Вы находитесь на странице: 1из 324

MD Nastran R2.

1
Release Guide
Corporate
MSC.Software Corporation
2 MacArthur Place
Santa Ana, CA 92707 USA
Telephone: (800) 345-2078
Fax: (714) 784-4056
Europe
MSC.Software GmbH
Am Moosfeld 13
81829 Munich, Germany
Telephone: (49) (89) 43 19 87 0
Fax: (49) (89) 43 61 71 6
Asia Pacific
MSC.Software Japan Ltd.
Shinjuku First West 8F
23-7 Nishi Shinjuku
1-Chome, Shinjyku-Ku
Tokyo 160-0023, JAPAN
Telephone: (03)-6911-1200
Fax: (03)-6911-1201
Worldwide Web
www.mscsoftware.com
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 2007 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.
This software may contain certain third-party software that is protected by copyright and licensed from
MSC.Software suppliers.
MSC, MD, Dytran, Marc, MSC Nastran, MD Nastran, Patran, MD Patran, the MSC.Software corporate logo, and
Simulating Reality are trademarks or registered trademarks of the MSC.Software Corporation in the United States
and/or other countries.
NASTRAN is a registered trademark of NASA. PAMCRASH is a trademark or registered trademark of ESI Group.
SAMCEF is a trademark or registered trademark of Samtech SA. LS-DYNA is a trademark or registered trademark
of Livermore Software Technology Corporation. ANSYS is a registered trademark of SAS IP, Inc., a wholly owned
subsidiary of ANSYS Inc. ABAQUS is a registered trademark of ABAQUS Inc. All other brand names, product
names or trademarks belong to their respective owners. PCGLSS 6.0, Copyright 1992-2005, Computational
Applications and System Integration Inc. All rights reserved. PCGLSS 6.0 is licensed from Computational
Applications and System Integration Inc.
MDNA:R2.1:Z:Z:Z:DC-REL
Cont ent s
MD Nastran R2.1 Release Guide
Preface to the MD Nastran R2.1 Release Guide xiv
A Word About Prerelease Features xv
List of Books xvi
Technical Support xvii
Internet Resources xix
1 Overview of MD Nastran R2.1
Overview 2
Contact in Linear and Nonlinear Solutions 2
Advanced Integrated Nonlinear Analysis 2
Implicit Nonlinear Analysis (SOL 600) 3
Explicit Nonlinear Analysis (SOL 700) 4
NVH and Acoustics 4
Numerical Enhancements 5
Elements and Connectors 5
Optimization 6
Rotor Dynamics and Aeroelasticity 6
List of Errors Resolved 6
List of MD Nastran Documents Released with MD Nastran R2.1 6
2 Contact in Linear and Nonlinear Solutions
Linear Contact, Permanent Glued Contact, and Nonlinear Contact 10
Introduction 10
Linear Contact Modeling in SOL 101 10
Permanent Glued Contact Modeling in SOLs 101, 103, 105, 107, 108, 109, 110,
111, and 112 10
Nonlinear Contact Modeling in SOL 400 10
Benefits 11
Input 11
Output 12
Guidelines and Limitations 12
Examples 17
MSC Nastran 2007
Release Guide
Table of Contents
MD Nastran R2.1 Release Guide

iv
BCONTACT=ALLBODY (Prerelease Only) 26
Introduction 26
Benefits 26
Input 26
Output 26
Limitation 26
Example 26
Performance Improvement to Contact Analysis 29
Introduction 29
Examples 29
3 Advanced Integrated Nonlinear Analysis
Advanced Integrated Nonlinear Analysis (SOL 400) 33
Introduction 33
Benefits 33
Limitations for the Current Release 34
A New General Nonlinear Contact Analysis Capability 34
Kinematic Elements 35
Introduction 35
Benefits 35
Types of Elements that can be Combined with the Kinematic Elements 35
The Scale Factors LMFACT and PENFN 36
Thermal Effects 36
RJOINT and the Pin Flags of the CBEAM Element 36
User Interfaces 37
Examples 38
SOL 400 Material Properties and Elements 43
Introduction to Nonlinear Property Enhancements 43
Benefits 43
Input 44
Output 45
Guidelines and Limitations 45
Examples 50
The Nonlinear CQUADR and CTRIAR Elements 51
Introduction 51
Benefits 51
Limitations 51
Types of Analysis 51
Drilling Degrees of Freedom 52
v Contents
Off-plane Shear Formulations 52
Green Strain 53
Stress Output Location 53
Differential Stiffness 54
CBAR Element to CBEAM Element Conversion 55
CSHEAR Element to CQUAD4 Element Conversion 56
SOL 400 Pretension Modeling Description 57
Introduction to Pretension Modeling 57
Benefits 57
Input 57
Output 57
Guidelines and Limitations 57
Examples 58
VCCT Virtual Crack Closure Technique 62
Cohesive Zone Modeling 63
Nonlinear Iteration Algorithms 66
Analysis Chaining SUBCASE, STEP, ANALYSIS, and NLIC 70
Adaptive Time Stepping Scheme 73
AUTO Step 73
Recycling Criterion 73
Exceptions 74
Post Files Output 74
Defaults 74
The Numerical Integration Method for Transient Analysis the HHT
Method 76
Nonlinear Iteration Summary Table for Nonlinear Transient Analysis in
SOL 400 77
Output Data Grouping: NLPACK 80
Restarts 80
File Management Commands 81
Case Control Modifications 81
Bulk Data Modifications 82
Initial Conditions 83
Transient Temperature Loads 85
Boundary Condition (SPC and MPC) Changes in MD Nastran R2.1 87
MD Nastran R2.1 Release Guide

vi
Introduction 87
Relative or Total Motion of a Step 87
Guidelines 88
Example 89
Direct Matrix Input Changes 93
Rotor Dynamics 94
Coordinate System for the Nonlinear Stresses of 3-D Elements 95
Outputs 95
Error Handling 95
User Interfaces 97
Examples 99
Differential/Follower Force Stiffness Processing in SOL 400 115
Introduction 115
Limitation 115
Input 115
Guidelines 115
Example 116
Contact Analysis and Kinematic Elements 118
Introduction 118
Example 118
4 Implicit and Explicit Nonlinear Analysis
MD Nastran Implicit Nonlinear - SOL 600 122
Heat Transfer 122
Heat Transfer Examples 125
Creep Simulation 125
Element Selection 125
Other SOL 600 Items 125
Fracture Mechanics J-Integral (LORENZI) 127
Delamination 127
MATEP Extensions 128
SOL 600 Failure Description MATF 128
Element Birth and Death 134
Unglue 134
Composite Element Numerical Analysis 135
PLOAD4 Extensions 135
Large Rotation RBE 136
Streaming Input 136
vii Contents
Connector Technology 136
Computational Enhancements 137
Superelements/DMIG 137
Contact Enhancements 138
New and Improved SOL 600 Features for MD Nastran R2.1. 138
New SOL 600 Parameters 139
Platform-Specific Notes 141
For 64-bit Windows EM64T Platforms 141
For 32-bit Windows Platforms 142
Supported Systems for SOL 600 in MD Nastran R2.1 142
MD Nastran Explicit Nonlinear - SOL 700 144
Introduction 144
Latest Capabilities of MD Nastran Explicit Nonlinear - SOL 700 144
Airbags and Occupant Safety 144
Defining Inflators in Airbags 146
Occupant Dummy Models 146
First Technology Safety Systems (FTSS) Anthropomorphic Test Dummy (ATD)
148
Time Domain NVH 148
Time Domain NVH Example 150
Prestressing (Implicit-to-Explicit Sequential Simulation) 159
Additional Capabilities of MD Nastran R2 Explicit Nonlinear - SOL 700 160
5 NVH and Acoustics
Frequency Response Function (FRF) and FRF-Based Assembly (FBA)
Feature in MD Nastran R2 164
Introduction 164
FRF Concept 164
FBA Concept 165
Using the FRF/FBA Feature 165
Exterior Acoustics 168
Introduction 168
Benefits 168
Input 168
Definition of Infinite Elements 168
Definition of Field Point Meshes 171
Case Control Commands 173
Output 175
Guidelines 176
Limitations 176
Input File 179
MD Nastran R2.1 Release Guide

viii
Excerpt of fluid1.bdf 180
Results 181
Rigid Porous Absorber 183
Introduction 183
Porous Materials 183
Inputs 184
Discussion 185
References 185
Example 185
6 Numerical Enhancements
New SPARSESOLVER Executive Control Statement 192
Introduction 192
Benefits 192
Method and Theory 192
Inputs 192
Outputs 192
Guidelines and Limitations 192
Improved Performance with New Sparse Solvers 195
Introduction 195
The TAUCS Sparse Solver 195
User Interface 195
Use Cases 195
Limitations 196
The UMFPACK Sparse Solver 197
UMFPACK Memory Guidelines (Updated for MD Nastran R2.1) 199
Known Limitations 200
Improved Memory Usage in Lanczos 201
New MAXRATIO Information Output (Prerelease) 202
Introduction 202
Benefits 202
Method and Theory 202
Inputs 202
Outputs 202
Guidelines and Limitations 203
Demonstration Example 203
Example Input Data 203
Example Output 205
ix Contents
New SPARSESOLVER MDTSTATS Information Output (Prerelease)
206
Introduction 206
Benefits 206
Method and Theory 206
Inputs 206
Outputs 206
Guidelines and Limitations 207
Demonstration Example 207
Example Input Data 207
Example Output 209
Performance Improvements and Expanded Capabilities for Automated
Component Modal Synthesis (ACMS) 211
Introduction 211
MDACMS for Upstream Superelements 211
Automatic FASTFR Decision Logic 212
Examples 213
Limitations for the FASTFR Method 214
Miscellaneous Performance Improvements 215
Relaxed Restrictions for CASI Solver Usage 218
System-Dependent Performance Improvements 219
Introduction 219
Model Description 219
Performance Improvements for Intel x86_64 220
Performance Improvements for AMD due to Atlas 220
Improved Selection of Reordering Methods 221
Iterative Solution Options for Nonlinear Analysis 223
Introduction 223
Benefits 223
Inputs 223
7 Elements and Connectors
Enhancements to Connector Elements 226
Introduction 226
CSEAM Element 226
SWLDPRM Enhancements 229
Displacement Output of GA and GB for CWELD and CFAST Elements 230
Composite Beam Using the Variational Asymptotic Method (VAM) 234
MD Nastran R2.1 Release Guide

x
Introduction 234
Benefits 234
Input 234
Output 236
Guidelines 237
Limitations 238
Example 238
References 240
Nonhomogeneous Multipoint Constraint 241
Introduction to MPCY and MPCD Entries 241
Benefits 241
Input 241
Output 241
Guidelines and Limitations 241
Theory 242
Examples 243
8 Optimization
Topology Optimization Enhancements 248
Introduction 248
Benefits 248
Modified Type One Responses Fractional Mass 250
Guidelines and Limitations 250
Example 1 Wheel (wheeltop.dat) 251
Automatic External Superelement Optimization (AESO) 254
Introduction 254
Benefits 254
Methodology 254
Input 255
Outputs 257
Guidelines and Limitations 260
Examples 262
Miscellaneous 265
Randomization of an Input Data File (Prerelease) 269
Introduction 269
Benefits 269
Guidelines and Limitations 270
Random Elimination of Element Types (Prerelease) 271
Introduction 271
xi Contents
Benefits 271
Guidelines and Limitations 271
9 Rotor Dynamics and Aeroelasticity
Changes to Rotor Dynamics for MD Nastran R2 274
UNBALNC Entry for Frequency Response 274
New Parameter and Hybrid Damping Specifications 274
Updating/Summing of Monitor Points 275
Introduction 275
Benefits 275
Input 275
Output 275
Examples (monsum.dat and monsum3.dat) 275
Guidelines and Limitations 276
Stripwise Aerodynamic Results 277
Introduction 277
Benefits 277
Input 277
Output 277
Examples (moncncm.dat and monbodi.dat) 278
Guidelines and Limitations 278
Input of an Aerodynamic Mesh 279
Introduction 279
Benefits 279
Input 279
Output 279
Example (aegridf.dat) 280
Guidelines and Limitations 281
Rigid Body Spline 282
Introduction 282
Benefits 282
Input 282
Output 282
Example (splinrb.dat) 282
Guidelines and Limitations 282
Wendland Spline Functions for the SPLINE4/5 283
Introduction 283
Benefits 283
Input 283
MD Nastran R2.1 Release Guide

xii
Example (ha144c_ris) 283
Guidelines and Limitations 283
Spline Blending 284
Introduction 284
Benefits 284
Input 284
Example (ha145e_blnd) 284
Guidelines and Limitations 284
Export of the Spline Matrix 286
Introduction 286
Benefits 286
Input 286
Output 286
Examples (splinopch and splinoop2) 286
External Spline Server 287
Introduction 287
Benefits 287
Inputs 288
Outputs 292
Guidelines and Limitations 292
Example 292
The 2005 New Template
Preface
s Preface to the MD Nastran R2.1 Release Guide
s A Word About Prerelease Features
s List of Books
s Technical Support
s Internet Resources
MD Nastran R2.1 Release Guide

xiv
Preface to the MD Nastran R2.1 Release Guide
This Release Guide contains descriptions for both the MD Nastran R2 and MD Nastran R2.1 versions,
and supersedes the MD Nastran R2 Release Guide.
xv
Preface
A Word About Prerelease Features
MD Nastran R2.1 contains a number of features that have been labeled as prerelease.
A prerelease feature or enhancement is defined as a feature or enhancement that has not yet completed
MSCs exhaustive verification and validation (V and V) testing and qualification process. Therefore,
prerelease features are to be used at the clients own risk.
MD Nastran R2.1 Release Guide

xvi
List of Books
Below is a list of some of the MD Nastran and MSC Nastran documents. You may order any of these
documents from the MSC.Software BooksMart site at http://store.mscsoftware.com/.
Installation and Release Guides
Installation and Operations Guide
Release Guide
Reference Books
Quick Reference Guide
DMAP Programmers Guide
Reference Manual
Users Guides
Getting Started
Linear Static Analysis
Basic Dynamic Analysis
Advanced Dynamic Analysis
Design Sensitivity and Optimization
Thermal Analysis
Numerical Methods
Aeroelastic Analysis
Superelement
User Modifiable
Toolkit
Implicit Nonlinear (SOL 600)
Explicit Nonlinear (SOL 700)
MD Users Guide
Topology Optimization
xvii
Preface
Technical Support
For help with installing or using an MSC.Software product, contact your local technical support services.
Our technical support provides the following services:
Resolution of installation problems
Advice on specific analysis capabilities
Advice on modeling techniques
Resolution of specific analysis problems (e.g., fatal messages)
Verification of code error.
If you have concerns about an analysis, we suggest that you contact us at an early stage.
You can reach technical support services on the web, by telephone, or e-mail.
Web Go to the MSC.Software website at www.mscsoftware.com, and click on Support. Here you can find
a wide variety of support resources including application examples, technical application notes, training
courses, and documentation updates at the MSC.Software Training, Technical Support, and
Documentation web page.
Phone
and
Fax
United States
Telephone: (800) 732-7284
Fax: (714) 784-4343
Frimley, Camberley
Surrey, United Kingdom
Telephone: (44) (1276) 60 19 00
Fax: (44) (1276) 69 11 11
Munich, Germany
Telephone: (49) (89) 43 19 87 0
Fax: (49) (89) 43 61 71 6
Tokyo, Japan
Telephone: (81) (03) 6911 1200
Fax: (81) (03) 6911 1201
Rome, Italy
Telephone: (390) (6) 5 91 64 50
Fax: (390) (6) 5 91 25 05
Paris, France
Telephone: (33) (1) 69 36 69 36
Fax: (33) (1) 69 36 45 17
Moscow, Russia
Telephone: (7) (095) 236 6177
Fax: (7) (095) 236 9762
Gouda, The Netherlands
Telephone: (31) (18) 2543700
Fax: (31) (18) 2543707
Madrid, Spain
Telephone: (34) (91) 5560919
Fax: (34) (91) 5567280
MD Nastran R2.1 Release Guide

xviii
Email Send a detailed description of the problem to the email address below that corresponds to the product you
are using. You should receive an acknowledgement that your message was received, followed by an
email from one of our Technical Support Engineers.
Training
The MSC Institute of Technology is the world's largest global supplier of CAD/CAM/CAE/PDM training
products and services for the product design, analysis, and manufacturing markets. We offer over 100
courses through a global network of education centers. The Institute is uniquely positioned to optimize
your investment in design and simulation software tools.
Our industry experienced expert staff is available to customize our course offerings to meet your unique
training requirements. For the most effective training, The Institute also offers many of our courses at our
customer's facilities.
The MSC Institute of Technology is located at:
2 MacArthur Place
Santa Ana, CA 92707
Phone: (800) 732-7211
Fax: (714) 784-4028
The Institute maintains state-of-the-art classroom facilities and individual computer graphics laboratories
at training centers throughout the world. All of our courses emphasize hands-on computer laboratory
work to facility skills development.
We specialize in customized training based on our evaluation of your design and simulation processes,
which yields courses that are geared to your business.
In addition to traditional instructor-led classes, we also offer video and DVD courses, interactive
multimedia training, web-based training, and a specialized instructor's program.
Course Information and Registration. For detailed course descriptions, schedule information,
and registration call the Training Specialist at (800) 732-7211 or visit www.mscsoftware.com.
MD Patran Support
MD Nastran Support
Dytran Support
MSC Fatigue Support
Marc Support
MSC Institute Course Information
mscpatran.support@mscsoftware.com
mscnastran.support@mscsoftware.com
mscdytran.support@mscsoftware.com
mscfatigue.support@mscsoftware.com
mscmarc.support@mscsoftware.com
msctraining.support@mscsoftware.com
xix
Preface
Internet Resources
MSC.Software (www.mscsoftware.com)
MSC.Software corporate site with information on the latest events, products, and services for the
CAD/CAE/CAM marketplace.
MD Nastran R2.1 Release Guide

xx
Ch. 1: Overview of MD Nastran R2.1 MD Nastran R2.1 Release Guide
1
Overview of MD Nastran R2.1
!
Overview
MD Nastran R2.1 Release Guide

2
Overview
MD Nastran R2 and R2.1 bring powerful new analysis features and enhancements in the areas of contact;
nonlinearity; noise, vibration, and harshness (NVH); acoustics; automotive powertrain; crash and
occupant safety; high performance computing; assembly modeling; optimization; rotor dynamics; and
aeroelasticity. The combined enhancements in MD Nastran R2 and R2.1 are summarized as follows.
Contact in Linear and Nonlinear Solutions
MD Nastran introduces contact in linear and nonlinear solutions, with the ability to easily convert from
linear contact in SOL 101 to nonlinear contact in SOL 400.
Nonlinear Contact in SOL 400. Provides general multibody contact under full nonlinear
conditions of large contact motion, nonlinear materials, large rotation, and large strain. SOL 400
features deformable-deformable contact, deformable-rigid contact, and contact with friction.
Linear Contact in SOL 101. Provides an algorithm similar to SOL 400 for use in SOL 101 linear
analysis. A SOL 101 model with linear contact can be converted to simulate nonlinear behavior
in SOL 400 without having to redefine contact.
Permanent Glued Contact in Linear Solution Sequences. Provides a special type of contact
model to be used when the contacting surfaces have no relative normal or tangential motion.
Permanent glued contact is available in most of the linear solutions (except SOL 200), and
provides a convenient way to join dissimilar meshes. Glued, linear, and nonlinear contact make
use of the same user interface.
(MD R2.1) Simplified Contact Definition. A new ALLBODY contact option saves significant
time in preparing the input definition.
(MD R2.1) Improved Performance. Contact performance has been greatly improved, particularly
for large models.
(MD R2.1) (Prerelease) Glued Contact in Optimization. A prerelease version of glued contact is
available in SOL 200.
Advanced Integrated Nonlinear Analysis
SOL 400 has been extensively enhanced for multistep nonlinear analysis of large strain and material
behavior, with or without contact.
Nonlinear Contact Modeling. A new contact algorithm has been implemented in SOL 400 to
perform general multibody contact for large contact motion, nonlinear materials, large rotations,
and large strains. It is available in both nonlinear static and nonlinear transient analysis.
3
CHAPTER 1
Overview of MD Nastran R2.1
Nonlinear Elements and Materials. Enhancements include orthotropic material properties for
three-dimensional and plane strain behavior, nonlinear gasket material properties for
compression behavior, elasto-plastic material properties for use in large deformation analysis,
and failure model properties for linear elastic material. Where appropriate, most elements in
MD Nastran have been extended to include the effects of finite strain. Also, layered
axisymmetric, shell, and solid composite elements are now available. The CQUADR and
CTRIAR elements have been extended to nonlinear analysis.
Nonlinear Procedures. New analysis chaining procedures have been added to conveniently
define multiple load steps, run multiple independent cases, and specify multiple and mixed types
of analyses in one job. For example, the user can pretension a bolt structure as a first load step in
a multistep nonlinear analysis. Adaptive stepping schemes have been introduced that
automatically modify load increment or time step size to accelerate convergence.
Crack and Delamination. A new virtual crack closure technique (VCCT) calculates energy
release rates for sharp cracks. In addition, a new library of interface elements (cohesive zone
modeling) can be used to simulate the onset and progress of delamination.
Kinematic Elements. New rigid elements with Lagrange formulations are now available for use
in geometric nonlinear analysis involving large rotations. These elements have been
implemented in both nonlinear static and nonlinear transient analysis.
(MD R2.1) Differential Stiffness. Improved nonlinear differential stiffness with expanded
options to selectively turn off differential stiffness (NLDIFF) for improved convergence is now
available.
(MD R2.1) Relative Boundary Conditions. A new relative enforced displacement capability (see
the SPCR Bulk Data entry) is now available; this is particularly useful for bolt modeling.
Implicit Nonlinear Analysis (SOL 600)
Heat Transfer. Advanced thermal analysis is now available in SOL 600, including an efficient
hemi-cube view factor calculation method, and automated procedures for thermal stress analysis
based upon a prior heat transfer simulation. In addition, thermal analysis of composites includes
calculation of the thermal gradient through the thickness.
Modeling Enhancements. Connector technology has been enhanced to include large deformation
formulations of the CFAST, CWELD, and CBUSH elements. Enhancements for facture
mechanics include the calculation of stress intensity factors using the VCCT or Lorenzi
methods, and the prediction of delamination.
Performance Improvements. A new streaming input option is available that eliminates transfer
files. Streaming input will be automatically turned off for analysis scenarios that are not
supported. The analysis of composite shells has been improved such that assembly time has
been reducedoften by a factor of tenand memory requirements have been substantially
reduced as well. Other performance improvements include an iterative solution that runs out-of-
core, and direct PLOAD4 support.
(MD R2.1) Glued Contact. Permanent glued contact is now supported for conveniently joining
dissimilar meshes.
MD Nastran R2.1 Release Guide

4
(MD R2.1) Solver Options. A mixed derivative solver has been added as an option.
Explicit Nonlinear Analysis (SOL 700)
Several new SOL 700 analysis capabilities are introduced in this release.
Airbags and Occupant Safety. Fluid structure interaction (FSI) is now available in SOL 700 to
simulate the complex behavior of multicompartment airbags and their interaction with
anthropomorphic test devices (ATDs) during crash scenarios. Airbag inflation is based on full
gas dynamics and finite volume (Eulerian) technology with adaptive meshing capability. The
airbag surface acts as a coupling surface which adapts itself and follows the expanding Eulerian
elements as the gas jet flows inside the bag. In addition to full gas dynamics, SOL 700 also
supports the conventional uniform pressure method to simulate airbag behavior.
Time Domain Noise, Vibration, and Harshness (NVH). This technique is based on the fast
Fourier transform (FFT) method to compute frequencies and mode shapes of highly nonlinear
and dynamic systems such as a vehicle running over a rough road. The advantage of time domain
NVH is its ability to capture the nonlinear effects of contact, spring, bushing, tire, and
suspension behavior.
Prestress. Prestressing is now performed by a double precision version of the implicit solver, and
the results are used as the prestate for explicit simulations such as bird strike and blade-out
applications.
Nastran Native Output. This release produces native MD Nastran output files for seamlessly
postprocessing explicit events with MD Patran.
New Material and Element Formulations. Thirteen new material models and five element
formulations are added in SOL 700.
Contact Features. SOL 700 now supports many new contact capabilities such as glued contact,
tied nodes to surface, tied shell edge to surface, tied surface to surface, spot weld contact, single
edge contact, and force transducers.
NVH and Acoustics
Frequency Response Functions (FRFs) and FRF Based Assembly (FBA). A powerful new
capability is available in MD Nastran R2 for frequency response analysis based on
representations of components by their FRFs, and on assemblies of such components by the FBA
process. FRF-based assemblies provide an effective means for revealing how excitations on one
component affect responses at other components in the assembly.
Exterior Acoustics. MD Nastran R2 is the production release for the exterior acoustics capability.
It extends acoustics applications to unbounded regions such as radiated acoustics from an
engine. A new sparse solver is available for the efficient solution of the unsymmetric matrices
encountered in exterior acoustics frequency response analysis.
5
CHAPTER 1
Overview of MD Nastran R2.1
Numerical Enhancements
Sparse Solvers. Two new sparse solvers have been introduced: TAUCS (statics) and UMFPACK
(unsymmetric). The UMFPACK solver provides scalable performance for exterior acoustics. In
addition, the Lanczos method has been enhanced to take advantage of available memory.
Automatic optimal reordering selection has been implemented for solid models to eliminate
requiring the user to set flags.
Iterative Solvers. Restrictions to the CASI iterative solver have been relaxed for statics of large
solid models (e.g., engines). This includes an expanded list of supported element types.
ACMS. Automated component modal synthesis (ACMS) has been extended to external
superelements to provide significant reductions in computational time, I/O, and scratch space.
One typical case study demonstrated improvement by an order of magnitude.
Other HPC Enhancements. MD Nastran R2 has been ported to the Microsoft Compute Cluster.
The compute kernels for x86_64 platforms have been optimized for both Intel and AMD-based
systems. Improved user diagnostics provide a pivot ratio bar chart to help localize model
singularities.
(MD R2.1) Iterative Solver for Nonlinear Analysis. An iterative solver option is now available
for nonlinear contact analysis of large solid models, as well as linear contact. Up to five-fold
increases in speed have been achieved.
(MD R2.1) Memory Management. Users can now control the amount of memory available for
the UMFPACK solver to ensure adequate memory at runtime.
(MD R2.1) (Prerelease) Sparse solver information output (MDTSTATS). Matrix diagonal term
statistical output is available via the new MDTSTATS feature for use in determining model
quality.
Elements and Connectors
Connectors. A new seam weld connector element (CSEAM) is now available for assembly
modeling. It features extended capabilities for connecting higher order elements; mesh-
independent connections to top/bottom shell patches defined either by property IDs or element
IDs; tailored parts connection; and support for anisotropic material properties. For spot weld
elements (CWELD, CFAST), end point displacement output can now be obtained to view the
relationship between the spot weld and the connecting shells. A new connector type, RBE2GS,
is introduced to optionally search and connect independent grid points of the two closest RBE2
elements within a specified search radius.
Composite Beam Using Variable Asymptotic Method (VAM). The arbitrary beam cross section
capability has been enhanced to support composite materials. A new three-noded composite
beam element (see the CBEAM3 Bulk Data entry) represents the cross-section and layup of
plies, and accounts for interaction among plies. It provides an efficient alternative to
conventional 3-D modeling for beam-like composite structures such as rotor blades and shell
stiffeners.
MD Nastran R2.1 Release Guide

6
Optimization
Topology Optimization. This release provides combined simultaneous topology, sizing, and
shape optimization to find potentially better designs. Different mass targets can now be applied
on multiple design parts of the structure. Symmetry constraints have been extended to cyclical
applications such as car wheels. An adjoint design sensitivity analysis method has been
implemented for inertia relief sizing optimization. Significant performance enhancement with
minimum member size control has been achieved, particularly for large numbers of sizing design
variables.
Automatic External Superelement Optimization (AESO). This new feature automatically
partitions the model into designed and nondesigned (external superelement) parts for efficient
optimization. An order of magnitude increase in speed can be achieved without requiring user
knowledge of superelements.
Randomization (Prerelease). This randomization capability provides a way to stochastically
introduce uncertainty into a model, such as tolerances in connectivity, properties, and loads. The
user selects the outputs to monitor. This beta capability is a first step in developing a multirun
environment to spawn multiple jobs, collect the results, and perform statistical postprocessing.
Rotor Dynamics and Aeroelasticity
Rotor Dynamics. Unbalanced loading can now be used for frequency response analysis with the
rotor dynamics option. Frequency response case control can be used directly in SOL 146, since
multiple RGYRO subcases can now be used in rotor dynamics. The specification of damping has
been simplified, and now allows for new damping formulations such as hybrid damping. In
addition, as a prerelease capability, the effects of rotor stiffness, mass, and damping can be
included in SOL 200 optimization.
Aeroelasticity. Monitoring points can now be updated and summed. A new type of monitoring
point (MONCNCM) has been introduced for monitoring stripwise aerodynamic results such as
lift and pitching moments. Other enhancements include various splining techniques for
aerodynamic structural applications.
(MD R2.1) Spline Application Programming Interfaces (APIs). Support has been added for user-
defined splines (SPLINEX) via an API.
List of Errors Resolved
The list of errors resolved in this release can be found at:
http://www.mscsoftwawre.com/support/prod_support/nastran/errorlist/files/error2007.lst
List of MD Nastran Documents Released with MD Nastran R2.1
Along with this Release Guide, the following documents are updated for the MD Nastran R2.1 release:
MD Nastran Installation and Operations Guide
7
CHAPTER 1
Overview of MD Nastran R2.1
MD Nastran Implicit Nonlinear (SOL 600) Users Guide
MD Nastran Explicit Nonlinear (SOL 700) Users Guide
MD Nastran Quick Reference Guide
MD Users Guide - Application Examples
MD Users Guide - Topology Optimization
MD Nastran R2.1 Release Guide

8
Ch. 2: Contact in Linear and Nonlinear Solutions MSC Nastran 2007 Release Guide
2
Contact in Linear and Nonlinear
Solutions
!
Linear Contact, Permanent Glued Contact, and Nonlinear Contact
!
BCONTACT=ALLBODY (Prerelease Only)
!
Performance Improvement to Contact Analysis
MD Nastran R2.1 Release Guide

10
Linear Contact, Permanent Glued Contact, and
Nonlinear Contact
Introduction
MD Nastran R2 introduces 3-D contact to linear analysis, and nonlinear general 3-D contact analysis
to SOL 400.
Linear Contact Modeling in SOL 101
Linear contact is defined as the full nonlinear contact algorithm of SOL 400 without material
nonlinear requirements, and without the usual linear requirements of small strain and small rotation
imposed. The contact bodies need not be in i nit ial contact, and multiple contact bodies are allowed.
Only surface-to-surface 3-D contact is currently supported. The grid points of the contacting bodies
need not be aligned, and the contact algorithm may be used to join dissimilar meshes. Both
deformable-deformable and deformable-rigid contact is allowed.
Bilinear Coulomb or bilinear shear friction is allowed. I n the general bilinear friction algorithms,
the relative tangental displacement is split into stick (elastic contribution) and slip (plastic
contribution) components. I n SOL 101, only an elastic (stick) contribution to relative tangental
displacement is computed.
Note that the linear GAP contact defined by the parameter PARAM,CDITER,n is still supported, but
should not be used in association with the surfaces defined for linear contact.
Permanent Glued Contact Modeling in SOLs 101, 103, 105,
107, 108, 109, 110, 111, and 112
Permanent glued contact is defined as a special type of contact model which imposes the condition
that there is no relative normal or tangental motion between the contacting surfaces.
For permanent glued contact, MD Nastran forms the required constraints without entering the full
nonlinear contact algorithm. I f there is no initial contact between the contacting bodies, these
constraints cannot be formed and the run will fail. I f i t i s not pos si bl e t o explicitly define the
contact surfaces, the user can set contact surface tolerances, and any grid points falling within these
tolerances will be glued. With this option, the user can also require that the algorithm bring the
surfaces into contact before applying the constraints.
Nonlinear Contact Modeling in SOL 400
A full nonlinear 3-D surface-to-surface contact algorithm and a 2-D solid edge-to-edge contact
algorithm is available in SOL 400, which works in conjunction with all the large strain-large
r ot at i on and material nonlinearity now available in SOL 400. The contact bodies need not be in
initial contact, and multiple contact bodies are allowed. Both deformable-deformable and
11
CHAPTER 2
Contact in Linear and Nonlinear Solutions
deformable-rigid contact is allowed. The algorithm works for both nonlinear static and nonlinear
transient analysis in SOL 400.
Bilinear Coulomb or bilinear shear friction is allowed. I n the general bilinear friction algorithms,
the relative tangental displacement is split into stick (elastic contribution) and slip (plastic
contribution) components.
The permanently glued contact is available for joining dissimilar meshes.
Benefits
The primary benefit of linear contact modeling in SOL 101 is that it permits the use of the full SOL 400
contact algorithm for linear structural models for which the only nonlinearity is contact. Also, if after
running such a model in SOL 101, the user determines that there are other nonlinear effects such as
material nonlinearity or large rotation, the model can simply be switched to SOL 400 without having to
redefine the contact surfaces, since both SOL 400 and SOL 101 use the same contact definitions.
The primary benefit of the permanent glued contact is the joining of two dissimilar meshes.
The primary benefit of nonlinear contact modeling in SOL 400 is that it allows the analysis of general
multibody contact under full nonlinear conditions of large contact motion, nonlinear materials, large
rotation, and large strain.
All of the new contact capabilities use the same user interface.
Input
Linear contact, permanent glued contact, and general nonlinear contact are all controlled by the Case
Control command BCONTACT. The following describes the appropriate user interface for SOL 101,
SOL 400, and permanent glued contact.
1. When used with SOL 101, SOL 400, or permanent glued contact in the previously described
linear solutions, only the BCONTACT = n or BCONTACT=NONE (default) entries are valid.
For BCONTACT = n, n can refer to the BCTABLE, BCHANGE, or BCMOVE entries.
2. For SOL 101 and SOL 400, the following associated contact Bulk Data entries are supported:
a. BCBODY is supported except for the following options:
HEAT, POLY, CYLIND, SPHERE, LINE, ARC, and NURBS2.
b. BCHANGE is supported except for the following option:
EXCLUDE
c. BCMOVE is supported.
d. BCPARA is supported except for the following options:
NBODIES, MAXENT, MAXNOD, ISPLIT, MAXSEP, ICHECK, and IPRINT;
FTYPE only supports 0 (default for SOL 400), 6, and 7;
MD Nastran R2.1 Release Guide

12
FKIND, BEAMB, FSSMULT, FSSTOL, LINQUAD, INITCON.
e. BCPROP is supported.
f. BCTABLE is supported, with the following not supported:
HHHB entry;
FK through TBLCID.
3. Prior to bodies coming into contact due to load application, it may be desirable to initially identify
contacting bodies or specify initial stress-free contact. Thus, Bulk Data entries such as BCPARA
have an ID field which may be 0 or blank. (For the BCPARA entry, BCPARA,0 is usually the
typical and only usage, since it is used to define initial contact parameters.) For other entries such
as a BCTABLE entry, a BCTABLE,0 entry would be used to identify the contact bodies that can
possibly contact each other at the start of the analysis. Then, BCTABLE, n with n > 0 and called
out in the Case Control Section by a BCONTACT = n entry is used to control contact during
loading. In SOL 400, a BCONTACT= 0 is allowed above all subcases, but is not required. Any
of the contact Bulk Data entries that allow a 0 and have a 0 value ID field are automatically
detected by SOL 400 with or without a BCONTACT = 0 Case Control command.
4. The parameters PARAM,LMFACT,value and PARAM,PENFN,value can be used to control the
penalty function value for contact.
5. The parameter PARAM, MARCREVR can be used to reverse the motion of rigid contact surfaces
if the user enters the patch information in the wrong order.
Output
The Case Control command BOUTPUT controls contact output. For the BOUTPUT Case Control
command in SOLs 101 and 400, SORT2 and PLOT are not currently supported.
Guidelines and Limitations
1. Surface-to-surface 3-D contact is supported in SOL 101 and SOL 400. Additionally, 2-D solid
contact is supported in SOL 400. For solids contacting solids, the contact surfaces contain the grid
points. For shell surfaces contacting each other, the contacting surfaces do not contain the grid
points. For contacting shell surfaces, the distance between grids is , where and are
the thicknesses of the two contacting shells. For deformable-rigid contact, contact occurs when
the position of the shell grid plus or minus half the thickness projected with the normal comes into
contact with the rigid segment. Note that if the shell grids are specified to be touching, the
algorithm will assume contact penetration.
2. SOL 101 allows linear contact. This means the full SOL 400 nonlinear contact algorithm
without material nonlinearity, and standard linear small strain and small rotation is assumed.
3. Permanent glued contact is defined when the IGLUE field of the BCTABLE entry is set to 1. In
this option, all degrees of freedom of the contact grids are constrained with MPCs in the case of
deformable-deformable contact once the grids have come in contact. The relative tangential
motion of a contact grid is zero in the case of deformable-rigid contact. Permanent glued contact
is available in SOLs 101, 103, 105, 107, 108, 109, 110, 111, and 112, as well as in SOL 400. If
t
1
t
2
+ ( ) 2 t
1
t
2
13
CHAPTER 2
Contact in Linear and Nonlinear Solutions
IGLUE=1 on the BCTABLE entry, MD Nastran will form the required constraints without
entering the full nonlinear contact algorithm. If there is no initial contact between the contacting
bodies, these constraints cannot be formed and the run will fail. Therefore, special cases arise
which are described as follows:
a. For a permanent glue contact analysis in SOL 101 or SOL 400, consider the case of no initial
contact between the contacting bodies. Use entry BCPARA , 0 , NLGLUE , 1 to turn
on the general SOL 400 contact algorithm. Use the Case Control command BCONTACT,
etc., to define possible contacting surfaces. The algorithm will then determine the contacting
surfaces and glue the bodies together.
b. If, in SOL 400 on the BCTABLE entry, there are multiple GLUE and non-GLUE entries
associated with different SLAVE entries, then BCPARA , 0 , NLGLUE , 1 must be
used.
c. Only SOL 101 and SOL 400 can call the general nonlinear contact algorithm. If, in SOLs 101
(optional method), 103, 105, 107, 108, 109, 110, 111, and 112, the user wishes to form a
permanent glue contact and there is no initial contact between the contacting bodies, it is
necessary to specify a value for the ERROR field on the BCTABLE,0 entry. Any grid within
this error tolerance will be considered to be in contact. If the user sets ICOORD=1 on the
BCTABLE,0 entry, then the grid points will be physically moved so that the surfaces are
actually in contact. The user can also use the Bulk Data entry BCPARA,0 field ISHELL to
have the algorithm ignore shell thickness when computing contact penetration.
Do not use BCPARA , 0 , NLGLUE , 1 with this option.
4. Each grid and element should be, at most, in one body. Solids and shells should not be mixed in
the same body.
5. Deformable bodies can contact rigid bodies. Contact between rigid bodies is not supported.
6. In the following text, examples of contact bulk data entries and their fields are given. Only one
entry, field pair is given. The field, however, may appear on multiple entries. Field values on the
BCPARA entry are overridden by field values on the BCTABLE entry.
7. When defining deformable-deformable contact, the default rule is that the grid points of a body
may contact any other body, and the searching starts with the body having the lowest ID number
and ends with the body having the highest ID number. Depending on the contact constraints
found, it is possible that grid points of a body with a higher ID number contact surfaces of bodies
with a lower ID number. This default behavior (so-called double-sided contact) can be overruled
by using the BCBODY option or the BCTABLE option.
Two different global search orders can be defined via the BCBODY option. The first is so-called
single-sided contact, which means that the search for contact for grid points of a deformable body
is only done with respect to surfaces of the body itself and to bodies with a higher ID number.
The other global search order activates a process which attempts to optimize the set of contacting
grid points and contacted surfaces, and is only recommended in the rare cases for which a proper
set of MASTER and SLAVE bodies cannot be defined by the user.
MD Nastran R2.1 Release Guide

14
The recommended way of overruling the default search process is via the BCTABLE option,
where a local search method can be defined for each body combination. The BCTABLE option
allows the user to define which pair of contact bodies can potentially come into contact and, if
they can come into contact, it is also possible to specify that grid points of the body indicated as
SLAVE can only contact surfaces of the body indicated as MASTER, and not the other way
around. In this way, the user can define single sided contact per body combination, thus overruling
the default double-sided contact.
8. When defining deformable-deformable contact, it is important to ensure that grid points of a body
with a finer mesh are contacting surfaces of a body with a coarser mesh. When defining
deformable-rigid contact, it is important that the deformable bodies all have ID numbers lower
than the rigid bodies.
9. Softer bodies should be designated as SLAVE; e.g., rubber should be a SLAVE and steel should
be a MASTER.
10. Avoid sharp corners in contacted surfaces. It is better to have a finer mesh and smooth surface
where possible.
11. In sliding contact, try to have smooth surfaces. Use the SPLINE option, which generates a Coons
surface (where the mesh surfaces are patched). This helps considerably by avoiding scattering
and accelerating convergence.
12. Because contact can entail large motion, it is recommended that a full Newton iteration scheme
be used. This can be facilitated on the NLPARM Bulk Data entry by the addition of the keyword
FNT to the KMETHOD field and on the TSTEPNL Bulk Data entry by the addition of the
keyword FNT to the METHOD field .
13. The motion of deformable bodies is prescribed by applying displacements or loads to the bodies.
It is recommended that enforced displacements or point loads not be placed on grids that might
come into contact with a rigid body.
14. It is recommended that MPC relationships be avoided for grid points that might come into contact.
15. In SOL 400 2-D solid edge-to-edge contact, it is not recommended that higher order CQUAD,
CTRIAX, or CQUADX elements be used with the contact algorithm unless they have an
associated PSHLN2 entry.
16. Rigid bodies may have prescribed velocity, position, or load. See the BCBODY Bulk Data entry.
17. During the contact process, it is unlikely that a grid point contacts a surface exactly. For this
reason, a contact tolerance is associated with each contact body pair. If a grid point is within the
tolerance, it is considered to be in contact. The algorithm computes the tolerance as the smaller of
5% of the smallest element size (solids), or 25% of the smallest element thickness (shells). The
user can manually set the tolerance (see BCTABLE entry, field ERROR). In general, the default
value is recommended.
18. The tolerance value lies equally above and below the contacted surface. That is to say, the
thickness of the contacted surface is twice the value of the tolerance. Often, the grid points are
almost touching the contacted surface. The algorithm provides for a biased tolerance with a
smaller distance on the outside (the side nearest the contacting surface) and a larger distance on
the inside. This avoids the close grid points from coming into contact and separating again. There
15
CHAPTER 2
Contact in Linear and Nonlinear Solutions
is a default bias of 0.9 in the SOL 400 algorithm. The user may set the bias (see BCTABLE entry
field BIAS). The outside contact area is computed as (1-BIAS)*ERROR and the inside contact
area is (1+BIAS)*ERROR. To turn off the bias, enter a small value on the BCTABLE entry for
BIAS such as 1.0-18. (Do not enter a value of 0.0.)
19. In SOL 101 and SOL 400, bilinear Coulomb friction (see BCPARA entry, FTYPE = 6) or bilinear
shear friction (see BCPARA entry, FTYPE = 7) is available. Bilinear shear is usually used for
forging processes. Bilinear Coulomb is used for most friction applications.
20. Currently, quadratic (eight-noded) shell elements are not recommended for use with contact.
21. Bilinear Coulomb friction is based on relative tangential displacements. The model assumes that
the stick and slip conditions correspond to reversible (elastic) and permanent (plastic) relative
displacements. The model defines a slip surface as
where is the tangential force, is the normal force, and is the friction coefficient (see the
BCTABLE entry, field FRIC). Stick is given by .
The rate of relative tangential displacement is split into elastic and plastic contributions according
to
and the rate of change of the friction force vector is related to the elastic tangential displacement
by
where the matrix is given by
where is the slip threshold below which sticking is simulated. The algorithm
determines as 0.0025 times the average edge length of the elements defining the deformable
contact bodies.
The physically impossible case of implies that the plastic or slip contribution must be
determined. The determining relationship is

ft = f
n

ft f
n

0 <
u

t u

t
e
= u

t
p
+
f

t Du
e
t =
D
D
f
n

-------- 0
0
f
n

--------
=

0 >
f
t

D u
t

u
t

\ .
| |
=
MD Nastran R2.1 Release Guide

16
The bilinear Coulomb model then assumes a slip flow potential given by , and further
assumes that the direction of the slip displacement rate is given by the normal to the slip flow
potential. Note that this is similar to a nonassociative rule, as and are different functions.
The bilinear Coulomb model then defines
where is the slip displacement rate. Also, since the friction force must always lie on the slip
surface,
Combining the preceding two equations, the slip displacement rate is given by
and the rate of change of friction force vector then becomes
The bilinear Coulomb model also uses the check on convergence to determine if convergence is
achieved when
where is the current total friction force vector; is the total friction force vector from the
previous iteration; and is the friction force tolerance, which has a value of 0.05.
22. The bilinear Coulomb model may not correlate well with experiment if the frictional tractions
exceed the flow stress or the stress of the material. If this occurs, the user can modify the frictional
stress limit (see BCTABLE entry, field FRLIM). In the case of the frictional stress limit , if
the surface traction reaches the limit value, the applied friction force is reduced so that the
maximum shear traction is given by . The user may also switch to the bilinear shear
friction model.
f
t
=

u
p
t

ft
-------- =


ft
---------
\ .
|
| |
T
f

t 0 = =

ft
---------
\ .
|
| |
T
Du

ft
---------
\ .
|
| |
T
D

ft
-------
------------------------------- =
f

t D
D

ft
--------

ft
---------
\ .
|
| |
T
D

ft
---------
\ .
|
| |
T
D

ft
--------
------------------------------------
\ .
|
|
|
|
|
|
| |
u

t D D

( )u

t = =
F
t
F
t
previous

F
t
----------------------------------------------- e
F
t
F
t
previous
e

t
limit
min
n

t
limit
, ( )
17
CHAPTER 2
Contact in Linear and Nonlinear Solutions
23. The bilinear shear friction model states that the frictional stress is a fraction of the equivalent
stress in the material:
for stick, and for slip
where is the surface tangent vector; is the friction coefficient (see the BCTABLE entry, field
FRIC); and is computed by
Examples
The following examples describe glued contact and linear contact. Other examples of contact can be
found in the MD Users Guide.
Example 1: Glued Contact (s103perg.dat)
As an example of permanent glued contact in SOL 103, Figure 2-1 shows a deformable cylinder made
up of ten-noded CTETRA elements and a bracket structure, also made up of ten-noded CTETRA
elements, sitting on top of the cylinder. We wish to glue the two structures together, and do a ground
check on the structures before proceeding with further analysis. The gluing is accomplished with the use
of the Case Control command BCONTACT=1 as shown in the abbreviated bulk data file. The
BCONTACT=1 command points to a BCTABLE with ID=1 and one value for the NGROUP field of the
BCTABLE entry, indicating that one pair of structures is to be glued.
Figure 2-1 Glued contact.

t


3
------- <
t


3
------- = t
t

t
min
n


3
--------- ,
\ .
|
| |
=
MD Nastran R2.1 Release Guide

18
The file for this example is shown as follows:
SOL 103
CEND
TITLE = Gluing with TETRA10 with flat interface
$
echo = sort(except ctetra,grid,bsurf,pload4)
autospc(noprint) = yes
SUBCASE 1
SUBTITLE= Modal analysis
LABEL = Ground Check for rigid body modes
BCONTACT = 1
METHOD=1
GROUNDCHECK(SET=ALL,DATAREC=YES)=YES
BEGIN BULK
PARAM,WTMASS,.00259
PARAM PRTMAXIM NO
BCPARA,0,BIAS,0.0
$
$ 1 >< 2 >< 3 >< 4 >< 5 >< 6 >< 7 >< 8 >< 9 >< 10 >
EIGRL120
BCTABLE 1 1
SLAVE 3 0.0001 0. 0. 0. 1
1
MASTERS 2
$ Deform Body Contact LBC set: cylinder
BCBODY 2 3D DEFORM 2 0
BSURF 2 1 2 3 4 5 6 7
8 9 10 11 12 13 14 15
.
.
.
4960 4961 4962 4963 4964 4965 4966 4967
4968 4969 4970 4971 4972 4973
$ Deform Body Contact LBC set: bracket
BCBODY 3 3D DEFORM 3 0
BSURF 3 4974 4975 4976 4977 4978 4979 4980
4981 4982 4983 4984 4985 4986 4987 4988
.
.
.
9413 9414 9415 9416 9417 9418 9419 9420
9421 9422 9423 9424 9425 9426 9427 9428
$ Model definition follows:
$
GRID 1 .446672 2.59695 -.063605
GRID 2 .50144 -.5775 -.085552
GRID 3 .323733 2.69367 -.571285
.
.
.
$
PSOLID 1 1 0
$ Pset: cylinder will be imported as: psolid.1
CTETRA 1 1 617 616 615 614 621 622
623 618 619 620
19
CHAPTER 2
Contact in Linear and Nonlinear Solutions
CTETRA 2 1 613 612 611 610 627 628
629 624 625 626
CTETRA 3 1 617 616 609 608 621 633
634 630 631 632
.
.
.
$ Material Record : al
$ 1 >< 2 >< 3 >< 4 >< 5 >< 6 >< 7 >< 8 >< 9 >< 10 >
MAT1 1 1.+7 .33 0.1
$
ENDDATA
The second line of the BCTABLE entry shows that the bracket structure is going to be the contact body
or the touching body. This is accomplished by the IDSLA1=3 field entry pointing to a BCBODY entry
with the same BID=3.
Most importantly, in field 8 of the SLAVE entry record, a value of 1 for the IGLUE field is set. This entry
directs SOL 103 to perform a permanent glued contact analysis.
The continuation entry after the SLAVE entry show a value of 1 for ISEARCH. This tells the algorithm
to search touching-to-touched. The entry MASTER defines the cylinder as the touched body. Its
IDMA1=2 field entry points to a corresponding BCBODY entry with a value of BID=2.
The BCBODY entries show that both structures are to be considered deformable, and that the 3-D
contact algorithm is to be used. (This is the only algorithm valid in linear analysis.) The BCBODY entry
with field BID=3 has on its field 5 BSID =3, which corresponds to a BSURF Bulk Data entry of the same
value for BSID. The BSURF entry then lists the surface elements of the bracket that can come into
contact.
A similar BSURF entry is present for the cylinder.
Note that the Bulk Data file also contains a BCPARA,0 entry. ID=0 on this entry sets parameter values
for the contact prior to the start of the analysis. If there is a BCONTACT=n in any Case Control
command, MD Nastran will automatically search for the existence of a BCPARA,0 entry to initialize the
contact algorithm.
The results for the model show that all GROUNDCHECK requirements pass. The first twenty modes are
given in the following output listing:
MD Nastran R2.1 Release Guide

20
*** USER INFORMATION MESSAGE 7570 (GPWG1D)
RESULTS OF RIGID BODY CHECKS OF MATRIX KGG (G-SET) FOLLOW:
PRINT RESULTS IN ALL SIX DIRECTIONS AGAINST THE LIMIT OF 2.674629E-03
DIRECTION STRAIN ENERGY PASS/FAIL
--------- ------------- ---------
1 1.100904E-07 PASS
2 7.325037E-08 PASS
3 4.130997E-08 PASS
4 5.545296E-07 PASS
5 1.048790E-07 PASS
6 3.550209E-07 PASS
*** USER INFORMATION MESSAGE 7570 (GPWG1D)
RESULTS OF RIGID BODY CHECKS OF MATRIX KAA1 (A-SET) FOLLOW:
PRINT RESULTS IN ALL SIX DIRECTIONS AGAINST THE LIMIT OF 2.674629E-03
DIRECTION STRAIN ENERGY PASS/FAIL
--------- ------------- ---------
1 1.096489E-07 PASS
2 7.120107E-08 PASS
3 3.338044E-08 PASS
4 5.495763E-07 PASS
5 1.451471E-07 PASS
6 5.887807E-07 PASS

R E A L E I G E N V A L U E S
MODE EXTRACTION EIGENVALUE RADIANS CYCLES GENERALIZED GENERALIZED
NO. ORDER MASS STIFFNESS
1 1 -3.343832E-05 5.782588E-03 9.203274E-04 1.000000E+00 -3.343832E-05
2 2 -1.705077E-05 4.129258E-03 6.571918E-04 1.000000E+00 -1.705077E-05
3 3 -1.452048E-06 1.205010E-03 1.917832E-04 1.000000E+00 -1.452048E-06
4 4 2.856250E-06 1.690044E-03 2.689789E-04 1.000000E+00 2.856250E-06
5 5 1.440756E-05 3.795729E-03 6.041091E-04 1.000000E+00 1.440756E-05
6 6 2.289412E-05 4.784780E-03 7.615214E-04 1.000000E+00 2.289412E-05
7 7 6.371218E+07 7.981991E+03 1.270373E+03 1.000000E+00 6.371218E+07
8 8 4.260703E+08 2.064147E+04 3.285192E+03 1.000000E+00 4.260703E+08
9 9 1.158678E+09 3.403937E+04 5.417533E+03 1.000000E+00 1.158678E+09
10 10 1.949990E+09 4.415869E+04 7.028074E+03 1.000000E+00 1.949990E+09
11 11 4.246696E+09 6.516668E+04 1.037160E+04 1.000000E+00 4.246696E+09
12 12 4.305021E+09 6.561266E+04 1.044258E+04 1.000000E+00 4.305021E+09
13 13 4.692808E+09 6.850407E+04 1.090276E+04 1.000000E+00 4.692808E+09
14 14 6.424801E+09 8.015486E+04 1.275704E+04 1.000000E+00 6.424801E+09
15 15 7.120649E+09 8.438394E+04 1.343012E+04 1.000000E+00 7.120649E+09
16 16 9.036405E+09 9.506001E+04 1.512927E+04 1.000000E+00 9.036405E+09
17 17 9.057384E+09 9.517029E+04 1.514682E+04 1.000000E+00 9.057384E+09
18 18 9.837994E+09 9.918666E+04 1.578605E+04 1.000000E+00 9.837994E+09
19 19 1.080527E+10 1.039484E+05 1.654390E+04 1.000000E+00 1.080527E+10
20 20 1.156105E+10 1.075223E+05 1.711271E+04 1.000000E+00 1.156105E+10

1 * * * END OF JOB * * *
21
CHAPTER 2
Contact in Linear and Nonlinear Solutions
Example 2: 3-D Contact (s101bmcn.dat)
This is an example of 3-D contact in SOL 101. Figure 2-2 shows two beams. The left beam is clamped
at its left end and is 10 units in length. The right beam is clamped at its right end and is 0.15 units above
the left beam. It is 10 units in length, and overlaps the right end of the left beam by two units of length.
Both beams are modeled using CQUAD4 elements. The upper right beam has a uniform pressure load
applied along its entire length. It is desired to know how the upper beam contacts the lower beam under
the assumption that the beams do not undergo large rotation, or experience large strain or nonlinear
material behavior. (If, on running such a model in SOL 101, it is determined that these assumptions are
incorrect, the model can be run in SOL 400 without changing the contact definition.)
Figure 2-2 3-D contact in SOL 101.
The following abbreviated Bulk Data file is used for discussion:
SOL 101
CEND
SUBCASE 1
BOUTPUT = ALL
BCONTACT = 1
NLPARM = 1
SPC = 2
LOAD = 3
DISPLACEMENT(SORT1,REAL)=ALL
SPCFORCES(SORT1,REAL)=ALL
STRESS(SORT1,REAL,VONMISES,BILIN)=ALL
FORCE = ALL
BEGIN BULK
MD Nastran R2.1 Release Guide

22
PARAM POST 0
PARAM PRTMAXIM YES
$
NLPARM 1 20 FNT
BCTABLE 1 1
SLAVE 3 0. 0. 0. 0. 0
1 0 0
MASTERS 4
$
$ Deform Body Contact LBC set: left
$
BCBODY 3 3D DEFORM 3 0
BSURF 3 1 2 3 4 5 6 7
8 9 10 11 12 13 14 15
16 17 18 19 20 21 22 23
24 25 26 27 28 29 30 31
32 33 34 35 36 37 38 39
40
$
$ Deform Body Contact LBC set: right
$
BCBODY 4 3D DEFORM 4 0
BSURF 4 101 102 103 104 105 106 107
108 109 110 111 112 113 114 115
116 117 118 119 120 121 122 123
124 125 126 127 128 129 130 131
132 133 134 135 136 137 138 139
140
$
$ Elements and Element Properties for left beam
$
PSHELL 1 1 .1 1 1
$
CQUAD4 1 1 1 2 23 22
CQUAD4 2 1 2 3 24 23
CQUAD4 3 1 3 4 25 24
.
.
.
$
$ Elements and Element Properties for right beam
$
PSHELL 2 2 .1 2 2
$ Pset: right will be imported as: pshell.2
CQUAD4 101 2 101 102 123 122
CQUAD4 102 2 102 103 124 123
CQUAD4 103 2 103 104 125 124
.
.
.
$
$ Material Record : left
MAT1 1 1.+6 .3 2.59-4
$ Material Record : right
MAT1 2 1.+6 .3 2.59-4
$ Nodes Left Beam
GRID 1 0. 0. 0.
GRID 2 .5 0. 0.
GRID 3 1. 0. 0.
GRID 4 1.5 0. 0.
23
CHAPTER 2
Contact in Linear and Nonlinear Solutions
.
.
.
$ Nodes Right Beam
GRID 101 8. 0. .15
GRID 102 8.5 0. .15
GRID 103 9. 0. .15
.
.
.
$ Loads and Boundary Contitions
$
SPCADD 2 1 3
LOAD 3 1. 1. 2
SPC1 1 123456 1 22 43
SPC1 3 123456 121 142 163
$ Pressure Loads right beam
PLOAD4 2 101 -.5 THRU 140
$ Referenced Coordinate Frames
CORD2R 1 0. 0. .05 0. 0. 1.05
1. 0. .05
ENDDATA
The BOUTPUT = ALL Case Control command requests contact force information.
The BCONTACT=1 Case Control command points to a BCTABLE with ID=1 and one value for the
NGROUP field of the BCTABLE entry, indicating that one pair of structures is to be glued (the number
of SLAVE,MASTERS groups).
The second line on the BCTABLE,1 Bulk Data entry identifies the SLAVE body as IDSLA1=3, and the
ISEARCH=1 value on the continuation line following the SLAVE entry line indicates to the search
algorithm that touching-to-touched is to be invoked. The IDSLA1=3 field entry points to a
corresponding BCBODY Bulk Data entry with a value of BID=3, indicating the touched body.
The Bulk Data entry MASTERS, with its IDMA1=4 field, points to a corresponding BCBODY entry
with BID=4, indicating the touched body.
The BCBODY entries show that both structures are to be considered deformable, and that the 3-D
contact algorithm is to be used. The BCBODY entry with field BID=3 has, in its field 5, BSID = 3, which
corresponds to a BSURF Bulk Data entry of the same value for BSID. The BSURF then lists the elements
of the left beam.
A similar BSURF is specified for the right beam.
One additional Case Control and Bulk Data pair shown is the NLPARM=1 Case Control command and
its corresponding NLPARM,1 Bulk Data entry. These entries are not required and MD Nastran will
generate internally, based on the presence of a BCONTACT=n command, a default NLPARM entry. The
presence of this NLPARM=1 Case Control command and its corresponding NLPARM,1 Bulk Data entry
allows the user to change the defaults as deemed necessary. In this case, 20 iterations are requested,
which is recommended for contact, as well as the full Newton technique (FNT).
MD Nastran R2.1 Release Guide

24
The following partial .f06 file shows some output unique to SOL 101 when running contact, and when
BOUTPUT is present in the Case Control Section. A status of 1 in the contact table indicates touching
nodes, and a status of 0 in the table indicates the grid points of the touched surface.
N O N - L I N E A R I T E R A T I O N M O D U L E O U T P U T
STIFFNESS UPDATE TIME-1892160000.00 SECONDS SUBCASE 1 STEP 0
ITERATION TIME 0.01 SECONDS
LOAD NO. - - ERROR FACTORS - - CONV ITR MAT NO. AVG TOTL - - - - - DISP - - - - - - LINE_S NO. TOT TOT
STEP INC ITR DISP LOAD WORK RATE DIV DIV BIS R_FORCE WORK AVG MAX AT GRID C FACT NO QNV KUD ITR
% 1.0000 20 3 4.09E-05 4.53E-03 5.91E-05 0.700 0 1 0 1.18E-05 2.731E+00 1.72E-01 -3.423E+00 42 3 1.00 0 0 0
64

0 SUBCASE 1
N O D A L F O R C E S A N D S T R E S S E S I N 3 D C O N T A C T A N A L Y S I S
LOAD STEP = 1.00000E+00 SUBCASE 1 STEP 0
GRID STATUS CONTACT FORCES - RESIDUALS FRICTION FORCES NORMAL FRICTION FRICTION
EXCLUDING FRICTION STRESS STRESS 1 STRESS 2
- IN GLOBAL SYSTEM - - IN GLOBAL SYSTEM - - IN CONTACT LOCAL SYSTEM -
17 1 1.1705E-01 -1.8131E-05 -4.4983E-01 0.0000E+00 0.0000E+00 0.0000E+00 3.8423E+00 0.0000E+00
0.0000E+00
38 1 8.6045E-02 -1.3307E-05 -3.3009E-01 0.0000E+00 0.0000E+00 0.0000E+00 1.4101E+00 0.0000E+00
0.0000E+00
59 1 1.1705E-01 1.8087E-05 -4.4983E-01 0.0000E+00 0.0000E+00 0.0000E+00 3.8423E+00 0.0000E+00
0.0000E+00
101 0 -1.1503E-01 1.7818E-05 4.4206E-01 0.0000E+00 0.0000E+00 0.0000E+00 7.5518E+00 0.0000E+00
0.0000E+00
102 0 -2.0232E-03 3.1340E-07 7.7754E-03 0.0000E+00 0.0000E+00 0.0000E+00 6.6415E-02 0.0000E+00
0.0000E+00
122 0 -8.4557E-02 1.3077E-05 3.2438E-01 0.0000E+00 0.0000E+00 0.0000E+00 2.7713E+00 0.0000E+00
0.0000E+00
123 0 -1.4867E-03 2.2991E-07 5.7031E-03 0.0000E+00 0.0000E+00 0.0000E+00 2.4363E-02 0.0000E+00
0.0000E+00
143 0 -1.1503E-01 -1.7774E-05 4.4205E-01 0.0000E+00 0.0000E+00 0.0000E+00 7.5518E+00 0.0000E+00
0.0000E+00
144 0 -2.0232E-03 -3.1263E-07 7.7753E-03 0.0000E+00 0.0000E+00 0.0000E+00 6.6414E-02 0.0000E+00
0.0000E+00
STATUS DEFINITION -----
= 0 node is a retained node
= 1 node is a tied node (touching node)
*** JOB CONVERGES FOR THE CURRENT STEP.
*** SUBCASE 1 STEP 0 IS COMPLETED.
25
CHAPTER 2
Contact in Linear and Nonlinear Solutions
Figure 2-3 shows the deformation of the two beams. Note that it appears that the two beams are not
actually touching. However, in 3-D contact, shells have actual thickness and the contact algorithm
accounts for this fact.
Figure 2-3 Deformed beams.
MD Nastran R2.1 Release Guide

