0.00000 DVPREL1,600,PBAR,3,5,1.0,,0.0,,+DVPF
GRID 10 500.0 750.00 0.00000 345 +DVPF,60,1.0
GRID 11 500.0 500.00 0.00000 345 $
GRID 12 500.0 250.00 0.00000 345
GRID 13 500.00 0.00000 0.00000
NAS 107, Workshop 3, May 2005 (Cont’d)
2. Compare the MSC.NASTRAN Input File (Cont.)
The completed MSC.NASTRAN input file for PART A is shown below (Cont.):
$
$ --- Design Responses ---
DRESP1,100,HD1,DISP,,10,1,,1
DRESP1,200,VD1,DISP,,11,2,,1
DRESP1,300,RD1,DISP,,12,6,,1
DRESP1,400,HD5,DISP,,20,1,,5
DRESP1,500,VD5,DISP,,21,2,,5
DRESP1,600,RD5,DISP,,22,6,,5
$
$ --- Equation to describe Objective function
$
$ These equations scale the objective to get it away from zero.
$
DRESP2,700,ERROR,10,,,,,,+DRESP2A
+DRESP2A,DTABLE,T1,T2,T3,T4,T5,T6,,+DRESP2B
+DRESP2B,DRESP1,100,200,300,400,500,600
$
DEQATN 10 OBJ(T1,T2,T3,T4,T5,T6,U1,U2,U3,U4,U5,U6) +
+ = ( (U1-T1)**2 + (U2-T2)**2 + (U3-T3)**2 + +
+ (U4-T4)**2 + (U5-T5)**2 + (U6-T6)**2 )/1.e-3
$
$ 2 3 4 5 6 7 8 9 10
DTABLE* T1 9.1449 E-01 T2 -9.56418E-02 +
* T3 -9.7538 E-04 T4 1.068138 +
* T5 9.56418E-02 T6 -1.0012 E-03 +
*
$
$ --- Screening and Optimization Control Data ---
DSCREEN,DISP,-0.5,1
$ 2 3 4 5 6 7 8 9 10
DOPTPRM DESMAX 20 DELP .8 DPMIN .1
ENDDATA


wkshp3a scr=yes. Monitor the run using the UNIX ps command:

4. View Results
When the run is completed, edit the wkshp3a.f06 file and search for the word FATAL. If no
matches exist, search for the word WARNING. Determine whether existing WARNING
messages indicate modeling errors.
While still editing wkshp3a.f06, search for the word “H I S T O R Y”. Compare the results obtained
in the .f06 file with the following:
***************************************************************
***************************************************************

---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 2.595013E+04 N/A

1 2.089655E+04 2.076502E+04 6.334271E-03 N/A
2 1.768919E+04 1.763351E+04 3.157606E-03 N/A
3 1.573590E+04 1.571276E+04 1.472791E-03 N/A
4 1.452311E+04 1.451312E+04 6.882241E-04 N/A
5 8.160316E+03 7.820767E+03 4.341637E-02 N/A
6 3.609501E+03 3.484466E+03 3.588381E-02 N/A
7 8.769125E+02 8.677664E+02 1.053991E-02 N/A
8 1.643453E+02 1.558075E+02 5.479699E-02 N/A
9 1.475247E+02 1.474937E+02 2.100113E-04 N/A
10 1.437667E+02 1.437606E+02 4.277453E-05 N/A
11 1.414624E+02 1.414595E+02 2.027896E-05 N/A
12 1.408562E+02 1.408562E+02 -1.083288E-07 N/A
13 1.408531E+02 1.408531E+02 0.000000E+00 N/A

4. View Results (Cont.)

----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 10 | AREAHB | 1.0000E+02 : 1.5000E+02 : 2.2500E+02 : 3.3750E+02 : 5.0625E+02 : 7.5937E+02 :
2 | 20 | AREAC1 | 5.0000E+01 : 7.5000E+01 : 1.1250E+02 : 1.6875E+02 : 2.5312E+02 : 3.7969E+02 :
3 | 30 | AREAC2 | 1.5000E+02 : 2.2500E+02 : 3.3750E+02 : 5.0625E+02 : 7.5937E+02 : 1.1391E+03 :
4 | 40 | IHB | 1.0000E+07 : 1.0000E+07 : 1.0000E+07 : 1.0000E+07 : 1.0000E+07 : 1.0025E+07 :
5 | 50 | IC1 | 1.0000E+06 : 1.0000E+06 : 1.0000E+06 : 1.0001E+06 : 1.0002E+06 : 1.4980E+06 :
6 | 60 | IC2 | 3.0000E+06 : 3.0000E+06 : 3.0000E+06 : 3.0000E+06 : 3.0001E+06 : 3.4156E+06 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 8 : 9 : 10 : 11 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 10 | AREAHB | 1.1391E+03 : 1.7086E+03 : 1.7558E+03 : 8.7791E+02 : 8.9285E+02 : 8.8541E+02 :
2 | 20 | AREAC1 | 5.6953E+02 : 8.5430E+02 : 1.2814E+03 : 1.9222E+03 : 2.8833E+03 : 4.3249E+03 :
3 | 30 | AREAC2 | 1.7086E+03 : 2.5629E+03 : 3.8443E+03 : 5.0000E+03 : 5.0000E+03 : 5.0000E+03 :
4 | 40 | IHB | 1.0084E+07 : 1.0369E+07 : 1.0763E+07 : 1.0764E+07 : 1.0764E+07 : 1.0764E+07 :
5 | 50 | IC1 | 2.2470E+06 : 3.3705E+06 : 5.0558E+06 : 5.0564E+06 : 5.0564E+06 : 5.0565E+06 :
6 | 60 | IC2 | 4.1130E+06 : 5.9846E+06 : 7.5850E+06 : 7.5854E+06 : 7.5855E+06 : 7.5855E+06 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 12 : 13 : 14 : 15 : 16 : 17 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 10 | AREAHB | 8.8226E+02 : 9.0826E+02 :
2 | 20 | AREAC1 | 5.0000E+03 : 5.0000E+03 :
3 | 30 | AREAC2 | 5.0000E+03 : 5.0000E+03 :
4 | 40 | IHB | 1.0764E+07 : 1.0764E+07 :
5 | 50 | IC1 | 5.0566E+06 : 5.0566E+06 :
6 | 60 | IC2 | 7.5856E+06 : 7.5856E+06 :

Moment of Inertia
Figure 3.1 Objective Function Figure 3.2 Design Variables (Moment of Inertia)
Figure 3.3 Design Variables (Area)

The completed MSC.NASTRAN input file for PART B is shown below:
$ soln3b.dat
$
ID MSC, D200X10
TIME 5
SOL 200
CEND
TITLE = DISPLACEMENT MATCHING EXERCISE
SUBTITLE = PORTAL FRAME SUBJECT TO HORIZONTAL LOAD PARTIAL SCALING
ECHO = SORT
SPC = 10
SET 10 = 1,5
DISP = 10
ANALYSIS = STATICS
DESOBJ = 700
SUBCASE 1
LOAD = 1
BEGIN BULK
PARAM,POST,-1
PARAM AUTOSPC YES
PARAM GRDPNT 0
$
$ Analysis Model Description
$
MAT1 1 69000.0 0.30000 2.798-6
SPC1 10 123456 9 13
FORCE 1 5 10000.0 1.00000 0.00000 0.00000
$
GRID 1 500.0 1000.0 0.00000 345
GRID 2 250.0 1000.0 0.00000 345
GRID 3 0.000 1000.0 0.00000 345
GRID 4 -250.0 1000.0 0.00000 345
GRID 5 -500.0 1000.0 0.00000 345
GRID 6 -500.0 750.00 0.00000 345
GRID 7 -500.0 500.00 0.00000 345
GRID 8 -500.0 250.00 0.00000 345
GRID 9 -500.0 0.0000 0.00000
GRID 10 500.0 750.00 0.00000 345
GRID 11 500.0 500.00 0.00000 345
GRID 12 500.0 250.00 0.00000 345
GRID 13 500.00 0.00000 0.00000
NAS 107, Workshop 3, May 2005 (Cont.)

6. Generate an Input File for MSC.NASTRAN Users (Cont.)
The completed MSC.NASTRAN input file for PART B is shown below (Cont.):
CBAR 1 1 1 2 0.000 1.00000 0.000
CBAR 2 1 2 3 0.000 1.00000 0.000
CBAR 3 1 3 4 0.000 1.00000 0.000
CBAR 4 1 4 5 0.000 1.00000 0.000
CBAR 5 2 5 6 1.00000 0.000 0.000
CBAR 6 2 6 7 1.00000 0.000 0.000
CBAR 7 2 7 8 1.00000 0.000 0.000
CBAR 8 2 8 9 1.00000 0.000 0.000
CBAR 9 3 1 10 1.00000 0.000 0.000
CBAR 10 3 10 11 1.00000 0.000 0.000
CBAR 11 3 11 12 1.00000 0.000 0.000
CBAR 12 3 12 13 1.00000 0.000 0.000
PBAR 1 1 100. 1.+7 1.+7 1.+7
PBAR 2 1 50. 1.+6 1.+6 1.+7
PBAR 3 1 150. 0.3+7 3.+6 1.+7
$
$ Design Model Description
DESVAR,10,AREAHB,100.0,0.05,5000.0
DESVAR,20,AREAC1,50.0,0.05,5000.0
DESVAR,30,AREAC2,150.0,0.05,5000.0
DESVAR,40,IHB,10.E6,0.1,1.0E9
DESVAR,50,IC1,1.E6,0.1,1.0E9
DESVAR,60,IC2,3.E6,0.1,1.0E9
$
$
$ --- Design Variable to Properties Relation ---
$
DVPREL1,100,PBAR,1,4,5.0,,0.0,,+DVPA
+DVPA,10,1.0
DVPREL1,200,PBAR,2,4,5.0,,0.0,,+DVPB
+DVPB,20,1.0
DVPREL1,300,PBAR,3,4,5.0,,0.0,,+DVPC
+DVPC,30,1.0
DVPREL1,400,PBAR,1,5,1.0,,0.0,,+DVPD
+DVPD,40,1.0
DVPREL1,500,PBAR,2,5,1.0,,0.0,,+DVPE
+DVPE,50,1.0
DVPREL1,600,PBAR,3,5,1.0,,0.0,,+DVPF
+DVPF,60,1.0
(Cont.)
The completed MSC.NASTRAN input file for PART B is shown below (Cont.):
$
DRESP1,100,HD1,DISP,,10,1,,1
DRESP1,200,VD1,DISP,,11,2,,1
DRESP1,300,RD1,DISP,,12,6,,1
DRESP1,400,HD5,DISP,,20,1,,5
DRESP1,500,VD5,DISP,,21,2,,5
DRESP1,600,RD5,DISP,,22,6,,5
$
$
+DRESP2B,DRESP1,100,200,300,400,500,600
$
$
$234567812345678
+ =( (1.-U1/T1)**2 + (1.-U2/T2)**2 + (1.-U3/T3)**2 + +
+ (1.-U4/T4)**2 + (1.-U5/T5)**2 + (1.-U6/T6)**2 )/1.e-3
$ 2 3 4 5 6 7 8 9 10
DTABLE* T1 9.1449 E-01 T2 -9.56418E-02 +
* T3 -9.7538 E-04 T4 1.068138 +
* T5 9.56418E-02 T6 -1.0012 E-03 +
*
$
DSCREEN,DISP,-0.5,1
$ 2 3 4 5 6 7 8 9 10
ENDDATA


wkshp3b scr=yes. Monitor the run using the UNIX ps command:

8. View Results
When the run is completed, edit the wkshp3b.f06 file and search for the word FATAL. If no
While still editing wkshp3b.f06, search for the word “H I S T O R Y”. Compare the results obtained
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
1 9.252209E+04 9.028305E+04 2.480030E-02 N/A
2 4.683422E+04 4.613752E+04 1.510049E-02 N/A
3 2.708072E+04 2.688389E+04 7.321415E-03 N/A
4 1.904624E+04 1.899343E+04 2.780052E-03 N/A
5 1.604644E+04 1.603576E+04 6.656264E-04 N/A
6 1.504332E+04 1.504140E+04 1.280970E-04 N/A
7 1.481502E+04 1.481492E+04 6.591750E-06 N/A
8 1.481185E+04 1.481185E+04 0.000000E+00 N/A


----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 10 | AREAHB | 1.0000E+02 : 1.5000E+02 : 2.2500E+02 : 3.3750E+02 : 5.0625E+02 : 7.5937E+02 :
2 | 20 | AREAC1 | 5.0000E+01 : 7.5000E+01 : 1.1250E+02 : 1.6875E+02 : 2.5312E+02 : 3.7969E+02 :
3 | 30 | AREAC2 | 1.5000E+02 : 2.2500E+02 : 3.3750E+02 : 5.0625E+02 : 7.5937E+02 : 8.7651E+02 :
4 | 40 | IHB | 1.0000E+07 : 1.0000E+07 : 1.0000E+07 : 1.0000E+07 : 1.0000E+07 : 1.0000E+07 :
5 | 50 | IC1 | 1.0000E+06 : 1.0000E+06 : 1.0000E+06 : 1.0000E+06 : 1.0000E+06 : 1.0000E+06 :
6 | 60 | IC2 | 3.0000E+06 : 3.0000E+06 : 3.0000E+06 : 3.0000E+06 : 3.0000E+06 : 3.0000E+06 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 8 : 9 : 10 : 11 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 10 | AREAHB | 1.1391E+03 : 1.7086E+03 : 1.7883E+03 :
2 | 20 | AREAC1 | 5.6953E+02 : 8.5430E+02 : 8.5599E+02 :
3 | 30 | AREAC2 | 8.7845E+02 : 8.9912E+02 : 8.9195E+02 :
4 | 40 | IHB | 1.0000E+07 : 1.0000E+07 : 1.0000E+07 :
5 | 50 | IC1 | 1.0001E+06 : 1.0002E+06 : 1.0002E+06 :
6 | 60 | IC2 | 3.0000E+06 : 3.0001E+06 : 3.0001E+06 :

9. XY Plot of Design Results
Objective Function
Area
Figure 3.4 Objective Function Figure 3.5 Design Variables (Area)

9. XY Plot of Design Results (Cont.)
Moment of Inertia
Figure 3.6 Design Variables (Moment of Inertia)

10. Generate an Input File for Part C
Generate an input file called wkshp3c.dat using wkshp3b.dat as a starting point. The completed
MSC.Nastran input file for PART C is shown below.
$ soln3c.dat
ID MSC, D200X10
TIME 5
SOL 200
CEND
SUBTITLE = PORTAL FRAME SUBJECT TO HORIZONTAL LOAD FULL SCALING
ECHO = SORT
SPC = 10
SET 10 = 1,5
DISP = 10
ANALYSIS = STATICS
DESOBJ = 700
SUBCASE 1
LOAD = 1
BEGIN BULK
PARAM,POST,-1
PARAM AUTOSPC YES
PARAM GRDPNT 0
MAT1 1 69000.0 0.30000 2.798-6
SPC1 10 123456 9 13
FORCE 1 5 10000.0 1.00000 0.00000 0.00000
GRID 1 500.0 1000.0 0.00000 345
GRID 2 250.0 1000.0 0.00000 345
GRID 3 0.000 1000.0 0.00000 345
GRID 4 -250.0 1000.0 0.00000 345
GRID 5 -500.0 1000.0 0.00000 345
GRID 6 -500.0 750.00 0.00000 345
GRID 7 -500.0 500.00 0.00000 345
GRID 8 -500.0 250.00 0.00000 345
GRID 9 -500.0 0.0000 0.00000
GRID 10 500.0 750.00 0.00000 345
GRID 11 500.0 500.00 0.00000 345
GRID 12 500.0 250.00 0.00000 345
GRID 13 500.00 0.00000 0.00000
(Cont.)
10. Generate an Input File for Part C (Cont.)
CBAR 1 1 1 2 0.000 1.00000 0.000
CBAR 2 1 2 3 0.000 1.00000 0.000
CBAR 3 1 3 4 0.000 1.00000 0.000
CBAR 4 1 4 5 0.000 1.00000 0.000
CBAR 5 2 5 6 1.00000 0.000 0.000
CBAR 6 2 6 7 1.00000 0.000 0.000
CBAR 7 2 7 8 1.00000 0.000 0.000
CBAR 8 2 8 9 1.00000 0.000 0.000
CBAR 9 3 1 10 1.00000 0.000 0.000
CBAR 10 3 10 11 1.00000 0.000 0.000
CBAR 11 3 11 12 1.00000 0.000 0.000
CBAR 12 3 12 13 1.00000 0.000 0.000
PBAR 1 1 100. 1.+7 1.+7 1.+7
PBAR 2 1 50. 1.+6 1.+6 1.+7
PBAR 3 1 150. 0.3+7 3.+6 1.+7
$
$ Note these are unit design variables
$ You will have to scale final dvs by coeff.
$
DESVAR,10,AREAHB,1.0,0.05,5000.0
DESVAR,20,AREAC1,1.0,0.05,5000.0
DESVAR,30,AREAC2,1.0,0.05,5000.0
DESVAR,40,IHB,1.0,0.1,1.0E9
DESVAR,50,IC1,1.0,0.1,1.0E9
DESVAR,60,IC2,1.0,0.1,1.0E9
$
DVPREL1,100,PBAR,1,4,5.0,,0.0,,+DVPA
+DVPA,10,100.0
DVPREL1,200,PBAR,2,4,5.0,,0.0,,+DVPB
+DVPB,20,50.0
DVPREL1,300,PBAR,3,4,5.0,,0.0,,+DVPC
+DVPC,30,150.0
DVPREL1,400,PBAR,1,5,1.0,,0.0,,+DVPD
+DVPD,40,10.0E6
DVPREL1,500,PBAR,2,5,1.0,,0.0,,+DVPE
+DVPE,50,1.0E6
DVPREL1,600,PBAR,3,5,1.0,,0.0,,+DVPF
+DVPF,60,3.0E6
$
NAS 107, Workshop 3, May 2005 (Cont.)

