10524280

Section 5
Model Verification
NAS101, Page 5 - 1
Modeling
PAGE
Overview 5
Singularities and Mechanisms 6
MSC.NASTRAN Automatic Checking Procedure 10
AUTOSPC 11
How AUTOSPC Works 12
Problems with Autospc 14
Autospc with CQUAD4s 18
Autospc Problems 19
Autospc Output 21
Autospc 23
Autospc Workshop 24
NAS101, Page 5 - 2
Modeling (cont.)
PAGE
Diagnosing Errors 34
Model Debugging 35
Debugging Workshop 39
Model Debugging 43
Debugging Workshop 44
Further Model Debugging 49
Analysis Health Checks 53
Essential QA Checks 55
Applied Load Summation 64
Reaction Load Summation 66
Actual Applied Load Check 67
NAS101, Page 5 - 3
Modeling (cont.)
PAGE
Essential QA Checks – Workshop 71

Good Modeling Practice 72
MultiPoint Constraints, Rigid Elements 80
Multipoint Constraints 81
Multipoint Constraints Workshop 84
R-type Elements 85
RIGID Elements 86
RIGID Elements Workshop 97
RIGID Elements 98
RIGID Elements Workshop 101
NAS101, Page 5 - 4
Modeling (cont.)
PAGE
Useful Model Summary Information 102

ELSUM 103
MAX/MIN DISPLACEMENT AND SPCFORCE OUTPUT 105
Element Geometry Checks 109
GEOMCHECK 111
Element Geometry Checks 113
Structural Symmetry 114
NAS101, Page 5 - 5
Overview
Typical errors; Singularities and Mechanisms

MSC.NASTRAN’s automatic checking procedure
Diagnosing errors
Health checks the user should be making
RIGID elements and MPC’s
Element Geometry Checks
Symmetry
NAS101, Page 5 - 6
Typical errors; Singularities and Mechanisms
NAS101, Page 5 - 7
Singularities and Mechanisms
A singularity is caused by a degree of freedom without
any structural stiffness present, or with a low stiffness
The Stiffness Matrix cannot be inverted if it is Singular
 Some examples of singularities are:

Forgetting to eliminate Rigid Body Motions
Connecting Dissimilar Element Types
Cross connecting wrong Degrees of Freedom
NAS101, Page 5 - 8
(cont.)
Rigid Body Motions
 Sufficient Nodal Displacement should be specified so that the 6
“rigid body” modes of movement are fixed.
 Rigid Body Motion Adequate Constraints
NAS101, Page 5 - 9
(cont.)
Rigid Body Motions
 Forgetting to ‘Equivalence’ in MSC.PATRAN or other

preprocessors is a very common error.
 Results in unconnected meshes.
NAS101, Page 5 - 10
MSC.NASTRAN’s automatic checking procedure
NAS101, Page 5 - 11
AUTOSPC
If obvious singularities exist, MSC/NASTRAN attempts

to automatically deal with them.
 The Bulk Data entry PARAM,AUTOSPC,YES instructs

the program to automatically apply SPCs to these
singularities.
PARAM, AUTOSPC, YES is the default in most of the

structured solutions.
NAS101, Page 5 - 12
How Autospc works
GRID 99 Stiffness Terms
T1
T2
Hexa Element
T3
GRID 99
R1
R2
R3
NAS101, Page 5 - 13
How Autospc works (cont.)
GRID 99 Stiffness Terms
Successful Elimination of
T1
Zero Stiffness terms
T2
Hexa Element
T3
GRID 99
R1
R2
R3
NAS101, Page 5 - 14
Problems with Autospc
Solid Bar
 No Elimination of Solid
Element Zero Stiffness terms
T1 T1
Bar Element T2 T2
GRID 99
T3 T3
R1 R1
Hexa Element R2 R2
R3 R3
NAS101, Page 5 - 15
Problems with Autospc (cont.)
 No Elimination of Solid Combined Stiffness Terms
Element Zero Stiffness terms T1
Bar Element T2
T3
GRID 99
R1
R2
Hexa Element
R3
NAS101, Page 5 - 16
 3 Mechanisms !
Bar Element
GRID 99
Hexa Element
NAS101, Page 5 - 17
Solutions
Manual SPC Bar Element
MPC’s (later)
GRID 99
Rigid Links (later)
Hexa Element
NAS101, Page 5 - 18
Autospc with CQUAD4’s
GRID 106 Stiffness
 CQUAD4 T1
R3 = 0.0
T2
T3
T3
GRID 106
R1
T1,
R1 R2
T2, R3
R2
NAS101, Page 5 - 19
Autospc Problems
GRID 106 Stiffness
 2 CQUAD4’s T1 T1
T3 T2 T2
T3 T3
R1 R1
R2 R2
Possible Mechanisms ! R3 ? R3 ?
NAS101, Page 5 - 20
Autospc Problems (cont.)
Solutions
PARAM,K6ROT, kvalue
Not Recommended – for Non Linear usage
PARAM,SNORM, angle
Recommended
2 * angle
All vectors within ‘angle’
are averaged
NAS101, Page 5 - 21
Autospc Output
Output includes a grid point singularity table. This table
should be inspected carefully for potential singularities.
Default stiffness
ratio = 1.0E-8
GRID POINT SINGULARITY TABLE