26
BCONTACT=ALLBODY (Prerelease Only)
Introduction
The Case Control command BCONTACT is used to request 3-D contact analysis in SOLs 101 and 400.
The format BCONTACT = n, where n is the ID number of all corresponding BCTABLE (required),
BCMOVE (optional), and BCHANGE (optional) Bulk Data entries, is supported in MD Nastran R2. In
this release, a new format of the BCONTACT=ALLBODY Case Control command is added to support
3-D Contact.
Benefits
The use of BCONTACT=ALLBODY can save considerable time in preparing the contact input (see the
following example).
Input
Unlike BCONTACT = n, which selects the contactable bodies on the BCTABLE Bulk Data entry when
BCONTACT=ALLBODY is specified in the Case Control Section, all BCBODYs listed in the file are
selected as contactable bodies to each other. Also, since there is no BCTABLE referenced, default values
are used in the BCTABLE fields.
To specify BCMOVE or BCHANGE Bulk Data entries when BCONTACT=ALLBODY, two new Case
Control commands, BCMOVE = n and BCHANGE = n, are introduced in this release. These new
commands can also be used to overwrite the SID = n from the BCONTACT = n case.
Output
There is no new output for BCONTACT=ALLBODY.
Limitation
Although potentially convenient, it is strongly recommended to use BCONTACT=ALLBODY carefully.
It is appropriate for simple models, or for checking out runs, to use BCONTACT=ALLBODY. Setting
BCONTACT=ALLBODY without careful study may produce unacceptable results and poor
convergence.
Example
The test problem nlc021a.dat can be used as an example to show the advantage of
BCONTACT=ALLBODY. This example is a transient analysis in 2-D contact that uses four deformable
contact bodies and one rigid contact body. Each of these bodies is contactable, which yields ten possible
27
CHAPTER 2
Contact in Linear and Nonlinear Solutions
combinations of contact. With four self-contacts of the deformable bodies, excluding the rigid body,
there are a total of fourteen possible combinations.
1. Bodies 6 and 13 contact.
2. Bodies 6 and 14 contact.
3. Bodies 6 and 16 contact.
4. Bodies 6 and 17 contact.
5. Bodies 13 and 14 contact.
6. Bodies 13 and 16 contact.
7. Bodies 13 and 17 contact.
8. Bodies 14 and 16 contact.
9. Bodies 14 and 17 contact.
10. Bodies 16 and 17 contact.
11. Body 13 self-contacts.
12. Body 14 self-contacts.
13. Body 16 self-contacts.
14. Body 17 self-contacts.
When BCONTACT=1 is specified, the following BCTABLE Bulk Data entry is required, which must
include fourteen slave and master pairs.
BCTABLE 1 14
SLAVE 16 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 16
SLAVE 17 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 16
SLAVE 17 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 17
SLAVE 16 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 14
SLAVE 17 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 14
SLAVE 14 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 14
SLAVE 16 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 13
SLAVE 17 0. 0. 0. 0. 0
0 0 0
MD Nastran R2.1 Release Guide

28
FBSH 1.+20 0. 0.
MASTERS 13
SLAVE 14 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 13
SLAVE 13 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 13
SLAVE 16 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 6
SLAVE 17 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 6
SLAVE 14 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 6
SLAVE 13 0. 0. 0. 0. 0
0 0 0
FBSH 1.+20 0. 0.
MASTERS 6
Note that this long BCTABLE Bulk Data entry can be eliminated by setting BCONTACT=ALLBODY.
29
CHAPTER 2
Contact in Linear and Nonlinear Solutions
Performance Improvement to Contact Analysis
Introduction
It has been found that 3-D contact analysis may consume excessive CPU time when the model has more
than one million degrees of freedom. This deficiency has been removed.
Examples
The main model in the study used to demonstrate performance improvement has
665069 grid points;
431580 CTETRA elements;
180800 active nodes; and
no friction.
The model runs 7560 CPU seconds in MD Nastran R2. With performance improvement, which occurs
automatically, it takes 5893 CPU seconds in MD Nastran R2.1.
The expected time saved on this kind of problem is 22%.
MD Nastran R2.1 Release Guide

30
Ch. 3: Advanced Integrated Nonlinear Analysis MD Nastran R2.1 Release Guide

3
Advanced Integrated Nonlinear
Analysis
!
Advanced Integrated Nonlinear Analysis (SOL 400)
!
Kinematic Elements
!
SOL 400 Material Properties and Elements
!
The Nonlinear CQUADR and CTRIAR Elements
!
CBAR Element to CBEAM Element Conversion
!
CSHEAR Element to CQUAD4 Element Conversion
!
SOL 400 Pretension Modeling Description
!
VCCT Virtual Crack Closure Technique
!
Cohesive Zone Modeling
!
Nonlinear Iteration Algorithms
!
Analysis Chaining SUBCASE, STEP, ANALYSIS, and NLIC
!
Adaptive Time Stepping Scheme
!
The Numerical Integration Method for Transient Analysis the HHT
Method
!
Nonlinear Iteration Summary Table for Nonlinear Transient Analysis
in SOL 400
!
Output Data Grouping: NLPACK
!
Transient Temperature Loads
!
Boundary Condition (SPC and MPC) Changes in MD Nastran R2.1
!
Direct Matrix Input Changes
!
Rotor Dynamics

3
Advanced Integrated Nonlinear
Analysis
(continued)
!
Coordinate System for the Nonlinear Stresses of 3-D Elements
!
User Interfaces
!
Examples
!
Differential/Follower Force Stiffness Processing in SOL 400
!
Contact Analysis and Kinematic Elements
33
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Advanced Integrated Nonlinear Analysis (SOL 400)
Introduction
This release extends the capabilities of SOL 400, including nonlinear static analysis and nonlinear
transient analysis. In this release guide, all new items pertaining to SOL 400 are discussed. Some of the
items may have also been discussed in the MD Nastran R1 Release Guide, but are included here for
completeness and clarity.
Benefits
The benefits of SOL 400 are discussed in this section for the current release. Some of the benefits may
be subject to limitations discussed in next section. The major benefits are:
A new general nonlinear contact analysis capability.
Extensive enhancements for nonlinear large strain and material behavior.
The kinematic elements, such as RBAR, RBE2, etc., can be combined with shell and beam
elements.
The CQUADR and CTRIAR elements are extended to nonlinear analysis in SOL 400.
The linear CBAR element can be converted into a geometric nonlinear CBEAM element.
The linear CSHEAR element can be converted into a geometric nonlinear CQUAD4 element.
Improved nonlinear iteration algorithms make the solution converge more quickly. These
algorithms include the ADAPT, AUTO, ITER, SEMI, FNT, and PFNT methods.
Analysis chaining SUBCASE, STEP, ANALYSIS, and NLIC Case Control commands allow
flexible loading and solution sequences. Also, different nonlinear analysis types can now
communicate with each other in a meaningful physical sequence in a SUBCASE
An adaptive time stepping scheme is implemented into SOL 400 to automatically modify the
size of a load increment or time step.
Other benefits include:
A transient solution integration method, the Hilbert-Hughes-Taylor (HHT) method, to give a
stable transient solution.
Comprehensive nonlinear summary tables for the static analysis and the transient analysis.
Support of initial conditions (ICs) for the transient analysis.
The Case Control command NLIC allows any previous converged nonlinear static solution to be
the preload for the first nonlinear transient step.
Boundary conditions (SPC and MPC) are allowed to change between load steps.
Direct matrix input, such as K2PP, M2PP, B2PP, and TFL, is allowed to change between load
steps.
MD Nastran R2.1 Release Guide

34
Support of grid-based reordering for faster decomposition after a time step size change or
stiffness update in nonlinear iterations.
Support of thermal loads in transient analysis. Two new Bulk Data entries, TTEMP and
TMPSET, have been added for this new feature in nonlinear transient analysis.
Support of rotor dynamics in nonlinear transient analysis.
A user-friendly restart procedure.
A more flexible output method for nonlinear transient analysis. The user can specify a new
parameter, NLPACK, to control the output and restart time steps.
The solution output is at user-requested time for transient analysis. Also, the program allows the
simulation of the same output logic and format of SOL 129 by specifying a negative value for
NO on the TSTEPNL Bulk Data entry.
Support of linear superelements.
The nonlinear stresses for 3-D elements selected by NLSTRESS can now be output in a user
selected material coordinate system.
OTIME output request is supported (except NLSTRESS).
Limitations for the Current Release
The followings are limitations for the current release of nonlinear transient analysis:
Omitted degrees of freedom (o-set) may not be used.
Nonlinear normal modes and nonlinear buckling analysis are not supported.
The arc-length method (input on the NLPCI Bulk Data entry) is not supported.
Restart is not available for contact analysis and the newly enhanced large strain and material
behavior (elements with a PSHNLi, PSLNL1, or PSHEARN Bulk Data entry).
A New General Nonlinear Contact Analysis Capability
A general nonlinear contact analysis is implemented into SOL 400, which is applicable for both nonlinear
static analysis and nonlinear transient analysis. This capability can be used with existing Nastran
nonlinear elements and the newly enhanced large strain and material behavior. Both deformable-
deformable and deformable-rigid contact is allowed. For details, refer to SOL 400 Material Properties
and Elements on page 43.
Extensive Enhancements for Nonlinear Large strain and Material Behavior.
For this release, extensive enhancements for nonlinear large strain and material behavior have been
implemented in SOL 400. For details, refer to section SOL 400 Material Properties and Elements on
page 43.
35
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Kinematic Elements
Introduction
The kinematic elements are the rigid elements with Lagrange formulation used in the geometric
nonlinear analysis of large rotations. They include RBAR, RBAR1, RJOINT, RBE1, RBE2, RBE3,
RTRPLT, and RTRPLT1. This type of element becomes a linkage if any degree of freedom connecting
to the element is released. Appropriate constraints must be provided; otherwise, the structure model will
be singular and the solution will diverge. In the follow sections, the terms kinematic elements and
rigid elements are used interchangeably. The kinematic elements have been implemented into both
nonlinear static analysis and nonlinear transient analysis for SOL 400.
Benefits
The benefits of the kinematic elements are as follows:
The kinematic elements can be combined with shell, beam, and solid elements.
Differential stiffness for the kinematic elements is computed to facilitate the convergence of the
solution.
Appropriate scale factors for the Lagrange multiplier and penalty function are computed
automatically.
Thermal loads are allowed to be used with the kinematic elements.
Both force and GPFORCE output can be requested.
Some of the benefits may be subject to the limitations discussed in the following section.
Types of Elements that can be Combined with the Kinematic
Elements
The following types of elements can be combined with the kinematic elements:
Shell and beam elements CQUAD4, CQUADR, CTRIA3, CTRIA3, and CBEAM.
Solid elements CHEXA, CPENTA, and both the four-noded and ten-noded CTETRA.
The newly enhanced shell and solid elementselements with the PSHLN1 and PSLDN1 Bulk
Data entries.
The CBAR element is a linear element, but can be converted into a CBEAM element by the
Bulk Data entry MDLPRM,BRTOBM,1. Example 2 on page 39 demonstrates this.
The CSHEAR element is a linear element, but can be converted into a CQUAD4 element by the
Bulk Data entry MDLPRM,SHRTOQ4,1.
MD Nastran R2.1 Release Guide

36
The Scale Factors LMFACT and PENFN
The parameters LMFACT and PENFN are the scale factor and penalty function, respectively, for the
kinematic elements. The purpose of LMFACT and PENFN is to make the values of the stiffness matrix
of the kinematic elements, and/or the contact components, about the same in relative magnitude as those
of the other elements in the model. Too small a value will produce inaccurate results, and too large a value
will produce numerical difficulties. The same value is usually assigned to both LMFACT and PENFN.
Under special requirements, the user may assign different values for LMFACT and PENFN. For
example, if PENFN=0.0 and LMFACT=0.0, then the solution method for the rigid elements becomes the
pure Lagrange multiplier method instead of the augmented Lagrangian method. However, the user must
exercise caution if different values are assigned to LMFACT and PENFN.
The default value for the kinematic element is computed automatically by SOL 400 based on the
geometry of the individual element, and on the average magnitude of the stiffness matrix. This means
that the actual scale factor is different for each kinematic element.
Note that PARAM,LMFACT,F1 and PARAM,PENFN,F2 can be used to modify the SOL 400 computed
value.
Thermal Effects
Thermal effects for the kinematic elements are computed if thermal loading is requested. In this case, the
ALPHA field on the Bulk Data entry for the element must be entered.
RJOINT and the Pin Flags of the CBEAM Element
The Bulk Data entry for RJOINT element is:
GA is the independent grid point, and all six degrees of freedom are the independent degrees of
freedom. GB is the dependent grid point. The length between grid points GA and GB must be
zero. Since the length is zero, the thermal load effect is not applicable.
If CB=123456 or blank, then grid point GB is constrained to move with grid point GA and the
two grid points move as a single point. If any degree of freedom is released on CB, then RJOINT
becomes a mechanical joint. A mechanical joint is a mechanical system that has two bodies
jointed at a point. The two bodies can rotate relatively about one, two, or three axes of a local
coordinate system at the point. For example:
HingeA hinge is a mechanical joint that rotates freely about one axis of the local coordinate
system. It can be simulated by RJOINT with one rotational degree of freedom released; i.e.,
CB = 12356, 12346, or 12345.
Universal jointA universal joint is a mechanical joint that rotates freely in two axes. It can be
simulated by RJOINT with two rotational degrees of freedom released; i.e., CB = 1234, 1235,
or 1236.
RJOINT EID GA GB CB
37
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Spherical jointA spherical joint is a mechanical joint that rotates freely about all three axes.
It can be simulated by a zero length RBAR with all rotational degrees of freedom released;
i.e., CB = 123.
Prismatic jointA prismatic joint is a mechanical system with two blocks that are constrained
to have the same rotations, but translate relatively with each other along a local axis. It can be
simulated by RJOINT with one translational degree of freedom released; i.e., CB = 23456,
13456, 12456.
Cylindrical jointA cylindrical joint is a mechanical system that allows two grid points have
relative translation along a moving axis and, at the same, have relative rotation about the same
axis. It can be simulated by RJOINT with one translational degree of freedom and one
rotational degree of freedom released; i.e., CB = 2356, 1346, 1245.
The pin flag on the CBEAM element should not be used in nonlinear analysis. It may or may not give
correct results, depending structural model. Also, the differential stiffness for a pin flag is not computed
to facilitate convergence. Instead, the RJOINT should be used to simulate the pin flag. Example EX04
demonstrates this.
User Interfaces
Case Control Command RIGID
The Case Control command RIGID selects the type of rigid element. It has the following format:
LINEAR will select the linear rigid elements, LAGRAN will select the Lagrange rigid element with the
Lagrange multiplier method, and LGELIM will select the Lagrange element with the Lagrange
elimination method.
If there is no RIGID command in the Case Control Section, for SOL 400, the selection
RIGID=LAGRAN is assumed. Example EX03 shows that RIGID=LAGRAN is not required in the Case
Control Section.
Note that for rotor dynamics, the underlying theory requires linear rigid elements. Otherwise, the
solution will not converge. In this case, RIGID=LINEAR must be specified in the Case Control Section.
For nonlinear solution sequences, the LGELIM field entry is not available. If requested, a User Fatal
Error will occur. Also, for the nonlinear solution sequences, RIGID=LAGRAN is implemented only for
SOL 400. A User Fatal Error will be issued if it is used in SOLs 106 or 129.
Force and GPFORCE Output
The force output for the kinematic elements is requested by the MPCF Case Control command.
RIGID
LINEAR
LAGRAN
LGELIM
)

`


=
MD Nastran R2.1 Release Guide

38
The GPFORCE output is requested by the GPFOR Case Control command.
Both FORCE and GPFORCE output is shown in Examples EX01 and EX02.
Bulk Data Entries
The kinematic elements are input using the following Bulk Data entries: RBAR, RBAR1, RBE1, RBE2,
RBE3, RTRPLT, and RTRPLT1. If thermal loads are used, a value in the ALPHA field of these entries
must be specified.
Examples
The following four examples show the inputs and capabilities of the rigid elements, in combination with
other elements, for nonlinear static or nonlinear transient analysis. The intention of these examples is to
show the input structure. The model itself and the detailed entries in the Bulk Data file are not important.
Example 1
Example 1, ex01.dat, is a cantilever beam with two CQUAD4 elements connected to an RBE2 element
with two branches. The load is applied at the end of the RBE2. This is a nonlinear static geometric
analysis with 3-D motion. The force and GPFORCE output for the RBE2 are requested by MPCF=all
and GPFOR=all.
ID MSC, EX01
SOL 400
CEND
TITLE = NONLINEAR STATIC ANALYSIS
SUBTITLE = QUAD4 + RBE2 LARGE ROTATION MPCF/GPFOR
RIGID = LAGRAN $ This entry is not required.
NLPARM = 10
SUBCASE 1
LOAD = 100
SPC = 10
DISP = ALL
SPCF = ALL
MPCF = ALL
GPFOR= ALL
BEGIN BULK
NLPARM,10,10,,,,,,,+NLP1
PARAM,LGDISP,+1
FORCE,100,7,0,50.0,0.0,0.0,-1.0
FORCE,100,8,0,50.0,0.0,0.0,-1.0
GRID,1,,0.0,0.0,0.0
GRID,2,,1.0,0.0,0.0
GRID,3,,2.0,0.0,0.0
GRID,4,,0.0,1.0,0.0
GRID,5,,1.0,1.0,0.0
GRID,6,,2.0,1.0,0.0
GRID,7,,2.0,2.0,0.0
GRID,8,,3.0,2.0,0.0
MAT1,1,4.+6,,0.3
CQUAD4,11,10,1,2,5,4
CQUAD4,12,10,2,3,6,5
PSHELL,10,1,0.1,1
RBE2,21,6,123456,7,8
39
CHAPTER 3
Advanced Integrated Nonlinear Analysis
SPC1,10,123456,1,4
ENDDATA
Example 2
Example 2, ex02.dat, is a cantilever frame with six CBAR elements and an RBE3 element connected to
three independent grid points. An extra grid point and RBAR element are added to give the direction of
follower force. The CBAR element is a linear element. The entry MDLPRM, BRTOBM, 1 is added to
convert the linear CBAR element into a nonlinear CBEAM element. This is a static geometric nonlinear
analysis with 3-D motion.
ID MSC, EX02
SOL 400
CEND
TITLE = NONLINEAR STATIC ANALYSIS
SUBTITLE = BEAM + RBE3 LARGE ROTATION FOLLOWER FORCE MPCF/GPFOR
RIGID = LAGR $ This entry is not required.
NLPARM = 10
SUBCASE 1
LOAD = 100
SPC = 10
DISP = ALL
SPCF = ALL
MPCF = ALL
GPFOR= ALL
BEGIN BULK
NLPARM,10,10,,,,,,,+NLP1
MDLPRM,BRTOBM,1
PARAM,LGDISP,+1
FORCE1,100,7,100.0,17,7
GRID,1,,0.0,0.0,0.0
GRID,2,,1.0,0.0,0.0
GRID,3,,2.0,0.0,0.0
GRID,4,,0.0,1.0,0.0
GRID,5,,1.0,1.0,0.0
GRID,6,,2.0,1.0,0.0
GRID,7,,3.0,2.0,0.0
GRID,17,,3.0,2.0,1.0
MAT1,1,1.+3,,0.3
CBAR,11,10,1,2,0.0,1.0,0.0
CBAR,12,10,2,3,0.0,1.0,0.0
CBAR,13,10,4,5,0.0,1.0,0.0
CBAR,14,10,5,6,0.0,1.0,0.0
CBAR,15,10,2,5,1.0,0.0,0.0
CBAR,16,10,3,6,1.0,0.0,0.0
PBAR,10,1,1.0,0.1,0.1,0.5
RBE3,21, ,7,123456,1.0,123456,3,5,+RBE3
+RBE3,6
RBAR,31,7,17,123456, , ,123456
SPC1,10,123456,1,4
ENDDATA
Example 3
Example 3, ex03.dat, is a model with solid elements, and RBE2 and RBE3 elements. This is a nonlinear
transient analysis with 3-D motion. There is no RIGID=LAGRAN in the Case Control Section, since
RIGID=LAGRAN is the default.
MD Nastran R2.1 Release Guide

40
ID MSC, EX03
SOL 400
CEND
TITLE = NONLINEAR TRANSIENT ANALYSIS
SUBTITLE = HEXA/BEAM + RBE2/3
TSTEPNL = 10
SET 1000 = 11, 21, 221, 231, 100000
ANALYSIS=NLTRAN
SUBCASE 1
DLOAD = 200
SPC = 10
DISP = 1000
BEGIN BULK
$
TSTEPNL 10 40 0.05 AUTO
PARAM,LGDISP,+1
$
$ BOUNDARY CONDITION
$
SPC1 10 123456 1 211 232 442
$
$ RBE ELEMENTS
$
RBE3 21 100000 123456 1.0 123 21 221 +RBE3
+RBE3 231
$
RBE2 252 252 123456 463
RBE2 357 357 123456 468
RBE2 462 462 123456 473
$
$ BEAM ELEMENTS
$
PBEAM 10 1 1.0 0.1 0.1 0.0 0.5
$
CBEAM 251 10 247 252 0.0 1.0 0.0
CBEAM 256 10 352 357 0.0 1.0 0.0
CBEAM 261 10 457 462 0.0 1.0 0.0
$
CBEAM 252 10 252 357 1.0 0.0 0.0
CBEAM 357 10 357 462 1.0 0.0 0.0
$
$ SOLID ELEMENTS AND ELEMENT PROPERTIES
$
PSOLID 1 1 0
$
CHEXA 1 1 1 6 111 106 232 237
342 337
CHEXA 2 1 6 11 116 111 237 242
347 342
CHEXA 3 1 11 16 121 116 242 247
352 347
CHEXA 4 1 16 21 126 121 247 252
357 352
CHEXA 5 1 106 111 216 211 337 342
447 442
CHEXA 6 1 111 116 221 216 342 347
452 447
CHEXA 7 1 116 121 226 221 347 352
457 452
CHEXA 8 1 121 126 231 226 352 357
462 457
41
CHAPTER 3
Advanced Integrated Nonlinear Analysis
$
$ Referenced Material Records
$
MAT1 1 2.+5 .3 0.10 0.10
$
$ NODES OF THE ENTIRE MODEL
$
GRID 100000 21. 11. 0.
GRID 1 0. 0. 0.
GRID 6 5. 0. 0.
GRID 11 10. 0. 0.
GRID 16 15. 0. 0.
GRID 21 20. 0. 0.
GRID 106 0. 5. 0.
GRID 111 5. 5. 0.
GRID 116 10. 5. 0.
GRID 121 15. 5. 0.
GRID 126 20. 5. 0.
GRID 211 0. 10. 0.
GRID 216 5. 10. 0.
GRID 221 10. 10. 0.
GRID 226 15. 10. 0.
GRID 231 20. 10. 0.
GRID 232 0. 0. 1.
GRID 237 5. 0. 1.
GRID 242 10. 0. 1.
GRID 247 15. 0. 1.
GRID 252 20. 0. 1.
GRID 337 0. 5. 1.
GRID 342 5. 5. 1.
GRID 347 10. 5. 1.
GRID 352 15. 5. 1.
GRID 357 20. 5. 1.
GRID 442 0. 10. 1.
GRID 447 5. 10. 1.
GRID 452 10. 10. 1.
GRID 457 15. 10. 1.
GRID 462 20. 10. 1.
GRID 463 0 21. 0. 1.
GRID 468 0 21. 5. 1.
GRID 473 0 21. 10. 1.
$
$ LOADS FOR LOAD CASE
$
$-------2-------3-------4-------5-------6-------7-------8-------9-------0---
----
TLOAD1 200 100 0 0 500
TABLED1 500 +TAB1
+TAB1 0. 0. 1. 1. 1.2 0. 10. 0. +TAB2
+TAB2 ENDT
$
FORCE 100 100000 0 100.0 0.0 0.0 -1.0
FORCE 100 463 0 12.0 0.0 0.0 -1.0
FORCE 100 468 0 20.0 0.0 0.0 -1.0
FORCE 100 473 0 12.0 0.0 0.0 -1.0
$
ENDDATA
MD Nastran R2.1 Release Guide

42
Example 4
Example 4, ex04.dat, is a frame modeled with CBEAM, RBE2, and RJOINT elements. The large motion
of the frame is enabled by the pin ends of the CBEAM element. However, in the nonlinear analysis, the
pin flag should not be used. Therefore, the RJOINT elements are used to simulate the PIN ends. This is
a static geometric analysis with 3-D motion.
ID MSC, EX04
SOL 400
CEND
TITLE = NONLINEAR STATIC ANALYSIS
SUBTITLE = BEAM + RBE2 + RJOINT LARGE ROTATION
NLPARM = 10
RIGID=LAGRAN
LOAD = 100
SPC = 10
DISP = ALL
NLPARM,10,10,,,,,,,+NLP1
PARAM,LGDISP,+1
$-------2-------3-------4-------5-------6-------7-------8-------9-------0---
----
RBE210 3 1234562
RBE220 6 1234567
RJOINT 23 3 23 1234
RJOINT 67 6 67 1234
$-------2-------3-------4-------5-------6-------7-------8-------9-------0---
----
GRID1 0.0.0.
GRID2 5.0 0.00.
GRID3 10.0 0.00.
GRID4 20.0 0.00.
GRID5 20.0 10.00.
GRID6 30.0 0.00.
GRID7 35.0 0.00.
GRID8 40.0 0.00.
GRID23 10.0 0.00.
GRID67 30.0 0.00.
$-------2-------3-------4-------5-------6-------7-------8-------9-------0---
----
CBEAM 1 10 1 2 0.0 0.0 1.0
CBEAM 2 10 23 4 0.0 0.0 1.0
$ 156
CBEAM 3 10 4 5 0.0 0.0 1.0
CBEAM 4 10 4 67 0.0 0.0 1.0
$ 56
CBEAM 5 10 7 8 0.0 0.0 1.0
$-------2-------3-------4-------5-------6-------7-------8-------9-------0---
----
PBEAM 10 1 1.0 0.1 0.1 0.0 0.5
MAT111.+6.3
$-------2-------3-------4-------5-------6-------7-------8-------9-------0---
----
SPC1 10 123456 1 8
$-------2-------3-------4-------5-------6-------7-------8-------9-------0---
----
FORCE1002 5000.0.0-1.00.0
FORCE1005 5000.0.0 0.0 1.0
FORCE1007 5000.0.0-1.00.0
ENDDATA
43
CHAPTER 3
Advanced Integrated Nonlinear Analysis
SOL 400 Material Properties and Elements
Introduction to Nonlinear Property Enhancements
MD Nastran R2 introduces into SOL 400 extensive enhancements for nonlinear large strain and material
behavior.
Now available are orthotropic material properties for three-dimensional and plane strain behavior via the
MATORT Bulk Data entry; nonlinear gasket material properties for compression behavior via the MATG
Bulk Data entry; elasto-plastic material properties for use in large deformation analysis via the MATEP
Bulk Data entry; and failure model properties for linear elastic material via the MATF Bulk Data entry.
To take advantage of these material descriptions, the new PSHLN1, PSHLN2, and PSLDN1 Bulk Data
entries are used.
The existing composite material capabilities in MD Nastran, invoked by the PCOMP or PCOMG Bulk
Data entries, are now enhanced. They allow the use of newly added material behavior capabilities by
associating these entries with a PSHLN1 Bulk Data entry.
Composites are now also available for plane strain and axisymmetric elements via the PLCOMP Bulk
Data entry, and for layered solid composites via the PCOMPLS Bulk Data entry.
Another enhancement is the introduction of hexahedral, pentahedral, quadrilateral, and axisymmetric
quadrilateral interface elements, which may be used to simulate the onset or progress of delamination.
Cohesive material is defined using cohesive energy (also called the critical energy release rate). Mixed
mode delamination is incorporated by converting the normal and shear components of relative
displacement into an equivalent relative displacement using the shear-normal weighting factor. This
capability is invoked via the CIFPENT, CIFHEX, CIFQUAD, and CIFQDX Bulk Data entries and their
associated PCOHE and MCOHE entries.
Another new capability, called the virtual crack closure technique (VCCT), is invoked via the VCCT
Bulk Data entry, and has been introduced for evaluating energy release rates. Multiple cracks can be
defined, and results will be obtained for each crack separately. Each crack consists of a crack tip grid for
shells, and a crack front for solids. A crack is also allowed to grow. This can occur if the crack is in a
glued contact interface. The user can enter a crack growth resistance (fracture toughness) for the crack.
If the calculated energy release rate is larger than this value, the crack will grow. This is done by
automatically releasing the glued contact interface segment-by-segment.
Benefits
With these material property enhancements, MD Nastran SOL 400 is in a better position to support the
analysis of modern manufacturing techniques that use composite materials and the study of nonlinear
material behavior dealing with failure and delamination.
Furthermore, most of the finite elements in MD Nastran are now extended to include the effects of finite
strain.
MD Nastran R2.1 Release Guide

44
Input
The following Bulk Data entries are used with the new large strain, material, and virtual crack closure
technique (VCCT) enhancements:
1. PSHLN1 Bulk Data entry: This entry extends the large strain and new material capabilities to the
general shells defined by the CQUAD4, CQUADR, CQUAD8, CTRIA3, and CTRIAR elements.
This entry must have the same property ID as the PSHELL, PCOMP, or PCOMPG entry
associated with the element. If any grid of a shell element is listed on the new VCCT Bulk Data
entry, that shell element must have a PSHLN1 entry associated with it.
2. PSHLN2 Bulk Data entry: This entry extends the large strain and new material capabilities to the
two-dimensional solid plane strain, plane stress, or axisymmetric elements defined by the
CQUAD4, CQUAD8, CTRIA6, CQUAD and CQUADX elements with either four or eight grid
points, or the CTRIAX elements with six grid points. This entry must have the same property ID
as the PLPLANE entry associated with the element. These elements must lie in the basic X-Y
plane. If any grid of a two-dimensional solid element is listed on the new VCCT Bulk Data entry,
that must have a PSHLN2 entry associated with it. This replaces previous hyperelastic element
technology.
3. PSLDN1 Bulk Data entry: This entry extends the large strain and new material capabilities to
three-dimensional solid elements defined by the CHEXA and CTETRA elements. This entry must
have the same property ID as the PSOLID associated with the element. If any grid of a three-
dimensional solid element is listed on the new VCCT Bulk Data entry, that three-dimensional
solid element must have a PSLDN1 entry associated with it.
4. PLCOMP Bulk Data entry: This entry extends composites to the two-dimensional solid plane
strain, plane stress, or axisymmetric elements defined by the CQUAD4, CQUAD8, CQUAD and
CQUADX elements with either four or eight grid points.
5. PCOMPLS Bulk Data entry: This entry extends composites to the three-dimensional solid
element. A solid shell formulation is available with this entry.
6. CIFPENT, CIFHEX, CIFQUAD, and CIFQDX Bulk Data entries: These are new MD Nastran
interface elements (currently valid only in SOL 400) used to simulate the onset or progress of
delamination.
7. PCOHE Bulk Data entry: This entry provides the property interface for the CIFPENT, CIFHEX,
CIFQUAD, and CIFQDX elements.
8. MCOHE Bulk Data entry: This entry specifies material cohesive properties used to simulate the
onset or progress of delamination.
9. NLMOPTS Bulk Data entry: This entry controls parameters associated with PSHLN1, PSHLN2,
PLCOMP, PCOMPLS, and PCOHE entries.
10. VCCT Case Control command: By specifying VCCT = n, this command selects the VCCT Bulk
Data entry to be used in a given STEP.
11. VCCT Bulk Data entry: This entry specifies the use of the virtual crack closure technique for
evaluating energy release rates.
45
CHAPTER 3
Advanced Integrated Nonlinear Analysis
12. PSHEARN Bulk Data entry: This entry extends large membrane rotation to the CSHEAR
element. Stringer effectiveness is ignored. The Bulk Data entry MDLPRM,SHRTOQ4,1 cannot
be used with this entry.
13. MATORT and MATG Bulk Data entries: These existing primary material entries have been
extended for use with SOL 400. Their associated MATTORT and MATTG entries are also valid
for specifying temperature-dependent materials.
14. MATEP and MATF Bulk Data entries: These existing associated material entries have been
extended for use with SOL 400. The associated MATTEP entry is also valid for specifying
temperature-dependent materials.
15. MAT3 Bulk Data entry: This existing entry may also be used in conjunction with PSHLN2 and
PLCOMP axisymmetric elements. The associated MATT3 entry is also valid for specifying
temperature-dependent materials.
16. CQUAD, CQUADX, and CTRIAX Bulk Data entries: These existing entries have had a
(THETA/MCID) field added to their description. This new field is only applicable if the
PLPLANE entry has an associated PSHLN2 entry. If the element has only a PLPLANE property
entry, the field is ignored.
Output
Element output is obtained by using standard MD Nastran STRESS = n and NLSTRESS = n Case
Control commands. Linear formatted stress/strain, nonlinear stress, and nonlinear stress/strain output
are available.
VCCT output data is automatically placed on file OFVCCT and on the .f06 file.
Sample VCCT output is shown as follows:
Guidelines and Limitations
1. For shell elements, two-dimensional solid elements, and three-dimensional solid elements, there
are two types of property entries:
a. The primary property entries are PSHELL, PCOMP, PCOMPG, PLPLANE, PSOLID,
PLCOMP, PCOMPLS, and PSHEAR.
b. The associated property entries are PSHLN1, PSHLN2, PSLDN1, and PSHEARN.
The associated property is matched to the primary property by having the same ID.
LOAD STEP = 1.00000E+00
V C C T C R A C K R E S U L T S
CRACK TIP ------------- ENERGY RELEASE RATE ------------ ESTIMATED CRACK GROWTH DIRECTION
CRACK ID GRID ID TOTAL MODE I MODE II MODE III X Y Z
100 1 4.5493E+01 4.5493E+01 7.6309E-16 0.0000E+00 1.0000E+00 1.3297E-16 0.0000E+00
100 2 4.5484E+01 4.5484E+01 1.4533E-15 0.0000E+00 1.0000E+00 1.0262E-16 0.0000E+00
100 3 4.5484E+01 4.5484E+01 4.3923E-17 0.0000E+00 1.0000E+00 1.6459E-16 0.0000E+00
100 4 4.5493E+01 4.5493E+01 1.6856E-15 0.0000E+00 1.0000E+00 9.2419E-17 0.0000E+00
MD Nastran R2.1 Release Guide

46
The PSHLN1 entry invokes the enhanced nonlinear capability for shell elements whose PID
points to a PSHELL, PCOMP, or PCOMPG entry. The PSHLN2 entry invokes the enhanced
nonlinear capability for two-dimensional solid elements whose PID points to a PLPLANE
entry. The PSLDN1 entry invokes the enhanced nonlinear capability for three-dimensional
solid elements whose PID points to a PSOLID entry.
2. MD Nastran has two types of material entries:
a. A primary material entry, whose ID may appear on an appropriate PSHELL, PLPLANE,
PSOLID, PCOMP(G), PSHLN1, PSHLN2, PSLDN1, PLCOMP, PCOMPLS, PSHEAR, etc.,
entry.
b. An associated material entry, whose ID must appropriately match one of the primary material
IDs.
The primary material entry MATORT ID may only appear on PSHLN2, PSLDN1, PLCOMP,
and PCOMPLS. If its ID appears, say, on a PSOLID entry in the MID field, it will be ignored
and the run will fail with a no material defined error. The primary material entry MATG ID
may only appear on PSHLN2 and/or PSLDN1 entries.
If the associated materials MATEP or MATF point to a primary material ID for shell elements,
and there is no associated PSHLN1 pointing to a PSHELL, PCOMP, or PCOMPG, the
associated material will not be used. If the associated materials MATEP or MATF point to a
primary material ID for two-dimensional solid elements that have a PLPLANE as their
primary property, and there is no associated PSHLN2 pointing to a PLPLANE, the associated
material will not be used. If the associated materials MATEP or MATF point to a primary
material ID for three-dimensional solid elements that have a PSOLID as their primary
property and there is no associated PSLDN1 pointing to a PSOLID, the associated material
will not be used.
47
CHAPTER 3
Advanced Integrated Nonlinear Analysis
3. The PSHLN1 entry allows the user to change the material ID associated with the MID1 or MID2
entries, or both, on the PSHELL entry. If these entries are left blank on the PSHLN1, then the
MID1 and MID2 values on the PSHELL entry are used. The following flow diagram shows the
relationship between the PSHLN1 entry and the shell elements.
4. The PSHLN2 entry allows the user to change the material ID associated with the MID on the
PLPLANE entry. There is no default. The PLPLANE entry requires a MATHP entry, and the user
must override with a MAT1, MAT2, MAT3, MAT8, MATORT, or, if appropriate, a MATG
entry. The following flow diagram shows the relationship between the PSHLN2 entry and the
two-dimensional solid using a PLPLANE entry as its primary property entry. On the PSHLN2
entry, the BEHi codes are sensitive to the required primary material used. MAT1 is applicable to
all BEHi codes. MAT2 anisotropic and MAT8 orthotropic are applicable to BEHi=PSTRS codes
only. MAT3 axisymmetric orthotropic is applicable to BEHi=AXSOLID code only. MATORT
orthotropic is applicable to BEHi=PLSTRN code only. MATG is applicable for BEH4=COMPS
or AXCOMP with INT4=L codes only. The BEH4=COMPS or AXCOMP with INT4=L should
PSHLN1
Property ID Material Used
PSHELL PCOMP or PCOMPG MAT1
MAT2, MAT8
MAT1 MAT2, MAT8 MAT1, MAT2, MAT8 MATEP, MATF, MATS1 MATEP, MATF
MATS1
Associated Material on PSHELL
Associated Material only if PSHLN1 is Present
Optional Primary Material Points to
MD Nastran R2.1 Release Guide

48
not be used with MAT1, MAT2, MAT3, MAT8, or MATORT, as they will undergo hour-
glassing. In SOL 400, if a PLPLANE entry has an associated PSNLN2 entry, it can directly refer
to an appropriate MAT1, MAT2, etc., material entry and not have a MATHP referral. However,
in this case, all elements referring to the PLPLANE entry will fail in all other solution sequences.
PSHLN2
Property ID Material Used
PLPLANE MAT1 MAT2, MAT3, MAT8, MATORT
MATEP, MATF, MATS1 MATEP, MATF MATHP
Required entry
Associated Material only if PSHLN2 is Present
Required Primary Material Override Points to
MATG
49
CHAPTER 3
Advanced Integrated Nonlinear Analysis
5. The PSLDN2 entry allows the user to change the material ID associated with the MID on the
PSOLID. If the MID field is left blank on the PSLDN1 entry, then the MID value on the PSLOID
entry is used. The following flow diagram shows the relationship between the PSHLN2 entry and
the three-dimensional solid using a PSOLID entry as its primary property description.
6. The C3, C4, and C8 strings on the continuation fields on the PSHLN1, PSSHLN2, PSLDN1,
PLCOMP, and PCOMPLS Bulk Data entries have default integration schemes, and need not be
used if the defaults are appropriate.
7. The MATG gasket material entry requires a special integration scheme. It is available only for
elements that use a PSHLN2 or PSLDN1 Bulk Data entry. For the PSHLN2 entry, the C4
keyword entry with BEH4=COMPS or AXCOMP and INT4=L would be required. For the solid
BEH8=SLCOMP, INT8=L would be required. For example, consider an eight-noded CHEXA
element with a PSOLID entry of PID=24, and that it is desired to use a MATG entry with an MID
of 13. The following would then be required:
PSLDN1
Property ID Material Used
PSOLID MAT1 MAT9, MATORT
MATEP, MATF, MATS1 MATEP, MATF MAT1
Associated Material
Associated Material only if PSLDN1 is Present
Optional Primary Material Points to
MATG
MAT9
MATS1
PSLDN1 24 13
C8 SLCOMP L
MD Nastran R2.1 Release Guide

50
8. For composites, a solid shell element formulation is available. The DIRECT field entry must be
DIRECT=1 (the default). For the linear and quadratic formulations, no C3, C4, or C8 continuation
entry is required. The following example shows the solid shell element request:
9. Because these new material features often involve large deformation, it is recommended that a
full Newton iteration scheme be used. This is facilitated on the NLPARM Bulk Data entry by the
specification of the keyword FNT in the KMETHOD field, and on the TSTEPNL Bulk Data
entry by the addition of the keyword FNT to the METHOD field.
10. Any shell element that has nonstructural mass (NSM), and that utilizes any PSHELN1 or
PSHLN2 entry, will lose the associated nonstructural mass.
Examples
Many examples of new element materials can be found in the MD Nastran R2 Users Guide.
PCOMPLS 782 3
C8 SLCOMP ASTN
100 171 .3 12.3
101 175 .7 77.7
51
CHAPTER 3
Advanced Integrated Nonlinear Analysis
The Nonlinear CQUADR and CTRIAR Elements
Introduction
The CQUADR and CTRIAR elements have been implemented in all solution sequences except the
nonlinear solution sequences. In this release, these elements are implemented in SOL 400 only. If
CQUADR and CTRIAR are requested in other nonlinear solution sequences (SOLs 106 or 129), a user
fatal error will be issued.
Benefits
The benefits for the CQUADR and CTRIAR elements are:
The drilling DOFs are defined for the shell elements so that parameter K6ROT is not required in
a geometric nonlinear analysis.
With the drilling DOFs, drilling loads can be applied.
The membrane performance of CQUADR and CTRIAR is superior to that of CQUAD4 and
CTRIA3.
For material nonlinear analysis, four (CQUADR) or three (CTRIAR) integration points in the
plane of the element are used. This will give better results.
Nonlinear temperature composites are implemented for both the smeared and the nonsmeared
methods
Green strains are implemented for CQUADR and CTRIAR.
The stress output location can be at either corner or integration points.
Limitations
The following nonlinear analysis material models have not been implemented for the
CQUADR/CTRIAR elements:
Creep.
Hyperelasticity.
Types of Analysis
The following types of analysis have been implemented for the CQUADR and CTRIAR elements:
Material nonlinear analysisthe material models are elastoplastic and nonlinear elastic. The
method to invoke the material nonlinear analysis is the same as that of CQUAD4 and CTRIA3.
See Bulk Data entry MATS1.
Geometric nonlinear analysisthis capability is invoked using the parameter PARAM, LGDISP,
1.
MD Nastran R2.1 Release Guide

52
Geometric nonlinear analysis for temperature-dependent compositesboth the smear method and
nonsmear method are implemented. This capability is invoked using the parameter PARAM,
COMPMATT, value and PARAM, EPSILONT, value. Refer to Section 4.1, Temperature-
Dependent Composites Support Extended to Unsymmetric Laminates, MSC.Nastran Release
Guide 2005, for details. SOL 400 must be used for this capability with CQUADR and CTRIAR,
although SOL 106 was specified in the 2005 Release Guide.
The CQUADR and CTRIAR elements have been implemented into both nonlinear static
analysis and nonlinear transient analysis.
Drilling Degrees of Freedom
The CQUADR and CTRIAR elements have rotational stiffness computed for drilling degrees of freedom
(DOFs). Options are provided to deactivate the drilling DOFs. Here, deactivate means that the
rotational stiffness for drilling DOFs is not computed. CQUADR and CTRIAR without drilling DOFs
become elements similar to the CQUAD4 and CTRIA3 elements. This option is provided by the Bulk
Data entry.
MDLPRM, QR6ROT, n
where
If the drilling DOF is deactivated, a small drilling DOF stiffness may be supplied by using parameter
PARAM, K6ROT, value. By default, value = 0.0 for a membrane element, and value = 100.0 for a shell
element. Note that if the drilling DOF is active, the K6ROT parameter has no effect on the CQUADR
and CTRIAR elements.
The QR6ROT option is available all solution sequences.
Off-plane Shear Formulations
For CQUADR and CTRIAR elements, there are two methods to compute the off-plane shear stiffness:
the stiffness method and the flexibility method. The stiffness method is a new method implemented in
the CQUADR element. The flexibility method was the method implemented in the CQUAD4 element.
Therefore, if the flexibility method is selected, the solution results of CQUADR are closer to those of
CQUAD4. An option is provided to section the off-plane shear computation method by the Bulk Data
entry
n = 0 The drilling DOF is active (Default).
n = 1 The drilling DOF is deactivated for all CQUADR and CTRIAR elements in the model.
n = 2 The drilling DOF is deactivated for those CQUADR and CTRIAR elements that have
membrane stiffness only (MID2 and MID3 are blank on the PSHELL entry).
53
CHAPTER 3
Advanced Integrated Nonlinear Analysis
MDLPRM, QRSHER, n
where
The QRSHEAR option is available for all solution sequences.
Green Strain
In addition to the small strain formulation, the Green strain formulation is also computed for the
CQUADR and CTRIAR elements. An option is provided to select the strain formulation using the Bulk
Data entry
MDLPRM, GNLSTN, n
where
The GNLSTN option is available for SOL 400 only.
Stress Output Location
For CQUADR and CTRIAR elements, the stress/strain/force output location can be either at the corner
points or at the integration points. An option is provided to select the output location by the Bulk Data
entry
MDLPRM, INTOUT, n
where
The INTOUT option is available for all solution sequences.
n = 0 Use the stiffness method if on the PSHELL Bulk Data entry. Use the flexibility
method if MID3 = 0 (Default n = 0).
n = 1 Use the flexibility method (regardless of the value of MID3).
n = 2 Use the stiffness method (regardless of the value of MID3).
n = 0 Use small strain (Default).
n = 1 Use Green strain.
n = 0 Use corner point output (Default).
n = 1 Use integration point output.
MID3 0
MD Nastran R2.1 Release Guide

54
Differential Stiffness
Differential stiffness is computed for the CQUADR and CTRIAR elements in a geometric nonlinear
analysis. However, the differential stiffness is not required to obtain the correct results. If the solution
converges, the results are correct. The differential stiffness will facilitate the convergence if the axial
forces are large. This is true for problems that compute the nonlinear buckling load, or problems with a
tension stiffening effect.
There are situations in which differential stiffness hinders convergence, or even leads to solution
divergence. For this reason, SOL 400 provides an option to turn off the generation of differential stiffness.
The option is given by the Bulk Data entry
MDLPRM, NLDIFF, n
where
The NLDIFF option is available only for SOL 400, and is available for other elements such as CQUAD4
and CTRIAR3.
n = 0 Differential stiffness is computed (Default).
n = 1 Turns off computation of differential stiffness.
55
CHAPTER 3
Advanced Integrated Nonlinear Analysis
CBAR Element to CBEAM Element Conversion
The CBAR element is a linear element. It can be converted to a geometric nonlinear CBEAM by the Bulk
Data entry
MDLPRM, BRTOBM, n
where
Note that after conversion, all outputs are in CBEAM format.
n = 0 Do not converted CBAR to CBEAM (Default).
n = 1 Convert CBAR to CBEAM.
n = -1 Same as n = 1, but print the converted Bulk Data entries on the .f06 file.
MD Nastran R2.1 Release Guide

56
CSHEAR Element to CQUAD4 Element Conversion
The CSHEAR element is a linear element. It can be converted to a geometric nonlinear CQUAD4 with
membrane shear stiffness only. The conversion is specified by the Bulk Data entry
MDLPRM, SHRTOQ4, n
where
Note that for irregular shapes, CSHEAR to CQUAD4 conversion is only an approximation. Also, after
conversion, all outputs are in CQUAD4 format.
n = 0 Do not convert CSHEAR to CQUAD4 (Default).
n = 1 Convert CSHEAR to CQUAD4.
n = 2 Same as n = 1, but print the converted Bulk Data entries on the .f06 file.
57
CHAPTER 3
Advanced Integrated Nonlinear Analysis
SOL 400 Pretension Modeling Description
Introduction to Pretension Modeling
It may be appropriate to use pretension as part of the structural model. This is done with a combination
of MPC and SPC entries and the use of so-called slack variables. In SOL 400 MD Nastran, this is
automated so that the user can apply in the first load step the pretension load as a physical load, and in
subsequent steps automatically pick up the associated displacement produced by the initial load to
maintain the initial pretension under other general structural loading conditions.
Benefits
In SOL 400, an automatic pretension capability is provided. This is useful, say, in applying a pretension
to a bolt structure.
Input
The method requires standard MPC and SPC input as described in the following Guidelines and
Limitations Section.
Output
Standard MD Nastran output is available.
Guidelines and Limitations
The procedure to apply pretension in SOL 400 is as follows:
1. No SPC or SPCD with a displacement value Di other than 0.0 is allowed in a residual structure
when this method is used.
2. The user identifies a cut between two surfaces where it is desired to pretension the structure.
3. The element grids that lie above the cut and lie on the cut plane will be identified as , and should
be in the independent set of MPC equations.
4. The element grids that are connected to the elements that lie below the cut and lie on the cut plane
will be identified as , and should be in the dependent set of MPC equations.
5. The user associates with the cut a single control grid point, identified as , that will be in the
independent set of MPC equations.
6. Within the first STEP Case Control command, apply a pretension load to the control (or slack)
grid point, in the appropriate direction, to apply the desired pretension load. While either SPOINT
or GRID Bulk Data entries can be used to define a slack variable, in SOL 400, a GRID entry is
recommended.
u
t
u
b
u
s
MD Nastran R2.1 Release Guide

58
7. In the first STEP Case Control command, and all subsequent STEP commands, write a series of
MPC equations for the cut plane in the form:
One such equation is written for each top and bottom grid point along the cut plane.
8. Within the first STEP Case Control command, also apply any SPCs necessary to constrain the
overall structure against rigid body motion. The Di value on any SPC or SPCD must be 0.0
9. In the second and any subsequent steps, apply the cut plane MPCs and, additionally, add to the
existing SPC applied in the first step an SPC for each control grid point in the direction of the load
application with a Di value of 0.0
10. MD Nastran will automatically incrementally apply the displacement of the control grid point in
the direction of the pretension applied in the first step.
Examples
Figure 3-1 shows two blocks of solid elements. The lower block is a 10x2x2 mesh. The upper block is
also a 10x2x2 mesh. The three spider arrangements of MPCs will apply a pretension of 100 load units in
each spider. At the center of each spider, a control grid is introduced. The pretension load for each spider
will be applied at these control grids. These control grids were placed at the center of each spider for
convenience only. They could have been placed at any reasonable location. The following partial input
file shows important modeling features for pretensioning.
Figure 3-1 Two block sections with pretension applied by MPC at each spider.
u
b
u
t
u
s
0 =
59
CHAPTER 3
Advanced Integrated Nonlinear Analysis
In step 10, a pretension load is applied at each spider. In step 20, the structure has a thermal load applied,
and the pretension displacement at each control grid is carried forward from step 10 to step 20, and to
any subsequent steps.
SOL 400
CEND
TEMP(INIT)=3001
DISP = ALL
MPCF = ALL
SPCF = ALL
NLPARM=1
ANALYSIS=NLSTAT
MPC = 22 <----------- MPCs carried through each STEP
STEP 10
SPC = 1
LOAD = 100
STEP 20
SPC = 3
TEMP(LOAD)=6002
BEGIN BULK
$
NLPARM,1,2,,ITER,1,10,,NO,+NLPRM1
+NLPRM1,,,,,0
$
TEMPD,3001,0.
TEMPD,6002,100.
$
GRDSET,,,,,,,456
$
$ Lower 10x2x2 mesh grids
$
GRID 1 0 0.0 0.0 0.0 0
.
.
.
GRID 99 0 10.000 2.000 2.000 0
$
$ Upper 10x2x2 mesh grids
$
GRID 1067 0 0.0 0.0 2.000 0
.
.
.
GRID 165 0 10.000 2.000 4.000 0
$
$ Properties and material for all CHEXA elements
$
PSOLID 1 1 0 0 0 0
$
MAT1* 1 2.0680E+05 0.28999999166+MA 1
*MA 1 1.00000000 1.169999996E-05 +MB 1
*MB 1 1500000.00 1500000.00 68000.00 +MC 1
*MC 1
$
MATS1 1 PLASTIC 1.+7 1 1 3.+4
$
$ Lower 10x2x2 mesh elements
$
MD Nastran R2.1 Release Guide

60
CHEXA 1 1 1 2 13 12 34 35
46 45
.
.
.
CHEXA 40 1 54 55 66 65 87 88
99 98
$
$ Upper 10x2x2 mesh elements
$
CHEXA 41 1 1067 1068 1079 1078 100 101
112 111
.
.
.
CHEXA 80 1 120 121 132 131 153 154
165 164
$
$ Clamp lower section of block
$
SPC 1 1 123 0.0
.
.
.
SPC 1 89 123 0.0
$
$ Simply support upper section of block
$
SPC 1 133 123 0.0
SPC 1 143 3 0.0
SPC 1 155 13 0.0
$
$ In STEP 20 apply supports plus SLACK variables to impose Pretension
$
SPCADD312
$
$ u_bottom - u_top - u_control = 0
$
$ Pretension input required for center Spider of figure
$
$ BOLT 2
$
$ Control grid
$
GRID 2083 0 5.000 1.0 2.000 0
$
$ Preload on 2083 applied in STEP 10
$
FORCE1002083-100.0.0.1.
$
$ Displacement overlapp - often call a SLACK variable in MD Nastran
$ Disp applied in STEP 20
$ -1.598639E-02 will be picked up from STEP 10
for this SPC
$
SPC2208330.0
$
$ MPCs for Pretension on center Spider of figure Spider
$
MPC 22 1071 1 1. 71 1 -1.
61
CHAPTER 3
Advanced Integrated Nonlinear Analysis
2083 1 -1.
.
.
.
MPC 22 1095 1 1. 95 1 -1.
2083 1 -1.
$
MPC 22 1071 2 1. 71 2 -1.
2083 2 -1.
.
.
.
MPC 22 1095 2 1. 95 2 -1.
2083 2 -1.
$
MPC 22 1071 3 1. 71 3 -1.
2083 3 -1.
.
.
.
MPC 22 1095 3 1. 95 3 -1.
2083 3 -1.
$
ENDDATA
MD Nastran R2.1 Release Guide

62
VCCT Virtual Crack Closure Technique
The virtual crack closure technique (VCCT) option provides a method to calculate the energy release rate
for sharp cracks. The energy release rate is equivalent to the so-called J-integral for the linear elastic case.
The minimum required user input is the crack tip in a 2-D analysis or for a shell with a line crack and the
grids defining the crack front for the 3-D case. The capability supports line cracks in 2-D solid elements
and for shell elements and surface cracks in 3-D and for stacked shells. The VCCT results are available
on the OFVCCT file. The printed output (.f06 file) consists of the energy release rate and an estimation
of the anticipated crack growth direction. The crack growth direction is calculated using the maximum
hoop stress criterion. In addition, the output table contains three vectors defining the current crack tip
system. This system is updated with the deformations, and follows the crack tip. The vectors are available
for postprocessing.
If the crack is placed in an interface between two glued contact bodies, it can propagate along this glued
interface. This is illustrated in Figure 3-2 for a simple 2-D case. The user specifies the grid point ID of
the initial crack tip and a crack growth resistance (fracture toughness or critical energy release rate).
When, during the analysis, the calculated energy release rate is larger than the limit, the crack will grow.
The program automatically determines how much of the glued interface should be released for each load
level. If multiple choices of grid release are available, the program will choose the one that is closest to
the estimated crack growth direction.
Figure 3-2 Design of model for crack propagation.
The 3-D case is similar, but the crack is defined as a crack front of multiple grid points and the glued
interface is a surface. Each point along the crack front can grow independently.
The crack growth resistance value can be given a variation with the distance the crack has grown. This
is done by associating it with a TABLEM1 entry in the VCCT Bulk Data entry. Thus, it is possible to
model a so-called r-curve behavior, where the resistance to further growth changes (usually increases) as
the crack grows. For a 3-D model, this is traced for each crack front point separately.
glued interface
initial crack tip
first contact body
second contact body
63
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Cohesive Zone Modeling
MD Nastrans new library of so-called interface elements can be used to simulate the onset and progress
of delamination. The constitutive behavior of these elements is expressed in terms of tractions versus
relative displacements between the top and bottom edge/surface of the elements.
Considering a 3-D interface element, the relative displacement components are given by one normal and
two shear components, expressed with respect to the local element system (see Volume B for the
definition of the local element systems):
(3-1)
Based on the relative displacement components, the effective opening displacement is defined as
(3-2)
Where is called the shear-normal weighting coefficient, which equals the ratio between the critical
shear and normal tractions. Equation 3-2 is strictly valid under the assumption that the interface element
is loaded in tension ( ). At the end of this section, the modification for compressive loading will
be discussed.
The effective traction is introduced as a function of the effective opening displacement, and is
characterized by an initial reversible response followed by an irreversible response as soon as a critical
effective opening displacement has been reached. Three standard functions are currently available:
bilinear, exponential, and a combined linear-exponential function (see Figure 3-3) described as follows:
if
if (bilinear) (3-3)
if
(exponential) (3-4)
if
v
n
u
1
t op
u
1
bot t om
=
v
s
u
2
t op
u
2
bot t om
=
v
t
u
3
t op
u
3
bot t om
=
v v
n
2

2
v
s
2

2
v
t
2
+ + =

v
n
0
t
v
c
t
2G
c
v
m
----------
v
v
c
----- = 0 v v
c

t
2G
c
v
m
----------
v
m
v
v
m
v
c
--------------------
\ .
|
| |
= v
c
v < v
m

t 0 = v v
m
>
t G
c
v
v
c
2
----- e
v v
c

=
t
2qG
c
v
c
q 2 + ( )
-----------------------
v
v
c
----- = 0 v v
c

MD Nastran R2.1 Release Guide

64
linear-exponential (3-5)
if
in which is the critical energy release rate (also called the cohesive energy), is the maximum
effective opening displacement (which is only used by the bilinear model), and is the exponential
decay factor (which is only used by the linear-exponential model).
Figure 3-3 Bilinear (left), exponential (middle), and linear-exponential (right) cohesive
material model.
It can be verified that the maximum effective traction , corresponding to the critical effective opening
displacement , is given by
(bilinear) (3-6)
(exponential) (3-7)
(linear-exponential) (3-8)
So, if the maximum effective traction is known, the critical energy release rate can be determined by
(linear) (3-9)
(exponential) (3-10)
t
2qG
c
v
c
q 2 + ( )
----------------------- e
q 1 v v
c
( )
= v v
c
>
G
c
v
m
q
v v
t t
v
c
v
c
v
m
v
t
v
c
t
c
v
c
t
c
2G
c
v
m
---------- =
t
c
G
c
ev
c
-------- =
t
c
2qG
c
v
c
q 2 + ( )
----------------------- =
G
c
t
c
v
m
2
---------- =
G
c
et
c
v
c
=
65
CHAPTER 3
Advanced Integrated Nonlinear Analysis
(linear-exponential) (3-11)
In order to avoid convergence problems in a finite element simulation of delamination, the user may
activate so-called viscous energy dissipation. The basic idea of the dissipation model is that when
delamination starts, the rate of deformation may suddenly increase. This increase is used to augment the
constitutive behavior with a viscous contribution equivalent to the rate of deformation:
(3-12)
where is the viscous energy factor, is effective opening displacement rate, and is the reference
value of the effective opening displacement rate. This reference value can either be user-defined, or
calculated by the program. In the latter case, the reference value is given by the maximum effective
opening displacement rate in any interface element, as long as the response in all the interface elements
is reversible. The viscous energy dissipation model does not directly have a physical background, but is
basically numerical in nature.
As previously stated, the assumption has been that the interface element is loaded in tension. Assuming
that in compression the behavior will remain reversible, Eq. 3-2 will be adapted as
Thus far, the constitutive behavior has been discussed in terms of an effective traction versus an effective
opening displacement. The traction components follow from the effective traction according to
(3-13)
(3-14)
G
c
t
c
v
c
q 2 + ( )
2q
---------------------------- =
t
vi s
t
c
v

0
---------- =
v

0
v max v
n
0 , ( ) [ ]
2

2
v
s
2

2
v
t
2
+ + =
t
n
t
v
v
n

-------- =
t
s t ,
t
v
v
s t ,

----------- =
MD Nastran R2.1 Release Guide

66
Nonlinear Iteration Algorithms
In this section, we discuss the nonlinear iteration algorithm used in SOL 400 to obtain an equilibrium
solution.
The equilibrium solution is achieved by a combination of the following methods:
Time step size adjustment for transient analysis at the beginning of a subcase or at convergence
of a time step, the time step size is adjusted based on the estimated current natural frequency of
the structural model or the input loads.
Load iteration at each time step, the quasi-Newton method and line search technique are
employed repeatedly to obtain the balance of internal forces and external loads. This is called the
load iteration, or simply the iteration. The time step or a load increment has converged when the
balance of internal forces and external loads is obtained. The user can deselect the quasi-Newton
method (MAXQN) or line search (MAXLS) by specifying certain fields on the TSTEPNL and
NLPARM Bulk Data entries.
Stiffness update in many situations, the load iteration will not be able to achieve equilibrium at
a time step. In this situation, the stiffness matrix can be recomputed using the current geometric
and material state of the structure model to facilitate convergence. This is called the stiffness
update. Consecutive stiffness updates without load iteration is called the full Newton method.
Bisection when the program senses that it is impossible for the solution to converge, a
procedure called divergence processing is employed. One technique used in this procedure is
bisection. Bisection means cutting the time step size or load increment size by half. Divergence
processing uses a combination of stiffness updates and bisections to facilitate convergence.
Externally, bisection and time step adjustment are very similar. However, they are initiated for different
reasons. Time step adjustment is based on the requirement of the natural frequency of the structure, which
may or may not be nonlinear. On the other hand, bisection is due to large nonlinearity in the solution.
Another difference is that the time step adjustment is performed at the beginning of a subcase or at a
converged time step. Bisection is performed during a time step when the program determines that the
solution is diverging.
When the program determines the large nonlinearity is gone and the solution is stabilized toward
convergence, reversal of the bisection is performed in order to maintain the time step size required by the
preceding time step adjustment method.
The nonlinear iteration algorithm is controlled by the NLPARM and TSTEPNL entries. These entries are
as follows:
1 2 3 4 5 6 7 8 9 10
NLPARM ID NINC DT KMET KSTEP MAXI CONV INTOU
EPSU EPSP EPSW MAXD MAXQ MAXL FSTRE LSTOL
MAXBIS MAXR RTOL MINIT
67
CHAPTER 3
Advanced Integrated Nonlinear Analysis
The KMETHOD selects the nonlinear iteration method. There are six options available:
ADAPT the convergence of a time step is obtained chiefly by the time step adjustment and the
load iteration. No stiffness update is performed during the normal iteration. Stiffness updates are
performed only for divergent processing and at solution convergence of a time step with number
of iterations greater than 3*MAXITR. MAXITR is the maximum number of iterations given on
the TSTEPNL bulk entry. Even for divergence processing, the main method to correct the
divergence is bisection. This method is available for the transient analysis only.
AUTO the convergence of a load increment or a time step is achieved by automatically
selecting the load iterations and the stiffness updates, in combination with the ADAPT method.
The ADAPT method can be deselected by setting ADJUST=0 on the TSTEPNL Bulk Data
entry. This is the default method. For divergence processing, the divergence is corrected by a
combination of stiffness updates and bisections. The AUTO method always tries to maintain the
time step size required by the time step adjustment method.
ITER the convergence of a load increment or a time step is achieved by performing a stiffness
update at every KSTEP load iterations, in combination with the ADAPT method. Again, the
ADAPT method can be deselected by setting ADJUST=0. The divergence processing is the
same as that of the AUTO method.
SEMI same as the AUTO method, except that a stiffness update is performed at the first
iteration of a new time step.
FNT the full Newton method is used to obtain the convergence. For this method, the stiffness
is updated at every iteration. In comparison with the PFNT method, the defaults for FNT are
EPSU=0.01, EPSW=0.01, and MAXLS=4.
PFNT the pure full Newton method is used to obtain the solution convergence. The PFTN is
the same as the FNT method except that the defaults for PFNT are EPSU=-0.01, EPSW=-0.01,
and MAXLS=0. This means that PFNT is more stringent than FNT, and the line search is not
selected by default for PFNT.
In performing the convergence tests, three error factors are computed: the displacement error (U), the
load error (P), and the work or energy error (W), which are printed in the nonlinear iteration summary
table. These three error factors must satisfy the error tolerance rules specified on the CONV, EPSU,
EPSP, and EPSW fields.
The CONV field selects convergence criteria to be used. It can select U, P, and/or W. It is highly
recommended that CONV=UPW, which is the default, should be used. If U is selected together with P
or W, the U is not checked at the first iteration of a load increment or time step.
For U and W, there are two ways to normalize the errors: the errors are computed with respect to the total
quantities, or to the incremental quantities of a load increment. In older nonlinear solution sequences
(SOLs 106 and 129), the errors are computed with respect to the total quantities. In SOL 400, both ways
1 2 3 4 5 6 7 8 9 10
TSTEPN ID NDT DT NO KMET KSTEP MAXI CONV
EPSU EPSP EPSW MAXD MAXQ MAXL FSTRE
MAXBIS ADJUST MSTEP RB MAXR UTOL RTOL MINIT
MD Nastran R2.1 Release Guide

