MSC - Nastran Numerical Methods User's Guide PDF

MSC.
Nastran 2001
Numerical Methods
Users Guide
by Dr. Louis Komzsik
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: (81) (03) 6911 1200
Fax: (81) (03) 6911 1201
Worldwide Web
www.mscsoftware.com
Disclaimer
This documentation, as well as the software described in it, is furnished under license and may be used only in accordance
with the terms of such license.
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 2004 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.
The software described herein may contain certain third-party software that is protected by copyright and licensed from
MSC.Software suppliers.
MSC, MSC/, MSC., MSC.Dytran, MSC.Fatigue, MSC.Marc, MSC.Patran, MSC.Patran Analysis Manager, MSC.Patran
CATXPRES, MSC.Patran FEA, MSC.Patran Laminate Modeler, MSC.Patran Materials, MSC.Patran Thermal, MSC.Patran
Queue Manager and PATRAN are trademarks or registered trademarks of MSC.Software Corporation in the United
States and/or other countries.
NASTRAN is a registered trademark of NASA. PAM-CRASH 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. ACIS is a registered trademark of Spatial
Technology, Inc. CATIA is a registered trademark of Dassault Systemes, SA. EUCLID is a registered trademark of Matra
Datavision Corporation. FLEXlm is a registered trademark of GLOBEtrotter Software, Inc. HPGL is a trademark of
Hewlett Packard. PostScript is a registered trademark of Adobe Systems, Inc. PTC, CADDS and Pro/ENGINEER are
trademarks or registered trademarks of Parametric Technology Corporation or its subsidiaries in the United States and/or
other countries.Unigraphics, Parasolid and I-DEAS are registered trademarks of Electronic Data Systems Corporation or
its subsidiaries in the United States and/or other countries. All other brand names, product names or trademarks belong
to their respective owners.
C O N T E N T S
MSC.Nastran Numerical Methods Users Guide
MSC.Nastran Numerical Methods
Users Guide
Preface
About this Book, 12
List of MSC.Nastran Books, 14
Technical Support, 15
, 17
Internet Resources, 18
, 19
Permission to Copy and Distribute MSC Documentation, 20
Utility Tools, 2
System Cells, 3
Diagnostic (DIAG) Flags, 6
Matrix Trailers, 7
Indexed Matrix File Structure, 9
Kernel Functions, 10
Timing Constants, 12
Time Estimates, 15
Storage Requirements, 17
Performance Analysis, 18
Parallel Processing, 19
Multiply-Add Equation, 22
Theory of Matrix Multiplication, 23

Method One (Dense x Dense), 23
Method Two (Sparse x Dense), 26
Method Three (Sparse x Sparse), 27
Method Four (Dense x Sparse), 28
Sparse Method, 29
Triple Multiply Method, 30
1
Utility Tools and
Functions
2
Matrix MultiplyAdd Module
Parallel Multiply Method, 33
MPYAD Methods, 34
DMAP User Interface, 36
Method Selection/Deselection, 37
Automatic Selection, 37
Automatic Deselection, 37
User Deselection, 38
User Selection, 39
Option Selection, 40
Diagnostics, 41
Performance Diagnostics, 41
Submethod Diagnostics, 41
Error Diagnostics, 42
MPYAD Estimates and Requirements, 44
Decomposition Process, 46
Theory of Decomposition, 47
Symmetric Decomposition Method, 47
Mathematical Algorithm, 47
Symbolic Phase, 48
Numeric Phase, 49
Numerical Reliability of Symmetric Decomposition, 50
Partial Decomposition, 51
User Interface, 52
Method Selection, 54
Minimum Front Option, 55
Reordering Options, 55
Compression Options, 55
Perturbation Option, 55
High Rank Options, 55
Diagnostic Options, 56
Diagnostics, 57
Numerical Diagnostics, 57
Statistical Diagnostics, 59
Decomposition Estimates and Requirements, 61
References, 62
Solution Process, 64
Theory of Forward-Backward Substitution, 65
3
Matrix
Decomposition
4
Direct Solution of
Linear Systems
Right-Handed Method, 65
Left-Handed Method, 65
Sparse Method, 65
Parallel Method, 66
User Interface, 67
FBS Method Selection, 68
Right-handed FBS Options, 69
Left-handed FBS Option, 69
Diagnostics, 70
Performance Messages, 70
FBS Estimates and Requirements, 72

Sparse FBS Estimates, 72
References, 73
Iterative Solver Solutions, 76

Methods, 76
Theory of the Conjugate Gradient Method, 77

Convergence Control, 77
Block Conjugate Gradient Method (BIC), 78
Real and Complex BIC, 79
Preconditioning Methods, 82
Scaling, 83
Numerical Reliability of Equation Solutions, 83
User Interface, 84
Iterative Method Selection, 87
Preconditioner Options, 88
Convergence Criterion Options, 89
Diagnostic Output Option, 89
Incomplete Cholesky Density Option, 90
Extraction Level Option for Incomplete Cholesky, 90
Recommendations, 90
Iterative Solution Diagnostics, 92

Accuracy Diagnostics, 92
Iterative Solver Estimates and Requirements, 95
References, 97
5
Iterative Solution
of Systems of
Linear Systems
6
Real Symmetric
Eigenvalue
Analysis
Real Eigenvalue Problems, 100
Theory of Real Eigenvalue Analysis, 101

Reduction (Tridiagonal) Method, 102
Solution Method Characteristics, 140
DMAP User Interface, 141

Normalization Options, 145
Frequency and Mode Options, 145
Performance Options, 145
Miscellaneous Options, 147
Mass Matrix Analysis Options, 147
Real Symmetric Eigenvalue Diagnostics, 149

Execution Diagnostics, 149
Table of Shifts, 149
Lanczos Diagnostics, 152
Real Lanczos Estimates and Requirements, 155
References, 156
Damped Models, 160
Theory of Complex Eigenvalue Analysis, 161

Canonical Transformation to Mathematical Form, 161
Dynamic Matrix Multiplication, 165
Physical Solution Diagnosis, 167
Hessenberg Method, 168
QR Iteration Using the Householder Matrices, 172
Eigenvector Computation, 174
The Complex Lanczos Method, 178
The Single Vector Method, 178
The Adaptive, Block Lanczos Method, 188
Singular Value Decomposition (SVD), 196
Solution Method Characteristics, 198
User Interface, 199

Damping Options, 204
Normalization Options, 204
Hessenberg and Lanczos Options, 204
Complex Eigenvalue Diagnostics, 207

Hessenberg Diagnostics, 207
7
Complex
Eigenvalue
Analysis
Complex Lanczos Internal Diagnostics, 207

Complex Lanczos Estimates and Requirements, 213
References, 214
Glossary of Terms
Bibliography
INDEX
MSC.Nastran Numerical Methods Users Guide, 223
Preface
About this Book

List of MSC.Nastran Books
Technical Support
Internet Resources
Permission to Copy and Distribute MSC Documentation
12
About this Book

Nastran is a general-purpose finite element program which solves a wide variety of engineering
problems. MSC.Nastran is developed, marketed, and supported by MSC.Software Corporation.
The MSC.Nastran Numerical Methods User's Guide is intended to help users of the program choose
among the different numerical methods and to tune these methods for optimal performance.
Another goal for this guide is to provide information about the accuracy, time, and space
requirements of these methods.
This edition covers the major numerical methods presented in MSC.Nastran Version 2001. The
main new material in this guide covers parallel execution of the eigenvalue analysis for the fast
solution of normal modes analyses.
The major change of this edition is related to the elimination of the descriptions of old methods
such as the active column decomposition or the inverse power method of eigenvalue analysis.
As in each edition, several known typographical errors from the earlier editions were corrected
and some of the comments were clarified.
This guide will continue to be revised to reflect future enhancements to MSC.Nastran's
numerical methods. Therefore, readers who would like to suggest changes or additions are
encouraged to send them to me.
I would like to thank Wendy Webb for her electronic publishing expertise in producing this
guide. Thanks are also due to Dr. Tom Kowalski for his assistance in preparing and reviewing
the technical content of this edition.
Dr. Louis Komzsik
March 2001
Introduction
This guide is designed to assist the users of MSC.Nastran with the method selection and time
estimations for the most important numerical modules. The guide is separated into seven
chapters:
Utility Tools and Functions on page 1

Matrix Multiply-Add Module on page 21
Matrix Decomposition on page 45
Direct Solution of Linear Systems on page 63
Iterative Solution of Systems of Linear Systems on page 75
Real Symmetric Eigenvalue Analysis on page 99
Complex Eigenvalue Analysis on page 159
Preface
These topics are selected because they have the biggest impact on the performance of
MSC.Nastran. To obtain the most accurate solutions, an analyst should read this guide carefully.
Some of the numerical solutions exhibit different characteristics with different problems. This
guide provides tools and recommendations for the user regarding how to select the best
solution.
Using This Guide

The reader of this guide should be familiar with the basic structure of MSC.Nastran. Experience
with the methods of linear statics and normal modes is helpful. A first-time reader of this guide
should read Chapter 1 to become familiar with the utility tools and functions. After that it is
possible to move directly to the chapters containing the topic the user is trying to apply and tune
(see Chapters 2 through 7). Each chapter contains general time estimates and performance
analysis information as well as resource estimation formulae for some of the methods discussed
in the chapter.
Since this guide also discusses the theory of numerical methods, it is a stand-alone document
except for a few references to the MSC.Nastran Quick Reference Guide.
13
14
List of MSC.Nastran Books

Below is a list of some of the MSC.Nastran documents. You may order any of these documents
from the MSC.Software BooksMart site at www.engineering-e.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
User Modifiable
Toolkit
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, available 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) 67 10 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) (3) 3505 02 66
Fax:
(81) (3) 3505 09 14
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
15
16
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.
MSC.Patran Support
MSC.Nastran Support
MSC.Nastran for Windows Support
MSC.visualNastran Desktop 2D Support
MSC.visualNastran Desktop 4D Support
MSC.Abaqus Support
MSC.Dytran Support
MSC.Fatigue Support
MSC.Interactive Physics Support
MSC.Marc Support
MSC.Mvision Support
MSC.SuperForge Support
MSC Institute Course Information
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
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 market. 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.
Preface
17
18
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.
Simulation Center (simulate.engineering-e.com)
Simulate Online. The Simulation Center provides all your simulation, FEA, and other
engineering tools over the Internet.
Engineering-e.com (www.engineering-e.com)
Engineering-e.com is the first virtual marketplace where clients can find engineering expertise,
and engineers can find the goods and services they need to do their job
MSC.Linux (www.mscsoftware.com/hpc)
Now with almost 40-years of unparalleled experience in scientific and technical computing,
MSC.Software is leveraging this knowledge to deliver its customers state-of-the-art, high
performance computing solutions based on clustered computing for running engineering and
life sciences applications.
CATIASOURCE (plm.mscsoftware.com)
Your SOURCE for Total Product Lifecycle Management Solutions.
Process Architecture Lab (PAL) (pal.mscsoftware.com/services/pal)
PAL is a virtual product development environment that enables PAL participants and
customers to define, validate, and demonstrate advanced tools, processes, and e-business
solutions.
Preface
19
20
Permission to Copy and Distribute MSC Documentation

If you wish to make copies of this documentation for distribution to co-workers, complete this
form and send it to MSC.Software Corporation. MSC will grant written permission if the
following conditions are met:
All copyright notices must be included on all copies.

Copies may be made only for fellow employees.
No copies of this manual, or excerpts thereof, will be given to anyone who is not an
employee of the requesting company.
Please complete and mail to MSC for approval:

Attention: Legal Department
2 MacArthur Place
Santa Ana, CA 92707
Name:_____________________________________________________________
Title: ______________________________________________________________
Company: _________________________________________________________
Address:___________________________________________________________
__________________________________________________________________
Telephone:_________________Email: __________________________________
Signature:______________________________ Date: ______________________
Please do not write below this line.

APPROVED: MSC.Software Corporation
Name:_____________________________________________________________
Title: ______________________________________________________________
Signature:______________________________ Date: ______________________
Page 21
Fold here
_______________________
Place
Stamp
Here
_______________________
_______________________
Attention: Legal Department
2 MacArthur Place
Santa Ana, CA 92707
Fold here
Page 22
CHAPTER
Utility Tools and Functions

Utility Tools
System Cells
Diagnostic (DIAG) Flags
Matrix Trailers
Kernel Functions
Timing Constants
Time Estimates
Storage Requirements
Performance Analysis
Parallel Processing
1.1
Utility Tools
In this chapter the following utility tools are discussed:
System cells
DIAG flags
Matrix trailers
Kernel functions
Timing constants
Since these utilities are of a general nature, they are used in the same way on different
computers and solution sequences. They are also used to select certain numerical methods
and request diagnostics information and timing data. For these reasons, the utility tools
are overviewed here before any specific numerical method is discussed.
CHAPTER 1
1.2
System Cells
One of the most important communication utilities in MSC.Nastran is the SYSTEM
common block. Elements of this common block are called system cells. Some of the system
cells have names associated with them. In those cases, the system cell can be referred to by
this name (commonly called a keyword).
Performance Cells. Some of the system cells related to general performance and timing
issues are
BUFFSIZE
= SYSTEM(1)
HICORE
= SYSTEM(57)
REAL
= SYSTEM(81)
IORATE
= SYSTEM(84)
BUFFPOOL
= SYSTEM(119)
Method Cells. System cells directly related to some numerical methods are
SOLVE
= SYSTEM(69) - mixed
MPYAD
= SYSTEM(66) - binary
FBSOPT
= SYSTEM(70) - decimal
Execution Cells. System cells related to execution types are

SHARED PARALLEL
SPARSE
DISTRIBUTED PARALLEL
USPARSE
The binary system cells are organized so that the options are selected by the decimal values
of the powers of 2. This organization makes the selection of multiple options possible by
adding up the specific option values. The decimal cells use integer numbers. The mixed
cells use both decimal and binary values.
The following several system cells are related to machine and solution accuracy:
MCHEPSS
= SYSTEM(102)
MCHEPSD
= SYSTEM(103)
MCHINF
= SYSTEM(98)
where MCHEPSS and MCHEPD are the machine epsilons for single- and double-precision,
respectively, and MCHINF is the exponent of the machine infinity.
Setting System Cells. The following are several ways a user can set a system cell to a
certain value:
NASTRAN Entry
NASTRAN SYSTEM (CELL) = value

NASTRAN KEYWORD = value
DMAP Program
PUTSYS (value, CELL)

PARAM //STSR/value/ CELL
The first pair of techniques is used on the NASTRAN entry, and the effect of these
techniques is global to the run. The second pair of techniques is used for local settings and
can be used anywhere in the DMAP program; PUTSYS is the recommended way.
To read the value of cell, use:
VARIABLE = GETSYS (TYPE, CELL)
or
VARIABLE = GETSYS (VARIABLE, CELL)
Sparse Keywords. The setting of the SPARSE keyword (SYSTEM(126)) is detailed below:
Value
Meaning
Enable SPMPYAD T and NT
Deselect SPMPYAD NT
Force SPMPYAD NT
Deselect SPMPYAD T
Force SPMPYAD T
Deselect SMPMYAD T and NT
Force SPMPYAD T and NT
Force SPDCMP
16
Force SPFBS
Combinations of values are valid. For example, SPARSE = 24 invokes a sparse run, except
for SPMPYAD.
In the table below, the following naming conventions are used
:
SPMPYAD
SPARSE matrix multiply
SPDCMP
SPARSE decomposition (symmetric)
The default value for SPARSE is 25.
CHAPTER 1
Another keyword (USPARSE) is used to control the unsymmetric sparse decomposition

and FBS. By setting USPARSE = 0 (the default is 1, meaning on), the user can deselect
sparse operation in the unsymmetric decomposition and forward-backward substitution
(FBS).
Shared Memory Parallel Keyword. The SMPARALLEL (or PARALLEL) keyword
controls the shared memory (low level) parallel options of various numerical modules.
The setting of the SMPARALLEL keyword (SYSTEM(107)) is as follows:
Value
1 1023
Meaning
No. of Processors
1024
Deselect FBS
2048
Deselect PDCOMP
4096
Deselect MPYAD
8192
Deselect MHOUS
16384
Unused
32768
Deselect READ
262144
Deselect SPDCMP
524288
Deselect SPFBS
Combinations are valid. For example, PARALLEL = 525314 means a parallel run with two
CPUs, except with FBS methods.
Module Naming Conventions. In the table above, the following naming conventions are
used:
FBS
Forward-backward substitution
PDCOMP
Parallel symmetric decomposition
MHOUS
Parallel modified Householder method
READ
Real eigenvalue module
SPFBS
Sparse FBS
MPYAD
Multiply-Add
SPDCMP
Sparse decomposition
Distributed Parallel Keyword. For distributed memory (high level) parallel processing
the DISTRIBUTED PARALLEL or DMP (SYSTEM (231)) keyword is used. In general, this
keyword describes the number of subdivisions or subdomains (in geometry or frequency)
used in the solution. Since the value of DMP in the distributed memory parallel execution
of MSC.Nastran defines the number of parallel Nastran jobs spawn on the computer or
over the network, its value may not be modified locally in some numerical modules.
1.3
Diagnostic (DIAG) Flags

To request internal diagnostics information from MSC.Nastran, the user can use DIAG
flags. The DIAG statement is an Executive Control statement.
DIAG Flags for Numerical Methods. The DIAG flags used in the numerical and
performance areas are:
DIAG
Print matrix trailers
12
Diagnostics from CEAD
13
Open core length
16
Diagnostics from READ
19
Multiply-Add time estimates
58
Print timing data
For other DIAG flags and solution sequence numbers, see the Executive Control
Statements on page 99 of the MSC.Nastran Quick Reference Guide.
Always use DIAG 8, as it helps to trace the evolution of the matrices throughout the
MSC.Nastran run, culminating in the final matrices given to the numerical solution
modules.
The module-related DIAGs 12, 16, 19 are useful depending on the particular solution
sequence; for example, 12 on SOL 107, 111s, 16 on SOL 103 and 19 on SOL 200 jobs.
DIAG 58 is to be used only at the time of installation and it helps the performance timing
of large jobs.
CHAPTER 1
1.4
Matrix Trailers
The matrix trailer is an information record following (i.e., trailing) a matrix containing the
main characteristics of the matrix.
Matrix Trailer Content. The matrix trailer of every matrix created during an
MSC.Nastran run is printed by requesting DIAG 8. The format of the basic trailer is as
follows:
Name of matrix
Number of columns: (COLS)
Number of rows: (ROWS)
Matrix form (F)
= 1 square matrix
= 2 rectangular
= 3 diagonal
= 4 lower triangular
= 5 upper triangular
= 6 symmetric
= 8 identity matrix
= 10 Cholesky factor
= 11 partial lower triangular factor
= 13 sparse symmetric factor
= 15 sparse unsymmetric factor
Matrix type (T)

= 1 for real, single precision
= 2 for real, double precision
= 3 for for complex, single precision
= 4 for complex, double precision
Number of nonzero words in the densest column: (NZWDS)

Density (DENS)
Calculated as:
number of terms
-------------------------------------------- 10,000
COLS ROWS
Trailer Extension. In addition, an extension of the trailer is available that contains the
following information:
Number of blocks needed to store the matrix (BLOCKS)
Average string length (STRL)

Number of strings in the matrix (NBRSTR)
Three unused entries (BNDL, NBRBND, ROW1)
Average bandwidth (BNDAVG)
Maximum bandwidth (BNDMAX)
Number of null columns (NULCOL)
This information is useful in making resource estimates. In parentheses the notation used
in the DIAG8 printout of the .f04 file is given.
The matrices of MSC.Nastran were conventionally stored as follows:
The matrix header record was followed by column records and concluded with a trailer
record. The columns contained a series of string headers, numerical terms of the string and
optionally a string trailer. The strings are consecutive nonzero terms. While this format
was not storing zero terms, a must in finite element analysis, it had the disadvantage of
storing topological integer information together with numerical real data. Currently, the
following indexed matrix storage scheme is used on most matrices:
Indexed Matrix Structure. An Indexed Matrix is made of three files, the Column, String
and Numeric files.
Each file consists of only two GINO Logical Records:
HEADER RECORD. For the Column file, it contains the Hollerith name of the
data block (NAME) plus application defined words. For the String file, it contains
the combined data block NAME and the Hollerith word STRING. For the
Numeric file, it contains the combined data block NAME and the Hollerith word
NUMERIC.
DATA RECORD. It contains the Column data (see Indexed Matrix Column data
Descriptions) for the Column file, the String data for the String file and the
numeric terms following each other for the Numeric file.
CHAPTER 1
Indexed Matrix File Structure

Column File
String File
Numeric File
Header
Record 0 as written by
application (Data block NAME
+ application defined words)
Data block NAME +

STRING
Data block
NAME +
NUMERIC
*2\1 word(s) per String

Entry
All matrix
numerical
terms following
each
other in one Logical
GINO Record
Data Record
*6\3 words per Column Entry
Word 1\first 1/2 of 1:
Column Number, negative if the
column is null
Word 1\first 1/2 of 1:

Row number of first term in
String
Word 2\second 1/2 of 1:

Number of Strings in Column
Word 2\second 1/2 of 1:

Number of terms in String
Word 3 and 4\2:

String Relative
Pointer to the first String of
Column
Word 5 and 6\3:
Relative Pointer to the first Term
of Column
Note: If null column, then
word(s) 3 to 4\2 points to the
last non null column
String Pointer, word(s) 5 to 6\3
points to the last non null
column Numeric Pointer
*n1\n2 words, where
n1 is the number of words on short word machines

n2 is the number of words on long words machines
10
1.5
Kernel Functions
The kernel functions are internal numerical and I/O routines commonly used by the
functional modules.
Numerical Kernel Functions. To ensure good performance on a variety of computers,
the numerical kernels used in MSC.Nastran are coded into regular and sparse kernels as
well as single-level and double-level kernels. The regular (or vector) kernels execute basic
linear algebraic operations in a dense vector mode. The sparse kernels deal with vectors
described via indices. Double-level kernels are block (or multicolumn) versions of the
regular or sparse kernels.
The AXPY kernel executes the following loop:
Y( i ) = s X( i) + Y( i )
where:
i = 1, 2, , n
s = a scalar
n = the vector length
The sparse version (called AXPl) of this loop is
( ( INDEX ( i ) ) = s X ( i ) ) + Y ( INDEX ( i ) )
where i = 1, 2, , n
In these kernels, X and Y are vectors. INDEX is an array of indices describing the sparsity
pattern of the X vector. A specific MSC.Nastran kernel used in many occasions is the block
AXPY (called XPY2 in MSC.Nastran).
Y ( i, j ) = S ( j ) X ( i, j ) + Y ( i, j )
where:
i = 1, 2, , n
j = 1, 2, , b
Here X , Y are blocks of vectors (rectangular matrices), S is an array of scalar multipliers,
and b is the number of vectors in the block.
Similar kernels are created by executing a DOT product loop as follows:
n
DOT: Sum =
i = 1
X(i) Y(i)
CHAPTER 1
DOT1: Sum =
X ( INDEX ( i ) ) Y ( i )
i = 1
n
DOT2: Sumj =
Xi, j Yi, j
i = 1
where:
b = the block size
j = 1, 2, , b
Indexed versions of the XPY2 and DOT2 kernels also exist.
To increase computational granularity and performance on hierarchic (cache) memory
architectures, the heavily used triangular matrix update kernels are organized in a triple
loop fashion.
The DFMQ kernel executes the following mathematics:
A = A + uv
where A is a triangular or trapezoidal matrix (a portion of the factor matrix) and u, v are
vectors.
The DFMR kernel executes a high rank update of form
A = A + UV
where now U and V are rectangular matrices. All real, complex, symmetric, and
unsymmetric variations of these kernels exist, but their description requires details beyond
the scope of this document.
Triple Look Kernels. Additional triple loop kernels are the triple DOT (DOT3) and
SAXPY (XPY3) routines. They are essentially executing matrix-matrix operations. They
are also very efficient on cache-machines as well as very amenable to parallel execution.
I/O Kernel Functions. Another category of kernels contains the I/O kernels. The
routines in this category are invoked when a data move is requested from the memory to
the I/O buffers.
Support Kernels. Additional support kernels frequently used in numerical modules are
ZEROC, which zeroes out a vector; MOVEW, which moves one vector into another; and
SORTC, which sorts the elements of a vector into the user-requested (ascending or
descending) order.
11
12
1.6
Timing Constants
Single Loop Kernel Performance. Timing constants are unit (per term) execution times
of numerical and I/O kernels. A typical numerical kernel vector performance curve shows
unit time T as a function of the loop length. A loop is a computer science structure that
executes the same operation repeatedly. The number of repetitions is called the loop
length.
Unit
Time: T
Figure 1-1
...
1024
Loop (vector)
Length s
Single-Level Kernel Performance Curve
The kernel performance curve can be described mathematically as