10. Generate an Input File for Part C (Cont.)

DRESP1,100,HD1,DISP,,10,1,,1
DRESP1,200,VD1,DISP,,11,2,,1
DRESP1,300,RD1,DISP,,12,6,,1
DRESP1,400,HD5,DISP,,20,1,,5
DRESP1,500,VD5,DISP,,21,2,,5
DRESP1,600,RD5,DISP,,22,6,,5
$
$
+DRESP2B,DRESP1,100,200,300,400,500,600
$
$
+ =( (1.-U1/T1)**2 + (1.-U2/T2)**2 + (1.-U3/T3)**2 + +
+ (1.-U4/T4)**2 + (1.-U5/T5)**2 + (1.-U6/T6)**2 )/1.e-3
$ 2 3 4 5 6 7 8 9 10
DTABLE* T1 9.1449 E-01 T2 -9.56418E-02 +
* T3 -9.7538 E-04 T4 1.068138 +
* T5 9.56418E-02 T6 -1.0012 E-03 +
*
$
DSCREEN,DISP,-0.5,1
$ 2 3 4 5 6 7 8 9 10
$DOPTPRM DESMAX 20
$ TRY THE ABOVE DOPTRM AND CHECK OUT DISPLACEMENTS
ENDDATA


wkshp3c scr=yes. Monitor the run using the UNIX ps command:

12. View Results
When the run is completed, edit the wkshp3c.f06 file and search for the word FATAL. If no
While still editing wkshp3c.f06, search for the word “H I S T O R Y”. Compare the results obtained
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------

1 6.804968E+04 6.841313E+04 -5.312622E-03 N/A
2 2.294679E+04 2.295197E+04 -2.258452E-04 N/A
3 6.495106E+03 6.494557E+03 8.458105E-05 N/A
4 1.201931E+03 1.202051E+03 -1.000284E-04 N/A
5 2.801859E+01 3.169961E+01 -1.161221E-01 N/A
6 2.332087E-01 5.317960E+01 -9.956147E-01 N/A
7 6.716316E-02 8.171839E-02 -1.781145E-01 N/A
8 2.312893E-04 5.673932E-04 -5.923651E-01 N/A
9 4.758398E-07 4.783054E-07 -5.154831E-03 N/A
10 4.783054E-07 4.783054E-07 0.000000E+00 N/A


----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 10 | AREAHB | 1.0000E+00 : 1.5000E+00 : 2.2500E+00 : 3.3750E+00 : 3.5171E+00 : 3.5419E+00 :
2 | 20 | AREAC1 | 1.0000E+00 : 1.5000E+00 : 2.2500E+00 : 3.3750E+00 : 5.0625E+00 : 7.5937E+00 :
3 | 30 | AREAC2 | 1.0000E+00 : 1.5000E+00 : 2.2500E+00 : 3.3750E+00 : 3.4002E+00 : 3.0024E+00 :
4 | 40 | IHB | 1.0000E+00 : 5.0000E-01 : 3.7034E-01 : 4.0011E-01 : 4.3056E-01 : 4.2357E-01 :
5 | 50 | IC1 | 1.0000E+00 : 1.5000E+00 : 2.2500E+00 : 3.3750E+00 : 5.0625E+00 : 7.5937E+00 :
6 | 60 | IC2 | 1.0000E+00 : 1.5000E+00 : 2.2500E+00 : 3.3750E+00 : 5.0625E+00 : 7.5937E+00 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 8 : 9 : 10 : 11 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 10 | AREAHB | 4.2234E+00 : 4.7694E+00 : 4.9785E+00 : 4.9993E+00 : 4.9993E+00 :
2 | 20 | AREAC1 | 1.1391E+01 : 9.9712E+00 : 9.9995E+00 : 1.0000E+01 : 1.0000E+01 :
3 | 30 | AREAC2 | 3.6283E+00 : 3.3220E+00 : 3.3333E+00 : 3.3333E+00 : 3.3333E+00 :
4 | 40 | IHB | 5.5075E-01 : 4.9831E-01 : 4.9999E-01 : 5.0000E-01 : 5.0000E-01 :
5 | 50 | IC1 | 1.1256E+01 : 9.8309E+00 : 9.9949E+00 : 9.9995E+00 : 9.9995E+00 :
6 | 60 | IC2 | 5.3253E+00 : 5.1368E+00 : 5.0065E+00 : 5.0003E+00 : 5.0003E+00 :

Objective Function
Area
Figure 3.7 Objective Function Figure 3.8 Design Variables (Area)

Moment Of Inertia
Figure 3.9 Design Variables (Area)

WORKSHOP 4
CALCULATION OF DESIGN SENSITIVITY
COEFFICIENTS
Objective: Calculate the design sensitivity coefficients for the following spring/mass
system

Model Description
X Design Variables – Spring constants K1, K2, K3
- Damping Coefficients GE1, GE2, GE3

X Constraints:
Displacements at Grid 1 (X direction) ≤ ± 0.2 m
Note:
For frequency responses, the default output is in terms of real and imaginary format. If the
magnitude/phase format is needed, use DISP( PHASE ) = n.

1. Generate an input file for MSC.Nastran Users.
2. Compare the MSC.Nastran Input File.
4. View Results
5. Perform Finite Difference Check
6. Changing DESVAR value.
7. View Results

Generate an input file using the data from pages 4-3 and 4-4. Use the following input file as a starting
point. $
$ wkshp4.dat
$
TIME 5
SOL 200 $
CEND
TITLE=TWO MASS DIRECT FREQUENCY SENSITIVITY
D108G7v
DISPL(PHASE) = ALL
$
$ ADD THE REST OF THE CASE CONTROL
$
.
.
$
BEGIN BULK
$
$ ANALYTICAL MODEL
$
CELAS1 20 20 1 1
CELAS1 21 21 1 1 2 1
CELAS1 22 22 2 1
CONM2 10 1 1.
CONM2 11 2 2.
DAREA 201 1 1 100. 2
1 100.
FREQ 100 4.0 6.0
GRID 1 1. 0. 0.
23456
GRID 2 2. 0. 0.
23456
MAT1 200 1.0E7 0.3 0.1
PELAS 20 1.0E3
PELAS 21 1.5E3
PELAS 22 2.0E3
RLOAD1 200 201 210
TABLED1 210
1. 1. 10. 1. ENDT
$
$ ADD DESIGN MODEL BELOW
$
.
.
$
ENDDATA
The completed MSC.NASTRAN input file is shown below:
$ DESVAR 3 K3 2.0E3 2.0E2 2.0E4

$ soln4.dat DESVAR 4 GE1 .1 .05 .15
$ DESVAR 5 GE2 .1 .05 .15
TIME 5 DESVAR 6 GE3 .1 .05 .15
SOL 200 $ DRESP1 10 U1 FRDISP 1
1
CEND
2
TITLE=TWO MASS DIRECT FREQUENCY SENSITIVITY DVPREL1 1 PELAS 20 3 1.0E3
D108G7v 1 1.0
DISPL(PHASE) = ALL DVPREL1 2 PELAS 21 3 1.5E3
SUBCASE 2 2 1.0
DVPREL1 3 PELAS 22 3 2.0E3
DESSUB = 2
3 1.0
ANALYSIS = DFREQ DVPREL1 4 PELAS 20 4 .05
FREQ = 100 4 1.0
DLOAD = 200 DVPREL1 5 PELAS 21 4 .05
dsaprt(end=sens) = all 5 1.0
DVPREL1 6 PELAS 22 4 .05
BEGIN BULK
6 1.0
$ FREQ 100 4.0 6.0
CELAS1 20 20 1 1 GRID 1 1. 0. 0. 23456
CELAS1 21 21 1 1 2 1 GRID 2 2. 0. 0. 23456
MAT1 200 1.0E7 0.3 0.1
CELAS1 22 22 2 1
PELAS 20 1.0E3
CONM2 10 1 1. PELAS 21 1.5E3
CONM2 11 2 2. PELAS 22 2.0E3
DAREA 201 1 1 100. 2 1 RLOAD1 200 201 210
100. TABLED1 210
DCONSTR 2 10 -2.0E-1 -2.0E-1 1. 1. 10. 1. ENDT
ENDDATA
DESVAR 1 K1 1.0E3 1.0E2 1.0E4
DESVAR 2 K2 1.5E3 1.5E2 1.5E4


Q Find an available UNIX shell window and at the command prompt enter nastran
wkshp4 scr=yes. Monitor the run using the UNIX ps command.

4. View Results
When the run is completed, edit the wkshp4.f06 file and search for the word FATAL. If no matches
exist, search for the word WARNING. Determine whether existing WARNING messages
While still editing wkshp4.f06, search for the word “M A T R I X”.

5. Perform Finite Difference Check
5. Perform a finite difference check of the results by:

a. Change the value of DESVAR 1 to 1.1E3
b. Set PARAM, OPTEXIT, 3 to exit after finite element analysis and
constraint evaluation.
c. Submit the run and determine the response of f = 4.0 hz, grid 1,
component 1. Let’s call this RPERT.
d. Compute a finite difference sensitivity
ΔR RPERT − RBASE
=
ΔX 1.1E 3 − 1.0 E 3
e. Compare with the computed sensitivity.

6. Changing DESVAR value
Repeat Step 5 by setting DESVAR 1 to 1.1E3

7. View Results
Compare the results obtained in the .f06 file with the following:
****************************************************************************
* *
* D E S I G N S E N S I T I V I T Y M A T R I X O U T P U T *
* *
* *
* R E S P O N S E S E N S I T I V I T Y C O E F F I C I E N T S *
* *
****************************************************************************
--------------------------------------------------------------------------------------------------------------------------------
DRESP1 ID= 10 RESPONSE TYPE= FRDISP GRID ID= 1 COMP NO= 1 SEID= 0
SUBCASE RESP VALUE FREQ/TIME DESIGN VARIABLE COEFFICIENT DESIGN VARIABLE COEFFICIENT DESIGN VARIABLE COEFFICIENT
--------------------------------------------------------------------------------------------------------------------------------
2 1.8718E-01 4.0000E+00 1 K1 -2.0951E-04 2 K2 -7.1772E-06 3 K3 -1.2402E-04
4 GE1 -4.8650E-02 5 GE2 9.2470E-05 6 GE3 -7.5393E-02
2 1.3771E-01 6.0000E+00 1 K1 5.1171E-05 2 K2 1.2517E-05 3 K3 1.5027E-04
4 GE1 -3.3135E-02 5 GE2 1.2923E-02 6 GE3 -2.5227E-02
--------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------
2 1.6842E-01 4.0000E+00 1 K1 -1.3863E-04 2 K2 3.5824E-06 3 K3 -1.5645E-04
4 GE1 -3.7918E-02 5 GE2 -1.2604E-04 6 GE3 -7.9682E-02
2 1.6226E-01 6.0000E+00 1 K1 1.2268E-04 2 K2 -6.5162E-06 3 K3 1.0405E-04
4 GE1 -2.9513E-02 5 GE2 -5.7992E-03 6 GE3 -6.2802E-02

WORKSHOP 5
MODE TRACKING EXERCISE
Objective: Keep track of the mode shapes for the first ten modes of the following fixed-
fixed beam.

Model Description
X Objective Function – Weight Minimization.
X Design Variables = B, H
X Constraints:
Eigenvalue Allowable: 9.8696 < λ1 < 11.942
39.4784 < λ2 < 43.525

Exercise Procedure
X Part A
For this exercise, define the BAR property using the PBAR entry and
specify the constraints using eigenvalues. Keep track of the mode shapes
for the first ten modes. The design variables are width (b) and height (h).
The properties on the PBAR entries are defined using the area (A) and
moment of inertia (I). They are related by the following equations:
b * h3
A = b* h I=
12
Use DVPREL2 to relate A, I, b and h by writing the equations given.
X Part B
Redo the previous exercise defining the BAR property using the PBARL entry. Specify
the constraints using the frequencies shown below.
Eigenvalue Allowable: 0.50 < f1 < 0.55

1.0 < f2 < 1.05

Generate an input file for Part A using the data from pages 5-3 and 5-5. Use the following input file as a
starting point.
$ wkshp5a.dat
$ PBAR, 30, 40, 0.5, 1.042E-2,4.167E-2
TIME 10 $
SOL 200 MAT1, 40, 1.0E+7, , 0.33, 0.1
CEND $
$ INPUT CASE CONTROL $IGRL SID V1 V2 ND
$ EIGRL, 50, , , 10
. $
. $------------------------------------------------------
. $ DESIGN MODEL:
$ $------------------------------------------------------
BEGIN BULK $
param,post,-1 $ INPUT DESIGN MODEL
$ $
$------------------------------------------------------------ .
$ ANALYSIS MODEL: .
$------------------------------------------------------------ .
$ $
$...FIXED-FIXED CANTILEVER, TWO PLANES OF MOTION: ENDDATA
GRID, 10, , 0., 0., 0., , 123456
GRID, 11, , 10., 0., 0., , 4
GRID, 12, , 20., 0., 0., , 4
GRID, 13, , 30., 0., 0., , 4
GRID, 14, , 40., 0., 0., , 4
GRID, 15, , 50., 0., 0., , 4
GRID, 16, , 60., 0., 0., , 4
GRID, 17, , 70., 0., 0., , 4
GRID, 18, , 80., 0., 0., , 4
GRID, 19, , 90., 0., 0., , 4
GRID, 20, , 100., 0., 0., , 123456
GRID, 21, , 0., 1., 0., , 123456
$
CBAR, 1, 30, 10, 11, 21
CBAR, 2, 30, 11, 12, 21
CBAR, 3, 30, 12, 13, 21
CBAR, 4, 30, 13, 14, 21
CBAR, 5, 30, 14, 15, 21
CBAR, 6, 30, 15, 16, 21
CBAR, 7, 30, 16, 17, 21
CBAR, 8, 30, 17, 18, 21
CBAR, 9, 30, 18, 19, 21
CBAR, 10, 30, 19, 20, 21

The completed MSC.NASTRAN input file for PART A is shown below:
$
$ soln5a.dat
$
TIME 10
SOL 200
CEND
analysis = MODES
dessub = 100
desobj(min) = 203
METHOD = 50
DISPL = ALL
MODTRAK=10
BEGIN BULK
PARAM,POST,-1
PARAM, NASPRT, 1
$------------------------------------------------------------------
$ ANALYSIS MODEL:
$------------------------------------------------------------------
$...FIXED-FIXED BEAM , TWO PLANES OF MOTION:
GRID, 10, , 0., 0., 0., , 123456
GRID, 11, , 10., 0., 0., , 4
GRID, 12, , 20., 0., 0., , 4
GRID, 13, , 30., 0., 0., , 4
GRID, 14, , 40., 0., 0., , 4
GRID, 15, , 50., 0., 0., , 4
GRID, 16, , 60., 0., 0., , 4
GRID, 17, , 70., 0., 0., , 4
GRID, 18, , 80., 0., 0., , 4
GRID, 19, , 90., 0., 0., , 4
GRID, 20, , 100., 0., 0., , 123456
GRID, 21, , 0., 1., 0., , 123456
$
CBAR, 1, 30, 10, 11, 21
CBAR, 2, 30, 11, 12, 21
CBAR, 3, 30, 12, 13, 21
CBAR, 4, 30, 13, 14, 21
CBAR, 5, 30, 14, 15, 21
CBAR, 6, 30, 15, 16, 21
CBAR, 7, 30, 16, 17, 21
CBAR, 8, 30, 17, 18, 21
CBAR, 9, 30, 18, 19, 21
CBAR, 10, 30, 19, 20, 21
NAS 107, Workshop 5, May 2005 (Cont’d)
$
PBAR, 30, 40, 0.5, 1.042E-2,4.167E-2
$
MAT1, 40, 1.0E+7, , 0.33, 0.1
$
$IGRL SID V1 V2 ND
EIGRL, 50, , , 10
$
$---------------------------------------------------------------------
$ DESIGN MODEL:
$---------------------------------------------------------------------
$
$ MODE TRACKING INPUT DATA:
$
MODTRAK 10 1 10 0.8
$
$...design variables, b & h (cross-sectional dimensions):
DESVAR, 101, B, 0.5, 0.1, 2.0
DESVAR, 102, H, 1.0, 0.1, 2.0
$
$...relate to changes in cross-sectional properties:
DVPREL2,110, PBAR, 30, 4, 1.E-3, , 121, , +
+, DESVAR, 101, 102
DVPREL2,111, PBAR, 30, 5, 1.E-3, , 122, , +
+, DESVAR, 102, 101
DVPREL2,112, PBAR, 30, 6, 1.E-3, , 122, , +
+, DESVAR, 101, 102
DEQATN 121 A(B,H) = B*H $ ...equations for
DEQATN 122 I(B,H) = B*H**3/12. $ rectangular sections
$
$...specify first two eigenvalues as design responses:
DRESP1, 201, LAMBDA1,EIGN, , , 1
DRESP1, 202, LAMBDA2,EIGN, , , 2
$
$...and constrain them.
DCONSTR,100, 202, 9.8696 11.942 $
DCONSTR,100, 201, 39.4784 43.525 $
$
$...define weight minimization as the objective
DRESP1, 203, W, WEIGHT
$
doptprm,desmax, 10, p1,1,p2,15
ENDDATA


wkshp5a scr=yes. Monitor the run using the UNIX ps command.

4. View Results
When the run is completed, edit the wkshp5a.f06 file and search for the word FATAL. If no
While still editing wkshp5a.f06, search for the word “T R A C K I N G”. Compare the results
obtained in the .f06 file with the following:
*******************************************
* *
* D E S I G N O P T I M I Z A T I O N *
* *
* M O D E T R A C K I N G *
* *
*******************************************
-----------------------------------------
| DESIGN CYCLE 3 |
-----------------------------------------
*** MODES FOR THIS SUBCASE IN THE CURRENT DESIGN CYCLE HAVE SUCCESSFULLY BEEN TRACKED ***
FOLLOWING IS A SUMMARY OF MODE TRACKING ACTIVITY WHERE:

INITIAL ORDER - MODE NUMBERS FROM INITIAL DESIGN, AS IDENTIFIED IN THE DESIGN MODEL
TRACKED AS - MODE NUMBERS AS TRACKED IN THE CURRENT DESIGN CYCLE
--------------------------------------------------------------------------------------------------------------------------
| INITIAL ORDER | 1 | 2 |
--------------------------------------------------------------------------------------------------------------------------
| TRACKED AS | 2 | 1 |
--------------------------------------------------------------------------------------------------------------------------

While still editing wkshp5a.f06, search for the word “H I S T O R Y”. Compare the results obtained
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 5.000000E+00 2.492728E+00
1 4.705015E+00 4.705064E+00 -1.043859E-05 8.363646E-01
2 4.388538E+00 4.388587E+00 -1.108270E-05 2.914904E-01
3 4.729199E+00 4.729098E+00 2.127523E-05 1.397610E-02
4 4.732932E+00 4.732931E+00 1.007488E-07 -5.643043E-05
5 4.732931E+00 4.732931E+00 0.000000E+00 -5.643043E-05

---------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------
:
--------------------------------------------------------------------------------------------------------------------------------
1 | 101 | B | 5.0000E-01 : 6.4889E-01 : 8.1890E-01 : 9.6606E-01 : 9.7291E-01 : 9.7291E-01
:
2 | 102 | H | 1.0000E+00 : 7.2510E-01 : 5.3591E-01 : 4.8953E-01 : 4.8647E-01 : 4.8647E-01
:

Objective Function
Dimensions
Figure 5.1 Objective Function Figure 5.2 Design Variables

Figure 5.3 Maximum Constraint

Generate an input file for Part B using the data from pages 5-3 and 5-5. Use the following input file as a
starting point.
$ wkshp5b.dat
TIME 10
SOL 200
CEND
$ INPUT CASE CONTROL
.
.
.
BEGIN BULK
param,post,-1
$---------------------------------------------------------------------
$ ANALYSIS MODEL:
$---------------------------------------------------------------------
$...FIXED-FIXED CANTILEVER, TWO PLANES OF MOTION:
GRID, 10, , 0., 0., 0., , 123456
GRID, 11, , 10., 0., 0., , 4
GRID, 12, , 20., 0., 0., , 4
GRID, 13, , 30., 0., 0., , 4
GRID, 14, , 40., 0., 0., , 4
GRID, 15, , 50., 0., 0., , 4
GRID, 16, , 60., 0., 0., , 4
GRID, 17, , 70., 0., 0., , 4
GRID, 18, , 80., 0., 0., , 4
GRID, 19, , 90., 0., 0., , 4
GRID, 20, , 100., 0., 0., , 123456
GRID, 21, , 0., 1., 0., , 123456
$
CBAR, 1, 30, 10, 11, 21
CBAR, 2, 30, 11, 12, 21
CBAR, 3, 30, 12, 13, 21
CBAR, 4, 30, 13, 14, 21
CBAR, 5, 30, 14, 15, 21
CBAR, 6, 30, 15, 16, 21
CBAR, 7, 30, 16, 17, 21
CBAR, 8, 30, 17, 18, 21
CBAR, 9, 30, 18, 19, 21
CBAR, 10, 30, 19, 20, 21
PBAR, 30, 40, 0.5, 1.042E-2,4.167E-2
MAT1, 40, 1.0E+7, , 0.33, 0.1
$
$IGRL SID V1 V2 ND
EIGRL, 50, , , 10
$---------------------------------------------------------------------
$ DESIGN MODEL:
$---------------------------------------------------------------------
$ INPUT DESIGN MODEL
.
.
.
$
ENDDATA

The completed MSC.NASTRAN input file for PART B is shown below:
$ soln3b.dat
ID MSC, D200X10
TIME 5
SOL 200
CEND
SUBTITLE = PORTAL FRAME SUBJECT TO HORIZONTAL LOAD PARTIAL SCALING
ECHO = SORT
SPC = 10
SET 10 = 1,5
DISP = 10
ANALYSIS = STATICS
DESOBJ = 700
SUBCASE 1
LOAD = 1
BEGIN BULK
PARAM,POST,-1
PARAM AUTOSPC YES
PARAM GRDPNT 0
MAT1 1 69000.0 0.30000 2.798-6
SPC1 10 123456 9 13
FORCE 1 5 10000.0 1.00000 0.00000 0.00000
$
GRID 1 500.0 1000.0 0.00000 345
GRID 2 250.0 1000.0 0.00000 345
GRID 3 0.000 1000.0 0.00000 345
GRID 4 -250.0 1000.0 0.00000 345
GRID 5 -500.0 1000.0 0.00000 345
GRID 6 -500.0 750.00 0.00000 345
GRID 7 -500.0 500.00 0.00000 345
GRID 8 -500.0 250.00 0.00000 345
GRID 9 -500.0 0.0000 0.00000
GRID 10 500.0 750.00 0.00000 345
GRID 11 500.0 500.00 0.00000 345
GRID 12 500.0 250.00 0.00000 345
GRID 13 500.00 0.00000 0.00000
(Cont.)
CBAR 1 1 1 2 0.000 1.00000 0.000