68
are computed. If EPSU > 0, the U error is computed with respect to the total displacements. If EPSU < 0,
the U error is computed with respect to the delta displacements of a load increment. Similarly, if
EPSW< 0, the W error is computed with respect to the total energy. If EPSW< 0, the W error is
computed to the delta energy of a load increment. If CONV=UPW, both methods give similar results for
most problems.
Another major modification in comparison to SOLs 106 or 129 is the computation of the work error. In
SOLs 106 or 129, the work error is computed based on the multiplication of the residual force and the
displacement change. During iteration, both the residual force and the displacement change become
smaller; therefore, the convergence rate of this value is proportional to the square of the convergence rate
of the solution. Thus, it becomes very small near convergence. Also, it does not have a counterpart in the
physical world. In SOL 400, the total work done to the structure model is computed at each iteration and
the work error is estimated based on the total work. In this way, the work error gives an estimation of the
error in the actual work done to the structural model. The total work for each iteration is printed on the
nonlinear iteration summary table. Note that this total work is only an approximation. Users who specify
the work error criterion and are switching from SOL 106/129 to SOL 400 may need to change the value.
The MAXITER and MAXDIV fields control how the solution will be ended if it is divergent. If
MAXDIV > 0 (default), SOL 400 will try to compute the best solutions for four consecutive
nonconvergent load increments and then terminate the run with a fatal error message. If MAXDIV <0,
the run will terminated immediately. However, in many cases, the user may want to compute the solution
to the end of a step even if the solution is divergent. In this case, setting MAXITER < 0 will result in the
solution continuing to the end of a step. The default value of MAXBIS is zero if MAXITER < 0. The
reason for this is so that the solution ends quickly without bisections.
MINITER is new field for SOL 400 only. It specifies the minimum iterations to be performed for each
load increment. The default is 1 for all cases except contact analysis, for which the default is 2.
AUTO and ADAPT are two major methods for transient analysis in SOL 400. From testing problems, it
is observed that the AUTO method tries to maintain the time step size required by the time adjustment
method or the user time step. The ADAPT method usually converges with a time step size smaller than
the time step required by the time step adjustment method or the user time step. For most problems, the
AUTO method gives better solutions; therefore, it is selected as the default method. However, if the time
step estimated by the time step adjustment method or the user time step is too large, the ADAPT method
gives a better solution. A better solution here means that the solution converges faster or does not diverge.
For highly nonlinear problems, FNT and PFNT are the recommended methods. For contact analysis and
new enhanced large strain or new material models (elements with PSHNLi, PSLDN1, or PSHEARN
Bulk Data entries), it is recommended to use FNT or PFNT. PFNT uses negative EPSU and EPSW as
defaults; therefore, PFNT is more stringent than FNT. However, if CONV=UPW, both methods give
similar results for most problems.
For transient analysis, the DT and NO fields on the TSTEPNL Bulk Data entry need further explanation.
The DT field defines the user time step size. NO (which may be positive or negative) specifies that the
results are output at every |NO| time steps. Since both time step size adjustment and bisection may modify
the time step size, options are available to select whether to output at user time step size DT, or at
internally computed time step size based upon the following:
69
CHAPTER 3
Advanced Integrated Nonlinear Analysis
If NO > 0, the output will be at the user time step size or multiples of the user time step size.
Also, the time step size computed by the time step adjustment will never be greater then the user
time step size DT. The default is NO=1.
If NO < 0, the output will be at the internally computed time step size, which may or may not be
at the user time point determined by DT. Also, the time step size computed by the time step
adjustment can be either greater than or less than the user time step size DT.
MD Nastran R2.1 Release Guide

70
Analysis Chaining SUBCASE, STEP, ANALYSIS, and
NLIC
The combination of SUBCASE, STEP, ANALYSIS and NLIC Case Control commands provide a
mechanism for defining multiple load steps, running multiple independent cases, and specifying multiple
(and mixed) types of analyses in one job.
The SUBCASE and STEP commands define load cases for a job. SUBCASE defines multiple load cases
which are independent from each other; i.e., the load history is not passed from one subcase to next. In a
subcase a number of steps can be defined. The solution of one step is a continuation of the solution of its
previous step.
The user can specify the type of analysis for each subcase and/or step by using the Case Control
command ANALYSIS. ANALYSIS supports the following two keywords for SOL 400. They are:
NLSTATIC Nonlinear static analysis (Default).
NLTRAN Nonlinear transient analysis.
NLIC selects the initial condition from a previous static analysis step for the current nonlinear transient
analysis step; see the Initial Conditions Section for details.
The following examples illustrate how the SUBCASE, STEP, and ANALYSIS commands are used.
With one SUBCASE and multiple steps, each step defines the total external load and other
characteristics for the step, which will be applied by the completion of the step. The solution of
any step is a continuation of the solution of its previous step. The following is a typical example:
SUBCASE 1 $ This line can be omitted
ANALYSIS = NLTRAN
TSTEPNL = 200
STEP 10
DLOAD = 10
STEP 20
DLOAD = 20
STEP 30
DLOAD = 30
In the preceding example, the output time will be continued from step 10 to step 20 and step 30.
Multiple subcases may be executed in one job where the types of analysis, loads, and boundary
conditions can be changed. All subcases are independent from each other; i.e., no load history
information is transmitted from one subcase to the next. At the start of each subcase, the
displacements, stresses, and strains throughout the model are zero if there is no initial condition
specified. For example:
SUBCASE 1
ANALYSIS = NLSTAT $ This line can be omitted
NLPARM = 100
STEP 110
LOAD = 110
STEP 120
LOAD = 120
71
CHAPTER 3
Advanced Integrated Nonlinear Analysis
SUBCASE 2
ANALYSIS = NLTRAN
TSTEPNL = 200
STEP 210
DLOAD = 210
STEP 220
DLOAD = 220
In the preceding example, the solutions of subcase 1 and subcase 2 are independent of each
other. If solution divergence is detected in a step, SOL 400 will terminate the solution of the
current subcase and jump to the next subcase.
A Case Control command placed below the step level allows that command to vary from one
step to another. If it is placed above the step level, the command becomes the default for all steps
in the subcase. Most of the Case Control commands, which can be placed below the subcase
level, can also placed below the step level. For example, all steps in the preceding examples use
the same Case Control command, NLPARM = 100, in subcase 1, and TSTEPNL = 200 in
subcase 2.
In the current release, NLSTATIC and NLTRAN analyses can be mixed in a single subcase. For
example:
SUBCASE 10
STEP 1
ANALYSIS = NLSTAT
LOAD = 10
NLAPRM = 110
STEP 2
ANALYSIS = NLSTAT
LOAD = 20
NLPARM = 120
STEP 3
ANALYSIS = NLTRAN
DLOAD = 30
TSTEPNL = 130
In the preceding example, subcase 10 has three steps: the first two steps request nonlinear static analysis
and the third step requests a nonlinear transient analysis. Since there is no NLIC request in the third step,
the final result of the second step automatically becomes the initial condition of the third step. Note that
in one subcase, all steps of NLSTAT must precede the steps of NLTRAN.
The meaning of multiple subcases without the STEP command is dependent on the system cell
NASTRAN SYSTEM (366) as follows:
0The solutions of all subcases are independent of each other. This is consistent with the new
SOL 400 procedure. SOL 400 will keep all subcase commands in the Case Control Section and
insert internally a step 1 for each subcase. The default is 0.
1The solution of each subcase is a continuation of the previous subcase. This is similar to the
procedure in SOLs 106 or 129. SOL 400 will internally convert all the subcase identification
numbers to STEP identification numbers and insert a SUBCASE 1 before the first STEP.
MD Nastran R2.1 Release Guide

72
Both SUBCASE and STEP are used in Examples 1, 2, 3, 4, 5, 7, and 8. SUBCASE only is used in
Example 9 on page 112. SUBCASE and STEP are not used in Example 6 on page 108.
73
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Adaptive Time Stepping Scheme
AUTO Step
An adaptive time stepping scheme has been introduced in SOL 400 using the NLAUTO method to
automatically modify the load increment size or the time step size. The primary control of the load
stepping is based upon the number of recycles needed to obtain convergence if the full Newton method
(FNT or PFNT) is used. Recycle means either a stiffness update or a load iteration. For modified Newton
methods (ADAPT, AUTO, ITER, or SEMI), both the number of recycles and the number of new stiffness
formations are taken into account.
Adaptive time stepping is invoked if an NLAUTO Bulk Data entry has the same ID as that of an
NLPARM or TSTEPNL Bulk Data entry.
Recycling Criterion
The default recycle-based criterion works as follows: The user specifies a desired number of recycles.
For most problems, it is sufficient to provide a value in the range of 3 to 5. For problems with severe
nonlinearities, or for problems with very small convergence tolerances, it may be necessary to increase
this number. This number is used as a target value for the load stepping scheme. If the number of recycles
required in the current increment is less than the desired number, the load step for the next increment is
increased. The time step increase is based on a factor, , that can also be specified by the user. Typical
values for are in the range of 1.2 to 1.5. While the time step increase is obviously more aggressive
with larger scale factors, it should be noted that there may be excessive recycling and cutbacks if sudden
nonlinearities are encountered. To avoid this, the following logic is used for higher scale factors: If the
actual number of recycles in an increment is greater than 60% of the desired number of recycles (i.e., the
current increment did not converge easily), the increase in scale factor for the next increment is limited
to 1.25 for scale factor values between 1.25 and 1.5625, and to 80% of the value for scale factors above
1.5625.
Time Step Cutback Scheme
The load step is never increased during an increment. If the number of recycles needed to obtain
convergence exceeds the desired number, the load step size is scaled back, the recycling cutback number
is incremented by 1, and the increment is performed again with the new load step. The scaleback
factor for the th cutback is taken as , where the factor is calculated from the expression

where is the maximum number of recycling related cutbacks for the increment and is calculated from
; is the time increment before any recycling related cutbacks occur for the
increment; and is the minimum possible time step for the increment. is equal to the value set by
S
u
S
u
N
r
N
r
s
Nr
s
s T
s
T
m
[ ]
2 N
rm
N
rm
1 + ( ) ( )
=
N
rm
N
rm
log
10
10
5
T
s
T
m
( ) = T
s
T
m
T
m
MD Nastran R2.1 Release Guide

74
the user ( by default) if there is no quasistatic inertial damping, and is equal to times the value
set by the user ( by default) if there is quasistatic inertial damping. The scaleback factor for any
cutback is the smaller of ( , ). This scheme guarantees that no matter what the starting time step
for an increment, the minimum time step is reached in a reasonable number of cutbacks if the increment
consistently fails to converge.
In addition, the averaged convergence ratios of the previous few iterations are stored and compared
against the current iterations residual. In case of divergence, the time step is cut down.
Exceptions
There are some exceptions to the previously outlined approach. If an increment is consistently
converging with the current load step, and the number of recycles exceeds the desired number, the
number of recycles is allowed to go beyond the desired number until convergence is achieved, or up to
the user-specified maximum number. The time step is then decreased for the next increment by . An
increment is determined to be converging if the convergence ratio was decreasing in three previous
recycles.
Special rules also apply in a contact analysis. During the recycles, the contact status can keep changing
(new nodes come in contact, nodes slide to new segments, nodes separate, etc.). Whenever the contact
status changes during an increment, a new set of contact constraints are incorporated into the equilibrium
equations, and more recycles are necessary in order to find equilibrium. These extra recycles due to
contact changes are not counted when the recycle number is checked against the desired number for
determining if the load step needs to be decreased within the increment. Thus, only true Newton-Raphson
iterations are taken into account. For the load step of the next increment, the accumulated number of
recycles during the previous increment is used. This ensures that the time step is not increased when there
are many changes in contact during the previous increment.
Post Files Output
In many analyses, it is convenient to obtain post file results at specified time intervals. This is naturally
obtained with a fixed load stepping scheme but not with an automatic scheme. Traditionally, the post
output frequency is given as every nth increment. Using the NLPARM option, you can request post
output to be obtained at equally spaced time intervals. In this case, the time step is temporarily modified
to exactly reach the time for output. The time step is then restored in the following increment.
Defaults
The defaults of the NLAUTO option are carefully chosen to be adequate in a wide variety of applications.
There are cases, however, when the settings may need to be modified. Assume that the default settings
are used, which means that the recycle-based control is active with an initial load of one percent of the
total. If the structure is weakly nonlinear, convergence is obtained in just a few recycles, and the time
steps for successive increments get progressively larger. This can lead to problems if the initially weakly
10
5
10
3
10
8
s
Nr
1 S
u

1 S
u

75
CHAPTER 3
Advanced Integrated Nonlinear Analysis
nonlinear structure suddenly exhibits stronger nonlinearities; for instance, occurrence of plasticity or
parts coming into contact. Possible remedies to this problem include:
Decreasing the time step scale factor to a smaller number so the step size does not grow so
rapidly;
Using the maximum time step to limit large steps; and
Decreasing the desired and maximum number of recycles to decrease the load step if more
recycles are needed.
Another situation is if the structure is highly nonlinear and convergence is slow. In this case, it may be
necessary to increase the desired number and maximum number of recycles. In general, there is a close
connection between the convergence tolerances used and the desired number and maximum number of
recycles.
MD Nastran R2.1 Release Guide

76
The Numerical Integration Method for Transient
Analysis the HHT Method
To solve the equation of motion for a nonlinear transient analysis, two issues must be resolved. The first
is how to integrate the equation of motion for the transient analysis, and the second is how to obtain an
equilibrium solution for the nonlinear analysis. This section discusses the numerical integration of the
equation of motion, and the section entitled Nonlinear Iteration Algorithms on page 66 discusses the
nonlinear iteration methods for obtaining the balanced solution.
For numerical integration, the second order Hilbert-Hughes-Taylor (HHT) method is used in nonlinear
transient analysis in SOL 400. It provides a user-definable parameter which defines a numerical
damping associated with higher frequency modes, but also maintains accuracy in the essential lower
frequency modes. Since is a key feature of the HHT method, it is sometimes called the HHT- method.
When , numerical damping is introduced into the system. This leads to an unconditionally stable
integration time scheme when it is in the following range:
When compared with other numerical damping methods even at (the maximum numerical
damping value), the HHT method still introduces less damping. Therefore, the solution is theoretically
more accurate. Note that when , this method is equivalent to the central finite difference method.
The numerical damping, , can be specified using parameter NDAMP (the default is -0.05 in SOL 400).
Parameter NDAMP is used in Example 1 on page 99, and Example 7 on page 109.

0 <
1
3
--- 0
1 3 =
0 =

77
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Nonlinear Iteration Summary Table for Nonlinear
Transient Analysis in SOL 400
In order to allow the user to track the solution sequence during nonlinear iteration, a detailed nonlinear
iteration summary table is output. A line for each iteration is output on the .f06 file. Printing the average
and the maximum displacements allows the user to know the solution status before the end of the job.
This is useful for large nonlinear problems. Even for small problems, the user will be able to know
approximately how the analysis of a structural model is behaving by examining this table. The summary
table is printed in real time; i.e., each line is printed in the .f06 file when the analysis for the line is
completed. Thus, for very large problems, the user can read the .f06 to monitor the progress of the
solution. Also, each line of the summary table begins with a %, which may be used to locate the table.
An example of this table is given as follows, and the descriptions of information given in the table are
shown in Table 3-1.
0 N O N - L I N E A R I T E R A T I O N M O D U L E O U T P U T
STIFFNESS UPDATE TIME 0.02 SECONDS SUBCASE 100 STEP 10
ITERATION TIME 0.00 SECONDS
- TIME STEP - - - ERROR FACTORS - - CONV ITR MAT AVG TOTL
TIME NO. BIS ADJUST ITR DISP LOAD WORK RATE DIV DIV R_FORCE WORK
1.00000E-01 10 0 1.0000 6 6.09E-05 4.59E-03 1.11E-04 0.36 0 1 2.4E-02 1.487E+00
1.10000E-01 11 0 1.0000 8 3.55E-05 8.29E-03 7.79E-05 0.33 0 1 4.9E-02 2.456E+00
1.20000E-01 12 0 1.0000 11 1.18E-05 4.51E-03 1.14E-05 0.40 0 1 3.1E-02 4.000E+00
1.30000E-01 13 0 1.0000 11 4.16E-04 1.07E-01 1.82E-04 0.38 0 1 6.9E-01 7.460E+00
- - - - - DISP - - - - - - NO. TOT TOT
AVG MAX AT GRID C QNV KUD ITR
7.60E-02 5.538E-01 10701 2 5 1 46
1.00E-01 7.398E-01 10701 2 12 1 54
1.29E-01 9.615E-01 10701 2 19 1 65
1.63E-01 1.221E+00 10701 2 16 1 76
Table 3-1 Information in Nonlinear Iteration Summary Table
TIME The current time. Starts from 0.0 at the beginning of the first STEP and
accumulates the value until the end of the last STEP.
For each STEP, the total time is determined by NDT and DT on the
TSTEPNL Bulk Data entry.
TIME STEP NO Number of time steps, including bisection. Initialized to 0 in the
beginning of each STEP.
TIME STEP BIS Number of bisections performed.
TIME STEP ADJUST The ratio of the current time increment to the original DT on the
TSTEPNL Bulk Data entry.
ITR Number of iterations at each time increment.
MD Nastran R2.1 Release Guide

78
Nonlinear Iteration Module Output
For a large problem, TIME STEP NO, TOT KUD, and TOT ITR in Table 3-1 may be too large to be
printed in the allocated fields, resulting in print overflow. If any of these values overflows, an additional
line is printed to show the offsets of these values. In the following output listing, TOT ITR of the first
line is 15,504. The offset of TOT ITR is shown as TOT ITR= 10000+XXX, where XXX is the number
shown under the TOT ITR column.
ERROR FACTORS:
DISP
LOAD WORK
There are three error factors: displacement, load, and work. In order for
an increment to converge, these factors must satisfy the error tolerance
rules specified by CONV, EPSU, EPSP, and EPSP on the TSTEPNL Bulk
Data entry.
CONV RATE Converge rate denotes how fast the solution converges for the current
increment. A value of 0.0 means fast convergence, and a value > 1.0
means that the solution will never converge.
ITR DIV Number of iteration divergences. Action to correct solution divergence
will be taken if ITRDIV > MAXDIV.
MAT DIV Number of material divergences + 1; i.e., it will be 1 if there is no material
divergence. Material divergence is due to bad creep strain or excessive
subincrements in plasticity.
AVG R_FORCE Average residual force. In order for a timestep to converge, this value
must become very small.
TOTAL WORK Accumulated total work done to the structure model. This value is only an
approximation.
DISP
AVG
MAX AT GRID C
The average displacement, the maximum displacement, and its grid point
identification number and component number.
NO. QNV Number of quasi-Newton vectors stored and used.
TOT KUD Total number of stiffness updates performed.
TOT ITR Total number of iterations performed, including the number of stiffness
updates and time steps.
Table 3-1 Information in Nonlinear Iteration Summary Table
79
CHAPTER 3
Advanced Integrated Nonlinear Analysis
For static analysis, the following example table is similar. Instead of TIME STEP, LOAD STEP and NO.
INC (number of increments) are given in the table.
0 N O N - L I N E A R I T E R A T I O N M O D U L E O U T P U T
STIFFNESS UPDATE TIME 0.02 SECONDS SUBCASE 100 STEP 10
ITERATION TIME 0.00 SECONDS
TIME STEP NO.= 0+XXX TOT KUD= 0+XXX TOT ITR= 10000+XXX
- TIME STEP - - - ERROR FACTORS - - CONV ITR MAT AVG TOTL
TIME NO. BIS ADJUST ITR DISP LOAD WORK RATE DIV DIV R_FORCE WORK
1.99906E+00 2374 0 0.0313 5 1.48E-06 3.21E-03 7.01E-07 0.21 0 1 4.3E-01 5.917E+05
1.99937E+00 2375 0 0.0313 5 1.37E-06 4.75E-03 7.24E-07 0.07 0 1 5.3E-01 5.934E+05
1.99969E+00 2376 0 0.0313 7 6.14E-07 8.13E-04 1.48E-07 0.13 0 1 9.5E-02 5.935E+05
2.00000E+00 2377 0 0.0313 5 1.40E-06 6.48E-03 2.09E-06 0.09 0 1 7.0E-01 5.918E+05
- - - - - DISP - - - - NO. TOT TOT
AVG MAX AT GRID C QNV KUD ITR
2.44E+00 2.022E+01 11100 2 12 614 5504
2.44E+00 2.020E+01 11100 2 13 614 5509
2.44E+00 2.023E+01 11200 2 16 614 5516
2.44E+00 2.026E+01 11200 2 11 614 5521
N O N - L I N E A R I T E R A T I O N M O D U L E O U T P U T
STIFFNESS UPDATE TIME 0.01 SECONDS SUBCASE 1 STEP 1
ITERATION TIME 0.02 SECONDS
LOAD NO. - - ERROR FACTORS - - CONV ITR MAT NO. AVG TOTL
STEP INC ITR DISP LOAD WORK RATE DIV DIV BIS R_FORCE WORK
0.1000 1 1 1.00E+00 1.62E+02 1.62E+02 1.000 0 1 0 1.56E+02 9.852E-01
0.1000 1 2 6.90E-01 2.66E+01 5.18E-01 0.164 0 1 0 3.12E+01 1.739E+00
0.1000 1 3 2.83E-01 1.63E+01 4.54E-01 0.227 0 1 0 1.91E+01 2.678E+00
0.1000 1 4 5.64E-02 6.76E+00 9.73E-02 0.537 0 1 0 7.91E+00 2.552E+00
0.1000 1 5 6.32E-02 5.88E+00 6.51E-02 0.549 0 1 0 6.82E+00 2.467E+00
0.1000 1 6 1.70E-01 6.23E+00 1.75E-02 0.796 0 1 0 6.99E+00 2.456E+00
- - - - - - - DISP - - - - - - - - LINE_S NO. TOT TOT
AVG MAX AT GRID C FACT NO QNV KUD ITR
1.32E-02 9.800E-02 8 1 0.36 3 0 0 1
2.31E-02 2.047E-01 6 1 0.04 1 1 0 2
2.96E-02 2.627E-01 6 1 1.00 0 2 0 3
2.89E-02 2.561E-01 6 1 1.00 0 3 0 4
2.88E-02 2.521E-01 6 1 2.00 1 4 0 5
2.97E-02 2.532E-01 6 1 2.00 1 5 0 6
MD Nastran R2.1 Release Guide

80
Output Data Grouping: NLPACK
For nonlinear transient analysis of a large problem with many time steps, the amount of data to be stored
in the database will be excessive without a scheme to group the output data. Data grouping will save CPU
time, IO time, and disk space to store the data. Another advantage is that, for a large problem, if an
accident happens in the middle of a run, data can be salvaged for a subsequent restart.
There are two types of data to be saved:
Output data are displacements, stresses, and strains, etc. at each output time step to be printed in
the .f06 file or plotted by the postprocessors. They are controlled by the NO field on the
TSTEPNL Bulk Data entry.
Restart data are used to described the material and geometric state of the structural model. These
data reconstruct the stiffness, mass, and damping matrices, and other tables required in
subsequent usage. Restart data are usually much larger than the output data if the structural
model is large. This type of data can be saved selectively without degrading the effectiveness of
the transient analysis.
For this purpose, in the current release, the user-modifiable parameter PARAM, NLPACK,N is available.
N means that SOL 400 will pack output data for N output timesteps, and restart data for the last timestep,
as a single data package. For example, if N=100 (the default), then one data package has output data for
100 output time steps and restart data for the last time step. Subsequent usage, such as a restart or an
initial condition for a later step, can be performed only at NLPACK data group boundaries.
Some values of N have special meaning:
N = -1: All output data for a step, and restart data for the end of the step, are grouped into a
single package. This is the SOL 129 grouping method. In this case, the restart can be performed
only at step boundaries.
N = 0: This is illegal.
N = 1: Each package of data on the database includes the output data for one output time step,
and restart data. This is the NLSTATIC grouping method. Therefore, for nonlinear static
analysis, the restart can be performed at each user output load increment, which is controlled by
INTOUT on the NLPARM Bulk Data entry.
NLPACK is used in Example 1.
For the nonlinear static analysis, NLPACK=1.
Restarts
The purpose of a nonlinear restart is to use the material and geometrical properties of a previously
converged solution as a new starting point to continue the analysis. This is useful to change the loading
sequence, the solution criteria, or to extend the analysis.
For SOL 400, a user-friendly restart procedure has been implemented. For restarts, the following
principles are noted:
81
CHAPTER 3
Advanced Integrated Nonlinear Analysis
The restart must be continued at a previous converged solution point in a nonlinear transient
analysis or a static analysis by specifying a SUBCASE, STEP, and/or TIME (LOADFAC). This
is accomplished by using the Case Control command NLRESTART; refer to the MD Nastran
Quick Reference Guide, Section 4A.
When the cold start is ANALYSIS=NLSTAT, it can be restarted at any user-specified output
load increment (controlled by NOUT on the NLPARM Bulk Data entry).
When the cold start is ANALYSIS=NLTRAN, it must be restarted from a saved or checkpointed
time step. The checkpoint times are dependent on the DT and NO values on the TESTEPNL
Bulk Data entry, and on PARAM, NLPACK of the cold start run. The checkpoint times are
integer multiples of (DT x NO) x NLPACK. For example, if DT=0.001 second, NO=10, and
NLPACK=100, the possible times that can be used for restart are at 1.0, 2.0, etc. If a requested
restart time does not match a checkpoint time, the closest checkpoint time will be used
The geometry and the initial material properties of the structural model cannot be modified,
because any modification to the geometry or the initial material properties would invalidate the
previous analysis and require the nonlinear solution to start from the beginning. In such cases, it
is simpler to initiate another cold start.
Performing a restart is described in the following subsections.
File Management Commands
For a restart, the data of the cold start must be made available through File Management commands. For
a nonlinear restart, two commands are needed: ASSIGN and RESTART. These two commands are
existing commands, and special requirements are not needed for SOL 400.
There are many methods to retrieve data for a restart. One method is given in the following example. For
other methods, refer to the File Management Section of the MD Nastran Quick Reference Guide, or
Chapter 12 of the MD Nastran Reference Manual.
Case Control Modifications
The presence of the Case Control command NLRESTART indicates that the current run is a restart
execution. The Case Control Section contains both subcases and steps which have been executed in the
cold start, and those that are to be executed in the restart. The first subcase, step, and/or load factor to be
executed in the restart is indicated by the options on the NLRESTART command. This is shown by the
following example:
NLRESTART SUBCASE 1, STEP 2, TIME 0.3
SUBCASE 1
STEP 1
LOAD = 10
STEP 2
LOAD = 20
STEP 3
MD Nastran R2.1 Release Guide

82
LOAD = 30
In this example, the first step through time = 0.3 of the second step has been previously executed. The
restart execution begins with time = 0.3 of the second step, and continues through the end of the third
step. If time = 0.3 is not a restart point saved by NLPACK on the cold start, SOL 400 will search for the
nearest restart point on the data base and use that point to begin the restart. For restart, the Case Control
file structure for SUBCASE and STEP commands must be the same as the cold start up to the restart
point. After the restart point, the user may modify the Case Control Section for SUBCASE and STEP
commands. For example, in this example, steps 1 and 2 must exist in the cold start. However, step 3 may
or may not exist in the cold start.
The following Case Control commands may be modified in a nonlinear restart:
Boundary conditions such as MPC and SPC.
Nonlinear solution control, NLPARM/TSTEPNL.
The LOAD requests.
Output request such as DISP and NLSTRESS.
The analysis type, ANALYSIS.
Depending on the option selected with the NLRESTART command, the nonlinear restart may be
logically divided into three types: a case restart, a step restart, or a time restart:
The case restart begins the execution with a SUBCASE. All five types of modification described
previously are legal for a case restart.
The step restart begins the execution with a STEP, which may be a new step or a previously
executed step. Although boundary condition and analysis type modifications are allowed, the
user has the responsibility to determine whether they are meaningful. Special attention should be
given to the analysis type modification; it may not be meaningful in many situations and may
therefore lead to erroneous results.
The time restart begins execution with a user-specified TIME. For a time restart, the user should
not modify the analysis type, boundary conditions, or load requests. The user needs to exercise
care when attempting other types of modification at this level. Also, in order to perform this type
of restart, the specified TIME must be at the NLPACK data group boundary. If it is not, SOL 400
will search for the nearest data boundary and use this boundary as the restart point.
Bulk Data Modifications
The Bulk Data Section for a nonlinear restart contains only those entries that are to be added to the cold
start. The deletion Bulk Data entry / cannot be used. This is to serve as a reminder that the geometry
and the initial properties cannot be modified. The user may make modifications to the Bulk Data file by
introducing new entries, which may be copies of the original entries with appropriate changes and new
identification numbers. The following list of entries can be added in a restart:
Load entries such as LOAD, FORCE, PLOAD4, and SPCD.
NLPARM and TSTEPNL entries.
83
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Boundary condition entries such as SPC, SPC1, and MPC.
Examples for cold start and restart are given in Example 4 on page 105 and Example 5 on page 106.
Initial Conditions
SOL 400 supports the traditional way of requesting initial conditions by using the Case Control
command IC, and its associated Bulk Data entry TIC, to assign initial displacements and
velocities.
In the current release, a user-friendly interface to define initial conditions is implemented. This
is the new Case Control command NLIC. It allows users to assign the result of any previous step
of NLSTAT analysis as the initial condition for the first step of NLTRAN analysis in a subcase.
The following is an example:
SUBCASE 10
STEP 1
ANALYSIS = NLSTAT
LOAD = 10
NLAPRM = 110
STEP 2
ANALYSIS = NLSTAT
LOAD = 20
NLPARM = 120
STEP 3
ANALYSIS = NLTRAN
NLIC STEP 1 LOADFAC 0.5
DLOAD = 30
TSTEPNL = 130
Step 3 uses the result of the first step at 50% load increment as the initial condition. Note that the NLIC
can only be defined at a load increment whose output flag is on an available restart point in static
analysis. Otherwise, a User Fatal Message will be issued and the job will be terminated. Here, SOL 400
will not search for the nearest available restart point, (the user must specify the precise restart point).
In the preceding example, if LOADFAC,0.5 is left out, then step 3 will take last state of step 1 as its
initial condition. In this case, the job will always run because the last state of a step is always a restart
point.
For NLIC, there are some rules and limitations:
SOL 400 requires that all steps of NLSTAT must be located before all steps of NLTRAN. In
other words, analysis change can only occur once between NLSTAT and NLTRAN in one
subcase.
MD Nastran R2.1 Release Guide

84
In one subcase, the beginning time of a transient step will be reset to 0.0 when the analysis type
is changed from a static step to transient step (or an NLIC Case Control command is detected).
If the user does not specify an NLIC command when the analysis type changes from NLSTAT to
NLTRAN, the final result of the last step of NLSTAT will be used as the initial condition.
The NLIC (or IC) can only appear in the first transient analysis step (ANALYSIS=NLTRAN) in
a subcase. Otherwise, it will be ignored.
The new Case Control command NLIC can only specify a restartable NLSTAT location as the
initial condition. (See the Restart section for the definition of Restartable.)
It is not allowed to use NLIC to select the initial condition from any previous step of NLTRAN;
SOL 400 will issue a User Fatal Message, and the user should run a restart job instead.
In the same step, the NLIC cannot appear together with an IC. A User Fatal Message will be
issued when NLIC and IC appear in the same step.
The NLIC command can only be used in SOL 400 (NONLIN).
Parameter ICOPT is used with the NLIC and IC Case Control Commands. At the beginning of an
NLTRAN step, the input loads may or may not be in equilibrium with the results of the previous preload
step. When ICOPT = 0, SOL 400 will compute the initial acceleration based on the users inputs.
Otherwise, it will be assumed that the initial acceleration is null. In other words, when ICOPT=1 (the
default), it is assumed that the whole structure is in equilibrium automatically. Theoretically, ICOPT = 0
gives better performance, since it guarantees equilibrium in the beginning to avoid a suddenly jump of
loads or displacements. The drawback of ICOPT = 0 comes from the characteristics of the mass matrix,
whose inverse matrix is required when computing the initial acceleration. The mass matrix is usually
highly singular for a lumped mass matrix, or for a model with only solid 3-D elements, and a large
amount of CPU time may be required. Therefore, the accuracy of the result may be in doubt.
An alternative way to resolve the problem of ICOPT = 1with a suddenly large jump of loads or
displacements is to insert an NLTRAN step in between the analysis step and its previous preload step.
This step will have very short time duration in comparison with the analysis step, and will provide a
transition between the preload step and the analysis step.
NLIC is used in Example 4 on page 105 and Example 5 on page 106. ICOPT is also used in these
two examples.
85
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Transient Temperature Loads
A new time-dependent dynamic thermal effect capability is now available in SOL 400. This capability
is invoked with the Case Control command ANALYSIS=NLTRAN, and is applied to all nonlinear
elements in the residual.
The time-dependent thermal-elastic equation can be written as follows:
where
For all nonlinear elements, the temperature effect in both static and transient analysis is directly handled
as thermal strain in SOL 400 when computing the element forces.
For nonlinear transient analysis, two new Bulk Data entries are created for the current release. They are
TTEMP and TMPSET. TTEMP defines a time-dependent dynamic thermal field , which includes
a spatial temperature distribution (TMPSET) and a time function (TABLEDi) in the same form as
TLOAD1. TMPSET defines the spatial distribution by referencing a set of grid points. The temperatures
of these grid points are defined by TEMPD, TEMP, TEMPP1, or TEMPRB entries in the normal
way. See the Bulk Data Section in the MD Nastran Quick Reference Guide for the details of these two
new entries. By using TTEMP and TMPSET, the entire model can be separated into finite subregions,
and each subregion can have its own temperature distribution pattern. If necessary, the user can also
make every grid point an independent subregion, or make the whole model into a single subregion.
As in nonlinear static analysis, TEMP(INIT) and TEMP(LOAD) commands are used in the Case Control
Section to define the temperature input in nonlinear transient analysis. The SID of TEMP(LOAD) can
refer to TTEMP to define the transient temperature load for a STEP. The spatial temperature distribution
defined by TEMP, TEMPD, etc., must have the same SID as that of the associated TTEMP. If
TEMP(INIT) refers to TTEMP entry, only the spatial temperature distribution of the entry is used, and
the time function is ignored.
The Case Control command TEMP(LOAD) can also refer to a spatial distribution (TEMP, etc.) directly
without a TTEMP Bulk Data entry. In this case, the temperature time functions are linearly interpolated
= The thermal strain;
= The current temperature defined by
where is the temperature field and is the time function;
= The reference temperature;
= The stress free temperature (initial temperature); and
= The coefficient of thermal expansion.

T
t ( ) T t ( ) ( ) = T t ( ) T
ref
( ) T
0
( ) T
0
T
ref
( )

T
t ( )
T t ( )
T t ( ) T
p
{ }f t ( ) =
T
p
{ } f t ( )
T
ref
T
0
T ( )
T f ( )
T
p
( )
MD Nastran R2.1 Release Guide

86
for the current step by using the last values of the previous step and the values of the spatial distribution
referenced by the TEMP(LOAD) command.
The thermal effects computed by the previously described method depend on the current material state
and geometric shape of the structural model. Therefore, they are called the nonlinear transient
temperature loads.
For all upstream superelements and all linear elements in the residual, the thermal effects are computed
using the conventional method the user can specify the DLOAD Bulk Data entry to combine multiple
TLOAD1 and TLOAD2 entries, whose EXCITE_ID references thermal loads. The DLOAD Bulk Data
entry must be referenced by a DLOAD Case Control command to be selected for analysis. The transient
temperature loads computed by this method depend on the initial stiffness matrix only, and are called the
linear transient temperature load. The TEMP(LOAD) and all its corresponding temperature-related Bulk
Data entries previously introduced can only describe the thermal effect for the nonlinear elements in the
residual. If there is no DLOAD Case Control command to select the temperature load for the linear
temperature load, the temperature effect for the linear part of the structure will be lost.
An example of the temperature load is given in Example 3 on page 101.
87
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Boundary Condition (SPC and MPC) Changes in
MD Nastran R2.1
Introduction
In SOL 400, SPC and MPC Case Control command IDs are allowed to change from one step to the next.
This is accomplished by placing the SPC or MPC Case Control command under the step level. In this
section, we will clarify the meaning of boundary condition changes, and describe the method to request
each type of boundary change.
Relative or Total Motion of a Step
There are two types of boundary conditions: relative and total. A relative boundary condition requests
the relative motion of the current step with respect to the motion of the previous step. This can be
illustrated by the following example:
Example: Relative Boundary Condition Changes
In this example, no SPC is prescribed for grid point 2 in the first step, and thus it is free to displace (0.02
for component 1). In step 2, all DOFs of grid point 2 are constrained. The displacement of component 1
for grid point 2 is frozen at 0.02.
For total boundary condition change, the displacement of a component is constrained to a specific value
regardless of the displacement of this component in the previous step. This can be illustrated by the
following example.
ANALYSIS=NLSTAT
NLPARM = 110
STEP 1
LOAD = 800
SPC = 200 $ grid 2 free to move, disp for grid 2 component 1 is 0.02
STEP 2
LOAD = 900
SPC = 400 $ grid 2 SPCD, disp for grid 2 component 1 is still 0.02,
$ which is frozen. The SPC change is incremental
BEGIN BULK
.
SPC1,200,1,1 $ no SPC requested on grid 2.
SPC1,400,1,1,2 $ SPC is requested on grid 2 for relative motion.
.
ENDDATA
MD Nastran R2.1 Release Guide

88
Example: Total Boundary Condition Changes
In this example, the motion of component 1 of grid point 2 is constrained to 0.0, although this component
moved to 0.02 in step 1.
Method to Specify Boundary Condition Changes
In SOL 400, the boundary condition change for an MPC for a step is always relative. There is no other
option.
For an SPC, the user can request either the relative boundary condition change or the total boundary
condition change as follows:
SPC1 and SPCR requests the relative boundary condition changes.
SPC and SPCD requests the total boundary condition changes.
Guidelines
For MD Nastran R2.1, SPC, SPC1, and SPCD are available for both nonlinear static analysis and
nonlinear transient analysis. SPCR is available for static analysis only.
SPC and SPC1 are requested by the SPC Case Control command, while SPCD and SPCR are
requested by the LOAD or DLOAD Case Control commands. A degree of freedom referenced
by SPCD must be also on an SPC or SPC1 entry. A degree of freedom referenced by SPCR must
be also on a SPC1, but cannot be on an SPC.
SPCR, SPCD, SPC, and SPC1 can be mixed in a step.
A User Fatal Message will be issued if (1) the degree of freedom specified on SPCD or SPCR is
not in the s-set, or (2) the same degree of freedom is specified on both SPCR and SPCD (or
SPC).
SUBCASE 100
ANALYSIS=NLSTAT
NLPARM = 110
STEP 1
LOAD = 800
SPC = 200 $ grid 2 free to move, disp for grid 2 component 1 is 0.02
STEP 2
LOAD = 900
SPC = 600 $ grid 2 SPCD, disp for grid 2 component 1 is 0.0,
$ This SPC change is total.
BEGIN BULK
.
SPC,200,1,1,0.0 $ no SPC requested on grid 2.
SPC,600,1,1,0.0
SPC,600,2,1,0.0 $ Grid 2 component 1 is constrained to be 0.0.
.
ENDDATA
89
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Example
The following example has three subcases; each has five steps to control the enforced motion. All three
subcases produce the same results by different combinations of SPC, SPC1, SPCD, or SPCR.
The purpose of the three subcases are:
Subcase 1 SPCD and SPC combination.
Subcase 2 SPCR and SPC1 combination.
Subcase 3 SPCD, SPCR, and SPC combination
ID MSC, NLSSPC1
SOL 400
TIME 10
DIAG 8,15
CEND
TITLE=SPCD/SPC COMBINATION TEST (NLSPCD=1: Force Method) NLSSPC1
SPCF=ALL
DISPL=ALL
$STRESS(PRINT,PUNCH)=10
$GPFORCE=ALL
$ESE=ALL
NLPARM=1
ANALYSIS=NLSTAT
$
SUBCASE 1
SUBTI=SPCD/SPC CREATE {Ys} INCLUDING Di BLANK(=0)
STEP 10
LABEL= T3(5)=0.01, T3(6)=0.01
SPC=110
LOAD=110
STEP 20
LABEL= T3(5)=0.01, T3(6)=0.01
SPC=120
LOAD=120
STEP 30
LABEL= T3(5)=0.04, T3(6)=0.04
SPC=130
LOAD=130
STEP 40
LABEL= T3(5)=0.00, T3(6)=0.00
SPC=140
LOAD=140
STEP 50
LABEL= T3(6)=0.00, T3(8)=0.01; S-SET Change
SPC=150
LOAD=150
$
SUBCASE 2
SUBTI=SPCR/SPC1 CREATE {Ys} - RESULTS ARE SAME AS SUBCASE 1
STEP 10
LABEL= T3(5)=0.01, T3(6)=0.01
SPC=210
LOAD=210
STEP 20
MD Nastran R2.1 Release Guide

90
LABEL= T3(5)=0.01, T3(6)=0.01
SPC=220
LOAD=220
STEP 30
LABEL= T3(5)=0.04, T3(6)=0.04
SPC=230
LOAD=230
STEP 40
LABEL= T3(5)=0.00, T3(6)=0.00
SPC=240
LOAD=240
STEP 50
LABEL= T3(6)=0.00, T3(8)=0.01; S-SET Change
SPC=250
LOAD=250
$
SUBCASE 3
SUBTI=SPCD/SPCR/SPC CREATE {Ys} - RESULTS ARE SAME AS SUBCASE 1
STEP 10
LABEL= T3(5)=0.01, T3(6)=0.01
SPC=310
LOAD=310
STEP 20
LABEL= T3(5)=0.01, T3(6)=0.01
SPC=320
LOAD=320
STEP 30
LABEL= T3(5)=0.04, T3(6)=0.04
SPC=330
LOAD=330
STEP 40
LABEL= T3(5)=0.00, T3(6)=0.00
SPC=340
LOAD=340
STEP 50
LABEL= T3(6)=0.00, T3(8)=0.01 ; S-SET Change
SPC=350
LOAD=350
$
BEGIN BULK
$mdlprm,nlspcd,0
$PARAM,GPFORCE,1
PARAM,GRDEQ,0
PARAM,LGDISP,1
NLPARM,1,4,,AUTO,,,,YES
,,1.-6,1.-6
$
$ 8 NODE CHEXA MODEL
$
GRID 1 0. 0. 0.
GRID 2 1.
GRID 3 1. 1.
GRID 4 1.
GRID 5 10.
GRID 6 10. 1.
GRID 7 10. 1. 1.
GRID 8 10. 1.
$
$ TEST SPC total_Di
$
91
CHAPTER 3
Advanced Integrated Nonlinear Analysis
SPC1 110 123 1 2 3 4
SPC 110 5 3 0.01
SPC 110 6 3 0.01
$
$ Test SPC1 as del_Di=0
$
SPC1 120 123 1 2 3 4
SPC1 120 3 5 6
$
$ TEST SPCD total_Di
$
SPC1 130 123 1 2 3 4
SPC1 130 3 5 6
SPCD 130 5 3 0.04
SPCD 130 6 3 0.04
$
$ TEST SPCD OVERWRITE SPC total_Di and BLANK Di=0.0
$
SPC1 140 123 1 2 3 4
SPC 140 5 3 0.01
SPC 140 6 3 0.01
SPCD 140 5 3 0.0
SPCD 140 6 3
$
$ SPC/SPCD MIXED CASE + S-SET CHANGE
$
SPC1 150 123 1 2 3 4
SPC 150 6 3 0.01
SPC 150 8 3 0.01
SPCD 150 6 3 0.0
$
$ TEST SPCR del_Di (1)
$
SPC1 210 123 1 2 3 4
SPC1 210 3 5 6
$SPC 210 5 3 0.01 <=== FATAL ERROR
$SPC 210 6 3 0.01 <=== FATAL ERROR
SPCR 210 5 3 0.01
SPCR 210 6 3 0.01
$
$ TEST SPCR del_Di (2)
$
SPC1 220 123 1 2 3 4
SPC1 220 3 5 6
SPCR 220 5 3
SPCR 220 6 3
$
$ TEST SPCR del_Di (3)
$
SPC1 230 123 1 2 3 4
SPC1 230 3 5 6
SPCR 230 5 3 0.03
SPCR 230 6 3 0.03
$
$ TEST SPCR del_Di (4)
$
SPC1 240 123 1 2 3 4
SPC1 240 3 5 6
SPCR 240 5 3 -0.04
SPCR 240 6 3 -0.04
MD Nastran R2.1 Release Guide

92
$
$ TEST SPCR del_Di (5)
$
SPC1 250 123 1 2 3 4
SPC1 250 3 6 8
SPCR 250 6 3
SPCR 250 8 3 0.01
$
$ TEST SPC/SPCR - MIXING tatal_Di and del_Di
$
SPC1 310 123 1 2 3 4
SPC1 310 3 6
SPC 310 5 3 0.01
SPCR 310 6 3 0.01
$
$ TEST SPC1/SPCR - BOTH ARE del_Di
$
SPC1 320 123 1 2 3 4
SPC1 320 3 5 6
SPCR 320 6 3 0.0
$
$ TEST SPCD/SPCR - MIXING tatal_Di and del_Di
$
SPC1 330 123 1 2 3 4
SPC1 330 3 5 6
SPCD 330 5 3 0.04
SPCR 330 6 3 0.03
$
$ TEST SPCD/SPCR - MIXING tatal_Di and del_Di
$
SPC1 340 123 1 2 3 4
SPC1 340 3 5 6
SPCR 340 5 3 -0.04
SPCD 340 6 3
$
$ TEST SPCR del_Di (5)
$
SPC1 350 123 1 2 3 4
SPC1 350 3 8
SPC 350 6 3 0.01
SPCD 350 6 3 0.0
SPCR 350 8 3 0.01
$
CHEXA 1 8 1 2 3 4 5 6 +CHEX1
+CHEX1 7 8
$
$ COMMON DATA FOR EACH PROBLEM
$
PSOLID 8 1 0
MAT1 1 3.+7 0.3
MATS1 1 PLASTIC 1.+7 1 1 3.+4
GRDSET 456
ENDDATA
93
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Direct Matrix Input Changes
In SOL 400 nonlinear transient analysis, the direct input matrices K2PP, M2PP, B2PP, and TFL are
allowed to change between steps. This is accomplished by placing the related Case Control commands
below step level.
An example of the direct matrix changes is given in Example 7 on page 109.
MD Nastran R2.1 Release Guide

94
Rotor Dynamics
Nonlinear transient rotor dynamics is implemented in SOL 400. For details, refer to Chapter 4, Rotor
Dynamics, in the MSC.Nastran Release Guide V2004.
A rotor dynamics example is given in Example 6 on page 108.
95
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Coordinate System for the Nonlinear Stresses of 3-D
Elements
Nonlinear stresses are selected by the NLSTRESS Case Control command. The coordinate system for
the nonlinear stresses of 3-D elements was previously limited to the element coordinates. In previous
releases, no user options were allowed. In this release, the coordinate system for nonlinear stresses can
be output in a user-selected material coordinate system. This coordinate system is defined by CORDM
on the PSOLID Bulk Data entry.
Note that the default coordinate system for CORDM is the basic coordinate system (0) instead of the
element coordinate system; therefore, the values for nonlinear stresses will change for any previous
analyses.
Outputs
Output is requested by using Case Control commands. All existing Case Control commands such
DISPLACEMENT, VELOCITY, ACCELERATION, STRESS, NLSTRESS, OLOAD, SPCFORCE,
etc., are also allowed in SOL 400 nonlinear analysis.
Two special outputs, Nonlinear Iteration Summary Table and PARAM, PH2OUT, are available for
both nonlinear static and transient analysis in SOL 400. In addition, a new output control, PARAM,
NLPACK, n, is added for the current release of nonlinear transient analysis. For static analysis, only
PARAM, NLPACK, 1 is allowed, and it is the default.
The new parameter NLPACK (100 is the default for NLTRAN), is used to control packed output in SOL
400. The value of NLPACK represents the total number of output time steps in one output package. SOL
400 processes the output procedure only after collecting all NLPACK output time steps, or at the end of
each STEP. For more detail, see the preceding section Output Data Grouping: NLPACK on page 80.
Because the volume of output created by the printing of matrix and table trailers may be extremely large,
the diagnostic output requests of DIAG 8 and DIAG 15 are automatically turned off when preparing the
output data in the solver of SOL 400 if NLPACK,1 is specified. This action can greatly reduce the .f04
file output size, especially when large output timesteps are requested. The user can turn them on by
setting DIAG 56, but it is not recommended.
Error Handling
In SOL 400, there are three types of fatal errors:
User or system fatal errors;
Fatal errors due to solution divergence; and
Fatal errors due to insufficient CPU time.
For user or system fatal errors, if the error occurs before the solution iteration phase, the run will
terminate immediately without outputting the stored data. If the error occurs during the solution iteration
MD Nastran R2.1 Release Guide

96
phase, SOL 400 will try to output all stored data for the current subcase and terminate the run. The
solution will not continue into the next subcase if there are multiple subcases.
For fatal errors due to solution divergence, SOL 400 will try to output all stored data for the current
subcase and terminate the subcase. The solution sequence will jump to attempt the next subcase if there
are multiple subcases.
For fatal errors due to insufficient CPU time, SOL 400 will attempt to output all stored data and terminate
the run.
97
CHAPTER 3
Advanced Integrated Nonlinear Analysis
User Interfaces
User interfaces which are new to nonlinear transient analyses in SOL 400 are summarized in this section.
For details, please refer to the MD Nastran Quick Reference Guide.
Nastran System Cells
The system cell STPFLG (SYSTEM (366)) selects the subcase or step layout when there are a number
of SUBCASE commands, and no STEP command in the Case Control Section.
The system cell TZEROMAX(SYSTEM (373)) controls initial time step adjustment in nonlinear
transient analysis.
File Management Commands
The following File Management statements are required for restarts. Refer to the File Management
Section of the MD Nastran Quick Reference Guide, or Chapter 12 of the MD Nastran Reference Manual,
for details.
ASSIGN Assigns physical file names to database files that are used by a Nastran data file to run a job.
RESTART Requests that data stored in a previous run be used in the current run.
Executive Control Statement
The Executive Control statements SOL 400 or SOL NONLIN request the SOL 400 advanced integrated
nonlinear solution sequence.
Parameters
PARAM, LGDISP Requests a geometric nonlinear analysis. Default = 2 (for no geometric nonlinear
effect).
PARAM, FOLLOWK Requests whether the follower force stiffness will be used in a geometric
nonlinear analysis. Default = YES.
PARMAM, FKSYMFAC Controls whether the symmetrical follower force stiffness will be used in a
geometric nonlinear analysis. Default = 0.24.
PARAM, LMFACT and PENFN Modifies the scale factors to be used in kinematic elements and/or
contact analysis.
PARAM, MAXLP Specifies the maximum number of iterations for element relaxation and the material
point subincrement process. Default = 10.
PARAM, NLAYERS Specifies the number of layers for integration through shell thickness of the
material nonlinear properties. Default = 5 for all elements except CQUADR and CTRIAR. For
CQUADR and CTRIAR, the default = 7.
MD Nastran R2.1 Release Guide

98
PARAM, NLTOL Selects defaults for CONV, EPSU, EPSP, and EPSW for the Bulk Data entries
NLPARM and TSTEPNL. Default = 2.
PARAM, PH2OUT Requests phase II outputs for a nonlinear analysis. Default = 0.
PARAM, NLPACK Controls the total output time step in one output package; see Output Data
Grouping: NLPACK on page 80. Default = 100.
PARAM, NDAMP Specifies the values when using the HHT- method in SOL 400. Default = -0.05.
PARAM, ICOPT Selects how to handle the equilibrium when dealing with the initial condition; see
Initial Conditions on page 83.
Case Control Commands
ANALYSIS Selects solution method for an analysis step.
NLIC Selects the initial condition from any static analysis for nonlinear transient analysis
NLPARM Selects the NLPARM Bulk Data entry.
NLRESTART Requests a restart execution at a specific solution point for SOL 400; see Restarts on
page 80.
NLSTRESS Requests the form and type of the nonlinear element stress output.
STEP Delimits and identifies an analysis step.
TSTEPNL Selects the TSTEPNL Bulk Data entry.
VCCT Selects the VCCT Bulk Data entry.
Bulk Data Entries
MATHP Specifies hyperelastic material properties for an element.
MATS1 Specifies stress-dependent material properties for an element.
MDLPRM Defines various model parameters for the analysis.
NLPARM Defines a set of parameters for the nonlinear static analysis iteration strategy.
TSTEPNL Defines a set of parameters for the nonlinear transient analysis iteration strategy.
TTEMP Defines a time-dependent temperature distribution for use in nonlinear transient response.
TMPSET Defines a spatial temperature distribution for use with the TTEMP Bulk Data entry.
VCCT Selects the virtual crack closure technique to evaluate energy release rates.
Also, the Bulk Data entries for the newly enhanced large strain and material models. Refer to SOL 400
Material Properties and Elements on page 43 for descriptions of these entries.
99
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Examples
The following nine examples show input formats and capabilities of SOL 400 nonlinear static and
transient analysis. The model itself, and the detailed entries in the Bulk Data file, are not important.
Example 1
Example 1, EX01, is a simplified version of a standard test problem file, NLTSUB02. This model has
only CQUAD4 elements. It has both material nonlinearity (MATS1) and geometrical nonlinearity
(PARAM, LGDISP, 1). The first step processes the output data at every five output time steps, and the
second step processes output only once because of the settings of the parameter NLPACK. All bold-font
statements are entries pertaining to the nonlinear analysis.
ID MSC, EX01 $
TIME 150 $
SOL 400 $
CEND
TITLE=ISOTROPIC MATERIAL & MATS1, ELLIPTIC CYLINDER UNDER EX01
SUBTITLE =SPC CHANGE IN EACH STEP, NLPACK's
SET 10 = 10000,11200
SET 20 = 101
SEALL = ALL
DISPL = ALL
STRESS = 20
$
SUBCASE 100
ANALYSIS=NLTRAN
STEP 10
PARAM,NLPACK,5
DLOAD = 100
SPC = 200
TSTEPNL = 310
STEP 20
PARAM,NLPACK,-1
DLOAD = 100
SPC = 400
TSTEPNL = 320
$
BEGIN BULK
PARAM NDMAP -0.05
PARAM LGDISP 1
TSTEPNL 310 100 0.01 10 AUTO
TSTEPNL 320 100 0.01 10 AUTO
$
PLOAD4 510 101 5. THRU 112
$
TLOAD1 100 510 0 0 120
TABLED1 120 +TBD1
+TBD1 0. 0. 5. 1. 16. 1. ENDT
MAT1 100 3.+7 0.3 .283-2
MAT1 101 3.+7 0.3 .283-2
MATS1 100 PLASTIC 3.+5 500000.
$
GRID 10000 100. 0.0 10. 345
GRID 10001 100. 0.0 0.0 345
GRID 10100 99.3625 3.30491 10. 345
GRID 10101 99.3625 3.30491 0.0 345
MD Nastran R2.1 Release Guide

100
GRID 10200 96.8149 6.51543 10. 345
GRID 10201 96.8149 6.51543 0.0 345
GRID 10300 92.5105 9.59323 10. 345
GRID 10301 92.5105 9.59323 0.0 345
GRID 10400 86.6025 12.5 10. 345
GRID 10401 86.6025 12.5 0.0 345
GRID 10500 79.2443 15.1974 10. 345
GRID 10501 79.2443 15.1974 0.0 345
GRID 10600 70.5889 17.6472 10. 345
GRID 10601 70.5889 17.6472 0.0 345
GRID 10700 60.7898 19.8111 10. 345
GRID 10701 60.7898 19.8111 0.0 345
GRID 10800 50. 21.6506 10. 345
GRID 10801 50. 21.6506 0.0 345
GRID 10900 38.3729 23.1276 10. 345
GRID 10901 38.3729 23.1276 0.0 345
GRID 11000 26.0617 24.2037 10. 345
GRID 11001 26.0617 24.2037 0.0 345
GRID 11100 13.2197 24.8406 10. 345
GRID 11101 13.2197 24.8406 0.0 345
GRID 11200 0.0 25. 10. 345
GRID 11201 0.0 25. 0.0 345
$
CQUAD4 101 100 10000 10001 10101 10100
CQUAD4 102 100 10100 10101 10201 10200
CQUAD4 103 100 10200 10201 10301 10300
CQUAD4 104 100 10300 10301 10401 10400
CQUAD4 105 100 10400 10401 10501 10500
CQUAD4 106 100 10500 10501 10601 10600
CQUAD4 107 100 10600 10601 10701 10700
CQUAD4 108 100 10700 10701 10801 10800
CQUAD4 109 100 10800 10801 10901 10900
CQUAD4 110 100 10900 10901 11001 11000
CQUAD4 111 100 11000 11001 11101 11100
CQUAD4 112 100 11100 11101 11201 11200
$
PSHELL 100 100 0.10 100 101
$
SPC1 200 16 11200 11201
SPC1 200 26 10000 10001
$
SPC1 400 16 11200 11201
SPC1 400 26 10000 10001
SPC1 400 1 10700
SPC1 400 2 10701
$
ENDDATA
Example 2
Example 2, EX02, is a modified version of a standard test problem file, NLTSUB02. It shows two
different types of analysis in the same job. This model is similar to Example 1 except it adds some static
loads and the required NLPARM entries. All bold-font statements are entries that show the difference in
the two different analysis types.
ID MSC, EX02 $
TIME 150 $
SOL 400 $
CEND
101
CHAPTER 3
Advanced Integrated Nonlinear Analysis
TITLE=TEST MIXED ANALYSES - NLSTAT AND NLTRAN EX02
SUBTITLE =SPC CHANGE IN THE STEPS IN EACH SUBCASE
SET 10 = 10000,11200
SET 20 = 101
SEALL = ALL
DISPL = ALL
STRESS = 20
$
SUBCASE 100
ANALYSIS=NLSTAT
STEP 10
LOAD = 800
SPC = 200
NLPARM = 110
STEP 20
LOAD = 900
SPC = 400
NLPARM = 120
$
SUBCASE 200
ANALYSIS=NLTRAN
STEP 10
DLOAD = 100
SPC = 200
TSTEPNL = 310
STEP 20
DLOAD = 100
SPC = 400
TSTEPNL = 320
$
BEGIN BULK
NLPARM 110 10 AUTO YES
NLPARM 120 10 AUTO YES
$
LOAD 800 0.01 1.0 510
LOAD 900 0.05 1.0 510
( The rest is same as what in the Bulk Data Deck in the 1st Example)
ENDDATA
Example 3
Example 3, EX03, is a modified form of a standard test problem file, NLTTL002. This model has one
CQUAD4 element and two CTRIA3 elements. Its purpose is to show the various combinations of
TTEMP and TMPSET Bulk Data entries representing thermal effects in nonlinear transient analysis. All
the bold-font statements are entries related to the temperature-related inputs.
ID MSC, EX03 $
SOL 400
DIAG 8,15
TIME 60
CEND
SEALL = ALL
SUPER = ALL
TITLE = THERMAL LOAD TEST FOR NONLINEAR TRANSIENT ANALYSIS EX03
SUBTITLE = Q4/T3 MODEL, TTEMP AND TMPSET
$ECHO = NONE
MAXLINES = 999999999
$
TEMPERATURE(INITIAL) = 1
SUBCASE 1
MD Nastran R2.1 Release Guide

102
analysis=NLTRAN
step 1
TSTEPNL= 1
SPC = 2
TEMPERATURE(LOAD) = 3
DISPLACEMENT(SORT1,REAL)=ALL
nlstress = all
stress = all
step 2
TSTEPNL= 1
SPC = 2
TEMPERATURE(LOAD) = 4
DISPLACEMENT(SORT1,REAL)=ALL
nlstress = all
stress = all
SUBCASE 2
analysis=NLTRAN
step 3
TSTEPNL= 1
SPC = 2
TEMPERATURE(LOAD) = 5
DISPLACEMENT(SORT1,REAL)=ALL
nlstress = all
stress = all
step 4
TSTEPNL= 1
SPC = 2
TEMPERATURE(LOAD) = 6
DISPLACEMENT(SORT1,REAL)=ALL
nlstress = all
stress = all
SUBCASE 3
analysis=NLTRAN
step 5
TSTEPNL= 1
SPC = 2
TEMPERATURE(LOAD) = 7
DISPLACEMENT(SORT1,REAL)=ALL
nlstress = all
stress = all
step 6
TSTEPNL= 1
SPC = 2
TEMPERATURE(LOAD) = 8
DISPLACEMENT(SORT1,REAL)=ALL
nlstress = all
stress = all
SUBCASE 4
analysis=NLTRAN
step 7
TSTEPNL= 1
SPC = 2
TEMPERATURE(LOAD) = 9
DISPLACEMENT(SORT1,REAL)=ALL
nlstress = all
stress = all
step 8
TSTEPNL= 1
SPC = 2
TEMPERATURE(LOAD) = 10
DISPLACEMENT(SORT1,REAL)=ALL
nlstress = all
stress = all
$
BEGIN BULK
PARAM POST -1
PARAM COUPMASS 1
PARAM LGDISP 1
103
CHAPTER 3
Advanced Integrated Nonlinear Analysis
PARAM K6ROT 100.
PARAM,NOCOMPS,-1
PARAM PRTMAXIM YES
PARAM,COMPMATT,YES
PARAM,EPSILONT,INTEGRAL
PARAM NLTOL 0
TSTEPNL,1,4,0.25,1,AUTO
$
PCOMP 1 79. 0.
* 1 .04875 0. YES
$
MAT8 1 7.15+6 2.9+6 .29 1.4+6 1.9-4
2.9-6 6.-6 79.
MATT8 1 3 5 4 6
1 2
$
TABLEM1 1 + CR
+ CR 60. 2.9-6 70. 2.9-6 80. 3.24-6 100. 3.86-6 + CS
+ CS 120. 4.01-6 140. 3.89-6 150. 3.78-6 160. 3.68-6 + CT
+ CT 180. 3.52-6 200. 3.47-6 220. 3.55-6 240. 3.76-6 + CU
+ CU 250. 3.87-6 260. 3.99-6 280. 4.12-6 300. 4.24-6 + CV
+ CV 320. 4.24-6 ENDT
$
TABLEM1 2 + CW
+ CW 60. 6.-6 70. 6.-6 80. 7.67-6 100. 1.168-5+ CX
+ CX 120. 1.341-5 140. 1.37-5 150. 1.349-5 160. 1.328-5+ CY
+ CY 180. 1.266-5 200. 1.222-5 220. 1.218-5 240. 1.259-5+ CZ
+ CZ 250. 1.296-5 260. 1.334-5 280. 1.415-5 300. 1.46-5 + DA
+ DA 320. 1.46-5 ENDT
$
TABLEM1 3 + BX
+ BX 60. 7.15+6 70. 7.15+6 80. 7.15+6 100. 7.13+6 + BY
+ BY 120. 7.11+6 140. 7.08+6 150. 7.07+6 160. 7.07+6 + BZ
+ BZ 180. 7.06+6 200. 7.05+6 220. 7.05+6 240. 7.04+6 + CA
+ CA 250. 7.04+6 260. 7.05+6 280. 7.06+6 300. 7.08+6 + CB
+ CB 320. 7.08+6 ENDT
$
TABLEM1 4 + CM
+ CM 60. .29 70. .29 80. .29 100. .29 + CN
+ CN 120. .29 140. .29 150. .29 160. .29 + CO
+ CO 180. .29 200. .29 220. .29 240. .29 + CP
+ CP 250. .29 260. .29 280. .29 300. .29 + CQ
+ CQ 320. .29 ENDT
$
TABLEM1 5 + CC
+ CC 60. 2.9+6 70. 2.9+6 80. 2.9+6 100. 2.82+6 + CD
+ CD 120. 2.75+6 140. 2.68+6 150. 2.64+6 160. 2.58+6 + CE
+ CE 180. 2.47+6 200. 2.35+6 220. 2.22+6 240. 2.09+6 + CF
+ CF 250. 2.03+6 260. 1.95+6 280. 1.8+6 300. 1.65+6 + CG
+ CG 320. 1.65+6 ENDT
$
TABLEM1 6 + CH
+ CH 60. 1.4+6 70. 1.4+6 80. 1.4+6 100. 1.34+6 + CI
+ CI 120. 1.29+6 140. 1.24+6 150. 1.22+6 160. 1.2+6 + CJ
+ CJ 180. 1.15+6 200. 1.1+6 220. 980000. 240. 870000.+ CK
+ CK 250. 810000. 260. 750000. 280. 620000. 300. 500000.+ CL
+ CL 320. 500000. ENDT
$
cquad4,1,1,1,2,5,4
ctria3,2,1,1,2,4
ctria3,3,1,2,5,4
$
MD Nastran R2.1 Release Guide