B
T = A + --s
Eq. 1-1
where the constant A is characteristic of the asymptotic performance of the curve since
T(s = ) = A
Eq. 1-2
The constant B represents the startup overhead of the loop as

T(s = 1) = A + B
Eq. 1-3
These constants for all the MSC.Nastran numerical kernels can be printed by using
DIAG 58.
Sometimes it is impossible to have a good fit for the datapoints given by only one curve.
In these cases, two or more segments are provided up to a certain break point in the
following forma:
CHAPTER 1
X Segments for Kernel Y

Segment 1
Segment 2
Break Point
Break Point
A1
A2
B1
B2
where X is the number of segments and Y is the name of the particular kernel.
Double Loop Kernel Performance. In the case of the double loop kernels, the unit time
is a function of both the inner loop length and the number of columns, which is the outer
loop length. The unit time is described by a surface as shown in Figure 1-2.
Unit Time: T
Outer Loop Curves
Outer Loop Length
Inner Loop Curves
1
2
.
.
1024
Inner Loop Length
Figure 1-2
Double-Level Kernel Performance Surface
The surface on Figure 1-2 is built from curves obtained by fixing a certain outer loop length
and varying the inner loop length. Intermediate values are found by interpolation.
Another set of curves is obtained by fixing the inner loop length and varying the outer loop
length.
I/O Kernels. There are many I/O kernels also in MSC.Nastran.
The unit time for these kernels for string or term operations is
T s = number of strings A + number of nonzeroes B
For column operations (PACK, UNPACK),
Eq. 1-4
13
14
T c = T s + rows columns A
Eq. 1-5
and the two values are given for real and complex A and B values.
Triple Loop Kernels. The triple loop kernels are now included in the time estimate
(GENTIM) process of MSC.Nastran.
While difficult to show diagramatically, the timing model for the triple loop kernels can be
thought of as families of double loop surfaces as shown in Figure 1-2. A family is generated
for specified lengths of the outermost loop. Values intermediate to these specified lengths
are determined by interpolation.
CHAPTER 1
1.7
Time Estimates
Calculating time estimates for a numerical operation in MSC.Nastran is based on analytic
and empirical data. The analytic data is an operation count that is typically the number of
multiply (add) operations required to execute the operation. In some cases the number of
data movements is counted also.
The empirical data is the unit time for a specific kernel function, which is taken from the
timetest tables obtained by DIAG 58 and explained in Timing Constants on page 12.
These tables are generated on the particular computer on which the program is installed
and stored in the database.
The operation count and the execution kernel length are calculated using information
contained in the matrix trailers. Sometimes trailer information from the output matrix
generated by the particular operation is required in advance. This information is
impossible to obtain without executing the operation. The parameters are then estimated
in such cases resulting in less reliable time estimates.
Available Time. Time estimates in most numerical modules are also compared with the
available time (TIME entry). Operations are not started or continued if insufficient time is
available.
I/O time estimates are based on the amount of data moved (an analytic data item) divided
by the IORATE and multiplied by the I/O kernel time. Since the user can overwrite the
default value of the IORATE parameter, it is possible to increase or decrease the I/O time
calculated, which also results in varying the method selection.
In most numerical modules, MSC.Nastran offers more than one solution method. The
method used can be selected by the user. The method automatically selected by
MSC.Nastran is based on time estimates. The estimated (CPU) execution time is calculated
by multiplying the number of numerical operations by the unit execution time of the
numerical kernel executing the particular operation. In addition, an estimation is given for
the (I/O) time required to move information between the memory and secondary storage.
After the estimates for the CPU execution time and the l/O time are added together,
MSC.Nastran selects the method that uses the least time.
Matrix Methods. Several methods are offered because each of them is best suited to
certain types of matrices. The difference in cost among the methods for specific cases can
be an order of magnitude or more. As each matrix is generated, the parameters describing
its size and the distribution of nonzero terms are stored in a matrix trailer. (The parameters
that define the properties of the matrices were described in Matrix Trailers on page 7.)
For each matrix, these parameters include the number of rows and columns, the form (for
example, square or symmetric), the type (for example, real or complex), the largest number
of nonzero words in any column, and the density. Some of the newer methods also record
the number of strings in the matrix. Other descriptive parameters may be added in the
future.
15
16
The only empirical data used in deriving the timing equations is the measurement of the
time per operation for the kernels. These measurements are computed at the time of
installation on each computer and are stored in the delivery database for later use. After
the system is installed, the measurements may be updated if faster hardware options are
installed on the computer. The remaining terms in the equations are derived from careful
operation counts, which account for both arithmetic and data storage operations.
Timing Equations. Timing equations are derived for all major numerical modules.
Conservative upper bounds are the best estimates that can be calculated. At present, these
estimates are not used for method selection. Instead, the user is required to input the total
amount of available CPU time to solve the total run. The amount of time remaining at the
start of the numerical solution modules is compared with the estimate. The run is
terminated before the numerical module starts execution if the amount of time remaining
is less than the estimate. The goal is to minimize wasted computer resources by
terminating expensive operations before they start rather than terminating them midway
through their operation before any output is available.
The many types of machine architecture which MSC.Nastran supports and the great
differences in operation between scalar, vector, and parallel computing operations result
in a challenge to the numerical method developers to provide correct estimation and
method selection. There are a number of diagnostic tools which can be used to print out
the estimates and the other parameters affecting computation cost. These tools are
generally activated by the DIAG flags described earlier.
CHAPTER 1
1.8
Storage Requirements
Main storage in MSC.Nastran is composed of the space used for the code, the space used
for the Executive System, and the actual working space used for numerical operations.
Working Space = DIAG 13. The actual working space available for a numerical operation
can be obtained using DIAG 13.
Disk storage is needed during the execution of an MSC.Nastran job to store temporary
(scratch) files as well as the permanent files containing the solution.
Memory Sections. The Executive System provides the tools needed to optimize the
execution using a trade-off between memory and disk. The main memory is organized as
follows:
Working Storage
Printed on DIAG 13
Executive
RAM
MEM
User-Controllable
BUFFPOOL
RAM, MEM, BUFFPOOL. The RAM area holds database files, while the SMEM area holds
scratch files. The BUFFPOOL area can act as a buffer memory. The user-selectable sizes of
these areas have an effect on the size of the working storage and provide a tool for tuning
the performance of an MSC.Nastran job by finding the best ratios.
A general (module-independent) user fatal message associated with storage requirements
is:
UFM 3008:
INSUFFICIENT CORE FOR MODULE XXXX
This message is self explanatory and is typically supported by messages from the module
prior to message 3008.
17
18
1.9
Performance Analysis
.f04 Event Statistics. The analysis of the performance of an MSC.Nastran run is
performed using the .f04 file.
Disk Usage. The final segment of the .f04 file is the database usage statistics. The part of
this output most relevant to numerical modules is the scratch space usage (the numerical
modules are large consumers of scratch space). SCR300 is the internal scratch space used
during a numerical operation and is released after its completion. The specific SCR300
table shows the largest consumer of internal scratch space, which is usually one of the
numerical modules. The output HIWATER BLOCK shows the maximum secondary
storage requirement during the execution of that module.
Memory Usage. Another table in this final segment shows the largest memory usage in
the run. The output HIWATER MEMORY shows the maximum memory requirement
combining working storage and executive system areas, described in Storage
Requirements on page 17.
CHAPTER 1
1.10
Parallel Processing
Parallel processing in MSC.Nastran numerical modules is a very specific tool. It is very
important in enhancing performance, although its possibilities in MSC.Nastran and in
specific numerical modules are theoretically limited.
The parallelization possibilities in MSC.Nastran consist of three different categories:
High level
Frequency domain
Medium level
Geometry domain
Low level
Block kernels (high rank updates)
The currently available methods of parallel processing in MSC.Nastran numerical modules
are:
Shared memory parallel

Medium, low level
MPYAD, DCMP, FBS modules
Distributed memory parallel

High, medium level
SOLVIT, DCMP, FBS, READ modules
Details of the various parallel methods are shown in the appropriate Modules sections
throughout.
19
20
CHAPTER
Matrix Multiply-Add Module

Multiply-Add Equation
Theory of Matrix Multiplication
MPYAD Methods
DMAP User Interface
Method Selection/Deselection
Option Selection
Diagnostics
MPYAD Estimates and Requirements
22
2.1
Multiply-Add Equation
The matrix multiply-add operation is conceptually simple. However, the wide variety of
matrix characteristics and type combinations require a multitude of methods.
The matrix multiply-add modules (MPYAD and SMPYAD) evaluate the following matrix
equations:
(MPYAD)
[ D ] = [ A ]
(T)
[B] [C]
Eq. 2-1
or
(SMPYAD)
[G] = [A]
(T)
[B]
(T)
[C] [D] E F
Eq. 2-2
The matrices must be compatible with respect to the rules of matrix multiplication. The ( T )
stands for (optional) transpose. The signs of the matrices are also user parameters. In
Eq. 2-2, any number (between 2 and 5) of input matrices can be present.
The detailed theory of matrix multiply-add operation is described in Theory of Matrix
Multiplication on page 23. Subsequent sections provide comprehensive discussions
regarding the selection and use of the various methods.
CHAPTER 2
2.2
Theory of Matrix Multiplication

The matrix multiplication module in MSC.Nastran evaluates the following matrix
equations:
D = [ A ] [ B ] [ C ]
Eq. 2-3
or
T
D = [ A ] [ B ] [ C ]
where A , B , C , and D are compatible matrices. The calculation of Eq. 2-3 is carried out by
the following summation:
n
d ij =
a ik b kj c ik
Eq. 2-4
k = 1
where the elements a , b , c , and d are the elements of the corresponding matrices, and n
is the column order of matrix A and the row order of matrix B . The sign of the matrices
and the transpose flag are assigned by user-defined parameters.
MSC.Nastran has four major ways to execute Eq. 2-3 and performs the selection among the
different methods automatically. The basis of the selection is the density pattern of
matrices A and B and the estimated time required for the different kernels.
Generally, these methods are able to handle any kind of input matrices (real, complex,
single, or double precision) and provide the appropriate result type. Mixed cases are also
allowed and are handled properly.
The effective execution of multiplication is accomplished by invoking the MSC.Nastran
kernel functions.
The four methods are summarized in the following table and explained in more detail
below.
Method
Combination
Dense Dense
Sparse Dense
Sparse Sparse
Dense Dense
Method One (Dense x Dense)

Method one consists of several submethods. The submethod designated as method one
storage 1 is also known as basic method one.
23
24
In basic method one, enough storage is allocated to hold as many non-null columns of
matrices B and D as memory allows. The columns of matrix C corresponding to the nonnull columns of B are initially read into the location of matrix D (the result). In basic
method one, matrix A is processed on a string-by-string basis. The complete multiplication
operation may take more than one pass when all the non-null columns of matrices B and
D cannot fit into memory. The number of passes can be calculated as follows:
N
N p = ------NB
Eq. 2-5
where:
N = order of problem
N p = number of passes
N B = number of non-null columns of B in memory
The basic procedure of method one (storage 1) can be viewed as follows:
D = C
j
a il b j
or
a li b j
a il
a li
bj
i
bN
b1
NB
NB
In Memory
In Memory
Figure 2-1
Method One
Note: The underlined quantities in Figure 2-1 represent vectors.

For each nonzero element of A , all combinatorial terms of B currently in memory are
multiplied and accumulated in D . N B columns of matrix D are calculated at the end of
one complete pass of matrix A through the processor. Then the completed columns of D
are packed out, along with any non-null columns of C that correspond to null columns of
B skipped in this pass (as they are columns of D ). This part is saved, N B non-null columns
of B and the corresponding columns of C are loaded, and the process continues. The
effective execution of the multiplication depends on whether or not the transpose flag is
set.
CHAPTER 2
Nontranspose:
d ij = a il b lj + d ij
Eq. 2-6
d ij = a li b lj + d
ij
Eq. 2-7
Transpose:
The other submethods provide for different handling of matrix A and for carrying out the
multiplication operations. The main features of the submethods vary depending on the
different ways of handling the strings (series of consecutive nonzero terms in a matrix
column). A comparison of the main method and the submethods is shown as follows:
Table 2-1 Nontranspose Cases
Storage A:
Unpacked columns of B and D

Processing A string by string
A is in the inner loop
Storage B:

A is in the outer loop
Storage C:
Unpacked partial rows of B

Storage D:
Partial rows of B in string format

Storage E:

Unpacked columns of A (band only)
Storage F:
Partial rows of B in string format

Unpacked columns of A
Storage 2:
Unpacked non-null columns of B and D

Loops are the same as storage 1, except
outermost loop pulled inside the kernel
(triple loop kernel)
25
26
Table 2-2 Transpose Cases

Storage A:

Storage B:
Unpacked columns of B
Partial rows of D
Storage C:

Unpacked rows of A (band only)
Storage D:
Unpacked columns of B
Partial rows of D
Unpacked rows of A
Storage 2:
Unpacked non-null columns of B and

D
Loops are the same as storage 1, except
outermost loop pulled inside the kernel
(triple loop kernel)
The effective execution of the multiplication operation in method one subcases (and other
methods with the exception of method three) is accomplished by involving the
MSC.Nastran kernel functions. In method one submethods, except storage 2, the double
loop kernels of DOT2 and XPY2 are used. In storage 2, the triple loop kernels of DOT3 are
used.
Depending on whether the length of the current string of A is N or M , the A string is in
the inner loop or in the outer loop. This explains the comments: " A is in the inner loop" or
" A is in the outer loop" in Table 2-1. The selection between the previous two possible
usages of the kernel functions depends on the density of matrices A and B . If A is sparse,
it is in the inner loop; otherwise, A is in the outer loop.
Method Two (Sparse x Dense)

In the nontranspose case of this method, a single term of B and one full column of the
partially formed D are in the main memory. The remaining main memory is filled with as
many columns of A as possible. These columns of matrix A are in string format. This
method is effective when A is large and sparse; otherwise, too many passes of B must be
taken. The number of passes in B is calculated from Eq. 2-5.
CHAPTER 2
The method can be graphically represented as follows:

A
D = C
j
b kj a j
b kj
NB
NA
In Memory
b kj In Memory
Figure 2-2
k-th Column In Memory
Method Two
When b kj is in memory, the k-th column of A is processed against it and the result is
accumulated into the k-th column of D . In the transpose case, one column of B is held in
memory while D holds only a single term at a time. This method provides an alternative
means of transposing matrix A by using the identity matrix B and the zero matrix C when
the transpose module of MSC.Nastran is inefficient.
Method Three (Sparse x Sparse)

In method three, the transpose and nontranspose cases are essentially the same except for
the initial transpose of matrix A in the transpose case. In both cases, matrix A is stored in
the same way as in method two, i.e., in string format. Matrix B is processed on an
element-by-element basis, and the products of each b kj term are calculated using the
corresponding terms of A in memory. However, in method three, the results and storage
are different from method two. In method three, a storage bin is set for the columns of
matrix D . The number of these bins is based on the anticipated density of matrix D and is
calculated as follows:
Number of bins = N
Eq. 2-8
where:
N = order of the problem
= density of D (estimated)
The size of the bins can be calculated as follows:
N
1
Size of bins = ------------------------------------- = --Number of bins
Eq. 2-9
This manner of storing the results takes advantage of the sparsity of matrix B . However,
it requires sorting in the bins before packing out the columns of D .
27
28
Method Four (Dense x Sparse)

This method has two distinct branches depending on the transpose flag.
Nontranspose Case
First, matrix B is transposed and written into the results file. This operation is performed
with the assumption that B is sparse. As many columns of A as possible are unpacked into
memory and the columns of B T (rows of B ) are interpreted on a term-by-term basis.
ak
b kj
NA
b kj a j
In Memory
b kj In Memory
Figure 2-3
NA
In Memory
In Scratch
Nontranspose Method Four
For each nonzero term of B , the scalar product with the columns of A is formed and
written into the scratch file. When all columns of A and rows of B are processed, the
scratch file contains one column for each nonzero term in B . Therefore, a final pass must
be made to generate matrix D .
Transpose Case
T
In this case, a set of rows of A (columns of A ) are stored in an unpacked form. These rows
build a segment. Matrix B is processed on a string-by-string basis, providing the DOT
products of the strings and the columns of the segments.
CHAPTER 2
a1
b1
bn
a1 b1
a1 bn
ar b1
ar bn
ar
Figure 2-4
Transpose Method Four
The results are sequentially written into the scratch file continuously. The structure of this
file is as follows:
CoI 1
CoI 2
Seg 1
Seg 1
...
CoI n
CoI 1
Seg 1
Seg 2
...
CoI n EOF
Seg n
The number of segments is

n
k = --r
Eq. 2-10
Finally, the product matrix must be assembled from the scratch file, and matrix C , if
present, must be added.
Sparse Method
The sparse multiply method is similar to regular method three. When the transpose case
is requested, matrix A is transposed prior to the numerical operations. This step is
typically not very expensive since A is sparse when this method is selected.
The significance of this method is that matrix A is stored in a new sparse form. Specifically,
all nonzero terms of a column are stored in a contiguous real memory region, and the row
indices are in a separate integer array. The sparse kernel AXPI is used for the numerical
work. The scheme of this method is shown on the following figure.
29
30
A, A
B
j
+C
In Memory
B kj
a ij +
d ij =
Figure 2-5
d ij
partial
Sparse Method
From the figure it is clear that matrix B is processed on an element-by-element basis. When
all the nonzero terms of A do not fit into memory at once, multiple passes are executed on
matrix B , and only partial results are obtained in each pass. In turn, these results are
summed up in a final pass.
The sparse method is advantageous when both matrices are sparse.
Triple Multiply Method

In MSC.Nastran a triple multiplication operation involving only two matrices occurs in
T
several places. The most common is the modal mass matrix calculation of M . Note
that in this multiplication operation the matrix in the middle is symmetric. Therefore, the
result is also symmetric. No symmetry or squareness is required for the matrices on the
sides. Historically, this operation was performed by two consecutive matrix
multiplications which did not take advantage of the symmetry or the fact that the first and
third matrices are the same.
The operation in matrix form is
T
C = [ A] [B ][A ] D
where:
A = order of m n ( m columns n rows )
B = n n symmetric matrix
Eq. 2-11
CHAPTER 2
C = m m symmetric matrix
D = m m symmetric matrix
Any element of matrix C can be calculated as follows:
n
C l, k = a jk a il b ij + d l, k
i = k
j = 1
n
Eq. 2-12
where:
k = the column index, 1 k m
l = the row index, 1 l m
It can be proven by symmetry that
n
n
n
a
b
=
a
a
b
lj im ij jm il ij
j = 1
j = 1
i = 1
i = j
n
Eq. 2-13
Based on the equality in Eq. 2-13, a significant amount of work can be saved by calculating
only one-half of matrix C .
When designing the storage requirements, advantage is taken of the fact that matrix B is
only needed once to calculate the internal sums. Based on this observation, it is not
necessary to have this matrix in the main memory. Matrix B can be transferred through
the main memory using only one string at a time.
T
The main memory is equally distributed among A , A B , and three vector buffers. One
of the vector buffers must be a full column in length. Therefore, the structure of the main
memory is as follows:
31
32
Vector
Buffer
Area 1
Area 2
k Columns of A
k Rows of A B
One Column
Long
k -Length
Buffers
I/O Buffers
Figure 2-6
Memory Distribution for Triple Multiply
Three l/O buffers must be reserved for the three simultaneously open files of A , B , and the
scratch file containing partial results. Therefore, the final main memory size is
n z = 2k n + n + 3 BUFFSIZE + 2k
Eq. 2-14
From Eq. 2-14, the number of A columns fitting into memory can be calculated as follows:
n z ( 3 BUFFSIZE ) n
k = -----------------------------------------------------------------2n + 2
Eq. 2-15
The number of passes is calculated as follows:

m m
---- if ---- = integer
k
k
p =
m
int m
---- + 1 if ---- integer
k
k
Eq. 2-16
which is equal to the number of times the triple multiply operation reads through the
matrix B .
The number of times the triple multiply operation reads through the matrix A can be
approximated as follows:
p
p A = --2
Eq. 2-17
The triple multiply method is implemented with a block spill logic where the result of the
matrix C is generated in k k blocks.
CHAPTER 2
Parallel Multiply Method

The parallel multiply method, which is only used when parallel processing is requested, is
basically a parallel version of method one designed to solve the CPU-intensive
multiplication of dense matrices.
The storage structure of this method is the same as that of method one. However, the
columns of matrix B are distributed among the processors. Consequently, although B is
stored in the main (shared) memory, the processors access different portions of it.
At the beginning of each pass, subtasks are created. These subtasks wait for the columns
of matrix A to be brought into memory. Once a column of A is in the main memory, all
subtasks process this column with their portion of matrix B . When all of the subtasks are
finished, the main processor brings in a new column of A and the process continues. The
parallel method is advantageous when matrix B is very dense and multiple processors are
available.
NT
C, D
B
CPU
1
2 3
In Memory
CPU
1
2 3
In Memory
Figure 2-7 Parallel Multiply Method

Algorithms 1TC and 1NTE (see MPYAD Methods on page 34) are executed in parallel
when parallel MPYAD is requested. The parallel methods are represented in Figure 2-7.
The triple loop kernels used in Method 1 Storage 2 are also parallelized by some machine
vendors providing another way to execute parallel MPYAD.
33
34
2.3
MPYAD Methods
The methods in the MPYAD module are divided into six main categories: four methods
for different density combinations, one method for parallel execution, and one method for
sparse operations. These methods are summarized in the following table:
Method
Combination
Dense Dense
Sparse Dense
Sparse Sparse
Dense Sparse
Dense Dense
Sparse Sparse
There are no fixed density boundaries set for these methods. Method selection is a complex
topic. Within method one, there are also ten submethods for the special handling of the
matrices.
Methods two and three are only selected in special cases. In most cases, the sparse method
replaces both methods two and three.
The parallel multiply method is aimed at shared memory parallel computers. It does not
run on distributed memory parallel computers.
The method 1 submethods (A-F) are automatically deselected in some cases. As example is
when the A and B matrices are the same, or another when any of the matrices are nonmachine precision.
MPYAD Method Identifiers. For selection and deselection purposes, identifiers are
assigned to certain methods. However, these identifiers are bit oriented, and in some cases
their decimal equivalents are used:
Method
Bit
Decimal
1NT
1T
2NT
2T
3NT
16
3T
32
4NT
64
4T
128
CHAPTER 2
Method
Bit
Decimal
1NTA
256
1NTB
512
1NTC
10
1024
1NTD
11
2048
1NTE
12
4096
1NTF
13
8192
1TA
14
16384
1TB
15
32768
1TC
16
65536
1TD
17
131072
SNT
18
262144
ST
19
524288
Deselect
20
1048576
DIAG
21
2097152
22
4194304
1NT2
23
8388608
1T2
24
16777216
AutoS2
25
33554432
In the above table, T represents transpose, NT indicates nontranspose, and A, B, C, D, E, F,

2 are submethod names when they appear as the last character. For example, 1NTD is a
method one, nontranspose case, D submethod operation.
Bit 21 (with the corresponding decimal value of 2097152) is reserved for submethod
diagnostics.
35
36
2.4
DMAP User Interface

The DMAP call for the MPYAD module executing the operation in Eq. 2-1 is
MPYAD
A,B,C/D/T/SIGNAB/SIGNC/PREC/FORM
where:
T = 0,1: Non-transpose or transpose (see Multiply-Add Equation on page 22)
PREC = 0,1,2: Machine, single, double, etc. (see Matrix Trailers on page 7)
FORM = 0,1,2: Auto, square, rectangular, etc. (see Matrix Trailers on page 7)
An alternative MPYAD call is to use the SMPYAD module as follows
:
SMPYAD
A,B,C,D,E,F/G/N/SIGNG/SIGNF/PREC/TA/TB/TC/TD/FORM
where:
N = number of matrices given
TA,TB,TC,TD = transpose flags
FORM = as above
This module executes the operation in Eq. 2-2.
CHAPTER 2
2.5
Method Selection/Deselection
MPYAD automatically selects the method with the lowest estimate of combined cpu and
I/O time from a subset of the available methods. Also, those methods that are
inappropriate for the users specific problem are automatically deselected. The user can
override the automatic selection and deselection process by manual selection, subject to
certain limitations. The details of both automatic and user selection and deselection criteria
are described below.
Automatic Selection
By default, methods 1 (all submethods), 3, 4, and Sparse are available for automatic
selection. If bit 25 of System Cell 66 has been set (decimal value 33554432), method 2 is
included in the automatic selection set of methods. Also, if all of the default methods have
been deselected, method 2 will be used, provided it was not deselected. If all methods have
been deselected, a fatal error occurs.
Automatic Deselection
If a method is determined to be inappropriate for the users problem, it will be
automatically deselected. Except in those cases noted below, an automatically deselected
method will be unavailable for either automatic selection or manual user selection. Note
that any method that has insufficient memory to execute the users problem will be
deselected. The other automatic deselection criteria are described below for each method.
In this discussion "mpassI" stands for the number of passes required by method I. A and
B represent the density of the A matrix and B matrix, respectively. Also, unless the
method name is qualified by NT (non-transpose) or T (transpose), the criteria applies to
both.
Method 1 All Submethods
If method S is not deselected and mpass1 is greater than

5 mpass3 and A and B are less than 10%.
If MPYAD was called by the transpose module.
Method 1 Storage A-F
If the type of matrix A is real and either B or C is

complex.
If the type of matrix A is complex and both B and C are
real.
Method 1NT Storage A, D,

and F
Unless explicitly user selected.
Method 1T Storage B
and C
1TB unless 1TA deselected.