CBAR 2 1 2 3 0.000 1.00000 0.000
CBAR 3 1 3 4 0.000 1.00000 0.000
CBAR 4 1 4 5 0.000 1.00000 0.000
CBAR 5 2 5 6 1.00000 0.000 0.000
CBAR 6 2 6 7 1.00000 0.000 0.000
CBAR 7 2 7 8 1.00000 0.000 0.000
CBAR 8 2 8 9 1.00000 0.000 0.000
CBAR 9 3 1 10 1.00000 0.000 0.000
CBAR 10 3 10 11 1.00000 0.000 0.000
CBAR 11 3 11 12 1.00000 0.000 0.000
CBAR 12 3 12 13 1.00000 0.000 0.000
PBAR 1 1 100. 1.+7 1.+7 1.+7
PBAR 2 1 50. 1.+6 1.+6 1.+7
PBAR 3 1 150. 0.3+7 3.+6 1.+7
$
DESVAR,10,AREAHB,100.0,0.05,5000.0
DESVAR,20,AREAC1,50.0,0.05,5000.0
DESVAR,30,AREAC2,150.0,0.05,5000.0
DESVAR,40,IHB,10.E6,0.1,1.0E9
DESVAR,50,IC1,1.E6,0.1,1.0E9
DESVAR,60,IC2,3.E6,0.1,1.0E9
$
$
$
DVPREL1,100,PBAR,1,4,5.0,,0.0,,+DVPA
+DVPA,10,1.0
DVPREL1,200,PBAR,2,4,5.0,,0.0,,+DVPB
+DVPB,20,1.0
DVPREL1,300,PBAR,3,4,5.0,,0.0,,+DVPC
+DVPC,30,1.0
DVPREL1,400,PBAR,1,5,1.0,,0.0,,+DVPD
+DVPD,40,1.0
DVPREL1,500,PBAR,2,5,1.0,,0.0,,+DVPE
+DVPE,50,1.0
DVPREL1,600,PBAR,3,5,1.0,,0.0,,+DVPF
+DVPF,60,1.0
(Cont.)
$
DRESP1,100,HD1,DISP,,10,1,,1
DRESP1,200,VD1,DISP,,11,2,,1
DRESP1,300,RD1,DISP,,12,6,,1
DRESP1,400,HD5,DISP,,20,1,,5
DRESP1,500,VD5,DISP,,21,2,,5
DRESP1,600,RD5,DISP,,22,6,,5
$
$
+DRESP2B,DRESP1,100,200,300,400,500,600
$
$
$234567812345678
+ =( (1.-U1/T1)**2 + (1.-U2/T2)**2 + (1.-U3/T3)**2 + +
+ (1.-U4/T4)**2 + (1.-U5/T5)**2 + (1.-U6/T6)**2 )/1.e-3
$ 2 3 4 5 6 7 8 9 10
DTABLE* T1 9.1449 E-01 T2 -9.56418E-02 +
* T3 -9.7538 E-04 T4 1.068138 +
* T5 9.56418E-02 T6 -1.0012 E-03 +
*
$
DSCREEN,DISP,-0.5,1
$ 2 3 4 5 6 7 8 9 10
ENDDATA


wkshp5b scr=yes. Monitor the run using the UNIX ps command.

8. View Results
While still editing wkshp5b.f06, search for the word “H I S T O R Y”. Compare the results obtained
***************************************************************
***************************************************************

---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 5.000000E+00 8.676274E-01
1 5.258287E+00 5.258245E+00 8.070851E-06 4.084157E-01

2 5.278220E+00 5.278224E+00 -8.130640E-07 1.370379E-01
3 5.209185E+00 5.209221E+00 -6.956822E-06 -2.185843E-02
4 4.839862E+00 4.839878E+00 -3.349767E-06 -3.792524E-03
5 4.738291E+00 4.738292E+00 -1.006348E-07 8.218288E-04
6 4.738292E+00 4.738292E+00 0.000000E+00 8.218288E-04
---------------------------------------------------------------------------------------------------------------
1 APRIL 26, 2005 MSC.NASTRAN 4/17/05 PAGE 153
--------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------
1 | 101 | B | 1.0000E+00 : 7.5390E-01 : 6.0856E-01 : 5.2348E-01 : 4.9529E-01 : 4.8713E-01
2 | 102 | H | 5.0000E-01 : 6.9747E-01 : 8.6733E-01 : 9.9511E-01 : 9.7719E-01 : 9.7269E-01
--------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 8 : 9 : 10 : 11
--------------------------------------------------------------------------------------------------------------------------------
1 | 101 | B | 4.8713E-01 :
2 | 102 | H | 9.7269E-01 :

9. XY Plot of Design Results
Objective Function
Dimensions

9. XY Plot of Design Results (Cont.)
Constraints
Figure 5.6 Maximum Constraints

WORKSHOP 6
SHAPE OPTIMIZATION OF A PLANAR
CULVERT
(See seminar notes for further detailed description of the model).

Model Description
X Objective Function – Volume Minimization.
X Constraints: Von Mises Stresses < 31,000 psi
X Design Variables: use the Analytical Boundary Shape method to define shape basis
vectors.
Hints
1. Create an auxiliary model by putting CBAR elements around the outside edge of the culvert
model.
2. For the auxiliary model, fix the right and top edges similar to the direct input of the shape
method.
3. Apply the appropriate boundary conditions at the left (symmetric b.c.) and bottom edges
(fixed b.c.)
4. Apply enforced displacements around the hold by creating seven subcases – one at each
grid point around the hole.
5. Include the proper BNDGRIDs and DVBSHAPs – one DVBSHAP for each design variable.
6. If a smoother shape is desired, reduce the size of the fractional change in the design
variable at the expense of more iterations.

point. GRID, 4,, 6.00000, 0.00000,.00
GRID, 5,, 2.89464, 0.78478,.00
$ wkshp6.dat GRID, 6,, 3.79369, 0.75885,.00
nastran q4skew=10.,q4taper=100. GRID, 7,, 4.69274, 0.73293,.00
$ culvert problem using the analytical boundary shape GRID, 8,, 5.59178, 0.70700,.00
SOL 200 $ GRID, 9,, 2.60164, 1.49178,.00
GRID, 10,, 3.46229, 1.46585,.00
TIME 100
GRID, 11,, 4.32293, 1.43993,.00
CEND GRID, 12,, 5.18357, 1.41400,.00
TITLE=CULVERT EXAMPLE USING EXTERNAL AUXILIARY STRUCTURE GRID, 13,, 0.00000, 3.00000,.00
SUBTITLE=THE PRIMARY STRUCTURE GRID, 14,, 0.78478, 2.89464,.00
ANALYSIS = STATICS GRID, 15,, 1.49178, 2.60164,.00
SPC=25 GRID, 16,, 2.12100, 2.12100,.00
LOAD=1 GRID, 17,, 3.00578, 2.12100,.00
DISP=ALL GRID, 18,, 3.89057, 2.12100,.00
STRESS=all GRID, 19,, 4.77535, 2.12100,.00
GRID, 20,, 0.00000, 3.73200,.00
$ call out objective function
GRID, 21,, 0.68985, 3.66176,.00
desobj=... GRID, 22,, 1.32785, 3.46643,.00
$ call out design constraint GRID, 23,, 1.91400, 3.14600,.00
dessub=... GRID, 24,, 2.67052, 3.14600,.00
$ ---------------------------------------------------- GRID, 25,, 3.42704, 3.14600,.00
$ START OF AUXILIARY MODEL SUBCASE GRID, 26,, 4.18357, 3.14600,.00
AUXCASE GRID, 27,, 0.00000, 4.46400,.00
AUXMODEL = 1 GRID, 28,, 0.59493, 4.42888,.00
GRID, 29,, 1.16393, 4.33122,.00
TITLE = AUXILIARY MODEL
GRID, 30,, 1.70700, 4.17100,.00
DISP=ALL GRID, 31,, 2.33526, 4.17100,.00
$ add auxiliary model subcases--one for each loading GRID, 32,, 2.96352, 4.17100,.00
condition GRID, 33,, 3.59178, 4.17100,.00
SUBCASE.. GRID, 34,, 0.00000, 5.19600,.00
LOAD=.. GRID, 35,, 0.50000, 5.19600,.00
SPC =.. GRID, 36,, 1.00000, 5.19600,.00
$ GRID, 37,, 1.50000, 5.19600,.00
. GRID, 38,, 2.00000, 5.19600,.00
GRID, 39,, 2.50000, 5.19600,.00
.
GRID, 40,, 3.00000, 5.19600,.00
$ CQUAD4, 1,101, 1, 2, 6, 5
BEGIN BULK
CQUAD4, 2,101, 2, 3, 7, 6
$ -----------------------------------------------------
CQUAD4, 3,101, 3, 4, 8, 7
$ ANALYTICAL MODEL
CQUAD4, 4,101, 5, 6, 10, 9
PARAM,POST,-1
CQUAD4, 5,101, 6, 7, 11, 10
PARAM,CDIF,NO
CQUAD4, 6,101, 7, 8, 12, 11
PARAM,NEWSEQ,-1
CQUAD4, 7,101, 9, 10, 17, 16
GRID, 1,, 3.00000, 0.00000,.00
CQUAD4, 8,101, 10, 11, 18, 17
GRID, 2,, 4.00000, 0.00000,.00
CQUAD4, 9,101, 11, 12, 19, 18
GRID, 3,, 5.00000, 0.00000,.00
Copyright© 2005 MSC.Software Corporation WS6-4 (cont.)
Input File (cont.) dresp1 2 von-mis stress pshell 9
101
CQUAD4, 10,101, 13, 14, 21, 20
DCONSTR 10 2 -3.100e43.100e4
CQUAD4, 11,101, 14, 15, 22, 21
doptprm DESMAX 25 APRCOD 1
CQUAD4, 12,101, 15, 16, 23, 22
param,nasprt,1
CQUAD4, 13,101, 20, 21, 28, 27
$
CQUAD4, 14,101, 21, 22, 29, 28
BEGIN BULK AUXMODEL = 1
CQUAD4, 15,101, 22, 23, 30, 29
$ -------------------------------------------------------
CQUAD4, 16,101, 27, 28, 35, 34
$ AUXILIARY MODEL
CQUAD4, 17,101, 28, 29, 36, 35
CBAR 1 1 1 2 1. 1.
CQUAD4, 18,101, 29, 30, 37, 36
CBAR 2 1 2 3 1. 1.
CQUAD4, 19,101, 16, 17, 24, 23
CBAR 3 1 3 4 1. 1.
CQUAD4, 20,101, 17, 18, 25, 24
$
CQUAD4, 21,101, 18, 19, 26, 25
CBAR 11 1 4 8 1. 1.
CQUAD4, 22,101, 23, 24, 31, 30
CBAR 12 1 8 12 1. 1.
CQUAD4, 23,101, 24, 25, 32, 31
CBAR 13 1 12 19 1. 1.
CQUAD4, 24,101, 25, 26, 33, 32
CBAR 14 1 19 26 1. 1.
CQUAD4, 25,101, 30, 31, 38, 37
CBAR 15 1 26 33 1. 1.
CQUAD4, 26,101, 31, 32, 39, 38
CBAR 16 1 33 40 1. 1.
CQUAD4, 27,101, 32, 33, 40, 39
$
FORCE 1 34 0 1250. -1.
CBAR 21 1 34 35 1. 1.
FORCE 1 35 0 2500. -1.
CBAR 22 1 35 36 1. 1.
FORCE 1 36 0 2500. -1.
CBAR 23 1 36 37 1. 1.
FORCE 1 37 0 2500. -1.
CBAR 24 1 37 38 1. 1.
FORCE 1 38 0 2500. -1.
CBAR 25 1 38 39 1. 1.
FORCE 1 39 0 2500.00 -1.
CBAR 26 1 39 40 1. 1.
FORCE 1 40 0 1250. -1.
$
PSHELL,101,102,.44
MAT1,102,2.+7,,.3,0.731-3
CBAR 31 1 1 5 1. 1.
SPC1,25,345,1,THRU,40 CBAR 32 1 5 9 1. 1.
SPC1,25,126,1,THRU,4 CBAR 33 1 9 16 1. 1.
spc1,25,6,6,7,10,11 CBAR 34 1 16 15 1. 1.
spc1,25,6,17,18 CBAR 35 1 15 14 1. 1.
spc1,25,6,21,thru,25 CBAR 36 1 14 13 1. 1.
spc1,25,6,28,thru,32 CBAR 41 1 13 20 1. 1.
SPC1,25,1,13,20,27,34 CBAR 42 1 20 27 1. 1.
$--------------------------------------------------------------
CBAR 43 1 27 34 1. 1.
$ DESIGN MODEL
bndgrid,... PBAR 1 102 20.0 1.0 1.0
. MAT1,102,2.+7,,.3,0.731-3
. $ ADD AUXILIARY MODEL LOADING
desvar,.... $ USE ENFORCED DISPLACEMENT
. .
. .
dvbshap,... .
.
ENDDATA
.
dresp1 5 volume volume

The completed MSC.Nastran input file is shown below:
$ soln6.dat BEGIN BULK
$ $
nastran q4skew=10.,q4taper=100. $ ANALYTICAL MODEL
$ $
$ culvert problem using the analytical boundary shape PARAM,POST,-1
$ PARAM,CDIF,NO
SOL 200 $ PARAM,NEWSEQ,-1
TIME 100 GRID, 1,, 3.00000, 0.00000,.00
CEND GRID, 2,, 4.00000, 0.00000,.00
TITLE=CULVERT EXAMPLE USING EXTERNAL AUXILIARY STRUCTURE GRID, 3,, 5.00000, 0.00000,.00
D200CS GRID, 4,, 6.00000, 0.00000,.00
SUBTITLE=THE PRIMARY STRUCTURE GRID, 5,, 2.89464, 0.78478,.00
ANALYSIS = STATICS GRID, 6,, 3.79369, 0.75885,.00
SPC=25 GRID, 7,, 4.69274, 0.73293,.00
LOAD=1 GRID, 8,, 5.59178, 0.70700,.00
DISP=ALL GRID, 9,, 2.60164, 1.49178,.00
STRESS=all GRID, 10,, 3.46229, 1.46585,.00
DESSUB = 10 GRID, 11,, 4.32293, 1.43993,.00
desobj = 5 GRID, 12,, 5.18357, 1.41400,.00
$ GRID, 13,, 0.00000, 3.00000,.00
AUXCASE GRID, 14,, 0.78478, 2.89464,.00
AUXMODEL = 1 GRID, 15,, 1.49178, 2.60164,.00
TITLE = AUXILIARY MODEL GRID, 16,, 2.12100, 2.12100,.00
$ SPC = 250 GRID, 17,, 3.00578, 2.12100,.00
DISP=ALL GRID, 18,, 3.89057, 2.12100,.00
$ GRID, 19,, 4.77535, 2.12100,.00
SUBCASE 1 GRID, 20,, 0.00000, 3.73200,.00
LOAD=100 GRID, 21,, 0.68985, 3.66176,.00
spc = 100 GRID, 22,, 1.32785, 3.46643,.00
SUBCASE 2 GRID, 23,, 1.91400, 3.14600,.00
LOAD=101 GRID, 24,, 2.67052, 3.14600,.00
spc = 101 GRID, 25,, 3.42704, 3.14600,.00
SUBCASE 3 GRID, 26,, 4.18357, 3.14600,.00
LOAD=102 GRID, 27,, 0.00000, 4.46400,.00
spc = 102 GRID, 28,, 0.59493, 4.42888,.00
SUBCASE 4 GRID, 29,, 1.16393, 4.33122,.00
LOAD=103 GRID, 30,, 1.70700, 4.17100,.00
spc = 103 GRID, 31,, 2.33526, 4.17100,.00
SUBCASE 5 GRID, 32,, 2.96352, 4.17100,.00
LOAD=104 GRID, 33,, 3.59178, 4.17100,.00
spc = 104 GRID, 34,, 0.00000, 5.19600,.00
SUBCASE 6 GRID, 35,, 0.50000, 5.19600,.00
LOAD=105 GRID, 36,, 1.00000, 5.19600,.00
spc = 105 GRID, 37,, 1.50000, 5.19600,.00
SUBCASE 7 GRID, 38,, 2.00000, 5.19600,.00
LOAD=106 GRID, 39,, 2.50000, 5.19600,.00
spc = 106 GRID, 40,, 3.00000, 5.19600,.00
NAS 107,
$ Workshop 6, May 2005
CQUAD4, 1,101, 1, 2, 6, 5
bndgrid 123456 34 thru 40
CQUAD4, 2,101, 2, 3, 7, 6
bndgrid 123456 4 8 12 19 26 33
CQUAD4, 3,101, 3, 4, 8, 7
CQUAD4, 4,101, 5, 6, 10, 9
bndgrid 2 2 3
CQUAD4, 5,101, 6, 7, 11, 10
bndgrid 1 20 27
CQUAD4, 6,101, 7, 8, 12, 11
bndgrid 12 1 5 9
CQUAD4, 7,101, 9, 10, 17, 16
CQUAD4, 8,101, 10, 11, 18, 17
$
CQUAD4, 9,101, 11, 12, 19, 18
desvar 1 b1 3. -1.e6 1.e6
CQUAD4, 10,101, 13, 14, 21, 20
desvar 2 b2 3. -1.e6 1.e6
CQUAD4, 11,101, 14, 15, 22, 21
desvar 3 b3 3. -1.e6 1.e6
CQUAD4, 12,101, 15, 16, 23, 22
desvar 4 b4 3. -1.e6 1.e6
CQUAD4, 13,101, 20, 21, 28, 27
desvar 5 b5 3. -1.e6 1.e6
CQUAD4, 14,101, 21, 22, 29, 28
desvar 6 b6 3. -1.e6 1.e6
CQUAD4, 15,101, 22, 23, 30, 29
desvar 7 b7 3. -1.e6 1.e6
CQUAD4, 16,101, 27, 28, 35, 34
$
CQUAD4, 17,101, 28, 29, 36, 35
dvbshap 1 1 1 .01
CQUAD4, 18,101, 29, 30, 37, 36
dvbshap 2 1 2 .01
CQUAD4, 19,101, 16, 17, 24, 23
dvbshap 3 1 3 .01
CQUAD4, 20,101, 17, 18, 25, 24
dvbshap 4 1 4 .01
CQUAD4, 21,101, 18, 19, 26, 25
dvbshap 5 1 5 .01
CQUAD4, 22,101, 23, 24, 31, 30
dvbshap 6 1 6 .01
CQUAD4, 23,101, 24, 25, 32, 31
dvbshap 7 1 7 .01
CQUAD4, 24,101, 25, 26, 33, 32
$
CQUAD4, 25,101, 30, 31, 38, 37
dresp1 5 volume volume
CQUAD4, 26,101, 31, 32, 39, 38
dresp1 2 von-mis stress pshell 9
CQUAD4, 27,101, 32, 33, 40, 39
101
FORCE 1 34 0 1250. -1.
DCONSTR 10 2 -3.100e43.100e4
FORCE 1 35 0 2500. -1.
doptprm DESMAX 25 APRCOD 1
FORCE 1 36 0 2500. -1.
param,nasprt,1
FORCE 1 37 0 2500. -1.
$
FORCE 1 38 0 2500. -1.
BEGIN BULK AUXMODEL = 1
FORCE 1 39 0 2500.00 -1.
$
FORCE 1 40 0 1250. -1.
CBAR 1 1 1 2 1. 1.
PSHELL,101,102,.44
CBAR 2 1 2 3 1. 1.
MAT1,102,2.+7,,.3,0.731-3
CBAR 3 1 3 4 1. 1.
SPC1,25,345,1,THRU,40
$
SPC1,25,126,1,THRU,4
CBAR 11 1 4 8 1. 1.
spc1,25,6,6,7,10,11
CBAR 12 1 8 12 1. 1.
spc1,25,6,17,18
CBAR 13 1 12 19 1. 1.
spc1,25,6,21,thru,25
CBAR 14 1 19 26 1. 1.
spc1,25,6,28,thru,32
CBAR 15 1 26 33 1. 1.
SPC1,25,1,13,20,27,34
CBAR 16 1 33 40 1. 1.
$
$ design model
$