POINT TYPE FAILED STIFFNESS OLD USET NEW USET
ID DIRECTION RATIO EXCLUSIVE UNION EXCLUSIVE UNION
1 G 6 0.00E+00 B F SB SB
2 G 6 0.00E+00 B F SB SB
3 G 6 0.00E+00 B F SB SB
4 G 6 0.00E+00 B F SB SB
NAS101, Page 5 - 22
Autospc Output (cont.)
 What does the USET term mean?
 Consider all grid point and scalar point degrees of freedom in a finite element model as
the members of a single displacement set. This global set is called the g-set; the
displacement set is known as ug.
 [Kgg] must be nonsingular in order to solve this equation. To achieve a nonsingular

stiffness matrix, the user can specify the independent subsets of {ug} to be partitioned
out during matrix reduction.
For example:
um Degrees of freedom eliminated by multipoint constraints
us Degrees of freedom eliminated by single-point constraints
 Elimination of the M and S sets results in the F (free) set, which is typically solved to
obtain the unknown displacements,
 For a thorough discussion of constraint and partitioned displacement sets, see Appendix B of the
MSC.Nastran Quick Reference Guide and MSC.NASTRAN Linear Static Analysis Users’ Guide.
NAS101, Page 5 - 23
Autospc
Controlling AUTOSPC
 All ‘failed’ DOF are written to Grid Point Singularity Table
 Can get very large – obscure real problems
Write out ‘failed’ DOF’s to .pch File

 Param,spcgen,1 param,checkout,yes
Re-use generated SPC1 data selectively
NAS101, Page 5 - 24
Autospc Workshop
 Run the MSC.NASTRAN input files
 section5_1.bdf set of solid elements
 section5_2.bdf set of plate elements
 Evaluate the Grid Point Singularity Tables
 Run the MSC.NASTRAN input files

 Section5_3.bdf solid/plate combination
 Section5_4.bdf plate/bar combination
 Evaluate the Tables and check the Fatal Messages
NAS101, Page 5 - 25
Autospc Workshop (cont.)
 section5_1.bdf
NAS101, Page 5 - 26
 section5_1.bdf

GRID POINT SINGULARITY TABLE
0 POINT TYPE FAILED STIFFNESS OLD USET NEW USET
1 G 4 0.00E+00 BF F SB S *
1 G 5 0.00E+00 BF F SB S *
1 G 6 0.00E+00 BF F SB S *
GRID 1
2 G 4 0.00E+00 BF F SB S *
2 G 5 0.00E+00 BF F SB S *
2 G 6 0.00E+00 BF F SB S *
3 G 4 0.00E+00 BF F SB S *
3 G 5 0.00E+00 BF F SB S *
3 G 6 0.00E+00 BF F SB S *
NAS101, Page 5 - 27
 section5_2.bdf
NAS101, Page 5 - 28
 section5_2.bdf
0 GRID POINT SINGULARITY TABLE

0 POINT TYPE FAILED STIFFNESS OLD USET NEW USET
2 G 6 0.00E+00 BF F SB S *
3 G 6 0.00E+00 BF F SB S *
4 G 6 0.00E+00 BF F SB S *
5 G 6 0.00E+00 BF F SB S *
6 G 6 0.00E+00 BF F SB S * R3 = 0.0
7 G 6 0.00E+00 BF F SB S *
8 G 6 0.00E+00 BF F SB S *
9 G 6 0.00E+00 BF F SB S *
10 G 6 0.00E+00 BF F SB S *
NAS101, Page 5 - 29
 section5_3.bdf
CQUAD4’s
HEXA’s
NAS101, Page 5 - 30
 section5_3.bdf
1 G 6 0.00E+00 BF F SB S *
2 G 6 0.00E+00 BF F SB S *
GRID 13 3 G 6 0.00E+00 BF F SB S *
4 G 6 0.00E+00 BF F SB S *
What
happens
62 G 4 0.00E+00 BF F SB S * along here !!
62 G 5 0.00E+00 BF F SB S *
62 G 6 0.00E+00 BF F SB S *
63 G 4 0.00E+00 BF F SB S *
63 G 5 0.00E+00 BF F SB S *
63 G 6 0.00E+00 BF F SB S *
64 G 4 0.00E+00 BF F SB S *
64 G 5 0.00E+00 BF F SB S *
64 G 6 0.00E+00 BF F SB S *
NAS101, Page 5 - 31
 section5_3.bdf
THE FOLLOWING DEGREES OF FREEDOM HAVE FACTOR DIAGONAL RATIOS GREATER THAN
1.00000E+07 OR HAVE NEGATIVE TERMS ON THE FACTOR DIAGONAL.
0 SUBCASE 1
0
GRID POINT ID DEGREE OF FREEDOM MATRIX/FACTOR DIAGONAL RATIO MATRIX DIAGONAL
13 R2 1.71146E+13 2.13419E+02
^^^ USER FATAL MESSAGE 9050 (SEKRRS)
^^^ RUN TERMINATED DUE TO EXCESSIVE PIVOT RATIOS IN MATRIX KLL.
^^^ USER ACTION: CONSTRAIN MECHANISMS WITH SPCI OR SUPORTI ENTRIES OR SPECIFY PARAM,BAILOUT,-1 TO
CONTINUE THE RUN WITH MECHANISMS.
NAS101, Page 5 - 32
 section5_4.bdf
1 G 6 0.00E+00 BF F SB S *
2 G 6 0.00E+00 BF F SB S *
3 G 6 0.00E+00 BF F SB S *
4 G 6 0.00E+00 BF F SB S *
10 G 6 0.00E+00 BF F SB S *
12 G 6 0.00E+00 BF F SB S *
14 G 6 0.00E+00 BF F SB S *
15 G 6 0.00E+00 BF F SB S *
No Grid Point Singularity terms for GRID 13 What happens

here !!
NAS101, Page 5 - 33
 section5_4.bdf
0 SUBCASE 1
0
13 R3 -7.02842E+15 4.09108E+05
^^^ USER ACTION: CONSTRAIN MECHANISMS WITH SPCI OR SUPORTI ENTRIES OR SPECIFY PARAM,BAILOUT,-1 TO
NAS101, Page 5 - 34
Diagnosing errors
NAS101, Page 5 - 35
Model Debugging
The previous workshop examples
 Section5_3.bdf
 Section5_4.bdf
 Resulted in FATAL errors due to the presence of