1TD unless 1TC is deselected.
Method 2
If method S is not deselected, unless user selected.
37
38
If the C matrix is non-null and the type of C is not

machine precision or the A , B , and C matrices are not all
real or all complex.
If MPYAD was called by the transpose module.
Method 3
If method S is not deselected, unless user selected.

If matrix A is real and the B and/or C matrix is complex.
Method 3NT
If the requested type of the D matrix is real and any of

the input matrices are complex, or if the requested type of
D is complex and all of the input matrices are real.
If methods 1, 2, 3, and S are not deselected and A
greater than 10%, unless method 4 has been user selected.
Method 4
If matrix C is non-null and its type is not equal to

machine precision or the B and C matrices are not both
real or not both complex.
Method 4T
If more than 100 "front-end" (R4) passes or more than 10

"back-end" (S4) passes.
Method Sparse
If the type of matrix B is not machine precision.

If matrix C is non-null and A and C are not both real or
both complex.
If matrix A is complex and matrix B is real.
Method Parallel NT
Unless 1NTE is not deselected.

If the number of columns per pass for 1NTE is less than
the number of available processors.
Method Parallel T
Unless 1TC is available.

If the number of columns per pass for 1TC is less than the
number of available processors.
User Deselection
For method deselection, the following is required:
Main Methods Deselection
SYSTEM(66) = decimal value of method
Submethods Deselection
SYSTEM(66) = decimal value of submethod
CHAPTER 2
Sparse Method Deselection

SYSTEM(126) = 0: Deselect all sparse methods
SYSTEM(126) = 2: Deselect sparse NT only
SYSTEM(126) = 4: Deselect sparse T only
where SYSTEM (126) is equivalent to the SPARSE keyword.
Parallel Method Deselection

SYSTEM(107) = 0: Deselect all parallel modules
SYSTEM(107) = 4096 + ncpu: Deselect parallel MPYAD only
where ncpu = number of CPUs and SYSTEM (107) is equivalent to the PARALLEL.
keyword.
The philosophy of method selection is to deselect all the methods except for the one being
selected.
Triple Loop Method Deselection
SYSTEM(252) = 0 or > 100: Do not use triple loops in 1T,1NT
User Selection
For method selection, the following is required:
Main Methods Selection
SYSTEM(66) = 255 decimal value of method identifier
Main or Submethods Selection
SYSTEM(66) = 1048576 + bit value of method or submethod identifier
Sparse Method Selection
SYSTEM(126) = 1: Auto selection (This is the default.)
SYSTEM(126) = 3: Force sparse NT method
SYSTEM(126) = 5: Force sparse T method
SYSTEM(126) = 7: Force either T or NT sparse
Parallel Method Selection
SYSTEM(107) > 0 and
SYSTEM(66) = 1048592(T) or 1048588(NT)
39
40
2.6
Option Selection
The following table shows the type combination options that are supported (R stands for
real, C for complex). These options are automatically selected based on matrix trailer
information. When the user selects one particular method with an option not supported
with that method, an alternate method is chosen by MPYAD unless all of them are
deselected.
Method
RR+R
CC+C
RC+R
RC+C
1T
YES
YES
YES
YES
1NT
YES
YES
YES
YES
2T
YES
YES
YES
YES
2NT
YES
YES
NO
NO
3T
YES
YES
NO
NO
3NT
YES
YES
NO
NO
4T
YES
YES
NO
YES
4NT
YES
YES
NO
NO
YES
YES
NO
NO
YES
YES
NO
NO
CHAPTER 2
2.7
Diagnostics
The MPYAD module outputs diagnostic information in two categories: performance
diagnostics and error diagnostics.
Performance Diagnostics
DIAG 19 Output. The following performance diagnostics is received by setting DIAG 19.
M MATRIX Trailer(COLS ROWS FORM TYPE
A
NZ DENS)
METHOD 1 Passes = XX CPU = XX

I/O = XX Total = XX
MATRIX BTrailer(COLS ROWS FORM TYPE

NZ DENS)

I/O = XX Total = XX
Y MATRIX Trailer(COLS ROWS FORM TYPE

C
NZ DENS)

I/O = XX Total = XX
A Working Memory = XX SYSTEM (66) = XX

I/O = XX Total = XX
D Transpose Flag = XX SYSTEM (126) = XX
METHOD S Passes = XX CPU = XX

I/O = XX Total = XX
Figure 2-8
Excerpt from the DIAG19 Output
In the above figure, passes means the number of partitions needed to create the result
matrix.
To prevent creating huge .f04 files when many MPYAD operations are executed,
MSC.Nastran has a machine-dependent time limit stored in SYSTEM(20). When a time
estimate is below this value, it is not printed. To print all time estimates, the user should
set SYSTEM(20) = 0.
Most of the diagnostics information mentioned in the above table is self explanatory.
Notice the presence of the MPYAD keyword (SYSTEM(66)) used to verify the method
selection/deselection operation.
Whenever a method is deselected, its time estimate is set to 999999.
Submethod Diagnostics
For special diagnostics on the submethods, the user must add 2097152 to the value of
SYSTEM(66) (i.e., turn on bit 21). The format of this diagnostic is shown in Table 2-3. The
last four columns contain the appropriate times.
Table 2-3 Method One Submethods
NEW1 = B
DESELECT
NCPP
PASSES
KERNEL
CPU
I/O
TOTAL
YES
NO
41
42
Table 2-3 Method One Submethods (continued)

NEW1 = B
DESELECT
NCPP
PASSES
KERNEL
CPU
I/O
TOTAL
NO
YES
YES
YES
YES
where:
NCPP = number of columns per pass
NEW1 = B indicates that submethod B is chosen
Error Diagnostics
Error messages are abbreviated as follows:
UFM
User Fatal Message
SFM
System Fatal Message
UWM
User Warning Messages
SWM
System Warning Messages
UIM
User Information Messages
SIM
System Information Messages
The following error-related messages may be received from MPYAD:

UFM 3055:
AN ATTEMPT TO MULTIPLY NONCONFORMABLE MATRICES.
The message is given if the number of columns of A is not equal to the number of rows in B,
the number of rows of C is not equal to the number of rows of A, or the number of columns
of C is not equal to the number of columns of B. This message is also given when MPYAD
is called from another module.
SFM 5423:
ATTEMPT TO MULTIPLY INCOMPATIBLE MATRICES.
The cause for this message is the same as for UFM 3055. However, this message is more
elaborate and prints the trailers for all matrices involved. This message comes from the
MPYAD module.
UFM 6199:
INSUFFICIENT CORE AVAILABLE FOR MATRIX MULTIPLY.
CHAPTER 2
This message results while using the sparse multiply method when the storage estimate
based on the trailer information is exceeded during the actual execution of the operation.
43
44
2.8
MPYAD Estimates and Requirements

The CPU time estimate for the sparse multiply-add method is based on the following input
matrix characteristics:
[ A ] : ( m n, [ B ] ) : ( n p, [ C ] : m p, A )
Eq. 2-18
Computation time (sec):

m n p A M
Eq. 2-19
n p p P n pass + m n A P * + ( 2 ) ( m p C, ( D ) P * )
Eq. 2-20
Data move time (sec):
where:
m n A
---------------------------------------W ( IPREC + 1 )
n pass =
W = workspace available in words

P * = one of either P s , P , or P i depending on the particular methods used
IPREC = machine precision (1 for short-word machines, 2 for long-word machines)
* = density of * matrix
Note:
M, P, and P * are defined in the Glossary of Terms.
The minimum storage requirements are as follows:

Disk:
Memory:
m n A + n p B + ( 2 )m p D
2 ( ( n + m ) IPREC )
CHAPTER
Matrix Decomposition
Decomposition Process
Theory of Decomposition
User Interface
Method Selection
Option Selection
Diagnostics
Decomposition Estimates and Requirements
References
46
3.1
Decomposition Process
The decomposition operation is the first step in solving large linear systems of equations.
For symmetric matrices:
[A] = [L][D][L]
Eq. 3-1
where:
[ A ] = system matrix
[ L ] = lower triangular factor
[ D ] = diagonal matrix
or
[A ] = [C ][C ]
Eq. 3-2
where:
[ C ] = Cholesky factor
For unsymmetric matrices:
[A] = [L ][U ]
where:
[ L ] = lower triangular factor
[ U ] = an upper triangular factor
Eq. 3-3
CHAPTER 3
3.2
Theory of Decomposition
Symmetric Decomposition Method
The symmetric decomposition algorithm of MSC.Nastran is a sparse algorithm. This
algorithm relies on fill-in reducing sequencing and sparse numerical kernels. The specific
implementation also allows for the indefiniteness of the input matrix. This method is based
on Duff, et al., 1982.
The factor has a specific storage scheme that can be interpreted only by the sparse FBS
method.
Mathematical Algorithm
Permute and partition A as follows:
PAP =
E C
C B
Eq. 3-4
where the assumption is that the inverse of the s by s submatrix E exists. lf A is indefinite,
appropriate pivoting is required to ensure the existence of the inverse. This requirement is
fulfilled by the presence of the P permutation matrices in the previous equation. The order
of E is either 1 or 2. Then the elimination of E is shown as
Is
PAP =
CE
Take A 2 = B CE
0
In 1
0 B CE
1 T
Is E
1 T
Eq. 3-5
0 In 1
1 T
C , permute and partition again to obtain the following:

T
PA 2 P =
E2 C2
C2 B2
Eq. 3-6
and continue the process until

order ( B k ) = 1 or 2
The final factored form of
PAP = LDL
Eq. 3-7
47
48
is then given by building the following:

I1
1
C1 E1
I2
I3
L =
C2 E2
C3 E3
Eq. 3-8
.
Ik
and
E1 0
D =
0 E2 0
0 0 E3
Eq. 3-9
.
1 T
0 Bk Ck Ek Ck
where D is built of 1 by 1 and 2 by 2 diagonal blocks. The I k identity submatrices are also
of order 1 or 2 and the submatrices are rectangular with one or two columns. The rows of
1
the C i E i matrices extend to the bottom of the L matrix.
The most important step is the proper selection of the E partition. This issue is addressed
later in this guide.
The module consists of two distinct phases: the symbolic phase and the numeric phase.
Symbolic Phase
This phase first reads the input matrix A and creates the following information: one vector
of length NZ (where NZ is the number of nonzero terms of the upper half of the input
matrix A ) contains the column indices, and another vector of the same length contains the
row indices of the nonzero terms of the upper triangular half of the matrix A . Both of these
vectors contain integers. Another responsibility of this phase is to eliminate the zero rows
and columns of the input matrix.
The selection of the E partition (i.e., the general elimination process) can be executed in
different sequences. The performance of the elimination using the different sequences is
obviously different. To find an effective elimination sequence, a symbolic decomposition
is also executed in the preface. An important criterion is to minimize the fill-in created in
each step of the elimination process, thereby reducing the numerical work and the l/O
requirements.
The reordering methods may be prefaced by a compression process based on grid-dof
connection or the so-called supermodal amalgamation principle (both available with any
reordering). Both contribute by making the reordering more efficient in time and profile.
CHAPTER 3
Several different reordering algorithms are available in the symbolic phase: multiple
minimum degree algorithm (MMD), Metis, EXTREME, MLV (multilevel vertex
partitioning), etc. These are described in the references.
Each of the methods can be selected by setting system cell 206 to the desired option; see
User Interface on page 52 for details.
In each of the above methods, the elimination of a chosen variable is performed based on
severing all the edges connected to it and connecting those nodes which were connected
through the eliminated variable. Severing these edges leaves a reduced graph where the
same process can continue until the reduced graph is a single node only. Then using this
term as the root, create a so-called assembly tree of the matrix. The final elimination
sequence is obtained by traversing the assembly tree. Note that this sequence may be
changed due to numerical reasons.
Finally, the elimination sequence is stored into an array of length 3 N (where N is the
order of the A matrix). Note that at this phase, the actual terms of the matrix are not
needed.
Numeric Phase
The mathematical decomposition process was described previously except for the details
of the pivot selection for numerical stability. The strategy applied is a variant of the BunchParlett method (1971) and is implemented as follows.
Let us assume that the next potential pivot row is the j-th. The diagonal entry of that row
is tested against all the other terms k = ( j + 1, , n ) as follows:
a jj > t a jk
where t is based on an input parameter. If the inequality is true, then the decomposition
process uses s = 1 (1 by 1 pivot) and a jj as the pivot term ( E matrix). If the inequality is
not true and the a jl term is the largest in the pivot row, then the following pivotal matrix
is tested for stability:
E2 =
a jj a jl
a lj a ll
If
2
max i
a lm
e ij < -----------t
j = 1
is satisfied, then E 2 is the pivot. Here a lm is the largest term in row l and e ij are the terms
1
of E 2 .
49
50
If both of the above pivots fail, then a search is performed in the remaining possible pivot
rows ( k > j ) for pivots. When one is found, that particular row is permuted into the j-th
row position (by putting ones into the proper locations of the P matrix), and the numerical
elimination proceeds.
The numerical work is executed primarily by vector kernels. The triangular kernels DFMR
and DFMQ are also used in addition to the conventional AXPY kernel.
Numerical Reliability of Symmetric Decomposition

The numerical reliability of the matrix decomposition process is monitored via the
matrix/factor diagonal ratio as follows:
a ii
-----di
Eq. 3-10
where a ii is the original diagonal term of the matrix and d i is the corresponding diagonal
term of the factor. The maximum value of these ratios is used to indicate how wellconditioned the original matrix was. The higher this ratio, the closer the matrix is to
singularity. As shown by the algorithm of the decomposition, small d i values are the cause
of numerical instability. Hence in the case of unusually high matrix/factor diagonal ratios,
the user should practice extreme care in evaluating the results.
The sources of high ratios are the mechanisms in MSC.Nastran. A mechanism is a group
of degrees of freedom that may move independently from the rest of the structure as a rigid
body.
For example, when unconstrained directions allow the entire model to move (a
mechanism) a high ratio occurs at the last grid point in the internal sequence. Another
possible cause of high ratios is connecting flexible elements to stiff elements. Finally,
missing elements can also cause high ratios.
7
In general, ratios below 10 are usually acceptable; however, the safety limit is
3
approximately 10 .
Unsymmetric Decomposition
The sparse unsymmetric decomposition algorithm is another variation of the Gaussian
elimination as follows:
For j = 1 , ... , n (loop on all columns)
For i = 1 , ... , n (loop on all rows)
If i j , then (elements of lower triangular)
min ( i 1, j 1 )
l ij = a ij
k = 1
If i < j , then
l ik u kj
CHAPTER 3
min ( i 1, j 1 )
a ij
l ik u kj
k = 1
u ij = -----------------------------------------------------------------l ii
For numerical reliability, the above elimination order is modified when necessary. The
pivoting step is based on the following convention:
a kk max ( a ik t )
i
Eq. 3-11
The meaning of Eq. 3-11 is that the a kk term is accepted as a possible pivot term if it is
larger than the maximum taken in the k-th column multiplied by a ratio of t (which is based
on a user-specified threshold parameter; see Eq. 3-12).
From the computer science aspect, the sparse unsymmetric decomposition is similar to the
symmetric decomposition, using indexed vector operations and frontal logic which are not
discussed here in detail. The sparse unsymmetric decomposition also has a distinct
symbolic and numeric phase similar to symmetric sparse decomposition.
Partial Decomposition
The sparse symmetric decomposition method may be used to decompose only a certain
partition specified by a partitioning vector (PARTVEC input data block for DCMP). In this
case the following decomposition is obtained:
A =
A oo A oa
A ao A aa
L oo
L ao I
D oo
A aa
T LT
L oo
ao
I
where
T
A aa = A aa L ao D oo L ao
The results of L oo, D oo, L ao are stored in the LD output data block and the aa is in the
LSCM data block, see User Interface on page 52.
51
52
3.3
User Interface
DCMP
Matrix decomposition with extended diagnostics
Decompose a square matrix [ A ] into upper and lower triangular factors [ U ] and [ L ] and
diagonal matrix [ D ] . DCMP also provides extended diagnostics.
[ A ] = [ L ] [ U ] for unsymmetric [ A ]
[A ] = [L ][D ][L ]
for symmetric [ A ]
Format:
DCMP
USET,SIL,EQEXIN,A,PARTVEC,EQMAP/
LD,U,LSCM/
S,N,KSYM/CHOLSKY/BAILOUT/MAXRATIO/SETNAME/F1/DECOMP/
DEBUG/THRESH/S,N,MINDIAG/S,N,DET/S,N,POWER/S,N,SING/
S,N,NBRCHG/S,N,ERR $
Input Data Blocks:

USET
Degree-of-freedom set membership table.
SIL
Scalar index list.
EQEXIN
Equivalence between external and internal numbers.
A square matrix (real or complex, symmetric or unsymmetric).
PARTVEC
Partitioning vector specified when A is a partition of SETNAME. Its rowsize

is indicated by SETNAME. A is the zero-th partition from PARTVEC. In the
partial decomposition case it defines A oo .
EQMAP
Table of degree-of-freedom global-to-local maps for domain decomposition.
Output Data Blocks:

LD
Nonstandard lower triangular factor [ L ] and diagonal matrix [ D ] or

Cholesky Factor. [ LD ] also contains [ L ao ] for partial decomposition.
Nonstandard upper triangular factor of [ A ] or Cholesky Factor.
LSCM
Schur complement matrix based on internal partitioning of A.
Parameters:
KSYM
Input/output-integer-default=1. See Method Selection on page 54
CHOLSKY
Input-integer-default=0. See Method Selection on page 54
CHAPTER 3
BAILOUT
Input-integer-default=0. If BAILOUT>0, then the module exits with error

message if factor to diagonal ratio exceeds MAXRATIO. If BAILOUT<-1,
then the module continues with warning message if factor to diagonal ratio
exceeds MAXRATIO.
MAXRATIO Input-real-default=1.E5. See the BAILOUT and ERR parameter.

SETNAME
Input-character-default=H. One or two letters indicating the set

membership of [ A ] .
F1
Input-real-default = 0.0. Tolerance for suppressing numbers of small

magnitude. Matrix elements with magnitudes less than F1 will be set to
zero.
DECOMP
Input-integer-default=-1. See Option Selection on page 55.
DEBUG
THRESH
MINDIAG
Output-real double precision-default=0.0D0.
DET
Output-complex-default=(0.0,0.0).
POWER
Output-integer-default=0.
SIGN
Output-integer-default=0. See Option Selection on page 55.
NBRCHG
Output-integer-default=0. See READ module.
ERR
Output-integer-default=-1. If BAILOUT=-1, this parameter always remains

at zero. If BAILOUT=0 and the factor to diagonal ratio is negative or greater
than MAXRATIO, ERR is reset to -1.
53
54
3.4
Method Selection
To select decomposition methods, the following parameters are used:
KSYM
1Use symmetric decomposition (default).

0Use unsymmetric decomposition.
-1Use decomposition consistent with form of [ A ] . KSYM will be reset to 0
or 1 consistent with actual decomposition type.
3Use symmetric partial decomposition.
CHOLSKY
If KSYM=1 or KSYM=-1 and [ A ] is symmetric then:

1Use Cholesky decomposition.
0Use standard decomposition (default).
If KSYM=3, then CHOLSKY is set to the number of degrees of freedom in the
o-set.
CHAPTER 3
3.5
Option Selection
Minimum Front Option
To increase performance of the sparse symmetric decomposition, the user may set
SYSTEM(198), the MINFRONT keyword to a value greater than 1. The appropriate value
is problem and machine dependent. Its meaning is to restrict the sparse strategy to above
a certain minimum front size (characteristically 8 - 16).
Reordering Options
The various reordering options are selected via SYSTEM(206) as follows:
SYSTEM(206)
DCMPSEQ
Method
EXTREME for 3D, Metis or MMD for 2D
MMD - definite matrices
MMD - indefinite matrices
No sequencing
Default, EXTREME
Metis
Better of Metis and MMD
32
MLV
64
Turn on Supernodal Compression Scheme
Compression Options
The supernodal compression scheme currently is available only with EXTREME and Metis.
The grid-based compression scheme is automatically executed when the datablocks
defining the grid-DOF connections (USET,SIL) are available to the module.
Perturbation Option
If DEBUG (= SYSTEM(60)) is set then an = 10 DEBUG is put on zero diagonal terms. If
not set, = 10 10 is used.
High Rank Options

An additional performance improvement is possible, with the high rank update in the
sparse decomposition methods. The rank of update in the various sparse decomposition
methods is set as:
55
56
SYSTEM
Sparse Decomposition
(205)
Symmetric, real
(219)
Symmetric, complex
(220)
Unsymmetric, real
(221)
Unsymmetric, complex
The defaults of these cells are set automatically, since they are machine dependent.
Diagnostic Options
These options of the sparse symmetric decomposition are requested as follows:
SYSTEM
Action
(69) = 64
(166) =
Stop after diagnostic phase

2
Provides internal diagnostics
Overwrites MAXRATIO by 1.0
Provides MAXRATIO vector in U
The THRESH parameter is used to control the pivoting for the unsymmetric and the sparse
(which also pivots) decompositions. The pivot threshold is
t = 10
THRESH
Eq. 3-12
In the unsymmetric case, pivoting occurs if a factor diagonal value is less than t . In the
symmetric case of the sparse decomposition, pivoting occurs when the ratio for the factor
diagonal value to the largest term in the pivot row is less than t .
The default value of THRESH is 6 for the sparse decomposition and 10 for the
unsymmetric decomposition. The latter may also be defined by SYSTEM(91) for the
unsymmetric case.
In the case of a READ module the THRESH parameter may be set by SYSTEM(89).
The shared memory parallel execution of the sparse symmetric decomposition can be
selected by
SYSTEM(126) = 8 and SYSTEM(107) > 1
and deselected by turning off either one of these system cells.
CHAPTER 3
3.6
Diagnostics
The diagnostics given by the decomposition module are organized into numerical,
performance, statistical, and error diagnostics.
Numerical Diagnostics
These diagnostics are related to the accuracy of the solution.
Singularity
Causes of Singularity. A matrix is singular if its inverse cannot be calculated. The SING
parameter is set to -1 if the matrix is singular. Singularity is a relative issue with respect to
stiffness ratios. However, some independent general reasons for singularity include:
Degree of freedom without stiffness

2-D problem, normal rotation unconstrained
3-D problem, rotational DOFs at solids unconstrained
Planar joint in space structure
Singularity Test
To avoid singularity, the user can enter the AUTOSPC keyword. If AUTOSPC is set to YES,
the singular degrees of freedom are automatically constrained out. Any degree of freedom
is considered to be singular if
A ij
-------------
A max
Eq. 3-13
where A ij is the term in the i-th row and the j-th column of A matrix, and A max is the
largest term in [ A elem ] .
8
The default for is 10 and can be changed by the keyword EPZERO. The SPC entries
constraining the singular DOFs are generated by setting the SPCGEN keyword to YES.
8
Parameter EPPRT (Bulk Data) (default = 10 ) is used to set a threshold below which all
potential singularities are listed. If EPPRT is greater than EPZERO, then the printing of
singularities with a ratio of exactly zero is suppressed.
Ill-Conditioning
Causes of Ill-Conditioning. The ill-conditioning of a matrix can be caused by any of the
following reasons:
Low stiffness in rotation

Large mass
Very stiff beam
Mechanisms
57
58
MAXRATIO Parameter. Ill-conditioning is diagnosed by the MAXRATIO diagnostics

parameter which is defined as
A ii
MAXRATIO = ------D ii
Eq. 3-14
where D ii is the i-th diagonal term in the D matrix.