104
GRID 1 0.00000 0.00000 0.00000
GRID 2 1.00000 0.00000 0.00000
GRID 4 0.00000 1.00000 0.00000
GRID 5 1.00000 1.00000 0.10000
$
SPCADD 2 1
SPC1 1 123456 1 2
spc1 1 123456 4
$
TTEMP,3,111,300
TMPSET,111,4,5
TTEMP,3,101,310
TMPSET,101,1,2
$
TTEMP,4,102,400
TMPSET,102,1,2,4,5,7,8,9,
,10,11,12
$
TTEMP,5,201,500
TMPSET,201,1,2,4,5
$
TTEMP,6,-1,400
$
TTEMP,7,202,700
TMPSET,202,1,2
$
TTEMP,8,204,800
TMPSET,204,1,2
$
TTEMP,9,402,900
TMPSET,402,1,2,4,5
$
TEMP 1 1 79.
TEMP 1 2 79.
TEMP 1 4 79.
TEMP 1 5 79.
$
TEMP 3 1 80.
TEMP 3 2 80.
TEMP 3 4 80.
TEMP 3 5 80.
TABLED1 300
0.0 .9875 1.0 1.0 ENDT
TABLED1 310
0.0 .9875 1.0 1.0 ENDT
$
TEMP 4 1 81.
TEMP 4 2 81.
TEMP 4 4 81.
TEMP 4 5 81.
TABLED1 400
1.0 .9876542 2.0 1.0 ENDT
$
TEMP 5 1 80.
TEMP 5 2 80.
TEMP 5 4 80.
TEMP 5 5 80.
TABLED1 500
0.0 .9875 1.0 1.0 ENDT
$
TEMP 6 1 81.
TEMP 6 2 81.
TEMP 6 4 81.
TEMP 6 5 81.
$
$
TEMP 7 1 80.
TEMP 7 2 80.
105
CHAPTER 3
Advanced Integrated Nonlinear Analysis
TEMP 7 4 80.
TEMP 7 5 80.
TABLED1 700
0.0 .9875 1.0 1.0 ENDT
$
TEMP 8 1 81.
TEMP 8 2 81.
TEMP 8 4 81.
TEMP 8 5 81.
TABLED1 800
1.0 .9876542 2.0 1.0 ENDT
$
TEMP 9 1 80.
TEMP 9 2 80.
TEMP 9 4 80.
TEMP 9 5 80.
TABLED1 900
0.0 .9875 1.0 1.0 ENDT
$
TEMP 10 1 81.
TEMP 10 2 81.
TEMP 10 4 81.
TEMP 10 5 81.
ENDDATA
Example 4
Example 4, EX04, is modified from a standard test problem file, NLTIC19. This model has one CHEXA
element. Its purpose is to show two different types of analysis in the same subcasethe model itself is not
important. All the bold-font statements are entries that show the difference between the analyses, and
how to set the initial condition for the nonlinear transient analysis after first static analysis. Note that the
nonlinear transient analysis does not use the final results of the closest static analysis as the initial
condition; instead, it uses the results of the 50% load increment in the first step as the initial condition.
Also, parameter ICPOT=0 is selected, which will compute the initial acceleration at the beginning
(t = 0.0) of the transient analysis when it is not in balance.
ID MSC, EX04 $
DIAG 8,15
TIME 60
SOL 400 $
CEND
TITLE= ELASTIC-PLASTIC STATIC & TRANSIENT RESPONSE, EX04
SUBTI= INITIAL ACCELERATION COMPUTED - PARAM,ICOPT,0
SET 1 = 1111
SET 2 = 100
DISP = 1
VELO = 1
ACCE = 1
OLOAD = 1
$ STRESS(PLOT) = 2
SUBCASE 1130
step 1
LABEL=UNIAXIAL TENSION (LOADING)
ANALYSIS=NLSTAT
SPC=100
LOAD=1130
NLPARM = 1
step 2
LABEL=UNIAXIAL TENSION (UNLOADING)
MD Nastran R2.1 Release Guide

106
ANALYSIS=NLSTAT
SPC=100
NLPARM = 1
step 10
LABEL=I.C. FROM THE FIRST NLSTAT STEP(50%) - UNBALANCED CASE (NLIC)
ANALYSIS=NLTRAN
NLIC STEP 1 LOADFAC 0.5
SPC=100
DLOAD=2130
TSTEPNL=10
param,icopt,0
BEGIN BULK
PARAM,LANGLE,3
PARAm,LGDISP,1
PARAM,W4,1.0
$
NLPARM 1 4 AUTO ALL
1.-6
TSTEPNL 10 2000 0.001 AUTO
$
MAT1 1 30.0+6 11.5+6 0.3 7.332-2 0.01
PSOLID 1 1
$
SPC1 100 123456 1000
SPC1 100 1 1010
SPC1 100 2 1001
SPC1 100 3 1100
GRDSET 456
$
TLOAD1 2130 2130 0 500
TABLED1 500 +TAB1
+TAB1 0. 0. 1. -1. 1.2 0. 10. 0. +TAB2
+TAB2 ENDT
$
LOAD 1130 -1.6 2. 121
LOAD 2130 -1.6 1. 121
$
GRID 1000 0. 0. 0.
GRID 1100 1. 0. 0.
GRID 1110 1. 1. 0.
GRID 1010 0. 1. 0.
GRID 1001 0. 0. 1.
GRID 1101 1. 0. 1.
GRID 1111 1. 1. 1.
GRID 1011 0. 1. 1.
$
CHEXA 100 1 1000 1100 1110 1010 1001 1101 +HX100
+HX100 1111 1011
$
PLOAD4 121 100 36.+3 1100 1111
PLOAD4 121 100 36.+3 1000 1011
ENDDATA
Example 5
Example 5, EX05, is a modified version of a standard test problem file, NLTIC19R. This model is a
restart run of Example 4, EX04. Since there is no structure change in SUBCASE 1130 and there is no
parameter in the NLRESTART command, this restart job will start from the second subcase. The purpose
of this example is to show how the Case Control commands NLRESTART and NLIC can work together.
107
CHAPTER 3
Advanced Integrated Nonlinear Analysis
All bold-font statements show the key Case Control commands in this example. Note that the nonlinear
transient analysis in the second subcase specifies the final results in the first step of the first subcase to
be the initial condition. Also, parameter ICOPT = 1 (the default) is selected, which will not compute the
initial acceleration; rather, it will compute loads in the beginning (t = 0.0) of the transient analysis.
ICOPT = 1 assumes the whole model is automatically in equilibrium when initial conditions are applied.
ASSIGN RSFILE='DBSDIR:ex04.MASTER' $
RESTART LOGICAL=RSFILE $
$
ID MSC, EX05 $
DIAG 8,15
TIME 60
SOL 400 $
CEND
TITLE= ELASTIC-PLASTIC STATIC & TRANSIENT RESPONSE, EX05
SUBTI= NO INITIAL ACCELERATION COMPUTED - PARAM,ICOPT1
SET 1 = 1111
SET 2 = 100
DISP = 1
VELO = 1
ACCE = 1
OLOAD = 1
$ STRESS = 2
NLRESTART
SUBCASE 1130
step 1
LABEL=UNIAXIAL TENSION (LOADING)
ANALYSIS=NLSTAT
SPC=100
LOAD=1130
NLPARM = 1
step 2
LABEL=UNIAXIAL TENSION (UNLOADING)
ANALYSIS=NLSTAT
SPC=100
NLPARM = 1
step 10
LABEL=I.C. FROM THE FIRST NLSTAT STEP - UNBALANCED CASE (NLIC)
ANALYSIS=NLTRAN
NLIC STEP 1 LOADFAC 0.5
SPC=100
DLOAD=2130
TSTEPNL=10
param,icopt,0
SUBCASE 1131
LABEL=I.C. FROM THE 1st NLSTAT STEP OF PREVIOUS SUBCASE (NLIC)
ANALYSIS=NLTRAN
NLIC SUBCASE 1130 STEP 1
SPC=100
DLOAD=2130
TSTEPNL=10
param,icopt,1
BEGIN BULKk
ENDDATA
MD Nastran R2.1 Release Guide

108
Example 6
Example 6, EX06, is modified from a standard test problem file, NLTROT01. This model shows how to
run rotor dynamics in SOL 400. All bold-font statements are the basic entries used in rotor dynamic
analysis. For details, refer to the MSC.Nastran 2004 Release Guide.
ID MSC, EX06 $
SOL 400
DIAG 8, 15
CEND
$
ANALYSIS=NLTRAN
RIGID=LINEAR
$
RGYRO= 100
TSTEP= 100
SET 99= 101
disp= 99
$
BEGIN BULK
$
UNBALNC, 100, 2.0, 101, 0., 1., 0.,
, 1.0, 0.0, 0.0, 0.0, 1000., none
$
TSTEPNL, 100, 5000, 3.0E-4, 10
$
ROTORG 10 101 THRU 103
$
RSPINT 10 101 102 FREQ 100
0.01
TABLED1, 100,
, 0.0, 22.5, 100.0, 22.5, endt
$
$ ROTOR 1
$
GRID, 101, , 0., 0., 0.
GRID, 102, , 1., 0., 0., , 14
GRID, 103, , 2., 0., 0.
GRID, 104, , 0., 0., 0.
GRID, 105, , 2., 0., 0.
$
RBE2, 1001, 102, 123456, 101, 103
RBE2, 1002, 101, 123456, 104
RBE2, 1003, 103, 123456, 105
$
CONM2, 1004, 102, , 50.,
, 5.0, , 15.0, , , 15.0
$
CELAS1, 1005, 1000, 104, 2
CELAS1, 1006, 1000, 104, 3
CELAS1, 1007, 1000, 105, 2
CELAS1, 1008, 1000, 105, 3
PELAS, 1000, 1.0E+5, 0.0
$
param, g, 0.05
param, w3, 141.3
$
enddata
109
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Example 7
Example 7, EX07, is modified from a standard test problem file, NLTK2PP1. This model shows how to
apply K2PP in SOL 400; for example, using different sets of K2PP in different steps. The first step
requests the K2MAT matrix as K2PP input, and the second step requests the combination of K2MAT and
K3MAT matrices as K2PP input. All the bold-font statements are first step entries required in this type
of analysis.
ID MSC, EX07 $
SOL 400 $
DIAG 8,15
TIME 50
CEND
$
TITLE = LINEAR TRANSIENT RESP. (DIRECT METHOD) TEST NLTK2PP1
SUBTITLE = COULOMB FRICTION LESS F3(0),F4(0)SEALL=ALL
$ ECHO = NONE
SET 10= 1,7
DISP=10
SEALL = ALL
MPC = 20
SPC = 10
analysis=NLTRAN
SUBCASE 1
STEP 1
DLOAD = 12
K2PP = K2MAT
TSTEPNL = 10
STEP 2
DLOAD = 12
K2PP = K2MAT, K3MAT
TSTEPNL = 10
BEGIN BULK
param,ndamp,-0.055
$
CORD2R 1 0 0. 0. 0. 0. 0. 1. +CORD
+CORD 1. 0. 0.
$
GRDSET 1 1 1246
GRID 1 -460. 0. 0.
GRID 3 -460. 0. 100.
GRID 5 -460. 0. 200.
GRID 2 290. 0. 0.
GRID 4 290. 0. 100.
GRID 6 290. 0. 200.
GRID 7 0. 0. 200.
$
CONM2 3 3 1 0.6867
CONM2 4 4 1 1.2735
CONM2 7 7 1 15.445 +CONM
+CONM 3.0E5
$
CELAS2 1 2200. 1 3 3 3
CDAMP2 1 3.2 1 3 3 3
CELAS2 2 4200. 2 3 4 3
CDAMP2 2 6.2 2 3 4 3
CELAS2 3 550. 3 3 5 3
CDAMP2 3 13.3 3 3 5 3
CELAS2 4 2000. 4 3 6 3
MD Nastran R2.1 Release Guide

110
CDAMP2 4 0. 4 3 6 3
$
MPC 10 7 3 750. 5 3 -290. +MPC10
+MPC10 6 3 -460.
MPC 11 7 5 750. 5 3 -1. +MPC11
+MPC11 6 3 1.
MPCADD 20 10 11
$
SPC1 10 5 1 THRU 6
$
TABLED1 12 +TBLD11
+TBLD11 -100. 0. 0. 0. 0.005 3. 100. 3. +TBLD12
+TBLD12 ENDT
$
EPOINT 101
$
DMIG K2MAT 0 1 1 1
DMIG K2MAT 1 3 101 2.
DMIG K2MAT 101 1 3 -2.
$
DMIG K3MAT 0 1 1 1
DMIG K3MAT 1 3 101 3.
DMIG K3MAT 101 1 3 -3.
$
DAREA 12 101 1.
DELAY 12 101 0.
$
EPOINT 202
$
DMIG K2MAT 2 3 202 2.
DMIG K2MAT 202 2 3 -2.
$
DMIG K3MAT 2 3 202 3.
DMIG K3MAT 202 2 3 -3.
$
DAREA 12 202 1.
DELAY 12 202 0.54
$
TLOAD1 12 12 12 12
$
TSTEPNL 10 1000 0.002 1 AUTO 10000
ENDDATA
Example 8
Example 8, EX08, is modified from a standard test problem file, NLTSTP06. This model shows the use
of the SPC-set change between STEPs. All the bold-font statements are required entries.
ID MSC, EX08 $
TIME 150 $
SOL 400 $
DIAG 8,15 $
CEND
TITLE=ANISOTROPIC MATERIAL & MATS1, ELLIPTIC CYLINDER UNDER NLTSTP06
SUBTITLE = TWO STEP'S TEST, SPC SET CHANGE AND SAME DT
SET 10 = 10000,11200
SET 20 = 101
SEALL = ALL
DISPL = ALL
STRESS = 20
111
CHAPTER 3
Advanced Integrated Nonlinear Analysis
SPCF = ALL
OLOAD = ALL
$
SUBCASE 100
ANALYSIS=NLTRAN
STEP 10
DLOAD = 100
SPC = 200
TSTEPNL = 310
STEP 20
DLOAD = 100
SPC = 400
TSTEPNL = 320
$
BEGIN BULK
PARAM LGDISP 1
TSTEPNL 310 100 0.01 10 AUTO +TS11
TSTEPNL 320 100 0.01 10 AUTO +TS21
$
PLOAD4 510 101 5. THRU 112
$
TLOAD1 100 510 0 0 120
TABLED1 120 +TBD1
+TBD1 0. 0. 5. 1. 16. 1. ENDT
MAT2 100 3.2967+79.8901+60. 3.2967+70. 1.1538+70.283-2
MAT2 101 3.2967+79.8901+60. 3.2967+70. 0.283-2
MATS1 100 PLASTIC 3.+5 500000.
$
GRID 10000 100. 0.0 10. 345
GRID 10001 100. 0.0 0.0 345
GRID 10100 99.3625 3.30491 10. 345
GRID 10101 99.3625 3.30491 0.0 345
GRID 10200 96.8149 6.51543 10. 345
GRID 10201 96.8149 6.51543 0.0 345
GRID 10300 92.5105 9.59323 10. 345
GRID 10301 92.5105 9.59323 0.0 345
GRID 10400 86.6025 12.5 10. 345
GRID 10401 86.6025 12.5 0.0 345
GRID 10500 79.2443 15.1974 10. 345
GRID 10501 79.2443 15.1974 0.0 345
GRID 10600 70.5889 17.6472 10. 345
GRID 10601 70.5889 17.6472 0.0 345
GRID 10700 60.7898 19.8111 10. 345
GRID 10701 60.7898 19.8111 0.0 345
GRID 10800 50. 21.6506 10. 345
GRID 10801 50. 21.6506 0.0 345
GRID 10900 38.3729 23.1276 10. 345
GRID 10901 38.3729 23.1276 0.0 345
GRID 11000 26.0617 24.2037 10. 345
GRID 11001 26.0617 24.2037 0.0 345
GRID 11100 13.2197 24.8406 10. 345
GRID 11101 13.2197 24.8406 0.0 345
GRID 11200 0.0 25. 10. 345
GRID 11201 0.0 25. 0.0 345
$GRID 20000 0. 0. 0. 123456
$GRID 20001 100. 0. 0. 123456
$GRID 20100 0. 100. 0. 123456
$
CQUAD4 101 100 10000 10001 10101 10100
CQUAD4 102 100 10100 10101 10201 10200
MD Nastran R2.1 Release Guide

112
CQUAD4 103 100 10200 10201 10301 10300
CQUAD4 104 100 10300 10301 10401 10400
CQUAD4 105 100 10400 10401 10501 10500
CQUAD4 106 100 10500 10501 10601 10600
CQUAD4 107 100 10600 10601 10701 10700
CQUAD4 108 100 10700 10701 10801 10800
CQUAD4 109 100 10800 10801 10901 10900
CQUAD4 110 100 10900 10901 11001 11000
CQUAD4 111 100 11000 11001 11101 11100
CQUAD4 112 100 11100 11101 11201 11200
$
PSHELL 100 100 0.10 100 101
$
SPC1 200 16 11200 11201
SPC1 200 26 10000 10001
$
SPC1 400 16 11200 11201
SPC1 400 26 10000 10001
SPC1 400 1 10700
SPC1 400 2 10701
$
ENDDATA
Example 9
Example 9, EX09, is modified from a standard test problem file, NLTSTP07. This model shows that the
upstream superelement can have output requests that are different from the residual. Also, Case Control
command OTIME is used here to select a subset of all output time steps, which is selected by the
TSTEPNL Bulk Data entry. It can reduce the output data dramatically. For example, in the following file,
TSTEPNL requests output at time=0.0, 0.1, , 1.0 seconds, but OTIME overwrites this request and only
generates output at time = 0.5 seconds. All the bold-font statements are the entries required to complete
the output requests used in this example.
ID MSC, EX09 $
TIME 150 $
SOL 400 $
DIAG 8,15 $
CEND
TITLE=Test For Upperstream Superelement Output Request EX09
SUBTITLE = OTIME Output Request
SET 1 = 0.5
SET 10 = 11200
SET 20 = 101
LOADSET = 500
DISPL = 10
OTIME = 1
$
SUBCASE 10
SUPER=10
METHOD=10
SET 10001 = 10001
DISP =10001
SUBCASE 100
ANALYSIS=NLTRAN
DLOAD = 100
SPC = 200
TSTEPNL = 310
$
113
CHAPTER 3
Advanced Integrated Nonlinear Analysis
BEGIN BULK
PARAM LGDISP 1
$
SESET,10,10000,thru,10301
SEQSET1 10 0 10500 THRU 11001
EIGRL 10
$
TSTEPNL 310 100 0.01 10 AUTO +TS11
$+TS11 1.E-2
LSEQ 500 110 510
PLOAD4 510 101 5. THRU 112
$
TLOAD1 100 110 0 0 120
TABLED1 120 +TBD1
+TBD1 0. 0. 5. 1. 16. 1. ENDT
$MAT1 100 3.+7 0.3 .283-2
MAT2 100 3.2967+79.8901+60. 3.2967+70. 1.1538+70.283-2
MAT2 101 3.2967+79.8901+60. 3.2967+70. 0.283-2
MATS1 100 PLASTIC 3.+5 500000.
$
GRID 10000 100. 0.0 10. 345
GRID 10001 100. 0.0 0.0 345
GRID 10100 99.3625 3.30491 10. 345
GRID 10101 99.3625 3.30491 0.0 345
GRID 10200 96.8149 6.51543 10. 345
GRID 10201 96.8149 6.51543 0.0 345
GRID 10300 92.5105 9.59323 10. 345
GRID 10301 92.5105 9.59323 0.0 345
GRID 10400 86.6025 12.5 10. 345
GRID 10401 86.6025 12.5 0.0 345
GRID 10500 79.2443 15.1974 10. 345
GRID 10501 79.2443 15.1974 0.0 345
GRID 10600 70.5889 17.6472 10. 345
GRID 10601 70.5889 17.6472 0.0 345
GRID 10700 60.7898 19.8111 10. 345
GRID 10701 60.7898 19.8111 0.0 345
GRID 10800 50. 21.6506 10. 345
GRID 10801 50. 21.6506 0.0 345
GRID 10900 38.3729 23.1276 10. 345
GRID 10901 38.3729 23.1276 0.0 345
GRID 11000 26.0617 24.2037 10. 345
GRID 11001 26.0617 24.2037 0.0 345
GRID 11100 13.2197 24.8406 10. 345
GRID 11101 13.2197 24.8406 0.0 345
GRID 11200 0.0 25. 10. 345
GRID 11201 0.0 25. 0.0 345
$
CQUAD4 101 100 10000 10001 10101 10100
CQUAD4 102 100 10100 10101 10201 10200
CQUAD4 103 100 10200 10201 10301 10300
CQUAD4 104 100 10300 10301 10401 10400
CQUAD4 105 100 10400 10401 10501 10500
CQUAD4 106 100 10500 10501 10601 10600
CQUAD4 107 100 10600 10601 10701 10700
CQUAD4 108 100 10700 10701 10801 10800
CQUAD4 109 100 10800 10801 10901 10900
CQUAD4 110 100 10900 10901 11001 11000
CQUAD4 111 100 11000 11001 11101 11100
CQUAD4 112 100 11100 11101 11201 11200
$
MD Nastran R2.1 Release Guide

114
PSHELL 100 100 0.10 100 101
$
SPC1 200 16 11200 11201
SPC1 200 26 10000 10001
$
SPC1 400 16 11200 11201
SPC1 400 26 10000 10001
SPC1 400 1 10700
SPC1 400 2 10701
$
ENDDATA
115
CHAPTER 3
Advanced Integrated Nonlinear Analysis
Differential/Follower Force Stiffness Processing in
SOL 400
Introduction
In SOL 400, by default, differential/follower force stiffness is computed for geometric nonlinear
analysis. However, these stiffnesses are not required to obtain the correct results. If the solution
converges, the results are correct. The differential stiffness will facilitate convergence if the axial forces
are large. This is true for problems that compute nonlinear buckling load, or for problems with a tension
stiffening effect.
However, there are situations in which the differential stiffness and/or the follower force stiffness hinders
convergence, or even lead to solution divergence. For this reason, SOL 400 provides an option, NLDIFF,
to turn off the generation of differential stiffness and follower force stiffness.
For MD Nastran R2, the option NLDIFF is available for CQUADR and CTRIAR elements only. For MD
Nastran R2.1, this option is extended to CBEAM, CQUADR, CTRIAR, CQUAD4, CTRIA3, CHEXA,
CPENTA, and CTETRA elements.
Limitation
The option NLDIFF has no effect on the advanced nonlinear elements. These are elements with
PSHNL1, PSHNL2, PSLNL1, or PSHEARN Bulk Data entries.
Input
The option is input by the Bulk Data entry:
MDLPRM, NLDIFF, n
where
Guidelines
For certain problems, especially post-buckling or near-buckling problems, n=1 or 2 will help
solution convergence using the Newton iteration methods (AUTO, FNT, PFNT, or ITER on the
NLPARM entry).
n = 0 Computes the differential stiffness/follower force stiffness (default);
= 1 Turns off the computation of differential/follower stiffness; and
= 2 Turns of the differential/follower force stiffness computation if the total tangential
stiffness is found to be negative-definite.
MD Nastran R2.1 Release Guide

116
For post-buckling or near-buckling problems with n=1 or 2, CONV=PW on the NLPARM entry
is usually the best convergence criterion. The reason is that this type of problem involves large
displacement jumps at snap-through. CONV=UPW are weighted average convergence criteria.
The weighting function used in these criteria is displacement.
Between n = 1 and n = 2, n = 1 is usually more efficient than n = 2 if the user knows that the
problem is a post-buckling problem.
Example
This example is a snap-through analysis of a long thin plate. NLDIFF=1 is used to obtain the solution
convergence.
id msc, nqr4ar1a.dat $
$ID QUADR,TEST
APP DISP
SOL 400
CEND
TITLE = GEOMETRIC NONLINEAR ANALYSIS
SUBTITLE = QAJOB-NQR6AR1A
LABEL = LARGE ROTATION OF A QUADR ELEMENT
$
$ SNAP-THROUGH TEST
$
$ STEP THEORY
$
$ P U
$ -------------------
$ 1 0.8 2.280
$ 2 4.0 -
$
$ NLDIFF=1 is used to get the solution converged.
$
SET 1 = 2
SET 2 = 10
SET 3 = 1,2
DISP(PRINT,SORT2) = 1
OLOAD(PRINT,SORT2) = 1
SPCF(PRINT,SORT2) = 3
FORCE(PRINT,SORT2) = 2
STRES(CENTER,SORT2) = 2
SPC = 1
SUBCASE 1
STEP 1
LOAD = 4
NLPARM = 1
STEP 2
LOAD = 6
NLPARM = 2
BEGIN BULK
$.......2.......3.......4.......5.......6.......7.......8.......9.......0...
....
MDLPRM NLDIFF 1
PARAM LGDISP 1
CQUADR 10 11 1 2 3 4
PSHELL 11 3 0.5 3 3
MAT1 3 2000.0 1.0-7
117
CHAPTER 3
Advanced Integrated Nonlinear Analysis
$MAT1NL1 3 PLASTIC VM 10.0 200.0
MATS1 3 PLASTIC 200.0 10.0
FORCE 4 2 0.4 0.0 -1.0 0.0
FORCE 4 3 0.4 0.0 -1.0 0.0
FORCE 6 2 2.0 0.0 -1.0 0.0
FORCE 6 3 2.0 0.0 -1.0 0.0
GRID 1 0.0
GRID 2 10.0 1.0
GRID 3 10.0 1.0 -2.0
GRID 4 0.0 0.0 -2.0
NLPARM 1 YES
NLPARM 2 YES
SPC1 1 123 1
SPC1 1 13 2
SPC1 1 13 3
SPC1 1 123 4
ENDDATA
MD Nastran R2.1 Release Guide

118
Contact Analysis and Kinematic Elements
Introduction
RBAR, RBE, and related kinematic elements can be used to constrain a contact surface. The kinematic
elements use Lagrange multipliers to enforce the constraints with additional Lagrange multiplier degrees
of freedom. Both the displacement degrees of freedoms and Lagrange multiplier degrees of freedoms are
in the solution set. The displacement degrees of freedom constrained by kinematic elements are just like
any other displacement degree of freedom. Therefore, there is no conflict between the contact surface and
the constraints applied by the kinematic elements. This is demonstrated by the following example
problem.
Example
This example demonstrates that contact surfaces can be constrained by kinematic elements. The example
consists of two blocks of different size, modeled by two CHEXA elements. There is an initial gap
between the two blocks. Enforced motion is used to make the blocks come into contact. The contact
surface is completely constrained by RBAR elements with temperature loads. The solution is both
correct and efficient.
ID MSC, NLSSPCD4 $
SOL 400
CEND
TITLE = Double sided contact with kinematic elements
$
$ A simple 2 hexa model to check SPCD with 3D contact.
$ The contact surface is completely constrained by RBAR.
$
DISPLACEMENT = ALL
TEMPERATURE(INITAL)=10
SUBCASE 1
NLPARM = 1
BCONTACT = 1
SPC = 1
LOAD = 1
MPCF = ALL
TEMPERATURE(LOAD)=20
$ Direct Text Input for this Subcase
BEGIN BULK
PARAM PRTMAXIM YES
BCPARA 0 NBODIES 2 MAXENT 36 MAXNOD 36
PARAM LGDISP 1
NLPARM 1 1 1.0 FNT 25 UW YES
BCTABLE 1 1
SLAVE 1 0. 0. .07 0. 0 0.
0 0 0
MASTERS 2
CHEXA,1,1,1,2,3,4,5,6,
,7,8
CHEXA,2,1,11,12,13,14,15,16,
,17,18
psolid,1,1
MAT1 1 30000. .3 1.0-4
119
CHAPTER 3
Advanced Integrated Nonlinear Analysis
rbar,101,5,1,123456,,,123,1.0-4
rbar,102,6,2,123456,,,123,1.0-4
rbar,103,7,3,123456,,,123,1.0-4
rbar,104,8,4,123456,,,123,1.0-4
rbar,105,11,15,123456,,,123,1.0-4
rbar,106,12,16,123456,,,123,1.0-4
rbar,107,13,17,123456,,,123,1.0-4
rbar,108,14,18,123456,,,123,1.0-4
grid,1,,0.25,0.25,0.
grid,2,,0.75,0.25,0.
grid,3,,0.75,0.75,0.
grid,4,,0.25,0.75,0.
grid,5,,0.25,0.25,1.
grid,6,,0.75,0.25,1.
grid,7,,0.75,0.75,1.
grid,8,,0.25,0.75,1.
grid,11,,0.,0.,-1.4
grid,12,,1.,0.,-1.4
grid,13,,1.,1.,-1.4
grid,14,,0.,1.,-1.4
grid,15,,0.,0.,-0.2
grid,16,,1.,0.,-0.2
grid,17,,1.,1.,-0.2
grid,18,,0.,1.,-0.2
FORCE 1 5 0. .57735 .57735 .57735
$SPC1,1,12,5,6,7,8
$SPC1,1,12,11,12,13,14
SPC1,1,456,1,2,3,4
SPC1,1,456,5,6,7,8
SPC1,1,456,11,12,13,14
SPC1,1,456,15,16,17,18
SPC,1,11,3,1.5,12,3,1.5
SPC,1,13,3,1.5,14,3,1.5
celas2,5,1.e3,5,3
celas2,6,1.e3,6,3
celas2,7,1.e3,7,3
celas2,8,1.e3,8,3
celas2,15,1.e3,5,1
celas2,16,1.e3,6,1
celas2,17,1.e3,7,1
celas2,18,1.e3,8,1
celas2,25,1.e3,5,2
celas2,26,1.e3,6,2
celas2,27,1.e3,7,2
celas2,28,1.e3,8,2
celas2,11,1.e3,11,1
celas2,12,1.e3,12,1
celas2,13,1.e3,13,1
celas2,14,1.e3,14,1
celas2,21,1.e3,11,2
celas2,22,1.e3,12,2
celas2,23,1.e3,13,2
celas2,24,1.e3,14,2
BCBODY 1 3D DEFORM 1 0 .07
BCBODY 2 3D DEFORM 2 0 .07
BSURF 1 1
BSURF 2 2
TEMPD 10 40.0
TEMPD 20 50.0
ENDDATA
MD Nastran R2.1 Release Guide

120
Ch. 4: Implicit and Explicit Nonlinear Analysis MD Nastran R2.1 Nastran Release Guide

4
Implicit and Explicit Nonlinear
Analysis
!
MD Nastran Implicit Nonlinear - SOL 600
!
MD Nastran Explicit Nonlinear - SOL 700
MD Nastran R2.1 Release Guide

122
MD Nastran Implicit Nonlinear - SOL 600
The MD Nastran R2 release contains significant enhancements to SOL 600 functionality and
performance. This includes the addition of heat transfer capability; support of MD Nastran fastener
technology (CBUSH, CFAST, and CWELD) in a native mode; improved support of PLOAD4, and
enhanced composite material capabilities. Fracture mechanics analysis may also be performed to obtain
the stress intensity factors at a crack. In addition, a new streaming input capability has been added for the
direct transfer of data.
Heat Transfer
Most of the heat transfer capabilities in SOLs 153 and 159 are supported by SOL 600, with the exception
of CHBDYP and forced convection, the equivalents of which are not available in Marc. The main
advantage of using SOL 600 over SOL 153 or 159 for heat transfer is that thermal contact is available
directly, and that radiation view factors might be calculated faster. The user needs to weigh the drawbacks
of not having CHBDYP and forced convection. Because of these alternatives, SOL 600 offers two ways
to perform a heat transfer analysis. The direct (new) method performs all of the calculations, and can
support thermal contact that varies during the run. The other (indirect) method is to calculate the thermal
contact conditions (if they are needed) at the start of the run, and performs the rest of the calculations
using SOLs 153 and 159. These approaches are addressed using new options on the SOL 600 entry,
TSOLVE=M or TSOLVE=N, respectively. A typical SOL 600 Executive Control statement for heat
transfer using the direct method would be:
SOL 600,153 TSOLVE=M
A typical Executive Control statement for heat transfer using the indirect method would be:
SOL 600,153 TSOLVE=N
To use the thermal contact capability released with the Nastran 2005 version, either TSOLVE=N should
be used, or the TSOLVE option should be left blank (which will support most existing input files).
SOL 600 heat transfer addresses conduction, free convection, radiation to space, cavity radiation, thermal
contact, and latent heat. Steady state or transient heat transfer calculations may be obtained. All material
properties may be temperature-dependent, and the material may be isotropic, orthotropic, or anisotropic.
For the direct method, table input is used for all applicable input items. The direct method requires
postprocessing using the Marc .t16 file. All standard output forms (.op2, .xdb, .f06, and/or punch) are
available using the indirect method.
The temperature history obtained may then be used in a subsequent thermal stress simulation by using
the MINSTAT and MCHSTAT Bulk Data entries to read the temperatures from the .t16 file. When used
in conjunction with the MTHERM Bulk Data entries, the time steps will be either subdivided or merged
to satisfy the accuracy and convergence requirements of the nonlinear mechanical analysis.
For the direct solutions, when CTRIA3 or CQUAD4 elements are used, thermal conductivity can be
based upon either of two methods which are selected on the PSHELL entry.
123
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
For the first method, the thermal behavior may be membrane-like only, in which case there is no thermal
gradient through the thickness. To support this, new heat transfer elements have been added. These new
elements are:
Three-noded bilinear heat transfer membrane (Marc element 196).
Six-noded biquadratic heat transfer membrane (Marc element 197).
Four-noded isoparametric heat transfer element (Marc element 198).
Eight-noded biquadratic heat transfer membrane (Marc element 199).
In the second method, it is assumed that the element has a thermal gradient through the thickness, which
may be required for composite simulation or thermal shock type problems. This is activated by
specifying a nonzero MID2 entry. The MPHEAT option used to specify whether the temperature gradient
is linear or quadratic through the total thickness of the shell, or linear or quadratic variation is specified
on a per layer basis. In the later case, if a composite shell has n layers, the number of degrees of freedom
per grid is n+1, or 2*n+1 for the quadratic case.
The MHEATSHL parameter may also be used to control this behavior.
Additions or changes to a standard MD Nastran SOLs 153 or 159 heat transfer input file are as follows:
Executive Control
Change SOL statement as previously described.
Case Control
No changes.
Bulk Data
BCBODY, or BCTABLE (BCPARA, if necessary to change defaults) if there is thermal contact.
MPHEAT A new entry that maps to Marcs HEAT parameter. See MPHEAT (SOL 600) on
page 2009 of the MD Nastran Quick Reference Guide for more information.
NLHEATC Defines numerical analysis parameters for SOL 600 heat transfer analysis. For more
information, see NLHEATC (SOL 600) on page 2028 of the MD Nastran Quick Reference Guide.
MCHSTAT Option to change state variables for SOL 600, used in SOL 600 only. For more information,
see MCHSTAT (SOL 600) on page 1961 of the MD Nastran Quick Reference Guide.
MINSTAT Option to define initial state variables for SOL 600, used in SOL 600 only. See MINSTAT
(SOL 600) on page 1976 of the MD Nastran Quick Reference Guide for more information.
Bulk Data Parameters
PARAM,MARCHEAT is obsolete starting with this release.
MD Nastran R2.1 Release Guide

124
Bulk Data
BCBODY or BCTABLE Bulk Data entries (BCPARA, if necessary, to change defaults) are used if there
is thermal contact. See the MD Nastran Quick Reference Guide for more details on entries DMIGOUT,
MCHSTAT, MINSTAT, MTHERM, MPHEAT, and NLHEATC.
PARAM,MARHTPRT Control heat transfer output in the Marc .out file (Integer).
0 = Do not print any output except for summary tables.
1 = Print the nodal temperatures.
2 = Print all possible nodal heat transfer output.
PARAM,MRADUNIT Controls the units used in radiation heat transfer for SOL 600 (Integer).
1 = Degrees Celsius.
2 = Degrees Kelvin (default if parameter not entered).
3 = Degrees Fahrenheit.
Remark: Degrees in Rankine is not available.
PARAM,MHEMIPIX Controls the number of pixels used in radiation heat transfer for SOL 600
using the hemi-cube method. The default, if this parameter is not entered,
is 500 (Integer).
PARAM,MARVFCUT Controls the fraction of the maximum view factor that is to be used as a
cutoff. View factors calculated below this cutoff are ignored. The default is
0.0001 if this parameter is not entered. Used in SOL 600 radiation heat
transfer only (Real).
PARAM,MRVFIMPL Controls the fraction of the maximum view factor that is to be treated
implicitly (contribute to operator matrix). View factor values smaller than
this cutoff are treated explicitly. The default is 0.01 if this parameter is not
entered. Using this parameter reduces the size of the heat transfer operator
matrix, which reduces the computational costs associated with
decomposition. Used in SOL 600 radiation heat transfer only (Real).
PARAM,MRSTEADY Controls the solution method for SOL 600 steady state heat transfer
(Integer).
1 = Marc STEADY STATE is used with TIME STEP of 1.0 (default if
parameter not entered). The specific heat matrix is not formed.
2 = AUTO STEP is used.
Remark: Requires that a sufficiently large time period be simulated for the
solution to reach steady state.
125
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Heat Transfer Examples
The following heat transfer examples are located in the tpl1 directory:
Creep Simulation
Creep is an important phenomena in high temperature applications. To facilitate creep analysis, several
new options have been added to SOL 600 in the MD Nastran R2 release.
The MPCREEP entry is used to select the procedure used to perform time integration. If the explicit
method is used, then the timestep must be small, but if the material is elastic and undergoes small
deformations, no reassembly of the stiffness matrix is required. If the implicit method is chosen, then
larger timesteps may be used, but note that reassembly occurs at every increment.
In thermal creep simulations, it is necessary that the timestep be chosen to satisfy accuracy of both the
rate-independent thermal stress problem and the rate-dependent creep problem. To ensure that this
occurs, the MTCREEP Bulk Data entry has been introduced.
Element Selection
To increase flexibility in performing simulations, the MRALIAS parameter has been augmented by the
ALIASM Bulk Data entry. This allows the user to map the Marc element type to be used for a selection
of elements. All Marc element types may be used if they are topologically similar to the Nastran element
type. See MSC.Marc Volume B: Element Library for more details.
This mapping process is analogous to the PCOMPLS, PLCOMP, PSHLN1, PSHLN2 and PSLDN1
options in SOL 400.
Other SOL 600 Items
Membrane Elements
SOL 600 now directly supports membrane-only shells without the user having to add any alias bulk data
entries or parameters.
Six-noded biquadratic isoparametric membrane (Marc element 200).
Conduction mhqbd1, mhqbd1a, mhqbd1c, mhqbd1s, mhqbd2, mhqbd2c, mhqbd2s,
mhbc01, mhbc02, mhtepe.
Free Convection mhcbv1, mhcbv1a, mhcbv1b, mhcbv1c, mhcbv1d, mhcbv1e.
Radiation to Space mhrad1, mhrad2, mhrad3.
Cavity Radiation mhrcv1, mhrcv1a, mhrcv2, mhrhx0, mhrdhx, mhrhx4, mhrc1t.
Thermal Contact mhcnoc, mhtc07, mhtc7a.
Latent Heat mtlat1, mtlat2.
MD Nastran R2.1 Release Guide

126
Element 200 completes the set of shell elements necessary to support membranes. The user should be
careful when using membranes with nonlinear analyses, as they are unstable under certain conditions
because of a lack of bending stiffness.
Fracture Mechanics
New capabilities have been added to SOL 600 to allow greater insight in determining fracture resistance.
These methods complement the material damage models that existed previously, and are entered through
the MATHED entry. The new capabilities include:
Calculation of energy release rates and stress intensity factors VCCT and LORENZI.
Option.
Crack Propagation VCCT.
Delamination COHESIV.
New failure criteria MATF.
Birth and death of elements.
Virtual Crack Closure Technique (VCCT)
The VCCT capability is fully supported by this release, which provides a new Case Control command,
VCCT, and a new Bulk Data entry, VCCT (see the MD Nastran Quick Reference Guide for details). This
option specifies that the virtual crack closure technique is to be used for evaluating energy release rates.
The user defines the grid point (in 2-D or for shells) or grid points (in 3-D) that define each crack tip. The
supported elements are lower- and higher-order 2-D solids and 3-D shells; lower- and higher-order 3-D
hexahedral solids; and lower order 3-D tetrahedral solids. For 3-D solids, it is important that a regular
mesh around the crack front is used.
Multiple cracks can be defined, and results obtained for each crack separately. A 2-D crack is modeled
with a single crack tip grid point and a set of shell elements. A 3-D crack is modeled with a line of crack
tip grid points and a set of solid elements. Shell elements can be used to define a 2-D line crack, and can
also be connected to the face of another shell or 3-D solid to form a 3-D surface crack.
The VCCT method is advantageous because it may be used with any material model including
orthotropic or anisotropic behavior, and because it automatically obtains the mode I, II, and III stress
intensity factors. This makes is applicable to composite structures.
For crack propagation, there are two modes of growth: fatigue and direct growth. For fatigue, the user
specifies a load sequence time period. During the load sequence, the largest energy release rate and the
corresponding estimated crack growth direction is recorded. At the end of the load sequence, the crack
is grown using the specified method. For direct growth, the crack grows as soon as the calculated energy
release rate is larger than the user-specified Gc. Note that Gc can be made a function of the accumulated
crack growth length in order to model crack growth resistance behavior. This release does not support
large crack propagation, which requires remeshing.
127
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Fracture Mechanics J-Integral (LORENZI)
The Lorenzi option gives an estimation of the J-Integral for a crack configuration using the domain
integration method. The domain integration method has the advantage that it can also be used for
problems with thermal behavior, and for dynamic analysis. This procedure is only available for
continuum elements. Only the grid points defining the crack front (crack tip in two dimensions) need to
be defined. The program automatically finds integration paths according to the format below. The
complete J-Integral is evaluated and output. For the case of linear elastic material with no external loads
on the crack faces, the program automatically separates mode I, mode II, and mode III (3-D only) stress
intensity factors from the J-Integral for isotropic materials.
A new Bulk Data entry, LORENZI, is used to activate this capability, and if entered applies to all
subcases in the analysis. For more details, see the MD Nastran Quick Reference Guide.
Delamination
An alternative method to model failure is to use the COHESIV Bulk Data entry in conjunction with
special delamination or interface elements. Three different models are available, along with a user
subroutine: bilinear, exponential, and linear-exponential, as shown in the following figures. The user
defines the traction versus the relative separation. The area under the curve is the cohesive energy, often
known as the critical energy release rate.
Bilinear model
Exponential Model
Linear-Exponential Model
Element Type Number of Grid Points Characteristic
186 4 Planar
187 8 Planar
188 8 3-D
189 20 3-D
190 4 Axisymmetric
MD Nastran R2.1 Release Guide

128
MATEP Extensions
Material description MATEP was extended to add Chaboche, Power Law, Kumar, Johnson Cook, and
other options. For more details, see the MD Nastran Quick Reference Guide.
SOL 600 Failure Description MATF
For SOL 600, failure indices or actual material failure is only described using the MATF entry. For this
release, MATF has been totally revised to accommodate additional types of failure, and to improve the
clarity of the input. Be aware that other Nastran solutions can specify failure index calculations on
various MAT entries. These specifications are not available in SOL 600 only MATF may be used. To
activate the new entries, PARAM,MRMATFSB, 1 must also be included. For more details on the MATF
entry, see the MD Nastran Quick Reference Guide.
An example of the MATF entry is shown in the following tables.
Primary Format (SOL 400, 600)
191 8 Axisymmetric
192 6 3-D
193 15 3-D
1 2 3 4 5 6 7 8 9 10
MATF MID ITYPE SB
CRI Criteria Xt Xc Yt Yc Zt Zc 1st
Sxy Syz Szx Find Fxy Fyz Fzx Ext
Exc Eyt Eyc Ezt Ezc Gxy Gyz Gzx
(SOL 600
Only)
PF A1 A2 A3 A4 A5 IC1 IC2
IC3 IC4 IC5 IC6 CI7 IC8 IC9
CRI Criteria Xt Xc Yt Yc Zt Zc 2nd
Sxy Syz Szx Find Fxy Fyz Fzx Ext
Exc Eyt Eyc Ezt Ezc Gxy Gyz Gzx
(SOL 600
Only)
PF A1 A2 A3 A4 A5 IC1 IC2
IC3 IC4 IC5 IC6 IC7 IC8 IC9
CRI Criteria Xt Xc Yt Yc Zt Zc 3rd
Sxy Syz Szx Find Fxy Fyz Fzx Ext
Exc Eyt Eyc Ezt Ezc Gxy Gyz Gzx
(SOL 600
Only)
PF A1 A2 A3 A4 A5 IC1 IC2
IC3 IC4 IC5 IC6 IC7 IC8 IC9
Element Type Number of Grid Points Characteristic
129
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Example 1 (Three criteria, no progressive failure (SOLs 400/600/700 only):
(Note: The fourth and sixth lines cannot be entirely blank, and the last line of the third criteria has been
omitted.)
Example 2 (With progressive failure, (SOL 600 only)):
Alternate Format (SOLs 600, 700):
Alternate Format Example:
MATF 100 0
CRI 1 2500. 4000. 2500. 4000. 2000. 3000. 1st
4500. 4500. 4500.
+
CRI 2 2nd
+
.11 .06 .1 .05 .075 .03 .03 .03
CRI 4 2500. 4000. 2500. 4000. 2000. 3000. 3rd
4500. 4500. 4500. 0.90
MATF 100 2
CRI 1 2500. 4000. 2500. 4000. 2000. 3000. 1st
4500. 4500. 4500.
+
PF .001 1 1
+ 1 1
1 2 3 4 5 6 7 8 9 10
MATF MID CRI67 Xt Xc Yt Yc Zt Zc
Sxy Syz Szx Find Fxy Fyz Fzx Ext
Exc Eyt Eyc Ezt Ezc Gxy Gyz Gzx
MATF 100 1 2500. 4000. 2500. 4000. 2000. 3000.
4500. 4500. 4500.
Note: Remarks are given in the MATF entry, MD Nastran Quick Reference Guide.
MD Nastran R2.1 Release Guide

130
Field Contents
MID Identification number of a MAT1, MAT2, MAT8, MATORT or MAT9 entry. See
Remarks 1. and 2. (Integer > 0; no Default)
ITYPE Flag to invoke progressive failure (Integer; Default = 0 for SOL 400 and 1 for SOL
600).
0 No progressive failure, compute failure indices only.
1 Standard (original) Marc method.
2 Gradual selective stiffness degradation. (MD Nastran R2.1 and subsequent
releasesSOL 600 only).
3 Immediate selective stiffness degradation. (MD Nastran R2.1 and subsequent
releasesSOL 600 only).
SB Allowable shear stress of bonding material between layers, composites only (Real,
no Default, SOL 600 only).
CRI For the primary failure model, warning may lead to slow convergence.
CRI67 Used with the alternate format only. (Integer; no Default; Required). It is highly
recommended that only one criterion be used. However, up to three criteria from the
following criteria list can be specified in a packed list as follows:
1000000*ITYPE+10000*C3+100*C2+C1
where C1, C2, C3 are integer values for the various criteria. For SOL 700, only
criteria number 6 is available, and progressive failure will occur if the limits are
reached regardless of the value of ITYPE.
Criteria Select an integer corresponding to the failure criteria to be applied (Integer; no
Default). Up to three failure criteria may be specified for each MID for SOLs 400
and 600. Only one failure criteria may be used for SOL 700, and the primary format
should be used.
1 for maximum stress criterion (SOLs 400, 600 only, see Remark 3.).
2 for maximum strain criterion (SOLs 400, 600 only, see Remark 4.).
3 for Hill failure criterion (SOLs 400, 600 only, see Remark 5.).
4 for Hoffman failure criterion (SOLs 400, 600 only, see Remark 6.).
5 Tsai-Wu failure criterion (SOLs 400, 600 only, see Remark 7.).
6 Chang-Chang failure criterion (SOL 700 only).
7 Hashin failure criteria (SOLs 400, 600 only, see Remark 8.).
8 Puck failure criteria (SOLs 400, 600 only, see Remark 9.).
9 User-defined failure criteria (SOL 600 only, see Remark 10.).
10 Hashin-Tape (SOLs 400, 600 only, see Remark 11.).
11 Hashin-Fabric (SOLs 400, 600 only, see Remark 12.).
For Hashin models or Puck model, see Remarks 9., 10., 11., 12. for meaning of
material data.
Xt Maximum tensile stress in x-direction (Real, 0.0, or blank, no Default).
131
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Xc Maximum compressive stress (absolute value) in x-direction (Real, 0.0, or blank, no
Default).
Yt Maximum tensile stress in y-direction (Real, 0.0, or blank, no Default).
Yc Maximum compressive stress (absolute value) in y-direction (Real, 0.0, or blank, no
Default).
Zt Maximum tensile stress in z-direction (Real, 0.0, or blank, no Default).
Zc Maximum compressive stress (absolute value) in z-direction (Real, 0.0, or blank, no
Default).
Sxy Maximum shear stress in xy-plane (Real, 0.0, or blank, no Default).
Syz Maximum shear stress in yz-plane (Real, 0.0, or blank, no Default).
Szx Maximum shear stress in zx-plane (Real, 0.0, or blank, no Default).
Find Failure index. See Remarks 5.-7. (Real, 0.0, or blank, no Default).
Fxy Interactive strength constant for xy-plane (Real, 0.0, or blank, no Default).
Fyz Interactive strength constant for yz-plane (Real, 0.0, or blank, no Default).
Fzx Interactive strength constant for zx-plane (Real, 0.0, or blank, no Default).
Ext Maximum tensile strain in x-direction (Real, 0.0, or blank, no Default).
Exc Maximum compressive strain (absolute value) in x-direction (Real, 0.0, or blank, no
Default).
Eyt Maximum tensile strain in y-direction (Real, 0.0, or blank, no Default).
Eyc Maximum compressive strain (absolute value) in y-direction (Real, 0.0, or blank, no
Default).
Ezt Maximum tensile strain in z-direction (Real, 0.0, or blank, no Default).
Ezc Maximum compressive strain (absolute value) in z-direction (Real, 0.0, or blank, no
Default).
Gxy Maximum shear strain in xy-plane (Real, 0.0, or blank, no Default).
Gyz Maximum shear strain in yz-plane (Real, 0.0, or blank, no Default).
Gzx Maximum shear strain in zx-plane (Real, 0.0, or blank, no Default).
PF
(SOL 600
only)
Enter the character string PF to start progressive failure input data if ITYPE is 2 or
3. If the defaults are to be taken, the PF line and the line following the PF line may
be omitted (Character, no Default).
A1
(SOL 600
only)
Residual stiffness fraction. For criteria = 3, this is the fraction of initial stiffness
upon failure. For criteria = 2, the stiffness is not reduced more than this fraction
(Real, Default = 0.01).
Field Contents
MD Nastran R2.1 Release Guide

132
A2
(SOL 600
only)
Must be 0.0 or blank except for the criteria values listed below.
Criteria = 7 (Hashin)A2 is the factor for E2 reduction due to matrix compression
failure. Takes values between 0.0 and 1.0 and defaults to 0.0 where E2 is reduced in
the same way as for matrix tension. A value of 1.0 leads to no E2 reduction due to
matrix compression failure.
Criteria = 10 (Hashin Tape)Same as for criteria = 7.
Criteria = 8 (Puck)Same as for criteria = 7.
A3
(SOL 600
only)
Must be 0.0 or blank except for the criteria values listed below.
Criteria = 7 (Hashin)A3 is the factor for G12 reduction relative to E2 reduction. It
takes values between 0.0 and 1.0 and defaults to 0.0 where G12 is reduced in the
same way as E2. A value of 1.0 leads to no G12 reduction.
Criteria = 10 (Hashin Tape)Same as for criteria = 7.
Criteria = 8 (Puck)Same as for criteria = 7.
A4
(SOL 600
only)
Must be 0.0 or blank except for the criteria values listed below.
Criteria = 7 (Hashin)A4 is the factor for E3 reduction due to fiber failure. It takes
values between 0.0 and 1.0 and defaults to 0.0 where E3 is reduced in the same way
as E1. A value of 1.0 leads to an E3 reduction due to E2 only. Values between 0.0
and 1.0 lead to a mixture of degradation from matrix and fiber failure.
Criteria = 10 (Hashin Tape)Factor for E3 reduction due to fiber failure. It takes
values between 0.0 and 1.0 and defaults to 0.0 where E3 is reduced in the same way
as E1. A value of 1.0 leads to an E3 reduction due to E2 only. Values between 0.0
and 1.0 lead to a mixture of degradation from matrix and fiber failure.
Criteria = 8 (Puck)Same as for criteria = 7.
A5
(SOL 600
only)
Must be 0.0 or blank except for the criteria values listed below.
Criteria = 7 (Hashin)A5 is the factor for G12 reduction from fiber failure and takes
values between 0.0 and 1.0. It defaults to 0.0 where G12 reduces to matrix failure.
A value of 1.0 leads to G12 reduction due to only fiber failure. Values between 0.0
and 1.0 lead to a mixture of degradation from matrix and fiber failure.
Criteria = 10 (Hashin Tape)Same as for criteria = 7.
Criteria = 8 (Puck)Same as for criteria = 7.
IC1
(SOL 600
only)
For all Criteria values except those listed below, IC1 is set to 1 if failure in the
positive x-direction is critical (leads to element deactivation).
Criteria = 7 (Hashin)Set IC1=1 if fiber tension is critical (leads to element
deactivation).
Criteria = 10 (Hashin Tape)Same as for criteria = 7.
Criteria = 8 (Puck)Same as for criteria = 7.
(Integer, Default = 0).
Field Contents
133
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
IC2
(SOL 600
only)
For all criteria values except those listed below, IC2 is set to 1 if failure in the
negative x-direction is critical.
Criteria = 7 (Hashin)Set IC2 = 1 if fiber compression is critical.
Criteria = 10 (Hashin Tape)Same as for criteria = 7.
Criteria = 8 (Puck)Same as for criteria = 7.
(Integer, Default = 0).
IC3
(SOL 600
only)
For all criteria values except those listed below, IC3 is set to 1 if failure in the positive
y--direction is critical.
Criteria = 7 (Hashin)Set IC3 = 1 if matrix tension is critical.
Criteria=10 (Hashin Tape)Same as for criteria = 7.
Criteria = 8 (Puck)Same as for criteria = 7.
(Integer, Default = 0).
IC4
(SOL 600
only)
For all criteria values except those listed below, IC4 is set to 1 if failure in the
negative y--direction is critical.
Criteria = 7 (Hashin)Set IC4 = 1 if matrix compression is critical.
Criteria = 10 (Hashin Tape)Same as for criteria = 7.
Criteria = 8 (Puck)Set IC4 = 1 if matrix compression mode B is critical.
(Integer, Default = 0).
IC5
(SOL 600
only)
For all criteria values except those listed below, IC5 is set to 1 if failure in the positive
z--direction is critical.
Criteria = 7 (Hashin)Not used, leave blank.
Criteria = 10 (Hashin Tape)Not used, leave blank.
Criteria=8 (Puck)Set IC5 = 1 if matrix compression mode C is critical.
(Integer, Default = 0).
IC6
(SOL 600
only)
For all criteria values except those listed below, IC6 is set to 1 if failure in the
negative z--direction is critical.
Criteria = 7 (Hashin)Not used, leave blank.
Criteria = 10 (Hashin Tape)Not used, leave blank.
Criteria = 8 (Puck)Not used, leave blank.
(Integer, Default = 0).
Field Contents
MD Nastran R2.1 Release Guide

134
Element Birth and Death
Starting with this release, it is possible to deactivate and reactivate elements in the model that have failed,
or for some other reason need to be deactivated or reactivated. This is accomplished using Case Control
commands DEACTEL and ACTIVAT, and matching Bulk Data entries DEACTEL and ACTIVAT,
respectively. Once an element is deactivated or activated, it stays that way during the entire subcase
unless it fails due to a MATF criteria. See the MD Nastran Quick Reference Guide for further details.
Unglue
In contact analysis, it is often known beforehand that two surfaces will never separate once they contact.
To prevent numerical chattering, contact between these surfaces can be described using glued contact. In
order to perform a VCCT analysis of such surfaces, it may be necessary to unglue those nodes near a
crack. A new Bulk Data entry, UNGLUE, is available for such purposes. See the MD Nastran Quick
Reference Guide for further details.
IC7
(SOL 600
only)
For all criteria values except those listed below, IC7 is set to 1 if failure in the xy
plane is critical.
Criteria = 7 (Hashin)Not used, leave blank.
Criteria = 10 (Hashin Tape)Not used, leave blank.
Criteria = 8 (Puck)Not used, leave blank.
(Integer, Default = 0).
IC8
(SOL 600
only)
For all criteria values except those listed below, IC8 is set to 1 if failure in the yz
plane is critical.
Criteria = 7 (Hashin)Not used, leave blank.
Criteria = 10 (Hashin Tape)Not used, leave blank.
Criteria = 8 (Puck)Not used, leave blank.
(Integer, Default = 0).
IC9
(SOL 600
only)
For all criteria values except those listed below, IC9 is set to 1 if failure in the zx
plane is critical.
Criteria = 7 (Hashin)Not used, leave blank.
Criteria = 10 (Hashin Tape)Not used, leave blank.
Criteria = 8 (Puck)Not used, leave blank.
(Integer, Default = 0).
Field Contents
135
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Composite Element Numerical Analysis
In previous versions, SOL 600 provided two options for composite analyses: (1) complete through-the-
thickness integration at every iteration, and (2) the smeared approach as used in other MD Nastran
solution sequences. The first approach is more accurate, particularly for nonlinear analyses where local
buckling takes place and the analysis needs to extend well into the post-buckling regime. The second
approach is usually satisfactory for small deformation linear static and dynamic analyses. Method 1,
complete integration through the thickness, has been modified such that the accuracy has been retained,
but the computational times and memory requirements have been significantly reduced. These are
known as fast integration techniques and are described by the new Bulk Data entry PCOMPF. The
limitation in using these fast integration procedures is that the material may not exhibit any nonlinear
behavior. Large deformation and buckling is supported using these procedures.
The following table indicates typical performance improvements with this release when using this
option.
See the MD Nastran Quick Reference Guide for further details.
PLOAD4 Extensions
Previous SOL 600 versions did not support the PLOAD4 continuation line. In addition, if corner
pressures with different values were entered, they were averaged. SOL 600 now fully supports different
corner pressurespressures specified by the CID, N1, N2, and N3 fields on the continuation line, and line
loads specified by the CID, N1, N2, and N3 fields. The SOLR field is fully supported. The LDIR field
is not supported. Line loads must be specified using the CID, N1, N2, and N3 fields rather than LDIR.
For SOL 600, the SORL field applies to CQUAD4, CTRIA3, CQUAD4R, and CTRIAR elements. The
CID field may reference an ID of any CORD1R, CORD1C, CORD1S, CORD2R, CORD2C, or
CORD2S entry, but not CORD3G.
All of these new PLOAD4 extensions are activated by entering PARAM,MRPLOAD4,2 in the Bulk Data
Section, or by placing this parameter in one of the RC files. For this release, these capabilities are not the
default and must be activated using this parameter.
Model
Number of
Element
Maximum Number of
Layers
Improvement in CPU
for Stiffness Matrix
Memory
Reduction
1 10000 46 1425 % 1550 %
2 88854 182 2013 % 957 %
3 48858 33 1085 % 795 %
4 420015 48 1263 % 886 %
MD Nastran R2.1 Release Guide