$ $
CBAR 21 1 34 35 1. 1. spcd,100,13,2,1.
CBAR 22 1 35 36 1. 1. spcd,101,14,1,.259,14,2,.9659
CBAR 23 1 36 37 1. 1. spcd,102,15,1,.5,15,2,.866
CBAR 24 1 37 38 1. 1. spcd,103,16,1,.707,16,2,.707
CBAR 25 1 38 39 1. 1. spcd,104,9,1,.866,9,2,.5
CBAR 26 1 39 40 1. 1. spcd,105,5,1,.9659,5,2,.259
$ spcd,106,1,1,1.
CBAR 31 1 1 5 1. 1. $
CBAR 32 1 5 9 1. 1. ENDDATA
CBAR 33 1 9 16 1. 1.
CBAR 34 1 16 15 1. 1.
CBAR 35 1 15 14 1. 1.
CBAR 36 1 14 13 1. 1.
$
CBAR 41 1 13 20 1. 1.
CBAR 42 1 20 27 1. 1.
CBAR 43 1 27 34 1. 1.
$
PBAR 1 102 20.0 1.0 1.0
MAT1,102,2.+7,,.3,0.731-3
$
spcadd,100,200,250
spcadd,101,201,250
spcadd,102,202,250
spcadd,103,203,250
spcadd,104,204,250
spcadd,105,205,250
spcadd,106,206,250
$
SPC1,250,345,1,THRU,40
SPC1,250,123456,34,THRU,40
SPC1,250,123456,4,8,12,19,26,33
SPC1,250,1,13,20,27
SPC1,250,2,1,THRU,4
$
spc,200,13,2
spc,201,14,12
spc,202,15,12
spc,203,16,12
spc,204,9,12
spc,205,5,12
spc,206,1,1



4. View Results
While still editing wkshp6.f06, search for the word “H I S T O R Y”. Compare the results obtained in
the .f06 file with the following:
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 7.214704E+00 -1.021607E-01
1 7.142582E+00 7.142212E+00 5.180827E-05 -7.427023E-02

2 7.032926E+00 7.032104E+00 1.168342E-04 -3.024553E-02
3 6.919731E+00 6.918852E+00 1.270169E-04 1.131867E-03
4 6.818840E+00 6.818259E+00 8.511129E-05 2.301096E-03
5 6.683523E+00 6.683002E+00 7.805772E-05 1.628654E-04
6 6.473311E+00 6.472531E+00 1.205256E-04 -3.052860E-03
7 6.266241E+00 6.264360E+00 3.002897E-04 7.978200E-04
8 5.998459E+00 5.996565E+00 3.158470E-04 2.060988E-03
9 5.872966E+00 5.874063E+00 -1.866253E-04 1.194506E-02
10 5.871826E+00 5.872206E+00 -6.471830E-05 8.337324E-04
11 5.865448E+00 5.865448E+00 -1.625919E-07 2.850365E-03
12 5.862948E+00 5.862957E+00 -1.545279E-06 2.839403E-03


----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | B1 | 3.0000E+00 : 4.5000E+00 : 6.7500E+00 : 1.0125E+01 : 1.5188E+01 : 2.2781E+01 :
2 | 2 | B2 | 3.0000E+00 : 4.5000E+00 : 6.7500E+00 : 1.0125E+01 : 1.5187E+01 : 2.2781E+01 :
3 | 3 | B3 | 3.0000E+00 : 4.5000E+00 : 6.7500E+00 : 1.0125E+01 : 1.5187E+01 : 2.2781E+01 :
4 | 4 | B4 | 3.0000E+00 : 4.5000E+00 : 6.7500E+00 : 1.0125E+01 : 6.0667E+00 : 3.0333E+00 :
5 | 5 | B5 | 3.0000E+00 : 4.5000E+00 : 6.7500E+00 : 4.2176E+00 : 2.1088E+00 : 1.0544E+00 :
6 | 6 | B6 | 3.0000E+00 : 4.5000E+00 : 6.7500E+00 : 1.0125E+01 : 1.5187E+01 : 1.4842E+01 :
7 | 7 | B7 | 3.0000E+00 : 4.5000E+00 : 6.7500E+00 : 1.0125E+01 : 1.5177E+01 : 2.2747E+01 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 8 : 9 : 10 : 11 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | B1 | 3.4172E+01 : 5.1258E+01 : 7.6887E+01 : 8.7081E+01 : 7.3393E+01 : 7.1651E+01 :
2 | 2 | B2 | 3.4172E+01 : 4.0456E+01 : 4.3075E+01 : 6.4613E+01 : 7.8186E+01 : 8.1804E+01 :
3 | 3 | B3 | 3.4172E+01 : 3.4356E+01 : 3.4571E+01 : 1.8982E+01 : 9.4912E+00 : 4.7456E+00 :
4 | 4 | B4 | 1.5167E+00 : 1.5381E+00 : 2.2850E+00 : 3.4275E+00 : 4.8840E+00 : 7.1672E+00 :
5 | 5 | B5 | 5.2720E-01 : 5.6416E-01 : 8.4625E-01 : 1.2579E+00 : 1.8400E+00 : 2.7401E+00 :
6 | 6 | B6 | 1.0364E+01 : 1.5546E+01 : 2.3319E+01 : 2.2530E+01 : 2.1741E+01 : 2.0927E+01 :
7 | 7 | B7 | 3.4120E+01 : 5.1180E+01 : 7.6770E+01 : 9.3267E+01 : 9.3646E+01 : 9.3927E+01 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 12 : 13 : 14 : 15 : 16 : 17 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | B1 | 7.2666E+01 :
2 | 2 | B2 | 8.0998E+01 :
3 | 3 | B3 | 2.3728E+00 :
4 | 4 | B4 | 1.0750E+01 :
5 | 5 | B5 | 3.7104E+00 :
6 | 6 | B6 | 1.8397E+01 :
7 | 7 | B7 | 9.5580E+01 :

Objective Function
Design Variables


WORKSHOP 7
MULTI-DISCIPLINARY OPTIMIZATION
(STATIC, NORMAL MODES, AND FREQUENCY
RESPONSE)

Model Description
X Multi-disciplinary analyses:
X Static loads of S1 = 30 and S2 = 40 pounds at Grid Points 1 and 2, respectively.
X Normal modes analysis.
X F1 = F2 = 10 pounds harmonic load at Grid Points 1 and 2.
X Objective Function – Minimize the SRSS response
( y f21 + y f22 + y f22 + .....)
at Grid Points 2 from 4 hz to 30 hz.
X Design Variables: cross-sectional area A1, A2, and A3
X Constraints:
Static subcase: -20,000 < Elem stresses < +20,000

Modal subcase: 17 hz < Freq1 < 20 hz

point. DAREA 201 1 1 10. 2 1 10.
$ FREQ1 100 4. .25 30
$ wkshp7.dat GRID 10 0. 0. 0. 123456
$ GRID 1 1. 0. 0. 23456
TIME 5 GRID 2 2. 0. 0. 23456
SOL 200 $ GRID 20 3. 0. 0. 123456
CEND PROD 20 200 1.E-3
TITLE=TWO MASS OPTIMIZATION PROBLEM PROD 21 200 1.5E-3
DISPL = ALL PROD 22 200 2.E-3
$ MAT1,200,1.0E7,,0.3,1.0 $ DUMMY DENSITY
SUBCASE 1 RLOAD1 200 201 210
LABEL = STATIC TABLED1 210
ANALYSIS = .... 1. 1. 10. 1. ENDT
LOAD = 101 param,g,.10
FORCE = ALL $
STRESS = ALL $ design model
DESSUB = .... $
$ DOPTPRM P1 1 P2 15 DESMAX 10
SUBCASE 2 .
LABEL = MODES .
ANALYSIS = .... .
METHOD = 201 ENDDATA
DESSUB = ....
$
SUBCASE 3
LABEL = DIRECT FREQUENCY RESPONSE
ANALYSIS = ....
DISP(PHASE)=ALL
FREQ = 100
DLOAD = 200
DESOBJ(MIN) = ....
$
BEGIN BULK
$
param,post,-1
$
FORCE 101 1 30. 1.
FORCE 101 2 40. 1.
EIGRL 201 2
$
CROD 20 20 10 1
CROD 21 21 1 2
CROD 22 22 2 20
CONM2 10 1 1.
CONM2 11 2 2.

$ soln7.dat
GRID 10 0. 0. 0. 123456
$
TIME 5 GRID 1 1. 0. 0. 23456
GRID 2 2. 0. 0. 23456
SOL 200 $
GRID 20 3. 0. 0. 123456
CEND
PROD 20 200 1.E-3
TITLE=TWO MASS OPTIMIZATION PROBLEM
PROD 21 200 1.5E-3
DISPL = ALL
$ PROD 22 200 1.5E-3
MAT1,200,1.0E7,,0.3,1.0 $ DUMMY DENSITY
SUBCASE 1
LABEL = STATIC RLOAD1 200 201 210
TABLED1 210
ANALYSIS = STATICS
1. 1. 10. 1. ENDT
LOAD = 101
FORCE = ALL param,g,.10
$
STRESS = ALL
DESSUB = 100 $ design model
$
$
DOPTPRM P1 1 P2 15 DESMAX 10
SUBCASE 2
DESVAR 1 A1 1.-3 1.-4 1.-2
LABEL = MODES
DESVAR 2 A2 1.5-3 1.5-4 1.5-2
ANALYSIS = MODES
METHOD = 201 DESVAR 3 A3 2.-3 2.0-4 2.0-2
DVPREL1 1 PROD 20 4 1.0-3
DESSUB = 200
$ 1 1.0
DVPREL1 2 PROD 21 4 1.5-3
SUBCASE 3
2 1.0
LABEL = DIRECT FREQUENCY RESPONSE
ANALYSIS = DFREQ DVPREL1 3 PROD 22 4 2.0-3
3 1.0
DISP(PHASE)=ALL
FREQ = 100 DCONSTR 100 101 -20000. 20000.
DCONSTR 100 102 -20000. 20000.
DLOAD = 200
DCONSTR 100 103 -20000. 20000.
DESOBJ(MIN) = 1000
DCONSTR 200 200 17. 20.
$
$
BEGIN BULK
$ DRESP1 101 STRELM1 STRESS PROD 2
param,post,-1 20
$ DRESP1 102 STRELM2 STRESS PROD 2
FORCE 101 1 30. 1. 21
FORCE 101 2 40. 1. DRESP1 103 STRELM3 STRESS PROD 2
EIGRL 201 2 22
$ DRESP1 200 MODE1 FREQ 1
CROD 20 20 10 1 $
CROD 21 21 1 2 dresp2,1000,srssg2c1,2000
CROD 22 22 2 20 ,dresp1,1001,1002,1003,1004,1005,1006,1007
CONM2 10 1 1. ,,*7,*7,*7,*7,*7,*7,*7
CONM2 11 2 2. =13
DAREA 201 1 1 10. 2 1 10. $
FREQ1 100 4. .25 104

DRESP1,1001,GRID2C1,FRDISP,,,1,4.0,2
=,*1,=,=,=,=,=,*(.25),=
=103
$
deqatn 2000 srssg2c1(f1,f2,f3,f4,f5,f6,f7,f8,f9,f10,f11,f12,f13,f14,
f15,f16,f17,f18,f19,f20,f21,f22,f23,f24,
f25,f26,f27,f28,f29,f30,f31,f32,f33,f34,
f35,f36,f37,f38,f39,f40,f41,f42,f43,f44,
f45,f46,f47,f48,f49,f50,f51,f52,f53,f54,
f55,f56,f57,f58,f59,f60,f61,f62,f63,f64,
f65,f66,f67,f68,f69,f70,f71,f72,f73,f74,
f75,f76,f77,f78,f79,f80,f81,f82,f83,f84,
f85,f86,f87,f88,f89,f90,f91,f92,f93,f94,
f95,f96,f97,f98,f99,f100,
f101,f102,f103,f104,f105) =
sqrt(f1**2+f2**2+f3**2+f4**2+f5**2+f6**2+f7**2+
f8**2+f9**2+f10**2+f11**2+f12**2+f13**2+f14**2+
f15**2+f16**2+f17**2+f18**2+f19**2+f20**2+
f21**2+f22**2+f23**2+f24**2+f25**2+f26**2+
f27**2+f28**2+f29**2+f30**2+f31**2+f32**2+
f33**2+f34**2+f35**2+f36**2+f37**2+f38**2+
f39**2+f40**2+f41**2+f42**2+f43**2+f44**2+
f45**2+f46**2+f47**2+f48**2+f49**2+f50**2+
f51**2+f52**2+f53**2+f54**2+f55**2+f56**2+
f57**2+f58**2+f59**2+f60**2+f61**2+f62**2+
f63**2+f64**2+f65**2+f66**2+f67**2+f68**2+
f69**2+f70**2+f71**2+f72**2+f73**2+f74**2+
f75**2+f76**2+f77**2+f78**2+f79**2+f80**2+
f81**2+f82**2+f83**2+f84**2+f85**2+f86**2+
f87**2+f88**2+f89**2+f90**2+f91**2+f92**2+
f93**2+f94**2+f95**2+f96**2+f97**2+f98**2+
f99**2+f100**2+
f101**2+f102**2+f103**2+f104**2+f105**2)
ENDDATA



4. View Results
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 2.092793E-02 2.692307E-01
1 2.204837E-02 1.858932E-02 1.860773E-01 -1.468428E-02
2 1.792609E-02 1.702995E-02 5.262158E-02 -2.815078E-02
3 1.614813E-02 1.526256E-02 5.802234E-02 -4.623711E-02
4 1.418893E-02 1.335493E-02 6.244885E-02 -5.594336E-03
5 1.263804E-02 1.243193E-02 1.657887E-02 1.052284E-03
6 1.243193E-02 1.243193E-02 0.000000E+00 1.052284E-03


-------------------------------------------------------------------------------------------------------------------------------
-------------------------------------------------------------------------------------------------------------------------------
1 | 1 | A1 | 1.0000E-03 : 1.4725E-03 : 1.2645E-03 : 9.9907E-04 : 1.0000E-03 : 1.0000E-03
2 | 2 | A2 | 1.5000E-03 : 2.0362E-03 : 2.4325E-03 : 2.7333E-03 : 1.4975E-03 : 1.4975E-03
3 | 3 | A3 | 2.0000E-03 : 2.1319E-03 : 2.6395E-03 : 3.3800E-03 : 3.8150E-03 : 4.1279E-03
-------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 8 : 9 : 10 : 11
-------------------------------------------------------------------------------------------------------------------------------
1 | 1 | A1 | 1.0000E-03 :
2 | 2 | A2 | 1.4975E-03 :
3 | 3 | A3 | 4.1279E-03 :

Design Variables
Objective Function


WORKSHOP 8
DESIGN STRUCTURE WITH STRENGTH AND

BUCKLING LIMITS SATISFYING BOTH THE
EULER’S AND JOHNSON CRITIRIA

Objectives
X Design a structure with strength and buckling limits
X Impose three buckling conditions
1. The MSC.NASTRAN buckling analysys
2. Euler Buckling Equation limit
− π 2E
σ ≥σ Eul
=
( L' / ρ )2
cr
3. Johnson Short Columns Design Formula
⎡ σ co ( L' / ρ )2 ⎤
σ ≥ σ crJohn = −σ co ⎢1.0 − ⎥
⎣ 4 π 2
E ⎦

Objectives (Cont.)
X Use the beam library in a design task:
For the pinned-pinned condition and a rod cross section type:

L' = L
ρ=
r ( ρ = radius of gyration)
2
Assume σ co = +80 ,000. ( σ co is an empirically determined column yield

stress)

Model Description:
X Subject to:
X Static load of –200,000 at x-direction of grid point 6
X Pins at points 1 and 6
X Buckling analysis
X Objective Functions: Minimize the Weight
X Design variable: Radius of rod
X Constraints:
Static Case: − 80 ,000 ≤ σ ≤ 100 ,000
σ ≥ σ crEUL ,
σ
* σ ≥ σ crJohn when −σ* > co
2
Buckling Subcase: λ ≥ 1 .0
* The Johnson Condition is the critical design condition only if the magnitude at the compressive stress is
greater than σ co / 2 . This can be implemented in MSC.Nastran by multiplying the response by a factor
that is a the maximum of 0.0 and ( −σ − σ co / 2 ). When the maximum is 0.0, the Johnson condition is
ignored.
point.
ID MSC, wkshp8a $ ELEMENT AND MATERIAL DATA
TIME 130 $ cbar 1 10 1 2
SOL 200 $ OPTIMIZATION 10
CEND cbar 2 10 2 3
TITLE = buckling test case 10
SUBTITLE = johnson/euler buckling case cbar 3 10 3 4
ECHO = SORT 10
spc = 100
cbar 4 10 4 5
DESOBJ = 20
10
SUBCASE 1
cbar 5 10 5 6
DESSUB = 1
10
LOAD = 300
MAT1 1 3.0e7
ANALYSIS = STATICS 0.33 0.1
DISP = ALL $ PROPERTY DATA
STRESS = ALL PbarL 10 1 mscbml0 rod
SUBCASE 2 1.0
DESSUB = 2 $ BOUNDARY CONDITION DATA
ANALYSIS = BUCK spc1 100 1 1
METHOD = 1
spc1 100 2 1 6
DISP = ALL
LABEL = BUCKLING FACTORS $ EXTERNAL LOADS DATA
$ FORCE 300 6
BEGIN BULK -2.0e5 1.0
param,post,-1 $ BUCKLING ANALYSIS DATA
$---------------------------------------------------------------------- EIGRL 1 .05
$ ANALYSIS MODEL 4
$----------------------------------------------------------------------
$
$
$ GRID DATA $-----------------------------------------------------------
$ 2 3 4 5 6 7 8 9 10 ---------
GRDSET 345 $ DESIGN MODEL
grid 1 0.0 0.0 0.0 $-----------------------------------------------------------
grid 2 8.0 0.0 ---------$
0.0 $ Beginning of Design Modeling Exercise for the Bulk Data
grid 3 16.0 0.0
0.0
Section
grid 4 24.0 0.0 $
0.0 .
grid 5 32.0 0.0 .
0.0 .
grid 6 40.0 0.0 $
0.0
ENDDATA
grid 10 0.0 0.0
100.0 123456
$

ID MSC, soln8a
TIME 130 $
SOL 200 $ OPTIMIZATION
CEND
TITLE = buckling test case
SUBTITLE = johnson/euler buckling case
ECHO = SORT
spc = 100
DESOBJ = 20
SUBCASE 1
DESSUB = 1
LOAD = 300
ANALYSIS = STATICS
DISP = ALL
STRESS = ALL
SUBCASE 2
DESSUB = 2
ANALYSIS = BUCK
METHOD = 1
DISP = ALL
LABEL = BUCKLING FACTORS
BEGIN BULK
param,post,-1
$------------------------------------------------------------------------
$ ANALYSIS MODEL
$------------------------------------------------------------------------
$ GRID DATA
$ 2 3 4 5 6 7 8 9 10
GRDSET 345
grid 1 0.0 0.0 0.0
grid 2 8.0 0.0 0.0
grid 3 16.0 0.0 0.0
grid 4 24.0 0.0 0.0
grid 5 32.0 0.0 0.0
grid 6 40.0 0.0 0.0
grid 10 0.0 0.0 100.0
123456
$ ELEMENT AND MATERIAL DATA
cbar 1 10 1 2 10
cbar 2 10 2 3 10
cbar 3 10 3 4 10
cbar 4 10 4 5 10
cbar 5 10 5 6 10
MAT1 1 3.0e7 0.33 0.1

$ PROPERTY DATA
PbarL 10 1 mscbml0 rod
1.0
$ BOUNDARY CONDITION DATA DVPREL1 10 pbarl
spc1 100 1 1 10 12
spc1 100 2 1 6 1 1.0
$ EXTERNAL LOADS DATA $
FORCE 300 6 -2.0e5 1.0 $ DESIGN FOR BUCKLING EIGENVALUE
$ BUCKLING ANALYSIS DATA DRESP1 1 BUCK1 LAMA
EIGRL 1 .05 4 1
$ dconstr 2 1
$------------------------------------------------------------------------ 1.0
$ DESIGN MODEL param dsnokd 1.0
$------------------------------------------------------------------------ $
$ ENDDATA
DESVAR 1 rg 1.0 0.01 10.0
DOPTPRM p2 15 DESMAX 20 DELP 0.5 GMAX 0.01