The maximum MAXRATIO of the decomposition is printed under User Information
7
Message 4158 or 4698 when its value is greater than 10 . This limit can be changed by
setting the keyword MAXRATIO.
In the case of 2 2 pivoting, this ratio is not calculated.
Negative Terms on Factor Diagonal

STURM Number. The NBRCHG parameter (DMAP call) gives the number of negative
terms on the factor diagonal (it is also called the STURM number). This diagnostics
information message is important when decomposition is used in an eigenvalue module.
In this case the number of negative terms provides the number of negative eigenvalues of
the matrix. Since the matrix decomposed in this case is usually a shifted matrix, the
NBRCHG gives the number of eigenvalues to the left of the shift. User Information
Messages 4158 and 5010 in the eigenvalue modules print the value of NBRCHG.
For symmetric decomposition, the following message (UIM 4157) appears:
MATRIX SIZE
NUMBER OF NONZEROES
NUMBER OF ZERO COLUMNS
NUMBER OF ZERO DIAGONALS
CPU TIME ESTIMATE
I/O TIME ESTIMATE
EST. MEMORY REQUIRED
MEMORY AVAILABLE
EST. INTEGER WORDS IN FACTOR
EST. NONZERO TERMS IN FACTOR
EST. MAX FRONT SIZE
RANK OF UPDATE
The integer words in factor are the row and column index information, and the real words
are the actual terms of the factor matrix [ L ] . In the sparse methods, the integer and real
words are stored in two separate records.
The 4157 message is followed by UIM 6439 for the sparse symmetric decomposition as
follows:
UIM 6439 (DFMSA)
ACTUAL MEMORY AND DISK SPACE REQUIREMENTS FOR

SPARSE SYMMETRIC DECOMPOSITION
CHAPTER 3
SPARSE DECOMP MEMORY REQUIRED
MAXIMUM FRONT SIZE
INTEGER WORDS IN FACTOR
NONZERO TERMS IN FACTOR
For unsymmetric sparse decomposition, UIM 4216 provides the following information.
MATRIX SIZE
NUMBER OF NONZEROES
NUMBER OF ZERO COLUMNS
NUMBER OF ZERO DIAGONAL TERMS
CPU TIME ESTIMATE
I/O TIME ESTIMATE
ESTIMATED MEMORY REQUIREMENT
MEMORY AVAILABLE
EST. INTEGER WORDS IN FACTOR
EST. NONZERO TERMS
ESTIMATED MAXIMUM FRONT SIZE
RANK OF UPDATE
This message is also followed by a UIM 6439 which gives actual values for the estimates in
UIM 4216.
Statistical Diagnostics
The following messages are self-explanatory.
UIM 4158:
STATISTICS FOR SYMMETRIC (PARALLEL AND/OR SPARSE) DECOMPOSITION OF
DATA BLOCK XX
NUMBER OF NEGATIVE TERMS ON FACTOR DIAGONAL.
MAXIMUM RATIO OF MATRIX DIAGONAL TO FACTOR DIAGONAL.
UIM 4367:
STATISTICS FOR UNSYMMETRIC DECOMPOSITION OF DATA BLOCK XX FOLLOW.
NUMBER OF PIVOT OPERATIONS = XX.
UWM 5221:
STATISTICS FOR DECOMPOSITION OF MATRIX XX.
THE FOLLOWING DEGREES OF FREEDOM HAVE NULL COLUMNS.
UWM 4698:
STATISTICS FOR DECOMPOSITION OF MATRIX XX.
THE FOLLOWING DEGREES OF FREEDOM HAVE FACTOR DIAGONAL RATIOS
GREATER THAN MAXRATIO OR HAVE NEGATIVE TERMS ON THE FACTOR
DIAGONAL.
Error Diagnostics
The following are messages from sparse decomposition and are described as follows:
SFM 4370:
DECOMPOSITION REQUIRES THAT PRECISION OF DATA BLOCK XX EQUAL
SYSTEM PRECISION.
59
60
This is a general limitation of the module.

UFM 3057:
MATRIX XX IS NOT POSITIVE DEFINITE.
A Cholesky option was requested by the user on a non-positive definite matrix.
SFM 4218:
UNSYMMETRIC DECOMPOSITION IS ABORTED DUE TO INSUFFICIENT MEMORY.
The preface of the unsymmetric decomposition module needs more memory to execute.
SFM 4255:
UNSYMMETRIC DECOMPOSITION OF DATA BLOCK XX FAILS AT ROW XX.
UNABLE TO PIVOT.
A singular matrix was given to the module.
UW(F)M 6136 (DFMSA):
INSUFFICIENT CORE FOR SYMBOLIC (NUMERIC) PHASE OF SPARSE
DECOMPOSITION.
USER ACTION: INCREASE CORE BY XX WORDS.
UFM 6133 (DFMSD):
SINGULAR MATRIX IN SPARSE DECOMPOSITION.
USER ACTION: CHECK MODEL
UFM 6137 (DFMSD):
INPUT MATRIX IS RANK DEFICIENT, RANK = XX.
USER ACTION: CHECK MODEL
CHAPTER 3
3.7
Decomposition Estimates and Requirements

The CPU time estimate for the sparse symmetric decomposition method includes:
2
1
--- N N front M
2
Eq. 3-15
1
--- N N front P s + 2 P s N z
2
Eq. 3-16
where:
N front = average number of connected DOFs
N z = approximate number of nonzeroes in the upper triangle of the original matrix
2
N 2+N
= density of the original matrix
Storage Requirements. The minimum storage requirements are as follows:
Disk:
( 1 + NWPT )N front N + N
Memory:
( 6 + 2 NWPT ) N
where:
NWPT = number of words per term
1 for 64 bit word real arithmetics
2 for 64 bit word complex arithmetics
2 for 32 bit word real arithmetics
4 for 32 bit word complex arithmetics
The CPU time estimate for the sparse unsymmetric decomposition method is
2
N N front M
Eq. 3-17
N N front P s + 4 P s N z
Eq. 3-18
61
62
3.8
References
Bunch, J. R.; Parlett, B. N. Direct Methods for Solving Symmetric Indefinite Systems of
Linear Equations. Society for Industrial and Applied Mathematics Journal of
Numerical Analysis, Volume 8, 1971.
Duff, I. S.; Reid, J. K. The Multifrontal Solution of Indefinite Sparse Symmetric Linear
Systems. Harwell Report CSS122, England, 1982.
George, A.; Liu, J. W. Computer Solutions of Large Sparse Positive Definite Systems.
Prentice Hall, 1981.
Goehlich, D.; Komzsik, L. Decomposition of Finite Element Matrices on Parallel
Computers. Proc. of the ASME Int. Computers in Engineering Conf., 1987.
Golub, G. H.; Van Loan, C. F. Matrix Computations. John Hopkins University Press, 1983.
Hendrickson, B., Rothberg, E. Improving the Runtime and Quality of Nested Dissection
Ordering. Silicon Graphics, Inc., Mountain View, CA, April 11, 1996.
Karypis, G., Kumar, V. METIS, A Software Package for Partitioning Unstructured
Graphs, Partitioning Meshes, and Computing Fill-Reducing Orderings of Sparse
Matrices, Version 3.0.3, University of Minnesota, Department of Computer
Sciences/Army HPC Research Center, Minneapolis, MN, November 5, 1997.
(http://www.cs.umn.edu/~karypis)
Pissanetzsky, S. Sparse Matrix Technology. Academic Press, 1984.
Shamsian, S.; Komzsik, L. Sparse Matrix Methods in MSC/NASTRAN. The MSC 1990
World Users Conf. Proc., Vol. II, Paper No. 64, March, 1990.
Rothberg, E. Ordering Sparse Matrices Using Approximate Minimum Local Fill, Silicon
Graphics, Inc., Mountain View, CA, April 11, 1996.
CHAPTER
Direct Solution of Linear Systems

Solution Process
Theory of Forward-Backward Substitution
User Interface
Method Selection
Option Selection
Diagnostics
FBS Estimates and Requirements
64
4.1
Solution Process
Solution of linear systems is an important and time-consuming component of
MSC.Nastran runs. Mathematically, the solution process represents a right-handed
[A][X] = [B]
Eq. 4-1
or a left-handed
T
[X] [A] = [B]
Eq. 4-2
direct solution.
The iterative solution directly solves Eq. 4-1 and is described in Iterative Solution of
Systems of Linear Systems on page 75.
The direct solution method contains two distinct parts: a forward and a backward
substitution, hence its name forward-backward substitution.
The forward-backward substitution is a follow-up operation to the decomposition process
(see Matrix Decomposition on page 45).
The right-handed direct solution step is executed using the triangular factors calculated in
the decomposition as a forward step of
[L ][Y ] = [B ]
Eq. 4-3
The backward step of the operation uses the intermediate result [ Y ] as follows:
T
[L] [X] = [D]
[Y]
Eq. 4-4
In the case of unsymmetric decomposition, the backward step is

[U][X] = [Y]
Eq. 4-5
The left-handed forward step to solve Eq. 4-2 is

T T
W L
= BT
Eq. 4-6
and
T
Z L = W
Eq. 4-7
In Eq. 4-6 and Eq. 4-7, the decomposition is assumed to be a Cholesky method. In the
LDL T case, Eq. 4-7 is modified as
T
Z L = D
WT
The theory covering the solution of systems of linear equations is described in the
following section.
Eq. 4-8
CHAPTER 4
4.2
Theory of Forward-Backward Substitution

Right-Handed Method
The elements of matrices [ Y ] and [ X ] of Eq. 4-3 and Eq. 4-4 are given by
i1
y ij = b ij
l ik y kj
Eq. 4-9
k = 1
and
y ij
x ij = ------
di
l ki x kj
Eq. 4-10
k = i+1
MSC.Nastran uses a sparse implementation of the above algorithm, as well as special

modifications for the parallel, and left-handed cases.
Left-Handed Method
The elements of Eq. 4-6 and Eq. 4-7 are:
i1
b ik
w jk l ij
j = 1
w ik = ------------------------------------------l ii
Eq. 4-11
and
n
w ik
w jk l ij
j = i+1
z ik = -----------------------------------------------------l ii
In the above equations, the A matrix is assumed to have been decomposed with the
Cholesky method.
Sparse Method
The sparse option of FBS executes the forward-backward substitution from the factor of the
sparse (multifrontal) decomposition. Thereby, one must consider the permutations
executed for pivoting given by P for the symmetric case, or P and Q for the unsymmetric
case, and then solve the following matrix equations. For the symmetric case,
T
LDL PX = PB
For the unsymmetric case,
Eq. 4-12
65
66
PLUQX = PB
Eq. 4-13
The usual forward pass solves the following equation for the symmetric case
LY = PB
Eq. 4-14
PLY = PB
Eq. 4-15
for Y . The backward pass gives X from the following equation for the symmetric case
T
DL ( PX ) = Y
Eq. 4-16
UQX = Y
Eq. 4-17
The storage requirements are real areas for the right-hand side B , the result X , and an
integer vector of length N holding the permutation information plus an area for the factor.
Parallel Method
A shared memory, parallel execution of the sparse method is also available. The
parallelization is either on the factor side of the equations, based on special shared memory
parallel kernels, or on the right-hand side when multiple loads exists.
CHAPTER 4
4.3
User Interface
To solve the matrix equation [ A ] [ X ] = [ B ] (right-handed solution) or
[ X ] T [ A ] = [ B ] T (left-handed solution) using the triangular factors computed by DCMP.
FBS
LD,U,B/X/KSYM/SIGN/FBTYP $
Input Data Blocks:

LD
Lower triangular factor/diagonal, or Cholesky factor.
Upper triangular factor. Purged unless [ A ] is unsymmetric.
Rectangular matrix.
Output Data Block:

X
Rectangular matrix having the same dimensions as [ B ] .
Parameters:
KSYM
Input-integer-default = -1. See Method Selection on page 68.
SIGN
Input-integer-default = 1. See Method Selection on page 68.
FBTYP
Input-integer-default = 0. See Option Selection on page 69
67
68
4.4
Method Selection
FBS Method Selection
The FBS method selection is executed via the following parameters:
KSYM
Symmetry flag.
-1choose symmetric if [ U ] is purged, otherwise unsymmetric (default).
0matrix [ A ] is unsymmetric.
1matrix [ A ] is symmetric.
2perform left-handed solution. See Option Selection on page 69.
SIGN
Sign of [ B ] .
1solve [ A ] [ X ] = [ B ] (default).
-1solve [ A ] [ X ] = [ B ] .
CHAPTER 4
4.5
Option Selection
Right-handed FBS Options
The user can control the flow of the forward-backward substitution via the FBTYP
parameter (DMAP call), which is also equivalent to SYSTEM(74). The default value is 0,
which indicates a full execution of both passes of the solution. Set FBTYP to +1 for a
forward pass only; set FBTYP to -1 for a backwards only partial execution. These options
are useful in eigenvalue modules.
T
SYSTEM(73) = +1 indicates the presence of Cholesky factor [ L ] [ L ] . The default is 0,

T
which indicates the regular [ L ] [ D ] [ L ] factor. To summarize this:
SYSTEM
(73)
(74)
Value
Action
+1
Cholesky factor: [ L ] [ L ] [ X ] = [ B ]
else
Regular factor: [ L ] [ D ] [ L ] [ X ] = [ B ] (default)
+1
Solve [ L ] [ Y ] = [ B ] or [ L ] [ D ] [ Y ] = [ B ]
1
0
Solve [ L ] [ X ] = [ Y ]
Full FBS (default)
Left-handed FBS Option

To indicate a left-handed solution, SYSTEM(70) = 2 is needed. Note that SYSTEM(70) is
equivalent to FBSOPT. The left-handed FBS obtains the solution by rows, as opposed to
columns. Since the solution is packed out via the usual GINO facilities, the solution rows
are stored as columns of the solution (X) matrix. The user may input a transposed or
untransposed right-hand side (B) matrix. To summarize this:
SYSTEM
(72)
Value
Action
=1
BT = XT A
B = XTA
The FBS options controlled by cells 73 and 74 apply to the left-handed solution
appropriately.
Parallel FBS Solution
The user needs FBSOPT = -2 and PARALLEL > 1 for the right-hand side shared memory
parallel execution. For the factor side parallel only PARALLEL > 1 is needed.
The parallel method can be deselected by setting PARALLEL = 1024 + ncpu where ncpu =
number of CPUs. This setting leaves the other parallel methods enabled.
69
70
4.6
Diagnostics
For the direct solution of systems of linear equations, diagnostics can also be classified as
follows: numerical diagnostics, performance messages, and error diagnostics.
UIM 5293:
FOR DATA BLOCK
....
LOADSEQ
EPSILON
EXTERNAL WORK
....
....
....
....
....
....
Performance Messages
UIM 4234:
UFBS TIME ESTIMATE TO FORM XX TYPE = X CPU = X I/0 = X TOTAL = X PASSES = X
UIM 4153:
FBS METHOD X TIME ESTIMATE TO FORM XX CPU = X I/0 = X TOTAL = X PASSES = X
These messages are printed only when the CPU time estimate is greater than the value of
SYSTEM(20) (default = machine-dependent). To force the printing of these messages, the
user must set SYSTEM(20) = 0.
Error Diagnostics
FBS 1(2, 3 OR 4) FATAL ERROR 20:
USER FATAL MESSAGE
This error should not occur under normal circumstances. The cause for the error is that
information from the GINO control block is incompatible with information from the buffer.
This occurs in methods 1 or 1A only.
SFM FBSUB LOGIC ERROR 10:
USER FATAL MESSAGE
This error is similar to the previous one from method 2.

SFM 6069 (LFBSS):
SYMMETRIC LEFT HANDED FBS IS CALLED TO SOLVE A COMPLEX SYSTEM WITH
CHOLESKY FACTOR.
This option is not supported.
SFM 6070 (LFBSS):
ERROR IN READING THE FACTOR IN SYMMETRIC LEFT HANDED FBS
These messages are from the left-handed method. They are either from symmetric,
unsymmetric, or from the timing phase of either one. The cause for this error is similar to
the cause for Error 20.
CHAPTER 4
SFM 6072 (LFBSU):

INCORRECT PIVOTING INSTRUCTIONS IN UNSYMMETRIC FACTOR DURING A
LEFT HANDED FBS.
This message is similar to Error 20 since it indicates inconsistency in a data block.
SFM 6073 (LFBSU):
ERROR IN READING THE FACTOR IN UNSYMMETRIC LEFT HANDED FBS
Similar causes as Error 20.
SFM 6201 (FBSQCK):
SPARSE FBS CANNOT BE EXECUTED WHEN THE FACTOR IS REAL AND THE RIGHT
HAND SIDE IS COMPLEX.
Recommendation: Do not select the sparse decomposition and FBS methods under these
circumstances.
UFM 6138 (DFMSB):
INSUFFICIENT CORE FOR SPARSE FBS.
USER ACTION: INCREASE CORE BY XX WORDS.
71
72
4.7
FBS Estimates and Requirements

Sparse FBS Estimates
The CPU time for sparse FBS is:
2 N RHS NZ FAC M
Eq. 4-18
2 N N RHS P + 2 NZ FAC P s N pass
Eq. 4-19
where:
N RHS = number of right-hand sides
NZ FAC = number of nonzeroes in the factor
N pass = number of passes
The minimum storage requirements of FBS are:
Disk:
NZ FAC + 2 N N RHS IPREC
Memory:
2
1 2 N front
IPREC + 2 N IPREC
where:
N front = average front size
IPREC = machine precision (1 for short-word machines, 2 for long-word machines)
CHAPTER 4
4.8
References
Komzsik, L. Parallel Processing in Finite Element Analysis. Finite Element News,
England, June, 1986.
Komzsik, L. Parallel Static Solution in Finite Element Analysis. The MSC 1987 World
Users Conf. Proc., Vol. I, Paper No. 17, March, 1987.
73
74
CHAPTER
Iterative Solution of Systems of Linear

Systems
Iterative Solver Solutions
Theory of the Conjugate Gradient Method
Preconditioning Methods
User Interface
Iterative Method Selection
Option Selection
Iterative Solution Diagnostics
Iterative Solver Estimates and Requirements
References
76
5.1
Iterative Solver Solutions

The previous chapter discussed the solution of linear systems using the direct method,
which consists of matrix decomposition followed by a forward-backward substitution. In
this chapter an alternative method, the iterative solver, is described.
Solution Sequences. The iterative solver can be used in the following solution
sequences:
Linear statics (SOLs 1, 101)

Nonlinear statics (SOL 106)
Direct frequency response (SOLs 8, 108)
Modal frequency response (SOLs 11, 111)
Parallel Execution. In addition to the above sequential versions, the iterative solver is
implemented for parallel execution on distributed memory machines in SOL 1 and
SOL 101 (linear statics).
Methods
For symmetric positive definite systems, several different versions of the preconditioned
conjugate gradient method are available in MSC.Nastran. For unsymmetric or indefinite
systems, the preconditioned bi-conjugate gradient or the preconditioned conjugate
residual method is used.
CHAPTER 5
Iterative Solution of Systems of Linear Systems
5.2
Theory of the Conjugate Gradient Method

The conjugate gradient method minimizes the error function
T
1 T
F ( x ) = --- x Ax x b
2
Eq. 5-1
The first derivative (gradient) of this function is

dF
------- = Ax b = r
dx
Eq. 5-2
which is the negative of the residual. An iterative procedure is obtained by calculating

consecutive versions of the approximate solution as follows:
xi + 1 = xi + i pi
Eq. 5-3
r i + 1 = r i i Ap i
Eq. 5-4
and
where the direction p and the distance are computed to minimize the above error
function.
Computationally efficient forms to calculate these quantities are
T
ri ri
i = --------------T
p i Ap i
Eq. 5-5
pi = ri + i pi 1
where
T
ri ri
i = ---------------------------T
ri 1 ri 1
See Hageman and Young, 1981 [1] for more details. The algorithmic formulation is fairly
straightforward.
Convergence Control
Convergence is achieved inside the iteration loop when the following inequality holds:
rn
----------- <
b
Eq. 5-6
77
78
where r n denotes the residual vector after n iterations, b denotes the initial right-hand
side vector, and is a user-defined parameter (default: 1.E-06). If this ratio does not
become smaller than within the maximum number of iterations, is compared to an
energy norm Eq. 5-7 after the iteration loop is finished. The solution is accepted if the
following inequality holds:
T
x r
------------- <
T
b x
Eq. 5-7
where x is the solution vector, r is the final residual vector, and b is the initial right-hand
side vector as before. Experience has shown that the convergence criterion inside the
iteration loop Eq. 5-6 is far more conservative than the convergence criterion outside the
iteration loop Eq. 5-7.
Based on user selection, an alternate convergence criterion is available for Jacobi, Cholesky
and RIC preconditioning:
x n < e 1 f n
Eq. 5-8
where:
x n = x n + 1 x n
fn =
x n
-----------------x n 1
See Conca, 1992 [2] for details.
Block Conjugate Gradient Method (BIC)

The BIC method is the most recent and most recommended method for an iterative
solution of the linear system in MSC.Nastran. It was introduced several years ago and has
gone through major improvements with regard to memory management and spill logic
since then.
BIC is a block version of the conjugate gradient method described above where block
refers to two characteristics of the method:
1. The efficient solution of the linear system with a block of right-hand sides where
the following inequality is expected to hold for m < 10 ( m = the number of righthand sides)
m
ti > tm
i = 1
where t i refers to the time needed to solve the linear system with just the i-th
right-hand side and t m refers to the time needed to solve the linear system with
the complete block of m right-hand sides.
CHAPTER 5
The method will still work for m 10 ; however, the inequality may not hold.
2. Both the matrix and the preconditioner are stored in block-structured form to
improve the performance.
The block of direction vectors is updated in every iteration as
T
p k = [ T k AT k ]
1 T
Tk
Rk
Eq. 5-9
and the blocks of residual and solution vectors are updated as

R k + 1 = R k AT k P k
Eq. 5-10
Xk + 1 = Xk Tk Pk
Eq. 5-11
where all matrices are n m size except for T k , which is obtained by the concatenation of
the two matrices as
Tk = [ Xk Xk 1 Rk ]
Eq. 5-12
and it is of size n 2m . Convergence is achieved when the following inequality holds

max X n + 1 X n
max R n + 1
max i ------------------------------------ , i ---------------------- <
Xn + 1
F
Eq. 5-13
where i is the column index, i m , R is the rectangular matrix of the r residual vectors,
F is the set of right-hand sides, and X is the set of solution vectors. See Babikov, 1995 [3]
for more details.
The algorithmic formulation of this method employed in the iterative solution module is
fairly straightforward, except for the very difficult issue of calculating the inverse matrix
in P k . However, the details of that calculation are beyond the scope of this guide.
Real and Complex BIC

The BIC methods for real and complex linear systems of equations are different and,
therefore, are described separately in the following paragraphs. The parameters are
described in User Interface on page 84.
79
80
The real BIC method follows the basic algorithm shown below:
X0 = X1 = 0
R0 = R1 = F
Loop on k = 1, 2, , ITSMAX
Wk = B
Rk
V k = AW k
Tk = [ Xk Xk 1 Wk ]
Pk = [ Rk Rk 1 Vk ]
T
Gk = ( Pk Tk )
Hk = Gk Tk Rk
Xk = Xk 1 Tk Hk
Rk = Rk 1 Pk Hk
IF ( error < ITSEP S ) then converged
End loop on k
In the above algorithm, B is the preconditioner, F is the block of right-hand side vectors,
X is the block of updated solution vectors, and R is the block of updated residual vectors.
The memory management and the spill logic are rather complex and only the basic steps
are listed below:
1. Execute a bandwidth-reducing permutation
2. Symbolic phase
Find the best memory usage given a certain amount of memory and an IPAD
value (please see User Interface on page 84 for details on IPAD). The following
logical steps are traversed in this order:
a. Check if matrix and factor fit in core; if yes, go to step Set up memory
accordingly. on page 81.
b. Check if factor fits in core with matrix being out-of-core; if yes, go to step Set
up memory accordingly. on page 81.
CHAPTER 5
c. Check if memory is sufficient when factor and matrix are both out-of-core; if
yes, go to step Set up memory accordingly. on page 81.
d. Decrease padding level: IPAD = IPAD - 1; go back to step Check if matrix and
factor fit in core; if yes, go to step Set up memory accordingly. on page 81.
on page 80.
e. Set up memory accordingly.
3. The numeric phase of the preconditioner calculation makes use of the following
techniques:
a. Calculation of preconditioner in double precision; storage of final
preconditioner in single precision.
b. Restarts with global shift regularization if incomplete factorization fails.
c. Post truncation for well-conditioned problems.
4. Using the techniques described above, more memory may be available than was
predicted by the symbolic phase. Unless both matrix and preconditioner were in
core, as much as possible is read from the scratch files and saved in memory.
The memory management and spill logic for the complex BIC methods are different.
There are two different BIC methods for the complex case that are selected via the IPAD
value. For IPAD < 5, the complex BIC algorithm is similar to the real BIC method.
However, for IPAD 5, a very different strategy is used, which is described below.
The solution of the system
AX = F
is based on its equivalent representation
A A
11 12
A A
12 22
1 = F1
X
F
2
2
where matrices A 11 and A 12 (= A 21 , since A is symmetric) have fully zero imaginary parts
and A 22 is truly complex. The solution is found using the Schur complement.
The spill logic for the first complex method (IPAD < 5) is very simple: if the matrix and
preconditioner do not both fit in core, then a fully out-of-core approach is used.
The spill logic for the second complex BIC method is more sophisticated. It determines
whether only A 11 , only A 22 , or A 11 and A 22 can be kept in core. A 12 is always kept in core
since it usually has very few nonzero terms.
The second complex BIC method is recommended and is also the default for complex
symmetric linear systems if the iterative solver was chosen.
81
82
5.3
Preconditioning Methods
The use of a preconditioner is recommended to reduce the number of iterations. The
disadvantage of this process is that the preconditioning calculation increases the amount
of work in each iteration. With the use of a preconditioner matrix, the original problem is
now written as
P
Ax = P
Eq. 5-14
where P is the preconditioning matrix. In this form, the preconditioner is applied in every
iteration step. This is called stepwise preconditioning. The Jacobi and Cholesky
preconditioners are stepwise. If P = A is chosen, then the problem is trivial:
P
Ax = Ix = A
Eq. 5-15
Of course, the cost of this preconditioning is equivalent to the direct solution of the system.
The following four major stepwise preconditioning strategies are supported in
MSC.Nastran.
Jacobi (J). For the Jacobi method, the P matrix is a diagonal matrix containing the
diagonal terms of the A matrix. The preconditioning step in every iteration is a simple
division of the current residual vector by these terms.
Cholesky (C). In the Cholesky method, the selection of the preconditioner matrix is
T
P = CC A
Eq. 5-16
where C is an incomplete Cholesky factor of A . Despite wide acceptance, the use of

standard Cholesky preconditioning in the conjugate gradient method is only proven to be
very good for finite difference discretization of partial differential equations. For example,
the conjugate gradient method has convergence problems when used in finite element
problems with high Poisson ratios.
Reduced Incomplete Cholesky (RIC). In the RIC method the preconditioner matrix is
calculated based on elemental information as
T
P = CC B = f ( A, A elem )
Eq. 5-17
where A elem is the elemental matrices used to assemble A .