136
Large Rotation RBE
Improved large rotation RBE capabilities have been added to SOL 600; however, it was decided to retain
the small rotation formulation as the default. This allows models used with previous versions to obtain
the same results. To activate the large rotation RBE capability, the following parameter is added to the
Bulk Data Section:
PARAM,marc7601,1
Streaming Input
A new capability available in this release is known as streaming input. Normally, SOL 600 will form
a Marc input file from the MD Nastran input file, then execute Marc to compute the results. With
streaming input, Marc is not executed. Instead, the Marc subroutines, which are now in MD Nastran,
are used directly and are called by the main SOL 600 routine. The Marc input file is still formed and saved
on disk for possible future use, but the same information is passed in memory from the main SOL 600
routine to the Marc initiation routine, thus saving computer time because a physical Marc input file on
disk does not need to be opened and read. Streaming input is activated using the parameter
PARAM,MRSTREAM,1
Streaming input is not available with DDM (parallel processing), or if user subroutines are necessary. For
those cases, do not include PARAM,MRSTREAM and run SOL 600 as before. Streaming input should
also not be used if a special version of Marc is necessary for your particular application. In that case,
specify one of the PATH options on the SOL 600 Executive Control statement in conjunction with a file
to point the analysis to the location of the version of Marc to be used.
Connector Technology
CBUSH, CWELD, and CFAST connection elements have been added to SOL 600 as nonlinear (large
deformation and rotation) elements. These formulations are now available in SOL 600 using
PARAM,MARCWELD (see the MD Nastran Quick Reference Guide for additional details). Be aware
that the CBUSH, CWELD, CFAST, RBE2, and RBE3 elements in SOL 600 are truly nonlinear elements,
while in their solution sequences they are linear small deformation elements. Thus, different results will
sometimes be obtained. In addition, solutions that converge with SOLs 106, 129 or 400 may not converge
with SOL 600 (the opposite may also occur). The user should be careful when using such elements to
make sure they are applicable to both linear and nonlinear solutions when constructing the model.
The new formulation of the CBUSH entry is activated using PARAM,MARCBUSH,-1. If CBUSH
entries are used with a nonzero CID, then PARAM,MRCOORDS,0 should also be included.
The new formulation of the CFAST and CWELD elements is activated using the parameter
MARCWELD,1.
The other new parameters that control CWELD behavior are:
MARIPROJ
MRCWANGL
137
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
MRFACEA
MRFACEB
MRHERRMN
MRITTYPE
The following CWELD parameters are not supported in SOL 600:
CWDIAGP
CWRANDEL
The following CWELD parameter is supported in SOL 600:
CWLDIGNR
If the new CWELD is used, then MSPEEDCW is ignored.
The following CFAST parameters are not supported in SOL 600:
CFDIAGP
CFRANDELT
Computational Enhancements
The direct iterative solver (MARCSOLV=2) can now be used with out-of-core assembly to allow larger
models to be analyzed. Note that there is a resulting decrease in performance. The program will
automatically use the out-of-core option if necessary, or the option can be selected using the
MARCOOCC parameter.
Superelements/DMIG
The use of superelements in conjunction with SOL 600 has increased over the last year, and several
changes have been made in this area. In addition to the MDMIOUT that may be used to create a reduced
stiffness matrix (superelement) or Adams .mnf file, the DMIGOUT option has been added.
The DMIGOUT option may be used to output either the complete global matrices, or individual element
matrices at the element levels. These global matrices include:
Stiffness matrix This is the total stiffness matrix, including geometric stiffness, follower force,
and friction contributions.
Differential stiffness This is only available in a buckling subcase.
Mass matrix Available in dynamics.
Damping matrix Available in dynamics.
Conductivity matrix Available in heat transfer.
Specific heat matrix Available in transient heat transfer.
MD Nastran R2.1 Release Guide

138
The element matrices correspond to the previously described matrices, and can be output in either the
basic system or in the transformed system. The global stiffness matrices are always output in the
transformed system.
Note that the output of these matrices may be extensive. To reduce the size of the DMIG files, it is
possible to filter out small values. However, doing this may influence subsequent calculations.
Contact Enhancements
The spline option used with deformable bodies via the BCBODY option can now be used with higher-
order elements. This improves the calculation of when contact is to occur, and of the normal to the
surface.
New and Improved SOL 600 Features for MD Nastran R2.1.
Permanent glue, similar to the capability in SOLs 101 and 10,3 was added in this release. This
can be used for all IDs in SOL 600 and is triggered by adding PERMGLUE on the SOL 600
entry (see the MD Nastran Quick Reference Guide for a complete discussion).
MARCOUT was changed to easily specify composite output for the top and bottom layers, or
top/bottom/middle layers, by adding a new variable in field 2 of the MARCOUT entry named
LAYCODE (see the MD Nastran Quick Reference Guide for a complete discussion). New post-
code selections for a variety of composite output, including progressive failure, may now be
selected using the MARCOUT entry. This new output is only available in the .t16 file.
A new Bulk Data entry, MLAYOUT, was added to easily specify which layers should be output
in addition to the top and bottom layers.
In the previous release, the COHESIV Bulk Data entry was not applicable to any interface
elements. The appropriate interface elements have now been added. When CQUAD4, CHEXA,
etc., elements refer to a PSHELL or PSOLID entry that references a COHSIV material ID, the
proper Marc interface elements are automatically generated.
A new Marc solver, number 10 (mixed directive/iterative solver), was added and is triggered
either by NLSTRAT ISOLVE=10, or by PARAM,MARCSOLV,10.
Additional soft spring-to-ground options were added by introducing a new Bulk Data entry,
MGRSPR.
An option was added to reverse tables which might have been entered backwards using a new
Bulk Data entry, MTABRV.
Marc rebar elements and properties were added using new Bulk Data entries CMREBAR,
CMREBAI, PMREBAR, and PMREBAI.
Several methods of entering multivariable (three-dimensional) tables were added in this release.
Refer to the MD Nastran Quick Reference Guide, TABL3Di entries, for further details.
MATF was extended to add new progressive failure options.
All applicable active Marc print parameter options were added using a new Bulk Data entry,
MARPRN.
139
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Fast composite shell integration was added. It may be specified with PARAM,MFASTCMP if
the integration method is the same for all composite shells, or with the Bulk Data entry
PCOMPF if different composite shells require different methods.
Streaming input is automatically turned off for analyses where it does not work properly, such as
parallel processing and for I8 versions.
The length of the command to execute Marc was increased to 640 characters to accommodate
MSC Sim Designer, and customers with extremely long paths to either MD Nastran or to the
input file being processed.
Tube elements are automatically changed to bar elements for cases where the Marc tube element
does not support various features.
Output in the .op2, .f06, and punch files is now available for multilayer composite shells and
axisymmetric models.
An error in PBUSHT was corrected. In addition, it is no longer necessary to enter
PARAM,MARCBUSH; however, leaving this parameter out or entering PARAM,MARCBUSH,-1
will produce the same results. Other values of MARCBUSH should no longer be used.
New SOL 600 Parameters
The following new SOL 600 parameters are described as follows. See the MD Nastran Quick Reference
Guide for more details:
PARAM,MARMTLCK Determines whether a check of various property-material combinations
for SOL 600 will be made or not. This slows down the input processing
slightly.
PARAM,MARLDCMB Determines whether extraneous loads in the input file will be combined to
save computer time.
PARAM,MARLDRMV Determines whether extraneous FORCE, MOMENT, and/or PLOAD4
entries will be filtered out at an early stage to save computer time.
PARAM,MARNOCID SOL 600, by default, does not support an MCID defined by cylindrical or
spherical coordinate systems. This parameter determines whether MCID
defined by cylindrical or spherical coordinate systems will be ignored or
cause fatal errors for shell and solid elements, depending on this
parameter. Inclusion of the parameter overcomes this problem.
PARAM,MRCPENTA Determines how CPENTA will be mapped to Marc degenerate solid
elements. Marc does not presently have wedge elements, so CPENTA
elements must be mapped to degenerate hexagonal elements such as type
7.
MD Nastran R2.1 Release Guide

140
PARAM,MRHERRMN Controls whether extra grid points created for such items as hyperelastic
Herrmann elements, CWELD, etc., are output in the .op2, .f06, punch,
and/or .xdb files. When Herrmann grid points are output, the displacement
value is actually pressure, which might be confusing when looking at an
.f06 file.
PARAM,MARCMID3 Controls whether MID3 will be set to the same value as MID2 when the
Marc PSHELL option is used (designated by PARAM,MRPSHELL,1 or
when the SMEAR option is used on the SOL 600 Executive Control
statement).
PARAM,MFORDUPE Controls how duplicate forces encountered for the same load case are
handled in SOL 600.
PARAM,MARBK105 Controls whether linear buckling or nonlinear buckling eigenvalues are
calculated for SOL 600.
PARAM,MRCTRIA3 Controls the default Marc element type for CTRIA3 elements in SOL
600.
PARAM,MRCQUAD4 Controls the default Marc element type for CQUAD4 elements in SOL
600.
PARAM,MARCDUPE Controls whether SOL 600 will check for duplicate entries for most types
of Bulk Data. SOL 600 does not allow duplicate entries, but the portion
of IFP that runs prior to spawning Marc does not usually check for
duplicate entries.
PARAM,MARCSTOP Controls whether a check model run will be performed; no actual
simulation will occur; and the analysis will stop with a Marc Exit 7.
PARAM,MRRSTOP2 Normally, .op2, .xdb, punch, and .f06 output is not available for SOL 600
restart analysis. Setting this parameter to 1 causes the program to attempt
to create one or more of these files. Only in limited cases will the job be
successful.
PARAM,MRBDYCVT Determines if CHBDYG is converted to CHBDYE for SOL 600 heat
transfer.
PARAM,MRMATFSB Determines whether or not the version of Marc being used supports the
new MATFSB field.
PARAM,MRDELTTT Determines how delta time is set for each step of a SOL 600 transient
nonlinear analysis.
PARAM,MARCFEAT,N Adds FEATURE,N to the Marc input file in the parameter section.
PARAM,MRCOORDS Determines whether Marc COORD SYSTEM will be added if any
CORD1i or CORD2i entries are in the model, and if CBUSH elements are
present in the model.
PARAM,MARCWELD Determines how CWELD/PWELD and CFAST/PFAST elements will be
translated to Marc.
141
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Platform-Specific Notes
For Linux IA64 and EM64T platforms
The default MPI for these platforms is HP MPI.
Intel MPI (iMPI) is also supported, and can be used by switching to it using the maintain script under the
tools directory.
To use the Intel MPI, note the following:
1. Create an .mpd.conf file in the home directory that contains the following line:
secretword=<your mpd password>
where <your mpd password> can be any arbitrary string.
Change permission of the .mpd.conf to 600 by entering
chmod 600 $HOME/.mpd.conf
2. Set up an mpd.hosts file in the home directory consisting of the names of nodes in your cluster (it
can have only one node, i.e., one line):
clusternode1
clusternode2
clusternode3
The rest is taken care of by the run_marc script.
If your cluster requires a password to perform ssh or rsh between nodes, you may need to enter your
password every time you run a parallel job. It may be possible to disable the password requirement by
consulting your system administrator.
For 64-bit Windows EM64T Platforms
The default MPI for this platform is MPICH2. MS MPI is not supported for this release of MD Nastran.
To install MPICH2, enter the mpichx64\bin directory and enter
smpd -install
PARAM,MRITTYPE Type of constraint used to connect the auxiliary grid points in all
CWELDs.
PARAM,MRCWANGL Angle in degrees through which to rotate the CWELD cross section about
the beam axis to obtain its final orientation.
PARAM,MARIPROJ Flag to determine if auxiliary nodes of a CWELD will be projected on the
model.
PARAM,MARFACEA Face number for A side of weld if welds are made of solid elements.
PARAM,MARFACEB Face number for B side of weld if welds are made of solid elements.
MD Nastran R2.1 Release Guide

142
For 32-bit Windows Platforms
The default MPI for this platform is Argonne National Labs MPICH2.
To use MPICH2, observe the following:
1. Enter the mpich2\bin directory and enter
smpd install
If smpd could not be installed, the system may need to be rebooted.
2. For the first run of a parallel job, enter your login ID and password.
If the system is rebooted or the command prompt is closed, repeat step 2.
Supported Systems for SOL 600 in MD Nastran R2.1
Vendor OS Hardware
FORTRAN
Version
C
Version
Parallel
Enabled Default MPI
Also Works
On
HP-Alpha
(DEC)
4
Tru64 5.1 Alpha
Server 4100
f90 5.5 cc 6.4 yes HP MPI 2.0
1
HP (64-bit)
2,4
HP (64-bit)
2,4
HPUX 11.0
HPUX 11.23
PA2.0
Itanium 2
f90 2.9.2
f90 2.8.7
C.03.50
A.06.02
yes
yes
HP MPI 2.0
HP MPI 2.2
IBM (64-bit)
4
AIX 5.2 RS/6000 &
RS/6000 SP
xlf 8.1.1 cc 6.0.0 yes MPICH
1
IBM POE 4.1
SGI (mips4
64-bit)
2,3,4
SGI (Altix 64-
bit)
2, 4
IRIX 6.5
Linux 2.4.21-
sgi303r2
R12000
Itanium 2
(Propack
3.0)
f90 7.4
Intel 8.1
cc 7.4
Intel 8.1
yes
yes
MPICH
1
SGI MPT 1.10.1 Propack 4.0
Sun (64-bit)
4
Solaris 2.8 UltraSPARC
III
f90 8.1 cc 5.7 yes MPICH
1
Linux (32-bit) Red Hat 9 Intel
Pentium or
equiv.
Intel 8.1 Intel 8.1 yes HP MPI 2.2.5 Red Hat AS 3.0
Linux (64-bit)

4,5
Red Hat AS 3.0 Itanium 2 Intel 8.1 Intel 8.1 yes HP MPI 2.2.5
5
Linux (64-bit)

4,5
Red Hat WS
3.0
Intel EM64T Intel 8.1 Intel 8.1 yes HP MPI 2.2.5
5
AMD Opteron,
Red Hat WS 4.0
Intel (32-bit)
Intel (64-bit)
4
Windows 2000
Windows
Server 2003
x64
Intel
Pentium or
equivalent
Intel EM64T
Intel 8.1
Intel 8.1
Intel 8.1
Intel 8.1
yes
yes
MPICH2
MPICH2
Windows XP,
Intel 9.1
MPICH2,
Windows XP 64
Note: 1. Hardware MPI version also available (via maintain in /tools directory).
2. Supports Solver 6.
143
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
3. Supports multi-threading.
4. Supports true 64-bit version.
5. Supports the Intel MPI 3.0
MD Nastran R2.1 Release Guide

144
MD Nastran Explicit Nonlinear - SOL 700
Introduction
SOL 700 offers advanced technology with which to analyze transient dynamic events of short duration
with severe geometric and material nonlinearities.
MD Nastran SOL 700 uses one common modeling environment featuring the same Bulk Data interface.
NVH, linear, and nonlinear models can be used for explicit applications such as crash, crush, drop test,
blade out, and bird strike simulations. This dramatically reduces the time spent to build different models
for implicit and explicit analyses, and reduces the chance of making mistakes while working with
different programs.
Latest Capabilities of MD Nastran Explicit Nonlinear - SOL 700
SOL 700 includes the following new capabilities in this release:
Fluid structure interactionairbag and occupant safety simulation.
Time domain noise, vibration, and harshness (NVH)Fast Fourier transform (FFT) method.
Prestress with implicit solver.
Many additional elements and material models.
Support for MD Nastran native output files.
Airbags and Occupant Safety
In this release, SOL 700 includes fluid structure interaction (FSI) capability based on the advanced finite
volume (Eulerian) and general coupling technology available in Dytran. However, the FSI capability is
limited to airbag and occupant safety simulation.
The objective of fluid-structure interaction using the coupling algorithm is to enable the material modeled
in Eulerian and Lagrangian meshes to interact. Initially, the two solvers are separate. Lagrangian
elements that lie within an Eulerian mesh do not affect the flow of the Eulerian material, and no forces
are transferred from the Eulerian material back to the Lagrangian structure. The coupling algorithm
computes the interaction between the two sets of elements, thereby allowing complex fluid-structure
interaction problems to be analyzed.
The first task in coupling the Eulerian and Lagrangian sections of a model is to create a surface on the
Lagrangian structure. This surface is used to transfer the forces between the two solver domains. The
surface acts as a boundary to the flow of material in the Eulerian mesh. At the same time, the stresses in
the Eulerian elements cause forces to act on the coupling surface, distorting the Lagrangian elements.
The method used for airbag simulation is based on general coupling with adaptive Euler. SOL 700
computes the fluid flow and coupling based on the Dytran solver, while the contact and fabric models are
cosimulated by LS-DYNA. Unlike other techniques such as the arbitrary Lagrange Euler (ALE)
technique, wherein the Eulerian mesh is fixed in space; or the GBAG method, wherein the gas flow is
145
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
modeled by applying a predetermined pressure profile to inflate the bag; in the general coupling
technique, the Eulerian mesh will adapt itself to the Lagrangian fabric model as the airbag is inflated.
In other words, initially, when the airbag is at the folded stage, there is a small Eulerian domain
encapsulating the Lagrangian mesh. When the airbag is inflated, the Eulerian mesh expands as the gas
jet flows through the airbag compartments and adapts itself to follow the airbag fabric. This technique is
unique to MD Nastran R2.1, and is considered the most accurate method to predict complex airbag
behavior such as out-of-position (OOP) simulation as required by FMVSS 208, wherein the occupant is
leaning forward when the airbag is inflated. The multicompartmented airbags are deployed using
multiple, fully automatic, adaptive Euler domains. The following capabilities are available:
Analysis of multiple compartments with the Eulerian fluid approach.
Simulation of flow from one Eulerian fluid domain into another Eulerian fluid domain.
The individual Eulerian fluid domains are dynamic and adaptive. The user does not need to
mesh the Eulerian fluid domains, nor worry about the size, since the Eulerian fluid domains will
automatically follow the deploying airbag compartments.
Accurate calculation of flow through both small and large holes.
Inflator Models
Figure 4-1 Shape of side curtain airbag at start of simulation.
Figure 4-2 Euler meshes for all six regions.
MD Nastran R2.1 Release Guide

146
Figure 4-3 Shape of bag after 20 milliseconds.
Defining Inflators in Airbags
There are several methods available to define an inflator in airbag analysis. The most general and
extended inflator definitions are:
INFLATR1 - Standard inflator defined by mass flow rate and static temperature of a single inflowing gas.
INFLHYB1 - Hybrid inflator defined by mass flow rate and static temperature of multiple inflowing
gasses.
Figure 4-4 Airbag and occupant safety using SOL 700.
The new Bulk Data entry AIRBAG in MD Nastran R2.1 facilitates the airbag simulation. Basically, all
required entries for airbags can be defined by using the AIRBAG Bulk Data entry.
Occupant Dummy Models
Occupant dummy models, also known as anthropomorphic test devices (ATDs), were introduced in the
previous release of SOL 700 for applications that do not involve airbags (many applications, such as sled
tests and aircraft seat design, or armored vehicle design where the occupant behavior is studied when a
land mine is detonated, do not require airbag simulation).
147
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
SOL 700 supports occupant dummies that are readily available in LS-DYNA *key file format. These
include:
LS-DYNA Public Domain Dummies
5th percentile deformable female dummy.
50th percentile deformable male dummy.
95th percentile deformable male dummy.
5th percentile rigid female dummy.
50th percentile rigid male dummy.
95th percentile rigid male dummy.
Engineering Technology Associates (ETA) Calibrated Dummies
The ETA calibrated dummies are similar to LS-DYNA dummies except that they are validated against a
set of standard tests. The calibration tests are conducted on all Hybrid III, SID, and EUROSID models
to ensure model fidelity.
5th percentile deformable female dummy.
50th percentile deformable male dummy.
95th percentile deformable male dummy.
5th percentile rigid female dummy.
50th percentile rigid male dummy.
95th percentile rigid male dummy.
EUROSID - 1 (Euro side impact dummy).
US Department of Transportation (DOT) SID.
European Enhanced Vehicle Safety Committee (EEVC) upper legform.
Federal Motor Vehicle Safety Standard (FMVSS) 201 headform.
Figure 4-5 Typical occupant dummies in virtual proving ground (VPG). Courtesy of ETA.
There are many other ATD models available through the ETA virtual proving ground (VPG) pre- and
postprocessor. In addition, VPG supports the following barriers:
MD Nastran R2.1 Release Guide

148
Barriers
Federal Motor Vehicle Safety Standard (FMVSS) &Economic Commission for Europe (ECE)
side impact barriers.
0 degree front rigid barrier.
30 degree rigid front barrier.
Front offset deformable barrier.
Rear impact barrier.
Impactors
Rams.
Pendulums.
Head forms.
For more details, contact Engineering Technology Associates in Troy, Michigan.
First Technology Safety Systems (FTSS) Anthropomorphic
Test Dummy (ATD)
The FTSS ATDs are high fidelity dummies and are available with additional licensing. The following
FTSS dummies are supported through MSC.Software SimX Crash or ETA VPG:
5% female dummy.
SID II.
Hybrid III, 3% child dummy (W.I.P.).
Hybrid III, 6% child dummy (W.I.P.).
BIOSID (W.I.P.).
Pre- and Postprocessing
Even though MD Patran supports SOL 700, it does not have dummy positioner or airbag folder
capabilities. Dummy positioning can be done using pre- and postprocessing tools such as MSC.Software
SimX Crash and ETA-VPG, both of which support SOL 700. For more information regarding airbag and
occupant analysis, refer to the MD Nastran Explicit Nonlinear (SOL 700) Users Guide and the MD
Nastran Quick Reference Guide.
Time Domain NVH
This is a new methodology in MD Nastran SOL 700 that is used to compute the natural frequencies of a
structure, similar to experimental modal identification using impact or other transient testing. The user
applies an impact, and simulates driving a vehicle over a rough road, or some other type of transient
loading to the structure, such that the loading is likely to excite all the important modes. A standard
SOL 700 analysis is performed to output accelerations, velocities, and/or displacements at selected grid
149
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
points using a reasonably fine output delta time. Upon completion of the nonlinear analysis, a
postprocessing operation is used to transform the selected outputs from the time domain to the frequency
domain using fast Fourier transforms (FFTs). Various criteria are available to select which peaks will be
approximately defined as modes (eigenvectors). This approximation is reasonably accurate if the
damping is small. The amplitudes of all the selected degrees of freedom for each of the selected modal
frequencies are normalized and become the eigenvectors, which are then saved on a file. A restart
capability is available to change the criteria selections. In addition, a user may pick modes from plots
made from data saved during the first run, and then compute the eigenvectors associated with the chosen
natural frequencies.
The advantages of time domain NVH include the ability to consider the material and structural
nonlinearity, damping, and contact interface between various components. Thus, instead of using a linear
implicit method, an explicit approach is undertaken to first predict the nonlinear behavior of the
structure, and then compute the NVH characteristics by the FFT method. The double precision version
of LS-DYNA is used to ensure higher fidelity of the solution.
The disadvantage of the time domain NVH technique is the excessive CPU time required, since an
analysis may be required to run for a few secondsconsidered a long runtime for an explicit simulation
before an FFT is performed to compute frequencies and mode shapes. However, this problem is
becoming less significant due to the advanced distributed memory parallel (DMP) technology in
SOL 700, and to dramatic hardware performance improvements.
Figure 4-6 Time domain NVH of vehicle running on proving grounds (courtesy of
Engineering Technology Associates).
a. Time domain data. b. Frequency domain data.
FFT Process
MD Nastran R2.1 Release Guide

150
To define a time domain NVH simulation, refer to the new entries of TIMNVH, TIMSML, and TIMNAT
in the MD Nastran Quick Reference Guide, and see examples in the MD Nastran SOL 700 Users Guide.
A simple example is given as follows.
Time Domain NVH Example
SOL 700 entries included:
SOL 700
TSTEPNL
PARAM,S700NVH1
TIMNVH
TIMNAT
Example: Time Domain NVH Example
Filename: timnvh.bdf, timnvh1.bdf
Description
This is an example of a virtual dynamic test. A plate was modeled and an impulse loading was applied at
one of the corner points, as shown in Figure 4-7. Time histories were obtained at six points, and they were
translated by the fast Fourier transformation method to the frequency domain in order to obtain modal
frequencies and shapes.
Model
A plate was modeled in this example. A fixed boundary condition was applied along one end of the plate.
The model has a total of 231 grid points and 200 quadratic elements. All shell elements use the
Belytschko-Wong-Chiang formulation. The impact loading at the corner was modeled by defining the
load time history shown in Figure 4-7. The simulation time is 1.024 seconds.
151
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Figure 4-7 Time domain NVH example.
MD Nastran R2.1 Release Guide

152
Time NVH scheme
Input
The loading, which is specified by FORCE and TABLE1 Bulk Data entries, was applied on grid point
231. Using TIMNVH, TIMNAT, and PARAM, S700NVH1, dynamic properties of the plate were
obtained. Three steps were used to acquire the dynamic properties, as follows:
Step 1: Find approximate modal properties using the TIMNVH Bulk Data entry.
Step 2: Check the obtained modal properties and select required natural frequencies.
Step 3: Rerun with selected natural frequencies.
Input File timnvh.bdf
SOL 700 is an Executive Control statement similar to SOL 600. It activates an explicit nonlinear transient
analysis.
SOL 700,NLTRAN path=3 stop=1
153
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
The Case Control Section is as follows.
CEND
TITLE = MD Nastran job created on 12-Dec-06 at 11:21:25
LOADSET = 1
$ Direct Text Input for Global Case Control Data
SUBCASE 1
TITLE=This is a default subcase.
TSTEPNL = 1
SPC = 2
DLOAD = 2
DISPLACEMENT(SORT1,REAL)=ALL
SPCFORCES(SORT1,REAL)=ALL
The Bulk Data Section is as follows:
BEGIN BULK
TSTEPNL 1 100 .01024 1 ADAPT 2 10
TIMNVH,1,,,1.0,500.,3,0.00005,2,+
+,0,3,1,0.015,0,3,13,.0030,+
+,11,21,116,126,221,231
TSTEPNL is a SOL 700 Bulk Data entry which describes the number of time steps (100) and time
increment (10.24 ms) of the simulation. End time is the product of the two entries. Notice that the time
increment is only for the first step. The actual number of time increments and the exact value of the time
steps is determined by LS-Dyna during the analysis. The time step is a function of the smallest element
dimension during the simulation.
TIMNVH is an entry for time NVH analysis.
The details of the TIMNVH entry are as follows:
TIMNVH, 1, , , 1.0, 500., 3, 0.00005, 2,+
The range of natural frequencies to be obtained is from 1.0 Hz to 500 Hz, and only the translational DOF
of the z-direction (3) is considered. The sampling rate is 0.00005 seconds. The peaking criterion is two,
which means a peak is assumed to be resonant if there are two points on each side of the peak that are of
increasing magnitude below the peaks frequency, and two points of decreasing magnitude above the
peaks frequency.
+, 0, 3, 1, 0.015, 0, 3, 13, .0030,+
Acceleration is selected for the response (0), and translational eigenvectors are only requested as ASCII
format (3). Eigenvalues are normalized by 1.0 (1) and 0.015 is selected as CLOSE in value, which means
that if there are two modes differing in frequency by less than 0.015 Hz, they are assumed to be the same
mode. ACII file formats of natural frequencies and eigenvalues are requested (0), and translational time
histories of the z-direction (3) are requested. Frequency-amplitude data of requested z-direction are
requested (13), and a peak whose amplitude is less than maximum amplitude is ignored
(.0030).
+,11,21,116,126,221,231
Only grid points 11, 21, 116, 126, 221, and 231 are considered for time NVH analysis.
0.0030 the
MD Nastran R2.1 Release Guide

154
The Bulk Data entries that define elements and grid points are:
CQUAD4 1 1 1 2 23 22
..
$ Nodes of the Entire Model
GRID 1 0. 0. 0.
..
The Bulk Data entry that defines material properties is:
$ Referenced Material Records
$ Material Record : steel
$ Description of Material : Date: 12-Dec-06 Time: 11:09:13
MAT1 1 2.+11 .3 7860.
MAT1 is an isotropic-elastic material Bulk Data entry.
The Bulk Data entry that defines properties for shell elements with .001 thickness.
PSHELL 1 1 .001 1 1
DLOAD 2 1. 1. 4
$ Nodal Forces of Load Set : impact
The Bulk Data entry that defines tables is shown below.
$ Dynamic Load Table : hammer
TABLED1 1
-10. 0. 0. 0. .001 1. .002 0.
10. 0. ENDT
Input File timnvh1.dat:
The following input refines the selection of modal frequencies and mode shapes.
PARAM,S700NVH1,1
TSTEPNL 1 100 .01024 1 ADAPT 2 10
TIMNVH,1,,,1.0,500.,3,0.00005,-2,+
+,0,3,1,0.015,0,3,13,.0030,+
PARAM, S700NVH is for the rerun of time NVH analysis. If PARAM,S700NVH=1, then time domain
NVH analysis is carried out without LS-DYNA rerunning.
The PEAK value (Default = 2) in the TIMNVH Bulk Data entry is changed to -2 to use the TIMNAT
Bulk Data entry.
The TIMNAT Bulk Data entry controls the natural frequency selection. In this run, 20.5, 88, 129, 285,
and 360 Hz are selected to obtain the results.
Results
There are three new types of results files from time NVH analysis:
1. mode.out: The selected natural frequencies and eigenvalues are restored.
2. ampl-freq-00000116-3.txt: Amplitude-frequency output of DOF =3 at grid point 116.
3. time-hist-00000116-3.txt: Time history output of DOF =3 at grid point 116.
155
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
From the ampl-freq-___ files, the resulting frequency-amplitude plots are as follows:
1.00E-02
1.00E-01
1.00E+00
1.00E+01
1.00E+02
1.00E+03
1.00E+04
0 100 200 300 400 500
Frequency (Hz)
A
m
p
l
i
t
u
d
e

(
a
c
c
e
l
e
r
a
t
i
o
n
)
node 11
node 21
node 116
node 126
node 221
node 231
1st mode = 20.5
2nd mode = 88.
3rd mode = 129.
4th mode = 285.
MD Nastran R2.1 Release Guide

156
The following results are obtained in the mode.out file:
A comparison of natural frequency results (Hz) between SOL 103 and SOL 700 is shown as follows:
Mode SOL 103 SOL 700 Diff(%)
1 2.0508E+01 2.0864E+01 1.74
2 8.7894E+01 8.8639E+01 0.85
3 1.2891E+02 1.2966E+02 0.58
4 2.8516E+02 2.8771E+02 0.89
5 3.6036E+02 3.6269E+02 0.65
MODES 1 18
EIGV 1 2.050821E+01
11-9.13892833E-06 3.86241316E-06 3.19331321E-01
21 2.14690732E-05-1.14613790E-05 9.95118635E-01
116-9.93770531E-06 2.13863912E-06 3.30875181E-01
126 7.27358928E-06-2.94809676E-06 1.00000000E+00
221-1.23613298E-05-3.06943202E-06 3.20993777E-01
231 1.35196599E-05 6.87379272E-06 9.99026322E-01
EIGV 2 8.789234E+01
11-5.69607103E-06 1.65922185E-06-6.09298304E-01
21 5.33656071E-06-7.67633319E-06-9.90453563E-01
116-3.47585317E-06 8.00428604E-07 1.97687095E-02
126 5.34868995E-06-3.96739670E-06-7.95414322E-03
221-4.07797686E-06-2.98546031E-06 5.90684541E-01
231 3.21720171E-06 7.30456344E-06 1.00000000E+00
Total number of modes
1st Natural Frequency
Node numbers
1st mode
2nd mode
2nd Natural Frequency
157
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
A comparison of mode shapes between SOL 103 and SOL 700 is shown in the following figures:
Mode SOL 103 SOL 700
1
2
MD Nastran R2.1 Release Guide

158
3
4
Mode SOL 103 SOL 700
159
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Prestressing (Implicit-to-Explicit Sequential Simulation)
Many applications require a prestress analysis prior to transient analysis. For example, in bird strike and
blade out analysis, the blades of a running jet engine have residual stresses due to high rotational
velocities. SOL 700 uses an implicit solver for prestress analysis to dramatically reduce the CPU time
required. Prestress calculations are performed by the LS-DYNA implicit double precision version where
the initial state of the model is written into a file that can only be used for a subsequent SOL 700 transient
run. This is an automated process, and there is no need for dynamic relaxation to eliminate the high
frequency oscillations.
Refer to the new PRESTRS Bulk Data entry in the MD Nastran Quick Reference Guide and in the MD
Nastran Explicit Nonlinear (SOL 700) Users Guide for more details.
5
6 points sampling
12 points sampling
Mode SOL 103 SOL 700
MD Nastran R2.1 Release Guide

160
Additional Capabilities of MD Nastran R2 Explicit Nonlinear -
SOL 700
The following are additional new capabilities in this release:
New Material Models
Force limited material MATD029.
Fabric material MATD034.
Nonlinear orthotropic material MATD040.
Concrete damage material MATD072.
Concrete damage material REL3 MATD72R.
Plasticity polymer MATD089.
Simplified Johnson Cook material MATD098.
Simplified Johnson Cook orthotropic damage material MATD099.
Finite elastic strain plasticity material MATD112.
Layered linear plasticity material MATD114.
Modified piecewise linear plasticity MATD123.
Rate sensitive composite fabric material MATD158.
General spring discrete spring material MATD196.
Elements and Boundary Conditions
Breakable join BJOIN.
Geometric rigid wall WALLGEO.
Spotweld beam property PBSPOT.
Spotweld material MATDSW1 to MATDSW5.
Connection spot weld CONSPOT.
Properties
Shell property PSHELLD.
Mass damping per property DAMPMAS.
Stiffness damping per property DAMPSTF.
Airbags
AIRBAG Defines an airbag.
PARAM, UGASC Defines a value for the universal gas constant.
PARAM, SBOLTZ Defines a value for the Stephan-Boltzmann constant.
PARAM, DYDEFAUL Controls the default setting of the simulation.
161
CHAPTER 4
Implicit and Explicit Nonlinear Analysis
Hybrid inflator gas fraction definition INFLFRC.
Gamma law gas equation of state EOSGAM.
Tabulated equation of state with compaction EOSTABC.
Tabulated equation of state EOSTAB.
Grid point in airbag reference geometry GRIA.
Prestress
ISTRSBE Initialize stresses and plastic strains in the Hughes-Liu beam elements.
ISTRSSH Initialize stresses, history variables, and the effective plastic strain for shell
elements.
ISTRSTS Initialize stresses, history variables, and the effective plastic strain for thick shell
elements.
ISTRSSO Initialize stresses and plastic strains for solid elements.
PRESTRS Perform prestress run to calculate an initially stressed model, and write out the
initial state to a file that can be used for a subsequent explicit SOL 700 run.
New Contact Features
IGLUE, JGLUE Glued contact features.
TIEDNSCO Tied nodes to surface constrained offset.
TIEDESCO Tied shell edge to surface constrained offset.
TIEDSSCO Tied surface to surface constrained offset.
SPOTWELD Spotweld contact.
SPOTWTOR Spotweld with torsion contact.
EDGE Single edge contact (master body not allowed).
FTRANSC Force transducer constraint.
FTRANSP Force transducer penalty.
Parameters
DYPARAM, LSDYNA, SOLID, [option], <value> Define control parameters that can
improve the accuracy of the calculation.
DYPARAM, LSDYNA, ACCURACY, [option], <value> Define control parameters that can
improve the accuracy of the calculation.
DYPARAM, LSDYNA, DATABASE, FORMAT, <value> Define the type of results output
files.
DYPARAM, LSDYNA, DATABASE, [option], <value> Define control parameters that
control output generation of the calculation.
MD Nastran R2.1 Release Guide

162
Miscellaneous
The use of ACC is no longer recommended, and it will be removed from the code in a future
version. Use ACCMETR instead.
The use of MATD100 is no longer recommended, and it will be removed from the code in a
future version. Use MATDSW1-5 in combination with PBSPOT.
Ch. 5: NVH and Acoustics MD Nastran R2.1 Release Guide

5
NVH and Acoustics
!
Frequency Response Function (FRF) and FRF-Based Assembly
(FBA) Feature in MD Nastran R2
!
Exterior Acoustics
!
Rigid Porous Absorber
MD Nastran R2.1 Release Guide

164
Frequency Response Function (FRF) and FRF-Based
Assembly (FBA) Feature in MD Nastran R2
Introduction
The frequency response function (FRF) and FRF-based assembly (FBA) are new capabilities in this
version. FRFs can be used to represent components. The FRFs of components can then be combined to
yield the FRFs of assemblies of the components. Such an FRF-based assembly process represents an
alternative to component mode synthesis (or modal substructuring). This method also lends itself to
tracking load paths or energy flows through a structure. This approach may be regarded as an alternative
to mode participation studies. This latter feature, commonly referred as transfer path analysis (TPA), is
not currently available, but will be incorporated in a post-MD Nastran R2 release.
FRF Concept
The equation of motion for frequency response analysis is given by
(5-1)
where is the dynamic stiffness matrix of the system.
Solving Eq. (5-1) for the response displacements yields
(5-2)
The matrix in Eq. (5-2) is the receptance matrix of the system. It is also commonly referred to as the
frequency response function (FRF) of the system. It should be noted that the FRF of a system is a function
of the forcing frequency . Thus, there is a different FRF for the system for each forcing frequency.
The rows of the FRF matrix represent response DOFs, and its columns represent excitation DOFs. Any
term of the FRF matrix represents the response displacement at the i-th DOF due to a unit load applied
to the j-th DOF. Thus, the rows in , the j-th column of the FRF, represent the displacements at all
response DOFs due to a unit load applied to the j-th DOF. Similarly, the columns in the i-th row of the
FRF represent the response displacements at the i-th DOF due to unit loads applied to all excitation
DOFs.
In practice, the FRF matrix is computed by solving Eq. (5-2) for a series of unit loads. Thus, consider the
following equation, which represents Eq. (5-2) for a series of unit loads on the right hand side:
(5-3)
The solution of Eq. (5-3) for the matrix of displacements yields the desired FRF matrix . The
matrix of displacements resulting from the solution of Eq. (5-3), which is equivalent to the
receptance or FRF matrix of Eq. (5-2), is square because Eq. (5-3) assumes that the responses are

2
M [ ] i B [ ] K [ ] + + ( ) u { } Z [ ] u { } P { } = =
Z [ ]
u { }
u { } Z [ ]
1
P { } H [ ] P { } = =
H [ ]

H
i j
H { }
j
U [ ] Z [ ]
1
I [ ] =
U [ ] H [ ]
U [ ]
H [ ]
165
CHAPTER 5
NVH and Acoustics
computed at all DOFs of the system, and also that unit loads are applied at all DOFs. However, this is
rarely the case in practice. In most practical cases, responses are computed at only a subset of the total
points in a configuration, and loads are applied only to a small subset of the total DOFs in the system.
Therefore, in practice, the actual FRF matrix to be considered for analysis purposes is really a subset (or
smaller partition) of the matrix given by Eq. (5-3). In general, this FRF matrix, which is what is
actually employed in FRF analysis, is rectangular and can be regarded as an (r x e) matrix where r, the
number of rows, is the number of actual response DOFs and e, the number of columns, is the number of
actual excitation DOFs.
FBA Concept
It is possible to obtain the FRFs of an assembly of components from the FRFs of the individual
components comprising the assembly. This process is referred to as FRF-based assembly (FBA).
The FRFs of an assembly of components is computed in MD Nastran by coupling the FRFs of its
constituent components and enforcing conditions of displacement compatibility and force equilibrium at
the connected DOFs. The final result of the FBA process is a set of FRF matrices for the assembly, with
each matrix corresponding to a forcing frequency. The number of rows of each such matrix will be equal
to the total number of response DOFs of the assembly, and the number of columns will be equal to the
total number of excitation DOFs of the assembly.
Using the FRFs resulting from the FBA process, the program recovers the FRFs for the individual FRF
components comprising the FRF assembly. These results will clearly reveal how excitations on one
component affect the responses at other components of the assembly.
Using the FRF/FBA Feature
The FRF/FBA capability is available in both SOL 108 (direct frequency response) and SOL 111 (modal
frequency response). The feature involves the use of a new Case Control command (FRF) and three new
Bulk Data entries (FRFCOMP, FRFXIT, and FRFXIT1). Descriptions of these entries are given in the
MD Nastran Quick Reference Guide.
The generation of FRFs for a component, and their use in a subsequent FBA process using the FRF Case
Control command, involves running a standard SOL 108 or SOL 111 job with the following additional
data:
1. The DOFs where unit loads are to be applied must be specified either indirectly via the DLOAD
Case Control request and/or directly via the FRFXIT/FRFXIT1 Bulk Data entries. The DLOAD
Case Control request points to appropriate Bulk Data loading entries. All DOFs with nonzero load
values will have unit loads applied to them. The FRFXIT entry permits specification of unit load
for a single DOF with a label. The FRFXIT1 entry permits specification of unit loads at multiple
DOFs.
2. There is no requirement that unit loading data described in (1.) be defined for every component
for which FRFs are generated, since some components in a configuration may not have any loads
applied to them.
U [ ]
MD Nastran R2.1 Release Guide

166
3. If the FRFs of a component are to be employed in a subsequent FBA process, the user must
specify the points where this component will be connected to other components in the FBA
process. This is done via the CONNPTS keyword in the FRF command, which references a set of
these connection points. Also, in this case, the FRFs and other related information for the
component are saved on the medium (database or .op2 file) specified in the FRF command, and
the assembly punch (.asm) file will be generated and saved with a single FRFCOMP Bulk Data
entry for subsequent use in the FBA process.
4. Regardless of whether an FRF component has unit loads explicitly specified for it, as in (1.), or
not, as in (2.), the program will internally apply unit loads automatically at all DOFs for all
connection points comprising the set referenced by the CONNPTS keyword of the FRF
command. This ensures that correct results are obtained from subsequent FBA processes.
5. The specific points for which FRFs are computed in an FRF generation run comprise the
following:
a. All points specified via DISP, VELO, and ACCE requests.
b. All points associated with elements for which STRESS/FORCE requests are specified.
c. All points where unit loads are applied (as per the scheme indicated in (1.)).
d. All points comprising the set referenced by the CONNPTS keyword in the FRF command.
6. A component ID of 0 is assigned to the assembled FRF configuration resulting from the FBA
process.
7. The FBA process generates a connection information table in the .f06 file, indicating the
relationship between the internal point IDs of the assembled FRF configuration (component 0)
and the external point IDs of the associated FRF components.
8. The normal output from the FRF generation and FBA process runs, implied by the default of
XITOUT = USER, gives the results only for unit loads explicitly specified by the user as indicated
in (1.) . If output for the unit loads internally applied automatically by the program as indicated in
(4.) is also desired, then XITOUT = ALL must be specified in the FRF command to obtain such
expanded output.
9. The normal output from an FBA process run, implied by the default of ASMOUT = COMP, gives
the results for the individual FRF components that comprise the assembled FRF configuration. If
output is also desired for the assembled configuration as a separate entity (component 0 as
indicated in (6.)), then ASMOUT = ALL must be specified in the FRF command to obtain such
expanded output. However, in this case, the output for component 0 will be limited to
displacements, velocities, and accelerations, and these will be output in terms of the internal point
IDs mentioned in (7.).
10. It is currently assumed that the FRFs of all FRF components are generated at the same forcing
frequencies, and that these are also the forcing frequencies at which the FBA process is
performed. As a result, the FBA process, as currently implemented, derives these forcing
frequencies from the saved data of the first of the FRF components being assembled, and employs
them for the FBA process. This restriction on forcing frequencies will be removed in a future
MD Nastran release.
167
CHAPTER 5
NVH and Acoustics
The usage of the FRF/FBA feature as described can produce one of four scenarios, as follows. All of
these scenarios are illustrated by job setup examples given with the description of the FRF Case Control
command in the MD Nastran Quick Reference Guide.
1. Generate the FRFs for a one-shot configuration with no subsequent FBA process.
2. Generate the FRFs for a component with the purpose of employing them in a subsequent FBA
process.
3. Compute the FRFs of an assembly of components by the FBA process by using the FRFs of
components that have been previously generated.
4. Generate the FRFs for a component and, in the same execution, compute the FRFs of an assembly
of components via the FBA process by combining the FRFs of this component with those of other
components whose FRFs have been previously generated.
Limitations of the FBA Process Implementation
The current implementation of the FBA process has several limitations, indicated as follows. These
limitations will be removed in future releases of MD Nastran.
The FBA process connects a particular point in an FRF component with points in other FRF
components only if such points are coincident (that is, their basic coordinates are the same).
There is no provision to allow for connection of noncoincident points.
All connections between and among FRF components are considered to be rigid. There is no
provision to allow for nonrigid (that is, spring or bushing) connections.
The FBA process has no provision for user-specified loads.
The FBA process assumes that the FRFs of all FRF components are generated at the same
forcing frequencies, and that these are also the forcing frequencies at which the FBA process is
performed.
MD Nastran R2.1 Release Guide

168
Exterior Acoustics
Introduction
MD Nastran has long been used to analyze interior acoustic problems. With interior acoustic problems,
the acoustic domain considered is bounded. A typical example is the determination of the sound pressure
inside a car.
In exterior acoustic problems, the acoustic domain is unbounded. An exterior acoustics analysis evaluates
the sound pressure in the vicinity of the vibrating body, or far away from the body. A further important
result is the acoustic power radiated from the vibrating structure. A typical example is the determination
of the radiated acoustic power of an engine.
In this version of MD Nastran, the infinite elements from MSC Actran have been integrated to allow the
analysis of exterior acoustic problems. The infinite elements are attached to the boundary of the acoustic
finite element mesh to provide the correct nonreflecting boundary condition.
Apart from standard acoustic results within the finite region, it is possible to compute acoustic results at
points within the infinite elements. These so-called field points may be connected by elements to form a
field point mesh. If a field point mesh is defined, the acoustic power through this field point mesh is also
computed.
Benefits
The integration of the proven and tested infinite elements from MSC Actran into MD Nastran facilitates
the analysis of exterior acoustic problems. It is no longer necessary to transfer data between different
programs, since the analysis can be done completely within MD Nastran.
Input
Two new Bulk Data entries, CACINF3 and CACINF4, are used to define the connectivity of the infinite
elements. The properties of the infinite elements are defined on PACINF Bulk Data entries.
Field points and field point meshes are defined in separate sections of the Bulk Data file. These sections
must follow the main Bulk Data Section.
Output of radiated power from the wetted surface, and intensities on the wetted surface, are controlled
by the Case Control commands ACPOWER and INTENSITY. Output of field point mesh results is
controlled by the Case Control command ACFPMRESULT.
Definition of Infinite Elements
The geometry of an infinite element is defined by the geometry of its base and the location of the pole,
as shown in Figure 5-1. The base of the infinite element is the part that is in contact with the finite domain.
The geometry of the base is defined by its connectivity and the location of the corner grid points. In order
169
CHAPTER 5
NVH and Acoustics
to avoid overlapping of the infinite elements, the surface the grid points are attached to must be convex.
However, it is not necessary that this surface be smooth.
Figure 5-1 Geometry of an infinite element.
Within an infinite element, the acoustic pressure is expanded into a power series of (1/r) where r is the
distance from the pole.The radial interpolation order is the number of terms kept in this series.
The connectivity of the base is defined on the CACINF3 and CACINF4 Bulk Data entries. The
orientation should be such that the normal vector on the base points into the radial direction. However,
MD Nastran will automatically change the orientation if the normal vector does not point away from the
pole. The location of the pole, as well as the radial interpolation order, are defined on PACINF Bulk Data
entries.
The CACINF3, CACINF4, and PACINF Bulk Data entries are described as follows:
MD Nastran R2.1 Release Guide

170
Defines an acoustic conjugate infinite element with triangular base.
Format:
Defines an acoustic conjugate infinite element with quadrilateral base.
Format:
CACINF3
Acoustic Conjugate Infinite Element Base Connection
1 2 3 4 5 6 7 8 9 10
CACINF3 EID PID G1 G2 G3
Field Contents
EID Element identification number. (Integer > 0)
PID Property identification number of a PACINF entry. (Integer > 0)
Gi Grid point identification numbers of element base connection points. (Integer > 0)
CACINF4
Acoustic Conjugate Infinite Element Base Connection
1 2 3 4 5 6 7 8 9 10
CACINF4 EID PID G1 G2 G3 G4
Field Contents
EID Element identification number. (Integer > 0)
PID Property identification number of a PACINF entry. (Integer > 0)
Gi Grid point identification numbers of element base connection points. (Integer > 0)
171
CHAPTER 5
NVH and Acoustics
Defines the properties of acoustic conjugate infinite elements.
Format:
Definition of Field Point Meshes
Acoustic field point meshes are defined in separate sections of the Bulk Data file. These sections follow
the main Bulk Data Section. Each of the sections begins with
BEGIN BULK AFPM = afpmid
or
BEGIN AFPM = afpmid
where afpmid is the acoustic field point mesh identifier (Integer > 0). Acoustic field points are defined
using the standard GRID Bulk Data entry. Because all grid points defined in this section of the Bulk Data
file are acoustic field points, it is not necessary to place a value of -1 into field 7.
The grid points can be connected by any type of elements. However, only CTRIA3 and CQUAD4
elements define a field point mesh that is used to compute normal components of the acoustic intensity
and power through the field point mesh. Valid property identifiers must be specified on the CTRIA3 and
CQUAD4 Bulk Data entries. However, the referenced PSHELL entries need not be defined.
If output to an .op2 file is requested, PARAM,POST must be defined within the section of the acoustic
field point mesh. Output of different field point meshes can be sent to different .op2 files using
PARAM,OUNIT2 within the sections of the acoustic field point meshes.
Example
BEGIN AFPM=100
$
PARAM, POST, -1
$
$ Isolated Field Points along a Line
$
PACINF
Acoustic Conjugate Infinite Element Property
1 2 3 4 5 6 7 8 9 10
PACINF PID MID RIO X1 X2 X3
Field Contents
PID Property identification number of PACINF entry. (Integer > 0)
MID Material identification number of a MAT10 entry. (Integer > 0)
XP, YP, ZP Coordinates of the pole of the infinite elements (in the basic coordinate system).
MD Nastran R2.1 Release Guide

172
GRID 1001 2 5. 0. 0.
GRID 1002 2 6. 0. 0.
GRID 1003 2 7. 0. 0.
GRID 1004 2 8. 0. 0.
GRID 1005 2 9. 0. 0.
GRID 1006 2 10. 0. 0.
CORD2R, 2,, 0., 0., 0., -.5, -.5, .707107
, .5, .5, .707107
$
BEGIN AFPM = 200
$
PARAM, POST, -1
$
GRID, 1,, -1., -1., 2.
GRID, 2,, 0., -1., 2.
GRID, 3,, 1., -1., 2.
GRID, 4,, -1., 0., 2.
GRID, 5,, 0., 0., 2.
GRID, 6,, 1., 0., 2.
GRID, 7,, -1., 1., 2.
GRID, 8,, 0., 1., 2.
GRID, 9,, 1., 1., 2.
$
GRID, 11,, -1., -1., 2.
GRID, 12,, 0., -1., 3.
GRID, 13,, 1., -1., 2.
GRID, 14,, -1., 0., 3.
GRID, 16,, 1., 0., 3.
GRID, 17,, -1., 1., 2.
GRID, 18,, 0., 1., 3.
GRID, 19,, 1., 1., 2.
$
CQUAD4, 1, 1, 1, 2, 5, 4
CQUAD4, 2, 1, 2, 3, 6, 5
CQUAD4, 3, 1, 4, 5, 8, 7
CQUAD4, 4, 1, 5, 6, 9, 8
$
CQUAD4, 5, 1, 12, 16, 18, 14
CTRIA3, 6, 1, 11, 12, 14
CTRIA3, 7, 1, 12, 13, 16
CTRIA3, 8, 1, 16, 19, 18
CTRIA3, 9, 1, 14, 18, 17
$
ENDDATA
173
CHAPTER 5
NVH and Acoustics
Case Control Commands
The following new Case Control commands control postprocessing related to exterior acoustics.
Requests output of the power radiated from the wetted surface.
Format:
ACPOWER
Acoustic Power Output Request
Describer Meaning
SORT1 Output will be presented as tabular listing of panels for each excitation frequency.
SORT2 Output will be presented as tabular listing of excitation frequencies for each panel.
(Default)
PRINT The printer will be the output medium. (Default)
PUNCH The punch file will be the output medium.
PLOT Results are generated but not output.
CSV Results will be written to a .csv file.
unit Unit of the .csv file as used on the ASSIGN statement.
ALL Specifies that radiated power will be processed for the wetted surface and all panels.
n Set identification of a previously defined set of panels. Radiated power will be
processed for the wetted surface and all panels in the referenced set.
NONE Radiated power will not be processed.
ACPOWER
SORT1
SORT2
PRINT,PUNCH
PLOT
CSV unit = [ ] , ,
\ .
|
| |
ALL
n
NONE
)

`


=
MD Nastran R2.1 Release Guide

174
Requests output of acoustic intensity on wetted surface.
Format:
INTENSITY
Acoustic Intensity Output Request
Describer Meaning
SORT1 Output will be presented as tabular listing of grid points for each excitation
frequency. (Default)
SORT2 Output will be presented as tabular listing of excitation frequencies for each grid
point.
PRINT The printer will be the output medium. (Default)
PUNCH The punch file will be the output medium.
PLOT Results will be generated but not output.
ALL Intensities will be computed for all grid points of the wetted surface.
n Set identification of a previously defined set of grid points. Intensities will be
computed for the grid points in this set only.
NONE Acoustic intensities will not be processed.
INTENSITY
SORT1
SORT2
PRINT,PUNCH
PLOT
,
\ .
|
| |
ALL
n
NONE
)

`


=
175
CHAPTER 5
NVH and Acoustics
Requests output of field point mesh results.
Format:
Output
In addition to the standard acoustic results, the following results may be requested:
ACFPMRESULT
Acoustic Field Point Mesh Results Output Request
Describer Meaning
SORT1 Output will be presented as tabular listing of grid points for each excitation
frequency. (Default)
SORT2 Output will be presented as tabular listing of excitation frequencies for each grid
point.
PRINT The printer will be the output medium. (Default)
PUNCH The punch file will be the output medium.
PLOT Results are generated but not output.
VELOCITY Requests output of particle velocities. (Default = NO)
REAL or IMAG Requests rectangular format (real and imaginary) of complex output. Use of either
REAL of IMAG yields the same output.
PHASE Requests polar format (magnitude and phase) of complex output. Phase output is in
degrees.
POWER Requests output of power through field point mesh. (Default = YES)
ALL Results of all field point meshes will be processed.
n Set identification of a previously defined set of field point mesh identifiers. Results
will be processed for the field point meshes in this set only.
NONE Field point mesh results will not be processed.
ACFPMRESULT
SORT1
SORT2
PRINT,PUNCH
PLOT
VELOCITY
YES
NO
)
`

= , ,
\

|
REAL or IMAG
PHASE
POWER
YES
NO )
`

=
.
|
|
,
ALL
n
NONE
)

`


=
MD Nastran R2.1 Release Guide

176
Acoustic energy radiated from the wetted surface or from panels.
Normal component of the acoustic intensity at grid points of the wetted surface.
Acoustic pressure and intensity at the field points.
Acoustic velocity at the field points.
Acoustic energy flowing through the field point mesh.
Field point mesh results can be written to the .f06 file or the .op2 file. The .xdb file is not yet supported.
Guidelines
1. The surface to which the infinite elements are attached must be convex but it need not be smooth.
Infinite elements must be connected to acoustic grid points. Thus, it is necessary to model the
vicinity of the vibrating structure with acoustic finite elements until a convex surface is reached.
2. The required radial interpolation order depends on the directivity of the pressure field. Usually, a
higher order is needed for higher frequencies.
3. It is recommended that the user study the sensitivity of the results with respect to the radial
interpolation order by repeating the analysis with a different radial interpolation order.
4. Infinite elements are supported in direct and modal frequency response analysis (SOL 108 and
SOL 111). Experience shows that a large number of fluid modes is needed to obtain converged
results. Thus, modal reduction of the fluid is not recommended. However, modal reduction can
be applied to the structure.
5. Efficiency can be increased if the iterative solver is used. The Jacobi preconditioner with an
epsilon of 10
-4
is recommended.
Limitations
1. Infinite elements are supported only in frequency response analysis (SOL 108 and SOL 111).
Example
In this example, the sound transmission through an elastic plate embedded in an infinite rigid wall is
analyzed. Figure 5-2 shows the geometry of the plate and of the fluid region meshed with finite elements.
The finite element mesh of the structure, and of one quarter of the fluid, can be seen in Figure 5-3.
177
CHAPTER 5
NVH and Acoustics
Figure 5-2 Geometry.
Figure 5-3 Structure and quarter of fluid mesh.
MD Nastran R2.1 Release Guide

178
Infinite elements are attached to the top and the four lateral faces, but not to the bottom. The pole of all
infinite elements is at the center of the plate. The finite element mesh of the structure, together with one
quarter of the infinite element mesh, is shown in Figure 5-4.
The field point mesh is a cube which completely encloses the fluid mesh. Thus, the power through this
field point mesh equals the total power radiated from the plate. Figure 5-5 shows the fluid mesh and one
quarter of the field point mesh.
Figure 5-4 Structure and quarter of infinite element mesh.
179
CHAPTER 5
NVH and Acoustics
Figure 5-5 Fluid mesh and quarter of field point mesh.
The excitation is a uniform pressure applied to the plate. The acoustic pressure in the finite element mesh,
and acoustic results at some discrete field points located along the z-axis and at the field points of the
field point mesh, are computed. Acoustic results at the field points include the acoustic pressure and the
acoustic intensities.
Input File
SOL 108
CEND
TITLE = Vibrating Plate Example
SUBTITLE = Direct Frequency Response
$
ECHO = SORT(EXCEPT, GRID, CHEXA, CQUAD4, CACINF4)
$
DLOAD = 10
FREQ = 20
SMETHOD = 30
SPC = 1
$
DISP(PLOT) = ALL
ACFPMRESULT(PHAS) = ALL
$
BEGIN BULK
$
PARAM, POST, -1
$
MD Nastran R2.1 Release Guide

180
ACMODL, IDENT
$
RLOAD1, 10, 200,,, 300
PLOAD2, 200, 1., 1, THRU, 150
TABLED1, 300
, 0., 1., 1000., 1., ENDT
$
FREQ, 20, 10.
$
ITER, 30
PRECOND = J,
ITSEPS = 1.0E-4
$
$ Fluid
$
GRDSET,,,,,, -1
INCLUDE 'fluid1.bdf'
$
$ Structure
$
INCLUDE 'structure.bdf'
$
BEGIN AFPM=100
$
$ Isolated Field Points
GRID, 10001,, 0., 0., 3.5
GRID, 10002,, 0., 0., 5.
GRID, 10003,, 0., 0., 10.
$
BEGIN AFPM=200
$
$ Field Point Mesh
INCLUDE 'fpm.bdf'
$
ENDDATA
Excerpt of fluid1.bdf
$ Exterior Acoustics - Vibrating Plate Example
$ Fluid Model of Domain 1: Finite and Infinite Elements
$
$ -------------------------------------------------------------------
$
$ PID MID RIO XP YP ZP
PACINF 10 20 10 0. 0. 0.
PSOLID 2 20 PFLUID
MAT10 20 1.21 340.
$
CHEXA 1001 2 1001 1002 1018 1017 1177 1178
1194 1193
CHEXA 1002 2 1002 1003 1019 1018 1178 1179
1195 1194

181
CHAPTER 5
NVH and Acoustics
$ Infinite Elements
$
CACINF4 13901 10 3641 3642 3658 3657
CACINF4 13902 10 3642 3643 3659 3658
CACINF4 13903 10 3643 3644 3660 3659
CACINF4 13904 10 3644 3645 3661 3660
CACINF4 13905 10 3645 3646 3662 3661
CACINF4 13906 10 3646 3647 3663 3662

Results
Figure 5-6 shows the acoustic pressure in the finite element mesh, and Figure 5-7 shows some results
from the .f06 file.
Figure 5-6 Acoustic pressure in the finite element mesh.
MD Nastran R2.1 Release Guide

182
Figure 5-7 Results from .f06 file.
FREQUENCY = 1.000000E+01 ACOUSTIC FIELD POINT MESH = 100
A C O U S T I C F I E L D P O I N T M E S H R E S U L T S

ACOUSTIC PRESSURE INTENSITY COMP. INTENSITY INTENSITY INTENSITY
POINT ID. MAGNITUDE PHASE NORMAL TO FPM X Y Z
10001 5.867592E-03 3.224626E+02 0.0 3.901852E-16 1.102822E-09 4.101243E-08
10002 4.148942E-03 3.068220E+02 0.0 1.168128E-16 3.946748E-10 2.071128E-08
10003 2.089879E-03 2.542063E+02 0.0 1.290767E-17 5.147245E-11 5.294778E-09
1 VIBRATING PLATE EXAMPLE DECEMBER 6, 2005 MSC.NASTRAN 12/ 5/05
PAGE 16
DIRECT FREQUENCY RESPONSE
0
*** USER INFORMATION MESSAGE 3119 (AFPINI)
DATA RECOVERY OF ACOUSTIC FIELD POINT MESH 200 INITIATED
1 VIBRATING PLATE EXAMPLE DECEMBER 6, 2005 MSC.NASTRAN 12/ 5/05
PAGE 17
DIRECT FREQUENCY RESPONSE
0
FREQUENCY = 1.000000E+01 ACOUSTIC FIELD POINT MESH
= 200
A C O U S T I C F I E L D P O I N T M E S H R E S U L T S

ACOUSTIC PRESSURE INTENSITY COMP. INTENSITY INTENSITY INTENSITY
POINT ID. MAGNITUDE PHASE NORMAL TO FPM X Y Z
46567 2.959861E-03 2.856447E+02 7.442431E-09 -7.647572E-09 7.442431E-09 1.545178E-10
46568 2.930327E-03 2.848879E+02 7.258169E-09 -7.408595E-09 7.258169E-09 1.340973E-09
46569 2.846820E-03 2.826632E+02 6.691689E-09 -6.762688E-09 6.691689E-09 2.608988E-09
46570 2.722373E-03 2.790949E+02 5.876647E-09 -5.877888E-09 5.876647E-09 3.492858E-09
46571 2.572942E-03 2.743544E+02 4.970530E-09 -4.929979E-09 4.970530E-09 3.970974E-09
46572 2.412938E-03 2.686276E+02 4.189022E-09 -4.042182E-09 4.189022E-09 4.023463E-09
46573 3.274509E-03 2.928127E+02 1.022698E-08 -8.121219E-09 1.022698E-08 2.107736E-10
46574 3.234508E-03 2.919741E+02 9.893179E-09 -7.820792E-09 9.893179E-09 1.809509E-09
183
CHAPTER 5
NVH and Acoustics
Rigid Porous Absorber
Introduction
A new capability to model basic rigid skeleton porous absorber characteristics in acoustic response
analysis is now available. The capability allows some types of absorbent material to be modeled, such
as vehicle seat structures, or lining materials which exhibit stiff carcasses. The absorber material is
considered using an equivalent fluid analogy and is this modeled in the same manner as a typical fluid,
using solid CHEXA, CPENTA, or CTETRA elements, the grid points of which have their CD field set
to -1. The porous absorber elements reference a PSOLID property entry with field 8 set to PFLUID in
the usual way, with field 3 of the PSOLID entry referencing a MAT10 entry which has been modified
with an additional field 7 dedicated to porous absorber materials.
Porous Materials
If a material is not completely solid, but contains voids or air pockets, then it is said to be porous. There
are many materials which exhibit porosity, the term given to the degree of openness of the material,
including materials generally considered solid like brick or stone. If the voids in the substance are
large enough, they may form an interconnected maze of passage ways allowing air (or any other fluid)
to pass through the material. However, depending on the degree of convolution in the passage ways
(known as tortuosity), the air will encounter some resistance as it passes through the substance, requiring
pressure to be exerted to push the air against the resistance. Sound waves striking the material do not
cause air to flow through the material, but they do cause local perturbations that exert pressure and cause
the air to move in the vicinity of the material; the oscillating movement of air caused by the sound waves
encounters resistance (called impedance) which uses up some of the sound energy and damps the level
of sound. This energy is eventually dissipated as heat.
If the porous material is enclosed in a frame which is considered to be rigid (as will be the case, for
example, for a porous medium which has a high skeleton density, very large elastic modulus, or weak
fluid-structure coupling), the porous material can be considered to be an equivalent fluid with both
density and bulk modulus being complex frequency-dependent parameters. It is possible to obtain values
for these parameters by empirical methods as introduced, for example, by Delany and Bazley
1
; these
methods have been widely used to describe sound propagation in fibrous materials.
In the equivalent fluid approach, the equation of motion is
where is the equivalent density, the equivalent bulk modulus, the complex pressure amplitude,
and the circular excitation frequency, for which it can be shown that
1

e
-----
2
P

2
B
e
------ P + 0 =

e
B
e
P

e
-----
1

--- 1 iGE + ( ) =
MD Nastran R2.1 Release Guide

184
and if it is assumed that the parameters are not frequency-dependent, a reasonable assumption for the
study of frequencies in a narrow band, we obtain
Here, , , and are the values of RHO, BULK, and GE, respectively, of the MAT10 entry for the
porous absorber material. A new field, field 7, has been added to the MAT10 entry to allow the value of
, the normalized admittance coefficient
2
, to be entered.
Inputs
Defines material properties for fluid elements in coupled fluid-structural analysis.
Format:
The following expressions may be used to calculate the equivalent fluid property values to be entered on
the MAT10 Bulk Data entry starting from the complex density and complex bulk modulus describing the
rigid porous absorber. If a complex density and complex speed of sound are determined for the porous
material, the complex bulk modulus must first be calculated.
MAT10 density
where is the complex density (real part) and is the complex density (imaginary part).
MAT10 damping coefficient
MAT10 bulk modulus
MAT10
Fluid Material Property Definition
1 2 3 4 5 6 7 8 9 10
MAT10 MID BULK RHO C GE ALPHA
1
B
e
------
1
B
--- = i
A