CONVDV 0.01 CONVPR 0.02 p1 1
DRESP1 20 W WEIGHT
DRESP1 23 S1 STRESS Pbar 7 10
DRESP1 24 S1 STRESS Pbar 8 10
DRESP1 25 S1 sTRESS Pbar 6 10
dresp2 31 EUL 31
DESVAR 1
DTABLE L E
DRESP1 25
$ EULER CONSTRAINT CONDITION
deqatn 31 rgyra(r,l,e,sigma) = r / 2.0;
euler = - sigma * ( L / rGYRA ) **2 / (PI(1)**2 * E)
dresp2 32 JOHNSON 32
DESVAR 1
DTABLE L E SIGMAC
DRESP1 25
$
$ JOHNSON CONSTRAINT CONDITION THAT IS ONLY ACTIVE WHEN FAC IS NONZERO
$
deqatn 32 LORHO2(r,l,e,SIGMAC,sigma) = (2.0 * L / R ) ** 2;
fac = max((-sigma - sigmac/2.0), 0.0);
JOHNSON = - FAC * sigma / ((-sigma - sigmac/2.0) *
SIGMAc * ( 1.0 - SIGMAC * LORHO2/
(4.0 * PI(1)**2 * E) ) )
DCONSTR 1 23 100000.
dconstr 1 24 -80000.
dconstr 1 31 1.0
dconstr 1 32 1.0
dtable e 30.0e6 l 40.0 sigmac
8.0e4



4. View Results
***************************************************************
***************************************************************


--------------------------------------------------------------------------------------------------------------
CYCLE OBJECTIVE FROM OBJECTIVE FROM FRACTIONAL ERROR MAXIMUM VALUE
NUMBER APPROXIMATE EXACT OF OF
OPTIMIZATION ANALYSIS APPROXIMATION CONSTRAINT
--------------------------------------------------------------------------------------------------------------
INITIAL 1.256637E+01 4.017613E-01
1 1.539844E+01 1.539775E+01 4.471775E-05 3.485799E-03
2 1.540468E+01 1.540468E+01 1.238162E-07 2.788901E-03

---------------------------------------------------------------------------------------------------------------
1 BUCKLING TEST CASE FEBRUARY 17, 2003 MSC.NASTRAN 4/ 9/01 PAGE 72
JOHNSON/EULER BUCKLING CASE
0 SUBCASE 2
-------------------------------------------------------------------------------------------------------------------------
-------------------------------------------------------------------------------------------------------------------------
1 | 1 | RG | 1.0000E+00 : 1.1069E+00 : 1.1072E+00 :

Use the output produced by DOPTPRM parameter p2 = 15 to determine the critical design case in
the .f06 file, search for FINAL. Then search again for DESIGN CONSTRAINTS. This will lead to a
table like:
----- DESIGN CONSTRAINTS ON RESPONSES -----
(MAXIMUM RESPONSE CONSTRAINTS MARKED WITH **)
---------------------------------------------------------------------------------------------------------
INTERNAL INTERNAL
INTERNAL DCONSTR RESPONSE RESPONSE L/U REGION SUBCASE
ID ID ID TYPE FLAG ID ID VALUE
---------------------------------------------------------------------------------------------------------
1 1 2 STRESS LOWER 1 1 -2.0423E-01
2 1 3 STRESS LOWER 1 1 -2.0423E-01
3 1 4 STRESS LOWER 1 1 -2.0423E-01
4 1 5 STRESS LOWER 1 1 -2.0423E-01
5 1 6 STRESS LOWER 1 1 -2.0423E-01
6 1 1 EQUA UPPER 31 1 3.7607E-01
7 1 2 EQUA UPPER 31 1 3.7607E-01
8 1 3 EQUA UPPER 31 1 3.7607E-01
9 1 4 EQUA UPPER 31 1 3.7607E-01
10 1 5 EQUA UPPER 31 1 3.7607E-01
11 1 6 EQUA UPPER 32 1 4.0176E-01**
12 1 7 EQUA UPPER 32 1 4.0176E-01**
13 1 8 EQUA UPPER 32 1 4.0176E-01**
14 1 9 EQUA UPPER 32 1 4.0176E-01**
15 1 10 EQUA UPPER 32 1 4.0176E-01**
16 2 12 LAMA LOWER 0 2 2.7638E-01
----- CONSTRAINTS ON DESIGNED PROPERTIES -----

This indicates that five constraints are tied at the maximum value and that they are associated with
an EQUA response type and have internal ID’s 6 thru 10.
Search for RETAINED DRESP2. This leads to a table like:
---- RETAINED DRESP2 RESPONSES ----
----------------------------------------------------------------------------------------------------------
INTERNAL DRESP2 RESPONSE EQUATION LOWER UPPER
ID ID LABEL ID BOUND VALUE BOUND
----------------------------------------------------------------------------------------------------------
1 31 EUL 31 N/A 1.3761E+00 1.0000E+00
2 31 EUL 31 N/A 1.3761E+00 1.0000E+00
3 31 EUL 31 N/A 1.3761E+00 1.0000E+00
4 31 EUL 31 N/A 1.3761E+00 1.0000E+00
5 31 EUL 31 N/A 1.3761E+00 1.0000E+00
6 32 JOHNSON 32 N/A 1.4018E+00 1.0000E+00
7 32 JOHNSON 32 N/A 1.4018E+00 1.0000E+00
8 32 JOHNSON 32 N/A 1.4018E+00 1.0000E+00
9 32 JOHNSON 32 N/A 1.4018E+00 1.0000E+00
10 32 JOHNSON 32 N/A 1.4018E+00 1.0000E+00
It is seen that the maximum constraints are associated with the Johnson condition.

5. Experiment With Different Length
Repeat the workshop problem with L = 60. How do the answers change and what is the critical
design condition?

WORKSHOP 9
RANDOM RESPONSE OPTIMIZATION

Objective:
Be familiar with the process to set up an random response
optimization task using a simulated car model.

Analysis Model Description
• Car body – modeled by two CBAR elements and three CONM2s
• Front and back wheels - modeled by two CONM2s
• The suspension components that connect car body and two wheels are
modeled by CBUSH elements
• Each wheel is loaded with a unit enforced displacement. Total two load
cases are applied.
• First two RANDPS entries represent the auto spectrum applied at the front
and back wheels. The third and fourth RANDPS entries represent the real
and imaginary parts of the cross spectrum between the front and back
wheels.

• Objective function:
Minimize the RMS displacement at CG of the vehicle in the frequency range
of 0.1Hz to 2.0Hz. The CG is located at component 2 of grid point 5. The
RANDPS entry is directly referenced on the ATTB field of the DRESP1
entry.
• Design Variables:
The shock spring and absorber of the suspension system
100 < K < 1000
1.0 < B < 10.
• Constraints:
The stresses at the supporting frame that connects grid points 1 and 5, and
5 and 2 must be less than 4500 across the whole frequency range.

starting point:
$ GRID 1 0. 0. 0.
$ wkshp9.dat - random input at wheels - enforced motion GRID 2 120. 0. 0.
$ direct method GRID 5 60. 0. 0.
$ stress constraints at discrete freq $ wheels
$ minimize acce at grid pt 5 $
$ cross spectrum GRID 3 0. -10. 0.
$ using new enforced motion GRID 4 120. -10. 0.
$ $
ID MSC, d200rd15.dat $ V2001 JML 8-Sep-2000 CBAR 5 11 1 5 0. 1. 0.
SOLa.200Generate
$ jml an input file using CBAR 6 11 5 2 0. 1. 0.
CEND $
TITLE =the dataCAR
SIMPLE from page
WITH 2-3.
RANDOM Use
INPUT PBAR 11 12 1.e2 1.e3 1.e3
100. 0.
the
SPC = 100 following
FREQUENCY = 130
input file as a -100. 0.
starting= point.
stress(phase) all MAT1 12 3.E+7 7.8e-4
disp(phase) = all $ CONSTRAIN TO A PLANAR PROBLEM
$ $
$ we perform direct frequency analysis SPC1 100 1345 1 2 5
analysis = ... SPC1 100 13456 3 4
$ spc1,100,2,3,4
$ calls out a RMS displacement response for minimization $
desobj= ... CONM2 10 1 2.5
$ CONM2 15 2 2.5
$ define a design case CONM2 20 5 5.0
dessub = ... CONM2 25 3 2.5
$ CONM2 26 4 2.5
$ this random case control callout is needed only CBUSH 100 1000 1 3
$ for the xyplot request when using optimization 0
$ CBUSH 200 1000 2 4
random = 1000 0
SUBCASE 1 PBUSH 1000 K 200.
DLOAD = 111 B 2.
SUBCASE 2 RLOAD2 111 222 444
DLOAD = 112 disp
$ spcd,222,3,2,1.
output (xyplot) TABLED1 444
xtitle = frequency (hz) 0. 1. 100. 1. ENDT
ytitle = disp psd at grid pt 5 $
xypunch disp psdf /5(t2) RLOAD2 112 223 444
$ disp
BEGIN BULK spcd,223,4,2,1.
$ $
$ car body FREQ2 130 0.1 2. 20
$

1. Generating an Input File for MSC.Nastran Users (Cont.)
FREQ2 130 0.1 2. 20

$ $---------------------------------------------------------
$ DEFINE THE INPUT PSD $ design model
$ $ ------------------------------------------------------$
RANDPS 1000 1 1 1. 0. 145 $ define sprint and absober as design variables
RANDPS 1000 2 2 1. 0. 145 $
RANDPS 1000 1 2 1. 0. 146 desvar,1...
RANDPS 1000 1 2 0. 1. 147 desvar,2...
TABRND1 145 $
.1 .1 5. 1. 10. .05 ENDT dvprel1,...
$
TABRND1 146 dvprel1,...
0.1000 0.0990 0.1096 0.1005 0.1202 0.1022 0.1318 0.1040
0.1445 0.1059 0.1585 0.1079 0.1738 0.1101 0.1905 0.1124 $
0.2089 0.1147 0.2291 0.1172 0.2512 0.1197 0.2754 0.1221 $ define RMS displacement response at the CG of the
0.3020 0.1246 0.3311 0.1269 0.3631 0.1289 0.3981 0.1305 $ vehicle as an objective
0.4365 0.1314 0.4786 0.1315 0.5248 0.1304 0.5754 0.1277 $
0.6310 0.1228 0.6918 0.1151 0.7586 0.1038 0.8318 0.0879 dresp1,...
0.9120 0.0665 1.0000 0.0382 1.0965 0.0020 1.2023 -0.0434 $
1.3183 -0.0986 1.4454 -0.1636 1.5849 -0.2372 1.7378 -0.3159 $ define dynamic stress responses at end B-points
1.9055 -0.3934 2.0893 -0.4593 2.2909 -0.4984 2.5119 -0.4910 $ C and E for pbar 11.
2.7542 -0.4149 3.0200 -0.2508 3.3113 0.0072 3.6308 0.3364 $
3.9811 0.6685 4.3652 0.8819 4.7863 0.8223 5.2481 0.3435 dresp1,...
5.7544 0.2965 6.3096 0.6842 6.9183 0.5743 7.5858 0.0891 dresp1,...
8.3176 0.2820 9.1201 0.1964 10.0000 0.0063 ENDT $
$ $ stress upper limit = 45000 across the freq. range
TABRND1 147 $
0.1000 0.0142 0.1096 0.0159 0.1202 0.0177 0.1318 0.0198 dconstr,...
0.1445 0.0221 0.1585 0.0248 0.1738 0.0279 0.1905 0.0313 dconstr,...
0.2089 0.0352 0.2291 0.0397 0.2512 0.0448 0.2754 0.0506 $
$ optimization parameter control. Notice SQP (method=3)
0.3020 0.0572 0.3311 0.0648 0.3631 0.0734 0.3981 0.0832 $ is selected for this job and delp=1.0 is used.
0.4365 0.0944 0.4786 0.1070 0.5248 0.1212 0.5754 0.1371 $
0.6310 0.1547 0.6918 0.1741 0.7586 0.1951 0.8318 0.2173 doptprm,desmax,20,p1,1,p2,15,iprint,7
0.9120 0.2401 1.0000 0.2625 1.0965 0.2830 1.2023 0.2993 ,method,3,delp,1.0
1.3183 0.3084 1.4454 0.3061 1.5849 0.2875 1.7378 0.2467 $
1.9055 0.1775 2.0893 0.0749 2.2909 -0.0631 2.5119 -0.2318 ENDDATA
2.7542 -0.4160 3.0200 -0.5848 3.3113 -0.6898 3.6308 -0.6687
3.9811 -0.4624 4.3652 -0.0505 4.7863 0.4968 5.2481 0.8888
5.7544 0.8037 6.3096 0.3101 6.9183 -0.2722 7.5858 -0.5008
8.3176 -0.2390 9.1201 0.0928 10.0000 0.0496 ENDT
$
$ -----------------------------------------------------------

$
$ d200rd15.dat - random input at wheels - CBAR 5 11 1 5 0. 1. 0.
enforced motion CBAR 6 11 5 2 0. 1. 0.
$ direct method PBAR 11 12 1.e2 1.e3 1.e3
$ stress constraints at discrete 100. 0. -100. 0.
freq MAT1 12 3.E+7 7.8e-4
$ minimize acce at grid pt 5 $
$ cross spectrum $ CONSTRAIN TO A PLANAR PROBLEM
$ using new enforced motion $
$ SPC1 100 1345 1 2 5
ID MSC, d200rd15.dat $ V2001 JML 8-Sep-2000 SPC1 100 13456 3 4
SOL 200 $ jml spc1,100,2,3,4
CEND $
TITLE = SIMPLE CAR WITH RANDOM INPUT CONM2 10 1 2.5
SPC = 100 CONM2 15 2 2.5
FREQUENCY = 130 CONM2 20 5 5.0
stress(phase) = all CONM2 25 3 2.5
disp(phase) = all CONM2 26 4 2.5
desobj= 101 $
analysis = dfreq CBUSH 100 1000 1 3 0
dessub = 800 CBUSH 200 1000 2 4 0
$ PBUSH 1000 K 200.
$ this random case control callout is needed only B 2.
$ for the xyplot request when using optimization $
$ RLOAD2 111 222 444 disp
random = 1000
SUBCASE 1 spcd,222,3,2,1.
DLOAD = 111 TABLED1 444
SUBCASE 2 0. 1. 100. 1. ENDT
DLOAD = 112 $
$ RLOAD2 112 223 444 disp
output (xyplot) spcd,223,4,2,1.
xtitle = frequency (hz) $
ytitle = disp psd at grid pt 5 FREQ2 130 0.1 2. 20
xypunch disp psdf /5(t2) $
$ $ DEFINE THE INPUT PSD
BEGIN BULK $
$ RANDPS 1000 1 1 1. 0. 145
$ car body RANDPS 1000 2 2 1. 0. 145
$ RANDPS 1000 1 2 1. 0. 146
GRID 1 0. 0. 0. RANDPS 1000 1 2 0. 1. 147
GRID 2 120. 0. 0. TABRND1 145
GRID 5 60. 0. 0. .1 .1 5. 1. 10. .05 ENDT
$
$ wheels
$
GRID 3 0. -10. 0.
GRID 4 120. -10. 0.
$

$
TABRND1 146
0.1000 0.0990 0.1096 0.1005 0.1202 0.1022 0.1318 0.1040 $
0.1445 0.1059 0.1585 0.1079 0.1738 0.1101 0.1905 0.1124 $ design model
0.2089 0.1147 0.2291 0.1172 0.2512 0.1197 0.2754 0.1221 $
0.3020 0.1246 0.3311 0.1269 0.3631 0.1289 0.3981 0.1305 desvar,1,k2,1.,0.5,5.
0.4365 0.1314 0.4786 0.1315 0.5248 0.1304 0.5754 0.1277 desvar,2,b2,1.,0.5,5.
0.6310 0.1228 0.6918 0.1151 0.7586 0.1038 0.8318 0.0879 $
0.9120 0.0665 1.0000 0.0382 1.0965 0.0020 1.2023 -0.0434 dvprel1,101,pbush,1000,k2,100.
1.3183 -0.0986 1.4454 -0.1636 1.5849 -0.2372 1.7378 -0.3159 ,1,200.
1.9055 -0.3934 2.0893 -0.4593 2.2909 -0.4984 2.5119 -0.4910 dvprel1,201,pbush,1000,b2,1.
2.7542 -0.4149 3.0200 -0.2508 3.3113 0.0072 3.6308 0.3364 ,2,2.0
3.9811 0.6685 4.3652 0.8819 4.7863 0.8223 5.2481 0.3435 $
5.7544 0.2965 6.3096 0.6842 6.9183 0.5743 7.5858 0.0891 dresp1,101,ming5t2,rmsdisp,,,2,1000,5
8.3176 0.2820 9.1201 0.1964 10.0000 0.0063 ENDT $
$ dconstr,800,801,-1.e20,4500.
TABRND1 147 dconstr,800,802,-1.e20,4500.
0.1000 0.0142 0.1096 0.0159 0.1202 0.0177 0.1318 0.0198 dresp1,801,e5ptc,frstre,pbar,,12,,11
0.1445 0.0221 0.1585 0.0248 0.1738 0.0279 0.1905 0.0313 dresp1,802,e5pte,frstre,pbar,,14,,11
0.2089 0.0352 0.2291 0.0397 0.2512 0.0448 0.2754 0.0506 $
0.3020 0.0572 0.3311 0.0648 0.3631 0.0734 0.3981 0.0832 doptprm,desmax,20,p1,1,p2,15,iprint,7
0.4365 0.0944 0.4786 0.1070 0.5248 0.1212 0.5754 0.1371 ,method,3,delp,1.0
0.6310 0.1547 0.6918 0.1741 0.7586 0.1951 0.8318 0.2173 $
0.9120 0.2401 1.0000 0.2625 1.0965 0.2830 1.2023 0.2993 ENDDATA
1.3183 0.3084 1.4454 0.3061 1.5849 0.2875 1.7378 0.2467
1.9055 0.1775 2.0893 0.0749 2.2909 -0.0631 2.5119 -0.2318
2.7542 -0.4160 3.0200 -0.5848 3.3113 -0.6898 3.6308 -0.6687
3.9811 -0.4624 4.3652 -0.0505 4.7863 0.4968 5.2481 0.8888
5.7544 0.8037 6.3096 0.3101 6.9183 -0.2722 7.5858 -0.5008
8.3176 -0.2390 9.1201 0.0928 10.0000 0.0496 ENDT
$



4. View Results
When the run is completed, edit the wkshp9.f06 file and search for the word FATAL. If no matches exist,
search for the word WARNING. Determine whether existing WARNING messages indicate modeling
errors.
While still editing wkshp9.f06, search for the word “H I S T O R Y”. Compare the results obtained in the
.f06 file with the following:
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 1.905990E+00 -1.686030E-01
1 -1.274773E+00 1.441520E+00 -1.884326E+00 -2.473581E-01
2 7.713596E-01 1.709733E+00 -5.488422E-01 -9.062972E-02
3 -6.315433E+00 8.359604E-01 -8.554703E+00 N/A
4 5.182346E-01 6.239330E-01 -1.694066E-01 N/A
5 6.239330E-01 6.239330E-01 0.000000E+00 N/A


----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | K2 | 1.0000E+00 : 1.5000E+00 : 1.2489E+00 : 6.2443E-01 : 5.0000E-01 : 5.0000E-01 :
2 | 2 | B2 | 1.0000E+00 : 1.5000E+00 : 2.2500E+00 : 3.3750E+00 : 5.0000E+00 : 5.0000E+00 :
*** USER INFORMATION MESSAGE 6464 (DOM12E)
RUN TERMINATED DUE TO HARD CONVERGENCE TO AN OPTIMUM AT CYCLE NUMBER = 5.

Fig. 9.1 – Objective Function History Fig. 9.2 – Design Variable History

Fig. 9. 3: Comparison of PSDF DISP RESP Before and After Optimization

WORKSHOP 10
DISCRETE VARIABLE OPTIMIZATION
Subcase 1 Subcase 2
X: -16,000 lbs X: 16,000 lbs
Y: -12,000 lbs Y: -12,000 lbs

Objective:
Be familiar with the process to set up a discrete optimization
task using the three bar truss problem.

X Objective Function: Weight minimization
X Design Variables: Cross sectional areas A1, A2 and A3. A1 and A3 are
linked as A1=A3.
Continuum bound: .1 < A1 < 100.