mechanisms.
 This section shows how to identify the type of error.
NAS101, Page 5 - 36
Model Debugging (cont.)
 from file section5_3.f06
 This is the most common FATAL ERROR
 Check the FATAL MESSAGE ID
 Check the Description
0 SUBCASE 1
0
GRID POINT ID DEGREE OF FREEDOM MATRIX/FACTOR DIAGONAL RATIO MATRIX
13 R2 1.71146E+13 2.13419E+02
^^^ USER ACTION: CONSTRAIN MECHANISMS WITH SPCI OR SUPORTI ENTRIES OR SPECIFY
PARAM,BAILOUT,-1 TO CONTINUE THE RUN WITH MECHANISMS.
NAS101, Page 5 - 37
 A Singularity or Mechanism is indicated
 It is seen at GRID 13, DOF R2
 Is anything special about this point?
0 SUBCASE 1
0
13 R2 1.71146E+13 2.13419E+02
PARAM,BAILOUT,-1 TO CONTINUE THE RUN WITH MECHANISMS.
NAS101, Page 5 - 38
 A Singularity or Mechanism is indicated
 It is seen at GRID 13, DOF R3
 Is anything special about this point?
0 SUBCASE 1
0
13 R3 -7.02842E+15 4.09108E+05
PARAM,BAILOUT,-1 TO
NAS101, Page 5 - 39
Debugging Workshop
 Using SPC or SPC1 entries,

 Correct the MSC.NASTRAN input files
 Section5_3.bdf (hint: GRIDS 1 5 9 13 form the joint)
 Section5_4.bdf
 Evaluate the results and the implication of the changes
NAS101, Page 5 - 40
Debugging Workshop (cont.)
 Corrections to Section5_3.bdf
Constrained DOF 4,5,6
Looks OK, but be careful!!
NAS101, Page 5 - 41
The displacement of the solid section implies small edge
rotations of connecting shells– we have constrained them
out
NAS101, Page 5 - 42
No complications seen
Constrained DOF 6
NAS101, Page 5 - 43
Model Debugging
So far we have seen two examples of FATAL errors
caused by excessive pivot ratios during an analysis
 Fatal Message 9050
In practice there are many checks that MSC.NASTRAN will

perform to flag errors prior to analysis in syntax checking
and general data checking, as well as during the analysis
The form of the FATAL message will be similar. The key

ingredients are the ID and the description
Further details on the meaning of the FATAL message can

be found in the Reference Guide, or more conveniently,
the On Line Encyclopedia
NAS101, Page 5 - 44
Debugging Workshop
 Run Section5_5.bdf
 Evaluate the messages and correct the analysis
NAS101, Page 5 - 45
 Section5_5.f06
 What is characteristic about the elements in the group
that are subject to warning?
 What does the WARNING message imply?
*** USER WARNING MESSAGE 5487 (EBCHKD)
ORIENTATION VECTOR DEFINED FOR THE ELEMENT ID = 7 IS NEARLY PARALLEL, IT MAY GIVE POOR RESULTS.
NAS101, Page 5 - 46
 Section5_5.f06
 Again, look at the group causing the FATAL error
 Look up the message ID in the On Line Encyclopedia
*** USER FATAL MESSAGE 2026 (EMG)
ELEMENT 7 GEOMETRY YIELDS UNREASONABLE MATRIX.
NAS101, Page 5 - 47
 On Line Encyclopedia – result of search for ‘2026’

2026 *** USER FATAL MESSAGE 2026, ELEMENT **** GEOMETRY OR
MATERIAL PROPERTY YIELDS UNREASONABLE MATRIX.
Referenced element geometry and/or properties yield a numerical result which
causes an element stiffness or mass matrix to be undefined. Possible causes
include, but are not limited to, (1) the length of a rod or bar is zero because the
end points have the same coordinates, (2) the sides of a triangle or
quadrilateral are collinear which leads to a zero cross product in defining an
element coordinate system, (3) the bar orientation vector is parallel to the bar
axis, or (4) a shear panel has zero thicknesss or modulus. Check GRID Bulk
Data entries defining element end points for bad data.
NAS101, Page 5 - 48
Debugging Workshop
 Section5_5.bdf Correct the bar data and rerun

$ bar elements follow
CBAR 1 1 1 2 0. 1. 0.
CBAR 2 1 2 3 0. 1. 0.
CBAR 3 1 3 4 0. 1. 0.
CBAR 4 1 4 5 0. 1. 0.
CBAR 5 1 5 6 0. 1. 0.
CBAR 6 1 6 7 0. 1. 0.
CBAR 7 1 8 9 0. 1. 0.
CBAR 8 1 9 10 0. 1. 0.
CBAR 9 1 10 12 0. 1. 0.
CBAR 10 1 12 1 0. 1. 0.
CBAR 11 1 14 15 0. 1. 0.
CBAR 12 1 15 16 0. 1. 0.
CBAR 13 1 16 18 0. 1. 0.
CBAR 14 1 18 7 0. 1. 0.
CBAR 15 1 10 21 0. 1. 0.
CBAR 16 1 21 22 0. 1. 0.
CBAR 17 1 22 23 0. 1. 0.
CBAR 18 1 23 24 0. 1. 0.
CBAR 19 1 24 25 0. 1. 0.
CBAR 20 1 25 16 0. 1. 0.
NAS101, Page 5 - 49
Further Model Debugging
 To enable deeper understanding of the files that can be
used for debugging the concept of DMAP needs to be
introduced
DMAP – Direct Matrix Abstraction Procedure
 The High Level Language that MSC.NASTRAN is written in
 Completely visible to Users
 Modifiable by Users
DMAP modules perform the mathematical operations