This method significantly increases the performance of the iterative solver for shell
problems. See Efrat, 1986 [4] for details.
Block Incomplete Cholesky (BIC). Another disadvantage of the Cholesky method is the
calculation cost for the incomplete or reduced factor and the memory requirement for parts
of the Cholesky factor. A specific block sparse implementation has reduced these
disadvantages. This proprietary method uses a global shift regularization strategy, a posttruncation technique for well-conditioned problems, and allow a certain level of fill-in
CHAPTER 5
based on delicate numerical considerations. The implementation of the block conjugate

gradient method accelerates convergence for problems involving multiple loads.
Moreover, a band reordering method is used for the matrix in the symbolic phase of the
iterative solver. See Babikov, 1995 for details.
Scaling
Another approach for preconditioning is to use the preconditioner as a transformation.
Then
P
APP
x = P
Eq. 5-18
is transformed into
Eq. 5-19
A x = b
In this case the solution of the transformed system has to be converted back to the original
solution as follows:
Eq. 5-20
x = Px
An example of this transformation approach is the diagonal scaling. Diagonal scaling is a

useful tool for matrices whose terms differ significantly in magnitude. In this method, the
following transformation is performed:
D
(A D
1
x = D b )
Eq. 5-21
where x = an intermediate result, such as

D =
a ii and x = D
Eq. 5-22
The diagonal terms of the scaled matrix are unity as a result of the diagonal scaling. This
scaling makes the Jacobi preconditioning step trivial. The other (Cholesky type)
preconditioning methods may be combined with scaling.
Numerical Reliability of Equation Solutions

The accuracy of the solution of the linear equation systems in MSC.Nastran is evaluated
with the following residual vector:
r = b Ax
Eq. 5-23
This residual vector is calculated for each solution vector in statics analysis. Then a scalar
value is calculated as follows:
T
x r
= --------T
x b
Eq. 5-24
The magnitude of this error ratio indicates the numerical accuracy of the solution vectors.
83
84
5.4
User Interface
Format for non-p-version analysis:
SOLVIT
A,B,XS,PC,USET,KGG,GM,SIL,EQEXIN,EDT,CASECC,EQMAP/
X,R,PC1,EPSSE/
SIGN/ITSOPT/ITSEPS/ITSMAX/IPAD/IEXT/ADPTINDX/
NSKIP/MSGLVL/PREFONLY/S,N,ITSERR/SEID $
Format for p-version analysis:

SOLVIT
A,B,XS,PS,USET,USET0,SIL0,SIL,EQEXIN,EDT,CASECC,
EQMAP/
X,R,PG,EPSSE/
SIGN/ITSOPT/ITSEPS/ITSMAX/IPAD/IEXT/ADPTINDX/
NSKIP/MSGLVL/PREFONLY/S,N,ITSERR/SEID $
Input Data Blocks:

A
Square matrix (real or complex, symmetric or unsymmetric).
Rectangular matrix (real or complex), the right-hand side.
XS
Optional starting vector, same type as B (may be purged).
PC
Optional stepwise preconditioner, same type as A (may be purged).
USET
Degree-of-freedom set membership table. See Remark 3.
KGG
Stiffness matrix - g-set. See Remark 3.
GM
Multipoint constraint transformation matrix. See Remark 3.
USET0
USET table from previous adaptivity index in p-version analysis.
SIL
Scalar index list.
SIL0
SIL table from previous adaptivity index in p-version analysis.
EQEXIN
Equivalence table between external and internal grid/scalar

identification numbers. Required for p-version preconditioning only.
EDT
Table of Bulk Data entry images related to element deformation,

aerodynamics, p-element analysis, divergence analysis, and the
iterative solver.
CASECC
Table of Case Control command images. Required if SMETHOD Case

Control command is used and NSKIP=-1.
EQMAP
Table of degree-of-freedom global-to-local maps for domain

decomposition.
CHAPTER 5
Output Data Blocks:

X
Solution matrix. Rectangular matrix having the same dimensions and

type as [B].
Residual matrix. Rectangular matrix having the same dimensions and

type as [B], the residual [R] = [B] - [A][X].
PC1
Updated stepwise preconditioner matrix. See Remark 6.
EPSSE
Table of epsilon and external work.
Parameters:
SIGN
Input-integer-default = 0. Sign flag for [B].

0 : + [B]
1 : - [B]
ITSOPT
Input-integer-default = 0. Preconditioner flag. See Option Selection

on page 88.
ITSEPS
Input-real-default = 1.0E-6. Convergence parameter epsilon.
ITSMAX
Input-integer-default = 0. Maximum number of iterations. The default

value implies N/4 (N = dimension of [A]).
IPAD
Input-integer-default = 0 (see table below). Padding level for reduced or

block incomplete Cholesky factorization (0, 1, 2, ...). See Option
Selection on page 88.
IEXT
Input-integer-default = 0. Extraction level in reduced or block

incomplete Cholesky factorization. See Option Selection on page 88.
ADPTINDX
Input-integer-default=0. P-version analysis adaptivity index. See

Remark 7.
NSKIP
Input-integer-default=1. Record number of current subcase in

CASECC and used only if the SMETHOD command selects the ITER
Bulk Data entry which specifies values for the desired iteration
parameters. If NSKIP=-1 then CASECC is not required and the values
are taken from the module specification of the values.
MSGLVL
Input-integer-default=0. Message level output. See Option Selection

on page 88.
PREFONLY
Input-integer-default=0. Preface execution only. If set to -1 then

SOLVIT is terminated after the preface information is computed and
printed.
ITSERR
Output-integer-default=0. Iterative solver return code.

1no convergence
85
86
2insufficent memory
SEID
Input-integer-default=0. Superelement identification number.
Remarks:
1. If ITSOPT = 3, the IPAD level is recommended to be 0, 1, or 2 (IEXT = 0) and
should be increased when IEXT is increased.
2. The amount of memory needed for ITSOPT = 3, 10, and 11 increases with the
increase of the parameters IPAD and IEXT.
3. For ITSOPT = 1 or 2, the input data blocks USET, KGG, and GM may be purged.
For ITSOPT = 3, USET must be specified. KGG and GM are necessary only if
IEXT = 2.
4. If the message *** USER FATAL MESSAGE 6288 (SITDRV): UNABLE TO
CONVERGE WITH ITERATIVE METHOD is issued, then results will still be
printed but may be inaccurate.
5. If data block PC1 is specified, the CPU time will increase slightly.
6. If SOLVIT is to be used for p-element analysis and ADPTINDX>1, then XS and
PC must be the solution matrix and pre-conditioner from the previous adaptivity
p-level. Also, the USET and SIL from the previous p-level are specified for U and
KGG and the USET and SIL from the current p-level are specified for GM and SIL.
7. For frequency response analysis with ITSOPT=10 or 11 (block incomplete
Cholesky), IEXT=0 is not available and IEXT=1 is used automatically.
CHAPTER 5
5.5
Iterative Method Selection

The NASTRAN keyword ITER (equivalent to SYSTEM(216)) is used to select the iterative
solution by ITER = YES. The default is ITER = NO.
The defaults for the iterative solver can be changed via the Bulk Data entry ITER, which
needs to be selected in the Case Control Section as
SMETHOD = SID
The Bulk Data entry ITER uses a free parameter format as follows:
ITER
SID
PRECOND=CHAR CONV=CHAR
MSGFLG=CHAR ITSEPS=REAL ITSMAX=INT
IPAD=INT IEXT=INT PREFONLY=INT
where CHAR = character type and INT = integer.

Note that the order and existence of the parameters is not mandatory. For example, the
following Bulk Data entry selected by SMETHOD=10 in the Case Control Section is valid:
ITER
10
ITSEPS=1.0E-04 PRECOND=J
This entry chooses a Jacobi preconditioner with 1.0E-04 accuracy.

Continuation lines must start in column 9, aligning with SID. Embedded spaces and
commas are allowed and the conventional continuation entry markers (+xx, +xx) are
obsolete.
All integer and real parameters correspond to the appropriate SOLVIT parameters shown
below. The character parameters used for method and option selections are also described
in Option Selection on page 88.
87
88
5.6
Option Selection
Preconditioner Options
The iterative solver preconditioner options are controlled via the IOPT and the PRECOND
parameters as shown below:
PRECOND
IOPT
Preconditioner
Type
J(S)
1(-1)
Jacobi
real, complex, symmetric,

unsymmetric
C(S)
2(-2)
Incomplete Cholesky
real, complex, symmetric,

unsymmetric
RIC(S)
3(-3)
Reduced Incomplete Cholesky
real, symmetric
USER
User given
real, complex, symmetric
The scaling option is chosen by adding an S to PRECOND or negative sign of IOPT. For
example, CS or -2 means incomplete Cholesky with scaling.
The option PRECOND=USER can be used in the following ways:
For direct frequency response (SOLs 8, 108) it will result in using the direct
method for the first frequency. The factor from this decomposition will then be
used for all subsequent frequencies as the preconditioner with the iterative solver.
In cases where several linear systems of the same size need to be solved, and
where the system matrices differ only slightly, this option can be used with a
DMAP ALTER. A possible scenario is to use the direct method to solve the first
linear system, and then use the SOLVIT module for the solution of the subsequent
systems. Specify the output data block from the decomposition containing the
factor as the 4th input data block (= user given preconditioner) to the SOLVIT
module. The following lines of DMAP show a simple example of the solution of
two linear systems:
AX=F
and
BY=G
where A and B do not differ too much
DECOMP
A/L, /$
FBS
L, , F/X/$
SOLVIT
B,G,,L,,,,,,,/Y,,//4//////-1 $
The hierarchic (for p-version elements) preconditioning strategies are chosen as:
CHAPTER 5
PRECOND
IOPT
Strategy
Type
PBCJ
Incomplete geometric, Jacobi hierarchic
real, symmetric
PBDJ
Complete geometric, Jacobi hierarchic

(default for p-version problems)
real, symmetric
PBDC
Complete geometric, Incomplete hierarchic
real, symmetric
The block incomplete Cholesky (BIC) technique options are:

PRECOND
IOPT
Technique
Type
BIC
11
Well conditioned problems (default for

real problems)
real, symmetric
BICCMPLX
10
Complex problems (default for complex

problems)
complex symmetric
For the distributed parallel SOLVIT only Jacobi preconditioning is available.
Convergence Criterion Options

The convergence criterion options of SOLVIT are selected by CONV as follows:
CONV
Criterion
AR
Eq. 5-6
GE
Eq. 5-8
AREX
Eq. 5-6 and Eq. 5-7
GEEX
Eq. 5-7 and Eq. 5-7
For the BIC method the convergence criterion is automatically set to Eq. 5-13.
Diagnostic Output Option

The diagnostic output from the iterative solver is controlled by MSGFLG or MSGLVL as
follows:
MSGFLG
MSGLVL
Action
89
90
no (default)
Only minimal output (i.e., UIM 6447 [see Option

Selection on page 88], information on whether
convergence was achieved). In case of multiple loads, only
if one or more loads did not converge, a final output is given
for each right hand side.
yes
For 1 RHS: Output above + convergence ratio and norm of

residual for each iteration are given.
For > 1 RHS: Minimal output + final output is given for
each RHS.
The parameter PREFONLY (default=0) can be set to -1, which will terminate the iterative
solver after the preface, giving some helpful information in UIM 4157 (.f04 file) such as
matrix size and optimal memory requirements for best performance.
Incomplete Cholesky Density Option

The density of the incomplete factor is controlled by the IPAD parameter as follows:
IPAD
Default
Method
ITSOPT
Model
Type
Type of
[A]
all
real
reduced incomplete
Cholesky
block incomplete Cholesky
10,11
3-D
real
10,11
2-D or
mixed
real
10,11
all
complex
Extraction Level Option for Incomplete Cholesky

IEXT
Default
Reduced
Block
solid bodies, no rigid elements.
Requires USET and SIL
shells only
Heuristic block structure

(default)
mixed including rigid elements
n/a
Recommendations
The recommended preconditioner options are as follows:
CHAPTER 5
Real symmetric positive definite systems

Sequential
PRECOND = BIC
Parallel
PRECOND = J
Complex symmetric positive definite systems

Sequential
PRECOND =
BICCMPLX
PRECOND = PBDJ
p-version analysis
Sequential
Direct frequency response (SOLs 8, 108)

Sequential
PRECOND = USER
Unsymmetric or indefinite systems

Sequential
Note:
PRECOND = J
Except for SOL 108, all of the above recommended options are also the defaults.
In the case where it is known in advance that a particular model needs to be run many
times (because the model needs to be modified in between, the load vector changes, or for
some other reason), it is highly recommended to determine the best parameter setting for
the iterative solver. To do this, one can simply vary the IPAD value from 1, ... , 4, while
monitoring the EST OPTIMAL MEMORY (see Performance Diagnostics), the size of the
SCRATCH FILE (see Performance Diagnostics), and the CPU time spent in SOLVIT. The
defaults have been selected to give the best performance for most problems, but since the
iterative solver is very sensitive to the conditioning of the matrix, they may not be best for
all problems.
To find the EST OPTIMAL MEMORY without having to go through a complete run,
PREFONLY=-1 can be set on the ITER Bulk Data entry. This option will give the desired
information and cause MSC.Nastran to exit after the symbolic phase of the iterative solver.
The distributed memory parallel execution of the SOLVIT module takes advantage of the
fact that the matrix multiply operation (the most time-consuming part of certain iterative
strategies) is also easily executed while the system matrix resides in parts on the local
memories. This is also the method of the iterative option of the STATICS supermodule.
Examples:
1. Solve [A][X]=[B] with Jacobi pre-conditioning with convergence established at
1.E-4 and maximum allowed iterations of 55 specified for the module parameters.
SOLVIT
A,B,,,,,,,,,/X,,//1/1.E-4/55///-1 $
2. Same as 1 except parameters are obtained from the SMETHOD command and
ITER entry.
SOLVIT
A,B,,,,,,,,EDT,CASECC/X,, $
91
92
5.7
Iterative Solution Diagnostics

There are several different types of diagnostics given by the iterative solver, including:
Accuracy diagnostics (.f06 file)

Performance and memory diagnostics (.f04 file)
Accuracy Diagnostics
In the .f06 output file of an MSC.Nastran run, the following accuracy diagnostics and
convergence messages may be found:
UIM 6447:
ITERATIVE SOLVER DIAGNOSTIC OUTPUT
The following example shows diagnostics for MSCFLG=yes and MSGFLG=no (default).
R
X ( I + 1 ) X ( I )
---------- and -------------------------------------------B
X(I)
are as given in Eq. 5-6 and Eq. 5-13, EPSILON is given in Eq. 5-24, and EXTERNAL WORK
is just the denominator of EPSILON.
MSGFLG=yes
*** USER INFORMATION MESSAGE 6447 (SITDR3)
IPS : 0.9999999975E-06
BIC PRECONDITIONING
ITERATION NUMBER
||R|| / ||B||
1
.1209539266E+00
2
.2251168868E-02
3
.8846335248E-04
4
.1581571489E-05
5
.5083508633E-07
*** USER INFORMATION MESSAGE 5293 (SBUT5 )
FOR DATA BLOCK KLL
LOAD SEQ. NO.
EPSILON
1
-4.3350458E-16
SOLUTION CONVERGED WITH ITERATIVE METHOD.
||X(I+1)-X(I)|| / ||X(I)||
.10000000E+01
.14229420E-01
.61378225E-03
.13831341E-04
.47836956E-06
EXTERNAL WORK
1.2827542E+05
CHAPTER 5
MSGFLG=no (Default)
EPS : 0.9999999975E-06
BIC PRECONDITIONING
ITERATION NUMBER
||R|| / ||B||
5
.5083508633E-07
*** USER INFORMATION MESSAGE 5293 (SBUT5 )
FOR DATA BLOCK KLL
LOAD SEQ. NO.
EPSILON
1
-4.3350458E-16
||X(I+1)-X(I)|| / ||X(I)||
.47836956E-06
EXTERNAL WORK
1.2827542E+05

This last portion of the diagnostic table prints either the last iteration number only or all the
iteration numbers based on the user selection of MSGFLG (see SOLVIT OPTIONS).
MSGFLG=YES
will print the information in every iteration.
MSGFLG=NO
must be set by the user to suppress the information (default).
UIM 6448:
UIM 6320:
SOLUTION WAS CONTINUED BECAUSE EXTERNAL CONVERGENCE CRITERION
WAS PASSED.
This message was printed if convergence is not achieved within the maximum number of
iterations, even though the solution is accepted due to the energy norm check (see
Iterative Solver Solutions on page 76).
UFM 6288:
UNABLE TO CONVERGE WITH ITERATIVE METHOD.
UIM 5293:
FOR DATA BLOCK XX
LOADSEQ NO EPSILON EXTERNAL WORK
Performance diagnostics as well as information about the system matrix and the memory
requirements are given in the .f04 MSC.Nastran output file with SIM 4157, which is also
used in the direct solution. An example is shown below with interpretation of the meaning
of the different variables:
93
94
*** SYSTEM INFORMATION MESSAGE 4157 (SITDR3)

PARAMETERS FOR THE ITERATIVE SOLUTION WITH DATA BLOCK KLL (TYPE = RDP ) FOLLOW
MATRIX SIZE = 134333 ROWS
DENSITY = .00056
STRING LENGTH =
5.19 AVG
NUMBER OF STRINGS =
1910 K
NONZERO TERMS =
10107 K
FULL BAND WIDTH =
6243 AVG
MEMORY AVAILABLE =
37359 K WORDS
IPAD =
2
NUMBER OF RHS =
1
BLOCK SIZE =
5
EST OPTIMAL MEMORY=
27347 K WORDS
EST MINIMUM MEMORY =
4217 K WORDS
MEMORY USED =
36639 K WORDS
PREFACE CPU TIME =
52.49 SECONDS
SCRATCH FILE =
0 K WORDS
AVG. CPU/ITER =
.7693 SECONDS
MATRIX SIZE, DENSITY, STRING LENGTH, NUMBER OF STRINGS, NONZERO

TERMS, and FULL BAND WIDTH are all obvious characteristics of the system matrix.
MEMORY AVAILABLE
K words of memory available to the iterative solver.
IPAD
padding level used (see User Interface on page 84).
NUMBER OF RHS
number of load vectors.
BLOCK SIZE
block size used to block-structure the matrix.
ESTIMATED OPTIMAL
MEMORY
memory needed by iterative solver to run in core;

ensures optimal performance.
ESTIMATED MINIMUM
MEMORY
absolute minimum memory needed for iterative solver

to run; will not give best performance.
MEMORY USED
memory that was actually used by iterative solver.
PREFACE CPU TIME
time required for memory estimation, re-ordering, and

preconditioner calculation.
SCRATCH FILE
=
>
0 in core run
0 amount of spill
The parameters IPAD, BLOCK SIZE, SCRATCH FILE, and MEMORY USED appear only
for BIC preconditioning.
For p-version analysis the additional information is printed in SIM 4157.
GEOMETRIC DOFs
= number of rows in stiffness matrix corresponding to the

geometric degrees of freedom.
HIERARCHIC DOFs
= number of rows in stiffness matrix corresponding to the

p degrees of freedom.
CHAPTER 5
5.8
Iterative Solver Estimates and Requirements

Time estimates for a complete iterative solution are not given, because the number of
operations per iteration is different for each preconditioner. Moreover, it is never known
in advance how many iterations are required to achieve convergence.
However, the following paragraph gives the computation time for one iteration using
Jacobi and BIC preconditioning. The calculation of the computation time t is based on the
operations necessary in each iteration using a particular preconditioner.
Jacobi:
-1 matrix/vector multiplication
-1 preconditioner application r i a ii
-3 dot products
?
t = M N RHS ( N + N + 6 N )
BIC:
-1 matrix/vector multiplication
1
-1 preconditioner application z = P r
-8 dot products
-2 saxpy
?
t = M N RHS ( N + 2 NZ p + 10 N )
where:
N = number of rows in matrix
P = preconditioner
NZ p = number of nonzero terms in P
M = average time of one multiply-add operation
The minimum and optimal memory estimates are equally difficult to determine, since they
also depend on the preconditioner used. Since BIC preconditioning is the most frequently
used option and since the estimates for Jacobi preconditioning are rather straightforward,
some memory estimates for those two options are given below:
Minimum memory in words (MINMEM)

Jacobi:
5 + ( ( 5N RHS + 2 ) N NWPT ) symmetric
5 + ( ( 8N RHS + 4 ) N NWPT ) unsymmetric
95
96
BIC:
0.5 NZA
Optimal memory in words (OPTMEM)

Jacobi:
OPTMEM = MINMEM + ( NZA 2 + N ) NWPT
BIC:
IPAD = 2, 3 OPTMEM 3 NZA
IPAD = 4 OPTMEM 4 NZA
where:
NWPT = number of words per term
1 for long word machines
2 for short word machines
NZA = number of nonzero terms in system matrix
N = number of rows in system matrix
Remark
Memory requirements for BIC are only very rough estimates because of the:
Automatic reduction of IPAD value.