.1 < A2 < 100.
.1 < A3 < 100.
All three design variable share the same discrete value set:
0.1 to 100. Incremented by 0.05
X Constraints:
Stress allowable: 20ksi tension and 15 ksi compression
Displacement at grid point 4:
|X direction| <= 0.2 in
|Y direction| <= 0.2 in

Model Description
X Part A:
Perform discrete variable processing after last design cycle
X Part B:
Perform discrete variable process with the round-up option after each
design cycle

1. Generate an Input File for Part A
2. Compare the Input File of Part A
3. Submit Input File to MSC.NASTRAN for analysis
4. View Results
5. Generate an Input File for Part B
6. Submit Input File to MSC.NASTRAN for analysis
7. View Results

1. Generating an Input File for Part A
Q A continuum optimization task has been prepared for you. The task is to define DESVAR
and DDVAL entries using the data from previous page and to activate the discrete variable
process. The portion of the file is shown here as a starting point.
C Executive and Case Control Sections are omitted
BEGIN BULK
$------------------------------------------------------------------------
$ Analysis model is omitted
$------------------------------------------------------------------------
$----------------------------------------------------------------------……….
$ DESIGN MODEL
$----------------------------------------------------------------------$
$...DESIGN VARIABLE DEFINITION
$...IMPOSE X3=X1 (LEADS TO A3=A1)
$DLINK, ID, DDVID, CO, CMULT, IDV1, C1, IDV2, C2, +
$+, IDV3, C3, ...
DLINK, 1, 3, 0.0, 1.0, 1, 1.00
$
$DESVAR,ID, LABEL, XINIT, XLB, XUB, DELXV(OPTIONAL), DDVAL
$
$ specify the ID of the discrete value list
$
DESVAR…
DESVAR…
DESVAR…
$
$DDVAL,ID, DVAL1, THRU, DVAL2, BY INC
$
$ define the discrete value set
$ Set 1: 0.1 to 100.0 incremented by 0.05
$
DDVAL...
$
$ constraint definitions are omitted.
$
$
$...OPTIMIZATION CONTROL:
$
$ Use Round-up for discrete processing after last design
cycle.
$
DOPTPRM,IPRINT, 5, DESMAX, 20, DELP, 0.5, P1,
1, +
+, P2, 15
$
ENDDATA

The completed MSC.NASTRAN input file for Part A is shown below:
$ Modified 15-sep-2000 xmy v2001
TIME 10 $
SOL 200 $ OPTIMIZATION
include 'dmap'
CEND
TITLE = SYMMETRIC THREE BAR DISCRETE DESIGN OPTIMIZATION
SUBTITLE = ROUNDED-UP APROACH (DISCOD = 3)
ECHO = SORT
SPC = 100
DISP = ALL
STRESS = ALL
DESOBJ(MIN) = 20 $ (DESIGN OBJECTIVE = DRESP ID)
DESSUB = 21 $ DEFINE CONSTRAINT SET FOR BOTH SUBCASES
ANALYSIS = STATICS
SUBCASE 1
LOAD = 300
SUBCASE 2
LOAD = 310
BEGIN BULK
$
$------------------------------------------------------------------------
$ ANALYSIS MODEL
$------------------------------------------------------------------------
$
$ GRID DATA
$ 2 3 4 5 6 7 8 9 10
GRID, 1, , -10.0 , 0.0, 0.0
GRID, 2, , 0.0 , 0.0, 0.0
GRID, 3, , 10.0 , 0.0, 0.0
GRID, 4, , 0.0 , -10.0, 0.0
$ SUPPORT DATA
SPC, 100, 1, 123456, , 2, 123456
SPC, 100, 3, 123456, , 4, 3456
$ ELEMENT DATA
CROD, 1, 11, 1, 4
CROD, 2, 12, 2, 4
CROD, 3, 13, 3, 4
$ PROPERTY DATA
PROD, 11, 1, 1.0
PROD, 12, 1, 2.0
PROD, 13, 1, 1.0
MAT1, 1, 1.0E+7, , 0.33, 0.1
$ EXTERNAL LOADS DATA
FORCE, 300, 4, , 20000., 0.8, -0.6
FORCE, 310, 4, , 20000., -0.8, -0.6
$
$------------------------------------------------------------------------
$ DESIGN MODEL
$------------------------------------------------------------------------
$...IMPOSE X3=X1 (LEADS TO A3=A1)

$DLINK, ID, DDVID, CO, CMULT, IDV1, C1, IDV2, C2, +
$+, IDV3, C3, ...
DLINK, 1, 3, 0.0, 1.0, 1, 1.00
$...DESIGN VARIABLE DEFINITION
$DESVAR,ID, LABEL, XINIT, XLB, XUB, DELXV(OPTIONAL), DDVAL
DESVAR, 1, A1, 1.0, 0.1, 100.0, , 1
DESVAR, 2, A2, 2.0, 0.1, 100.0, , 1
DESVAR, 3, A3, 1.0, 0.1, 100.0, , 1
$
$Define a discrete value set
$
$DDVAL,ID, DVAL1, THRU, DVAL2, BY INC
$ 2 3 4 5 6 7 8
DDVAL 1 0.1 THRU 100. BY 0.05
$...DEFINITION OF DESIGN VARIABLE TO ANALYSIS MODEL PARAMETER RELATIONS
$DVPREL1,ID, TYPE, PID, FID, PMIN, PMAX, C0, , +
$+, DVID1, COEF1, DVID2, COEF2, ...
DVPREL1,10, PROD, 11, 4, , , , , +DP1
+DP1, 1, 1.0
DVPREL1,20, PROD, 12, 4, , , , , +DP2
+DP2, 2, 1.0
DVPREL1,30, PROD, 13, 4, , , , , +DP3
+DP3, 3, 1.0
$
$...STRUCTURAL RESPONSE IDENTIFICATION
$DRESP1,ID, LABEL, RTYPE, PTYPE, REGION, ATTA, ATTB, ATT1, +
$+, ATT2, ...
DRESP1, 20, W , WEIGHT
DRESP1, 21, U4, DISP , , , 1, , 4
DRESP1, 22, V4, DISP , , , 2, , 4
DRESP1, 23, S1, STRESS, PROD, , 2, , 11
$
$...CONSTRAINTS
$DCONSTR,DCID, RID, LALLOW, UALLOW
DCONSTR,21, 21, -0.20 ,0.20
DCONSTR,21, 22, -0.20 ,0.20
DCONSTR,21, 23, -15000.,20000.
DCONSTR,21, 24, -15000.,20000.
DCONSTR,21, 25, -15000.,20000.
$
$...OPTIMIZATION CONTROL:
$
DOPTPRM,IPRINT, 5, DESMAX, 20, DELP, 0.5, P1, 1, +
+, P2, 15, DISCOD, 3
$
$.......2.......3.......4.......5.......6.......7.......8.......9.......0
param,post,-1
ENDDATA


wkshp10a scr=yes. Monitor the run using the UNIX ps command:

4. View Results
When the run is completed, edit the wkshp10a.f06 file and search for the word FATAL. If no matches
exist, search for the word WARNING. Determine whether existing WARNING messages indicate modeling
errors.
While still editing wkshp10a.f06, search for the word “H I S T O R Y”. Compare the results obtained in the
***************************************************************
***************************************************************
(SOFT FEASIBLE DISCRETE DESIGN OBTAINED)
(HARD FEASIBLE DISCRETE DESIGN OBTAINED)
NUMBER OF DISCRETE PROCESSING ANALYSES COMPLETED 1
---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 4.828427E+00 -3.234952E-01
1 3.007052E+00 3.007354E+00 -1.004459E-04 -3.352051E-03
2 2.821493E+00 2.821564E+00 -2.534962E-05 -1.881016E-02
3 2.731738E+00 2.731768E+00 -1.108409E-05 -2.713281E-03
4 2.708183E+00 2.708187E+00 -1.320544E-06 8.052734E-04
5 2.699788E+00 2.699798E+00 -3.444082E-06 1.659180E-03
6 2.695010E+00 2.695005E+00 2.034737E-06 2.547559E-03
7 2.694282E+00 2.694286E+00 -1.327357E-06 2.586133E-03
7D 2.754172E+00 2.754163E+00 3.289531E-06 -1.904385E-02

---------------------------------------------------------------------------------------------------------------


----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | A1 | 1.0000E+00 : 7.0992E-01 : 7.8247E-01 : 7.9916E-01 : 8.1306E-01 : 8.2427E-01 :
2 | 2 | A2 | 2.0000E+00 : 9.9941E-01 : 6.0839E-01 : 4.7140E-01 : 4.0851E-01 : 3.6842E-01 :
3 | 3 | A3 | 1.0000E+00 : 7.0992E-01 : 7.8247E-01 : 7.9916E-01 : 8.1306E-01 : 8.2427E-01 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 7D : 8 : 8D : 9 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | A1 | 8.3105E-01 : 8.3374E-01 : 8.5000E-01 :
2 | 2 | A2 | 3.4443E-01 : 3.3612E-01 : 3.5000E-01 :
3 | 3 | A3 | 8.3105E-01 : 8.3374E-01 : 8.5000E-01 :
AND HARD FEASIBLE DISCRETE DESIGN OBTAINED

5. Generating an Input File for Part B
Q Modify the DOPTPRM entry in the wkshp10b_fillin.dat file so that Round-up discrete
variable processing will be performed after each design cycle.
Q The solution is as follow:
DOPTPRM,IPRINT, 5, DESMAX, 20, DELP, 0.5, P1, 1

, P2, 15, DISCOD, 3 , DISBEG, 1


wkshp10b scr=yes. Monitor the run using the UNIX ps command:

7. View Results
When the run is completed, edit the wkshp10b.f06 file and search for the word FATAL. If no matches exist, search for the
word WARNING. Determine whether existing WARNING messages indicate modeling errors.
While still editing wkshp10b.f06, search for the word “H I S T O R Y”. Compare the results obtained in the .f06 file with the
following:
***************************************************************
***************************************************************
NUMBER OF DISCRETE PROCESSING ANALYSES COMPLETED 7
---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 4.828427E+00 -3.234952E-01
1 3.012073E+00 3.012374E+00 -9.996191E-05 -5.330859E-03
1D 3.121351E+00 3.121320E+00 9.700753E-06 -4.971670E-02
2 2.823109E+00 2.823122E+00 -4.391509E-06 -1.420059E-02
2D 2.912727E+00 2.912742E+00 -5.074927E-06 -4.611885E-02
3 2.737844E+00 2.737881E+00 -1.384594E-05 -3.985351E-03
3D 2.762745E+00 2.762742E+00 1.121872E-06 -1.138428E-02
4 2.712771E+00 2.712811E+00 -1.458911E-05 -1.378711E-03
4D 2.854178E+00 2.854163E+00 5.346152E-06 -4.905635E-02
5 2.697974E+00 2.697962E+00 4.595234E-06 2.384277E-03
5D 2.804153E+00 2.804163E+00 -3.485946E-06 -3.479961E-02
6 2.695038E+00 2.695050E+00 -4.688663E-06 2.654395E-03
6D 2.754164E+00 2.754163E+00 8.656661E-08 -1.904385E-02
7 2.694296E+00 2.694297E+00 -6.194305E-07 2.684863E-03
7D 2.754147E+00 2.754163E+00 -5.973096E-06 -1.904385E-02


----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | INITIAL : 1 : 1D : 2 : 2D : 3 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | A1 | 1.0000E+00 : 7.1148E-01 : 7.5000E-01 : 7.7204E-01 : 8.0000E-01 : 7.9776E-01 :
2 | 2 | A2 | 2.0000E+00 : 1.0000E+00 : 1.0000E+00 : 6.3946E-01 : 6.5000E-01 : 4.8147E-01 :
3 | 3 | A3 | 1.0000E+00 : 7.1148E-01 : 7.5000E-01 : 7.7204E-01 : 8.0000E-01 : 7.9776E-01 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 3D : 4 : 4D : 5 : 5D : 6 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | A1 | 8.0000E-01 : 8.1695E-01 : 8.5000E-01 : 8.2337E-01 : 8.5000E-01 : 8.2969E-01 :
2 | 2 | A2 | 5.0000E-01 : 4.0213E-01 : 4.5000E-01 : 3.6911E-01 : 4.0000E-01 : 3.4832E-01 :
3 | 3 | A3 | 8.0000E-01 : 8.1695E-01 : 8.5000E-01 : 8.2337E-01 : 8.5000E-01 : 8.2969E-01 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6D : 7 : 7D : 8 : 8D : 9 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | A1 | 8.5000E-01 : 8.3236E-01 : 8.5000E-01 :
2 | 2 | A2 | 3.5000E-01 : 3.4002E-01 : 3.5000E-01 :
3 | 3 | A3 | 8.5000E-01 : 8.3236E-01 : 8.5000E-01 :

WORKSHOP 11
ENHANCEMENT OF DYNAMIC RESPONSE

CREATION

Objective:
Learn different ways to define a synthetic dynamic response.
Use the same mass-sprint-damp system for workshop 4

Suggested Exercise Steps
2. Compare the Sensitivity Printout
3. Generate an Input File for Part A.
4. Compare an Input File for Part A.
5. Generate an Input File for Part B.
6. Compare an Input File for Part B.
7. Generate an Input File for Part C.
8. Compare an Input File for Part C.

X Objective Function:
Minimize an SSQ response of displacements at grid 1 (X-direction) at

frequencies = 4.,4.5,5.,5.5 and 6.0 Hz.
X Design Variables:
Spring constants: K1, K2 and K3
Damping coefficients: GE1, GE2 and GE3

Model Description
X Part A:
Define the SSQ response using the new DEQATN intrinsic function
X Part B:
Define the SSQ response using direct DRESP2 function reference
X Part C:
Define the SSQ response using direct DRESP1 function reference

1. Generating an Input File for MSC.NASTRAN Users
Q The following file contains the complete listing of the input file. However, the SSQ
response is explicitly defined using the traditional DEQATN entry.
TIME 5
SOL 200 $
CEND
TITLE=TWO MASS DIRECT FREQUENCY SENSITIVITY D108G7v $
DISPL(PHASE) = ALL DVPREL1 1 PELAS 20 3 1.0E3
SUBCASE 2 1 1.0
DESOBJ = 100 DVPREL1 2 PELAS 21 3 1.5E3
ANALYSIS = DFREQ 2 1.0
FREQ = 100 DVPREL1 3 PELAS 22 3 2.0E3
DLOAD = 200 3 1.0
dsaprt(end=sens) = all DVPREL1 4 PELAS 20 4 .05
BEGIN BULK 4 1.0
$ DVPREL1 5 PELAS 21 4 .05
CELAS1 20 20 1 1 5 1.0
CELAS1 21 21 1 1 2 1 DVPREL1 6 PELAS 22 4 .05
CELAS1 22 22 2 1 6 1.0
CONM2 10 1 1. FREQ 100 4.0 4.5 5.0 5.5 6.0
CONM2 11 2 2. GRID 1 1. 0. 0.
DAREA 201 1 1 100. 2 1 100. 23456
DESVAR 1 K1 1.0E3 1.0E2 1.0E4 GRID 2 2. 0. 0.
DESVAR 2 K2 1.5E3 1.5E2 1.5E4 23456
DESVAR 3 K3 2.0E3 2.0E2 2.0E4 MAT1 200 1.0E7 0.3 0.1
DESVAR 4 GE1 .1 .05 .15 PELAS 20 1.0E3
DESVAR 5 GE2 .1 .05 .15 PELAS 21 1.5E3
DESVAR 6 GE3 .1 .05 .15 PELAS 22 2.0E3
$ 23 4 5 6 7 RLOAD1 200 201 210
8 9 0 TABLED1 210
1. 1. 10. 1. ENDT
DRESP1 11 U1 FRDISP 1 4.0 1
ENDDATA
$
$
$ Compute FDSSQ = (R1**2+R2**2+R3**2+R4**2+R5**2)
$ Explicitly write out a SSQ equation with DEQATN
$
DRESP2 100 FDSSQ 200
DRESP1 11 12 13 14
15
DEQATN 200 FDSSQ(R1,R2,R3,R4,R5)=(R1**2+R2**2+R3**2+R4**2+R5**2)
$

2. Compare the Sensitivity Printout
The sensitivity print out for the SSQ response
****************************************************************************
* *
* D E S I G N S E N S I T I V I T Y M A T R I X O U T P U T *
* *
* *
* R E S P O N S E S E N S I T I V I T Y C O E F F I C I E N T S *
* *
****************************************************************************
--------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------
2 1.8718E-01 4.0000E+00 1 K1 -2.0951E-04 2 K2 -7.1772E-06 3 K3 -1.2402E-04
4 GE1 -4.8650E-02 5 GE2 9.2470E-05 6 GE3 -7.5393E-02
--------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------
2 3.1026E-01 4.5000E+00 1 K1 -5.1749E-04 2 K2 -7.8940E-06 3 K3 -3.6389E-04
4 GE1 -2.1688E-01 5 GE2 2.2538E-03 6 GE3 -3.9102E-01
--------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------
2 6.6514E-01 5.0000E+00 1 K1 -6.9685E-04 2 K2 -3.1737E-06 3 K3 -3.3339E-04
4 GE1 -2.2073E+00 5 GE2 1.3004E-02 6 GE3 -4.3026E+00
--------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------
2 2.9241E-01 5.5000E+00 1 K1 2.8222E-04 2 K2 8.7088E-06 3 K3 4.6273E-04
4 GE1 -2.1884E-01 5 GE2 1.2674E-02 6 GE3 -3.7439E-01
1 TWO MASS DIRECT FREQUENCY SENSITIVITY D108G7V NOVEMBER 8, 2002 MSC.NASTRAN 11/ 1/02 PAGE 19
--------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------
2 1.3771E-01 6.0000E+00 1 K1 5.1171E-05 2 K2 1.2517E-05 3 K3 1.5027E-04
4 GE1 -3.3135E-02 5 GE2 1.2923E-02 6 GE3 -2.5227E-02
--------------------------------------------------------------------------------------------------------------------------------
DRESP2 ID= 100 RESPONSE TYPE= SYNTHETIC SEID= 0
--------------------------------------------------------------------------------------------------------------------------------
2 6.7817E-01 4.0000E+00 1 K1 -1.1474E-03 2 K2 -3.2665E-06 3 K3 -4.0372E-04
4 GE1 -3.2262E+00 5 GE2 2.9704E-02 6 GE3 -6.2204E+00

3. Generate an Input File for Part A
Modify the input file, wkshp11a_fillin.dat to complete the exercise. The following is the
portion of the file to be filled in.

$ Method a: use new DEQATN intrinsic function: SSQ(X1,...,Xn)
$
$ First identify dynamic displacements at component1, grid 1 and at
$ frequencies=4.,4.5,5.,5.5,6.
$
DRESP1 ...
DRESP1 ...
DRESP1 ...
DRESP1 ...
DRESP1 ...
$
$ define a SSQ function using new DEQATN intrinsic function
$
DRESP2 ...
$
DEQATN ...

4. Compare an Input File for Part A
The completed portion of Part A is shown below:
$
$ Method a: use new DEQATN intrinsic function: SSQ(X1,...,Xn)
$
$
DRESP2 100 FDSSQ 200
DRESP1 11 12 13 14
15
DEQATN 200 FDSSQ(R1,R2,R3,R4,R5)=SSQ(R1,R2,R3,R4,R5)
$
Compare the results with that provided in page 11-7.

5. Generating an Input File for Part B
Q Modify the input file, wkshp11b_fillin.dat to complete the exercise. The following is the portion of
the file to be filled in.

$
$ First identify dynamic displacements at component1, grid 1 and at
$ frequencies=4.,4.5,5.,5.5,6.
$
DRESP1 ...
DRESP1 ...
DRESP1 ...
DRESP1 ...
DRESP1 ...
$
$ define a SSQ function using direct DRESP2 function reference
$
DRESP2 ...
$

6. Compare an Input File for Part B
The completed portion for the Part B exercise is shown here:
$ Method b: use durect DRESP2 function reference
$
$
DRESP2 100 FDSSQ SSQ
DRESP1 11 12 13 14
15
Compare the results with that provided in page 11-7.

7. Generating an Input File for Part C
Q Modify the input file, wkshp11c_fillin.dat to complete the exercise. The following is the portion of
the file to be filled in.
$
$
$ Identify dynamic dislacement response at grid1, component 1
$ and use direct DRESP1 function reference
$
DRESP1 ...

8. Compare Input File for Part C
The completed portion for the Part C exercise is shown here:
$
$
$ Identify dynamic dislacement response at grid1, component 1
$ and use direct DRESP1 function reference
$
DRESP1 100 U1 FRDISP 1 SSQ 1
Compare the results with that provided in page 11-7. Notice that only the SSQ response
sensitivities are printed out since the five individual displacement responses are not
explicitly defined.