required to perform the requested solution sequence.
NAS101, Page 5 - 50
Further Model Debugging (cont.)
 Schematic of DMAP – Direct Matrix Abstraction Procedure
Solution Sequence DMAP Modules Source Code
Interpret Input
Syntax Check
Partition SPC’s
Partition MPC’s
NAS101, Page 5 - 51
 .f04 File Description
 The ".f04" file contains a "road-map" of the MSC/NASTRAN

modules used in the solution of your problem.
 Each time a DMAP module is executed, a line is created in the .f04

file contain in the module name and information on the current
state of the run.
NAS101, Page 5 - 52
 The cause of the error was clear in the portal frame example. If the
messages do not provide enough explanation, the user should also
check the .f04 file to determine at what point during data processing
the analysis terminated.
 In Practice, understanding the DMAP Module flow may be difficult, but

it will provide valuable information to MSC Support.
10:33:06 0:00 33.0 0.0 0.6 0.0 PHASE1DR 104 (S)DBSETOFF BEGN
10:33:06 0:00 33.0 0.0 0.6 0.0 PHASE1DR 106 (S)PHASE1A BEGN
10:33:06 0:00 33.0 0.0 0.7 0.0 PHASE1A 42 TA1 BEGN SubDMAPs
10:33:06 0:00 34.0 1.0 0.7 0.0 PHASE1A 51 MSGHAN BEGN *
10:33:06 0:00 34.0 0.0 0.7 0.0 PHASE1A 52 (S)SEMG BEGN
10:33:06 0:00 34.0 0.0 0.7 0.0 SEMG 22 ELTPRT BEGN
10:33:06 0:00 34.0 0.0 0.7 0.0 SEMG 28 EMG BEGN
10:33:06 0:00 34.0 0.0 0.7 0.0 SEMG 36 (S)ERRPH1 BEGN
DMAP MODULES
10:33:06 0:00 34.0 0.0 0.7 0.0 ERRPH1 19 (S)PRTSUM BEGN
10:33:06 0:00 34.0 0.0 0.7 0.0 PRTSUM 24 PROJVER BEGN
10:33:06 0:00 34.0 0.0 0.7 0.0 PRTSUM 25 DBDICT BEGN
10:33:06 0:00 34.0 0.0 0.7 0.0 PRTSUM 26 PRTPARM BEGN
10:33:06 0:00 34.0 0.0 0.7 0.0 ERRPH1 20 EXIT BEGN
10:33:06 0:00 34.0 0.0 0.7 0.0 XSEMDR END Analysis stops here
NAS101, Page 5 - 53
Analysis Health Checks
NAS101, Page 5 - 54
Analysis Health Checks
 In the previous section we discussed the CURE:
 Debugging
 In this section we discuss PREVENTION

 Essential QA checks
 Good Modeling Practice
NAS101, Page 5 - 55
Essential QA checks
Pre Analysis
 Element Distortion
• Use your pre-processor to visually check distortion
• Use WARNING messages in .f06 to confirm
 Consistent Units
• Check with Force = Mass * Acceleration
NAS101, Page 5 - 56
Essential QA checks (cont.)
Element Distortions
 Aspect ratio a
a
b
b
• Aspect ratio should be less than about 4:1 (much less in regions
where stress levels change rapidly). In cases of nearly-uniaxial stress
fields, larger aspect ratios are acceptable.
NAS101, Page 5 - 57
 Skew Check
• Quadrilateral elements should be kept as square as possible
NAS101, Page 5 - 58
 Taper Check
NAS101, Page 5 - 59
 Warp
• Up to ~ 5% is normally acceptable. No real limit, but the element does

not include warpage.
h
a
h
NAS101, Page 5 - 60
Post Analysis
 Epsilon Value
 Applied Load Summation
 Reaction Summation
 Strain Energy Values
 Peak Deflections
NAS101, Page 5 - 61
 Post Analysis - Epsilon Value
Standard Solution Expression P  K δ
Assumes no Rounding Errors P  K δ  0
In Practice there is a Residual

P  K  δ  P*
(This on its own has dimensions)
Make it an Energy Term δT  P *


1 PT  δ
Compare it to the System Energy
2
NAS101, Page 5 - 62
 If 10-6 ,or greater ,it could be a sign of ill-conditioning.
 For your type of Structure, Idealization and Analysis
 Establish typical values for Epsilon
 Establish an agreed threshold
*** USER INFORMATION MESSAGE 5293 (SSG3A)