----
1
B
--- 1
i

-----
\ .
| |
=
B GE


r
2

i
2
+

r
------------------- =

r

i
GE
GE

r
----- =
B
B
B
r
2
B
i
2
+
B
r
------------------- - =
185
CHAPTER 5
NVH and Acoustics
where is the complex bulk modulus (real part) and is the complex bulk modulus (imaginary part).
MAT10 normalized admittance coefficient
Discussion
The implementation implies that if the complex density and bulk modulus are constant, the normalized
admittance coefficient is a function of frequency. However, a frequency-dependent is not supported,
so a reference frequency must be selected. Typically, this frequency will be either in the midrange of the
desired frequency range to be studied, or will correspond to the frequency at a response peak of interest.
As frequencies progressively farther away from the reference frequency are considered, the response
becomes increasingly subject to the limitations of the frequency-independent formulation; the extent will
depend somewhat on the nature of the problem, and it may be necessary to study discrete frequency
bands in order to mitigate against this effect.
The use of a nonzero value in field 7 of the MAT10 entry causes the generation of a damping matrix,
because the normalized admittance coefficient is multiplied by the imaginary operator i. Consequently,
the use of modal methods on the fluid are not appropriate and frequency response analysis must be
carried out using the direct method, at least for the fluid.
References
1. M.E. Delany and E.N. Bazley, Acoustical Characteristics of Fibrous Absorbent Materials,
National Physics Laboratory, Aerodynamics Division, NPL Aero Report Ac 37, March 1969.
2. J. Wandinger, Possible Implementations of Porous Absorbers in Nastran, MSC internal memo,
April 2006.
Example
Consider unbounded fluid (air) and porous absorber medium domains as shown in Figure 5-8. An
acoustic source is placed at the location indicated, and the acoustic response (pressure) at the center of
the fluid is monitored.
B
r
B
i

B
i
B
r
----- =

MD Nastran R2.1 Release Guide



186
Figure 5-8 Porous medium.
The following properties were determined using experimental methods.
Air
Porous Absorber
Density 1.225 + i0
Speed of sound 340.0 + i3.4
Density 3.8663 + i14.2204
Speed of sound 92.7076 + i70.2854
Porous medium
Response
Acoustic source
Air
kg m
3

m s
kg m
3

m s
187
CHAPTER 5
NVH and Acoustics
From which the following equivalent bulk moduli were obtained
Air
Porous Absorber
A frequency of 250 Hz was selected to calculate the values of alpha for air and the porous absorber.
Using the previously described equations (see Inputs section), the following data is entered on the
MAT10 Bulk Data entries:
MAT10 for air
MAT10 for the porous absorber material
Notice that the values of bulk modulus, GE damping coefficient, and alpha are all negative; this is a
normal characteristic of the implementation.
The response at the center of the air domain is calculated using the new capability, and the results are
compared with the same model run in Actran. 20-noded CHEXA (Figure 5-9) and 8-noded CHEXA
(Figure 5-10) elements are compared.
Bulk modulus 141595.8 + i2832.2 Pa
Bulk modulus -171190.0 + i102356.3 Pa
MID BULK RHO C GE ALPHA
MAT10 1 141652.5 1.225 0.0 31.41907
MID BULK RHO C GE ALPHA
MAT10 2 -232389. 56.16948 -3.67804 -939.196
MD Nastran R2.1 Release Guide

188

Figure 5-9 20-noded CHEXA element comparison.
Figure 5-10 8-noded CHEXA element comparison.
HEXA-20
60
80
100
120
140
160
180
0 100 200 300 400 500
Fr e que nc y ( Hz )
Actran Nastran
P
r
e
s
s
u
r
e

(
r
e
a
l
)
HEXA-8
60
80
100
120
140
160
180
0 100 200 300 400 500
Fr e que nc y ( Hz )
Actran Nastran
P
r
e
s
s
u
r
e

(
r
e
a
l
)
189
CHAPTER 5
NVH and Acoustics
The results compare very favorably with those from Actran. Notice the gradual departure from the
expected Actran response for frequencies progressively farther away from the reference frequency of
250 Hz, as the response becomes increasingly subject to the limitations of the frequency-independent
formulation.
MD Nastran R2.1 Release Guide

190
Ch. 6: Numerical Enhancements MD Nastran R2.1 Release Guide

6
Numerical Enhancements
!
New SPARSESOLVER Executive Control Statement
!
Improved Performance with New Sparse Solvers
!
Improved Memory Usage in Lanczos
!
New MAXRATIO Information Output (Prerelease)
!
New SPARSESOLVER MDTSTATS Information Output (Prerelease)
!
Performance Improvements and Expanded Capabilities for
Automated Component Modal Synthesis (ACMS)
!
Relaxed Restrictions for CASI Solver Usage
!
System-Dependent Performance Improvements
!
Improved Selection of Reordering Methods
!
Iterative Solution Options for Nonlinear Analysis
MD Nastran R2.1 Release Guide

192
New SPARSESOLVER Executive Control Statement
Introduction
A new interface, the SPARSESOLVER Executive Control statement, is now available to better control
the options and processes associated with sparse matrix solution methods. The interface specifies the
solver, matrix reordering methods, and compression techniques to be used. Furthermore, improved
control of matrix diagonal term ratio output is provided.
Benefits
SPARSESOLVER provides more efficient control over the process than the existing method selection
that usually involves defining values for some SYSTEM cells via the NASTRAN statement. In addition,
a new output data option is available for matrix diagonal term ratios in the form of a simple bar chart that
provides a more comprehensive view of the ratio data.
Method and Theory
No new theory is involved. The method simply involves the specification of solver options to be used by
various DMAP modules during the solution process. The specified options are checked against feature
availability tables to ensure that they do not conflict with any limitations posed by feature combinations.
For example, the specification of a particular ordering method may not be available for a particular solver
specification. The options are available only for the DCMP, DECOMP, FRRD1, READ, SOLVE, and
TRD1 modules.
Inputs
The sparse solution options are controlled by keywords specified on the SPARSESOLVER Executive
Control statement. See the MD Nastran Quick Reference Guide for a complete description of this
statement.
Outputs
There are no new outputs associated with this feature other than informational and diagnostic messages.
Guidelines and Limitations
The ability to specify particular sparse matrix solution options is sometimes useful in determining
whether one method is more effective than another in obtaining the solution. Other features can also be
useful, such as in obtaining diagnostic data output. For example, one might be interested in reviewing
matrix diagonal term ratios. In general, high ratios indicate a loss of accuracy. The feature can be used
by taking all of the program defaults for the various control variables. These defaults produce both the
193
CHAPTER 6
Numerical Enhancements
table and bar outputs. The table is limited to 25 ratios that exceed 1.0E+05. The bar chart uses powers of
ten for segment widths. This can be done by adding
SPARSESOLVER DCMP ( MDTRATIO )
to the Executive Control Section. The use of this new feature is currently limited to sparse symmetric
matrix operations in the DCMP module.
Several sparse matrix factorization methods are available. The following table summarizes the
advantages and disadvantages of the various factorization methods.
Similarly, reordering methods can be specified, but this should be done only after consideration of the
potential effects. The following table summarizes the advantages and disadvantages of the various
reordering methods.
Method Advantages Limitation
MSCLDL Small memory requirements; handles
indefinite matrices.
None.
MSCLU Small memory requirements. None.
TAUCSCHL Generally superior performance
compared to MSCLDL.
High memory requirements; real positive
definite matrices only.
UMFLU Generally superior performance
compared to MSCLU.
High memory requirements.
Method Advantage Limitation
AMF Similar to BEND, but with smaller
memory requirements.
Produces less optimal reordering
compared to BEND.
BEND Optimal for small matrices, and for large
matrices from models dominated by 1-D
and 2-D elements.
None.
MD Very low memory requirements. Produces suboptimal reordering.
METIS Good for large matrices dominated by 3-D
elements.
Large memory requirements.
MMD Low memory requirements. Produces suboptimal reordering.
MD Nastran R2.1 Release Guide

194
Similarly, compression methods can also be specified, but this should be done only after consideration
of the potential effects. The following table summarizes the advantages and disadvantages of the various
compression methods.
Method Advantage Limitation
GRID Utilizes USET and SIL information from the
MD Nastran Database.
Relies on USET and SIL.
SUPER Does not rely on USET and SIL tables; can
produce better orderings for models
dominated by 3-D elements.
None.
GRDSUPER Can produce better orderings for models
dominated by 3-D elements.
Requires USET and SIL tables.
195
CHAPTER 6
Numerical Enhancements
Improved Performance with New Sparse Solvers
Introduction
MSC.Software conducts surveys of solver technologies from industry and academia. This has lead to the
integration of two solvers developed outside of MSC.Software: TAUCS and UMFPACK. In addition,
The Lanczos eigensolver has been enhanced to take better advantage of available memory.
The TAUCS Sparse Solver
A new sparse Cholesky solver is available, derived from the TAUCS package of linear solvers.
User Interface
The TAUCS solver can be selected for the DCMP, DECOMP, and SOLVE modules through the
SPARSESOLVER Executive Control statement:
SPARSESOLVE DCMP (FACTMETH=TAUCSCHL)
It may also be selected by setting system cell 166 to 8192.
Use Cases
The TAUCS solver keeps the matrix data in memory, and may show improved performance over the
default solver for models which are dominated by three-dimensional elements.
The four models described as follows demonstrate the potential performance enhancement with the
TAUCS solver.
Each model was run on one node with two dual core 2.4 GHz Opteron processors, 8 GB of memory and
56 GB of scratch space. Each job was run with mem=7 GB, and the CPU time of the DCMP module is
displayed in the following chart.
Model Name DOFs Elements
zzz3 537876 108675 CTETRAs
vl0sst 408999 710768 CTETRAs
xl0rst 739815 113217 CHEXAs
Solidcyl 604800 176400 CHEXAs
MD Nastran R2.1 Release Guide

196

Limitations
The new solver only works for real, symmetric, positive-definite matrices. Modeling techniques which
lead to indefinite matrices, such as Lagrange multipliers, are not supported by the new solver. If the
TAUCS solver detects an indefinite matrix, has insufficient memory to perform the factorization, or
encounters any other error, the out-of-core LDL
T
solver is used.
The TAUCS solver is not recommended for multiple superelement analysis.
The TAUCS code is used by MSC.Software under the following license agreement. Note that in the
agreement, this program and this software refers only to the TAUCS code, available at
http://www.tau.ac.il/~stoledo/taucs.
TAUCS version 2.0, November 29, 2001. Copyright 2001 by Sivan Toledo, Tel-Aviv University,
stoledo@tau.ac.il. All Rights Reserved.
Used by permission.
THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED OR
IMPLIED. ANY USE IS AT YOUR OWN RISK.
Permission is hereby granted to use or copy this program, provided that the Copyright, this License, and
the Availability of the original version is retained on all copies. User documentation of any code that uses
this code or any derivative code must site the Copyright, this License, the Availability note, and Used
by permission. If this code or any derivative code is accessible from within MATLAB, then typing help
taucs must cite the Copyright, and the type taucs must also cite this License and the Availability note.
Permission to modify the code and to distribute modified code is granted, provided that the Copyright,
this License, and the Availability note are retained, and a notice that the code was modified is included.
This software is provided free of charge.
zzz3 vl0sst xl0rst solidcyl
0
50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
New Cholesky Solver Performance
MSCLDL
TAUCSCHL
Model
D
C
M
P

t
i
m
e

(
c
p
u

s
e
c
)
197
CHAPTER 6
Numerical Enhancements
The UMFPACK Sparse Solver
UMFPACK is a set of routines for solving unsymmetric sparse linear systems using an unsymmetric
multifrontal method. The UMFPACK factorization has been implemented in MD Nastran as a licensed
software product from the University of Florida. It operates on both real and complex matrices.
UMFPACK has been implemented for solving unsymmetric frequency response problems. In addition,
it is implemented in the SOLVE and DECOMP modules. It is activated by an explicit user request via
the SPARSESOLVER Executive Control statement. For example, the statement
SPARSESOLVER FRRD1 (FACTMETH=UMFLU)
specifies that UMFPACK is to be run in the FRRD1 module, which is used for frequency response
analysis problems. UMFPACK may also be selected by setting system cell 209 to 16.
The UMFPACK solver has shown to be very useful for problems in exterior acoustics, which generates
unsymmetric matrices for frequency response.
Examples:
Two simple example models (Models 1 and 2 shown in Figure 6-1 (a) and (b)), in exterior acoustics are
shown to demonstrate the usefulness of UMFPACK for unsymmetric applications. Model 2 is meshed
twice: once with a 64 x 64 plate structure, and once with a 128 x 128 plate structure.
Figure 6-1 UFMPACK example problems.
The model properties are described in the following table. Each model was run with SOL 111. There are
200 forcing frequencies in each run.
a. Model 1. b. Model 2.
MD Nastran R2.1 Release Guide

198
All jobs were run on a 2.6 GHz Intel Woodcrest processor system with 16 GB main memory, running
under Linux.
In order to use UMFPACK in a user-written DMAP sequence via the DECOMP module, permutations
must be applied by the DMAP writer using MATMOD option 45. In the following DMAP sequence
fragment, the matrix equation [A][x]=[b] is solved.
DECOMP A/L,U,ROWPERM,COLPERM/4 $ Option 4 for UMFPACK
MATMOD B,ROWPERM,,,,/BPERM,/45/1 $ Permute the RHS
FBS L,U,BPERM/XPERM $ Solve for permuted X
MATMOD XPERM,CPERM,,,,/X,/45/-1 $ Un-permute X
Element Distribution
Elements Model 1 Model 2a Model 2b
Air 6566 13440 13440
Infinite 768 1664 1664
Air_low n/a 8192 8192
Plate 4096 4096 16384
Total 11520 27392 39680
Note: For the DECOMP module to call UMFPACK, the first parameter (KSYM) must be set to 4.
199
CHAPTER 6
Numerical Enhancements
UMFPACK Memory Guidelines (Updated for MD Nastran R2.1)
The UMFPACK solver operates entirely in memory. Therefore, memory requirements for UMFPACK
may be considerably higher than for the default MD Nastran sparse direct unsymmetric solver.
Previously, in MD Nastran R2, UMFPACK operated entirely outside the control of MD Nastran memory
management. For MD Nastran R2, it was not possible to allocate additional memory for UMFPACK via
the nastran command line option mem=. Rather, it was the user's responsibility to ensure that
sufficient unsubscribed memory was available from the operating system for UMFPACK to complete
successfully.
This memory limitation has been removed for MD Nastran R2.1, so that all UMFPACK memory
operations are now controlled by Nastran memory management. This offers more control to the user via
the mem= command line option.
User Information Messages (UIM) and User Fatal Messages (UFM) related to UMFPACK memory
management are presented as follows.
User Information Message 4220
The first phase of a UMFPACK factorization is to store the input matrix entirely in memory. At the
successful completion of this phase, UIM 4220 is printed in the .f04 file.
The INITIAL MEMORY REQUIREMENT is the amount of memory required to store the input matrix.
The TOTAL MEMORY AVAILABLE is the current open core length of available memory. MEMORY
AVAILABLE FOR UMFPACK is the difference between these two amounts.
User Information Message 4226
Upon successful completion of a UMFPACK factorization, UIM 4226 is printed in the .f04 file.
The MINIMUM MEMORY REQUIREMENT is the amount of open core memory required for
UMFPACK to complete successfully. Note that it is necessary to add executive memory areas (SMEM,
BUFFPOOL, etc.) to this amount in order to arrive at the total amount of memory which should be
specified via the mem= command line option. Unless the user has increased the defaults for these
executive memory areas, these amounts should be quite small in size.
*** USER INFORMATION MESSAGE 4220 (UMFDRV)
PARAMETERS FOR SPARSE UNSYMMETRIC DECOMPOSITION OF DATA BLOCK SCRATCH (TYPE=CDP ) FOLLOW
MATRIX SIZE = 29556 ROWS NUMBER OF NONZEROS = 1811711 TERMS
INITIAL MEMORY REQUIREMENT = 9147229 WORDS ( 35 MB)
MEMORY AVAILABLE FOR UMFPACK = 247868331 WORDS ( 946 MB)
TOTAL MEMORY AVAILABLE = 257015559 WORDS ( 981 MB)
*** USER INFORMATION MESSAGE 4226 (UMFDRV)
ACTUAL MEMORY AND DISK SPACE REQUIREMENTS FOR SPARSE UNSYMMETRIC DECOMPOSITION
NUMBER OF NONZERO TERMS IN UPPER FACTOR = 16864850
NUMBER OF NONZERO TERMS IN LOWER FACTOR = 16864850
MINIMUM MEMORY REQUIREMENT = 195788859 WORDS ( 747 MB)
MD Nastran R2.1 Release Guide

200
User Fatal Message 3008
Notification of insufficient memory for UMFPACK is printed to the .f06 output file with a customized
version of the more general UFM 3008.
The UMFPACK PEAK MEMORY ESTIMATE is an estimate of the upper bound on memory
requirements. In practice, this requirement is very rarely the actual requirement. An attempt at a more
realistic requirement is printed as UMFPACK REALISTIC MEMORY EST. This figure is roughly 20%
of the upper bound. It is not unusual for actual memory requirements to be less than the realistic estimate.
Known Limitations
UMFPACK is an entirely memory-resident factorization method. Therefore, memory requirements for
problem sizes exceeding 100,000 DOF (complex matrices) are likely to exceeded 8 GB. If memory
requirements exceed 8 GB, the ILP-64 version of MD Nastran, evoked via the mode=i8 command line
option, is required.
Currently, exterior acoustics is the main target application for UMFPACK. Observed memory
requirements indicate that most problems of this nature will solve comfortably in 8 GB of memory.
Attention should be paid to the radial interpolation order input quantity on the PACINF Bulk Data entry
(RIO, field 4). Increasing this value will sharply increase the problem size, and thus memory required
for a UMFPACK solution.
*** USER FATAL MESSAGE 3008 (UMFDRV)
INSUFFICIENT MEMORY FOR UMFPACK FACTORIZATION
UMFPACK PEAK MEMORY ESTIMATE = 6896. MB
UMFPACK REALISTIC MEMORY EST. = 1380 MB
MEMORY AVAILABLE = 781 MB
201
CHAPTER 6
Numerical Enhancements
Improved Memory Usage in Lanczos
The Lanczos method has been improved to better take advantage of available memory.
This prerelease feature may be selected by setting system cell 146 to -1.
To illustrate improvements in performance, the following models were run on an SGI Altix with four 1.4
GHz itanium2 processors, 20 GB of RAM, and 470 GB of disk space. Each model was run with
mem=4200 MB.
Model Name DOFs Elements
xl0u 595303 55925 CHEXAs, 18483 CPENTAs
xl0h 335680 54509 CQUAD4s, 15523 CTRIA3s
xl0j 590474 94979 CQUAD4s, 13406 CTRIA3s
xl0m 529199 96185 CQUAD4s
xl0u xl0j xl0h xl0m
0
50
100
150
200
250
300
350
400
450
500
550
600
650
700
Improved Lanczos Performance
2006R1
2006R2
Model
R
E
A
D

t
i
m
e

(
c
p
u

s
e
c
)
MD Nastran R2.1 Release Guide

202
New MAXRATIO Information Output (Prerelease)
Introduction
A new interface is now available for analysts to better control the generation of matrix diagonal term ratio
statistics produced by the sparse symmetric matrix decomposition process in the DCMP module. The
matrix diagonal term ratio statistics are sometimes useful in determining the quality of the matrix
decomposition process. In general, for linear static analysis, high or negative ratios indicate a loss of
accuracy and could be indicative of a modeling error.
Benefits
The new interface provides analysts with more control over the process than the existing method of
supplying a value for the MAXRATIO DMAP parameter. In addition, a new output data option is
available in the form of a simple bar chart that provides a more comprehensive view of the ratio data.
Method and Theory
No new theory is involved. The method simply involves the computation of a ratio defined as the original
matrix diagonal term divided by the decomposed matrix diagonal term. These ratios are placed in a table
together with the external identifier associated with the row/column of the term. This table is then
processed according to the options requested by the user.
Inputs
The matrix diagonal term ratio output options are controlled by keywords specified on the
SPARSESOLVER Executive Control statement. See New SPARSESOLVER Executive Control
Statement on page 192 for a complete description of this statement.
Outputs
The matrix diagonal term ratios can be presented in two different views. The first view is the table view,
in which each ratio is listed together with the external identifier of the row/column of the matrix, as well
as the original input matrix diagonal term. This format is virtually identical to that produced by the
previous version when any ratio exceeds the value of the MAXRATIO input parameter.
The second view of the ratios is statistical in nature. It is similar to a bar chart. A series of bar segments
are generated. There are two options for specifying the segment widths of the bars. The default option
uses powers of 10 as the widths (e.g., 10.0 to 100.0, and 100.0 to 1000.0). The second option allows the
user to specify how many segments are desired. The program will compute the segment width using the
maximum and minimum ratios. For each bar in the chart, the total number of terms in the range is
tabulated together with a visual indication of the percentage number of terms in that particular bar.
203
CHAPTER 6
Numerical Enhancements
Note that when negative matrix diagonal term ratios are detected, they will always be output if the
TABLE option is specified.
These new views of the ratios do not replace any existing diagnostics generated by the DCMP module if
a problem is detected. Under these conditions, output from the table view may duplicate previous output
generated by DCMP module error processing.
Guidelines and Limitations
The matrix diagonal term ratio statistics are sometimes useful in determining the quality of the matrix
decomposition process. In general, high ratios indicate a loss of accuracy. The feature can be used by
taking all of the program defaults for the various control variables. These defaults produce both the table
and bar outputs. The table is limited to 25 ratios that exceed 1.0E+05. The bar chart uses powers of ten
for segment widths. This can be done by adding
SPARSESOLVER DCMP (MDTRATIO)
to the Executive Control Section of the input data file.
The use of this new feature is currently limited to sparse symmetric matrix operations in the DCMP
module.
If there are scalar-type points present in the problem, the degrees of freedom associated with these points
will be grouped into the results for the translational degrees of freedom output.
Demonstration Example
A simple example is presented that demonstrates the use of some of the new features available for output
of the matrix diagonal term ratios. The SPARSESOLVER Executive Control statement is used to specify
the desired features. The example is for demonstration purposes only, and is not representative any
particular modeling situation. The model data consists of a simple plate structure subject to an end load.
Example Input Data
$
$ Example problem to demonstrate matrix diagonal term ratios
$
id test,case
sol 101
SPARSESOLVER DCMP (MDTRATIO)
cend
spc=100
load=1000
disp=all
begin bulk
grdset,,,,,,,6
cquad4,101,101,1,2,52,51
cquad4,102,101,2,3,53,52
cquad4,103,101,3,4,54,53
cquad4,104,101,4,5,55,54
MD Nastran R2.1 Release Guide

204
cquad4,105,101,5,6,56,55
cquad4,106,101,6,7,57,56
cquad4,107,101,7,8,58,57
cquad4,108,101,8,9,59,58
cquad4,109,101,9,10,60,59
cquadr,1101,101,1,2,52,51
cquadr,1102,101,2,3,53,52
cquadr,1103,101,3,4,54,53
cquadr,1104,101,4,5,55,54
cquadr,1105,101,5,6,56,55
cquadr,1106,101,6,7,57,56
cquadr,1107,101,7,8,58,57
cquadr,1108,101,8,9,59,58
cquadr,1109,101,9,10,60,59
grid, 1,, 0.0,0.0,0.0
grid, 2,, 1.0,0.0,0.0
grid, 3,, 2.0,0.0,0.0
grid, 4,, 3.0,0.0,0.0
grid, 5,, 4.0,0.0,0.0
grid, 6,, 5.0,0.0,0.0
grid, 7,, 6.0,0.0,0.0
grid, 8,, 7.0,0.0,0.0
grid, 9,, 8.0,0.0,0.0
grid,10,, 9.0,0.0,0.0
grid,51,, 0.0,1.0,0.0
grid,52,, 2.4,1.0,0.0
grid,53,, 3.5,1.0,0.0
grid,54,, 4.6,1.0,0.0
grid,55,, 5.7,1.0,0.0
grid,56,, 6.8,1.0,0.0
grid,57,, 7.9,1.0,0.0
grid,58,, 9.0,1.0,0.0
grid,59,,10.1,1.0,0.0
grid,60,,11.2,1.0,0.0
$
ctria3,201,101,101,102,151
ctria3,202,101,102,152,151
ctria3,203,101,102,103,152
ctria3,204,101,103,153,152
ctria3,205,101,103,104,153
ctria3,206,101,104,154,153
ctria3,207,101,104,105,154
ctria3,208,101,105,155,154
ctriar,1201,101,101,102,151
ctriar,1202,101,102,152,151
ctriar,1203,101,102,103,152
ctriar,1204,101,103,153,152
ctriar,1205,101,103,104,153
ctriar,1206,101,104,154,153
ctriar,1207,101,104,105,154
ctriar,1208,101,105,155,154
grid,101,, 0.0,0.0,0.0
grid,102,, 1.0,0.0,0.0
grid,103,, 2.0,0.0,0.0
grid,104,, 3.0,0.0,0.0
grid,105,, 4.0,0.0,0.0
grid,151,, 0.0,1.0,0.0
grid,152,, 3.4,1.0,0.0
grid,153,, 4.5,1.0,0.0
grid,154,, 5.6,1.0,0.0
205
CHAPTER 6
Numerical Enhancements
grid,155,, 6.7,1.0,0.0
$
pshell,101,1,0.05,1
mat1,1,10.+6,,0.33
spc1,100,123,1,101
spc1,100,3,5,55,105,155
spc1,100,1,55,155
spc1,100,2,1,101
force,1000,10,,1000.0,1.0,0.0,0.0
force,1000,60,,1000.0,1.0,0.0,0.0
force,1000,105,,1000.0,1.0,0.0,0.0
force,1000,155,,1000.0,1.0,0.0,0.0
enddata
Example Output
The output generated by the example is shown as follows. Notice that there are two separate sections of
output: one for translational degrees of freedom, and one for rotational degrees of freedom. Within each
section, both a bar chart and table of matrix diagonal term ratios are output.
TRANSLATIONAL DOF DIAGONAL TERM RATIO STATISTICS CHART FOLLOWS FOR THE DECOMPOSITION OF MATRIX KLL
------------------------------------------------|---------------------------------------------------------------------------
DIAGONAL TERM RATIO RANGE #TERMS % TOT |MAXIMUM RATIO = 6.90963E+02 MINIMUM RATIO = 1.00000E+00
------------------------------------------------|---------------------------------------------------------------------------
1.0000E+00 TO 1.0000E+01 62 79.49 |**************************************************************************>
1.0000E+01 TO 1.0000E+02 12 15.38 |***************
1.0000E+02 TO 1.0000E+03 4 5.13 |*****


0
0 MATRIX/FACTOR DIAGONAL TERMS RATIO SUMMARY TABLE FOR TRANSLATIONAL DOF SORTED ON DIAGONAL RATIO
GRID POINT ID DEGREE OF FREEDOM MATRIX/FACTOR DIAGONAL RATIO MATRIX DIAGONAL
(TOP 1 RATIOS>MAXRAT= 6.90963E+02)
58 T3 6.90963E+02 5.65535E+04


ROTATIONAL DOF DIAGONAL TERM RATIO STATISTICS CHART FOLLOWS FOR THE DECOMPOSITION OF MATRIX KLL
------------------------------------------------|---------------------------------------------------------------------------
DIAGONAL TERM RATIO RANGE #TERMS % TOT |MAXIMUM RATIO = 3.35974E+02 MINIMUM RATIO = 1.00000E+00
------------------------------------------------|---------------------------------------------------------------------------
1.0000E+00 TO 1.0000E+01 38 63.33 |***************************************************************
1.0000E+01 TO 1.0000E+02 18 30.00 |******************************
1.0000E+02 TO 1.0000E+03 4 6.67 |*******
0
0 MATRIX/FACTOR DIAGONAL TERMS RATIO SUMMARY TABLE FOR ROTATIONAL DOF SORTED ON DIAGONAL RATIO
GRID POINT ID DEGREE OF FREEDOM MATRIX/FACTOR DIAGONAL RATIO MATRIX DIAGONAL
(TOP 1 RATIOS>MAXRAT= 3.35974E+02)
58 R2 3.35974E+02 2.14135E+04
MD Nastran R2.1 Release Guide

206
New SPARSESOLVER MDTSTATS Information Output
(Prerelease)
Introduction
A new interface is now available to control the generation of matrix diagonal term statistics for the input
matrix to the sparse symmetric matrix decomposition process in the DCMP module. The matrix diagonal
term statistics can be useful in determining the quality of the model in regions that produce unusually
large or small terms. In general, for linear static analysis, model degrees of freedom with small
stiffnesses could indicate areas where loads will produce large displacements. This feature complements
the MDTRATIO option that controls MAXRATIO output described previously.
Benefits
The new interface provides another means of identifying potential modeling errors other than monitoring
the MAXRATIO statistics. One of the new output data options is a simple bar chart that provides a more
comprehensive view of the diagonal term data.
Method and Theory
No new theory is involved. The method involves adding the original matrix diagonal term to the ratio
table where the computation of the ratio is defined to be the original matrix diagonal term divided by the
decomposed matrix diagonal term. These terms are placed together in a table with the external identifier
associated with the row/column of the term. This table is then processed according to the options
requested by the user.
Inputs
The matrix diagonal term statistical output options are controlled by keywords specified on the
SPARSESOLVER Executive Control statement. See the MD Nastran Quick Reference Guide for a
complete description of this statement.
Outputs
The matrix diagonal term statistics can be presented in two different views. The first is the table view in
which each diagonal term is listed together with the external identifier of the row/column of the matrix,
as well as with the Aii/Lii diagonal term ratio. This format is almost identical to that produced now when
any ratio exceeds the value of the MAXRATIO input parameter. The second view of the diagonal terms
is statistical in nature, similar to a bar chart. A series of bar segments is generated. There are two options
for specifying the segment widths of the bars. The default option uses powers of 10 as the widths (e.g.,
10.0 to 100.0, and 100.0 to 1000.0). The second option allows the user to specify how many segments
are desired. The program will compute the segment width using the maximum and minimum diagonal
207
CHAPTER 6
Numerical Enhancements
terms. For each bar in the chart, the total number of terms in the range is tabulated together with a visual
indication of the percentage number of terms in that particular bar.
These new views of the diagonal terms do not replace any existing diagnostics generated by the DCMP
module if a problem is detected. Under these conditions, output from the table view may duplicate
previous output generated by DCMP module error processing.
Guidelines and Limitations
The matrix diagonal term statistics are sometimes useful in determining areas of the model that may pose
problems during the decomposition process, or afterwards during the solution of equations that produce
displacements. In general, unusually large or small values could indicate a modeling problem. The
feature can be used by taking all of the program defaults for the various control variables. These defaults
produce both the table and bar outputs. The table is limited to the 25 largest terms that exceed 1.0E+10,
and the 25 smallest terms less than 1.0. The bar chart uses powers of ten for segment widths. This can
be done by adding
SPARSESOLVER DCMP ( MDTSTATS )
to the Executive Control Section of the input data file.
The use of this new feature is currently limited to sparse symmetric matrix operations in the DCMP
module.
If there are scalar-type points present in the problem, the degrees of freedom associated with these points
will be grouped into the results for the translational degrees of freedom output.
Demonstration Example
A simple example is presented that demonstrates the use of some of the new features available for output
of the matrix diagonal term statistics. The SPARSESOLVER Executive Control statement is used to
specify the desired features. The example problem is used for demonstration purposes only, and is not
representative of any particular model. The model data consists of a simple plate structure subject to an
end load. The model properties have been designed to indicate a potential problem in the bending
properties at grid points 4 and/or 54.
Example Input Data
$
$ Example problem to demonstrate matrix diagonal term statistics
$
id test,case
sol 101
$ Note: SPARSOLVER DCMP options must be enclosed in ()
$ Note also that MDTSTATS options must also be enclosed in their own
()
SPARSESOLVER DCMP ( MDTSTATS = ( CHART,
TABLET, NMAXVALT=10, MAXVALT=1.0e+08,
MD Nastran R2.1 Release Guide

208
NMINVALT=20, MINVALT=1.0,
TABLER, NMINVALR=30, MINVALR=100.0 ) )
cend
spc=100
load=1000
disp=all
begin bulk
grdset,,,,,,,6
cquad4,101,101,1,2,52,51
cquad4,102,101,2,3,53,52
cquad4,103,102,3,4,54,53
cquad4,104,102,4,5,55,54
cquad4,105,101,5,6,56,55
cquad4,106,101,6,7,57,56
cquad4,107,101,7,8,58,57
cquad4,108,101,8,9,59,58
cquad4,109,101,9,10,60,59
cquadr,1101,101,1,2,52,51
cquadr,1102,101,2,3,53,52
cquadr,1103,102,3,4,54,53
cquadr,1104,102,4,5,55,54
cquadr,1105,101,5,6,56,55
cquadr,1106,101,6,7,57,56
cquadr,1107,101,7,8,58,57
cquadr,1108,101,8,9,59,58
cquadr,1109,101,9,10,60,59
grid, 1,, 0.0,0.0,0.0
grid, 2,, 1.0,0.0,0.0
grid, 3,, 2.0,0.0,0.0
grid, 4,, 3.0,0.0,0.0
grid, 5,, 4.0,0.0,0.0
grid, 6,, 5.0,0.0,0.0
grid, 7,, 6.0,0.0,0.0
grid, 8,, 7.0,0.0,0.0
grid, 9,, 8.0,0.0,0.0
grid,10,, 9.0,0.0,0.0
grid,51,, 0.0,1.0,0.0
grid,52,, 2.4,1.0,0.0
grid,53,, 3.5,1.0,0.0
grid,54,, 4.6,1.0,0.0
grid,55,, 5.7,1.0,0.0
grid,56,, 6.8,1.0,0.0
grid,57,, 7.9,1.0,0.0
grid,58,, 9.0,1.0,0.0
grid,59,,10.1,1.0,0.0
grid,60,,11.2,1.0,0.0
$
ctria3,201,101,101,102,151
ctria3,202,101,102,152,151
ctria3,203,101,102,103,152
ctria3,204,101,103,153,152
ctria3,205,101,103,104,153
ctria3,206,101,104,154,153
ctria3,207,101,104,105,154
209
CHAPTER 6
Numerical Enhancements
ctria3,208,101,105,155,154
ctriar,1201,101,101,102,151
ctriar,1202,101,102,152,151
ctriar,1203,101,102,103,152
ctriar,1204,101,103,153,152
ctriar,1205,101,103,104,153
ctriar,1206,101,104,154,153
ctriar,1207,101,104,105,154
ctriar,1208,101,105,155,154
grid,101,, 0.0,0.0,0.0
grid,102,, 1.0,0.0,0.0
grid,103,, 2.0,0.0,0.0
grid,104,, 3.0,0.0,0.0
grid,105,, 4.0,0.0,0.0
grid,151,, 0.0,1.0,0.0
grid,152,, 3.4,1.0,0.0
grid,153,, 4.5,1.0,0.0
grid,154,, 5.6,1.0,0.0
grid,155,, 6.7,1.0,0.0
$
pshell,101,1,0.05,1
pshell,102,1,0.05,2
mat1,1,10.+6,,0.33
mat1,2,10.+1,,0.33
spc1,100,123,1,101
spc1,100,3,5,55,105,155
spc1,100,1,55,155
spc1,100,2,1,101
force,1000,10,,1000.0,1.0,0.0,0.0
force,1000,60,,1000.0,1.0,0.0,0.0
force,1000,105,,1000.0,1.0,0.0,0.0
force,1000,155,,1000.0,1.0,0.0,0.0
enddata
Example Output
The output generated by the example is shown as follows. There are two separate sections of output: one
for translational degrees of freedom and one for rotational. Within each section, both a bar chart and
table of matrix diagonal terms are output.
===============================================================================================================================
TRANSLATIONAL DOF Aii DIAGONAL TERMS STATISTICS CHART FOLLOWS FOR MATRIX KLL Matrix Trace(Aii) = 1.27351E+08
------------------------------------------------|---------------------------------------------------------------------------
| MAXIMUM VALUE = 5.94121E+06 MINIMUM VALUE = 4.07806E-01
MATRIX DIAGONAL TERM RANGE #TERMS % TOT | GRID ID = 104, DOF = T2 GRID ID = 54, DOF = T3
------------------------------------------------|---------------------------------------------------------------------------
1.0000E-01 TO 1.0000E+00 2 2.56 |***
1.0000E+03 TO 1.0000E+04 2 2.56 |***
1.0000E+04 TO 1.0000E+05 18 23.08 |***********************
1.0000E+05 TO 1.0000E+06 6 7.69 |********
1.0000E+06 TO 1.0000E+07 50 64.10 |****************************************************************
===============================================================================================================================
MD Nastran R2.1 Release Guide

210
0 MATRIX/FACTOR DIAGONAL TERMS SUMMARY TABLE FOR TRANSLATIONAL DOF SORTED ON Aii DIAGONAL
GRID POINT ID DEGREE OF FREEDOM Aii TERM Lii TERM Aii/Lii RATIO
(TOP 1 VALUES > 5.94121E+06)
104 T2 5.94121E+06 3.57301E+06 1.66280E+00
0 MATRIX/FACTOR DIAGONAL TERMS SUMMARY TABLE FOR TRANSLATIONAL DOF SORTED ON Aii DIAGONAL
GRID POINT ID DEGREE OF FREEDOM Aii TERM Lii TERM Aii/Lii RATIO
(TOP 2 VALUES < 1.00000E+00)
54 T3 4.07806E-01 4.07806E-01 1.00000E+00
4 T3 4.70350E-01 2.84250E-01 1.65471E+00
===============================================================================================================================
ROTATIONAL DOF Aii DIAGONAL TERMS STATISTICS CHART FOLLOWS FOR MATRIX KLL Matrix Trace(Aii) = 4.52211E+05
------------------------------------------------|---------------------------------------------------------------------------
| MAXIMUM VALUE = 2.34493E+04 MINIMUM VALUE = 4.71107E-02
MATRIX DIAGONAL TERM RANGE #TERMS % TOT | GRID ID = 9, DOF = R2 GRID ID = 54, DOF = R1
------------------------------------------------|---------------------------------------------------------------------------
1.0000E-02 TO 1.0000E-01 2 3.33 |***
1.0000E-01 TO 1.0000E+00 2 3.33 |***
1.0000E+02 TO 1.0000E+03 3 5.00 |*****
1.0000E+03 TO 1.0000E+04 36 60.00 |************************************************************
1.0000E+04 TO 1.0000E+05 17 28.33 |****************************

===============================================================================================================================
0 MATRIX/FACTOR DIAGONAL TERMS SUMMARY TABLE FOR ROTATIONAL DOF SORTED ON Aii DIAGONAL
GRID POINT ID DEGREE OF FREEDOM Aii TERM Lii TERM Aii/Lii RATIO
(TOP 1 VALUES > 2.34493E+04)
9 R2 2.34493E+04 4.87230E+03 4.81277E+00
0 MATRIX/FACTOR DIAGONAL TERMS SUMMARY TABLE FOR ROTATIONAL DOF SORTED ON Aii DIAGONAL
GRID POINT ID DEGREE OF FREEDOM Aii TERM Lii TERM Aii/Lii RATIO
(TOP 4 VALUES < 1.00000E+02)
54 R1 4.71107E-02 3.08701E-02 1.52610E+00
4 R1 4.71334E-02 9.03202E-03 5.21847E+00
54 R2 1.48139E-01 8.62427E-02 1.71770E+00
4 R2 1.48232E-01 7.96377E-02 1.86133E+00
211
CHAPTER 6
Numerical Enhancements
Performance Improvements and Expanded
Capabilities for Automated Component Modal
Synthesis (ACMS)
Introduction
Automated component modal synthesis (ACMS) is a powerful tool for a variety of large modal-based
analyses. ACMS functions in two domains, the matrix domain and the geometric domain. Since its
introduction in 2005, matrix domain ACMS (MDACMS) has become the default ACMS method.
Developments described in this section apply to MDACMS.
MDACMS has been extended to more thoroughly interact with the MD Nastran external superelement
capability. See the following section, MDACMS for Upstream Superelements.
Automatic decision logic has been implemented to select which path to take for frequency response
calculations. See the following section, Automatic FASTFR Decision Logic.
For other enhancements, see the following section, Miscellaneous Performance Improvements.
MDACMS for Upstream Superelements
MDACMS has been extended to more thoroughly interact with the MD Nastran external superelement
capability. Specifically, MDACMS reduction calculations have been expanded to include all quantities
being reduced to the residual; namely, stiffness, mass, and damping. (Previously, only the component
modes were computed by MDACMS.) By reducing these quantities while simultaneously computing
component modes, a significant reduction in compute time, disk I/O, and scratch disk space is realized.
A typical case study is presented here. Note that in this example, the resource requirements for MD
Nastran R1 exceeded availability and the job did not complete, even though it ran five times longer than
the current release at the time of its failure.
No. of Grid
Points
No. of
Global DOFs
No. of O-set
DOFs
No. of A-set
DOFs Analysis Description
375,502 2.25 million 1.9 million 3526 Normal modes, frequency range
0-550 Hz (1627 modes).
MD Nastran R2.1 Release Guide

212
The jobs illustrated in the following figures were run on an IBM Power4 computer running the AIX
operating system:
Automatic FASTFR Decision Logic
In MSC.Nastran 2004, the FASTFR method was introduced for modal frequency response analysis. It
can be selected via the Bulk Data entry PARAM,FASTFR,YES and shows significant performance
improvement for certain models in the mid-frequency range. However, in prior versions the user had to
make the decision to use the new method, or to run with the standard FRRD1 method with or without the
iterative solver, before starting the job. With this release, automatic decision logic has been implemented
which eliminates the need for the user to make that decision. The program will decide automatically
which solution method will be most efficient for the frequency response portion in a SOL 111 analysis.
Based on the size of the modal space and some other heuristic criteria, either the FASTFR solution
method will be run, or the FRRD1 module with or without the iterative solver will be used.
Currently, the FASTFR method is selected or deselected in the Bulk Data Section using
PARAM,FASTFR,YES or PARAM,FASTFR,NO. To activate the automatic decision logic, specify
PARAM, FASTFR, AUTO
in the Bulk Data Section. This will cause the program to run as though the FASTFR method was selected
until it reaches the point where the decision will be made. If the program decides it would be faster not
to run with the FASTFR method, the following message will be printed in the .f06 file, and the program
will continue with the standard frequency response method.
Also, if the FASTFR method is deselected for a different reason, a similar message will be printed. If the
FASTFR method will actually be used, then no message will be printed.
^^^ SYSTEM INFORMATION MESSAGE 9157 (GMA)
^^^ FASTFR OPTION REQUESTED, BUT THE MODEL DID NOT MEET THE FOLLOWING CRITERIA:
^^^ THE FASTFR OPTION MAY BE TOO EXPENSIVE.
^^^ USER INFORMATION: STANDARD FREQUENCY RESPONSE METHOD WILL BE USED
213
CHAPTER 6
Numerical Enhancements
This new option for PARAM,FASTFR is helpful for all modal frequency response analysis jobs,
especially in the mid- to upper frequency range.
Examples
The following charts show the benefit of the automatic decision logic for three examples. The times
given are the elapsed minutes for the complete job.
Example 1
For this automotive example, the automatic decision logic decided that the FASTFR method might be
too expensive, so it switched to the FRRD1 module with the iterative solver.

Example 2
For this automotive example, the automatic decision logic decided to go with the FASTFR method, and
rightly soboth theFRRD1 module with and without the iterative solver proved to be much slower.
Job Statistics: 1,208,135 grid points;
matrix size in FRRD1 (i.e., H-size) is 9,974;
modes below 1,000 Hz;
two frequencies.
Job Statistics: 233,094 grid points;
matrix size in FRRD1 is 4,350;
modes below 600 Hz;
260 frequencies.
0
200
400
600
800
1000
1200
1400
FASTFR FRRD1 (iter) FRRD1
(direct)
Decision
Time
MD Nastran R2.1 Release Guide

214
Example 3
For this automotive example, the automatic decision logic selected the FASTFR method, which proved
to be significantly faster than the conventional method (i.e., FRRD1 with the direct method). The
iterative method inside FRRD1 is not practical for this job because there are 147 loads, which would
significantly slow the iterative solver.
Limitations for the FASTFR Method
1. Fluid damping can only be specified via the PARAM,GFL parameter and/or the
SDAMPING(FLUID) command.
a. The CAABSF element (acoustic absorber) is not supported.
b. The FASTFR method will be turned off automatically for fluid K4 and for fluid viscous
damping.
2. All matrices must be symmetric.
Job Statistics: 321,597 grid points;
matrix size in FRRD1 is 5,424;
modes below 600 Hz;
501 frequencies.
0
50
100
150
200
250
300
350
FASTFR FRRD1 (iter) FRRD1
(direct)
Decision
Time
0
500
1000
1500
2000
2500
FASTFR FRRD1(direct) Decision
Time
215
CHAPTER 6
Numerical Enhancements
a. Unsymmetric formulations for acoustic coupling are not supported.
b. EPOINT Bulk Data entries are not supported.
3. The FASTFR method works only in SOL 111, and in SOL 200 with ANAL=MFREQ.
4. SESDAMP and FASTFR are not allowed in the same run.
Miscellaneous Performance Improvements
MDACMS has been enhanced to reduce the amount of disk I/O required for a successful analysis. This
happens automatically, and there is no user action required to realize the reduction in I/O. Typically,
reduced I/O requirements result in reduced elapsed time and greater CPU utilization, which increases the
effectiveness of shared memory parallel (SMP) computations.
To demonstrate some of these performance improvements, three example jobs were run and the
performance results were compared to the previous version. The results are as follows.
Model Description:
Example 1 shows time and I/O resource requirements for model 1 run on an IBM Power4 system under
AIX.
Model
No.
No. of Grid
Points
No. of Global
DOFs Analysis Description
1 44,314 265884 Auto body normal modes run w/ACMS. Frequency range
0-400 Hz (360 modes).
2 44,314 265884 Model (1) run in SOL 111 with ACMS; 75 load cases, 250
forcing frequencies.
3 1,243,651 7.4 million Mid-frequency acoustic analysis of automotive trimmed
body. 9389 structure modes below 750 Hz; 1486 fluid
modes below 1200 Hz; four load cases, 290 forcing
frequencies.
MD Nastran R2.1 Release Guide

216
Example 2 shows DMP scalability improvement for model 2. These jobs were run on an Altix machine
with IA64 processors under Linux.
217
CHAPTER 6
Numerical Enhancements
Example 3 shows I/O and elapsed time improvement for model 3. These jobs were run on an IBM
Power4 system under AIX.
MD Nastran R2.1 Release Guide

218
Relaxed Restrictions for CASI Solver Usage
The use of the element-based CASI iterative solver is limited by several restrictions noted in the remarks
for the ITER Bulk Data entry. In the current version, two of these restrictions have been relaxed so that
the solver is usable over a wider range of problems.
The first improvement is a reduction in the number of element types that the solver does not recognize.
The solver recognizes only a subset of all of the element types available, but the list has been expanded
to support the CFAST, CSEAM, and CWELD element types.
The second improvement is a relaxation of the restriction associated with the processing of direct input
matrices via the K2GG Case Control command. K2GG matrix input is now allowed by the CASI solver
interface. However, the matrix size is limited to 100 grid/scalar points since it must be treated as an
unknown finite element type. This should be sufficient for most cases where the input K2GG matrix
represents some form of attachment to, or base for, the structural model being analyzed.
219
CHAPTER 6
Numerical Enhancements
System-Dependent Performance Improvements
Introduction
MSC.Software has implemented new versions of the basic linear algebra subroutine library (BLAS) for
MD Nastran, on platforms supported by the x86-64 architecture.
Version 9 of Intels math kernel library (MKL) provides optimal performance for analysis solutions that
make good use of the BLAS. This includes the ACMS solution for automotive NVH and acoustic
analysis, and the exterior acoustics capability in MD Nastran. In addition, MKL provides automatic
multithreading support. Shared memory parallel speedup is available via the SMP=n command supplied
to the NASTRAN command processor.
The automatically tuned linear algebra software project (ATLAS) provides a portable, efficient version
of the BLAS. An ATLAS version of the BLAS library has been implemented for use on AMD Opteron
processors. ATLAS is public domain software available on the internet at
http://math-atlas.sourceforge.net/.
For demonstration purposes, two analysis models were selected and run with the current and prior release
of MD Nastran. Performance results are shown as follows:
Model Description
Model
No.
No. of Grid
Points
No. of Global
DOFs Analysis Description
1 268,486 1.6 million Auto body normal modes run w/ACMS. Frequency range
0-200 Hz (1043 modes).
2 603,266 2.8 million Low frequency acoustic analysis of automotive trimmed
body.
MD Nastran R2.1 Release Guide

220
Performance Improvements for Intel x86_64
Performance Improvements for AMD due to Atlas
Processor Clock Speed OS Real Memory Scratch Filesystem
Woodcrest 2666 MHz Linux RHWS4 16 GB ext2 not striped
Processor Clock Speed OS Real Memory Scratch Filesystem
Opteron 2400 MHz Linux SuSE 10 8 GB xfs not striped
Intel Woodcrest Performance Comparison
Model 2
0
2000
4000
6000
8000
10000
12000
14000
MD R1 MD R2 MD R2 2smp
17%CPU Ti me Speedup(seri al )
T
i
m

i
n

S
e
c
o
n
d
s
elapsed
cpu
Intel Woodcrest Performance Comparison
Model 1
0
500
1000
1500
2000
MD R1 MD R2 MD R2 2smp
17%CPU Ti me Speedup (seri al )
T
i
m

i
n

S
e
c
o
n
d
s
elapsed
cpu
AMD Opteron Performance Comparison
Model 1
0
500
1000
1500
2000
2500
3000
MD R1 MD R2
17%CPU Ti me Speedup
T
i
m
e

i
n

s
e
c
o
n
d
s elapsed
cpu
AMD Opteron Performance Comparison
Model 2
0
5000
10000
15000
20000
MD R1 MD R2
18% CPU Time Speedup
T
i
m
e

i
n

s
e
c
o
n
d
s
elapsed
cpu
221
CHAPTER 6
Numerical Enhancements
Improved Selection of Reordering Methods
MD Nastran contains five methods for reordering (permuting) sparse matrices in preparation for a sparse
symmetric factorization. The five methods are minimum degree (MD), multiple minimum degree
(MMD), and three nested dissection/minimum degree hybrid methods called BEND, AMF, and METIS.
A judicious choice of a reordering method can lead to a dramatic improvement in the performance of the
sparse factorization.
The hybrid methods (BEND, AMF, and METIS) are generally superior to MD and MMD. The previous
default behavior was to use BEND unless the USET and SIL tables were not available, in which case
MD was used. If the BEND algorithm failed, MD was also used in this case. The new default behavior
is to use BEND except for large (> 50,000 DOF) models, which are dominated by three-dimensional
elements; for those models, METIS is the default. If the default method fails, the program will attempt
to use another hybrid method (for example, if BEND fails, the program tries METIS). The minimum
degree algorithms are selected only as a last resort.
To illustrate the performance improvements, the following models were run on an SGI Altix with four
1.4 GHz itanium2 processors, 20 GB of ram, and 470 GB of disk space.
The following chart shows the factorization time for each job. The improvement for the xx0e, xx0d, and
xx0f jobs stems from the automatic selection of METIS rather than BEND reordering. The improvement
in lg0q is due to the selection of METIS rather than MD.
Model Name Solution DOFs Elements
xx0e Normal Modes 3308298 656570 CTETRAs
xx0d Normal Modes 1920855 402441 CTETRAs
xx0f Statics 1237999 246066 CTETRAs
lg0q Direct Freq. Response 93375 28140 CHEXAs
MD Nastran R2.1 Release Guide

222
xx0e xx0f xx0d lg0q
0
250
500
750
1000
1250
1500
1750
2000
2250
2500
2750
Reordering Select ion Performance Improvement
2006R1
2006R2
Model
F
a
c
t
o
r
i
z
a
t
i
o
n

t
i
m
e

(
c
p
u

s
e
c
)
223
CHAPTER 6
Numerical Enhancements
Iterative Solution Options for Nonlinear Analysis
Introduction
Nonlinear analysis is available in MD Nastran SOLs 400 and 101. At each nonlinear iteration, a solution
of the equations is performed using sparse direct factorization and forward-backward
substitution (FBS). For large models composed primarily of solid finite elements, the bulk of the
solution time is spent computing . For MD Nastran R2.1, iterative solutions are available to
compute for models which will benefit from an iterative solution, such as solid models.
Benefits
Users will experience significant performance increase by selecting an iterative solver for nonlinear
analysis of large solid models. Depending on the number of nonlinear iterations in the overall analysis,
the expected speedup is from two to five times.
Inputs
To select an iterative solver, specify the SMETHOD command in the Case Control Section. The
following SMETHOD command selects the CASI element-based iterative solver:
SMETHOD = ELEMENT
To select a matrix-based iterative solver, specify
SMETHOD = MATRIX
The ELEMENT solver is recommended for best performance.
To modify the specific parameter settings for one of the preceding iterative solvers, the SMETHOD
command can specify the ID of an ITER Bulk Data entry. The user interface for the iterative solver is
the same as it is for linear static analysis in SOL 101. Refer to the MD Nastran Quick Reference Guide,
Case Control command SMETHOD, and Bulk Data entry ITER, for more information.
Outputs
There are no new engineering outputs associated with this feature other than informational and
diagnostic messages. In addition, a PCS output text file contains additional diagnostic output.
Guidelines and Limitations
These new iterative solution options are primarily intended for use in nonlinear contact analysis of large
solid models exceeding one million DOFs. In addition, they are available for linear contact in SOL 101.
Iterative solution techniques are available only for nonlinear static analysis. Nonlinear transient analysis
must still be performed using sparse direct factorization and FBS in order to solve for .
A [ ] x { } b { } =
x { }
x { }
x { }
MD Nastran R2.1 Release Guide

224
The follower stiffness feature cannot currently be used if the CASI solver is selected due to the
unsymmetrical nature of the follower stiffness matrix. Follower stiffness is generated automatically by
default if any of the following Bulk Data entries are present: MOMENTi, FORCEi, PLOADi, and
RFORCE. In order to use the CASI solver with any of these entries, PARAM,FOLLOWK,NO must be
specified in the Bulk Data Section. However, this will alter the analysis results by not including follower
stiffness. The presence of these Bulk Data entries will generate User Fatal Message 9192 unless
PARAM,FOLLOWK,NO is also specified.
Ch. 7: Elements and Connectors MD Nastran R2.1 Release Guide

7
Elements and Connectors
!
Enhancements to Connector Elements
!
Composite Beam Using the Variational Asymptotic Method (VAM)
!
Nonhomogeneous Multipoint Constraint
MD Nastran R2.1 Release Guide

226
Enhancements to Connector Elements
Introduction
In MD Nastran R2, a new seam weld element, CSEAM, is introduced to define a seam line. This element
replaces the existing CWSEAM element, and is modeled using the new CSEAM and PSEAM Bulk Data
entries. In addition, the SWLDPRM Bulk Data entry is enhanced to support element type specific control
parameters and different diagnostic output format.
For the spot weld elements CWELD and CFAST, the displacements of the projected grid points GA and
GB are now computed and output in this release. These data will help users view the relationship between
the spot weld and the connecting shells.
CSEAM Element
The new CSEAM element provides the following key features to extend the analysis capabilities, provide
more flexibility in modeling, and improve the robustness of the analysis results.
A seam line is considered continuous between two CSEAM elements that have a common face
based on either common GS/GE grid points or XYZ coordinates. Note that the SMLN label on a
CSEAM element does not determine the definition of a seam line; it is only intended for ease of
seam line visualization.
The CSEAM element can connect up to 64 shell grid points, which allows the connection of
higher order shell elements.
Besides selecting the connected surface patches by property IDs, the user may define the
connection by specifying shell element IDs directly.
Each of the eight auxiliary grid points must have a projection onto the connected shell element.
This requirement prevents the generation of an ill-conditioned stiffness matrix in downstream
processing.
The user can model tailored blank tapering by specifying different property IDs at the start and
end points.
The CSEAM element type supports MAT9 anisotropic material properties.
Inputs
The seam connection is modeled by the CSEAM and PSEAM Bulk Data entries and the modified
SWLDPRM Bulk Date entry. The details of these new entries are described in the MD Nastran Quick
Reference Guide.
Outputs
The connecting record based on the seam line label is written as a SEAMLN record in the GEOM2 data
block, which is included in the .op2 file for postprocessing. The contents of the SEAMLN record for each
seam line label are listed in the following table:
227
CHAPTER 7
Elements and Connectors
Limitations
Each CSEAM element can connect a maximum of three shell elements on patch A, and three
shell elements on patch B.
Only a line type of seam is supported.
Superelement modeling is not supported.
FORCE, STRESS, and STRAIN output requests are not supported.
Example A Symmetric Hat Profile (cseam_hut.dat)
This example demonstrates the application of CSEAM elements to analyze an unconstrained symmetric
hat profile model, see Figure 7-1. Each edge of the hat is connected by 27 CSEAM elements. The grid
points with identification numbers 6000 to 6028, and 8000 to 8028, are used as the piercing points to
define the seams, see Figure 7-2.
Hat Profile
Figure 7-1 Hat Profile.
Word Type Description
1-2 C Seam line ID.
3 I Total number of elements for this seam line.
4 I CSEAM element ID.
5 I GS grid point ID for the start point.
6 I GE grid point ID for the end point.
: I Repeat words 4-6 for each CSEAM element.
: I End the seam line data with -1.
MD Nastran R2.1 Release Guide

228
Piercing Points and Seams
Figure 7-2 Piercing Points and Seams.
The Bulk Data input for the seam welds is listed as follows:
CSEAM 7000 500 SMLN_A PSHELL 100 200
6000 6001
CSEAM 7001 500 SMLN_A PSHELL 100 200
6001 6002
:
CSEAM 7026 500 SMLN_A PSHELL 100 200
6026 6027
CSEAM 7027 500 SMLN_A PSHELL 100 200
6027 6028
$
CSEAM 9000 500 SMLN_B PSHELL 100 200
8000 8001
CSEAM 9001 500 SMLN_B PSHELL 100 200
8001 8002
:
CSEAM 9026 500 SMLN_B PSHELL 100 200
8026 8027
CSEAM 9027 500 SMLN_B PSHELL 100 200
8027 8028
$
PSEAM 500 1 LINE 1.0
229
CHAPTER 7
Elements and Connectors
The normal mode analysis results are shown in the following table:
SWLDPRM Enhancements
Element-Specific Control Parameters
The SWLDPRM Bulk Data entry is enhanced to support element type specific control parameters. Two
new parameters, CNRAGLI and CNRAGLO, are introduced to define the angle limits for checking the
geometry of seam elements. These parameters replace the GSPROJ parameter to define the allowable
angles for corner check. The GSPROJ parameter specified in the preceding seam weld models must be
replaced by CNRAGLI and CNRAGLO to obtain identical results.
For example, if the SWLDPRM Bulk Data entry is defined as
SWLDPRM,PROJTOL,0.0,GSMOVE,0,GSPROJ,-1.,NREDIA,4,
,GMCHK,1,CHKRUN,1
then this entry must be changed into
SWLDPRM,PROJTOL,0.0,GSMOVE,0,GSPROJ,-1.,NREDIA,4,
,GMCHK,1,CHKRUN,1,cnraglo,-1.
to get the same results as the results produced by the previous version.