WORKSHOP 12
EXTERNAL RESPONSE (DRESP3)

Objective:
Learn how to set up an external response optimization task using
the same box design problem used in Workshop 2. You will write Fortran
routines to define the volume and surface area instead of using DEQATN
entries.

Suggested Exercise Steps
3. Modify DRESP3 Server Routine, r3sgrt
4. Compare the modified server routine, r3sgrt
5. Modify server routine to define volume and surface area
6. Compare modified server routine, r3svald
7. Build the response server
8. Submit Input file for MSC.Nastran for analysis
9. View Results

Model Description
X Objective Function: Minimize the material surface area.
S = 2 * ( W * H + L * H + 2.0 * W * L )
X Design Variables: W (Width), H (Height), L (Length)
X Constraints:
Volume: V = H * W * L ≥ 2.0
X Initial Design Optimum Design

H = 1.5 H = 2.0
W = 1.5 W = 1.0
L = 1.5 L = 1.0
S = 18.0 S = 12.0
V = 3.375 V = 2.0

1. Generating an Input File for MSC.NASTRAN Users
Q Following the instructions given in the input file, wkshp12_fillin.dat to complete the
exercise. The following input file is provided as a starting point.
$ Define an external response group using a CONNECT entry. Notice that

$ the group name must be referenced by the subsequent DRESP3 entries
CONNECT …
TIME 5
diag 8,56 $
SOL 200
CEND
TITLE = BOX MATERIAL MINIMIZATION WITH A VOLUME CONSTRAINT D200X8
DESOBJ = 2
DESSUB = 1
SUBTITLE = INITIAL DESIGN: W=1.5, L=1.5, AND H=1.5
$DISPLACEMENT = ALL
STRESS = NONE
SPC = 1
ANALYSIS = STATICS
LOAD = 1
BEGIN BULK
$ DUMMY STRUCTURAL MODEL (NOT USED)
FORCE 1 2 1.0E7 1.0 0.0 0.0
GRID 1 0.0 0.0 0.0
GRID 2 1.0 0.0 0.0
MAT1 1 1.0E7 0.3 0.1
CROD 1 1 1 2
PROD 1 1 1.0
SPC1 1 23456 2
SPC1 1 123456 1
$
$ DUMMY DESIGN PROPERTY RELATIONS
$
DVPREL1 1 PROD 1 4
3 1.0
$
$ BOX DESIGN PROBLEM DEFINITION
$
DESVAR DESVAR 2 L 1.5 0.1 10.0
DESVAR 3 H 1.5 0.1 10.0
$ Define volume and surface area using DRESP3 entries
DRESP3…
DRESP3…
DCONSTR 1 1 2.0 1.0E35
DSCREEN DRESP3 -100.0 1
DOPTPRM DESMAX 10 delp 1.0
gmax 1.e-16 ctmin 1.e-16
ENDDATA

The completed input file is shown below:
CONNECT DRESP3 WKSHOP EXTRESP $

TIME 5 $ BOX DESIGN PROBLEM DEFINITION
diag 8,56 $ $
SOL 200 DESVAR 1 W 1.5 0.1 10.0
CEND DESVAR 2 L 1.5 0.1 10.0
TITLE = BOX MATERIAL MINIMIZATION WITH A VOLUME DESVAR 3 H 1.5 0.1 10.0
CONSTRAINT D200X8 DRESP3 1 VOLUME WKSHOP VOLUME
DESOBJ = 2 DESVAR 1 2 3
DESSUB = 1 DRESP3 2 SURFACE WKSHOP SURFACE
SUBTITLE = INITIAL DESIGN: W=1.5, L=1.5, AND H=1.5 DESVAR 1 2 3
$DISPLACEMENT = ALL DCONSTR 1 1 2.0 1.0E35
STRESS = NONE DSCREEN DRESP3 -100.0 1
SPC = 1 DOPTPRM DESMAX 10 delp 1.0
ANALYSIS = STATICS gmax 1.e-16 ctmin 1.e-16
LOAD = 1 ENDDATA
BEGIN BULK
$
$ DUMMY STRUCTURAL MODEL (NOT USED)
$
FORCE 1 2 1.0E7 1.0 0.0 0.0
GRID 1 0.0 0.0 0.0
GRID 2 1.0 0.0 0.0
MAT1 1 1.0E7 0.3 0.1
CROD 1 1 1 2
PROD 1 1 1.0
SPC1 1 23456 2
SPC1 1 123456 1
$
$ DUMMY DESIGN PROPERTY RELATIONS
$
DVPREL1 1 PROD 1 4
3 1.0

3. Modify DRESP3 Server Routine, R3SGRT
Modify DRESP3 server routine, r3grt to define response types.
SUBROUTINE R3SGRT(GRPID,TYPNAM,ERROR)
C 09/18/00;14:24:09 cdp 2001t1 na001130
C Update dr3srv per SHZ
C ----------------------------------------------------------------------
C PURPOSE: VERIFY THE EXTERNAL RESPONSE TYPE
C Arguments
C GRPID input integer - Group ID
C TYPNAM input character*8 - Name of external response type
C ERROR input/output integer -error code for the call.
C
C Method
C Match the user input: typnam with the list of available
C external response types. If no match is found, set error code.
C Called by
C R3CGRT
C NOTE:
C The writer of this routine is responsible to specify
C NTYPES and R3TYPE.
C ----------------------------------------------------------------------
C VARIABLES PASSED IN
INTEGER GRPID, ERROR
CHARACTER*8 TYPNAM
C LOCAL VARIABLES
INTEGER BADTYP
C
C declaration for character array R3TYPE and initialize character
C array R3TYPE with corresponding R3TYPE from DRESP3s
…
DATA BADTYP/7554/
ERROR = 0
DO 100 ITYPE = 1, NTYPES
IF (TYPNAM .EQ. R3TYPE(ITYPE)) THEN
GOTO 200
END IF
100 CONTINUE
ERROR = BADTYP
200 CONTINUE
RETURN
END

4. Compare the Server Routine, R3SGRT
The listing of server routine, r3grt is shown here:
SUBROUTINE R3SGRT(GRPID,TYPNAM,ERROR)
C 09/18/00;14:24:09 cdp 2001t1 na001130
C Update dr3srv per SHZ
C ----------------------------------------------------------------------
C
C PURPOSE: VERIFY THE EXTERNAL RESPONSE TYPE
C
C GRPID input integer - Group ID
C
C Method
C Match the user input: typnam with the list of available
C external response types. If no match is found, set error code.
C
C Called by
C R3CGRT
C
C NOTE:
C The writer of this routine is responsible to specify
C NTYPES and R3TYPE.
C ----------------------------------------------------------------------
C
C
INTEGER GRPID, ERROR
CHARACTER*8 TYPNAM
C
C LOCAL VARIABLES
C
INTEGER NTYPES, BADTYP
PARAMETER(NTYPES=2)
CHARACTER*8 R3TYPE(NTYPES)
C
DATA BADTYP/7554/
DATA R3TYPE/'VOLUME','SURFACE'/
ERROR = 0
DO 100 ITYPE = 1, NTYPES
IF (TYPNAM .EQ. R3TYPE(ITYPE)) THEN
GOTO 200
END IF
100 CONTINUE
ERROR = BADTYP
200 CONTINUE
C
RETURN
END

5. Modify Server Routine to Define Volume and Surface Area
Q Modify DRESP3 server routine, r3vald to define volume and surface area.
SUBROUTINE R3SVALD(GRPID,TYPNAM,
. NITEMS,ARGLIS,
. NSIZE, ARGVAL,
. NWRDA8,ARGCHR,
. DR3VAL,ERROR)
C ----------------------------------------------------------------------
C
C PURPOSE: COMPUTE THE EXTERNAL RESPONSE
C
C GRPID input integer - Group id
C NITEMS input integer - dimension of array ARGLIS
C NSIZE input integer - dimension of array ARGVAL
C NWRDA8 input integer - dimension of character array ARGCHR
C ARGLIS input integer - array of no. of items for each argument
C ARGVAL input double - array of all argument values
C ARGCHR input character*8 - array of characters
C DR3VAL output double - value of the external response
C 0 = print error messages
C 1 = do not print error messages.
C
C Portion of routine is omitted
C
C LOCAL VARIABLES
C
INTEGER BADTYP
C declare user defined variables
…
C
DATA BADTYP /7554/
C
C THE USER-SUPPLIED EQUATION TO DEFINE THE EXTERNAL RESPONSE
C DEQATN 1 F(W,L,H)=W*L*H
C DEQATN 2 F(W,L,H)=2.0*(W*H+L*H+2.0*W*L)
C
ERROR = 0
C
C define equations of volume and surface area
C
…
RETURN
END

6. Compare the modified Server Routine, R3SVALD
The listing of server routine, r3vald is shown here. Notice there are many ways to modify
this routine:
SUBROUTINE R3SVALD(GRPID,TYPNAM,
. NITEMS,ARGLIS,
. NSIZE, ARGVAL,
. NWRDA8,ARGCHR,
. DR3VAL,ERROR)
C 07/02/01;11:30:11 cdp V2001 nxf01800
C Update per SHZ
C ----------------------------------------------------------------------
C
C PURPOSE: COMPUTE THE EXTERNAL RESPONSE
C
C GRPID input integer - Group id
C NITEMS input integer - dimension of array ARGLIS
C NSIZE input integer - dimension of array ARGVAL
C NWRDA8 input integer - dimension of character array ARGCHR
C ARGLIS input integer - array of no. of items for each argument
C ARGVAL input double - array of all argument values
C ARGCHR input character*8 - array of characters
C DR3VAL output double - value of the external response
C 0 = print error messages
C 1 = do not print error messages.
C
C Method
C A)Set up various parameters from the argument list
C B)Evaluate the external response based on the given TYPNAM
C C)Return BADTYP error if TYPNAM is not matched here.
C
C Called by
C SendR3svald
C ----------------------------------------------------------------------
C

6. Compare the modified Server Routine (Cont.)
C
C
CHARACTER*8 TYPNAM, ARGCHR(NWRDA8)
INTEGER GRPID, NITEMS, NSIZE, ARGLIS(NITEMS), ERROR
DOUBLE PRECISION ARGVAL(NSIZE), DR3VAL
C
C LOCAL VARIABLES
C
INTEGER BADTYP
DOUBLE PRECISION W,L,H
C
DATA BADTYP /7554/
C
C THE USER-SUPPLIED EQUATION TO DEFINE THE EXTERNAL RESPONSE
C DEQATN 1 F(W,L,H)=W*L*H
C DEQATN 2 F(W,L,H)=2.0*(W*H+L*H+2.0*W*L)
C
ERROR = 0
C
C set up parameters for various argument items
C
W = argval(1)
L = argval(2)
H = argval(3)
IF (TYPNAM .EQ. 'VOLUME') THEN
VOL = W*L*H
dr3val = VOL
ELSE IF (TYPNAM .EQ. 'SURFACE') THEN
SURF = 2.D0*(W*H+L*H+2.D0*W*L)
DR3VAL = SURF
ELSE
ERROR = BADTYP
END IF
RETURN
END

7. Build Response Server
Q Build the response server using the following command:
msc2001 ./dr3srv build
Q Generate a connection evaluator file, called wkshp12_eval that associates the external
response group with the actual server program. The full content of file, wkshp12_eval is
shown here.
EXTRESP, - , /home/wkshop/server/dr3serv
Q Submit the Nastran job by adding a new keyword, gmconn where gmconn refers to the
evaluator file name.


Q find an available UNIX shell window and at the command prompt enter
Nastran fn scr=yes gmconn=wkshp12_eval

9. View Results
When the run is completed, edit the wkshp12.f06 file and search for the word FATAL. If no matches exist, search for the
word WARNING. Determine whether existing WARNING messages indicate modeling errors.
While still editing wkshp12.f06, search for the word “H I S T O R Y”. Compare the results obtained in the .f06 file with the
following:
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 1.800000E+01 -6.875000E-01
1 1.200000E+01 1.200000E+01 0.000000E+00 -5.960464E-07
2 1.200000E+01 1.200000E+01 0.000000E+00 -5.960464E-07

---------------------------------------------------------------------------------------------------------------
1 BOX MATERIAL MINIMIZATION WITH A VOLUME CONSTRAINT D200X8 NOVEMBER 7, 2002 MSC.NASTRAN 11/ 7/02 PAGE 20
INITIAL DESIGN: W=1.5, L=1.5, AND H=1.5
0
----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | W | 1.5000E+00 : 1.0000E+00 : 1.0000E+00 :
2 | 2 | L | 1.5000E+00 : 1.0000E+00 : 1.0000E+00 :
3 | 3 | H | 1.5000E+00 : 2.0000E+00 : 2.0000E+00 :

WORKSHOP 13
AVOIDING A FREQUENCY RANGE
t1 t2 t3

Objective:
Design a cantilever structure to avoid frequency ranges using
synthetic response (DRESP2) to reduce dynamic amplification at
resonance.

Analysis Model Description
• A cantilever beam with 3 regions: t1, t2, t3 modeled with QUAD4
elements. Three PSHELL entries 1, 2 and 3 are used.
• It has three initial frequencies: F1=31.8Hz, F2=197.1 Hz and
F3=313.4Hz that are to be avoided.

• Objective function:
Minimize the structural weight.
• Design Variables:
Thickness of three PSHELL entries, 1, 2 and 3.
0.1 < X1 < 1.0
0.1 < X2 < 1 .0
0.1 < X3 < 1.0
X10 = 0.1, X20 = 0.1, X30 = 0.1
• Constraints:
Avoid three given frequencies by imposing the following:
F1 < 30 and F1 > 34, F2 < 100 and F2 > 200, F3 < 275 and F3 > 325
• In a separate run, replace F2<100 with F2<190. Run the new job and
compare the results.

Special Procedure to Avoid Frequency Ranges
Q The constraints can be defined algebraically with the following:
if f < flb or f > fub and flb < fub

2 ⎛ 2
⎛ f + f ⎞ f − f ⎞
then R2 = ⎜⎜ f − lb ub ⎟⎟ > ⎜⎜ ub lb ⎟⎟ = R2L
⎜ 2 ⎟ ⎜ 2 ⎟
⎝ ⎠ ⎝ ⎠
where
f = calculated frequency
flb = lower bound frequency
fub = upper bound frequency
Q In SOL200, the above constraints can be defined with the following steps:
Q First, use DRESP1 to identify three natural frequencies (f)
Q Next, use DABLE to specify lower and upper frequency bounds (flb, fub)
Q Then, use DRESP2 and DEQATN to define three R2.
Q Finally, use DCONSTR to specify lower limit, R2L to each R2.

starting point:
SOL 200
CEND
$
TITLE = CANTILEVERED BEAM MADE OF PLATES
$
$ CQUAD4 1 1 101 102 202 201
$ DISP = ALL CQUAD4 2 1 102 103 203 202
$ fill in response ID in DESOBJ command CQUAD4 3 1 103 104 204 203
CQUAD4 4 2 104 105 205 204
desobj(min)= $ minimize weight
CQUAD4 5 2 105 106 206 205
$
CQUAD4 6 2 106 107 207 206
$ fill in design case ID in DESSUB command CQUAD4 7 2 107 108 208 207
dessub = $ keep frequency away from certain values, see below CQUAD4 8 3 108 109 209 208
$ CQUAD4 9 3 109 110 210 209
$ CQUAD4 10 3 110 111 211 210
SUBCASE 10301 $ Normal Modes MAT1 1 10.E6 .3 2.588-4 1.E-6 0.
ANALYSIS=MODES PSHELL 1 1 .1 1
METHOD = 10301 PSHELL 2 1 .1 1
BEGIN BULK PSHELL 3 1 .1 1
$ $.......2.......3.......4.......5.......6.......7.......8......
$ ANALYSIS MODEL CREATION .9.......0
$ $ Eigenvalue Method for SOL 103/Normal Modes
$.......2.......3.......4.......5.......6.......7.......8.......9....... EIGRL 10301 -.1 2000. 10
0 $
$ $ DESIGN MODEL CREATION
$ Cantilevered Beam Made of Plates Model $
$
$ Fill in Three thickness design variables
$ \201 202 203 204 205 206 207 208 209 210 211
$ Y \*----*----*----*----*----*----*----*----*----*----* $ and relate thee design variables to the
$ ^ \| | | | | | | | | | | thickness of three PSHELL entries
$ | \| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | $
$ | \| | | | | | | | | | | DESVAR...
$ +--->X \*----*----*----*----*----*----*----*----*----*----* DVPREL1...
$ \101 102 103 104 105 106 107 108 109 110 111 $
$ $ Fill in for define a WEIGHT Response
$.......2.......3.......4.......5.......6.......7.......8.......9....... $
0 DRESP1...
GRID 101 0. 0. 0. 123456
GRID 102 1. 0. 0.
= *1 = *1. ==
=8
GRID 201 0. 1. 0. 123456
GRID 202 1. 1. 0.
= *1 = *1. ==
=8
$

$
$ Fill in: Define first three natural frequency responses
$
DRESP1...
...
$
$ Fill in: Specify table constants for three sets of lower and upper frequency limits.
$
DTABLE...
$
$ Fill in: Define synthetic response to avoid the frequency range.
$
DRESP2...
...
$
$ Fill in: Apply the lower bound for three synthetic responses
$
$ Lower bound = ((fub-flb)/2)**2
$ For F1: ((34-30)/2)) **2 = 4
$ For F2: ((200-100)/2)**2 = 2500
$ For F3: ((325-275)/2)**2 = 625
$
DCONSTR...
...
$
$ set DOPTPRM parameters: desmax=20, p1=1,p2=15
$
doptprm p1 1 p2 15 DESMAX 20
ENDDATA

SOL 200
CEND
$ CBAR 5 11 1 5 0. 1. 0.
TITLE = CANTILEVERED BEAM MADE OF PLATES CBAR 6 11 5 2 0. 1. 0.
$ PBAR 11 12 1.e2 1.e3 1.e3
$
$ DISP = ALL 100. 0. -100. 0.
desobj(min)=101 $ minimize weight MAT1 12 3.E+7 7.8e-4
dessub=200 $ keep frequency away from certain values, see below $
$
SUBCASE 10301 $ Normal Modes $ CONSTRAIN TO A PLANAR PROBLEM
ANALYSIS=MODES $
METHOD = 10301 SPC1 100 1345 1 2 5
BEGIN BULK
$ SPC1 100 13456 3 4
$ ANALYSIS MODEL CREATION spc1,100,2,3,4
$ $
$.......2.......3.......4.......5.......6.......7.......8.......9......
.0
CONM2 10 1 2.5
$ CONM2 15 2 2.5
$ Cantilevered Beam Made of Plates Model CONM2 20 5 5.0
$
$ \201 202 203 204 205 206 207 208 209 210 211
CONM2 25 3 2.5
$ Y \*----*----*----*----*----*----*----*----*----*----* CONM2 26 4 2.5
$ ^ \| | | | | | | | | | | $
$ | \| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
$ | \| | | | | | | | | | |
CBUSH 100 1000 1 3 0
$ +--->X \*----*----*----*----*----*----*----*----*----*----* CBUSH 200 1000 2 4 0
$ \101 102 103 104 105 106 107 108 109 110 111 PBUSH 1000 K 200.
$
$.......2.......3.......4.......5.......6.......7.......8.......9....…0
B 2.
GRID 101 0. 0. 0. 123456 $
GRID 102 1. 0. 0. RLOAD2 111 222 444 disp
= *1 = *1. ==
=8
GRID 201 0. 1. 0. 123456 spcd,222,3,2,1.
GRID 202 1. 1. 0. TABLED1 444
= *1 = *1. == 0. 1. 100. 1. ENDT
=8
$ $
CQUAD4 1 1 101 102 202 201 RLOAD2 112 223 444 disp
CQUAD4 2 1 102 103 203 202 spcd,223,4,2,1.
CQUAD4 3 1 103 104 204 203
CQUAD4 4 2 104 105 205 204 $
CQUAD4 5 2 105 106 206 205 FREQ2 130 0.1 2. 20
CQUAD4 6 2 106 107 207 206 $
CQUAD4 7 2 107 108 208 207
CQUAD4 8 3 108 109 209 208 $ DEFINE THE INPUT PSD
CQUAD4 9 3 109 110 210 209 $
CQUAD4 10 3 110 111 211 210 RANDPS 1000 1 1 1. 0. 145
MAT1 1 10.E6 .3 2.588-4 1.E-6 0.
PSHELL 1 1 .1 1 RANDPS 1000 2 2 1. 0. 145
PSHELL 2 1 .1 1 RANDPS 1000 1 2 1. 0. 146
PSHELL 3 1 .1 1 RANDPS 1000 1 2 0. 1. 147
$.......2.......3.......4.......5.......6.......7.......8.......9......
.0 TABRND1 145
$ Eigenvalue Method for SOL 103/Normal Modes .1 .1 5. 1. 10. .05 ENDT
EIGRL 10301 -.1 2000. 10
$