Automatic decision of post-truncation of the preconditioner for well-conditioned
problems.
Recommendation
To find reliable estimates, use:
PREFONLY = -1 on ITER Bulk Data entry.
CHAPTER 5
5.9
References
Babikov, P. & Babikova, M. An Improved Version of Iterative Solvers for Positive Definite
Symmetric Real and Non-Hermitian Symmetric Complex Problems. ASTE, JA-A81,
INTECO 1995.
Conca, J.M.G. Computational Assessment of Numerical Solutions. ISNA 92, Praque, 1992.
Efrat, I.; Tismenetsky, M. Parallel iterative solvers for oil reservoir models. IBM J. Res.
Dev. 30 (2), 1986.
Hageman & Young. Applied Iterative Methods. Academic Press, 1981.
Manteuffel, T. A. An Incomplete Factorization Technique for Positve Definite Linear
Systems, Math. of Computation, Vol 34, #150, 1980.
McCormick, Caleb W., Review of NASTRAN Development Relative to Efficiency of
Execution. NASTRAN: Users Exper., pp. 7-28, September, 1973, (NASA TM X2893).
Poschmann, P.; Komzsik, L. Iterative Solution Technique for Finite Element Applications.
Journal of Finite Element Analysis and Design, 19, 1993.
Poschmann, P.; Komzsik, L., Sharapov, I. Preconditioning Techniques for Indefinite Linear
Systems. Journal of Finite Element Analysis and Design, 21, 1997.
97
98
CHAPTER
Real Symmetric Eigenvalue Analysis

Real Eigenvalue Problems
Theory of Real Eigenvalue Analysis
Solution Method Characteristics
DMAP User Interface
Method Selection
Option Selection
Real Symmetric Eigenvalue Diagnostics
Real Lanczos Estimates and Requirements
References
100
6.1
Real Eigenvalue Problems

The solution of the real eigenvalue problem is very important for many analysis solution
sequences. The problem is numerically difficult and time consuming; therefore,
MSC.Nastran offers methods in two different categories: the reduction (tridiagonal)
method, and the iterative (Lanczos) method.
The problem of normal modes analysis is of form:
[ K ]x = [ M ]x
Eq. 6-1
The problem of the buckling analysis is stated as

[ K ]x = [ K d ]x
Eq. 6-2
where:
[ K ] = the stiffness
[ K d ] = differential stiffness
[ M ] = mass matrices
= eigenvalue
x = eigenvector
These problems may be solved with a reduction type method by transforming to a
canonical form and reducing the whole matrix to tridiagonal form. An iterative method
usually does not modify the matrices K and M ; it may use their linear combination of
[ K ] + s [ M ] where s is a shift value. The Lanczos method, as implemented in
MSC.Nastran, is a method using this technique. The detailed theory of real eigenvalue
analysis is discussed in Theory of Real Eigenvalue Analysis on page 101.
Although the methods are mathematically interchangeable, the Lanczos method is
recommended for the solution of large buckling and normal modes problems, for example,
those arising in the analysis of complete vehicles. The reduction methods are useful for
small normal modes problems in analysis of structural components.
CHAPTER 6
6.2
Theory of Real Eigenvalue Analysis

Two main methods of real eigenvalue extraction are provided in MSC.Nastran in order to
solve the wide variety of problems arising in finite element analysis applications:
1. Reduction (Tridiagonal) Method
2. Iterative (Lanczos) Method
In a reduction method, the matrix of coefficients is first transformed while preserving its
eigenvalues into a special form (diagonal, tridiagonal, or upper Hessenberg) from which
the eigenvalues may easily be extracted. In an iterative method, a certain number of roots
are extracted at a time by iterative procedures applied to the original dynamic matrix. One
of the methods used in MSC.Nastran is a transformation method (tridiagonal method), and
one (shifted block Lanczos method) is an iterative method.
The preliminary transformation procedure of the transformation methods requires that the
major share of the total effort be expended prior to the extraction of the first eigenvalue.
Thus, the total effort is not strongly dependent on the number of extracted eigenvalues. In
marked contrast, the total effort in the iterative methods is linearly proportional to the
number of extracted eigenvalues. Therefore, it follows that the iterative methods are more
efficient when only a few eigenvalues are required and less efficient when a high
proportion of eigenvalues are required.
The general characteristics of the real methods used in MSC.Nastran are compared in
Table 6-1. The tridiagonal method is available only for the evaluation of the vibration
modes of conservative systems and not for buckling analysis due to restrictions on the
matrix form. The Lanczos method is available for all vibration modes and buckling
problems currently solved by MSC.Nastran.
It may be noted from Table 6-1 that a narrow bandwidth and a small proportion of
extracted roots tend to favor the Lanczos method. An example of such a problem is the
evaluation of the lowest few modes of a structure. When the bandwidth is relatively large,
and/or when a high proportion of the eigenvalues are required, the tridiagonal method is
probably more efficient, assuming the problem size is not too large.
The main advantage of including two methods is to provide a backup method if one
method should fail (as sometimes happens with all methods of eigenvalue extraction).
101
102
Table 6-1 Comparison of Methods of Real Eigenvalue Extraction

Tridiagonal
Method
Characteristic/Method
Matrix pencil
Lanczos
Method
( A, I )
( M ( K M )
K
A = ----M
or
or
( K ( K K d )
M, M )
K, K )
M
A = -------------------K + M
Restrictions on matrix
character
A real, symmetrical,
constant
M positive
semidefinite
M Singular
or
or
K positive
semidefinite
K + M Singular
Obtains eigenvalues in order
All at once
Several--nearest to
the shift point
Takes advantage of bandwidth

or sparsity
No
Yes
Number of calculations
0(n )
0 ( nb E )
Recommended
All modes
Few modes
where:
n = number of equations
b = semi-bandwidth or similar decomposition parameter (such as average front size)
E = number of extracted eigenvalues
Reduction (Tridiagonal) Method

The reduction method of MSC.Nastran offers Givens or Householder tridiagonalization
options.
Transformation to Canonical Form

In the tridiagonal method when the eigenvalue problem is solved in a canonical
mathematical form, a Cholesky decomposition is performed as follows:
[M] = [C][C]
Eq. 6-3
CHAPTER 6
where [ C ] is a lower triangular matrix. The procedure used to obtain the factors is
described in Decomposition Process on page 46. The symmetric matrix [ A ] is then
obtained by the following transformation:
[C]
[K ]{ u } [C]
[ C][ C] { u } = 0
Eq. 6-4
Let
{u} = [C]
1, T
{x}
Eq. 6-5
where { x } is the transformed vector. Then

A = [C]
[K][C]
1, T
Eq. 6-6
After the eigenvalue of Eq. 6-4 is found, the eigenvectors can be calculated by using Eq. 6-5.
Tridiagonal Method
Tridiagonal methods are particularly effective for obtaining vibration modes when all or a
substantial fraction of the eigenvalues and eigenvectors of a real symmetric matrix are
desired. The general restrictions on the use of the method within MSC.Nastran are
described in Table 6-1. The basic steps employed in the method are as follows. First, the
canonical matrix is transformed into a tridiagonal matrix
A At
Eq. 6-7
Next A t is transformed to diagonal form:

A t diag ( )
Eq. 6-8
Finally, the eigenvectors are computed over a given frequency range or for a given number
of eigenvalues and are converted to physical form.
Givens Tridiagonalization Method

The most recognized and stable tridiagonalization methods are the Givens and
Householder methods. In the tridiagonal method of MSC.Nastran, both the Givens and
Householder solutions are used. This section describes the Givens solutions, and the next
section describes the Householder solutions.
T
The Givens method depends on orthogonal transformations T] [ A ] [ T ] of a symmetric

matrix [ A ] . An orthogonal transformation is one whose matrix [ T ] satisfies the following:
[T ][ T]
= [T] [T] = [I]
Eq. 6-9
The eigenvalues of a matrix are preserved under an orthogonal transformation since

[T ]([ A] [I ])[ T]
= [ T][A ][T ] [I ]
Eq. 6-10
T
Consequently, if det ( [ A ] [ I ] ) vanishes, then det ( [ T ] [ A ] ( T ) [ I ] ) also vanishes.
103
104
The effect of a series of orthogonal transformations on the eigenvectors of a matrix is a

succession of multiplications by orthogonal matrices. If
[ A]{x} = {x}
Eq. 6-11
and if 1 ] , [ T 2 ] , , [ T are orthogonal matrices, then

[ T r ] [ T r 1 ] [ T 2 ] [ T 1 ] [ A ] { x } = [ T r ] [ T r 1 ] [ T 2 ] [ T 1 ] { x }
Eq. 6-12
Through the substitution

T
{ x } = [ T1 ] [ T2 ] [ Tr 1 ] [ Tr ] { y }
Eq. 6-13
we obtain
T
[ T r ] [ T r 1 ] [ T 2 ] [ T 1 ] [ A ] [ T 1 ] [ T 2 ] [ T r 1 ] [ T r ] { y }
= [ T r ] [ T r 1 ] [ T 2 ] [ T 1 ] [ T 1 ] [ T 2 ] [ T r 1 ] [ T r ] { y }
= {y}
Eq. 6-14
where Eq. 6-9 is applied repeatedly to obtain the final form. Here { y } is an eigenvector of
the transformed matrix:
T
[ T r ] [ T r 1 ] [ T 2 ] [ T 1 ] [ A ] [ T 1 ] [ T 2 ] [ T r 1 ] [ T r ]
and that { x } may be obtained from { y } by:

T
{ x } = [ T1 ] [ T2 ] [ Tr 1 ] [ Tr ] { y }
Eq. 6-15
The Givens method uses orthogonal matrices [ T ] , which are identical with the unit
matrix [ I ] except for the four elements:
t i + 1, i + 1 = t j, j = cos i + 1, j
t i + 1, j = t j, i + 1 = sin i + 1, j
Eq. 6-16
The orthogonal transformation T] [ A ] [ T ] leaves all the elements of [ A ] unchanged except

those in the ( i + 1 ) st and j-th rows and columns, the so-called plane of rotation. The four
pivotal elements of the transformed matrix are:
2
a i + 1, i + 1 = a i + 1, i + 1 cos i + 1, j + a j, j sin i + 1, j + a i + 1, j sin 2 i + 1, j
a j, j = a i + 1, i + 1 sin 2 i + 1, j + a j, j cos i + 1, j a i + 1, j sin 2 i + 1, j Eq. 6-17
1
a i + 1, i = a j, i + 1 = a i + 1, j cos 2 i + 1, j --- ( a i + 1, i + 1 a j, j ) sin 2 i + 1, j
2
where a j, j , etc., are elements of the untransformed matrix. The other elements of the
( i + 1 ) st and j-th rows and columns of the transformed matrix are:
CHAPTER 6
a i + 1, s = a s, i + 1 = a i + 1,
a j, s = a s, j = a i + 1,
s cos i + 1, j
+ a j, s sin i + 1, j
s sin i + 1, j + a j, s cos i + 1, j
Eq. 6-18
In the Givens method, i + 1, j is chosen such that a i, j vanishes, which happens when
a i, j
tan i + 1, j = -----------------a i, i + 1
Eq. 6-19
The calculation of i + 1, j followed by the orthogonal transformation

[A
[m]
] = [ Tm ] [ A
(m 1)
] [ Tm ]
Eq. 6-20
(0)
is carried out for a sequence of iterations with [ A ] = [ A ] . The values of i used in Eq. 68, Eq. 6-9, Eq. 6-10, and Eq. 6-11 are 1, 2, 3, , ( n 1 ) . For each i , a set of ( n i 1 )
transformations is performed with j assuming the values of i + 2, i + 3, , n before the
next value of i is used. As a result, the elements in the matrix positions ( 1, 3 ), ( 1, 4 ), ,
( 1, n ), ( 2, 4 ), ( 2, 5 ), , ( 2, n ), , ( n 2, n ) are successively reduced to zero together
with their transposes, the ( 3, 1 ), ( 4, 1 ), , ( n, n 1 ) elements. Thus, the set of
transformations reduces the matrix to tridiagonal form.
MSC.Nastran employs a procedure introduced by Wilkinson (1965) in which the Givens
method is modified by grouping the ( n i 1 ) transformations together, which produces
zeroes in the i-th row and column. This procedure should not be confused with the
Householder method which eliminates a row and column at a time. The Wilkinson process
is particularly advantageous when the matrix [ A ] is so large that all the elements cannot
be held in memory at one time. This process only requires ( n 2 ) transfers of the matrix
to and from auxiliary storage instead of the ( n 1 ) ( n 2 ) 2 transfers required by the
unmodified Givens method. This method requires 4n memory locations for working space
that are divided into four groups of n storage locations each. The first ( i 1 ) rows and
columns play no part in the i-th major step. This step has five substeps as follows:
1. The i-th rows of [ A ] are transferred to the set of memory locations in group 1.
2. The values of cos i + 1, i + 2 , sin i + 1, i + 2, , cos i + 1, n , sin i + 1, n are
computed successively from the following:
(j 1)
a i, j + 1
cos i + 1, j = -------------------------------------------------------(j 1) 2
2
( a i, i + 1 ) + a i, j 1
a i, i
sin i + 1, j = -----------------------------------------------(j 1) 2
2
( a i, i + 1 ) + a i, j
Eq. 6-21
where the superscripted term is computed by the following:

(j 1)
a i, i + 1 =
(j 2) 2
( a i, i + 1 ) + a i, j 1
Eq. 6-22
105
106
and the starting value for j = i + 2 is as follows:

(i + 1)
a i, i + 1 = a i, i + 1
Eq. 6-23
The cos i, j may be overwritten on those elements of the untransformed matrix

a i, j which are no longer required, and the sin i + 1, j is stored in the group 2
storage locations.
3. The ( i + 1 ) st row of [ A ] is transferred to the group 3 storage locations. Only
those elements on and above the diagonal are used in this and succeeding rows.
For k = i + 2, i + 3, , n , in turn, the operations in substeps 4 and 5 are carried
out.
4. The k-th row [ A ] is transferred to the group 4 storage locations. The elements
a i + 1, i + 1 , a i + 1, k , and a k, k are subjected to the row and column operations
involving cos i, k and sin i + 1, k . For i = k + 1, k + 2, , n in turn, the part
of the row transformation involving cos i + 1, k and sin i + 1, k is performed on
a i + 1 and a k . At this point, all the transformations involving the i-th major step
were performed on all the elements of row k and on the elements i + 2 , i + 3, ,
n of row ( i + 1 ) .
5. The completed k-th row is transferred to auxiliary storage.
When substeps 4 and 5 are complete for all appropriate values of k , the work on row
( i + 1 ) is also complete. The values of cos i + 1, k and sin i + 1, k for k = i + 2 , i + 3 ,
, n are transferred to the auxiliary storage, and row ( i + 1 ) is transferred to the group 1
core storage locations. Since the ( i + 1 ) row plays the same part in the next major step as
in the i-th in the step just described, then the process is ready for substep 2 in the next major
step. In fact, substep 1 is only required in the first major step because the appropriate row
is already in the storage locations of group 1 while performing subsequent steps.
Householder Tridiagonalization Method

The Householder tridiagonalization method uses the following transformation:
Ar = Pr Ar 1 Pr
Eq. 6-24
where:
A0 = A
T
Pr
T
wr wr
= I 2 wr wr
=1
The elements of w r are chosen so that A r has zeroes in the r-th row except the three
diagonal positions. The configuration and partitioning of the A r 1 can be shown as:
CHAPTER 6
x x
x x x
x x
x
x
x
r
Ar 1 =
nr
0
Cr 1
x
x
x
x
x
x
x
x
x
x
x
x
br 1
=
0
br 1
Br 1
where:
C r 1 = a tridiagonal matrix of order r (partial result)
B r 1 = a square matrix of order n r (part of original matrix)
b r 1 = a vector having n r components
The P r transformation matrix can be partitioned as
Pr =
I 0
nr
0 Qr
0 I 2v r v r
where v r = a vector of order n r

By executing the multiplication given in the right-hand side of Eq. 6-24, the following is
obtained:
0
r
Ar 1 =
nr
Cr 1
0
cr
cr Qr Br 1 Qr
where c r = Q r B
.
r1
If we have chosen v r so that c r is null except for its first component, then the first ( r + 1 )
rows and columns of A r are tridiagonal.
An algorithm formulation can be developed by writing P r in the following form:
T
P r = I 2w r w r
Eq. 6-25
where w r is a vector of order n with zeroes as the first r elements. Adding further changes
as follows:
107
108
ur ur
P r = I ----------T
2K r
Eq. 6-26
where:
u r = 0 i = 1, 2, , r
u r + 1, r = a r, r + 1 S r
u ir = a ri ni = r + 2, , n
2
Sr
2
2K r
a ri
i 2= r + 1
= S r a r, r + 1 S r
and a ij is the ( i, j ) element of A . Substituting Eq. 6-26 into Eq. 6-24, the following
equation results:
T
T
ur ur
ur ur
A r = I ----------- A r 1 I -----------
2
2
2K r
2K r
Eq. 6-27
with a new notation as follows:

Ar 1 ur
p r = -------------------2
2K r
Eq. 6-28
Now the form becomes the following:

T
Ar = Ar 1
T
ur pr
T
pr ur
u r ( u r p r )u r
+ ----------------------------2
2K r
Eq. 6-29
By introducing one more intermediate element

T
1 u r p r
q r = p r --- u r ----------2 2K 2
Eq. 6-30
the final algorithmic form is obtained

T
Ar = Ar 1 ur qr qr ur
which takes advantage of the symmetry.
Eq. 6-31
CHAPTER 6
Modified Tridiagonal Methods

The modified tridiagonal methods (i.e., modified Givens or Householder) are used if M is
singular. These methods have a different transformation scheme to obtain the canonical
form. First, perform a Cholesky decomposition on a shifted matrix as follows:
( [ K ] + s [ M ] ) = [ C ] [ C ]
Eq. 6-32
where s is a positive shift value obtained from the diagonal terms of [ K ] and [ M ]
matrices. The new form of Eq. 6-7 is now
[ A ] I { x } = 0
Eq. 6-33
where:
1
--------------- =
+ s
1
1, T
[A ] = [C ] [M ][C]
The s shift value is calculated as:
1
s = --------------------------------------n
M
12
ii
-------n
K ii
i = 1
where M ii and K ii are the diagonal elements of M and K respectively. If M ii = 0 or

8
K ii = 0 or M ii K ii > 10 , the term is omitted from the summation. If all terms are
omitted, s is set to 0.001. If the s value calculated does not result in a stable Cholesky
decomposition, the value is slightly perturbed, and the decomposition is repeated up to 2
times. After three unsuccessful shifts, a fatal error is given.
The details of the above equation can be seen by rearranging Eq. 6-7 as follows:
( [ K ] + s [ M ] ( + s ) [ M ] ) { u } = 0
Eq. 6-34
Then by premultiplying 1 ( + s ) and substituting Eq. 6-32 the following is obtained:

T
[C ][C ]
[ M ] ---------------------- { u } = 0
+ s
Finally, premultiplying by [ C ]
and substituting the term

T
{x} = [C] {u}

gives the form of Eq. 6-33.
Eq. 6-35
Eq. 6-36
109
110
QR Method of Eigenvalue Extraction

In the tridiagonal methods of Givens or Householder, MSC.Nastran employs the QR
transformation of Francis (1962) which is an orthogonal transformation
T
[ A r + 1 ] = [ Q r ] [ A r ] [ Q r ] so that [ A r ] may be factored into the product [ Q r ] [ R r ]
where [ R r ] is an upper triangular matrix. Thus,
[ Ar ] = [ Qr ] [ Rr ]
Eq. 6-37
and
T
[ Ar + 1 ] = [ Qr ] [ Ar ] [ Qr ]
T
= [ Qr ] [ Qr ] [ Rr ] [ Qr ]
T
Now [ Q r ] [ Q r ] = [ I ] by virtue of the orthogonality property; therefore,

[ Ar + 1 ] = [ Rr ] [ Qr ]
Eq. 6-38
It follows that [ A r + 1 ] is determined from [ A r ] by performing in succession the

decomposition given by Eq. 6-37 and multiplication. Francis has shown that if a matrix
[ A ] = [ A 1 ] is nonsingular, then [ A r ] approaches an upper triangular matrix as r .
Since eigenvalues are preserved under orthogonal transformation, it follows that the
diagonal elements of the limiting matrix are the eigenvalues of the original matrix [ A ] .
Although the method can be applied to any matrix, it is particularly suitable for tridiagonal
matrices because the bandwidth of the matrix can be preserved as will be shown. In the
case where [ A ] is symmetric, the matrix [ A r ] tends to take a diagonal form as r .
Even though the upper triangular matrix [ R r ] and the orthogonal matrix [ Q r ] are unique,
several methods are available for performing the decomposition. In the method of
calculation devised by Francis, [ Q r ] is expressed as a product of ( n 1 ) elementary
rotation matrices where n is the order of [ A r ] as follows:
[ Qr ] = [ T
(1)
][T
(2)
] [ T
(n 1)
Eq. 6-39
The nonzero elements of the j-th elementary rotation matrix are the following:
(j)
(j)
(j)
(j)
t j + 1, j = t j, j + s
(j)
k j or j + 1
t k, k = 1
t j, j = t j + 1, j + 1 = c j
Eq. 6-40
The manner in which the c j and s j coefficients are obtained from the elements of [ A r ] are
shown later. From the orthogonality property
CHAPTER 6
[ Rr ] = [ Qr ]
= [T
[ Ar ] = [ Qr ] [ Ar ]
(n 1) T
] [T
(n 2) T
] [T
(2) T
] [T
(1)
Eq. 6-41
] [ Ar ]
we can define the nonzero elements of [ A r ] , [ A r + 1 ] and [ R r ] as follows:

a1
b1
b2
a2
[ Ar ] =
0
.
b3
.
bn 1 an 1
bn
a1
b1
b2
a2
.
[ Ar + 1 ] =
Eq. 6-42
bn
an
0
b3
Eq. 6-43
.
bn 1 an 1
0
bn
bn
an
and
r1
q1
t1
r2
q2
0
t2
.
[ Rr ] =
rn 1
0
Eq. 6-44
.
qn 1
tn 1
rn
The coefficients of the elementary rotation matrices are selected to reduce the subdiagonal
terms of [ R r ] to zero. Specifically,
pj
c j = ------------------------------------------
12
2
2
( pj + bj + 1 )
bj + 1
s j = -----------------------------------------12
2
2
( pj + bj + 1 )
j = 1, 2, , n 1
Eq. 6-45
111
112
and
p1 = a1
p2 = c1 a 2 s1 b 2
pj = cj 1 a sj 1 cj 2 bj
j
Eq. 6-46
j = 3, 4, , n 1
Substitution yields the elements of [ R r ] as follows:

j = 1, 2, , n 1
rj = cj p j + sj bj + 1
rn = p n
q1 = c1 b2 + s1 a2
qj = sj a j + 1 + cj cj 1 bj + 1
tj = sj bj + 2
j = 2, 3, , n 1
j = 1, 2, , n 2
Eq. 6-47
From Eq. 6-38, the elements of A r + 1 are the following:
j = 2, 3, , n 1
j = 1, 2, , n 2
a1 = c1 r1 + s1 q1
aj = cj 1 cj rj + sj qj
an = cn 1 rn
bj + 1 = s r
Eq. 6-48
MSC.Nastran uses a variation of Franciss original method that avoids the calculation of
square roots. This variation uses the following equations in place of Eq. 6-48 as follows:
2
a j = ( 1 + s j )g j + s j a j + 1
an = gn
2
bj + 1 = sj ( pj + 1 + bj + 2 )
2
bn = sn 1 pn
and
j = 1, 2, , n 1
j = 1, 2, , n 2
Eq. 6-49
Eq. 6-50
CHAPTER 6
g1 = a1
gj = cj 1 pj = aj
2
2
sj 1 ( aj
g1 = a1
g2
j
-----------
2
2
g1 = cj 1
c2 b2
j2 j
if c j 1 0
if c j 1 = 0
+ gj 1 )
j = 2, 3, , n
j = 2, 3, , n
j = 2, 3, , n
Eq. 6-51
The reason that the use of Eq. 6-49, Eq. 6-50, and Eq. 6-51 in place of Eq. 6-48 avoids the
calculation of square roots can best be seen by considering the terms input to and produced
by these equations. For Eq. 6-48, the input terms consist of the elements of [ A r ] (which are
a 1, a 2, , a n ) and b 2, b 3, , b n (which are the elements of [ A r + 1 ] ). This completes one
iteration step but involves the calculation of square roots. However, for Eq. 6-49, Eq. 6-50,
2 2
2
and Eq. 6-51, the input terms consist of a 1, a 2, , a n and b 1, b 2, , b n . The data
produced is a 1, a 2, , a n and b 2, b 3, , b n , which serve as the input to the next iteration
step. No square roots require computation here.
Convergence of the tridiagonal matrix to a diagonal form is speeded up by origin shifting.
MSC.Nastran employs a procedure suggested by Wilkinsons quadratic shift which has
shown that when the eigenvalues are real, the best shift strategy is to subtract a n from each
diagonal element of [ A ] and thereby reduce each eigenvalue by a n .
Another device useful in speeding up the determination of eigenvalues of tridiagonal
matrices takes advantage of zeroes that may occur in the off-diagonal rows. Let [ A r ]
matrix be represented in partitioned form as follows:
113
114
a1 b1
b1 a2 b2
.
.
.
aj 2
bj 2
bj 2
aj 1
aj
bj
bj aj + 1 bj + 1
.
.
.
Eq. 6-52
bm 2 am 1 bm 1
bm 1 am 1
am + 1
.
an
In this matrix the diagonal terms in the lower right-hand partition are eigenvalues that
were determined in previous iterations. The j-th is the next lowest row in which the offdiagonal term b j 1 is zero thereby uncoupling the equations in the first j 1 rows from
the rest. As a result, the eigenvalues of the matrix in the central block may be obtained
separately. Other uncoupled blocks may be found in the upper left partition.
2
The iteration described by Eq. 6-49 and Eq. 6-50 is continued until ( b m 1 ) decreases to a
satisfactory level of accuracy so that a m may be accepted as an eigenvalue of the shifted
2
2
2
2
matrix. ( b m 1 ) must be negligible compared to ( a m ) , that is, ( b m 1 ) + ( a m ) must
2
approximately equal ( a m ) . Then ( a m ) is transferred to the lower partition, and the
process is continued until all the eigenvalues of the partitioned matrix are extracted.
Computation of Eigenvectors
The eigenvector corresponding to any eigenvalue i of the tridiagonal matrix may be
determined by solving n 1 of the following equations:
( a 1 i )x 1 + b 2 x 2 = 0
b 2 x 1 + ( a 2 i )x 2 + b 3 x 3 = 0
.
.
.
bn 1 xn 1 + ( an 1 ) xn 1 + bn xn = 0
i
b n x n 1 + ( a n i )x n = 0
Eq. 6-53
CHAPTER 6
If the first ( n 1 ) equations are used, the solution is obtained by taking x 1 to be unity and
substituting in the equations to obtain values of x 2, x 3, , x n . Wilkinson (1958) shows that
this method is unstable and suggests the following approach, which is used in
MSC.Nastran.
The tridiagonal matrix [ A i I ] is factored into the product of a lower unit triangle [ L i ]
and an upper triangle [ U i ] . Partial pivoting is used in this decomposition; i.e., the pivotal
row at each stage is selected to be the equation with the largest coefficient of the variable
being eliminated. At each stage, there are only be two equations containing that variable.
The eigenvector [ i ] is then obtained from the solution of the equation as follows:
[ Ui ] [ i ] = [ C ]
Eq. 6-54
where [ C ] is randomly selected. The solution is easily obtained by backward substitution

because [ U i ] takes the form as follows:
p1 q1 r1
p2 q2
[ Ui ] =
r2
.
Eq. 6-55
pn 2 q n 2 rn 2
pn 1 qn 1
pn
An improved solution is obtained by repeated application of Eq. 6-54 using the current
estimate of [ i ] on the right-hand side. Thus,
(n)
(n 1)
[ Ui ] i = i
Eq. 6-56
where
0
1 = [ C ]

Wilkinson showed that, if the computed eigenvalue i is a close approximation to the true
eigenvalue, convergence is so rapid that not more than two iterations are required. The
applied test is for the maximum component of the eigenvector to remain unchanged (to
single precision accuracy) in one iteration. The initial vector [ C ] is chosen so that each
element is unity.
In the case of a double eigenvalue, the above method gives one eigenvector { 1 } . If you
start with any initial vector { b } orthogonal to { 1 } and apply the previous algorithm,
convergence to the other eigenvector { 2 } results. The following procedure is used in
MSC.Nastran to compute eigenvectors associated with multiple eigenvalues. If
T
eigenvectors { 1 }, { 2 }, , { m } with elements { d } = { a 1s, a 2s, , ans } are obtained,
115
116
an initial vector { b } orthogonal to each of these eigenvectors is obtained by taking the

components b m + 1, b m + 2, , bn as unity and solving the simultaneous equations as
follows:
b 1 a 11 + b 2 a 21 + + b m a m1 = ( a s1 )
s = m+1
b 1 a 12 + b 2 a 22 + + b m a m2 = ( a s2 )
s = m+1
.
.
.
b 1 a 1m + b 2m a 2m + + b m a mm = ( a sm )
s = m+1
Eq. 6-57
Accumulated round-off errors result in the computed multiple eigenvectors since they are
not exactly orthogonal to one another. The following Gram-Schmidt algorithm is used to
produce an orthogonal set of k eigenvectors { y s } from the almost orthogonal set { xs } . For
s = 1 , select as follows:
{ x1 }
{ y 1 } = ------------------{ x1 }
Eq. 6-58
Then for 1 < s k , calculate as follows:

s1
{ zs } = { xs }
( { xs } { yt } ) { yt }
Eq. 6-59
t = 1
zs
{ y s } = -----------------{ zs }
where { z s } denotes the Euclidean norm
2
z s1 + z s2 + + z sn
T
of vector { z s1, z s2, , z sn } and { x s } { y t } is a scalar product of the vectors { x s } and { y t } .