R E A L E I G E N V A L U E S
MODE EXTRACTION EIGENVALUE RADIANS CYCLES GENERALIZED GENERALIZED
NO. ORDER MASS STIFFNESS
1 1 -5.927181E-05 7.698819E-03 1.225305E-03 1.000000E+00 -5.927181E-05
2 2 -2.941204E-05 5.423287E-03 8.631429E-04 1.000000E+00 -2.941204E-05
3 3 2.051645E-05 4.529509E-03 7.208938E-04 1.000000E+00 2.051645E-05
4 4 2.372748E-05 4.871086E-03 7.752574E-04 1.000000E+00 2.372748E-05
5 5 5.330599E-05 7.301095E-03 1.162005E-03 1.000000E+00 5.330599E-05
6 6 1.118197E-04 1.057448E-02 1.682981E-03 1.000000E+00 1.118197E-04
7 7 1.154947E+07 3.398451E+03 5.408803E+02 1.000000E+00 1.154947E+07
8 8 1.585229E+07 3.981493E+03 6.336742E+02 1.000000E+00 1.585229E+07
9 9 2.653947E+07 5.151647E+03 8.199102E+02 1.000000E+00 2.653947E+07
10 10 2.959784E+07 5.440389E+03 8.658648E+02 1.000000E+00 2.959784E+07
11 11 3.002028E+07 5.479077E+03 8.720222E+02 1.000000E+00 3.002028E+07
12 12 3.041736E+07 5.515193E+03 8.777703E+02 1.000000E+00 3.041736E+07
13 13 3.328896E+07 5.769658E+03 9.182696E+02 1.000000E+00 3.328896E+07
14 14 3.757468E+07 6.129819E+03 9.755909E+02 1.000000E+00 3.757468E+07
15 15 4.259932E+07 6.526814E+03 1.038775E+03 1.000000E+00 4.259932E+07
16 16 4.738684E+07 6.883810E+03 1.095592E+03 1.000000E+00 4.738684E+07
MD Nastran R2.1 Release Guide

230
Real Format of Diagnostic Output
For the PRTSW parameter of the SWLDPRM Bulk Data entry, two options are added to support real
format of diagnostic output so that more significant digits will show (error report 1-25861871). This
parameter now has five options.
Displacement Output of GA and GB for CWELD and CFAST
Elements
The displacements of the projected grid points GA and GB for CWELD elements (with GRIDID,
ELEMID, ELPAT, or PARTPAT format) and CFAST elements are computed to display the relationship
between these elements and their connecting shell elements. As a result, the displacement output of GA
and GB for a CWELD element with GRIDID or ELEMID format and MSET=OFF are no longer dummy
zero values. These displacements are calculated from the constraint equations described in the following
Basic Theory and Methods section.
If GA or GB are not specified for CFAST elements, or for CWELD elements with the ELPAT or
PARTPAT option, the program will create a grid point internally, with the grid point ID number starting
from OSWPPT+1 (OSWPPT is specified by a PARAM Bulk Data entry). If PRTSW > 0 on the
SWLDPRM Bulk Data entry, then the grid point ID of GA or GB will be shown in the diagnostic output.
Input
The displacements of GA and GB are requested using the CONNECTOR keyword of the
DISPLACEMENT Case Control command. See the MD Nastran Quick Reference Guide for a detailed
description of this command.
Outputs
The output is integrated with the displacements of the general grid points.
Basic Theory and Methods
For CWELD elements with GRIDID or ELEMID format, the displacements of the projected grid points
GA and GB in basic coordinates are computed using the following equations:
PRTSW Output
PRTSW = 0 No diagnostic output.
PRTSW = 1 Print diagnostic output in exponential format to .f06 file.
PRTSW = 2 Punch diagnostic output in exponential format to .pch file.
PRTSW = 11 Print diagnostic output in real format to .f06 file.
PRTSW = 12 Punch diagnostic output in real format to .pch file.
231
CHAPTER 7
Elements and Connectors
where and are displacements at GA and GB, and are constraint matrices, and and are
displacements of the connected shell grid points.
For CWELD elements with ELPAT or PARTPAT format, the displacements of the auxiliary grid points
in basic coordinates are calculated first using
where and are displacements of the auxiliary grid points, and are RBE3-type constraint
matrices, and and are displacements of the connected shell grid points.
Then, the displacements of the projected grid points GA and GB in basic coordinates are computed by
the same constraint equations used for the GRIDID and ELEMID options. The displacements are given
by
where and are displacements at GA and GB, and are constraint matrices, and and are
displacements of the connected auxiliary points.
Current Limitation
The displacement output of GA and GB are only available in solution sequences 101 and 103.
Example
This example demonstrates the various displacement output requests and their results for a small model
with two CWELD elements.
The input file is as follows:
nastran mesh
SOL 101
TIME 60
CEND
d
A
A
I
u
I
I

=
d
B
B
J
u
J
J

=
d
A
d
B
A
I
B
J
u
I
u
J
u
I
G
I K
u
K
K

=
u
J
G
JL
u
L
L

=
u
I
u
J
G
IK
G
JL
u
K
u
L
d
A
A
I
u
I
I

=
d
B
B
J
u
J
J

=
d
A
d
B
A
I
B
J
u
I
u
J
MD Nastran R2.1 Release Guide

232
TITLE= two elements, identical location for GA, GB, GS
OLOAD= ALL
FORCE = ALL
SUBCASE 1
SUBTITLE= shear the weld
SPC= 1
LOAD= 1
DISP= ALL
SUBCASE 2
SUBTITLE= in plane twist
set 21 = 1002,1003,2011,thru,2014
spc= 1
LOAD= 2
DISP(CONN=ALL)=21
SUBCASE 3
SUBTITLE= upper bending
set 32 = 4
set 33 = 1012,1013,2001,thru,2004
spc= 1
LOAD= 3
DISP(CONN=32)=33
BEGIN BULK
$
$ Grids of lower shell
grid, 1001, , 0., 0., 0.
grid, 1002, , 20., 0., 0.
:
grid, 1013, , 20., 10., 5.
grid, 1014, , 0., 10., 5.
$ Grids of upper shell
grid, 2001, , 0., 0., 0.0
grid, 2002, , 20., 0., 0.0
:
grid, 2013, , 20., 10., 6.0
grid, 2014, , 0., 10., 6.0
$ spot weld grid
grid, 3001, , 10.0, 5.0, 0.0
grid, 3011, , 10.0, 5.0, 10.0
$ quad4s
cquad4, 4001, 10, 1001, 1002, 1003, 1004
cquad4, 5001, 10, 2001, 2002, 2003, 2004
cquad4, 4011, 10, 1011, 1012, 1013, 1014
cquad4, 5011, 10, 2011, 2012, 2013, 2014
:
$ property and material
pshell, 10, 10, 1.0, 10
mat1, 10, 2.e+5 , , 0.3, 0.785e-8
$
$ spot welds
$
cweld, 4, 4, 3001, elemid, , , , , +cw4
+cw4, 4001, 5001
cweld, 5, 4, 3011, elemid, , , , , +cw5
+cw5, 4011, 5011
pweld, 4, 10, 5.0
$
enddata
233
CHAPTER 7
Elements and Connectors
The displacement results are shown as follows:
0 SUBCASE 1
D I S P L A C E M E N T V E C T O R
POINT ID. TYPE T1 T2 T3 R1 R2 R3
1001 G 0.0 0.0 0.0 0.0 0.0 0.0
1002 G 4.887498E-03 5.624979E-04 0.0 0.0 0.0 1.406245E-05
1003 G 4.887498E-03 -5.624979E-04 0.0 0.0 0.0 -1.406245E-05
1004 G 0.0 0.0 0.0 0.0 0.0 0.0
1011 G 0.0 0.0 0.0 0.0 0.0 0.0
1012 G 4.887498E-03 5.624979E-04 -7.685000E-01 -6.750000E-03 5.864999E-02 1.406245E-05
1013 G 4.887498E-03 -5.624979E-04 -7.685000E-01 6.750000E-03 5.864999E-02 -1.406245E-05
1014 G 0.0 0.0 0.0 0.0 0.0 0.0
2001 G 6.058339E-04 3.750000E-04 0.0 0.0 0.0 7.733938E-19
2002 G 5.605834E-03 3.750000E-04 0.0 0.0 0.0 1.084202E-1
2003 G 5.605834E-03 -3.750000E-04 0.0 0.0 0.0 1.191827E-18
2004 G 6.058339E-04 -3.750000E-04 0.0 0.0 0.0 7.623363E-19
2011 G 3.907157E-02 3.750000E-04 8.148733E-04 -2.562901E-16 3.850649E-02 2.836059E-17
2012 G 4.407157E-02 3.750000E-04 -7.693148E-01 -5.232793E-15 3.850649E-02 2.851452E-17
2013 G 4.407157E-02 -3.750000E-04 -7.693148E-01 -5.356609E-15 3.850649E-02 2.956961E-17
2014 G 3.907157E-02 -3.750000E-04 8.148733E-04 -2.636780E-16 3.850649E-02 2.803595E-17
3001 G 0.0 0.0 0.0 0.0 0.0 0.0
3011 G 0.0 0.0 0.0 0.0 0.0 0.0
101000001 G 2.443749E-03 8.917563E-18 0.0 0.0 0.0 8.402567E-19
101000002 G 3.105834E-03 8.944668E-18 0.0 0.0 0.0 9.774760E-19
101000003 G 2.443749E-03 2.517788E-16 -3.842500E-01 -2.775558E-15 3.842500E-02 2.303930E-17
101000004 G 4.157158E-02 3.029749E-15 -3.842500E-01 -2.789435E-15 3.850649E-02 2.786400E-17
0 SUBCASE 2
D I S P L A C E M E N T V E C T O R
POINT ID. TYPE T1 T2 T3 R1 R2 R3
1002 G 1.499616E-02 3.324066E-02 0.0 0.0 0.0 2.330633E-03
1003 G -1.499616E-02 3.324066E-02 0.0 0.0 0.0 2.330633E-03
2011 G 6.107051E-03 -8.968769E-03 -1.542189E-16 3.303767E-17 1.479144E-16 1.890160E-03
2012 G 2.110705E-02 4.220943E-02 -3.132043E-15 5.117434E-17 1.498367E-16 3.390160E-03
2013 G -2.110705E-02 4.220943E-02 -2.615099E-15 5.202362E-17 1.383761E-16 3.390160E-03
2014 G -6.107051E-03 -8.968769E-03 1.645747E-16 3.118436E-17 1.398878E-16 1.890160E-03
101000001 G 6.505213E-18 1.662033E-02 0.0 0.0 0.0 1.580825E-03
101000002 G 6.722053E-18 1.662033E-02 0.0 0.0 0.0 2.640160E-03
101000003 G 3.035766E-18 1.662033E-02 -1.434168E-15 4.168935E-17 1.434168E-16 1.580825E-03
101000004 G 1.461505E-16 1.662033E-02 -1.434197E-15 4.178690E-17 1.439374E-16 2.640160E-03
0 SUBCASE 3
D I S P L A C E M E N T V E C T O R
POINT ID. TYPE T1 T2 T3 R1 R2 R3
1012 G 3.460202E-16 9.473882E-16 -1.537000E+00 -1.350000E-02 1.173000E-01 6.294140E-17
1013 G -4.478185E-16 9.592263E-16 -1.537000E+00 1.350000E-02 1.173000E-01 6.130147E-17
2001 G 0.0 0.0 8.021409E-04 -1.965454E-14 7.693148E-02 0.0
2002 G 0.0 0.0 -1.537827E+00 6.397660E-15 7.693148E-02 0.0
2003 G 0.0 0.0 -1.537827E+00 6.758049E-15 7.693148E-02 0.0
2004 G 0.0 0.0 8.021409E-04 -2.021495E-14 7.693148E-02 0.0
101000001 G 0.0 0.0 -7.685000E-01 -6.591949E-15 7.685000E-02 0.0
101000002 G 0.0 0.0 -7.685127E-01 -6.664110E-15 7.693148E-02 0.0
MD Nastran R2.1 Release Guide

234
Composite Beam Using the Variational Asymptotic
Method (VAM)
Introduction
Arbitrary beam cross section (ABCS) capability was implemented in MSC.Nastran 2005. With the
increasing popularity of composite materials in the aerospace and automotive industries, composite
support for an arbitrary beam cross section has been requested by users of MD Nastran. The variational
asymptotic method (VAM) is used to compute the beam properties of an arbitrary cross section. VAM is
available for both isotropic and composite materials. The theoretical background of VAM can be found
in many publications; literature relevant to this implementation is listed in the references given at the end
of this section.
Benefits
Beam-like structures made of composite materials, such as rotor blades, can be modeled with 3-D
elements such as CHEXA, CPENTA, and CTETRA, which can account for interaction among plies. An
obvious drawback to the use of 3-D elements to model composite beams is the increase in model size,
which can decrease CAE productivity. The composite beam using VAM provides an alternative to
conventional 3-D modeling technique, and permits the use of the CBEAM3 three-noded beam element
to model composite beams. The layup of composite beam plies is described on the PCOMP/PCOMPG
Bulk Data entries. In addition, the cross section of the composite beam can be expressed conveniently
with the CP/OP options of the PBMSECT Bulk Data entry.
Input
Modification of PBMSECT for the Composite Beam
The following keywords are added to the PBMSECT Bulk Data entry:
Core(id)=[PCID,PT=(pid1,pid2)]
Specifies the composite layup for the core part of the composite. PCID is the ID of a PCOMPi/PCOMPG
Bulk Data entry. PT=(pid1,pid2) defines the start and end points of line segment(s) which utilize PCID.
Layer(id)=[PCID,SETID]
Specifies the composite layup for additional layer(s) placed on the top or bottom of core. PCID is the ID
of a PCOMP/PCOMPG Bulk Data entry. SETID selects a SET1/SET3 with POINT IDs.
OUTM=value(Integer > 0)
Points to the ID of BEGIN BULK ARBMODEL. OUTM is designed specifically for ABCS with finite
element discretization already available. Note that OUTM must not appear together with other keywords,
such as OUTP or INP, on a PBMSECT entry.
It should be noted that:
235
CHAPTER 7
Elements and Connectors
1. Keyword Core can be abbreviated as c. Similarly, keyword Layer can be abbreviated as L.
2. If PT=(pid1,pid2) clause is utilized, pid1 and pid2 do not have to be two neighboring grid
points in a set. However, PID1 must be ahead of PID2 in the sequence.
3. If Core and/or Layer appears in PBMSECT, the PID of PBMSECT should not be referenced on
a CBEAM entry. Instead, it should be referenced on a CBEAM3 entry.
4. Keywords T (for thickness of isotropic beam) and Core (and Layer, for composite beam) cannot
coexist on a PBMSECT entry.
5. If OUTM=arbid is used on PBMSECT, element connection, grid location, PSHELL, and material
bulk data entries must be provided after the entry BEGIN ARBMODEL=arbid.
6. The arbid that is specified on BEGIN ARBMODEL is considered global, and can be
referenced by PBMSECT with OUTM=arbid in a different BEGIN SUPER Bulk Data Section
for part superelements.
7. PBMSECT with Core or Layer must be used along with PARAM,ARBMSTYP,TIMOSHEN
in the Bulk Data Section.
Selection of VAM for Properties of the Arbitrary Beam
PARAM,ARBMSTYP,TIMOSHEN
The default value for parameter ARBMSTYP is VKI which does not support the composite beam.
PARAM,ARBMASP,2
The default value for parameter ARBMASP is 2 (integer). Since the thickness of a ply is usually the
smaller dimension, ARBMASP affects the size of a CQUAD4 lengthwise along a segment.
PARAM,ARBMASP is functional only when PARAM,ARBMSTYP=TIMOSHEN.
Full Stress Recovery for Cross Section
Similar to ABCS with isotropic material, full stress recovery must be requested. The required input for
ABCS logic to perform stress recovery for the entire cross section is the element force. Hence, to activate
stress recovery for the entire cross section, the following must be present in the input file:
1. ELFORCE (or simply FORCE) command in the Case Control Section; and
2. PARAM,ARBMSS,YES as a Bulk Data entry (Default = NO).
Due to the potential for a large amount of output, these stresses are only available in OUTPUT2 format.
In addition, the direct stresses corresponding to the maximum failure index computed during full ABCS
stress recovery are collected for each grid point of a CBEAM3 element, and made available in an .f06
file.
MD Nastran R2.1 Release Guide

236
Output
Generation of Finite Element Model for Cross Sections
Similar to ABCS with isotropic material, the composite beam can be requested with the parameter
PARAM,ARBMFEM,YES. The file generated has following naming convention:
AAA_xxyyy_zz.bdf
where AAA, by default, assumes the input file name. To alter AAA to a name a file other than the input
file name, use
ASSIGN opcase=any character string
As the naming implies, there should be a cross sectional FEM referenced by a PBRSECT/PBMSECT
Bulk Data entry. Note that PBMSECT supports constant section beam only. For a composite beam, a
PSHELL entry for each ply of PCOMP/PCOMPG is also part of the FEM. However, the material entries,
such as MAT1, MAT2, and/or MAT8, are not available in the FEM Bulk Data Sections.
The model can be modified (with material entries inserted) and used via ARBMODEL=arbid under the
OUTM option of PBMSECT. A word of caution on modification: CQUAD4 elements are ordered in such
a way that a line from GID1 to GID2 defines the x-axis of the element coordinate system. Current
implementation does not support reordering of element connections based on the user-provided
coordinate system.
Stress Output for the Entire Cross Section
The stresses for the entire cross section are written to a file in .op2 format and has the same naming
convention as the FEM for the cross section. For stresses, the file has the extension of .op2.
For screened stresses, an output example is shown as follows:
xx Character string of BR for PBRSECT and BM for PBMSECT.
yyy ID of PBRSECT or PBMSECT.
zz Station ID. 01 for end A of PBMSECT. No zz section for PBRSECT.
1 VAB ALGORITHM USING CORE ON PBMSECT JANUARY 8, 2007 MD NASTRAN 1/ 8/07 PAGE 819
TRANSVERSE TIP LOAD
0 SUBCASE 1

S T R E S S E S I N L A Y E R E D C O M P O S I T E E L E M E N T S ( B E A M 3 )
ELEMENT GRID PLY D I R E C T S T R E S S E S FAILURE MAXIMUM STRENGTH
ID ID ID NORMAL-1 NORMAL-2 NORMAL-3 SHEAR-12 SHEAR-23 SHEAR-13 THEORY FAIL. INDEX RATIO FLAG
2 302 2 2.468E+01 1.601E+01 2.570E+00 2.323E+01 4.991E-01 3.724E+00 TSAI-WU 7.161E-04 4.035E+02
102 2 1.685E+01 1.610E+01 -7.230E-01 1.993E+01 -1.377E-01 -5.572E-01 TSAI-WU 7.258E-04 4.470E+02
1301 2 1.558E+01 1.594E+01 -7.167E-01 1.938E+01 -1.162E-01 -5.280E-01 TSAI-WU 7.193E-04 4.569E+02
237
CHAPTER 7
Elements and Connectors
Guidelines
1. CORE (or C) should point to a PCOMP/PCOMPG entry that has the common plies for most, if
not all, segments. For example,
2. PBMSECT,32 is a box beam made of composite material. All segments have a common
CORE=204 with four plies. Segments from POINTs 2 through 5 have one ply on top,
layer=(210,101), and one ply at the bottom, L(2)=(210,103).
3. The negative numbers in SET1,103 indicate that the top segment traverses in opposite sequence
to the bottom segment.
4. The centerline of a profile defined by OUTP and BRP must fall in between plies.
5. Plies of a PCOMP/PCOMPG are always placed from bottom to top.
6. During transition from OUTP to BRP (or vice versa), all plies above or below the centerline,
defined via OUTP/BRP/SET1 entries, must be carried over.
7. Interior ply insertion or deletion between neighboring segments is not supported. A ply or plies
added on top of the outermost ply can be made via the Layer keyword of PBMSECT.
8. A PCOMP/PCOMPG entry referenced on a PBMSECT entry via Core/Layer is converted to
multiple PSHELL entries. The value in the THETA field on the PCOMP/PCOMPG entry is
placed in the T (thickness) field of the PSHELL entry. Due to the nonnegative requirement of the
T field of a PSHELL entry, a negative value of THETA is converted to a supplementary positive
value. An example of this conversion is shown as follows:
$.......2.......3.......4.......5.......6.......7.......8.......9.......10.....
PBMSECT 32 OP +
OUTP=101,BRP(1)=102,CORE=204, +
layer=(210,101),L(2)=(210,103) +
PCOMP 204 -0.254
501 0.127 -15.0 501 0.127 -15.0
501 0.127 -15.0 501 0.127 -15.0
PCOMP 210 0.0
501 0.127 -15.0
point 2 11.724 6.349
point 3 11.724 -6.349
point 4 -11.724 -6.349
point 5 -11.724 6.349
SET1 101 2 thru 5
SET1 102 5 2
SET1 103 -5 -4 -3 -2
$
pcomp 101 -0.1
501 0.05 0.0 501 0.05 90.0
501 0.05 -45.0 501 0.05 45.0
501 0.05 0.0
PSHELL 10101 501 0.000
PSHELL 10102 501 90.000
$PSHELL 10103 501 -45.000
PSHELL 10103 501 135.000
PSHELL 10104 501 45.000
PSHELL 10105 501 0.000
MD Nastran R2.1 Release Guide

238
9. Use only matching .bdf and .op2 files from the same job to visualize the stress pattern.
Limitations
1. Although CBEAM3 has curved beam capability, CBEAM3 must be straight when used as a
composite beam.
2. TEMPRB is not supported with the current implementation of VAM.
3. SOL 200 does not yet support the composite beam.
Example
A simple file, vabcore1, with two PBMSECT entries is used here to demonstrate the features
implemented. A bracket used in auto industry is modeled with PBMSECT,31 with isotropic material, and
PBMSECT,32 with composite material. Some key Bulk Data entries are shown as follows:
$
$
param,arbmstyp,timoshen
param,arbmss ,yes
.
.
$ Elements and Element Properties for region : prop2
CBEAM3 2 32 302 102 1301 1204
.
.
$
$<<<rocker at b pillar>>>
$ U profile
point 1 0.8 14.5
point 2 0.9 13.5
point 3 1.0 13.0
point 4 1.1 12.5
point 5 1.5 12.0
point 6 1.75 11.5
point 7 2.6 10.9
point 8 2.7 10.0
point 9 2.6 9.0
point 10 2.4 8.0
point 11 1.9 6.0
point 12 1.1 4.0
point 13 0.1 1.4
point 14 3.0 1.8
point 15 4.5 2.0
point 16 6.1 2.2
point 17 6.7 5.6
point 18 7.4 9.56
point 19 7.35 10.4
point 20 7.7 13.0
point 21 7.6 15.4
SET1 101 1 thru 21
$
$ center branch
$15 point 31 4.5 2.0
point 32 5.5 2.8
239
CHAPTER 7
Elements and Connectors
point 33 6.8 9.6
point 34 6.6 15.1
SET1 103 15 32 33 34
$
$ Isotropic case
PBMSECT 31 1 OP 0.015
OUTP=101,T=0.1,brp=103
$
$ Composite case
PBMSECT 32 1 OP 0.015
OUTP=101,C=101,brp=103,c(1)=[201,pt=(15,34)]
pcomp 101 -0.1
501 0.05 0.0 501 0.05 90.0
501 0.05 -45.0 501 0.05 45.0
501 0.05 0.0
pcomp 201 SYM
501 0.05 -45.0 501 0.05 45.0
501 0.05 0.0
MAT1 501 3.+6 .3
.
This model uses the default values for PARAM,ARBMPS and PARAM,ARBMFEM. The finite element
model for the composite beam of PBMSECT,32 is shown as follows:
The properties of composite PBMSECT,32 computed as a 6x6 D matrix. A PBEAM3 is created
internally and printed in .f06 if requested. The equivalent PBEAM3 for PBMSECT,32 is shown as
follows:
MD Nastran R2.1 Release Guide

240
The 6x6 D matrix starts in the Wi(j) section of PBEAM3 (the seventh line of the previously described
PBEAM3). Note that the previously described PBEAM3 is for information only. It does not have a
punched version for later use.
The stress output for the composite CBEAM3, if PARAM,ARBMSS,YES is specified, is shown as
follows:
References
1. R. Palacios, Asymptotic Models of Integrally-Strained Slender Structures for High-Fidelity
Nonlinear Aeroelastic Analysis, Ph.D. Dissertation, 2005, Department of Aerospace
Engineering at University of Michigan, Ann Arbor, Michigan.
2. C.E.S. Cesnik, R. Palacios, UM/VABS Theoretical Manual, Release 1.20, May 2004, University
of Michigan.
*** USER INFORMATION MESSAGE 4379 (IFP9B)
THE USER SUPPLIED PBMSECT BULK DATA ENTRIES ARE REPLACED BY THE FOLLOWING PBEAM3 ENTRIES.
CONVERSION METHOD FOR PBARL/PBEAML - .
PBEAM3 32 0 4.7202E+00 8.3059E+01 2.9578E+01 -1.5664E+01 3.2316E+01 0.0000E+00
1.8014E+01 4.2136E+00 1.7100E+01 -2.7858E+00 3.8881E+00 -3.5404E+00 4.7202E+00 2.6994E+00
0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00
0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00
0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00 0.0000E+00
0.0000E+00 0.0000E+00
1.2253E+08 -2.1160E+05 8.1193E+04 -2.4761E+06 -3.7193E+06 7.9040E+06 -2.1160E+05 2.1792E+06
-1.7859E+06 1.9780E+07 5.4643E+05 -3.5845E+05 8.1193E+04 -1.7859E+06 2.7228E+07 1.7190E+07
2.9835E+04 2.1407E+06 -2.4761E+06 1.9780E+07 1.7190E+07 2.2332E+08 5.8182E+06 -1.2186E+06
-3.7193E+06 5.4643E+05 2.9835E+04 5.8182E+06 2.1349E+09 -4.0706E+08 7.9040E+06 -3.5845E+05
2.1407E+06 -1.2186E+06 -4.0706E+08 7.5602E+08
1 VAB ALGORITHM USING CORE ON PBMSECT JANUARY 8, 2007 MD NASTRAN 1/ 8/07 PAGE 819
TRANSVERSE TIP LOAD
0 SUBCASE 1

S T R E S S E S I N L A Y E R E D C O M P O S I T E E L E M E N T S ( B E A M 3 )
ELEMENT GRID PLY D I R E C T S T R E S S E S FAILURE MAXIMUM STRENGTH
ID ID ID NORMAL-1 NORMAL-2 NORMAL-3 SHEAR-12 SHEAR-23 SHEAR-13 THEORY FAIL. INDEX RATIO FLAG
2 302 2 2.468E+01 1.601E+01 2.570E+00 2.323E+01 4.991E-01 3.724E+00 TSAI-WU 7.161E-04 4.035E+02
102 2 1.685E+01 1.610E+01 -7.230E-01 1.993E+01 -1.377E-01 -5.572E-01 TSAI-WU 7.258E-04 4.470E+02
1301 2 1.558E+01 1.594E+01 -7.167E-01 1.938E+01 -1.162E-01 -5.280E-01 TSAI-WU 7.193E-04 4.569E+02
1 VAB ALGORITHM USING CORE ON PBMSECT JANUARY 8, 2007 MD NASTRAN 1/ 8/07 PAGE 820
LATERAL TIP LOAD
0 SUBCASE 2

S T R E S S E S I N L A Y E R E D C O M P O S I T E E L E M E N T S ( B E A M 3 )
ELEMENT GRID PLY D I R E C T S T R E S S E S FAILURE MAXIMUM STRENGTH
ID ID ID NORMAL-1 NORMAL-2 NORMAL-3 SHEAR-12 SHEAR-23 SHEAR-13 THEORY FAIL. INDEX RATIO FLAG
2 302 2 1.206E+02 1.181E+02 -4.676E+00 1.455E+02 -1.072E+00 -7.170E+00 TSAI-WU 5.479E-03 6.114E+01
102 2 1.242E+02 1.185E+02 -4.690E+00 1.473E+02 -1.119E+00 -7.235E+00 TSAI-WU 5.500E-03 6.059E+01
1301 2 1.224E+02 1.183E+02 -4.683E+00 1.464E+02 -1.096E+00 -7.202E+00 TSAI-WU 5.490E-03 6.086E+01
241
CHAPTER 7
Elements and Connectors
Nonhomogeneous Multipoint Constraint
Introduction to MPCY and MPCD Entries
Previously, if a nonhomogeneous multipoint constraint was desired, then the method of SLACK
variable was required. That is, the nonhomogeneous right hand side of the MPC equation was written
using a scalar or grid point, with an SPC or SPCD specifying the right hand side value. In this release,
an MPCY Bulk Data entry is introduced that allows the user to enter a right hand side value directly into
the MPC equation. The entry defines an equation of the form
See the description of the MPC Bulk Data entry in the MD Nastran Quick Reference Guide for a
description of the terms on the left hand side of the preceding equation.
Associated with the MPCY entry is another new Bulk Data entry, MPCD, used to define a load-
selectable value of for a nonhomogeneous multipoint constraint.
Benefits
The user can define a nonhomogeneous multipoint constraint directly.
Input
1. The MPCY Bulk Data entry is used to define a nonhomogeneous multipoint constraint directly.
This entry can also be used to define a standard homogenous multipoint constraint.
2. The MPCD Bulk Data entry is used to define a load selectable value for in a nonhomogeneous
multipoint constraint.
Output
Standard MD Nastran MPCF output is available.
Guidelines and Limitations
Currently, this method is not supported in the dynamic solution sequences because the dynamic load
generation modules have not yet been taught to automatically expand into the required number of
load columns.
A
m
u
m
A
i
u
i
i

+ Y
m
=
Y
m
Y
m
Y
m
MD Nastran R2.1 Release Guide

242
Theory
The unconstrained stiffness matrix equation in MD Nastran is
(7-1)
If we apply a nonhomogeneous MPC constraint
(7-2)
then Eq. (7-1) becomes
(7-3)
where and are the forces of constraint necessary to impose the conditions described in Eq. (7-2).
We partition Eq. (7-2) as
(7-4)
and solve for to obtain
(7-5)
where and .
Substitution of Eq. (7-5) into Eq. (7-3) yields
(7-6)
or
(7-7)
K
gg
[ ] U
g
{ } P
g
{ } =
R
g
[ ] U
g
{ } Y

{ } =
K
NN
K
NM
K
MN
K
MM
U
N
U
M
)

`


P
N
Q
N
+
P
M
Q
M
+
)

`


=
Q
N
Q
M
R
N
R
M
U
N
U
M
)

`


Y

{ } =
U
M
U
M
{ }
M
G
MN
U
N
+ { } =

M
R
M
1
Y

= G
MN
R
M
1
R
N
=
K
MN
K
MM
G
MN
+ [ ] U
N
{ } P
M
Q
M
K
MM

M
+ { } =
Q
M
{ } K
MN
K
MM
G
MN
+ [ ] U
N
{ } P
M
{ } = K
MM

M
{ } +

current new
243
CHAPTER 7
Elements and Connectors
Since the terms of Eq. (7-5) are prescribed, any virtual variation of Eq. (7-5) results, by definition, in
. Hence, the standard conjugate force transformation holds.
Or,
(7-8)
where the minus sign comes from the fact that we are imposing constraints, and the forces are
reactive to the constraint forces.
Combining the first row of Eq. (7-3) with Eq. (7-8) yields
and using Eq. (7-7) to expand the result, we obtain
or
(7-9)
Examples
As a simple example, consider a horizontal rod structure. The left and right ends are clamped. However,
as Listing 7-1 shows (see diagram and rod model in the Bulk Data listing), there is a break in the structure
of 0.1 units of length. If the grid point ID at the left end of the break is A, and the grid point ID at the
right end of the break is B, we can tie the structure by the MPC equation
In the example, we show how to write the MPC equation in two ways. The first way uses the standard
SLACK variable method. The second way uses the new MPCY entry.
Listing 7-1 Example Rod Structure for MPCY
SOL 101
CEND
TITLE = DEMONSTRATE USE OF MPCY
SPC = 300
MPC = 300

M
0 =
Q
N
{ } G
MN
T
Q
M

)
`

=
Q
N
Q
M
K
NN
U
N
K
NM
G
MN
U
N

M
+ [ ] + { } P
N
G
MN
T
Q
M

)
`

=
K
NN
K
NM
G
MN
G
MN
T
K
MN
G
MN
T
K
MM
G
MN
+ + + [ ] U
N
{ } P
N
G
MN
T
P
M
+ { } = K
NM
G
MN
T
K
MM
[ ]
M
{ }
K
NN
[ ] U
N
{ } P
N
{ } =

current
K
NM
G
MN
T
K
MM
+ [ ]
M
{ }

new
U
B
U
A
0.1 =
MD Nastran R2.1 Release Guide

244
LOAD= 300
DISPL= ALL
OLOAD= ALL
MPCFO = ALL
SPCFO= ALL
ELFOR = ALL
BEGIN BULK
$
PROD 1 1 1.
MAT1 1 1.+7 0.
GRDSET 3456
$
$
$ STANDARD SLACK VARIABLE METHOD
$
$ |1 2 3 4 5|
$ |O---------------O---------------O O---------------O|
$ | 1 2 3 |
$
$ GRIDs 3 and 4 have a 0.1 unit gap between them
$ We wish to impose relationship in x-direction
$ U4 = U3 + 0.1
$ When GRID 3 has a x-direction load of 1.+5
$
GRID 1 0. 0. 0. 123456
GRID 1 100. 0. 0. 23456
GRID 1 200. 0. 0. 23456
GRID 1 200.1 0. 0. 23456
GRID 1 300.1 0. 0. 123456
$
SPOINT 6
$
CROD 1 1 1 2
CROD 2 1 2 3
CROD 3 1 4 5
$
FORCE 300 3 1.+5 1.
$
SPC 300 6 1 0.1
MPC 300 4 1 -1. 3 1 1.
6 1 1.
$ ^
$ |
$ |
$ |
$ SPOINT with 0.1 in displacement gap
$
$
$ ----------------> NEW MPCY METHOD <----------------
$
$ |11 12 13 14 15|
$ |O---------------O---------------O O---------------O|
$ | 11 12 13 |
$
$ GRIDs 13 and 14 have a 0.1 unit gap between them
$ We wish to impose relationship in x-direction
$ U14 = U13 + 0.1
$ When GRID 13 has a x-direction load of 1.+5
$
GRID 11 0. 1. 0. 123456
245
CHAPTER 7
Elements and Connectors
GRID 12 100. 1. 0. 23456
GRID 13 200. 1. 0. 23456
GRID 14 200.1 1. 0. 23456
GRID 15 300.1 1. 0. 123456
$
CROD 11 1 11 12
CROD 12 1 12 13
CROD 13 1 14 15
$
FORCE 300 13 1.+5 1.
$
MPCY 300 14 1 1. 0.1
13 1 -1.
ENDDATA
The results are:
D I S P L A C E M E N T V E C T O R

POINT ID. TYPE T1 T2 T3 R1 R2 R3
1 G 0.0 0.0 0.0 0.0 0.0 0.0
2 G 3.000000E-01 0.0 0.0 0.0 0.0 0.0
3 G 6.000000E-01 0.0 0.0 0.0 0.0 0.0
4 G 7.000000E-01 0.0 0.0 0.0 0.0 0.0
5 G 0.0 0.0 0.0 0.0 0.0 0.0
6 S 1.000000E-01
11 G 0.0 0.0 0.0 0.0 0.0 0.0
12 G 3.000000E-01 0.0 0.0 0.0 0.0 0.0
13 G 6.000000E-01 0.0 0.0 0.0 0.0 0.0
14 G 7.000000E-01 0.0 0.0 0.0 0.0 0.0
15 G 0.0 0.0 0.0 0.0 0.0 0.0

L O A D V E C T O R

POINT ID. TYPE T1 T2 T3 R1 R2 R3
3 G 1.000000E+05 0.0 0.0 0.0 0.0 0.0
6 S 0.0
13 G 1.000000E+05 0.0 0.0 0.0 0.0 0.0
F O R C E S O F S I N G L E - P O I N T C O N S T R A I N T

POINT ID. TYPE T1 T2 T3 R1 R2 R3
1 G -3.000000E+04 0.0 0.0 0.0 0.0 0.0
5 G -7.000000E+04 0.0 0.0 0.0 0.0 0.0
6 S 7.000000E+04
11 G -3.000000E+04 0.0 0.0 0.0 0.0 0.0
15 G -7.000000E+04 0.0 0.0 0.0 0.0 0.0

F O R C E S I N R O D E L E M E N T S ( C R O D )
ELEMENT AXIAL ELEMENT AXIAL
ID. FORCE TORQUE ID. FORCE TORQUE
1 3.000000E+04 0.0 2 3.000000E+04 0.0
3 -7.000000E+04 0.0 11 3.000000E+04 0.0
12 3.000000E+04 0.0 13 -7.000000E+04 0.0

F O R C E S O F M U L T I P O I N T C O N S T R A I N T

POINT ID. TYPE T1 T2 T3 R1 R2 R3
3 G -7.000000E+04 0.0 0.0 0.0 0.0 0.0
4 G 7.000000E+04 0.0 0.0 0.0 0.0 0.0
6 S -7.000000E+04
13 G -7.000000E+04 0.0 0.0 0.0 0.0 0.0
14 G 7.000000E+04 0.0 0.0 0.0 0.0 0.0

* * * END OF JOB * * *
MD Nastran R2.1 Release Guide

246
Ch. 8: Optimization MD Nastran R2 Release Guide

8
Optimization
!
Topology Optimization Enhancements
!
Automatic External Superelement Optimization (AESO)
!
Randomization of an Input Data File (Prerelease)
!
Random Elimination of Element Types (Prerelease)
MD Nastran R2.1 Release Guide

248
Topology Optimization Enhancements
Introduction
Topology optimization capability was first released in MD Nastran 2005 R2, and a number of
manufacturability constraints were added to MD Nastran R1. New features have been added to this
release based on feedback from clients. With these enhancements, MD Nastran R2 SOL 200 is able to
support combined topology, sizing and shape optimization, multiple mass reduction targets, cyclical
symmetry constraints, and adjoint design sensitivity analysis for inertia relief of static analyses. A major
performance enhancement of module DOPR1 has been made to speed up minimum member size control
and sizing optimization with many thousands of variables.
Benefits
Combined Topology, Sizing, and Shape Optimization
It is often recommend that topology optimization first be used to find efficient design concepts at the
early design stage, whereas sizing and/or shape optimization is used for detail design based on the
topology design proposals at a later design stage. The use of topology, sizing, and shape optimization
simultaneously may find a potentially better design, since the interaction of sizing and/or shape variables
with topology optimization is considered during a single design optimization process. Another benefit of
this feature is that the DRESP2 BETA function (which minimizes the maximum responses) is now
available in topology optimization.
Cyclical Symmetry Constraints
A mirror symmetry constraint was added to MD Nastran R1 to enable the design of a rotational
symmetric component or system. By using cyclical symmetry constraints in topology optimization, a
rotational symmetric design can be obtained regardless of the boundary conditions or loads. This cyclical
symmetric constraint capability can be used for irregular finite element meshes.
Adjoint Design Sensitivity Support for Inertia Relief
A direct design sensitivity analysis method is only available for inertia relief of static analyses in previous
versions. The direct method is not computationally affordable for topology optimization, since many
thousands of variables are often involved. The adjoint design sensitivity analysis method has been
developed to benefit not only inertia relief topology optimization, but also inertia relief sizing
optimization with many design variables. The adjoint method is automatically selected when it is more
efficient computationally.
Multiple Mass Target
Type one response FRMASS (DRESP1=FRMASS) was previously the total fraction mass of topological
designed properties. User feedback showed that it is desirable to set up different mass reduction targets
on multiple designed parts for built-up structures. This enhancement reflects this requirement.
249
CHAPTER 8
Optimization
Module DOPR1 Performance Enhancements
When a large value is given for minimum member size TDMIN, significant CPU time was spent on
module DOPR1 in previous versions. With this enhancement, a substantial performance speedup (3-100
times) is achieved for minimum member size control. Module DOPR1 is also enhanced to efficiently
support optimization problems with many thousands of sizing design variables (for example, for a sizing
optimization file with 320,000 design variables, the DOPR1 module in previous MD Nastran versions
and other Nastran products required 16,000 seconds, whereas the MD Nastran R2 DOPR1 module
requires only 10 seconds).
Input
The TOPVAR Bulk Data entry has been enhanced to provide cyclical symmetry constraints. To select a
topologically designable region, the user needs to specify a group of elements using the TOPVAR Bulk
Data entry. The cyclical symmetry constraints are then applied on all elements referencing the given
property on the TOPVAR entry.
The enhanced TOPVAR format is specified as follows:
Remarks:
1. The topologically designable element properties include PROD, PBAR, PBARL, PBEND,
PBEAM, PBEAML, PSHELL, PSHEAR, PSOLID, and PWELD.
2. Multiple TOPVARs are allowed in a single file.
1 2 3 4 5 6 7 8 9 10
TOPVAR ID LABEL PTYPE XINIT XLB DELXV POWER PID
SYM CID MSi MSi MSi CS NCS
CAST CID DDi DIE
EXT: CID EDi
TDMIN TV
Field Contents
PID Property entry identifier (Integer > 0). This PID must be unique for PIDs
referenced by other TOPVAR, DVPREL1, and DVPREL2 entries. Topology
and sizing variables cannot share the same properties. (Integer > 0)
SYM Indicates that this line defines symmetry constraints.
CID Rectangular coordinate system ID used for specifying manufacturing
constraints. See Remark 5. (Blank or Integer > 0; Default = 0)
CS Cyclical symmetry axis (character X, Y, Z). See Remark 6.
NCS Number of cyclical symmetric segments in 360 degrees. (Integer > 0). The
angle for one segment is calculated by 360/NCS. See Remark 6.
MD Nastran R2.1 Release Guide

250
3. Combined topology, sizing, and shape optimization is supported in a single file.
4. TOPVAR cannot be used with DVMREL1 and DVMREL2 entries.
5. Only CORD1R and CORD2R can be used as referenced coordinate systems to specify topology
manufacturing constraints. Only one reference coordinate system CID is allowed for each
TOPVAR entry.
6. The first cyclical symmetry segment starts at the X-axis when CS = Z (at the Z-axis when CS = Y,
and at the Y-axis when CS = X). One cyclical symmetry constraint can be combined with one
mirror symmetry constraint if the axis of cyclic symmetry is normal to the plane of mirror
symmetry. For example, MSi = YZ and CS = X; MSi = XZ and CS = Y; and MSi = XY and
CS = Z. This feature can also be used for rotational parts that span less than 360 degrees, but NCS
must be specified as if the part spanned 360 degrees. For example, if three segments are desired
for a 60 degree part then NCS = 18.
7. For EXT constraints, possible combinations are (ED=X, MSi=XY, and/or ZX or CS=X),
(ED=Y, MSi=YZ, and/or XY or CS=Y), (ED=Z, MSi=ZX, and/or YZ or CS=Z).
8. For CAST constraints, possible combinations are (DD=X or X-, MSi=XY and/or ZX or CS=X),
(DD=Y or Y-, MSi=YZ and/or XY or CS=Y), (DD=Z or Z-, MSi=ZX and/or YZ or CS=Z).
Modified Type One Responses Fractional Mass
To allow the user to put different mass reduction constraints on multiple designed properties, the
RTYPE=FRMASS entry has been extended to provide property ID (PID) at attribute field ATTi. The
format for this extension is shown in Table 8-1.
Remark:
1. RTYPE= FRMASS (mass fraction of topological designed elements) entries are used for
topology optimization or combined topology, sizing, and/or shape optimization. ATTi=Blank is
for the total mass fraction of all topologically designed properties. ATTi=PID is the mass fraction
of the topologically designed property PID.
Guidelines and Limitations
1. Although combined topology and sizing optimization is supported, TOPVAR, DVPREL1, and
DVPREL2 entries cannot reference the same property ID (PID).
Table 8-1 Modified RTYPE=FRMASS
Response Type
(RTYPE)
Response Attributes
ATTA (Integer >
0)
ATTB (Integer > 0 or
Real > 0.0) ATTI (Integer > 0)
FRMASS
Remark 1
BLANK BLANK BLANK or Property
ID (PID)
251
CHAPTER 8
Optimization
2. Whereas FRMASS is calculated for topologically designed properties only, RTYPE=WEIGHT
computes total weight including all designed and nondesigned parts. For combined topology and
sizing/shape optimization problems, it is recommended that RTYPE=FRMASS be used for
topologically designed property mass reduction constraints, and RTYPE=WEIGHT be used for
total mass reduction constraints.
Example 1 Wheel (wheeltop.dat)
The wheel model shown in Figure 8-1 is used to demonstrate MD Nastran topology optimization cyclical
symmetry capabilities. The wheel is modeled with six-sided solid elements (CHEXAs). The wheel outer
layers and bolts are nondesignable. One load case is considered. The structural compliance is minimized
(i.e., minimize the total strain energy of the structure) with a mass target 0.1 (i.e., remove 90% of the
material). Although the load is not cyclically symmetric about the Y-axis, the design is required to be
cyclically symmetric about the Y-axis with five segments.
Figure 8-1 Wheel example.
Input
The input data for this example that is related to the topology optimization model is given in Listing 8-1.
The coordinate system CORD2R = 1 is created to specify cyclical symmetric constraints. The field CS
on the SYM line is the Y-axis with NCS=5.
MD Nastran R2.1 Release Guide

252
Listing 8-1 Input File for Example 1
DESOBJ = 10
DESGLB = 1
ANALYSIS = STATICS
SMETHOD = ELEMENT
SUBCASE 1
SPC = 2
LOAD = 2
BEGIN BULK
CORD2R 1 10.512 33.3312 12.9921 -22.209833.3312 4.88385
28.388 33.3313 -19.7297
DCONSTR 1 2 .1
TOPVAR 1 PSOLID PSOLID .1 2
SYM 1 Y 5
DRESP1 2 FRM FRMASS
DRESP1 10 COMP COMP
Output
Figure 8-2 shows the topology-optimized result that is smoothed by using MD Patran. Note that cyclical
symmetry is obtained even though the loading is not cyclically symmetric.
253
CHAPTER 8
Optimization
Figure 8-2 Wheel topology design.
MD Nastran R2.1 Release Guide

254
Automatic External Superelement Optimization
(AESO)
Introduction
The manual external superelement optimization (MESO) capability has been available since the release
of MSC.Nastran 2004. In this technique, the user manually partitions the analysis model into two parts:
a designed part and a nondesigned part. The latter is treated as one external part superelement, and the
former is defined as a residual structure. A creation run is performed which applies component mode
synthesis (CMS) or static condensation to the part superelement, and stores the resulting boundary
matrices in a database or a punch file. The original optimization task is performed on the assembly run
that assembles those boundary matrices into the residual model for solving system solutions. The strategy
is most efficient when the size of the design model (or residual model) is much smaller than the size of
the original analysis model. However, although the feature is efficient in terms of CPU time, since both
files of both creation and assembly runs must be created by the user, significant effort in manual
partitioning the model might outweigh the performance gain.
The AESO capability presented here extends MESO in an important way: rather than requiring the user
to segregate a large model into a designed and nondesigned part, the process does this automatically by
identifying which parts of the finite element model are affected by the design task. In essence, the new
AESO capability provides an efficient and accurate solution in a user-friendly way.
Benefits
The new AESO capability:
1. Eliminates a tedious and error-prone task by preparing the input data.
2. Does not require the user to be knowledgeable in the specialized area of superelements in general,
and external superelements in particular.
3. Provides an efficient and accurate approach for large-scale design optimization tasks.
4. Enables the performance of various design studies rapidly once the model has been divided into
designed and nondesigned parts. Examples of this are the setting up of different design constraints
and objectives in the studies to gain insight into the design and the available tradeoffs, and the
application of various frequency excitation loadings in frequency response analysis.
Methodology
A complete AESO task involves two separate MD Nastran runs. The first run is an AESO creation run
(or simply creation run) whose logical flow is described in Figure 7-3. The second run is an AESO
assembly run (or simply assembly run) whose logical flow is described in Figure 7-5.
As shown in Figure 7-3, the creation run automatically partitions the original analysis model into the
residual (the designed part) and external SE (the nondesigned part). This automatic partition procedure
will assign the following grid points to the residual:
255
CHAPTER 8
Optimization
1. All grid points that belong to a design model consisting of DRESP1, DVGRID, DVPRELi,
DVMRELi, and DVCRELi entries;
2. All grid points that are referenced on all static or dynamic loading entries such as DAREA,
DPHASE, FORCE, MOMENT, PLOADi, and TEMP entries;
3. All grid points for rigid elements that have one or more connecting grid point(s) belonging to the
residual; and
4. Any grid point that is a dependent grid on an MPC entry.
After the automatic partition procedure, a new user input file is created from the residual (Figure 8-4).
Then, the remainder of the creation run is the application of static condensation and/or component modes
synthesis procedures to produce stiffness, mass, and damping boundary matrices. After the creation run
is complete, a Nastran database is saved to store the boundary matrices and an .asm file is created to
include superelement boundary connection information.
The assembly run is similar to a conventional SOL 200 task, as shown in Figure 8-5, by using all three
types of data generated from the creation run (Figure 8-4). The original optimization problem is solved
by assembling the boundary matrices into the residual for the system solutions.
Input
1. Two parameters are added on the DOPTPRM entry for an AESO job:
AUTOSE flag to request an AESO job (Integer 0, 1, Default = 0). AUTOSE = 1 activates an
AESO creation run.
Optimizer
Automatic
Model Partition
External SE
CMS to create
boundary matrices
Residual
MSC.Nastran
EXTSE
Database
.asm file
New Input File
for
Assembly Run

Initial User
Input File
Initial User
Input File
Figure 8-3
AESO creation run.
Figure 8-4
Output of creation
run and input of
assembly run.
Figure 8-5
AESO assembly run.
MD Nastran R2.1 Release Guide

256
DRATIO the threshold value that is used to turn off an active AESO job if the ratio of the size
of the design model to that of the analysis model is greater than DRATIO (Real > 0.0; Default =
0.1).
2. An ASSIGN statement with a logical key name AESO is placed in the File Management
Section to specify an input file name for the assembly run.
To illustrate ideas behind the input and output for an AESO task, a test problem (aeso1.dat) is used here.
Figure 8-6 shows a sample model whose upper left portion covering elements 18 to 42 (SE 1) is the
nondesigned part, and the rest of the structure is the residual structure.
The following listing is a condensed version of the creation run file (aeso1.dat) that only shows the
required user input to invoke an AESO creation run. The ASSIGN AESO=AESOL_2.DAT statement is
specified in the FMS Section, and AUTOSE,1 and DRATIO,0.9 are specified on the DOPTPRM entry.
Here, the DRATIO=0.9 specification overrides the default.

Figure 8-6 A sample model.
SE1
257
CHAPTER 8
Optimization
Figure 8-7 Condensed version of the creation run file, dseoptl.dat.
Outputs
As shown in Figure 8-4, three types of data generated from the creation run are saved in the working
directory: a Nastran database, an .asm file, and a new input file for the assembly run (or an assembly
file). This section describes each of these items and explains how they are used in the assembly run. In
addition, some special print outputs from the creation run are shown that display the model partition
information.
Nastran Database Files
Two Nastran database files, aeso1.MASTER and aeso1.DBALL, are automatically saved when the
creation run is submitted with the SCR=NO option. Notice that the size of the database is much smaller
than the regular Nastran database because only boundary matrices are stored. The master file will be
referenced by an ASSIGN statement in the new input file described as follows.
The .asm File
The aeso1.asm file that includes the boundary connection information is described as follows. As shown
in Figure 8-6, this problem has only two boundary grid points, 35 and 36, between the residual and the
external superelement 1. This file is accessed through an INCLUDE command in the assembly file.
assign aeso='aeso1_2.dat'
SOL 200
CEND
desobj(max) = 1
analysis = modes
..
Begin Bulk

doptprm desmax 10 autose 1 dratio 0.9


enddata
MD Nastran R2.1 Release Guide

258
New Input File for the Assembly Run
Notice that the name of this file, aeso1_2.dat, is specified on the ASSIGN AESO statement in the creation
run (Figure 8-7). It is a standard MD Nastran input file.
Figure 8-8 Highlights of the assembly run file, aeso1_2.dat.
The AESO-specific contents in the Executive Control, Case Control, and Bulk Data Sections are listed
in Figure 8-8 and are described as follows.
1. Executive Control Section
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$
$ ASSEMBLY PUNCH (.ASM) FILE FOR EXTERNAL SUPERELEMENT 1
$
$ THIS FILE CONTAINING BULK DATA ENTRIES PERTAINING TO
$ EXTERNAL SUPERELEMENT 1 IS MEANT FOR INCLUSION
$ ANYWHERE IN THE MAIN BULK DATA PORTION OF THE ASSEMBLY RUN
$
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
SEBULK 1EXTERNAL MANUAL
$
SECONCT 1 0 5.0E-05
35 35 36 36
$
$ BOUNDARY GRID DATA
$
GRID 35 -3.6 6. 0.
GRID 36 -2.8 6. 0.
$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
nastran rseqcont=1
assign se1= './aeso1.MASTER'
dblocate datablk(EXTDB) logical=se1,
CONVERT(SEID=1)
SOL 200
CEND
Executive Control section
desobj(max) = 1
analysis = modes
.
subcase 10
method = 1
spc = 10
$
Case Control section
begin bulk
include './aeso1.asm'
.
doptprm desmax 10
enddata
Bulk Data section
}
}
}
259
CHAPTER 8
Optimization
The Nastran RSEQCONT=1 statement instructs the input file processor to ignore all continuation
fields. This statement is automatically created in this file regardless of whether the creation run
has it or not.
The next two statements assign the Nastran Master database file, and locates the EXTDB
datablock that stores various boundary matrices.
Notice that the other statements in Executive Control Section of the creation run are not retained.
2. Case Control Section
The entire Case Control Section of the creation run is retained in the assembly file.
3. Bulk Data Section
This section completely defines the residual structure. The INCLUDE./aeso1.asm
command allows the assembly run to access the .asm file created from the creation run. In
addition, the AUTOSE 1 and DRATIO 0.9 specifications have been removed from the
DOPTPRM entry of the creation run.
Special Print Outputs from the Creation Run
The following output is taken from the aeso1.f06 file. It displays detailed information about the model
partition. You may use Figure 8-6 to help read the printout here. Notice that superelement 1 covers the
nondesigned part, and the residual (or superelement 0) covers the designed part.
MD Nastran R2.1 Release Guide

260
Listing 8-2 Printout showing model partition of designed and nondesigned parts.
Guidelines and Limitations
1. DRATIO may be adjusted to allow an assembly run with a larger or smaller residual model. UIM
7824 provides summary information about the sizes of the analysis and design models in terms of
the number of the grid points.
BOUNDARY SEQUENCE ASSIGNMENT TABLE
BOUNDARY
SEQUENCE ID -------------------------- ASSIGNED TO POINT ID (SUPERELEMENT) ---------------------------------
1B 35 ( 0) 35 ( 1)
2B 36 ( 0) 36 ( 1)
SUPERELEMENT 0
LIST OF INTERIOR POINTS ( TOTAL NO. OF INTERIOR POINT = 70 )
INDEX -1- -2- -3- -4- -5- -6- -7- -8- -9- -10-
1 1 2 3 4 5 6 7 8 9 10
11 11 12 13 14 15 16 17 18 19 20
21 21 22 23 24 25 26 27 28 29 30
31 31 32 39 40 41 42 43 44 51 52
41 53 54 55 56 63 64 65 66 67 68
51 75 76 77 78 79 80 87 88 89 90
61 91 92 99 100 101 102 103 104 1B 2B
SUPERELEMENT 0
LIST OF ELEMENTS ( TOTAL NO. OF ELEMENTS = 42 )
INDEX -1- -2- -3- -4- -5- -6- -7- -8- -9- -10-
1 1 2 3 4 5 6 7 8 9 10
11 11 12 13 14 15 16 17 43 44 45
21 46 47 48 49 50 51 52 53 54 55
31 56 57 58 59 60 61 62 63 64 65
41 66 67
SUPERELEMENT 1
LIST OF EXTERIOR POINTS ( TOTAL NO. OF EXTERIOR POINT = 2 )
INDEX -1- -2- -3- -4- -5- -6- -7- -8- -9- -10-
1 1B 2B
SUPERELEMENT 1
LIST OF INTERIOR POINTS ( TOTAL NO. OF INTERIOR POINT = 34 )
INDEX -1- -2- -3- -4- -5- -6- -7- -8- -9- -10-
1 33 34 37 38 45 46 47 48 49 50
11 57 58 59 60 61 62 69 70 71 72
21 73 74 81 82 83 84 85 86 93 94
31 95 96 97 98
SUPERELEMENT 1
LIST OF ELEMENTS ( TOTAL NO. OF ELEMENTS = 25 )
INDEX -1- -2- -3- -4- -5- -6- -7- -8- -9- -10-
1 18 19 20 21 22 23 24 25 26 27
11 28 29 30 31 32 33 34 35 36 37
21 38 39 40 41 42
*** USER INFORMATION MESSAGE 7824 (DSGRDM)
THE NUMBER OF GRID POINTS IN THE ANALYSIS MODEL = 104.
THE NUMBER OF GRID POINTS IN THE DESIGN MODEL = 70.
THE DESIGN MODEL COMPRISES 67.3 PERCENT OF THE ANALYSIS MODEL.
261
CHAPTER 8
Optimization
2. For an AESO job with ANALYSIS=MODES or MFREQ, it is recommended to activate the
matrix domain-based decomposition with DOMAINSOLVER ACMS(PARTOPT=DOF) in the
Executive Control Section to speed up the CMS procedure.
3. Always specify the ASSIGN AESO=filename.ext statement in the creation run to define the
name of the assembly file. Directly assigning the original job name to filename should be
avoided, and will cause the assembly run to fail with User Fatal Message 713. A good practice is
to add a suffix to the original filename such as myjob_2nd.dat, where myjob is the original
filename.
4. When submitting the AESO creation run, use the SCR=NO option. Otherwise, the Nastran
database will not be retained after the creation run is done. However, this is optional for
submitting an assembly run.
5. After the creation run is complete, check the following User Information Message in the .f06 file
to ensure that the job is terminated successfully:
6. The assembly input file may be modified to perform various design studies as long as the changes
do not affect the boundary matrices stored in the database.
7. If the AESO creation run includes a GRAV Bulk Data entry, it will be terminated with the
following message. The same applies to a TEMPD entry. They should be removed from the file
if they are temporarily inactive for the current task. Gravity and TEMPD loads are not supported
with AESO, since they apply loads to the entire structure and therefore block any partitioning.
8. Duplicate GRID entries are allowed in the .asm file and in the assembly input file. They will be
automatically removed during the assembly run within the location tolerance specified by the
TOL field on the SECONCT entry. For AESO jobs, the default of the location tolerance has been
increased to 5.E-5 from the original 1.E-5. However, due to numerical imperfection, this
tolerance may need to be adjusted, particularly for the cases in which the boundary grid points
are defined in one coordinate system or in multiple levels of coordinate systems.
*** USER FATAL MESSAGE 732 (OPFUNT)
LOGICAL NAMES 'INPUT ' AND 'AESO ' ARE ASSIGNED TO THE SAME PHYSICAL FILE.
USER INFORMATION: PHYSICAL FILE NAME 1: ./abc.aeso
PHYSICAL FILE NAME 2: ./abc.aeso
USER ACTION: CHANGE FILE NAME ON ONE OF THE ASSOCIATED ASSIGN STATEMENTS.
^^^
^^^ USER INFORMATION MESSAGE 9181 (FEA)
^^^ THE JOB IS TERMINATED FOR AN AUTO EXTERNAL CREATION RUN
^^^
*** USER FATAL MESSAGE 7699 (DSGRDM)
A GRAV Bulk Data entry is specified in an AESO creation run.
USER INFORMATION: The AESO run does not support the GRAV entry.
MD Nastran R2.1 Release Guide

262
9. Bulk Data parameter SEMAPPRT can be used to control the printout of the model partition
information shown in Listing 8-2. For example, setting SEMAPPRT to 0 will turn off the printout.
10. If the AESO task includes DVGRID entries, make sure that the grid points referenced by
DVGRID entries are inside the residual. Since the grid points on DVGRID entries vary during the
design process, including them as part of boundary grid points will invalidate the invariance of
those boundary matrices. Currently, the grid points on the DVGRID entries will not be
automatically assigned to be inside the residual. ASET and ASET1 entries can be used to create
an enclosure or a barrier to ensure the grid points referenced on the DVGRID entries are always
placed inside the residual.
11. Since all the CORD1i entries are automatically converted to CORD2i entries during the AESO
creation run, the DVGRID entry should not reference the grid points that define the CORD1i
entry.
12. The AESO tasks do not support acoustic response.
Examples
Road Response Optimization with CAMARO Model with Analysis=MFREQ
Figure 8-9 Camaro model.
A Camaro model provided by General Motors Corporation, shown in Figure 8-9, is modeled with 23,000
grid points, 37,000 elements, and 137,000 DOFs, with random inputs applied on left and right
suspension, including cross-correlation, to simulate road conditions.
The design task is to vary nine spring constants of engine mounts modeled by elastic elements in two
design cases:
Case A: Minimize the sum of RMS acceleration at the drivers seat and passengers seat, and limit the
PSD response at steering column; and
Case B: Minimize the RMS acceleration at the drivers seat and maintain frequency-dependent limits on
the drivers seat.
263
CHAPTER 8
Optimization
Because these design variables and responses are limited to a small part of the total vehicle, the size ratio
of the design model versus the analysis model is <1.% (the residual structure has 298 elements, 163
interior points, and 130 boundary points, and the external superelement has 33345 elements and 22761
interior points).
To demonstrate the efficiency of the AESO capability, Table 8-2 and Table 8-3 and present the results
from a regular run and the corresponding AESO run. First, the same final designs are achieved by both
jobs. However, the AESO job takes one-fifth of the time to complete Case A. Furthermore, since Case
B only changes the design objective and constraint formulation, the boundary matrices from the creation
run are invariant. Therefore, only the assembly run for Case B is required. Comparing the time spent on
this assembly run with the single run, as shown in Case B on the bottom half of the table, the increase in
speed is sixteen-fold.
Road Response Optimization with Camaro Model with Analysis=DFREQ
In general, the direct frequency approach is more expensive than the modal approach. However, when
the size of the residual model is small, the direct frequency approach can also be performed efficiently
as shown here. Case A of the same road response optimization problem is solved with both a regular
Table 8-2 Regular Run, Case A.
Case A
Initial
OBJ
Final
OBJ
Init Max
Const.
Final Max
Const.
# Design
Cycle
Clock Time
(Minutes)
Full model run 0.1534 -0.0639 0.1329 -0.2102 9 76
AESO creation
run
N/A N/A N/A N/A N/A 10
AESO assembly
run
0.1534 -0.0639 0.1329 -0.2102 9 7
Total time of two
AESO runs
16
Performance ratio 5
Table 8-3 AESO Run, Case B.
Case B
Initial
OBJ
Final
OBJ
Init Max
Const.
Final Max
Const.
# Design
Cycle
Clock Time
(Minutes)
Full model
run
0.0713 0.0586 0.2855 0.0081 14 110
AESO
creation run
N/A N/A N/A N/A N/A 0
AESO
assembly run
0.0713 0.0584 0.2855 -0.0108 14 7
MD Nastran R2.1 Release Guide

264
direct frequency response optimization and with the direct frequency AESO. The following table
presents the results. Again, the same final designs are achieved. However, the increase in speed obtained
by the AESO over the regular job is twenty seven-fold.
Figure 8-10 shows the plot of the total clock time spent by two runs versus design cycle, and it clearly
shows that the AESO is much more efficient than the regular run because the cost per design cycle for
AESO (or the slope of the curve for the AESO run) is much smaller than that for a regular run (the slope
of the curve for the regular run). Although, for this example, the one-time cost for the AESO due to the
creation run is smaller than that for a regular run, in general, this one-time cost may be larger because the
time spent by the CMS procedure on mass matrix reductions could be expensive.
Figure 8-10 Clocktime versus design cycle.
Case A
Initial
OBJ
Final
OBJ
Init Max
Const
Final Max
Const
# Design
Cycles
Clock Time
(Minute)
Full model run 0.1535 0.1327 -0.0631 -0.2073 9 388
AESO creation run N/A N/A N/A N/A N/A 10
AESO assembly run 0.1534 0.1327 -0.0636 -0.2062 9 5
Total time of two
AESO runs
15
Performance ratio 27
0
50
100
150
200
250
300
350
400
450
0 2 4 6 8 10
Design Cycle
C
l
o
c
k

T
i
m
e

(
M
i
n
u
t
e
s
)
Total Time (Regular Run) Total Time (AESO)
265
CHAPTER 8
Optimization
Miscellaneous
Prior to this release, SOL 200 supported a limited part optimization capability while it fully supported
traditional superelement optimization. Notice that traditional superelement optimization is defined with
a SESET Bulk Data entry, whereas limited part optimization is defined with a BEGIN SUPER Bulk Data
entry. In addition, the part may be defined either as external or internal.
The enhancement added in this release provides a robust part optimization capability as long as the
design model is in the residual (an upstream part cannot be designed or constrained). One example would
be an airplane wing modeled as an internal part, and the engine modeled as an external part. These two
components are then attached to the fuselage, which is the residual structure.
The table below summarizes the features of superelement optimization and part optimization in SOL 200
in this release.
To illustrate how to create a part optimization job, the sample model shown in Figure 8-6 is used here.
The top left squared region is modeled as Part 1 (elements 18 to 42) and the top right squared region is
modeled as Part 2 (elements 43 to 67). The remainder is modeled as a residual (elements 1 to 17). Grid
points 1 and 2 are fully constrained. The design task is to maximize the first natural frequency by varying
the thickness of elements 1 to 17 within a given range.
Listing 8-3 is a condensed version of the dseopt18.dat file that shows three major Bulk Data sections. (Its
full version can be obtained from the Test Problem Library.) The unique feature of a part optimization
job is the presence of multiple Bulk Data Sections, and the way in which the design model definition is
placed. The three sections in this job are:
1. Main section (between Begin Bulk and Begin Super = 1). This section defines the residual model.
Notice that all the designed entries (such as DESVAR, DVPREL1, and DRESP1) are defined in
this section because the design model must be within the residual.
2. Section for Part 1 (between BEGIN SUPER = 1 and BEGIN SUPER =2). This section defines the
Part 1 model.
3. Section for Part 2 (between BEGIN SUPER = 2 and ENDDATA). This section defines the Part
2 model.
Item Feature Support in SOL 200
Design Model in Upstream
SE or Part
1 Traditional SE Yes Yes
2 External Part Yes No
3 Internal Part Yes No
4 2+3 Yes No
1+2,1+3,1+2+3 No No
MD Nastran R2.1 Release Guide