$ DESIGN MODEL CREATION DRESP2 152 F2star 99

$
$ Three thickness design variables DTABLE f2lb f2ub
$
DESVAR 1 T1 .1 .06 1.0 DRESP1 103
DVPREL1 11 PSHELL 1 T .01 1.0 DRESP2 153 F3star 99
1 1.0
DESVAR 2 T2 .1 .06 1.0 DTABLE f3lb f3ub
DVPREL1 12 PSHELL 2 T .01 1.0
2 1.0 DRESP1 104
DESVAR 3 T3 .1 .06 1.0
DEQATN 99 Fstar(flb,fub,f) = (f-(flb+fub)/2.)**2
DVPREL1 13 PSHELL 3 T .01 1.0
3 1.0 $
$
$ Define a WEIGHT Response $ Apply the lower bound for three synthetic responses
$
$
DRESP1 101 WEIGHT WEIGHT
$ $ Lower bound = ((fub-flb)/2)**2
$ Define first three natural frequency responses
$ $ For F1: ((34-30)/2)) **2 = 4
DRESP1 102 FREQ1 FREQ 1
DRESP1 103 FREQ2 FREQ 2 $ For F2: ((200-100)/2)**2 = 2500
DRESP1 104 FREQ3 FREQ 3 $ For F3: ((325-275)/2)**2 = 625
$
$ Specify table constants for three sets of lower and upper frequency limits. $
$
DTABLE f1lb 30. f1ub 34. f2lb 100. f2ub 200. DCONSTR 201 151 4.
f3lb 275. f3ub 325. DCONSTR 202 152 2500.
$
DCONSTR 203 153 625.
$ Define synthetic response to avoid the frequency range.
$
$
DCONADD 200 201 202 203
DRESP2 151 F1star 99
$
DTABLE f1lb f1ub
doptprm p1 1 p2 15 DESMAX 20
DRESP1 102
ENDDATA



4. View Results
errors.
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 2.588000E-04 9.869587E-01
1 2.467133E-04 2.467117E-04 6.370216E-06 2.828887E-02
2 2.418629E-04 2.418635E-04 -2.466799E-06 9.159404E-02
3 2.433889E-04 2.433896E-04 -3.109005E-06 1.650176E-02
4 2.437233E-04 2.437244E-04 -4.418277E-06 1.500293E-03
5 2.436734E-04 2.436738E-04 -1.672127E-06 2.750781E-03

---------------------------------------------------------------------------------------------------------------

0 SUBCASE 10301 $ NO
----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | T1 | 1.0000E-01 : 9.9828E-02 : 1.0161E-01 : 1.0020E-01 : 9.9108E-02 : 9.8211E-02 :
2 | 2 | T2 | 1.0000E-01 : 1.0345E-01 : 1.0949E-01 : 1.0723E-01 : 1.0973E-01 : 1.0945E-01 :
3 | 3 | T3 | 1.0000E-01 : 8.0000E-02 : 6.3924E-02 : 7.0305E-02 : 6.8505E-02 : 6.9702E-02 :
Initial Frequencies
F1=31.8, F2=197.1, F3=313.3
Final Frequencies
F1=35.8, F2=199.9, F3=353.0


Fig. 13. 3: Maximum Constraint History

WORKSHOP 14
STIFFENED PLATE OPTIMIZATION WITH
OBJECTIVE FORMED USING CROSSING-
SUBCASE RESPONSES

Objective:
Design the stiffened plate structure by minimizing an objective
formed by responses crossing subcases and satisfying additional stress
and displacement constraints.

Analysis Model Description (See Section 7.4 of
DS&O User Guide for more information)
• 16 QUAD4 elements to model uniform thickness base plate
• 4 CBAR elements to model the hat stiffener. The HAT type is selected
on the PBARL entry.
• Material: E=1.0E7 psi
Poisson Ratio = 0.33
Weight density = 0.283 lbs/in3
• Loadings: Subcase 1: In-plane load of Nx=1000 lbs/in
Uniform pressure load of 50 psi over the plate in
the positive z-direction
Subcase 2: Lumped vertical load of 10,000 lbs at Grid 10203

• Objective (Minimize)
2 2
F = 10W + ( ∑ ( SA j1 + SA j 2 + SA j3 + SA j 4 ) + ∑ (SB j1 + SB j 2 + SB j3 + SB j 4 )) /1000
j =1 j =1
where W is the weight and SAJI and SBJI are maximum stresses at ends A and B
of Bar elements. J is the subcase index and I is the element index.
• Design Variables (X1,X2) relates analysis model properties by
t plate = 1.0 X1
that = 0.1X 2
w3 = 1.5 + 0.5 X1
X10=0.15, X20=1.0
• Constraints:
Strength: von Mises stress <=25,000 at the centroid of plate elements
Displacement: |Vertical displacement at grid point 10302 for Subcase 1|<=0.1 in
|Vertical displacement at grid point 10203 for Subcase 2|<=0.03 in
starting point:
BEGIN BULK
ID MSC DSOUG4 $ v2004 ehj 25-Jun-2003 param post -1
TIME 10 PARAM NASPRT 1
SOL 200 $ OPTIMIZATION $
CEND $------------------------------------------------------------------------
$ $ ANALYSIS MODEL:
TITLE = STATIC ANALYSIS OF A STIFFENED PLATE $------------------------------------------------------------------------
DSOUG4 $
ECHO = BOTH GRID 10000 0.0 0.0 0.0
DISP = ALL …
STRESS = ALL GRID 10404 10.0 10.0 0.0
SPC = 1 $
ANALYSIS = STATICS CQUAD4 1 1 10000 10001 10101 10100
DSAPRT(START=1,END=LAST)=ALL …
$ CQUAD4 16 1 10303 10304 10404 10403
$ use DESOBJ command to reference the synthetic response $
$ CBAR 31 3 10200 10201 0.0 1.0 0.0
DESOBJ(MIN) = 1.575 1.575
$ CBAR 32 3 10201 10202 0.0 1.0 0.0
SUBCASE 1 1.575 1.575
$ CBAR 33 3 10202 10203 0.0 1.0 0.0
$ Identify stress DRESP1 responses for 1st subcase using 1.575 1.575
SET and DRSPAN CBAR 34 3 10203 10204 0.0 1.0 0.0
set 21 = 21, 22, 23, 24, 31, 32, 33, 34 1.575 1.575
DRSPAN = 21 $
LABEL = LOAD CONDITION 1 PSHELL 1 1 0.15 1
LOAD = 1 PSHELL 2 1 0.2 1
$ PBARL 3 1 hat
$ Set up a design case for 1st subcase 3.0 0.1 2.0 0.9
$ $
DESSUB = 100 $ CONSTRAINT DEFININITION MAT1 1 1.0E+7 0.33 0.283
SUBCASE 2 $
$ FORCE 1 10004 2000.0 1.0 0.0 0.0
$ Identify stress DRESP1 responses for 2nd subcase using …
SET and DRSPAN FORCE 2 10203 10000.0 0.0 0.0 1.0
set 22 = 25, 26, 27, 28, 35, 36, 37, 38 PLOAD2 1 50. 1 THRU 16
DRSPAN = 22 $
LABEL = LOAD CONDITION 2 SPC1 1 1236 10000
LOAD = 2 …
$ SPC1 1 6 10301 10302 10303 10304
$ Set up a design case for 2nd subcase $
$ PARAM GRDPNT 1
DESSUB = 200 $ CONSTRAINT DEFININITION PARAM WTMASS 0.00259
PARAM AUTOSPC YES

$------------------------------------------------------------------------
$ DESIGN MODEL:
$------------------------------------------------------------------------
$
$...Define the design variables:
$
$DESVAR ID LABEL XINIT XLB XUB DELXV
DESVAR
DESVAR
$
$...Relate the design variables to analysis model properties
$ (linear relations so use DVPREL1)
$
$...Express shell thicknesses as functions of x1 x2:
$DVPREL1 ID TYPE PID FID PMIN PMAX C0 +
$+ DVIDD1 COEF1 DVID2 COEF2 ...
DVPREL1
$
$...Express BOX THICKNESS as a function of x2:
DVPREL1
$. EXPRESS BOX OFFSET LOCATIONS AS A FUNCTION OF PLATE THICKNESS AND
$ FIXED BOX DIMENSIONS
$ For z-direction of the OFFSET at both ENDs A and B for 4 BAR elements
DVCREL1
...
...
$...Identify the design responses:
$
$DRESP1 ID LABEL RTYPE PTYPE REGION ATTA ATTB ATT1 +
$+ ATT2 ...
$
$ Identify weight response
DRESP1 15 W WEIGHT
$
$ Identify maximum stress at end A in 1st subcase
$
DRESP1 21
DRESP1 22
DRESP1 23
DRESP1 24
$

$ Identify maximum stress at end A in 2nd subcase

$
DRESP1 25
DRESP1 26
DRESP1 27
DRESP1 28
$
$ Identify maximum stress at end B in 1st subcase
$
DRESP1 31
DRESP1 32
DRESP1 33
DRESP1 34
$
$ Identify maximum stress at end B in 2nd subcase
$
DRESP1 35
DRESP1 36
DRESP1 37
DRESP1 38
$
$ define synthetic response with summation of Weight and Bar Stresses
$ that is used as the objective
$
DRESP2
DEQATN 101 OBJ(W,S11,S12,S13,S14,S21,S22,S23,S24,S31,S32,S33,

S34,S41,S42,S43,S44) =
10.*W + (S11+S12+S13+S14+S21+S22+S23+S24+
S31+S32+S33+S34+S41+S42+S43+S44)*1.E-3
$

$
$ Identify von Mises stress/displacement responses
$
DRESP1 3
DRESP1 6
DRESP1 13
DRESP1 14
$
$...Place bounds on the responses:
$
$DCONSTR DCID RID LALLOW UALLOW
DCONSTR
...
$
$ Use DCONADD for design cases 1 & 2
$DCONADD DCID DC1 DC2 ...
$ summed constraint set for subcase 1
DCONADD
DCONADD
$
$...Optional override of optimization parameters:
$
DOPTPRM IPRINT 7 DESMAX 20 DELP 0.5 P1 1 +
+ P2 15
$ (DELP=0.5 allows larger moves thus overcoming constraint
$ violations quicker)
$
ENDDATA

BEGIN BULK
ID MSC DSOUG4 $ v2004 ehj 25-Jun-2003 param post -1
TIME 10 PARAM NASPRT 1
SOL 200 $ OPTIMIZATION $
CEND $------------------------------------------------------------------------
$ $ ANALYSIS MODEL:
TITLE = STATIC ANALYSIS OF A STIFFENED PLATE $------------------------------------------------------------------------
DSOUG4 $
ECHO = BOTH GRID 10000 0.0 0.0 0.0
DISP = ALL …
STRESS = ALL GRID 10404 10.0 10.0 0.0
SPC = 1 $
ANALYSIS = STATICS CQUAD4 1 1 10000 10001 10101 10100
DSAPRT(START=1,END=LAST)=ALL …
$ CQUAD4 16 1 10303 10304 10404 10403
$ use DESOBJ command to reference the synthetic response $
$ CBAR 31 3 10200 10201 0.0 1.0 0.0
DESOBJ(MIN) = 101 1.575 1.575
$ CBAR 32 3 10201 10202 0.0 1.0 0.0
SUBCASE 1 1.575 1.575
$ CBAR 33 3 10202 10203 0.0 1.0 0.0
$ Identify stress DRESP1 responses for 1st subcase using 1.575 1.575
SET and DRSPAN CBAR 34 3 10203 10204 0.0 1.0 0.0
$ 1.575 1.575
set 21 = 21, 22, 23, 24, 31, 32, 33, 34 $
DRSPAN = 21 PSHELL 1 1 0.15 1
LABEL = LOAD CONDITION 1 PSHELL 2 1 0.2 1
LOAD = 1 PBARL 3 1 hat
$ 3.0 0.1 2.0 0.9
$ Set up a design case for 1st subcase $
$ MAT1 1 1.0E+7 0.33 0.283
DESSUB = 100 $ CONSTRAINT DEFININITION $
SUBCASE 2 FORCE 1 10004 2000.0 1.0 0.0 0.0
$ …
$ Identify stress DRESP1 responses for 2nd subcase using FORCE 2 10203 10000.0 0.0 0.0 1.0
SET and DRSPAN PLOAD2 1 50. 1 THRU 16
$ $
set 22 = 25, 26, 27, 28, 35, 36, 37, 38 SPC1 1 1236 10000
DRSPAN = 22 …
LABEL = LOAD CONDITION 2 SPC1 1 6 10301 10302 10303 10304
LOAD = 2 $
$ PARAM GRDPNT 1
$ Set up a design case for 2nd subcase PARAM WTMASS 0.00259
$ PARAM AUTOSPC YES
DESSUB = 200 $ CONSTRAINT DEFININITION

$------------------------------------------------------------------------
$ DESIGN MODEL:
$------------------------------------------------------------------------
$...Define the design variables:
$DESVAR ID LABEL XINIT XLB XUB DELXV
DESVAR 1 T-PLATE 0.15 0.001 10.0
DESVAR 2 T-BOX 1.0 0.001 10.0
$...Relate the design variables to analysis model properties
$ (linear relations so use DVPREL1)
$...Express shell thicknesses as functions of x1 x2:
$DVPREL1 ID TYPE PID FID PMIN PMAX C0 +
$+ DVIDD1 COEF1 DVID2 COEF2 ...
DVPREL1 1 PSHELL 1 T 0.01 +DP1
+DP1 1 1.0
$...Express BOX THICKNESS as a function of x2:
DVPREL1 3 PBARL 3 DIM2
2 0.1
$. EXPRESS BOX OFFSET LOCATIONS AS A FUNCTION OF PLATE THICKNESS AND
$ FIXED BOX DIMENSIONS
DVCREL1 10 CBAR 31 W3A 1.5
1 0.5
1 0.5
1 0.5
1 0.5
DVCREL1 50 CBAR 31 W3B 1.5
1 0.5
1 0.5
1 0.5
1 0.5
$...Identify the design responses:
$DRESP1 ID LABEL RTYPE PTYPE REGION ATTA ATTB ATT1 +
$+ ATT2 ...
DRESP1 15 W WEIGHT
$ Maximum stress at end A in 1st subcase
DRESP1 21 SBARA STRESS ELEM 7 31

$ Maximum stress at end A in 2nd subcase

$
$
$ Maximum stress at end B in 1st subcase
$
DRESP1 31 SBARB STRESS ELEM 14 31
$
$ Maximum stress at end B in 2nd subcase
$
$
$ define synthetic response with summation of Weight and Bar Stresses
$
DRESP2 101 SAAVRG 101
DRESP1 15 21 22 23 24
25 26
27 28 31 32 33 34
35
36 37 38
DEQATN 101 OBJ(W,S11,S12,S13,S14,S21,S22,S23,S24,S31,S32,S33,
S34,S41,S42,S43,S44) =
10.*W + (S11+S12+S13+S14+S21+S22+S23+S24+
S31+S32+S33+S34+S41+S42+S43+S44)*1.E-3

$
$ Identify von Mises stress/displacement responses
$
DRESP1 3 S13 STRESS PSHELL 9 1
DRESP1 6 S16 STRESS PSHELL 17 1
DRESP1 13 D1 DISP 3 10302
DRESP1 14 D2 DISP 3 10203
$
$...Place bounds on the responses:
$
$DCONSTR DCID RID LALLOW UALLOW
DCONSTR 10 3 25000.
DCONSTR 10 6 25000.
DCONSTR 20 13 -0.1 0.1
DCONSTR 30 14 -0.03 0.03
$
$DCONADD DCID DC1 DC2 ...
DCONADD 100 10 20
DCONADD 200 10 30
$
$...Optional override of optimization parameters:
$
DOPTPRM IPRINT 7 DESMAX 20 DELP 0.5 P1 1 +
+ P2 15
$ (DELP=0.5 allows larger moves thus overcoming constraint
$ violations quicker)
$
ENDDATA



4. View Results
errors.
***************************************************************
***************************************************************


---------------------------------------------------------------------------------------------------------------
---------------------------------------------------------------------------------------------------------------
INITIAL 2.229414E+02 -3.326168E-01
1 1.958472E+02 1.935345E+02 1.195020E-02 -2.657733E-01
2 1.844786E+02 1.840256E+02 2.461796E-03 2.372789E-02
3 1.826974E+02 1.826720E+02 1.393298E-04 5.035938E-04
4 1.818164E+02 1.817764E+02 2.201816E-04 5.689063E-04
5 1.814872E+02 1.814534E+02 1.863480E-04 5.310156E-04
6 1.814494E+02 1.814492E+02 6.727518E-07 5.257812E-04

---------------------------------------------------------------------------------------------------------------


----------------------------------------------------------------------------------------------------------------------------------
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | T-PLATE | 1.5000E-01 : 1.3376E-01 : 1.1109E-01 : 1.1247E-01 : 1.1244E-01 : 1.1242E-01 :
2 | 2 | T-BOX | 1.0000E+00 : 1.5000E+00 : 1.8752E+00 : 2.0221E+00 : 2.1807E+00 : 2.3429E+00 :
----------------------------------------------------------------------------------------------------------------------------------
DV. ID. | DV. ID. | LABEL | 6 : 7 : 8 : 9 : 10 : 11 :
----------------------------------------------------------------------------------------------------------------------------------
1 | 1 | T-PLATE | 1.1242E-01 :
2 | 2 | T-BOX | 2.3621E+00 :
*** USER INFORMATION MESSAGE 4114 (OUTPX2)


Fig. 14. 3: Maximum Constraint History


Design Sensitivity and Optimization in MSC - Nastran

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Design Sensitivity and Optimization in MSC - Nastran

Uploaded by

Copyright:

Available Formats

MSC.

Tokyo, Japan United States Munich, Germany

Design Sensitivity and Optimization in

NAS 107, Workshop 1, May 2005

X Design Variables: Cross-sectional area A1 and A2

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

Figure 1.1 – Geometry and Finite Elements

Elastic Modulus 10e6 psi

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

Show labels and change the view to

Whenever possible, deselect Auto

NAS 107, Workshop 1, May 2005

Create new nodes

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

Define a material using the specified b

NAS 107, Workshop 1, May 2005

Create a 1D rod with aluminum e

NAS 107, Workshop 1, May 2005

Create another property with the new b e

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

g Note: The viewport stays the same.

Use Tools to create the Design

Use Tools to create the Design

NAS 107, Workshop 1, May 2005 g

Create Stress Constraints for the

Create Design Study and set its

Review contents in a Design Study.

NAS 107, Workshop 1, May 2005 g

Generate an input file and set its

Select the Design Study b

NAS 107, Workshop 1, May 2005

Select a Global Objective

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

Generate an input file and sets its

NAS 107, Workshop 1, May 2005

$ Pset: "prop_1" will be imported as: "prod.1"

NAS 107, Workshop 1, May 2005

$ ...STRUCTURAL RESPONSE IDENTIFICATION

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

(HARD CONVERGENCE ACHIEVED)

NUMBER OF FINITE ELEMENT ANALYSES COMPLETED 7

OBJECTIVE AND MAXIMUM CONSTRAINT HISTORY

INITIAL 4.828427E+00 -3.234952E-01

1 2.887955E+00 2.888277E+00 -1.114384E-04 -2.479180E-02

2 2.754460E+00 2.754398E+00 2.250540E-05 -4.326953E-03

3 2.717907E+00 2.717877E+00 1.105302E-05 6.591797E-05

4 2.705185E+00 2.705179E+00 2.203353E-06 1.682617E-04

5 2.697357E+00 2.697340E+00 6.098927E-06 1.448145E-03

6 2.695266E+00 2.695282E+00 -6.015127E-06 1.754883E-03

NAS 107, Workshop 1, May 2005

NAS 107, Workshop 1, May 2005

This ends this exercise.

NAS 107, Workshop 1, May 2005