When all the eigenvectors of the tridiagonal matrix are extracted, the back transformations
are carried out to obtain the eigenvectors of the original matrix [ A ] . Finally, for all of the
tridiagonal methods, the eigenvectors are given one last stage of orthogonalization as
follows:
T
M modal = M
CHAPTER 6
Find L , such that

T
L L = M modal
by Cholesky factorization. Find the reorthogonalized eigenvectors by a forward solution
T
T
pass of L = . It can be shown that this operation is equivalent to orthogonalizing the
second vector with respect to the first, orthogonalizing the third with respect to the
purified second and first vector, , and orthogonalizing the last vector with respect to all
of the other purified vectors. If M modal is poorly conditioned with respect to the Cholesky
decomposition, a fatal error message is produced.
Shared Memory Parallel Householder Method

The most time-consuming part of the reduction type methods is the reduction to
tridiagonal form. Besides being more easily vectorizable, the Householder method also
lends itself easily to parallelization. The computer science aspects of the parallel
Householder method (i.e., creating subtasks, waiting, etc.) are similar to those of the
parallel methods mentioned previously and are not detailed here. The necessary
mathematical details follow.
A logical suggestion is to aggregate the Householder transformations, i.e., work in
column blocks instead of columns. For example, by observing the first k steps, the
following can be written:
[ A ] k = ( [ P ] k [ P ] k 1, , [ P ] 1 )A ( [ P ] k [ P ] k 1, , [ P ] 1 )
T
T
= [ Q ]k [ A ] [ K ]k
where:
[ Q ]k = I + [ W ] [ Y ]T
k
k
[ W ] and [ Y ] = n k matrices ( WY representation)
Completing the process in k-size blocks, produces the following:
T
[ A ]j = [ Q ]j [ A ] [ Q ]j
where [ Q ] j = [ P ] jk [ P ] jk 1 , [ P jk k + 1 ]
j = 1, 2, , n k
117
118
Before the j-th block step, the structure of A is as follows:
[A ] =
[ Cj 1 ]
[ Bj 1 ]
(j 1) k
(j 1)
[ A1
n ( jk )
where:
C j 1 = tridiagonal matrix of order n ( j 1 )k
B j 1 = general matrix of order n k
A
(j 1)
= general matrix of order n ( n jk )

T
The process is then to generate I + W j Y j so that B j 1 assumes tridiagonal form and

afterwards update the remainder of the matrix as shown:
A
(j)
= ( I + W j Y j )A
(j 1)
T T
( I + Wj Yj )
This block update can be parallelized and forms the parallel portion of the MSC.Nastran
Householder code. It is intended for shared memory parallel computers. It does not run on
distributed memory computers. A similar approach can also be used in the reduction to
Hessenbergs form, which is one of MSC.Nastrans complex eigenvalue methods.
However, this approach is not implemented as yet.
Real Symmetric Lanczos Method

The basic Lanczos recurrence (Lanczos, 1950) is a transformation process to tridiagonal
form. However, the Lanczos algorithm truncates the tridiagonalization process and
provides approximations to the eigenpairs (eigenvalues and eigenvectors) of the original
matrix. The block representation increases performance in general and reliability on
CHAPTER 6
problems with multiple roots. The matrices used in the Lanczos method are specifically
selected to allow the best possible formulation of the Lanczos iteration and are described
in Table 6-1 of Theory of Real Eigenvalue Analysis on page 101.
Basic Lanczos Recurrence
The basic Lanczos algorithm solves the problem as follows:
Ax = x
Eq. 6-60
and can be formulated as follows:

1. Initialization
a. A starting vector q 1 is chosen with q 1
b. 1 = 0 and q 0 = 0 are set.
= 1.
2. lteration
For j = 1, 2, 3, , iterate until convergence as follows:
r j + 1 = Aq j j q j j q j 1
where:
j = q T Aq
j
j
j + 1 =
rj + 1
qj + 1 = rj + 1 j + 1
If this algorithm is carried out without round-off errors, the vectors q 1, q 2, q 3, , are
orthonormal. These vectors can be considered to be the column of an orthogonal matrix
Q j = ( q 1, q 2, , q j ) . The scalars j and j can be combined in a tridiagonal matrix T j
as follows:
1 2
2 2 3
3 3 4
.
.
Tj =
.
i i i + 1
.
.
.
j
j j
119
120
With this notation, the first j steps of the Lanczos algorithm can be combined in matrix
form as follows:
T
AQ j Q j T j = j + 1 q j + 1 e j
Eq. 6-61
where the vector e j is the unit vector with zero entries except for the j-th row which holds
the value of one.
This mathematical presentation is not sufficient for a practical implementation of the
Lanczos algorithm that is capable of coping with finite precision arithmetic. The
computation of the Lanczos loop can be carried out in many mathematically equivalent
ways, one of which is the best for finite arithmetic, while several others do not work in
finite arithmetic (Grimes, R.G., et al.). Even with the recurrence written in the proper form,
round-off errors can cause the algorithm to behave quite differently from the ideal
description given above. More details on the effects of round-off errors are given in the
section on preserving orthogonality.
The Lanczos algorithm is valuable for solving sparse eigenvalue problems because it needs
only a partial tridiagonalization and does not actually modify the matrix A . The matrix A
enters the algorithm only in the formation of the matrix vector product Aq . In practice, this
means that the sparsity of A can be exploited in a natural way since all that is needed is an
efficient subroutine to compute Av from v .
The eigenvalues and corresponding eigenvectors of A are computed from those of T . Let
and s be an eigenpair of the tridiagonal matrix T j ; that is, let
T j s = s
Eq. 6-62
Then and y with y = Q j is an approximate eigenpair for the original problem. Note
that in a typical application, the order of matrix A may be in the tens of thousands, whereas
the order of T j is about 20 to 30. Hence, it is much easier to solve the eigenvalue problem
of Eq. 6-62 than of Eq. 6-61. But the question is: How good is the approximate eigenpair
, y , or more directly, when should the Lanczos recurrence be stopped so that the
eigenpairs are satisfactorily accurate?
The correct answer requires the norm of the residual for the eigenpair Ay y , which
requires the full computation of a tentative eigenpair. In fact, this residual norm can be
obtained from Eq. 6-62 by observing that
Ay y = j + 1 q j + 1 s j
Eq. 6-63
where s j is the j-th or bottom element of the eigenvector s . Hence, the norm of the residual
is given by j + 1 s j and the quantity can be computed without computing the eigenvector
y explicitly. A small extra effort allows the convergence of the eigenpairs to be monitored
and terminates the Lanczos loop whenever the required number of sufficiently accurate
eigenpairs is found. The eigenvector y is computed only once at the end.
CHAPTER 6
Why should it be expected that some of the eigenvalues of the tridiagonal matrix T j
converge quickly to the eigenvalues of A ? This question is answered by some intricate
mathematical theorems which guarantee that, under reasonable assumptions on the
properties of the starting vector, approximations to some of the extreme eigenvalues
appear in the tridiagonal matrix very quickly. Two factors can impede this rapid
convergence. If the starting vector has no components in the direction of one of the
eigenvectors, the convergence of this eigenvector is delayed. A cluster of poorly separated
eigenvalues also causes slow convergence.
Shifted Algorithm
The basic Lanczos algorithm as discussed in the previous section must be modified in two
respects to solve the practical vibration problem. The vibration problem is a symmetric
generalized eigenvalue problem using the following form:
Kx = Mx
Eq. 6-64
where K is a symmetric matrix, M is a symmetric positive semidefinite matrix, and is

the eigenvalue. If M is positive definite, the Cholesky factorization of M can be computed
and Eq. 6-64 can be reduced to an ordinary eigenvalue equation. However, there are
considerably more difficulties when M is only semidefinite (nonnegative definite). The
implementation of the Lanczos algorithm in MSC.Nastran avoids the Cholesky
factorization of M to use the more general form of the problem.
The typical eigenvalue problem is to compute the smallest (closest to zero) eigenvalues and
eigenvectors of Eq. 6-64. These eigenvalues are usually clustered closely together when
compared to the largest eigenvalues. The Lanczos algorithm applied directly to Eq. 6-65
yields precise approximations to these large, well-separated, but uninteresting
eigenvalues, and poor approximations to the small, clustered, and interesting eigenvalues.
In order to overcome this convergence difficulty, the Lanczos algorithm is applied to the
shifted and inverted eigenvalue problem in the following equation:
M ( K M )
1
Mx = -------------- Mx
Eq. 6-65
In Eq. 6-65 the shift s is chosen close to the eigenvalues of interest. This formulation of the
shifted and inverted eigenvalue problem is only one of several possible ones. Eq. 6-65 is the
preferred choice for vibration problems because of its improved rate of convergence to the
desired eigenvalues and the fact that the eigenvectors of Eq. 6-65 are also eigenvectors of
Eq. 6-64.
The relationship between the Lanczos algorithm of Eq. 6-64 and Eq. 6-65 is comparable to
the relationship between the power method and the inverse power method (inverse
iteration). The choice of shift in the inverse power method directs the convergence toward
the eigenpair closest to the shift. For well-chosen shifts, the convergence can be very rapid.
The same is true for the shifted and inverted Lanczos algorithm. A properly chosen shift
guarantees rapid convergence not only to the closest eigenpair but also to a whole group
of eigenpairs in the vicinity of the shift. The cost of this accelerated convergence towards
the desirable eigenvalues is the factorization of ( k M ) which, for large matrices, can be
121
122
the dominating cost in either algorithm. The computational superiority of the Lanczos
algorithm derives from its more thorough use of this expensive factorization. Even a
relatively short run of the shifted and inverted Lanczos algorithm can extract many
eigenvalues using only one factorization.
The Lanczos algorithm applied to the vibration problem of Eq. 6-65 can be formulated as
follows:
1. Initialization
a. A starting vector r 1 is chosen.
1
b. r 1 = ( K M ) Mr 1 is computed.
T
12
c. q 1 = r 1 ( r Mr 1 )
is normalized.
d. 1 = 0 and q 0 = 0 .
2. lteration
For j = 1, 2, 3, iteration is performed until convergence occurs.
r j + 1 = ( K M )
Mq j j q j j q j 1
Eq. 6-66
where:
j
j + 1
q j M ( K M )
T
( r j + 1 Mr j + 1 )
Mq j
12
qj + 1 = rj + 1 j + 1
There are two remarkable facts about this formulation of the Lanczos algorithm for the
eigenvalue problem. First, no factorization of the possibly singular mass matrix M is
required. Second, the Lanczos vectors q 1, q 2, are orthogonal with respect to the inner
product defined by M . When M is semidefinite, it does not define a true inner product. In
that case, the initialization step guarantees that all Lanczos vectors are orthogonal to the
null space of M . Thus, the Lanczos algorithm works implicitly only with the components
of M that are orthogonal to its null space. This restriction of M to a positive definite matrix
and the use of the M inner product causes no problem. An extra benefit is that the Lanczos
algorithm in the version above generates no approximations to the infinite eigenvalues of
Eq. 6-65 that arise when M is semidefinite.
The tridiagonal matrix T j defined in Eq. 6-62 only contains approximations to the finite
eigenvalues of Eq. 6-65 via the so-called spectral transformation. For example, if is an
eigenvalue of T j , then
1
= + --
Eq. 6-67
is an approximate eigenvalue of Eq. 6-65. The approximate eigenvector y can be computed

in the same way as in the previous section. Since the Lanczos vectors are M -orthogonal, the
eigenvectors corresponding to different eigenvalues are also M -orthogonal.
CHAPTER 6
To solve the buckling problem of the form

Kx = K d x
Eq. 6-68
a different shifting and inverting strategy is required. In Eq. 6-69, the differential stiffness
matrix K d is merely symmetric and has no definite properties, whereas the stiffness matrix
K is positive semidefinite. The semidefinite property of M in the vibration analysis is
critical to the Lanczos algorithm. The shifting strategy applied in the vibration case cannot
be applied to the buckling case by substituting K d for M .
Since the eigenvalues closest to zero are usually wanted, a simple approach is to
interchange the roles of K and K d and then compute the largest eigenvalues of the
problem. Therefore, by applying the Lanczos algorithm without shift,
Kd = K x
Eq. 6-69
where = 1 . This approach has two drawbacks. First, it requires the factorization of
the possibly semidefinite matrix K , thereby preventing its use in problems with rigid body
modes. Second, it does not allow any shifting for other eigenvalues.
A general shifting and inverting strategy is possible for the buckling problem. As shown
previously, the operator K K d is factored for an arbitrary shift, but the Lanczos
recurrence is carried out using K -orthogonality among the Lanczos vectors. Each
multiplication by the mass matrix M in the vibration case is replaced by a multiplication
by the stiffness matrix in the buckling case. The rest of the Lanczos recurrence remains the
same. Hence, in the buckling case the Lanczos algorithm works with the operator
1
( K K d ) K and K -orthogonality.
This shifted and inverted operator allows for arbitrary shifts with the exception of a shift
at zero that reduces the problem to an identity matrix. For all other shifts, an eigenvalue
of T j can be transformed as
= ------------1
Eq. 6-70
to yield an approximate eigenvalue of Eq. 6-69. Eigenvectors are computed as before

without any additional back transformation resulting in a K -orthogonal set. Since the
stiffness matrix K is used in the initialization step in the same way as M is used in the
vibration problem, the sequence of Lanczos vectors and hence the eigenvectors are
orthogonal to the null space of K . Therefore, T j does not yield any approximation to the
exact zero eigenvalues of the buckling problem. The rigid body modes of the structure are
not computed during the buckling analysis.
Block Method
In exact arithmetic, the simple Lanczos algorithm can only compute one eigenvector in a
degenerate set. Because of round-off errors introduced into the Lanczos recurrence,
additional eigenvectors of multiple eigenvalues eventually appear in the tridiagonal
matrix T j . A second eigenvector of a multiple eigenvalue only converges a number of steps
123
124
after the first copy converged. (Effectively, this is the case where the starting vector is
orthogonal to the desired eigenvector.) Thus, the simple Lanczos algorithm has difficulties
with eigenvalues of high multiplicity.
Each step of the shifted Lanczos recurrence requires the solution of a sparse linear system
of equations of the form ( K M )x and one multiplication by the matrix M . In
MSC.Nastran these operations require accessing matrices stored on disk files and thus
entail significant I/O costs.
Block Lanczos algorithms have been developed in which the basic Lanczos recurrence is
carried out for p vectors simultaneously. If the idea of a block code is combined with the
shifted and inverted Lanczos algorithm, the following recurrence is obtained for the
vibration problem:
1. Initialization
a. A starting block of p column vectors R 1 is chosen.
1
b. R 1 = ( K M ) MR 1 is computed.
c. An upper triangular p p matrix B 0 and an M -orthonormal n p matrix
Q 1 are computed so that R 1 = Q 1 B 0 .
d. The upper triangular p p matrix B 1 is set to 0 as well as Q 0 = 0 .
2. lteration
a. For j = 1, 2, 3, , the process iterates until convergence as follows:
R j + 1 = ( K M )
T
MQ j Q j A j Q j 1 B j
where A j = Q j M ( K M )
MQ j
b. The following factorization is computed

Rj + 1 = Qj + 1 Bj + 1
where Q j + 1 is an n p matrix with M -orthonormal columns and B j + 1 is
a p p upper triangular matrix.
Using the block version of the Lanczos algorithm provides the benefit that multiple
eigenvalues of the original problem can be computed more easily. All degenerate
eigenvectors can be found together when the multiplicity of an eigenvalue is less than or
equal to the block size p . Further, the amount of I/O per column is reduced by a factor of
1/ p since it is possible to solve and perform the matrix multiplication for p vectors
simultaneously.
The description of the block Lanczos code is incomplete without a description of the
factorization R = QB and consideration of the fact that T j is now block tridiagonal. The
factorization R = QB is required to obtain M -orthonormal vectors in the recurrence. In
this factorization, R and Q are n p matrices and B is an upper triangular p p matrix,
which is chosen so that the columns of Q are M -orthonormal. This procedure is
implemented using a generalization of the modified Gram-Schmidt procedure that avoids
CHAPTER 6
repeated multiplications by matrix M . Both vectors in R and a second set, initially set to
MR , are updated during the Gram-Schmidt orthogonalization. At the completion of the
procedure, the first set of vectors are transformed to Q and the second set to MQ . A
multiplication by M is required at the beginning of the procedure. A second multiplication
is made at the end to ensure that MQ is accurate, but no additional multiplications by M
are required during the procedure.
Another consequence of using a block Lanczos algorithm is that matrix T j , from which the
approximations to the eigenvalues are computed, is now a block tridiagonal matrix
represented as follows:
T
A1 B2
B2 A2 B3
B3 A3 B4
.
.
.
Tj =
Bi Ai Bi + 1
.
.
T
Bj
Bj Aj
Since the values for B j are upper triangular, T j is actually a band matrix with bandwidth
p + 1.
The eigenvalues and eigenvectors of T j are computed by the following procedure. First,
reduce T j to a tridiagonal form T . An orthogonal matrix Q H is found so that
T
QH Tj QH = T
Eq. 6-71
Then perform an eigenextraction for T . An orthogonal matrix Q T is found so that

T
QT T QT =
Eq. 6-72
where is the diagonal matrix of the eigenvalues of T . Then, by combining Eq. 6-72 and
Eq. 6-73, the following is obtained:
T
( Q T Q H )T j ( Q H Q T ) =
Eq. 6-73
where the orthogonal matrix ( Q H Q T ) is the eigenvector matrix for H .