266
Listing 8-3 Optimization Example.
SOL 200
diag 8,15,56
CEND
TITLE = Test for Internal Part Optimization
echo = sort
desobj(max) = 1
analysis = modes
subcase 10
spc = 10
method = 1
$
$ residual structure model
$
BEGIN BULK
eigrl,1,,,9
param,post,0
PARAM,GRDPNT,0
PARAM,WTMASS,.00259
CQUAD4 5 1 13 14 24 23
$
$GRDSET 6
GRID 13 -.4 3.6 0.
GRID 14 .4 3.6 0.
GRID 23 -.4 4.4 0.
GRID 24 .4 4.4 0.
$
MAT1,1,30.+6,,.3,.283
PSHELL,1,1,.05,1,,1
$
CQUAD4 14 1 19 20 30 29
CQUAD4 15 1 29 30 36 35
$$
spc1,10,123456,1,2
$
$ design model definition must be in the main BULK Data section
$ 2 3 4 5
desvar 1 X1 0.05 0.01 2.0
dvprel11 pshell 1 T
1 1.0
DRESP1 1 F1 FREQ 1
Main Bulk
Data Section
267
CHAPTER 8
Optimization
Listing 8-3. Optimization Example (Cont.)
$
$ Part 1
$
begin super = 1
$
$ define modal coordinates for CMS
$
$ define q-set for component modes and residual vectors
SPOINT 11001 THRU 11020
QSET1 0 11001 THRU 11020
$
$ define which dofs will be retained (i.e. which dofs
$ will form the attachment to the system model when we
$ create SE10 in se10.dat)
$
ASET1 123456 35 36
$
$ part1.dat
$
CQUAD4 18 9 33 34 46 45
CQUAD4 42 9 85 86 98 97
$
param,k6rot,100.
$
$ boundary grids
$
GRID 35 -3.6 6. 0.
GRID 36 -2.8 6. 0.
$
GRID 97 -2. 10. 0.
GRID 98 -1.2 10. 0.
$
MAT1,1,30.+6,,.3,.283
PARAM,WTMASS,.00259
PARAM,AUTOSPC,YES
PSHELL,9,1,.05,1,,1
$
$ plotels to outline component in assembly run
$
plotel,101,33,35
plotel,102,33,93
plotel,103,93,98
plotel,104,98,38
plotel,105,38,36
plotel,106,35,36
eigrl,1,,,9
Section for
Part 1 Model
MD Nastran R2.1 Release Guide

268
Listing 8-3. Optimization Example (Cont.)
$ Part 2
$
begin super = 2
$
$ define modal coordinates for CMS
$
$ define q-set for component modes and residual vectors
SPOINT 21001 THRU 21020
QSET1 0 21001 THRU 21020
$
$ define which dofs will be retained (i.e. which dofs
$ will form the attachment to the system model when we
$ create SE10 in se10.dat)
$
ASET1 123456 4142
$
$
$ part2.dat
$
CQUAD4 43 2 39 40 52 51

CQUAD4 66 2 90 91 103 102
CQUAD4 67 2 91 92 104 103
$
param,k6rot,100.
$
$ boundary grids
$
GRID 41 2.8 6. 0.
GRID 42 3.6 6. 0.
$

GRID 103 4.4 10. 0.
GRID 104 5.2 10. 0.
$
$
$
MAT1,1,30.+6,,.3,.283
PARAM,WTMASS,.00259
PARAM,AUTOSPC,YES
PSHELL,2,1,.05,1,,1
$
$ plotels to outline component in assembly run
$
plotel,201,39,41
plotel,202,39,99
plotel,203,99,104
plotel,204,104,44
plotel,205,44,42
plotel,206,41,42
$
eigrl,1,,,9
enddata
Section for
Part 2 Model
269
CHAPTER 8
Optimization
Randomization of an Input Data File (Prerelease)
Introduction
The stochastic capability in MD Nastran is the first step toward a complete and automatic self-
randomization of a finite element model. The current capability offers the possibility to automatically
distribute tolerances and uncertainties with minimum effort. This dramatically reduces the complexity
of large-scale stochastic simulations. In fact, once the stochastic option is triggered, the entire Bulk Data
file is automatically randomized without further user intervention. The resulting model, which needs to
be incorporated in a Monte Carlo Simulation loop (there are numerous off-the-shelf products which
support this capability) possesses unprecedented levels of realism.
In order to make full use of this new capability, it is necessary to use a multirun environment which can
spawn a certain number of independent MD Nastran executions, collect the results, and perform
statistical postprocessing. With the self-randomization capability in MD Nastran, the user need only
define the outputs to be monitored, such as stresses, Eigenfrequencies, temperatures, displacements, etc.
There is no need to define inputs, as these are defined automatically by MD Nastran.
Benefits
It is sometimes assumed that the inputs to an MD Nastran analysis are known exactly, and thus the
computed responses are exact. This is an invalid assumption since there will always be some uncertainty
in the input values with a corresponding variation in the results. MD Nastran R2 provides a way of
introducing this uncertainty into the analysis process by automatically randomizing user input real
numbers based on the input values and statistical quantities that characterize the variation.
Input
The randomization capability is driven by a new STOCHASTICS Case Control command, as described
in the MD Nastran Quick Reference Guide. If STOCHASTICS=ALL is used, all real quantities on
connectivity (those starting with C), material, and property Bulk Data entries, as well as any loads and
SPCD quantities, are modified based on a covariance factor of 0.05. A Gaussian distribution is used to
randomly select the perturbed quantity with the restriction than the value can be no more that a specified
number of standard deviations from the user input mean value. The default number of maximum
standard deviations is three.
Alternatively, the STOCHASTICS Case Control command can point to a STOCHAS Bulk Data entry
that provides the ability to selectively randomize different types of input quantities by means of user-
specified covariance values and user-prescribed numbers of allowed standard deviations. In this case,
only the types of input specified are randomized so that, for example, it is possible to randomize the load
inputs while leaving the property values unchanged.
Output
There is no new output produced by this capability.
MD Nastran R2.1 Release Guide

270
Guidelines and Limitations
The randomization algorithm involves using a random number generator, a Gaussian distribution, a
prescribed covariance, and a mean value based on user input to determine a randomized value that is to
be used in the analysis. In order to avoid physically meaningless properties, the random value is
prescribed to be within m standard deviations of the input value, where m is a user input value with a
default value of 3.0.
The product of m * COV should not be greater than 1.0 to eliminate the possibility of the property
changing sign.
Any real value in the Bulk Data file will be randomized unless otherwise specified by the user. To keep
a particular field or fields from being randomized, the user must set them equal to a value of 0.0.
271
CHAPTER 8
Optimization
Random Elimination of Element Types (Prerelease)
Introduction
There has been a longstanding capability in MD Nastran that allows the user to specify the random
elimination of a specified percentage of the CWELD elements contained in a bulk data file. This was
done using the PARAM CWRANDEL entry, with an additional CWDIAGP PARAM providing the
option of printing the IDs of the deleted elements. In the current release, this capability has been extended
to the CELASi, CFAST, CSEAM, and 1-D mass (CMASSi, CONM1, and CONM2) elements. In
addition, the user interface has been changed from the NASTRAN statement to the MDLPRM entry.
Benefits
The ability to randomly delete various 1-D elements provides the user with some assessment of the
integrity of the design being modeled. For instance, if randomly deleting 20% of, say, of the CWELD
elements from a model caused a negligible change in the first ten natural frequencies, this was taken as
an indication of the robustness of the structure. Extending this approach to other element types provides
more options in this type of analysis. Placing the input on the MDLPRM entry consolidates that input so
that the user does not have to deal with the PARAM entry.
Input
The MDLPRM entry has ten new PARAMi names that support this capability. Five of these names (e.g,
DELELAS) select the element type to which the random elimination applies and the ratio to be deleted,
while an additional five names (e.g., PRTELAS) provide control as to whether the IDs of the deleted
elements are to be printed. The default is that the IDs will not be printed.
Output
There is no new output produced by this capability.
Guidelines and Limitations
The deletion ratio is input as a real number between 0.0 and 1.0, with 0.0 indicating that no deletion is
to take place, while 1.0 eliminates all elements of the specified type.
It is possible that the elimination of a series of elements will introduce mechanisms in the structure that
will cause the analysis to fail. It is the users responsibility to determine whether this failure has occurred.
A likely scenario for the use of this capability would be to submit the same file multiple times and
determine the variation in the results. MSC does not offer an automated way of doing this at this time.
MD Nastran R2.1 Release Guide

272
Ch. 9: Rotor Dynamics and Aeroelasticity MD Nastran R2.1 Release Guide
9
Rotor Dynamics and
Aeroelasticity
!
Changes to Rotor Dynamics for MD Nastran R2
!
Updating/Summing of Monitor Points
!
Stripwise Aerodynamic Results
!
Input of an Aerodynamic Mesh
!
Rigid Body Spline
!
Wendland Spline Functions for the SPLINE4/5
!
Spline Blending
!
Export of the Spline Matrix
!
External Spline Server
MD Nastran R2.1 Release Guide

274
Changes to Rotor Dynamics for MD Nastran R2
UNBALNC Entry for Frequency Response
The UNBALNC loading has been available for rotor dynamics transient response for several versions.
In prior versions, unbalance loading for frequency response had to be input using RLOADi entries, which
was prone to error. Backward (opposite to the rotor spin direction) loading was easily input. To provide
a more natural input, the UNBALNC Bulk Data entry can now be used for frequency response with the
rotor dynamics option. The entry is selected using the DLOAD Case Control command (note that the
transient response solutions must still specify the transient unbalance using the RGYRO Case Control
command).
The format for the UNBALNC entry in frequency response analysis is the same for transient response,
except the continuations for force output via EPOINTs are ignored. For further information on the
UNBALNC Bulk Data entry, see the MD Nastran Quick Reference Guide.
New Parameter and Hybrid Damping Specifications
In prior versions, the specification of damping using parameter input (PARAM,G for structural damping
and PARAMs ALPHA1 and ALPHA2 for Rayleigh damping) often resulted in users not knowing what
damping, if any, was actually being specified in the model. This is because parameters can be entered
anywhere in the Case Control or Bulk Data Sections. For large models with many include files, a
rigorous search would be required. To make the damping specifications easier, and to allow new damping
formulations such as hybrid damping, a new damping entry has been developed that is selected in the
Case Control section. The new entry allows selection of parameter damping, such as structural damping
(PARAM, G) and Rayleigh damping (PARAMs ALPHA1 and ALPHA2), and a new hybrid damping.
This entry also allows the scaling of material damping (GE on material entries, MATi). The new Bulk
Data entry is named DAMPING, and is selected by new Case Control commands:
For specifying damping of superelements, use
SEDAMP= n
For specifying damping of the residual structure, use
RSDAMP(STRUCTURE, FLUID, or BOTH)= n (Default= STRUCTURE)
where n specifies the DAMPING Bulk Data entry.
For more information on the DAMPING Bulk Data entry, see the MD Nastran Quick Reference Guide.
275
CHAPTER 9
Rotor Dynamics and Aeroelasticity
Updating/Summing of Monitor Points
Introduction
MD Nastran R1 greatly expanded the concept of monitor points by modifying the MONPNT1 entry, and
by introducing the MONPNT2, MONPNT3, and MONDSP1 entries. The release of MD Nastran R2
further expands the monitor points approach by providing two additional capabilities:
1. Allowing for the modification of existing component results for MONDSP1, MONPNT1, and
MONPNT3 by a scalar multiple to allow, for example, a change in sign or units.
2. Enabling the weighted summation or two or more MONDSP1, MONPNT1, or MONPNT3
entries that are of the same type.
Benefits
Updating an existing monitor point result is achieved by applying factors to specified components. As
an example of the utility of this feature, consider a model that has been constructed in one set of units
(e.g., centimeters), where it is desired to see the results in a different set of units (e.g, millimeters).
In such a case, the ability to sum existing monitor point results could also be used, for example, to better
present running results along a wing or fuselage.
Input
The new MONSUM Bulk Data entry is used to implement both the updating and summing of monitor
point results. The description of this entry in the MD Nastran Quick Reference Guide includes examples
and numerous remarks that aid in understanding the entrys capabilities.
Output
The output of the monitor point results from the MONSUM Bulk Data entry is identical to that of the
underlying monitor point. In the updating scenario, if the NAME field on the entry is the same as the
NAMEij fields, only the final updated result is given. If a new NAME is used, both results are given.
Examples (monsum.dat and monsum3.dat)
Two small test files (monsum.dat and monsum3.dat) demonstrate the use of the MONSUM Bulk Data
entry. They are both variations of the familiar forward swept wing, and are described as follows:
monusm.dat- contains some MONSUM Bulk Data entries that are simple enough that the results can be
checked by hand.
monsum3.dat sums the MONPNT3 results. An interesting feature of this example is that it
demonstrates the DMAP enhancements that were required to sum MONPNT3 results produced by using
what is referred to as the mini-ema method with results using the grid point force recovery method.
MD Nastran R2.1 Release Guide

276
Guidelines and Limitations
The remarks for the MONSUM Bulk Data entry in the MD Nastran Quick Release Guide provide a
number of guidelines that should be reviewed before applying this new entry. Because MONSUM
performs both an update-in-place and a combining function, it represents a powerful, compact
capability.
The two examples shown in the MD Nastran Quick Reference Guide provide a further explanation of this
entry. In the first example, a new monitor point is being synthesized from three existing monitor point
results to provide the user with a blended result that has special meaning to the designer. In this case,
results at three stations along the root chord are being summed with different weights for the desired shear
and bending moment resultants. Since the COEFi fields are left blank for the shear components, the
default of 1.0 is applied. In the second example, the MONSUM is used to convert the units of the monitor
point output from English to metric units. Note that in the first case, monitor point results will be
presented for the summed quantity as well as for each of the original MONPNT1 quantities that
contribute to the sum. In the second case, only the final result is presented since its NAME is shared with
the NAMEij attributes. If output in both unit systems is desired, one needs to simply provide a unique
NAME in order to get both results.
277
CHAPTER 9
Rotor Dynamics and Aeroelasticity
Stripwise Aerodynamic Results
Introduction
An additional monitor point, MONCNCM, has been provided in the current release. MONCNCM
streamlines the task of providing stripwise lift and pitching moment results for a doublet-lattice type of
aerodynamic model.
Benefits
Stripwise aerodynamics; i.e., aerodynamic lift and pitching moments, can be quite useful in visualizing
the aerodynamic results on lifting surfaces. In particular, these can be compared with wind tunnel results
with the possibility of weighting the computed aerodynamics to match the test results. It is possible to
generate these results using an aerodynamic MONPNT1, but the preparation of the input data is a
tedious, error-prone process. The new entry automates this process so that it is possible to obtain the
results for each aerodynamic strip in the model with a single MONCNCM entry.
Input
The new MONCNCM Bulk Data entry is used to provide the stripwise aerodynamic results. The remarks
for this entry in the MD Nastran Quick Release Guide provide guidance in its use. In particular, use of
CAERID1=ALL produces results at each aerodynamic strip. The code internally determines if the flat
plate panels are abutting and, if they are, considers this a single strip from the leading edge to the trailing
edge of the surface made up of the multiple CAERO1 entries.
Output
A new table in the .f06 file is included as part of the aerodynamic monitor point output. A sample output
from the moncncm.dat test case, discussed in the following Examples section, is shown here:
A E R O D Y N A M I C M O N I T O R P O I N T I N T E G R A T E D L O A D S
CONFIGURATION = AEROSG2D XY-SYMMETRY = ASYMMETRIC XZ-SYMMETRY = SYMMETRIC
MACH = 9.000000E-01 Q = 4.000000E+01
MONCNCM NAME = ALL INT GROUP ID = 1 CLASS = STRIP
LABEL = ALL STRIPS FOR THE AIRPLANE
CP = 1 AERODYNAMIC COORDINATE SYSTEM
FORCE
MOMENT
STRP YS ZS XREF REFC REFS RIGID ELASTIC RIGID ELASTIC
____ __________ __________ __________ ___________ ___________ ___________ ___________ ___________ ___________
1 1.250E+00 0.000E+00 0.000E+00 1.0000E+01 2.5000E+01 1.5715E+00 1.5735E+00 1.4352E-01 1.4300E-01
2 1.250E+00 0.000E+00 1.428E+01 1.0000E+01 2.5000E+01 -3.1044E-01 -2.9946E-01 -6.2361E-02 -6.1159E-02
3 3.750E+00 0.000E+00 0.000E+00 1.0000E+01 2.5000E+01 1.2319E+00 1.2336E+00 1.1194E-01 1.1147E-01
4 3.750E+00 0.000E+00 1.283E+01 1.0000E+01 2.5000E+01 4.9708E-03 1.6573E-02 -9.2159E-03 -8.0990E-03
5 6.250E+00 0.000E+00 1.139E+01 1.0000E+01 2.5000E+01 1.1896E+00 1.2036E+00 1.7237E-01 1.7369E-01
6 8.750E+00 0.000E+00 9.948E+00 1.0000E+01 2.5000E+01 1.1173E+00 1.1319E+00 9.4866E-02 9.5348E-02
7 1.125E+01 0.000E+00 8.505E+00 1.0000E+01 2.5000E+01 1.0168E+00 1.0308E+00 3.8398E-02 3.8538E-02
8 1.375E+01 0.000E+00 7.061E+00 1.0000E+01 2.5000E+01 8.8463E-01 8.9719E-01 3.4616E-03 3.4251E-03
9 1.625E+01 0.000E+00 5.618E+00 1.0000E+01 2.5000E+01 7.1419E-01 7.2468E-01 -1.3503E-02 -1.3636E-02
10 1.875E+01 0.000E+00 4.175E+00 1.0000E+01 2.5000E+01 4.8032E-01 4.8760E-01 -1.6023E-02 -1.6161E-02
MD Nastran R2.1 Release Guide

278
The strips are arranged based on the increasing y-coordinate of the strip in the aerodynamic coordinate
system. The z coordinate and the x location of the point at which the pitching moment is computed (based
on the MREF field of the MONCNCM entry) is given, as are the local chord length and strip area. This
is followed by the lift and pitching moment at the trim state for the rigid vehicle and for the elastically
deforming vehicle.
Examples (moncncm.dat and monbodi.dat)
Two small test files demonstrate the use of the MONCNCM Bulk Data entry. They are both variations of
the familiar forward swept wing.
moncncm.dat provides tests of various ways of merging data across strips that may not be practical from
an engineering standpoint, but that exercise different code paths.
monbodi.dat demonstrates the application of MONCNCM when there are multiple interference groups
and aerodynamic bodies. It is seen that different interference groups are not merged across aerodynamic
panels, and that there are no results given for CAERO2 entries.
Guidelines and Limitations
The MONCNCM entry only supports doublet lattice-like models and only for lifting surfaces.
Aerodynamic bodies are not supported. The output is in the aerodynamic panel coordinate systems so
that the sign of the forces and moments are a function of the numbering of the corners of the panel (see
the description of the CAERO1 entry in the MD Nastran Quick Reference Guide). The results are given
at the aeroelastic trim state.
279
CHAPTER 9
Rotor Dynamics and Aeroelasticity
Input of an Aerodynamic Mesh
Introduction
Four new Bulk Data entries have been provided in MD Nastran R2: AEGRID, AEQUADD4, AETRIA3,
and AESCALE. These entries allow the user to define an aerodynamic mesh that is distinct from the
traditional aerodynamic model input using a combination of CAEROi/PAEROi entries.
Benefits
The ability to input an aerodynamic mesh into MD Nastran opens up several capabilities that are
considered preliminary at this point.
First, it enables the viewing of an external aero mesh on an MD Nastran structural model using a tool
such as MD Patran.
Second, it is possible to spline results from the structural mesh to the new aerodynamic mesh, and vice
versa. The spline export feature described in a subsequent section provides this spline for manipulation
outside of MD Nastran.
Third, the aerodynamic grid points can be loaded using the AEFORCE entry, thereby enabling the import
of rigid aerodynamic loads from an external aerodynamics program.
In addition, the AESCALE entry, described as follows, opens up some interesting possibilities for
morphing models that have yet to be explored.
Input
The aerodynamic mesh can be input using four new bulk data entries:
The AEGRID entry provides the coordinates of the mesh;
The AEQUAD4 entry connects the mesh using a quadrilateral element;
The AETRIA3 connects the mesh using a triangular element; and
The AESCALE entry allows scaling the AEGRID data on a grid-by-grid basis.
Note that the element data (AEQUAD4 and AETRIA3) are strictly for postprocessing purposes only, and
do not participate in the creation of any aerodynamic matrices.
Output
There is no new output.
MD Nastran R2.1 Release Guide

280
Example (aegridf.dat)
This simple example is a variation on the 15 degree swept wing model named ha144c.dat in the
MSC.Nastran Users Guide for Aeroelasticity. Relative to the example in the guide, the aegridf.dat
example has the following characteristics:
Removes the doublet lattice model.
Adds AEGRID quantities that coincide with all the structural grids in the model, plus four more
that represent an aerodynamic stabilizer.
Uses a SPLINE4 entry to spline the wing forces at the AEGRIDs to the structure.
Uses a SPLINRB entry to spline the AEGRIDs on the stabilizer to the support point.
Removes the FORCE/GRAV artifice used to simulate a constrained wind tunnel condition.
Instead, a 1 g cruise condition is specified with the aircraft free to pitch and plunge.
AEPARM entries specify ALPHA and STAB states.
AEFORCE entries are used to set up load vectors for the intercept, unit angle of attack, and unit
stabilizer setting.
The stability derivative print from running the example includes the following:
N O N - D I M E N S I O N A L S T A B I L I T Y A N D C O N T R O L D E R I V A T I V E C O E F F I C I E N T S
CONFIGURATION = AEROSG2D XY-SYMMETRY = ASYMMETRIC XZ-SYMMETRY = SYMMETRIC
MACH = 4.5000E-01 Q = 2.0000E+00
CHORD = 2.0705E+00 SPAN = 1.1050E+01 AREA = 1.1440E+01
CONTROLLER STATE: INTERCEPT ONLY, ALL CONTROLLERS ARE ZERO
TRIM VARIABLE COEFFICIENT RIGID ELASTIC INERTIAL
UNSPLINED SPLINED RESTRAINED UNRESTRAINED RESTRAINED UNRESTRAINED
REF. COEFF. CX 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CY 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
0.000000E+00
CZ 3.496504E-02 3.496504E-02 3.496504E-02 3.496504E-02 0.000000E+00
3.496504E-02
CMX 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
0.000000E+00
CMY -1.249997E-02 -1.249997E-02 -1.249997E-02 -1.249997E-02 0.000000E+00 -
1.249997E-02
CMZ 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
0.000000E+00
ALPHA CX 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CY 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CZ 2.596154E+00 2.596154E+00 2.596154E+00 2.596154E+00 0.000000E+00 2.596154E+00
CMX 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CMY -1.296420E+00 -1.296420E+00 -1.296420E+00 -1.296420E+00 0.000000E+00 -1.296420E+00
CMZ 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
STAB CX 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CY 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CZ 7.517483E-01 7.517483E-01 7.517483E-01 7.517483E-01 0.000000E+00 7.517483E-01
CMX 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CMY -2.760425E+00 -2.760425E+00 -2.760425E+00 -2.760425E+00 0.000000E+00 -2.760425E+00
CMZ 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
URDD3 CX 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CY 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CZ 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 1.793749E-03 0.000000E+00
CMX 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
CMY 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 -6.412643E-04 0.000000E+00
CMZ 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00 0.000000E+00
281
CHAPTER 9
Rotor Dynamics and Aeroelasticity
It is seen that the splined and unsplined rigid aerodynamic terms are identical, which means the splines
are functioning well, and that the rigid and elastic results are the same. These observations reflect the fact
that there is no aeroelastic feedback in this case.
Trim results from the run include:
It is seen that the very lightweight structure is in trim based on intercept aerodynamics, a negative angle
of attack, and a slightly positive stabilizer.
Guidelines and Limitations
The aerodynamic mesh cannot be used in conjunction with the existing CAEROi format for creating an
aero mesh.
The CMPID listed on the aerodynamic elements can be referenced on an AECOMP that is used in the
aerodynamic splining. However, the splining is only done based on the AEGRID data that are provided
on an AELIST entry identified with the AECOMP.
Unlike the CAEROi format, there is no central grid point on an aerodynamic element. The only
aerodynamic grid points are at the element vertices.
There is no integration of user input pressures or downwashes, so the AEFORCE entry is the only means
of loading the aerodynamic mesh and, again, this can only be done at the AEGRID locations.
AICs are not supported for this type of mesh.
A E R O S T A T I C D A T A R E C O V E R Y O U T P U T T A B L E S
CONFIGURATION = AEROSG2D XY-SYMMETRY = ASYMMETRIC XZ-SYMMETRY = SYMMETRIC
MACH = 4.500000E-01 Q = 2.000000E+00
CHORD = 2.0705E+00 SPAN = 1.1050E+01 AREA = 1.1440E+01
TRIM ALGORITHM USED: LINEAR TRIM SOLUTION WITHOUT REDUNDANT CONTROL SURFACES.
AEROELASTIC TRIM VARIABLES
ID LABEL TYPE TRIM STATUS VALUE OF UX
INTERCEPT RIGID BODY FIXED 1.000000E+00
501 ALPHA GENERAL CONTROL FREE -1.334841E-02 RADIANS
502 STAB GENERAL CONTROL FREE 1.973043E-03 RADIANS
503 URDD3 RIGID BODY FIXED 1.000000E+00
LOAD FACTOR
505 URDD5 RIGID BODY FIXED 0.000000E+00
RAD/S/S PER G
MD Nastran R2.1 Release Guide

282
Rigid Body Spline
Introduction
A new spline technique has been introduced that enables the splining of aerodynamic data to exactly six
structural degrees of freedom in a rigid body fashion.
Benefits
The primary application of the rigid body spline is to introduce aerodynamic loads into the structural
analysis process when there is no structure underlying the aero mesh. For example, this enables
introducing loads from the tail component in a wing design when only the wing has structural detail.
Input
The new rigid spline method uses the SPLINRB Bulk Data entry.
Output
There is no new output.
Example (splinrb.dat)
A very simple example is provided that is based on the ha144e test case of Section 7.5 of the
MSC.Nastran Users Guide for Aeroelastic Analysis. The example in the Users Guide uses separate
1-D splines (the SPLINE2 entry) to distribute the aerodynamic load on the beam structure. This example
replaces these two splines with SPLINRB entries that spline all the loads to the support point. The result
is that these loads no longer create an aeroelastic affect, and the rigid and splined stability derivatives are
identical.
Guidelines and Limitations
In addition to the loads being transferred to the six structural degrees of freedom, it should also be noted
that the motion of the aerodynamic surface corresponds to the rigid body motion defined by these six
DOFs.
283
CHAPTER 9
Rotor Dynamics and Aeroelasticity
Wendland Spline Functions for the SPLINE4/5
Introduction
Options have been added to the SPLINE4 and SPLINE5 Bulk Data entries to support the selection of
radial interpolation functions. Reference 1 contains the mathematical development of these functions,
and Reference 2 further develops them in the context of an aeroelastic application.
Benefits
The references indicate that these alternative splining techniques should provide smooth results with
improved performance relative to the existing methods. As indicated in the following Guidelines and
Limitations section, MSC does not yet have enough experience with the new methods to quantify these
benefits.
Input
The METH field for the SPLINE4 entry now supports RIS in addition to the existing IPS, TPS, and FPS
methods. If RIS is selected, the order of the interpolation method is either defaulted to WF2 or can be set
to WF0. RCORE is used to define the radius of support. For the SPLINE5, METH is a new field, with
BEAM (infinite beam spline) as the default. If METH=RIS is specified, FTYPE (default=WF2) and
RCORE also require input.
Example (ha144c_ris)
The ha144c example has been modified to change the spline from SPLINE1 to SPLINE4 with
METH=RIS, FTYPE=WF2, and RCORE=5.0. Running the example produces answers that are
equivalent, but not identical to, ha144c. The differences are comparable to those that results from
replacing the infinite plate spline with a finite plate spline so that it is deemed these results are
reasonable.
Guidelines and Limitations
MSC has limited experience with the radial interpolation functions and therefore SPLINE4/5 should be
used in conjunction with the Spline Verify option of MSC.Flightloads so that a visual check of the quality
of the fit can be made.
Because the radial interpolation functions act over a limited area, specified using RCORE, it is believed
that a single spline can be effectively applied to a large area. For example, with a high aspect ratio wing,
a single spline may suffice.
If the linear SPLINE5 includes the rotational flexibility (DTHX/DTHY > -1.0) from the structural
deformations, the FTYPE =WF2 will be used regardless of the user input value of FTYPE.
MD Nastran R2.1 Release Guide

284
Spline Blending
Introduction
A capability has been added to blend spline effects across multiple splines.
Benefits
Prior to the MD Nastran R2 release, an error occurred if the same aerodynamic grid was invoked on
multiple splines. This could result in discontinuities at the panel boundaries; therefore, this rule has been
relaxed so that an aero grid can be selected on multiple splines, and the results blended in some user
specified fashion. This results in smoother displacement patterns which will be of particular benefit as
aeroelastic analysis includes more CFD aerodynamics.
Input
A parameter has been provided to indicate that blending is allowed to occur. The parameter is
MPTSPLIN, and is input using the MDLPRM Bulk Data entry. This parameter has the following
meaning:
This parameter does not apply to aerodynamic grid points defined with AEGRID discussed previously.
These grid points can always be referenced on multiple splines.
Two new Bulk Data entries SPBLND1 and SPBLND2, have been provided to support spline blending.
If neither of these is used, the splines are averaged. The SPBLND1 performs strip-based blending that
can be either averaged (with a user defined weighting), linear, or cubic in nature. The SPBLND2
performs a curve-based blending based on the users input curve.
Example (ha145e_blnd)
The existing ha145e test file has a single SPLINE1 entry. In this example, this has been replaced by two
splines that overlap the two center rows of boxes. MDLPRM,MPTSPLIN,1 is used to enable this overlap
when using the CAERO1 generated aerodynamic meshes. A SPBLND1 entry is then used to blend the
results in a linear way. If the SPBLND1 entry had not been used, the splines would have been averaged.
For this simple example, the blending has little affect on the flutter results. The intent of the example is
to demonstrated the new features.
Guidelines and Limitations
This is an area where MSC has limited experience, so there is limited guidance. The blending
developments have been made in concert with alternative aeromeshingsee Input of an Aerodynamic
MPTSPLIN = 0 Do not allow an aerodynamic grid to be referenced by more than one spline
(Default).
= 1 Allow an aerodynamic grid to be referenced by multiple splines.
285
CHAPTER 9
Rotor Dynamics and Aeroelasticity
Mesh on page 279and are therefore pointed at applications that use external aerodynamics. Blending
will be most useful in areas such as wing fuselage junctions, where the standard splining techniques are
likely to result in gaps in the aerodynamic deformations.
MD Nastran R2.1 Release Guide

286
Export of the Spline Matrix
Introduction
A capability has been provided to export the spline matrix to either an .op2 file or to the punch file.
Benefits
Users can now easily obtain the spline matrix in a format that can be readily applied in another procedure,
or reintroduced into another MD Nastran run.
Input
The SPLINOUT Case Control command is used to produce the exported spline matrix. If the .op2 format
is used, the user should also assign a file that will contain the matrix in OUTPUT2 format.
Output
If the punch output is requested, the GPGK spline matrix is written to the .pch file using the DMIG
format. The DMIG print is preceded by direct table input (DTI) data that provides a map of the matrix
columns to the aerodynamic degrees of freedom. With the default .op2 method, an output2 file is written.
Examples (splinopch and splinoop2)
These two examples show the use of the .pch and .op2 options for exporting a spline mat.
References:
H. Wendland, Piecewise Polynomial, Positive Definite and Compactly Supported Radial Functions of
Minimal Degree, Adv. Computational Mathematics, 4(1995), pp. 389-396.
A. Beckert and H. Wendland, Multivariate Interpolation for Fluid-Structure-Interaction Problems using
Radial Basis Functions, Aerospace Sci. Technology, 00 (2001), pp. 1-11.
287
CHAPTER 9
Rotor Dynamics and Aeroelasticity
External Spline Server
Introduction
The external spline evaluation capability will allow the user to create custom spline evaluation programs,
and use them within MD Nastran without modifying the MD Nastran code.
This new spline type may be used in any of the existing solution sequences that currently support splines.
The external spline evaluation feature is implemented using the same client/server technology that is
used for the beam cross section library, P-element geometry evaluator, and DRESP3 external response
evaluator in MD Nastran. MD Nastran acts as the client, and the external spline evaluation program acts
as the server. Whenever the MD Nastran spline module receives a request for an externally-evaluated
spline, it sends the request to the spline server. The spline server computes the spline matrix and returns
it to the spline module. The spline module then assembles the externally-evaluated spline matrix into the
global spline matrix. Communication between the client and server programs is established using
application programming interface (API) routines. Multiple servers may be used in a single MD Nastran
run.
The following figure illustrates the data flow to and from the spline server.
Benefits
This new capability opens up MD Nastran to custom spline methods. The server executable is
independent from the MD Nastran program, so that server programs may be customized easily without
changes to MD Nastran.
MD Nastran
Spline Server
User function sxsevd.c
Spline Matrix
Integer Data
Real Data
Character Data
Grid Locations
Element Connectivity
API
API
MD Nastran
Spline Server
User function sxsevd.c
Spline Matrix
Integer Data
Real Data
Character Data
Grid Locations
Element Connectivity
API
API
MD Nastran R2.1 Release Guide

288
Inputs
Two new Bulk Data entries have been created: SPLINEX and AELISTC. These will be discussed in
subsequent sections.
Since an external spline is computed using a user-built server executable, defining an external spline
requires additional steps to preparing the regular bulk data entries. The purpose of the additional steps is
to establish the connection between MD Nastran and the external spline program. Five steps are required
in setting up the external spline:
1. Write FORTRAN or C routines and build a server executable based on the spline server template
routine.
2. Create a SPLINEX Bulk Data entry to define the spline.
3. Create a CONNECT entry to define an external spline group.
4. Create an evaluator connection file to associate the external evaluator with the server program.
5. Submit the MD Nastran job with the gmconn keyword referencing the evaluator connection file.
The descriptions of these steps are shown in the following sections.
Step 1. Build an External Spline Server
First, copy the server template routines (sxsevd.c and sxmsg.c) from the MD Nastran
install_dir/md2007/spxsrv directory. The routine sxmsg.c defines error messages. It is required even if
there are no messages to define. The stub version from the delivery is sufficient to satisfy the linker. The
sxsevd.c routine is the main interface to the users spline code. The user may either modify sxsevd.c to
perform the calculations, or simply use it as a conduit to pass data between MD Nastran and the users
spline evaluation subroutine.
Next, copy two script files: spxsrv and makefile. These two files are required to build the spline server
program. To build the server program, type
md2007 ./spxsrv build.
Step 2. Create a SPLINEX entry
A SPLINEX Bulk Data entry will define which points are to be splined and which spline server to use.
At a minimum, two sets of points, one on each mesh, must be defined.
Defines the input for a spline that will be evaluated with a user-supplied procedure.
SPLINEX
Externally-Evaluated Spline
289
CHAPTER 9
Rotor Dynamics and Aeroelasticity
Format:
Example:
Remarks:
1. The SPLINEX EID must be unique with respect to all SPLINEi entries.
2. The USAGE field allows you to specify that the particular spline interpolant defined in this entry
is to be used for just the force transformation (FORCE), or just the displacement transformation
(DISP), or both (BOTH).
(FORCE/BOTH splines are in the transform).
(DISP/BOTH splines are in the transform).
1 2 3 4 5 6 7 8 9 10
SPLINEX EID GROUP
DGCOMP IGCOMP DECOMP IECOMP
USAGE
AELIST
AEFACT AELISTC
SPLINEX 3
SPLNGRP4 GWNG1A GWNG1S
BOTH
101 201 301
Field Contents
EID Element identification number. (Integer > 0)
GROUP Group name to which the external spline type belongs. (Character; no Default)
DGCOMP The name of an AECOMP or AECOMPL entry that defines the set of points for the
dependent mesh. (Character; Default = Blank). See Remarks 3. and 4.
IGCOMP The name of an AECOMP or AECOMPL entry that defines the set of points for the
independent mesh. (Character; Default = Blank). See Remarks 3. and 4.
DECOMP The name of an AECOMP or AECOMPL entry that defines the set of elements for
the dependent mesh. (Character; Default = Blank). See Remarks 3., 4. and 5.
IECOMP The name of an AECOMP or AECOMPL entry that defines the set of elements for
the independent mesh. (Character; Default = Blank). See Remarks 3., 4. and 5.
USAGE Spline usage flag to determine whether this spline applies to the force
transformation, displacement transformation, or both. FORCE, DISP or BOTH. See
Remark 2. (Character, Default = BOTH)
AELIST ID of an AELIST that contains a list of user-defined integer data. (Integer; no
Default). See Remark 6.
AEFACT ID of an AEFACT that contains a list of user-defined real data. (Integer; no Default).
See Remark 6.
AELISTC ID of an AELISTC that contains a list of user-defined character data. (Integer; no
Default). See Remark 6.
Fg GPkg [ ]
T
Pk { } =
Uk GDkg [ ] Ug { } =
MD Nastran R2.1 Release Guide

290
In general, the two transforms are done with distinct matrices. Only when ALL splines are of type
BOTH is the familiar transpose relationship satisfied. The default behavior
(BOTH for all splines) is compatible with versions of MD Nastran prior to Version 70.5.
In general, the USAGE field can be used to apply aerodynamic forces to the structure from
aerodynamic panels that are intended not to move (USAGE=FORCE) or to apply structural
displacements to aerodynamic grids whose forces are not to be applied to the structure
(USAGE=DISP). The DISP option is somewhat esoteric in that its use implies the aeroelastic
effect of the surface is important while its forces are not. (In other words, only the forces arising
from its effects on other surfaces is important.) While there may be circumstances where this is
true, it is unlikely.
Take care that you included all the FORCEs from aerodynamic panels that are important by
including them in either FORCE or BOTH spline(s). MD Nastran will not issue a warning unless
all forces are omitted. All displacements may be omitted without warning (and is a means to
perform rigid aerodynamic analyses).
3. Typically, for aero-to-structure splines, dependent means aerodynamic, and independent
means structural.
4. If the component defines a structural mesh, then the grid component may be left blank, and the
list of grids will be obtained from the element component members connectivity. Both may not
be left blank.
5. Structural elements referenced by DECOMP and IECOMP are limited to the following element
types: CQUAD4, CQUADR, CTRIA3, CTRIAR. In one list, elements from the different types
may not share the same ID.
6. The data that are defined on the AELIST, AEFACT, and AELISTC have no meaning to MD
Nastran. These lists are generic containers for data that has meaning to the spline server. Note that
the AELIST is limited to numbers greater than zero.
Defines a list of eight-character strings.
Format:
Example:
AELISTC
Character item list
1 2 3 4 5 6 7 8 9 10
AELISTC SID C1 C2 C3 C4 C5 C6 C7
C8 -etc.-
AELISTC 101 FBS
STRING12
GPgk [ ]
T
GDkg [ ] =
291
CHAPTER 9
Rotor Dynamics and Aeroelasticity
Remark:
1. Intervening blank fields are not allowed.
Step 3. Create a CONNECT Statement
The CONNECT statement, specified in the File Management Section (FMS) of the MD Nastran input
file, defines the external spline group. Multiple CONNECT statements may be defined to define multiple
groups. For the spline server, the CONNECT statement should have the following form:
CONNECT SPLINEX <group name> <evaluator>
where:
Example:
CONNECT SPLINEX SPLNGRP4 EXTSPLN
The group name must match the one specified on a SPLINEX Bulk Data entry. The evaluator value is
the evaluator name specified in the evaluator connection file (see next step).
Step 4. Create an Evaluator Connection File
The evaluator connection file defines the association between the external evaluator and the
corresponding server program.
Format:
Evaluator name, the connection option, the path of the server program
Example:
EXTSPLN,-,/local/user/extspline/spxserv
EXTSPLN is the evaluator name. - indicates that the pipe option is used (i.e., a server executable
resides in any net-mounted computer). The path name points to the server executable.
Field Contents
SID Set identification number. (Integer > 0)
Ci List of eight-character strings.
group name = Group name referenced by the GROUP field on the SPLINEX Bulk Data entry.
evaluator = Evaluator name that is referenced in the evaluator connection file (see next step).
MD Nastran R2.1 Release Guide

292
Step 5. Submit an MD Nastran job with the GMCONN keyword
The keyword gmconn is used on the command line to reference the name of the evaluator connection file.
The connection file is processed at the MD Nastran initialization stage to establish a link between the MD
Nastran program and the server program(s).
Outputs
The spline matrix is created by the server program and passed back to MD Nastran. This matrix must be
of the size (number of independent degrees of freedom) x (number of dependent degrees of freedom).
Guidelines and Limitations
1. Each group of external splines requires one CONNECT entry.
2. The external spline, the DRESP3, the beam library, and the geometry evaluator may be mixed.
3. The AELIST entry may only contain numbers greater than zero.
Example
The following example uses a simple spline server that has the spline matrix hard coded to be identical
to that created by the SPLINE4. This spline server it not usable outside of this test case. Rather, the user
should take from this example the knowledge of how to set up and use an external spline program with
MD Nastran.
To keep the spline matrix very small, a new model was created. This model has only four quadrilateral
elements, two chordwise and two spanwise, connected with nine grid points. The aerodynamic model
contains nine boxes: three chordwise and three spanwise.
Step 1 - Build an External Spline Server
The source code for this example is copied from the MD Nastran installation. There are five source code
files required to build this server: sxsevdb.c, sxsevdb.h, sxmsg.c, spxaport.h, and mkgmat.F. This
example shows how to use a spline server that has the primary calculations performed using Fortran.
Another example included with the MD Nastran installation, sxsevda.c, uses only C for the spline server.
sxsevdb.c
sxsevdb.c, shown in the following listing, is the main interface routine between the server program and
the MD Nastran client. For this example, the primary spline matrix calculations are performed in the
Fortran subroutine, so this routine just acts as a conduit to pass the data between the MD Nastran spline
module and the Fortran subroutine. The output gmat variable contains the computed spline matrix.
293
CHAPTER 9
Rotor Dynamics and Aeroelasticity
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "stdmsc.h"
#include "ftncalls.h" /* String passing from C->Fortran */
#include "spxaport.h" /* FORTRAN -> C Bindings */
#include "sxsevdb.h"
/*
**.......................................................................
**.......................................................................
**.......................................................................
*/
void sxsevd (INTEGER group_id, /* Group id */
INTEGER spline_id, /* Spline id */
INTEGER *usage, /* Usage string (stored as hollerith) */
INTEGER n_int_data, /* Number of integer data */
INTEGER *int_data, /* Integer data */
INTEGER n_real_data, /* Number of real data */
MACHINEPRECISION *real_data, /* Real data */
INTEGER n_char_data, /* Number of character data */
INTEGER *char_data, /* Character data (stored as hollerith) */
INTEGER n_dep_grid, /* Number of dependent grids */
INTEGER *dep_grid_id, /* Dependent grid ids */
MACHINEPRECISION *dep_grid_xyz, /* Dependent grid x,y,z locations */
INTEGER n_indep_grid, /* Number of independent grids */
INTEGER *indep_grid_id, /* Independent grid ids */
MACHINEPRECISION *indep_grid_xyz, /* Independent grid x,y,z locations */
INTEGER n_dep_elem, /* Number of dependent elements */
INTEGER *dep_elem, /* Dependent element table */
INTEGER n_indep_elem, /* Number of independent elements */
INTEGER *indep_elem, /* Independent element table */
char *command_line, /* Optional command line argument */
char *connect_data, /* Optional connect data */
MACHINEPRECISION *gmat, /* The computed spline matrix */
INTEGER *error) { /* Error code */
/*.......................................................................
*
* PURPOSE:
*
* COMPUTE THE EXTERNAL SPLINE
*
MD Nastran R2.1 Release Guide

294
*.......................................................................
*
* CALLED BY:
*
* SendSXsvald
*
*.......................................................................
*
* NOTES:
*
* 1. THE ELEMENT TABLES HAVE THE FOLLOWING FORMAT:
*
* | ID | NG | ID1 | ID2 | ID3 | ID4 |
*
* WHERE:
*
* ID = ELEMENT ID
* NG = NUMBER OF GRIDS USED TO DEFINE THE ELEMENT
* IDi = GRID ID USED TO CONNECT THE ELEMENT. IF NG = 3,
* ID4 = 0.
*
*.......................................................................
*/
/* ..........................................................................
**
** This is a really dumb server. It hard codes the spline matrix to
** be used with the threeby test case. The spline matrix should be
** identical to that produced by the spline4.
*/
/* declarations */
int cml_len;
int cnd_len;
int n_dep_dof;
int n_indep_dof;
FTN_CHAR_ARG_DCL_AUTO(cml_descrip, command_line)
FTN_CHAR_ARG_DCL_AUTO(cnd_descrip, connect_data)
/* .......................................................................... */
n_dep_dof = n_dep_grid * 6;
n_indep_dof = n_indep_grid * 6;
/* set up character string passing to fortran */
cml_len = strlen(command_line);
FTN_CHAR_ARG_DEFLEN(cml_descrip, command_line, cml_len);
cnd_len = strlen(connect_data);
FTN_CHAR_ARG_DEFLEN(cnd_descrip, connect_data, cnd_len);
/* ..........................................................................
**
** Call the fortran routine to set the spline matrix. */
# if ( _FTN_CALL_CHAR_FORMAT == _FTN_CALL_CHAR_DESCRIP ) || \
( _FTN_CALL_CHAR_FORMAT == _FTN_CALL_CHAR_MIXED_LEN ) || \
( _FTN_CALL_CHAR_FORMAT == _FTN_CALL_CHAR_SHADOW_CLSQ )
_FTN_CALL_TYPE_PROTO MKGMAT (&group_id, &spline_id, usage,
&n_int_data, int_data,
&n_real_data, real_data,
&n_char_data, char_data,
&n_dep_grid, dep_grid_id, dep_grid_xyz,
&n_indep_grid, indep_grid_id, indep_grid_xyz,
&n_dep_elem, dep_elem,
&n_indep_elem, indep_elem,
FTN_CHAR_ARG(cml_descrip,command_line,cml_len),
FTN_CHAR_ARG(cnd_descrip,connect_data,cnd_len),
&n_indep_dof, &n_dep_dof, gmat,
error);
295
CHAPTER 9
Rotor Dynamics and Aeroelasticity
sxsevdb.h
This file, shown in the following listing, contains function prototypes used in the spline server driver
routine. The many if statement branches (# if) are required for portability.
# elif _FTN_CALL_CHAR_FORMAT == _FTN_CALL_CHAR_ADD_LEN
_FTN_CALL_TYPE_PROTO MKGMAT (&group_id, &spline_id, usage,
&n_int_data, int_data,
&n_real_data, real_data,
&n_char_data, char_data,
&n_dep_grid, dep_grid_id, dep_grid_xyz,
&n_indep_grid, indep_grid_id, indep_grid_xyz,
&n_dep_elem, dep_elem,
&n_indep_elem, indep_elem,
FTN_CHAR_ARG(cml_descrip,command_line,cml_len),
FTN_CHAR_ARG(cnd_descrip,connect_data,cnd_len),
&n_indep_dof, &n_dep_dof, gmat,
error,
(FTNIMPLICITLEN) cml_len,
(FTNIMPLICITLEN) cnd_len);
#endif
/* .......................................................................... */
return;
}
#ifndef SXSEVD_H
#define SXSEVD_H
#include "ftncalls.h"
#if _FTN_CALL_CHAR_FORMAT == _FTN_CALL_CHAR_DESCRIP
/* Purpose: computes spline matrix */
void _FTN_CALL_TYPE_PROTO MKGMAT (INTEGER *group_id, /* Group id */
INTEGER *spline_id, /* Spline id */
INTEGER *usage, /* Usage string (stored as hollerith) */
INTEGER *n_int_data, /* Number of integer data */
INTEGER *int_data, /* Integer data */
INTEGER *n_real_data, /* Number of real data */
MACHINEPRECISION *real_data, /* Real data */
INTEGER *n_char_data, /* Number of character data */
INTEGER *char_data, /* Character data (stored as hollerith) */
INTEGER *n_dep_grid, /* Number of dependent grids */
INTEGER *dep_grid_id, /* Dependent grid ids */
MACHINEPRECISION *dep_grid_xyz, /* Dependent grid x,y,z locations */
INTEGER *n_indep_grid, /* Number of independent grids */
INTEGER *indep_grid_id, /* Independent grid ids */
MACHINEPRECISION *indep_grid_xyz, /* Independent grid x,y,z locations */
INTEGER *n_dep_elem, /* Number of dependent elements */
INTEGER *dep_elem, /* Dependent element table */
INTEGER *n_indep_elem, /* Number of independent elements */
INTEGER *indep_elem, /* Independent element table */
CHARACTER , /* Command line string */
CHARACTER , /* Data from the CONNECT statement */
INTEGER *n_indep_dof, /* Number of independent degrees of freedom */
INTEGER *n_dep_dof, /* Number of dependent degrees of freedom */
MACHINEPRECISION *gmat, /* The computed spline matrix */
INTEGER *error); /* Error code */
#elif _FTN_CALL_CHAR_FORMAT == _FTN_CALL_CHAR_ADD_LEN
/* Purpose: computes spline matrix */
MD Nastran R2.1 Release Guide

296
void _FTN_CALL_TYPE_PROTO MKGMAT (INTEGER *group_id, /* Group id */
INTEGER *spline_id, /* Spline id */
INTEGER *usage, /* Usage string (stored as hollerith) */
INTEGER *n_int_data, /* Number of integer data */
INTEGER *int_data, /* Integer data */
INTEGER *n_real_data, /* Number of real data */
MACHINEPRECISION *real_data, /* Real data */
INTEGER *n_char_data, /* Number of character data */
INTEGER *char_data, /* Character data (stored as hollerith) */
INTEGER *n_dep_grid, /* Number of dependent grids */
INTEGER *dep_grid_id, /* Dependent grid ids */
MACHINEPRECISION *dep_grid_xyz, /* Dependent grid x,y,z locations */
INTEGER *n_indep_grid, /* Number of independent grids */
INTEGER *indep_grid_id, /* Independent grid ids */
MACHINEPRECISION *indep_grid_xyz, /* Independent grid x,y,z locations */
INTEGER *n_dep_elem, /* Number of dependent elements */
INTEGER *dep_elem, /* Dependent element table */
INTEGER *n_indep_elem, /* Number of independent elements */
INTEGER *indep_elem, /* Independent element table */
CHARACTER , /* Command line string */
CHARACTER , /* Data from the CONNECT statement */
INTEGER *n_indep_dof, /* Number of independent degrees of freedom */
INTEGER *n_dep_dof, /* Number of dependent degrees of freedom */
MACHINEPRECISION *gmat, /* The computed spline matrix */
INTEGER *error, /* Error code */
FTNIMPLICITLEN , /* String length of command_line */
FTNIMPLICITLEN ); /* String length of connect_data */
#elif _FTN_CALL_CHAR_FORMAT == _FTN_CALL_CHAR_MIXED_LEN
/* Purpose: computes spline matrix */
void _FTN_CALL_TYPE_PROTO MKGMAT (INTEGER *group_id, /* Group id */
INTEGER *spline_id, /* Spline id */
INTEGER *usage, /* Usage string (stored as hollerith) */
INTEGER *n_int_data, /* Number of integer data */
INTEGER *int_data, /* Integer data */
INTEGER *n_real_data, /* Number of real data */
MACHINEPRECISION *real_data, /* Real data */
INTEGER *n_char_data, /* Number of character data */
INTEGER *char_data, /* Character data (stored as hollerith) */
INTEGER *n_dep_grid, /* Number of dependent grids */
INTEGER *dep_grid_id, /* Dependent grid ids */
MACHINEPRECISION *dep_grid_xyz, /* Dependent grid x,y,z locations */
INTEGER *n_indep_grid, /* Number of independent grids */
INTEGER *indep_grid_id, /* Independent grid ids */
MACHINEPRECISION *indep_grid_xyz, /* Independent grid x,y,z locations */
INTEGER *n_dep_elem, /* Number of dependent elements */
INTEGER *dep_elem, /* Dependent element table */
INTEGER *n_indep_elem, /* Number of independent elements */
INTEGER *indep_elem, /* Independent element table */
CHARACTER , /* Command line string */
FTNIMPLICITLEN , /* String length of command_line */
CHARACTER , /* Data from the CONNECT statement */
297
CHAPTER 9
Rotor Dynamics and Aeroelasticity
sxmsg.c
sxmsg.c is used by the user to define error messages. This version has only one dummy message defined.
This file must exist, even if only in this stub form, for the linker to be satisfied.
FTNIMPLICITLEN , /* String length of connect_data */
INTEGER *n_indep_dof, /* Number of independent degrees of freedom */
INTEGER *n_dep_dof, /* Number of dependent degrees of freedom */
MACHINEPRECISION *gmat, /* The computed spline matrix */
INTEGER *error); /* Error code */
#elif _FTN_CALL_CHAR_FORMAT == _FTN_CALL_CHAR_SHADOW_CLSQ
/* Purpose: computes spline matrix */
void _FTN_CALL_TYPE_PROTO MKGMAT();
#endif /* _FTN_CALL_CHAR_FORMAT */
#endif /* SXSEVD_H */
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "stdmsc.h"
void sxmsg (INTEGER grpid,
INTEGER errcod,
INTEGER mxlen,
char *message,
INTEGER *error) {
/*.......................................................................
**
** Purpose
** To handle the error message of 'EXTSPLN' group
**
** This is a stub version, intended to be modified by the user
** to add their own messages.
**
**.......................................................................
** Arguments
**
** GRPID in int ID of group or group name
** ERRCOD in int Error message number if any found
** MXLEN out int Maximum length of the message that can be passed
** Z out char Array to contain the message return
** ERROR out int The code returned indicates the type of error
**
**....................................................................... */
/* local variables */
int n_messages = 0;
/* ..........................................................................
**
** Initializations */
*message = '\0';
if (grpid != 1) return;
if (n_messages < 1) return;
/* ..........................................................................
**
** Assign message text to error codes */
if (errcod == 1) {
strncpy (message, "Error code = 1", mxlen);
*error = errcod;
} else {
strncpy (message, "Unknown error code for spline server.", mxlen);
*error = -1;
}
return;
}
MD Nastran R2.1 Release Guide

298
spxaport.h
spxaport.h defines the name of the Fortran subroutine in a machine portable way.
/*
:::::::::::::::::::::::::::::::::::::::::::::
: :
: FILE spxaport.h :
: :
:::::::::::::::::::::::::::::::::::::::::::::
*/
#ifndef SPXEXPORT_H
#define SPXEXPORT_H
/* FORTRAN to C bindings */
#if defined(_UC_FTN_NAME)
/* The upper case names are correct. */
#elif defined(_LC_FTN_NAME)
#define MKGMAT mkgmat
#elif defined(_TU_FTN_NAME)
#define MKGMAT mkgmat_
#elif defined(_DOS_PRAGMA_FTN_NAME)
#pragma aux MKGMAT "^";
#endif /* ..._FTN_NAME */
#endif /* SPXEXPORT_H */
299
CHAPTER 9
Rotor Dynamics and Aeroelasticity
mkgmat.F
mkgmat.F creates the spline matrix.
SUBROUTINE MKGMAT (GRPID, SPLID, USAGE,
+ NIDATA, IDATA,
+ NRDATA, RDATA,
+ NCDATA, CDATA,
+ NDGLST, DGID, DGXYZ,
+ NIGLST, IGID, IGXYZ,
+ NDELEM, DELEM,
+ NIELEM, IELEM,
+ CMDLINE, CNCTDAT,
+ NIDOF, NDDOF, GMAT,
+ ERROR)
C.......................................................................
C
C PURPOSE:
C
C CREATE THE SPLINE MATRIX. THIS SAMPLE SPLINE SERVER IS VERY SIMPLE,
C WITH A HARD CODED SPLINE MATRIX THAT GOES WITH THE
C THREEBY TEST CASE.
C
MD Nastran R2.1 Release Guide

300
C ARGUMENTS:
C
C GRPID INT IN GROUP ID
C SPLID INT IN EXTERNAL SPLINE ID
C USAGE INT IN USAGE STRING (STORED AS HOLLERITH)
C
C NIDATA INT IN NUMBER OF INTEGER DATA
C IDATA INT IN INTEGER DATA
C
C NRDATA INT IN NUMBER OF REAL DATA
C RDATA RDP IN REAL DATA
C
C NCDATA INT IN NUMBER OF CHARACTER DATA
C CDATA INT IN CHARACTER DATA (STORED AS HOLLERITH)
C
C NDGLST INT IN NUMBER OF DEPENDENT GRIDS
C DGID INT IN DEPENDENT GRID IDS
C DGXYZ RDP IN DEPENDENT GRID X,Y,Z LOCATIONS
C
C NIGLST INT IN NUMBER OF INDEPENDENT GRIDS
C IGID INT IN INDEPENDENT GRID IDS
C IGXYZ RDP IN INDEPENDENT GRID X,Y,Z LOCATIONS
C
C NDELEM INT IN NUMBER OF DEPENDENT ELEMENTS
C DELEM INT IN DEPENDENT ELEMENT CONNECTIVITY
C
C NIELEM INT IN NUMBER OF INDEPENDENT ELEMENTS
C IELEM INT IN INDEPENDENT ELEMENT CONNECTIVITY
C
C NIDOF INT IN NUMBER OF INDEPENEDENT DEGREES OF FREEDOM
C NDDOF INT IN NUMBER OF DEPENEDENT DEGREES OF FREEDOM
C GMAT RDP OUT SPLINE MATRIX
C
C ERROR INT IN/OUT ERROR CODE FOR THE CALL.
C 0 = PRINT ERROR MESSAGES
C 1 = DO NOT PRINT ERROR MESSAGES.
C
C.......................................................................
IMPLICIT NONE
C ... I/O VARIABLES
INTEGER GRPID
INTEGER SPLID
INTEGER USAGE
INTEGER NIDATA
INTEGER IDATA(NIDATA)
INTEGER NRDATA
DOUBLE PRECISION RDATA(NRDATA)
INTEGER NCDATA
INTEGER CDATA(2,NCDATA)
INTEGER NDGLST
INTEGER DGID(NDGLST)
DOUBLE PRECISION DGXYZ(3,NDGLST)
INTEGER NIGLST
INTEGER IGID(NIGLST)
DOUBLE PRECISION IGXYZ(3,NIGLST)
301
CHAPTER 9
Rotor Dynamics and Aeroelasticity
INTEGER NDELEM
INTEGER DELEM(NDELEM)
INTEGER NIELEM
INTEGER IELEM(NIELEM)
CHARACTER*(*) CMDLINE
CHARACTER*(*) CNCTDAT
C NOTE: NIDOF and NDDOF are computed in the calling routine so that
c this routine can use a two-dimensional array instead of
c a 2D array stored in a 1D vector.
INTEGER NIDOF
INTEGER NDDOF
DOUBLE PRECISION GMAT(NIDOF,NDDOF)
INTEGER ERROR
C ... LOCAL VARIABLES
INTEGER I
INTEGER J
C.......................................................................
C
C INITIALIZE TO ZERO
DO I=1,NDDOF
DO J=1,NDDOF
GMAT(I,J) = 0.0D0
ENDDO
ENDDO
C.......................................................................
C
C SET THE MATRIX
GMAT( 3, 3) = 3.9638920267487D-01
GMAT( 9, 3) = 2.7548794986679D-01
GMAT(15, 3) = -6.6741703054081D-02
GMAT(21, 3) = 2.7548794986679D-01
GMAT(27, 3) = 1.9818500443149D-01
GMAT(33, 3) = -1.7277202501352D-02
GMAT(39, 3) = -6.6741703054081D-02
GMAT(45, 3) = -1.7277202501352D-02
GMAT(51, 3) = 2.2487704270919D-02
. .
. .
. .
Note: Many lines have been deleted here for brevity. See the mkgmat.F file in the MD Nastran
delivery for a full listing.
MD Nastran R2.1 Release Guide

302
Step 2. Create a SPLINEX entry
The following listing is the MD Nastran input file. The SPLINEX entry defines the connection between
the structural model, whose grid points are defined by the SWING AECOMP, and the aerodynamic
model, whose grid points are defined by the AWING AECOMP. Note that the AECOMP mechanism give
the user a way to define both grid points and elements. For this example, only grid points are needed.
CONNECT SPLINEX HARD4 EXTSPLN
SOL 144
COMPILE AERO0 LIST NOREF
ALTER 'GI aero'
MATPRN GPGK0 $
CEND
TITLE = Square Wing Test Model
ECHO = NONE
SUBCASE 1
SUBTITLE=Default
SPC = 1
SUPORT1=1
TRIM=1
DISPLACEMENT(SORT1,REAL)=ALL
BEGIN BULK
$
PARAM POST 0
$
EIGRL 1 10
$
GRID 1 0. 0. 0.
GRID 2 5. 0. 0.
GRID 3 10. 0. 0.
GRID 4 0. 5. 0.
GRID 5 5. 5. 0.
GRID 6 10. 5. 0.
GRID 7 0. 10. 0.
GRID 8 5. 10. 0.
GRID 9 10. 10. 0.
$
PSHELL 1 1 .2 1 1
$
CQUAD4 1 1 1 2 5 4
CQUAD4 2 1 2 3 6 5
CQUAD4 3 1 4 5 8 7
CQUAD4 4 1 5 6 9 8
$
MAT1 1 1.07+7 4.053+6 .32 .1
$
SPC1 1 12456 1 2 3
SUPORT1 1 2 3
$
AEROS 0 1 10. 10. 100.
$
CORD2R 1 0 5. 0. 0. 5. 0. 1.
6.0 0.0 0.0
$
PAERO1 1001
$
. .
. .
. .
C.......................................................................
RETURN
END
303
CHAPTER 9
Rotor Dynamics and Aeroelasticity
CAERO1 1001 1001 0 3 3 1
0. 0. 0. 10. 0. 10. 0. 10.
$
$ 01 .. 02 .. 03 .. 04 .. 05 .. 06 .. 07 .. 08 .. 09 .. 10 .
$SPLINEX EID GROUP GCOMP1 GCOMP2 ECOMP1 ECOMP2 USAGE +
$+ AELIST AECOMP AELISTC
SPLINEX 7 HARD4 AWING SWING
$
$ 01 .. 02 .. 03 .. 04 .. 05 .. 06 .. 07 .. 08 .. 09 .. 10 .
$AECOMP NAME LISTTYPELISTID1 LISTID2 ...
AECOMP AWING AELIST 1
AECOMP SWING SET1 1
$
AELIST 1 1001 THRU 1009
$
SET1 1 1 THRU 9
$
AESTAT 1 ANGLEA
AESTAT 2 URDD3
$
TRIM 1 0.2 0.4115 URDD3 1.0
$
ENDDATA
Step 3. Create a CONNECT Statement
See the input file listing in step 2. The CONNECT statement defines spline group HARD4, which is also
referenced on the SPLINEX entry, and an evaluator name of EXTSPLN, which is also used in the eval
connect file (see step 4).
Step 4. Create an Evaluator Connection File
An eval connect file defines where the server program resides. It has the form:
EXTSPLN,-,/home/user/spline_server/spxservb
where EXTSPLN is the evaluator name given in the CONNECT statement in step 3. The - indicates
that the server exists on any network mounted server. This is followed by the name of the file that
contains the compiled server program.
Step 5. Submit an MD Nastran job with the GMCONN keyword
Now that the files have been assembled, the job is submitted using the GMCONN option. For example:
nastran spxsrvb.dat gmconn=evalconnect
where evalconnect is a file that contains the line listed in step 4.
MD Nastran R2.1 Release Guide

304

Вам также может понравиться