FOR DATA BLOCK KLL
LOAD SEQ. NO. EPSILON EXTERNAL WORK EPSILONS LARGER THAN 0.001 ARE FLAGGED WITH
ASTERISKS
1 -1.3760919E-13 3.6560133E+04
NAS101, Page 5 - 63
 Post Analysis - Applied Load Summation
 Use OLOAD request in Case Control
 Particularly Important For:
 Inertia Loading
 Complex Pressure Loading
 Complex Distributed Loading
NAS101, Page 5 - 64
Applied Load Summation
0 RESULTANTS ABOUT ORIGIN OF SUPERELEMENT BASIC COORDINATE SYSTEM IN SUPERELEMENT BASIC SYSTEM COORDINATES.
0 OLOAD RESULTANT
SUBCASE/ LOAD
DAREA ID TYPE T1 T2 T3 R1 R2 R3
0 1 FX -3.900000E+03 ---- ---- ---- 0.000000E+00 3.744000E+05
FY ---- -4.500000E+03 ---- 0.000000E+00 ---- -1.296000E+06
FZ ---- ---- 0.000000E+00 0.000000E+00 0.000000E+00 ----
MX ---- ---- ---- 0.000000E+00 ---- ----
MY ---- ---- ---- ---- 0.000000E+00 ----
MZ ---- ---- ---- ---- ---- 0.000000E+00
TOTALS -3.900000E+03 -4.500000E+03 0.000000E+00 0.000000E+00 0.000000E+00 -9.216000E+05
0 20 FX 0.000000E+00 ---- ---- ---- 0.000000E+00 0.000000E+00
FY ---- -1.818989E-12 ---- 0.000000E+00 ---- 9.313226E-10
FZ ---- ---- 0.000000E+00 0.000000E+00 0.000000E+00 ----
MX ---- ---- ---- 0.000000E+00 ---- ----
MY ---- ---- ---- ---- 0.000000E+00 ----
MZ ---- ---- ---- ---- ---- 0.000000E+00
TOTALS 0.000000E+00 -1.818989E-12 0.000000E+00 0.000000E+00 0.000000E+00 9.313226E-10
0 30 FX 0.000000E+00 ---- ---- ---- 0.000000E+00 0.000000E+00
FY ---- -2.123938E+03 ---- 0.000000E+00 ---- -6.116941E+05
FZ ---- ---- 0.000000E+00 0.000000E+00 0.000000E+00 ----
MX ---- ---- ---- 0.000000E+00 ---- ----
MY ---- ---- ---- ---- 0.000000E+00 ----
MZ ---- ---- ---- ---- ---- 0.000000E+00
TOTALS 0.000000E+00 -2.123938E+03 0.000000E+00 0.000000E+00 0.000000E+00 -6.116941E+05
NAS101, Page 5 - 65
 Post Analysis - Reaction Summation
 Check that sense opposes and balances OLOAD summary
NAS101, Page 5 - 66
Reaction Load Summation
RESULTANTS ABOUT ORIGIN OF SUPERELEMENT BASIC COORDINATE SYSTEM IN SUPERELEMENT BASIC SYSTEM COORDINATES.
0 SPCFORCE RESULTANT
SUBCASE/ LOAD
DAREA ID TYPE T1 T2 T3 R1 R2 R3
0 1 FX 3.900000E+03 ---- ---- ---- 0.000000E+00 0.000000E+00
FY ---- 4.500000E+03 ---- 0.000000E+00 ---- 9.216000E+05
FZ ---- ---- 0.000000E+00 0.000000E+00 0.000000E+00 ----
MX ---- ---- ---- 0.000000E+00 ---- ----
MY ---- ---- ---- ---- 0.000000E+00 ----
MZ ---- ---- ---- ---- ---- 0.000000E+00
TOTALS 3.900000E+03 4.500000E+03 0.000000E+00 0.000000E+00 0.000000E+00 9.216000E+05
0 20 FX 0.000000E+00 ---- ---- ---- 0.000000E+00 0.000000E+00
FY ---- 7.275958E-12 ---- 0.000000E+00 ---- 2.095476E-09
FZ ---- ---- 0.000000E+00 0.000000E+00 0.000000E+00 ----
MX ---- ---- ---- 0.000000E+00 ---- ----
MY ---- ---- ---- ---- 0.000000E+00 ----
MZ ---- ---- ---- ---- ---- 0.000000E+00
TOTALS 0.000000E+00 7.275958E-12 0.000000E+00 0.000000E+00 0.000000E+00 2.095476E-09
0 30 FX -3.410605E-12 ---- ---- ---- 0.000000E+00 0.000000E+00
FY ---- 2.123938E+03 ---- 0.000000E+00 ---- 6.116941E+05
FZ ---- ---- 0.000000E+00 0.000000E+00 0.000000E+00 ----
MX ---- ---- ---- 0.000000E+00 ---- ----
MY ---- ---- ---- ---- 0.000000E+00 ----
MZ ---- ---- ---- ---- ---- 0.000000E+00
TOTALS -3.410605E-12 2.123938E+03 0.000000E+00 0.000000E+00 0.000000E+00 6.116941E+05
NAS101, Page 5 - 67
Actual Applied Load Check
Request printout of applied loads at actual load

application points
OLOAD = n
May produce large output for certain type loadings

 e.g., gravity load on a large model
NAS101, Page 5 - 68
Actual Applied Load Check (cont.)
NAS101, Page 5 - 69
 Post Analysis - Strain Energy Values
*** USER INFORMATION MESSAGE 5293 (SSG3A)

FOR DATA BLOCK KLL
LOAD SEQ. NO. EPSILON EXTERNAL WORK EPSILONS LARGER THAN 0.001 ARE FLAGGED WITH
ASTERISKS
1 -1.3760919E-13 3.6560133E+04
Work = ½ Total Force * Total Deflection

= ( approx) ½ OLOAD * Peak Deflection
( if peak deflection is near mean line of loading action)
NAS101, Page 5 - 70
 Post Analysis - Peak Deflections
 Set PARAM,PRTMAXIM,YES to output this
 The GRID ID is not included, and can be different for each DOF
0
0 MAXIMUM DISPLACEMENTS
0 T1 T2 T3 R1 R2 R3
0 1 3.0938861E-07 4.1483727E-08 3.6560131E+01 7.2180829E+00 5.6827263E+01 0.0000000E+00
Value !!! Work = ( approx) ½ OLOAD * Peak Deflection

( 2e3 * 36.5 *.5 = 36.5e3 )
NAS101, Page 5 - 71
Essential QA checks - Workshop
 Run section5_6.bdf
 Go through the Essential QA checks:
• Epsilon Value
• Applied Load Summation
• Reaction Summation
• Strain Energy Values
• Peak Deflections
NAS101, Page 5 - 72
Good Modeling Practice
NAS101, Page 5 - 73
Good Modeling Practice
 Essentials
 Mesh Density – fit for purpose
 Mesh Quality – fit for purpose
 Loading Boundary Conditions
 Displacement Boundary Conditions
NAS101, Page 5 - 74
Good Modeling Practice (cont.)
 Mesh Density – fit for purpose

NAS101, Page 5 - 75
 Mesh Quality – fit for purpose

NAS101, Page 5 - 76
 Loading Boundary Conditions
 Simple Point Loading?
Poor stress
Good stress distribution locally
distribution locally
NAS101, Page 5 - 77
Loading Boundary Conditions
 More sophisticated loading?
NAS101, Page 5 - 78
Displacement Constraint Considerations
 Wrongly defining Output Coordinate Systems at SPC’s, MPC’s,

RIGID Elements etc. can wreck a model
 Over constraining a model can give rise to Poisson contraction

stresses that distort the true stress field.
 A single nodal constraint (or force) produces a singularity in the

stress field - stress results at this point are likely to be erroneous.
NAS101, Page 5 - 79
 Displacement Constraint Considerations
 A special technique called inertia relief is available to perform quasi-static

analysis on unconstrained (free) structures under uniform (i.e., zero or
constant) acceleration.
Aero Loads
Inertia Loads
NAS101, Page 5 - 80
MultiPoint Constraints, Rigid Elements
NAS101, Page 5 - 81
Multipoint Constraints
 A multipoint constraint (MPC) is a user-imposed linear equation that relates
displacement degrees of freedom.
 MPCs are useful to
 Define the relative motion between two or more grid points as a degree of
freedom
 Join dissimilar elements; for example, to join elements with rotational

degrees of freedom to elements which have only translational degrees of
freedom (e.g., to join shell elements to solid elements)
 Distribute loads to several points in a structure
 Model rigid connections between grid points
NAS101, Page 5 - 82
Multipoint Constraints (cont.)
 Assume GRID 145 and 146 are forced to move together in
x and y ( a bracket may connect them)
-1.0*Ux145 + 1.0*Ux146 = 0.0
146 -1.0*Uy145 + 1.0*Uy146 = 0.0
145
General Form is S ai*Ui= 0.0
where a = constraint coefficient
u = displacement degree of
freedom
NAS101, Page 5 - 83
Multipoint Constraints (cont.)
Case Control
SUBCASE 1
SUBTITLE=edge
MPC = 1
SPC = 2
LOAD = 2
…….
Bulk Data
$ SID GRID DOF A1 GRID DOF A2

MPC 1 145 1 -1. 146 1 1.
MPC 1 145 2 -1. 146 2 1.
 The first component (C) defined in the equation is considered to be

the dependent coordinate and is placed in the Um set. This component
cannot belong to any other subset of ug.
NAS101, Page 5 - 84
Multipoint Constraints Workshop
 Rework section5_4.bdf
 Use an MPC to remove the singularity, rather than an SPC
0.25
0.25
NAS101, Page 5 - 85
R-TYPE ELEMENTS
MSC.Nastran contains several commonly used MPC

relationships defined in the form of various R-type elements.
To avoid possible errors, it is strongly recommended that the
user who is unfamiliar with writing MPC equations use rigid
elements whenever possible.
Unlike MPCs, R-type elements are not selected in the Case

Control. They are defined only in the Bulk Data on the
following entries:
NAS101, Page 5 - 86
RIGID ELEMENTS
RBAR - Rigid bar with six degrees of freedom at
each end
RBE2 - A rigid body connected to an arbitrary
number of grid points
RBE3 - Defines a constraint relation in which the
motion at a “reference” grid point is the
weighted average of the motions at other
grid points
NAS101, Page 5 - 87
RIGID ELEMENTS (cont.)
RSPLINE Defines a constraint relation whose coefficients
are derived from the deflections and slopes of a flexible
tubular beam connected to the referenced grid points
RSSCON Used to connect plate elements to solid

elements
See Section 2.10 of the MSC/NASTRAN Application

Manual for 10 examples that use rigid elements and two
examples that use MPCs.
NAS101, Page 5 - 88
 RBAR Example
 The explicit MPC in section5_4.bdf could be replaced by an RBAR

 Internally MPC equations are created
NAS101, Page 5 - 89
RBAR EXAMPLE
 An RBAR is preferred over creating a ‘stiff beam’ as
it has no relative stiffness side effects
$ RBAR Creation
$23456781234567812345678123456781234567812345678
$ ID GRID1 GRID2 DOF1 DOF2
RBAR 21 13 18 123456 123456
NAS101, Page 5 - 90
RBE2 Example
The explicit MPC in section5_4.bdf could be replaced
by an RBE2
 Internally MPC equations are created
 Acts like a rigid ‘spider’
NAS101, Page 5 - 91
RBE2 Example
 The DOF at the centre of the ‘spider’ is the Independent DOF
 The others are Dependent DOF’s and must not be cross coupled
NAS101, Page 5 - 92
RBE2 Example
RBE2 method SPC method
RBE2 constrains the ‘backing structure’
NAS101, Page 5 - 93
RBE2 Examples
 Connect in a component approximation

• Engine Block
• Antennae Dish
 Connect different mesh regions

• Detailed fitting in a shaft
NAS101, Page 5 - 94
RBE3 Example
 The explicit MPC in section5_4.bdf could be replaced
by an RBE3
 Mathematically very complex – a single dependent
moves as a average of multi independents
 Acts like a soft ‘spider’
NAS101, Page 5 - 95
RBE3 Example
 The DOF at the centre of the ‘spider’ is the Dependent DOF
 The others are Independent DOF’s and can be cross coupled
NAS101, Page 5 - 96
RBE3 Examples
Connect in a component approximation

•Antennae Dish – flexible backing structure
 Connect different mesh regions

• Fuselage beam to shell – flexible ovalisation of
fuselage
 Connect Payload
• Smears payload to required attachments
NAS101, Page 5 - 97
RIGID ELEMENTS Workshop
Use file section5_3.bdf
 Modify to try out:
• RBAR
• RBE2
• RBE3
 Compare the displacement distributions
NAS101, Page 5 - 98
RIGID ELEMENTS
RSSCON Example
Solid to Shell Connection
Writes MPC’s
NAS101, Page 5 - 99
RSSCON Example – Element Method
NAS101, Page 5 - 100

RSSCON Example – GRID Method
109
102 47
46 108
101
RSSCON,110,GRID,46,101,102,47,108,109
NAS101, Page 5 - 101

RIGID ELEMENTS Workshop
Use file section5_4.bdf
 Modify to try out:
• RSSCON – GRID
• RSSCON - ELEMENT
 Compare the displacement distributions
NAS101, Page 5 - 102

Useful Model Summary Information
NAS101, Page 5 - 103

ELSUM
The ELSUM Case Control command provides a summary
description of the requested elements
Output includes
 Element ID
 Material ID
 Length or thickness
 Area
 Volume
 Structural Mass
 Non-Structural Mass
 Total Mass
 Total Weight
NAS101, Page 5 - 104

ELSUM
Format:
ELSUM = I
Where I – a set number or ‘ALL’
Limitations:
Mass data is produced only for CBAR, CBEAM, CBEND,
CHEXA, CONROD, CPENTA, CQUAD4, CQUAD8, CQUADR,
CROD, CSHEAR, CTETRA, CTRIAR, CTRIA3, CTRIAX6,
and CTUBE elements
NAS101, Page 5 - 105

MAX/MIN DISPLACEMENT AND
SPCFORCE OUTPUT
In SOL 101, there is an option to obtain MAX/MIN output
summaries by SUBCASE for displacements and SPC
forces
If requested, this output is in addition to the standard
output
NAS101, Page 5 - 106

NAS101, Page 5 - 107
SPCFORCE OUTPUT
$ file maxmin.dat GRID 1 0 0.0 0.0 0.0 0
sol 101 GRID 2 0 12.5 0.0 0.0 0
cend GRID 3 0 25. 0.0 0.0 0
title = cantilever beam model GRID 4 0 37.5 0.0 0.0 0
subtitle = OLOAD OUTPUT GRID 5 0 50. 0.0 0.0 0
spc = 1 GRID 6 0 62.5 0.0 0.0 0
disp=all GRID 7 0 75. 0.0 0.0 0
maxmin(vmag=2,disp,spcf)=all GRID 8 0 87.5 0.0 0.0 0
subcase 1 GRID 9 0 100. 0.0 0.0 1
label = pload1 $
load = 1 CBEAM 1 1 1 2 1.
subcase 2 CBEAM 2 1 2 3 1.
label = load in x, y, and z CBEAM 3 1 3 4 1.
load = 2 CBEAM 4 1 4 5 1.
begin bulk CBEAM 5 1 5 6 1.
pload1,1,1,fy,fr,0.,1.,1.,1. CBEAM 6 1 6 7 1.
=,=,*(1),== CBEAM 7 1 7 8 1.
=(6) CBEAM 8 1 8 9 1.
force,2,9,,1.,1.,1.,1. $
PARAM GRDPNT 0 SPC 1 1 123456 0.0
PARAM POST -1 $
$ PBEAML 1 1 BAR
cord2r,1,,0.,0.,0.,0.,1.,0. 1. 2.
,1.,0.,1. $
MAT1 1 1.+7 .3 .1
GRID point 9 uses CORD2R 1 for output $
ENDDATA
NAS101, Page 5 - 108

SPCFORCE OUTPUT
0 *** T1 *** D I S P L A C E M E N T M A X / M I N V A L U E S U M M A R Y RESULTS FOR SUBCASE 1
MAXMIN OPTIONS: SET=ALL, CID=BASIC, VMAG=2, VMAG=2, COMP=T1
POINT ID. TYPE CID ***TMAG*** T2 T3 R1 R2 R3
1 G BASIC 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
2 G BASIC 5.404634E-02 0.000000E+00 0.000000E+00 8.251953E-03 0.000000E+00 0.000000E+00
8 G BASIC 1.563420E+00 0.000000E+00 0.000000E+00 2.495117E-02 0.000000E+00 0.000000E+00
9 G BASIC 1.875780E+00 0.000000E+00 0.000000E+00 2.500000E-02 0.000000E+00 0.000000E+00
9 G 1 0.000000E+00 0.000000E+00 1.875780E+00 1.767767E-02 -1.767767E-02 0.000000E+00
1 CANTILEVER BEAM MODEL FEBRUARY 13, 2001 MSC.NASTRAN 1/17/01 PAGE 12

OLOAD OUTPUT
0
Maximum Translations
0 *** R1 *** D I S P L A C E M E N T M A X / M I N V A L U E S U M M A R Y RESULTS FOR SUBCASE 1

MAXMIN OPTIONS: SET=ALL, CID=BASIC, VMAG=2, VMAG=2, COMP=R1
POINT ID. TYPE CID T1 T2 T3 ***RMAG*** R2 R3
1 G BASIC 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
2 G BASIC 5.404634E-02 0.000000E+00 0.000000E+00 8.251953E-03 0.000000E+00 0.000000E+00
8 G BASIC 1.563420E+00 0.000000E+00 0.000000E+00 2.495117E-02 0.000000E+00 0.000000E+00
9 G BASIC 1.875780E+00 0.000000E+00 0.000000E+00 2.500000E-02 0.000000E+00 0.000000E+00
9 G 1 0.000000E+00 0.000000E+00 1.875780E+00 1.767767E-02 -1.767767E-02 0.000000E+00
Displacements in
Maximum Rotations
System 1 for GRID 9
Vector Magnitude for GRID 9
NAS101, Page 5 - 109

Element Geometry Checks
NAS101, Page 5 - 110

Pre-processors are notorious for generating elements with
bad geometry (aspect ratio, taper, warp, skew, etc)
In the past, there was a separate message for each
element which did not satisfy the recommendations for
MSC.Nastran (this often resulted in pages upon pages of
messages which most users ignored)
There is now an option which provides user-control of
these messages (so if you want to ignore them, you can
disable their printout – NOT RECOMMENDED)
This is done by the GEOMCHECK executive control
statement
NAS101, Page 5 - 111

NAS101, Page 5 - 112
NAS101, Page 5 - 113
NAS101, Page 5 - 114

Structural Symmetry
Reflective symmetry can often be employed in the
modeling process to reduce the cost of the analysis.
NAS101, Page 5 - 115

Structural Symmetry (cont.)
The following example demonstrates the use of symmetric
modeling techniques to analyze the frame.
5000 lb • Full Model
2 4
C B
3 A
1 5
X
NAS101, Page 5 - 116
Symmetric Model
2500 lb P P
2 2
SUBCASE 1
2 3
SPC DOF 1,5,6

at Grid Point 3
for symmetry
1
NAS101, Page 5 - 117

Anti-Symmetric Model
Displaced Shape
2500 lb
SUBCASE 2
2 3
SPC DOF 2,3,4 at

Grid Point 3 for
antisymmetry
1
NAS101, Page 5 - 118

 D SYM,EX
TIME 5
SOL 101
CEND
$
TITLE =EXAMPLE OF USING SYMMETRY/ANTISYMMETRY CONSTRAINTS
DISP = ALL
$
SUBCASE 1
LABEL = SYMMETRY CONSTRAINTS
SPC = 1
LOAD = 1
$
SUBCASE 2
LABEL = ANTISYMMETRY CONSTRAINTS
SPC = 2
LOAD = 1
$
SUBCOM 3
LABEL = LEFT SIDE OF MODEL
SUBSEQ 1.0, 1.0
$
SUBCOM 4
LABEL = RIGHT SIDE OF MODEL
SUBSEQ 1.0, -1.0
$
BEGIN BULK
$
GRID 1 0.0 0.0 0.0 123456
GRID 2 0.0 10.0 0.0 345
GRID 3 5.0 10.0 0.0 34
$
CBAR 1 100 1 2 -1.0 0.0 0.0
CBAR 2 100 2 3 0.0 1.0 0.0
PBAR 100 1 5.0 5.0 5.0
$
NAS101, Page 5 - 119
 MAT1 1 3.E+7 0.3
$
FORCE 1 2 2500. 0.0 -1.0 0.0
$
SPC1 1 156 3
SPC1 2 2 3
$
ENDDATA
NAS101, Page 5 - 120

SUBCASE 1 + SUBCASE 2 = SUBCOM 3
NAS101, Page 5 - 121

SUBCOM 4 results in the displaced shape of the right side
of the model.
B C
- =
SUBCASE 1 - SUBCASE 2 = SUBCOM 4
NAS101, Page 5 - 122

The output for SUBCOM3 and SUBCOM 4 represent the
full model.
C B
A
NAS101, Page 5 - 123

Blank Page
NAS101, Page 5 - 124

10524280

Uploaded by

Document Information

Original Description:

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

10524280

Uploaded by

Copyright:

Available Formats

Section 5

Essential QA Checks – Workshop 71

Useful Model Summary Information 102

Typical errors; Singularities and Mechanisms

The Stiffness Matrix cannot be inverted if it is Singular

 Some examples of singularities are:

 Rigid Body Motion Adequate Constraints

 Forgetting to ‘Equivalence’ in MSC.PATRAN or other

If obvious singularities exist, MSC/NASTRAN attempts

 The Bulk Data entry PARAM,AUTOSPC,YES instructs

PARAM, AUTOSPC, YES is the default in most of the

GRID POINT SINGULARITY TABLE

 [Kgg] must be nonsingular in order to solve this equation. To achieve a nonsingular

Write out ‘failed’ DOF’s to .pch File

Re-use generated SPC1 data selectively

 Evaluate the Grid Point Singularity Tables

 Run the MSC.NASTRAN input files

 Evaluate the Tables and check the Fatal Messages

0 GRID POINT SINGULARITY TABLE

No Grid Point Singularity terms for GRID 13 What happens

 Resulted in FATAL errors due to the presence of

 This section shows how to identify the type of error.

 Using SPC or SPC1 entries,

 Evaluate the results and the implication of the changes

Constrained DOF 4,5,6

Looks OK, but be careful!!

 Fatal Message 9050

In practice there are many checks that MSC.NASTRAN will

The form of the FATAL message will be similar. The key

Further details on the meaning of the FATAL message can

 On Line Encyclopedia – result of search for ‘2026’

 Section5_5.bdf Correct the bar data and rerun

 Completely visible to Users

DMAP modules perform the mathematical operations

 The ".f04" file contains a "road-map" of the MSC/NASTRAN

 Each time a DMAP module is executed, a line is created in the .f04

 In Practice, understanding the DMAP Module flow may be difficult, but

 In this section we discuss PREVENTION

 Good Modeling Practice

• Use your pre-processor to visually check distortion

• Use WARNING messages in .f06 to confirm

• Check with Force = Mass * Acceleration

• Quadrilateral elements should be kept as square as possible

• Up to ~ 5% is normally acceptable. No real limit, but the element does

 Applied Load Summation

 Strain Energy Values

Assumes no Rounding Errors P  K δ  0

In Practice there is a Residual

Make it an Energy Term δT  P *

 For your type of Structure, Idealization and Analysis

 Establish typical values for Epsilon

 Establish an agreed threshold

*** USER INFORMATION MESSAGE 5293 (SSG3A)

 Use OLOAD request in Case Control

 Particularly Important For:

 Complex Pressure Loading

 Complex Distributed Loading

 Check that sense opposes and balances OLOAD summary

Request printout of applied loads at actual load

May produce large output for certain type loadings

*** USER INFORMATION MESSAGE 5293 (SSG3A)

Work = ½ Total Force * Total Deflection

-1.0Ux145 + 1.0Ux146 = 0.0

146 -1.0Uy145 + 1.0Uy146 = 0.0

0 * R1 * D I S P L A C E M E N T M A X / M I N V A L U E S U M M A R Y RESULTS FOR SUBCASE 1