The orthogonal matrices Q H and Q T are products of simple orthogonal matrices (Givens
rotations)
125
126
Q H Q H Q H
1
2
S
or
Q T Q T Q T
1
2
R
These product matrices are accumulated by beginning with the identity matrix I and
successively multiplying on the right by
QH
where: i = 1, , r
.
Note: The eigenvector matrix is formed by products that take linear combinations of the
columns. In the intermediate steps when only the last p rows of the eigenvector
matrix are desired, the leading rows are ignored and the rotations, are applied
only to the last p rows of the identity matrix.
The algorithms used in Eq. 6-72 and Eq. 6-73 are standard and numerically stable.
The actual implementation of the band reduction algorithm of Eq. 6-73 uses modified
(square root free) Givens rotations to reduce the overall cost by eliminating the square root
computations associated with each rotation.
The extraction of the eigenvalues and eigenvectors of T (see Eq. 6-73) is essentially the
same procedure used by the QR algorithm (see QR Method of Eigenvalue Extraction
on page 110). The procedure used here is an explicitly shifted QL algorithm using ordinary
Givens rotations. The square root
2
a +b
required for the calculation of the Givens rotations is computed by a special recursion
2
based on the Pythagorean Theorem. This recursion avoids overflow by not forming a or
2
b and avoids destructive underflow occurring with the implicitly shifted QL algorithm
for the matrices produced by the Lanczos shifted block. The spectral transformation is
applied as before to find approximate eigenvalues for the original problem of Eq. 6-65. If s
is an eigenvector of T j , then y = Q j s with Q j = ( Q 1, Q 2, , Q j ) is an approximate
eigenvector of Eq. 6-65. The residual norms of the eigenpairs are now given by B j + 1 s j
where the values for s j are the last p components of the eigenvector s .
The expense of l/O operations suggests that block size p should be as large as possible. The
available memory precludes choosing very large block sizes. However, large block sizes
also entail other costs because the M -orthogonal factorization requires additional
computation and the overall convergence depends highly on the number of Lanczos steps
and less on the dimension of T j . Therefore, the actual block size is taken as a compromise
among the reduction of l/O costs, the possible increase in arithmetic operations, and the
largest multiplicity of the expected eigenvalues.
CHAPTER 6
Orthogonalization
The Lanczos algorithm produces an orthonormal (or M -orthonormal) set of Lanczos
vectors in exact arithmetic. (For simplicity, the orthogonality referred to always implies
orthogonality with respect to the inner product defined by M ). The numerical algorithm is
affected by round-off errors that cause the Lanczos vectors to lose their orthogonality.
Maintenance of orthogonality is essential for preserving the convergence properties of the
Lanczos algorithm. Early implementations of the Lanczos algorithm employed a full
reorthogonalization scheme in which each new block of Lanczos vectors was explicitly
reorthogonalized against all previous Lanczos vectors. This process required pj 2 vector
operations at step j as well as access to all previous vectors (usually stored out-of-core).
Instead of this expensive scheme, the Lanczos algorithm in MSC.Nastran employs a
combination of several efficient reorthogonalization mechanisms that together accomplish
the computation of a sufficiently orthonormal set of Lanczos vectors.
Loss of orthogonalarity can occur in four different areas:
1. Within a given block of Lanczos vectors (internal).
2. With respect to the previous two blocks of Lanczos vectors (local).
3. With respect to the whole set of previously computed Lanczos vectors (global).
4. With respect to eigenvectors from different shifts (external).
Problems with orthogonality within a block of Lanczos vectors can arise if the vectors of
R j + 1 are almost linearly dependent. For example, this problem occurs when the shift 0 is
extremely close to an eigenvalue. In this case, one step of the generalized modified GramSchmidt procedure is not sufficient to produce vectors that are orthogonal to working
precision. Gram-Schmidt is considered an iterative procedure that possibly can be repeated
several times until the Lanczos vectors are orthogonal. The Gram-Schmidt procedure is
referred to as internal reorthogonalization, which also requires updating the elements of
Bj + 1 .
The local loss of orthogonality involves the previous two blocks of Lanczos vectors. The
recurrence can be considered an orthogonalization of the block R j + 1 against the blocks
Q j and Q j 1 . The block R j + 1 , computed from the Lanczos recurrence, may not be
orthogonal to full working precision to Q j and Q j 1 . Investigations indicate that the
orthogonality between R j + 1 and Q j is crucial for the correct continuation of the Lanczos
process. Therefore, one step of local reorthogonalization is carried out; i.e., the block R j + 1
is reorthogonalized against the block Q j . This procedure may be repeated until the two
blocks are orthogonal to working precision. This local reorthogonalization also requires
updating the elements of A j .
The global loss of orthogonality between the block Q j + 1 and previous Lanczos blocks is
of a different nature. The Lanczos recurrence involves only three blocks of Lanczos vectors.
The beauty and efficiency of this recurrence lies in the fact that the three-term recurrence
in exact arithmetic is enough to assure global orthogonality among all of the Lanczos
vectors. Unfortunately, this is no longer true under the influence of round-off errors. Once
some tiny error is introduced into the recurrence, it becomes amplified in the course of the
next Lanczos steps and soon the global orthogonality property is lost. The mechanisms of
127
128
this loss of orthogonality have been investigated in the last decade by several researchers
and are now well understood. There are two important insights from the theoretical works
that provide for an efficient implementation of the global reorthogonalization. First, it is
possible to monitor the loss of orthogonalization inexpensively by updating certain
estimates for loss of orthogonality at every Lanczos step. Second, it is sufficient to maintain
orthogonality at the level of the square root of the machines precision (semi-orthogonality)
and still obtain fully accurate eigenvalues. These two observations give rise to the scheme
of partial reorthogonalization that is generalized to the block Lanczos algorithm in the
MSC.Nastran implementation.
The fourth type of loss of orthogonality can only occur in the context of the shifted and
inverted algorithm and has nothing to do with the Lanczos recurrence directly. The
MSC.Nastran Lanczos algorithm begins by computing some eigenvalues and eigenvectors
with an initial shift 1 . If not all of the desired eigenvalues are found, then a second
Lanczos run with a new shift 2 is made. For reasons of efficiency and simplicity in
bookkeeping, the previously computed eigenvalues are prevented from being
recomputed. This benefit is achieved by the external selective orthogonalization
implemented in MSC.Nastran. For each new shift, a set of critical converged eigenvalues is
determined. The Lanczos vectors are then kept orthogonal against the corresponding
eigenvectors of this selective group of computed eigenvalues using a modification of the
technique of selective orthogonalization. Estimates for the loss of orthogonality with
respect to the computed eigenvectors are updated at each Lanczos step, and
reorthogonalizations are performed only when semi-orthogonality is about to be lost.
Shift Strategy
The eigenanalysis problem in MSC.Nastran is to compute either the lowest m eigenvalues
in magnitude or all eigenvalues in an interval [ a, b ] where a and b can be any real
numbers, such as a < b . The shift strategy incorporated in the block shifted Lanczos code
allows the shifts selected in the interval [ a, b ] to rapidly find the desired eigenvalues
without wasted factorizations. The main objective for the shift strategy is to encompass the
region containing the eigenvalues of interest with a trust region in which all the
eigenvalues have been computed and the number of eigenvalues has been verified with a
Sturm sequence check.
The first shift is chosen at the primary point of interest (usually 0 in vibration or -1 in
buckling) or the left endpoint if it is finite. Additional shifts are chosen to form or extend a
trust region until the requested eigenvalues are computed. These shifts are selected based
on the information computed during previous Lanczos runs. A trust region can be
extended in either direction.
The shift strategy is designed to place the final shift at the correct place to simultaneously
compute the last required eigenvalues and to make a Sturm sequence check for the entire
interval. However, a final shift may be required for a separate final Sturm sequence check.
This check is designed conservatively so that shifts are not taken so far as to skip past part
of the spectrum. However, if this does happen, the Sturm sequence count indicates that not
all the eigenvalues are computed between two shift points. The shift strategy immediately
attempts to find the missing eigenvalues rather than to extend the trust region further.
CHAPTER 6
Two values, called the left and right sentinels, are associated with all shifts. A right sentinel
is defined as the rightmost accepted eigenvalue that has no unaccepted eigenvalue
approximations between the shift and itself. The left sentinel is defined similarly for the
accepted and unaccepted approximations to the left of the shift. When eigenvalues are
missing between two shifts, it is usually because the last shift was too far from the previous
shift. In this case, the missing eigenvalues should be between the right sentinel of the
leftmost of the two shifts and the left sentinel of the rightmost of the two shifts. A new shift
is selected to bisect the interval between the above two sentinels. In the rare case when the
sentinels overlap each other, the new shift bisects the full interval between the two
previous shifts. (The usual reason that such an overlap occurs is that the multiplicity of
eigenvalues is greater than the Lanczos block size.)
There are several ways in which the shifting deviates from the description above. Four
major special cases are listed in this section. The first special case is the setting of a
number . Initially, is an approximation to the first nonrigid body mode of the structure.
In the case of an initial shift at 0 for a structure with rigid body modes, only the rigid body
modes in the initial Lanczos run are computed. In such cases the shift strategy does not
update the value of , but instead uses the initial to move away from 0 towards the first
nonzero eigenvalue. There are other special cases related to the rigid body modes where
the Lanczos algorithm may terminate with no new information available for the next shift
selection. In these cases is not updated, and the previous value is maintained.
The second special case is when no new information is computed at two consecutive shifts.
This case may occur if there is a very large gap between modes of a structure or all
remaining modes are infinite. The shift strategy expands its step size to cross the suspected
hole in the spectrum. If the user-specified interval [ a, b ] has a finite endpoint that was not
used previously, the shift strategy selects the new shift at that endpoint. If the interval has
only infinite endpoints remaining, then the new shift is chosen at 10 past the last shift. If
no new information is computed at this shift, the next shift is chosen at 100 past the last
shift. If there is still no new information, then the Lanczos procedure terminates with the
assumption that the remaining eigenvalues are infinite. An appropriate warning is
returned with those eigenvalues that were computed.
The third special case is the buckling problem. The operator used in the buckling problem
for the Lanczos iteration is ill-posed for shifts at or near 0. The shift strategy process of the
buckling problem is similar to the vibration problem with the exception that the shifts near
0 are perturbed from 0.
The fourth major special case is the processing of factorization errors. In rare cases when
the selected shift is exactly an eigenvalue, the operator K M is singular, and the
factorization can fail. If this occurs, the shift is perturbed, and an additional factorization is
performed. If three factorization errors occur consecutively, then the Lanczos procedure
terminates with an appropriate error message and returns any computed eigenvalues and
eigenvectors. Presumably, the cause of such a failure is an improperly posed eigenvalue
problem for which the mass and stiffness matrices have a common null space (massless
mechanisms).
129
130
Summary of Procedure
The general Lanczos procedure implemented in MSC.Nastran can be summarized in the
following figures. There are two major levels in the procedure: the outer level in the shift
strategy and administration, and the inner level in the actual Lanczos iteration. Figure 6-1
describes the outer level and Figure 6-2 describes the inner level.
Initialization
Select a Shift
Value
Decompose Shifted
Matrix
Run Inner Level

Flowchart 2
No
Yes
Use the
Same
Shift?
Administration
Accepted Eigenpairs
Prepare Output
Figure 6-1
Yes
Done
No
Outer Level of the Lanczos Procedure
CHAPTER 6
Initialization of Inner Level
Lanczos Loop:
for
j = 1, 2, , maxstp do
Lanczos Step:
Compute
a j, q j + 1, j + 1
Analyze
Tj :
Compute the Spectrum of
Tj
Did Sufficiently
Many Eigenvalues
Converge?
No
Yes
Postprocessing of Eigenvectors
Exit Lanczos Loop
Compute Eigenvectors of A
End of Lanczos Run

(Return to Outer Level)
Figure 6-2
Inner Level of the Lanczos Procedure
Segmented Lanczos Method

The segmented version of Lanczos is intended to alleviate the orthogonality problems
encountered on very wide frequency range runs. The orthogonality may be lost when very
long Lanczos runs are executed while trying to span a wide frequency range of interest.
The segment version forces intermittent shifts at semi-regular intervals resulting in more
frequent restarts of the Lanczos process. While it has significantly improved the solution
quality, i.e., we avoid aborts due to loss of orthogonality, the number of shifts is usually
more than the non-segmented run.
131
132
Frequency Domain Decomposition-Based Distributed Parallel Lanczos

Method
The frequency domain decomposition-based normal modes analysis is built on the
frequency segment approach of the Lanczos method. The frequency range of interest given
by the user is subdivided automatically. The user also has the choice of introducing
intermediate frequencies directly on the EIGRL continuation card. This may be especially
advantageous in case of repeated runs (the most practical situation), where the user may
be able to enforce better load balancing than the automatic intermediate frequency
generation method. This process is executed in a master/slave paradigm. One (the master)
of the processors will execute the coordination and collection of the separately calculated
eigenvectors into an appropriately ordered set. This guarantees the continuation and
proper exit of the master process. The slave processes will contain only the eigenvalues and
eigenvectors they found upon exiting the READ module.
Geometric Domain Decomposition-Based Distributed Parallel Lanczos

The principle of geometric domain decomposition is not confined completely to the READ
module as the frequency domain decomposition. The geometric domain decomposition
principle transcends many DMAP modules prior to the READ module and the
eigenproblem presented to READ is only a subset of the original problem.
It is important to emphasize that this version still solves the global (see Eq. 6-1) eigenvalue
problem as "exactly" as the single processor run. The significant difference is that the
geometry and therefore the global matrices are partitioned and the local processors see
only the local portions. In the current implementation, we restrict ourselves to as many
subdomains as we have processors. One might call this a "distributed global solution
technique."
In general, there are two alternatives to execute this process. One is to keep only short, local
n oj , n aj size vectors and another is to also add a global n a size vector to the local Lanczos
processes. The first method, while minimal in storage requirements, requires interprocess
communication even when executing vector operations such as the Lanczos step. The latter
technique requires communications only when matrix operations are executed, while it has
n a word redundancy on each processor, so this seems more attractive, and is used in our
implementation.
In the geometry domain parallel Lanczos method, the stiffness and mass matrices are
partitioned into submatrices based on the geometry subdomains as follows:
1
K oo
1
K oa
2
K oo
K =
2
K oa
.
j
K oo
K oa
.
1 K2 .
K ao
ao
j
K ao
. K aa
Eq. 6-74
CHAPTER 6
1
M oo
1
M oa
2
M oo
2
M oa
M =
.
j
M oo
.
j
M ao
1 M2 .
M ao
ao
Eq. 6-75
M oa
.
. M aa
where the superscript j refers to the j-th subdomain, subscript a to the common boundary
of the subdomains, and s is the number of the subdomains, so j = 1, 2, , s . The size of
these global matrices as well as the eigenvectors is N. For the presentation of the algorithm,
let us partition the Lanczos vectors accordingly:
x o1
x o2
.
x =
Eq. 6-76
j
xo
.
xa
Furthermore, the boundary portion may be partitioned as:
x a1
x a2
.
xa =
Eq. 6-77
xa
.
x as
i.e., the a set is the global boundary set; s is the number of subdomains. Finally, the j-th
processor will receive the j-th subdomain information of:
j
K oo K oa
,
j
j
K ao K aa
M oo M oa
,
j
j
M ao M aa
xo
j
xa
Eq. 6-78
where the submatrices are partitioned into local o and a-sets and the appropriate sizes are n o
j
and n a .
The main computational elements of the Lanczos method are executed in the distributed
form as follows.
133
134
Simultaneous Shifted Matrix Generation. Since the inputs to the READ module are the
j
j
K and M matrices, the task of creating the local system matrix will be simultaneously
executed on each of the processors serially:
A oo A oa
A ao A aa
j
j
j
j
K oo K oa
M oo M oa
=
0
j
j
Kj Kj
M
M
ao
aa
ao aa
Eq. 6-79
The A matrix is the distributed submatrix on the local memory of the j-th processor (node).
Naturally, the local mass matrix component will also be saved locally and, since it is
needed in each Lanczos step, it will be stored in local memory if possible. The shifted
stiffness matrix will be stored on the local scratch disk of the node.
Important: The shift is calculated from local information such as matrix norms and
0
runtime measures; therefore, it is processor dependent. Hence, some
communication between the nodes is required to assure that the shift is
uniform across the processors.
Distributed Factorization. The parallel, distributed implementation will execute the
following steps:
1. The partial decomposition will formally decompose this j-th subdomain as
follows:
A oo A oa
j
A ao
j
A aa
L oo 0
j
L ao
D oo
j
A aa
Tj
Tj
L oo L oa
0
I
Eq. 6-80
where the identity matrices are not computed; they are only presented to make
j
the matrix equation algebraically correct and the A aa submatrix is the boundary
submatrix of the j-th partition updated by the factorization contributions from the
interior as:
j
T, j
A aa = A aa L ao D oo L oo
Eq. 6-81
2. The boundary submatrices are summed up as:

s
A aa =
Eq. 6-82
A aa
j = 1
3. The boundary is decomposed as:

T
A aa = L aa D aa L aa
Eq. 6-83
The distributed decomposition step will be executed outside the Lanczos

recurrence logic and the resulting:
CHAPTER 6
j
LL ao
L oo
Eq. 6-84
L ao
partial factor and the global boundary L aa factor will be utilized inside the
recurrence.
Since one important function of the decomposition operation in the Lanczos
framework is to establish the Sturm count ( S ), another interprocessor
communication is needed:
s
j
Sg = Sa +
So
j = 1
where the Sturm counts in the equation above are the global, boundary, and local
interior Sturm counts in that order.
j
Distributed Matrix Multiply. This is the first operation where the fact of having n o + n a
j
j
+ n a long Lanczos vectors in all local processes, but executing a local n o + n a size matrix
operation requires extra care. The basic local matrix-vector multiply operation is:
yo
ya
M oo M oa
j
xo
M ao M aa
j j
= Mx
Eq. 6-85
xa
where the local vector partitions are defined as:

j
xo
x =
x
xa
xa
Eq. 6-86
xa
and
j
yo
y =
y
ya
ya
Eq. 6-87
ya
Note that the x a, y a boundary partitions are the local subsets of the x a, y a . The operation
may be executed in the following steps:
1. Execute y
j
j j
= Mx
2. Scatter y a to y a
3. Send y a to master processor
135
136
4. Master sums up y a
5. Receive y a from master processor
j
6. Gather y a from y a
Operation 1. will be executed by the existing architecture, which consists of two phases. In
subroutine phase 1 we save the M matrix in a sparse form, storing the indices and
numerical terms separately. This in turn is read in and interpreted by the 2nd phase that
executes the actual numerical operation by calling the XPI2R* indexed, double SAXPY
kernel. This phase does not need to understand the local versus boundary structure of the
M matrix.
In addition, the existing option of keeping a portion or all of the mass matrix in memory is
also operational in the distributed environment. The operations (3. through 5.) execute the
communication and summing up of the boundary segments between the processors
sharing the boundary. Finally, the scattering and gathering of the local boundary segments
(2., 6.) is straightforward.
Upon completing this step, all processors have a shared and complete ya vector, identical
in content. This is necessary to proceed with the local Lanczos process.
Distributed F-B Substitution. This phase will contain the following elements:
T, j
T, j
T, j
1. The LL ao = [ LL oo LL ao ] factorization of the interior of the j-th subdomain is

independent of the boundary allowing a forward substitution on the interior as:
T, j
L oo z o + L oo x a = [ L oo D oo ]
yo = zo
Eq. 6-88
It is important to note that in this step the overlapping boundary regions are
zeroed out on all subprocesses, except for one, which will do the update.
2. The global boundary solution is a complete forward-backward substitution of:
s
T
L aa D aa L aa
za =
( y a L ao D oo z o )
Eq. 6-89
j = 1
3. Finally, the interior solution is a backward only substitution:

T, j
z o = L oo
T, j
[ z o L ao z a ]
Eq. 6-90
The local z vector is again partitioned as:

j
zo
z =
z
za
za
Eq. 6-91
za
therefore, a scattering and gathering operation described in steps 2. and 6. of

Distributed Matrix Multiply on page 135
CHAPTER 6
j j
Simultaneous Lanczos Step. Since at this stage all processors have their local y k = M x k
j
j j
and z k = A x k vectors ( k refers to the Lanczos step number) as well as the last two Lanczos
j j
vectors x k, x k 1 , the Lanczos step will be executed simultaneously as follows:
1. Execute local inner product:
T, j j
k = yk zk
Eq. 6-92
2. Create global inner product via communication:

s
k =
Eq. 6-93
j = 1
3. Execute local saxpy:

j
xk + 1 = zk k xk k 1 xk 1
Eq. 6-94
In order to normalize the next Lanczos vector, another matrix vector

multiplication is needed inside of the Lanczos step (using the same method
described in Distributed Matrix Multiply on page 135.
j
yk + 1 = Mj xk + 1
Eq. 6-95
The calculation of the normalization parameter follows steps 1. through 3. above.

T, j
k = ( xk + 1 yk + 1 )
12
and
Eq. 6-96
s
k =
j 12
( k )
j = 1
This leads to the next normalized Lanczos vector

j
xk + 1 = xk + 1 k
Eq. 6-97
and the step is completed.

Distributed Orthogonalization Scheme. While this operation does not explicitly
involve matrix operations, its performance impact is such that we have a distributed
implementation. The main orthogonalization tasks are: against the last two Lanczos blocks;
against earlier found eigenvectors at the initial phase and during the iteration, respectively;
and, finally, within the current block.
The efficiency of the distributed orthogonalization scheme is coming from the fact that we
can distribute O ( k n ) (vector) operations but need to communicate only O ( k ) (scalar)
data.
The distributed orthogonalization is executed in the following steps:
137
138
1. Calculate local inner products.

2. Exchange and sum up local inner products.
3. Execute local saxpy step.
In step 1., each processor is to calculate:
T, j
i = xk + 1 M xi
Eq. 6-98
where i is the set of selected Lanczos vectors. In step 2., the following global sum is needed
on each processor:
s
i =
Eq. 6-99
j = 1
j
Finally, step 3. is executed in two segments, x a, x o , as

j
x o, k + 1 = x o, k + 1 i x o , i
Eq. 6-100
and
x a, k + 1 = x a, k + 1 i x a, i
Eq. 6-101
followed by a gathering into the x a, k + 1 array from x a, k + 1

After this step, the Lanczos process may be continued again from the Distributed Matrix
Multiply on page 135. The process continues on all the nodes until a certain number of
steps are executed or convergence achieved, which decision again needs communication..
Simultaneous Tridiagonal Solution Generation. Since all processors have the Lanczos
blocks, and hence the same block tridiagonal matrix, the solution of the tridiagonal
problem of Eq. 6-6 will be executed on all nodes simultaneously. Once the tridiagonal
solution has been completed, selected will be accepted as eigenvalues of the original
problem and the corresponding u k will become the basis to compute eigenvectors. Note
that the u k vectors are the same ( k ) size on all the nodes. Since all the nodes have
j
j
n o + n a + n a length Lanczos vectors x i, i = 1, 2, , k , the eigenvector computation of the
local segments will not require additional communication.
The decision to stop the Lanczos run is based on analyzing the T k matrix and cost analysis.
Since the terms of this matrix are identical on all nodes, the stop must be simultaneous, but
is synchronized for safety. If more eigenvectors are needed, another Lanczos run may be
executed or ultimately the complete process starting from Simultaneous Shifted Matrix
Generation on page 134 may be repeated with a different 0 shift value. If all required
eigenvectors are found, the final orthogonality test will be executed on the local matrices
and vectors and the READ module will complete by setting ND to the number of accepted
eigenvectors. Note that upon exiting the READ modules, all the processors have a
complete LAMA table; however, the PHIA matrix contains only the local rows of the
eigenvectors.
CHAPTER 6
The data recovery for each domain is done independently on each processor on the local
segments of the eigenvectors provided by READ. As this does not require any
communication, it helps in the overall speedup and disk-space savings during that part of
the run. Following data recovery, the output may or may not be merged based on the
mergeofp keyword.
139
140
6.3

The real eigenvalue solution methods are categorized as shown in Table 6-2:
Table 6-2 Real Eigenvalue Methods
Method
Type
Identifier
Givens
Reduction
GIV
All Roots
M Positive Definite
Householder
Reduction
HOU
All Roots
M Positive Definite
Modified Reduction
Reduction
M HOU, A HOU
All Roots
[ K ] + s [ M ]
Lanczos
Iteration
GIV
GIV
LANC
Application
Restriction
Not Singular
Small
Number of
Roots
[ K ] + s [ M ]
Not Singular
M Positive
Semidefinite
K Symmetric
CHAPTER 6
6.4
DMAP User Interface

READ
KAA,MAA,MR,DAR,DYNAMIC,USET,CASECC,
PARTVEC,SIL,VACOMP,INVEC,LLL,EQEXIN,GAPAR/
LAMA,PHIA,MI,OEIGS,LAMMAT,OUTVEC/
FORMAT/S,N,NEIGV/NSKIP/FLUID/SETNAME/SID/METH/
F1/F2/NE/ND/MSGLVL/MAXSET/SHFSCL/NORM/PRTSUM/MAXRATIO $
Input Data Blocks:

KAA
Stiffness matrix.
MAA
Mass matrix.
MR
Rigid body mass matrix
DAR
Rigid body transformation matrix.
DYNAMIC Eigenvalue Extraction Data (output by IFP module).

USET
Degree-of-freedom set membership table.
CASECC
Case Control Data Table (selects EIGR, EIGRL, or EIGB entries, output by
IFP module).
PARTVEC
Partitioning vector with values of 1.0 at the rows corresponding to degrees

of freedom which were eliminated in the partition to obtain KAA and MAA.
Required for maximum efficiency. See SETNAME parameter description
below.
SIL
Scalar index list. Required for maximum efficiency.
VACOMP
Partitioning vector of size of a-set with a value of 1.0 at the rows

corresponding to r-set degrees-of-freedom. The USET table may be specified
here as well. If VACOMP is purged and DAR does not have the same
number of rows as KAA, then the partitioning vector will be determined
from the size of MR.
INVEC
Starting vector(s) for Lanczos method only or EQMAP data blocks for
geometry domain parallel.
LLL
Lower triangular factor from decomposition of KAA. Use to enhance shift

logic for buckling eigenvalue extraction or VF01: interior boundary
partitioning vector for geometry domain parallel Lanczos method.
EQEXIN
Equivalence between external and internal grid identification numbers.

Required for maximum efficiency.
GAPAR
Local-global boundary partitioning vector for geometry domain parallel

Lanczos method.
141
142
Output Data Blocks:

LAMA
Normal modes eigenvalue summary table.
PHIA
Normal modes eigenvector matrix in the a-set.
OEIGS
Real eigenvalue extraction summary.
MI
Modal mass matrix.
LAMMAT
Diagonal matrix containing eigenvalues on the diagonal (Lanczos and

QLHOU only).
OUTVEC
Last vector block (Lanczos only).
Parameters:
FORMAT
Input-Character-no default. If FORMAT MODES, READ will solve a

d
buckling problem of ( [ K ] + [ K ] ) .
NEIGV
Output-integer-no default. NEIGV is the number of eigenvectors found. If

none were found, NEIGV = 0. If m modes were found (but error
encountered), NEIGV = -m. If m modes were found, NEIGV = m.
NSKIP
Input-integer-default=1. The method used by READ is taken from the

NSKIP record of CASECC.
FLUID
Input-logical-default=FALSE. If FLUID = TRUE, then the EIGRL or EIGR

entry is selected from METHOD(FLUID) Case Control command.
SETNAME
Input-character-default='A'. For maximum efficiency, the rows and

columns KAA and MAA must correspond to or be a partition of the
displacement set specified by SETNAME. If KAA and MAA are a partition
then PARTVEC must also be specified.
SID
Input-integer-default=0. Alternate set identification number.

If SID=0, the set identification number is obtained from the METHOD
command in CASECC and used to select the EIGR or EIGRL entries in
DYNAMIC.
If SID>0, then METHOD command is ignored and the EIGR or EIGRL is
selected by this parameter value. All subsequent parameter values (METH,
F1, etc.) are ignored.
If SID<0, then both the METHOD command and all EIGR or EIGRL entries
are ignored and the subsequent parameter values (METH, F1, etc.) will be
used to control the eigenvalue extraction.
METH
Input-character-default='LAN'. If SID<0, then METH specifies the method

of eigenvalue extraction.
LAN
Lanczos
GIV
Givens
MGIV
Modified Givens
CHAPTER 6
HOU
Householder
MHOU
Modified Householder
AGIV
Automatic selection of GIV or MGIV
AHOU
Automatic selection of HOU or MHOU
F1
Input-real-default=0.0. The lower frequency bound.
F2
Input-real-default=0.0. The upper frequency bound. The default value of

0.0 implies machine infinity.
NE
Input-integer-default=0. The number of estimated eigenvalues for nonLanczos methods only. For the Lanczos method, NE is the problem size
(default=20) below which the QL Householder option is used if it is
enabled.
ND
Input-integer-default=0. The number of desired eigenvalues.
MSGLVL
Input-integer-default=0. The level of diagnostic output for the Lanczos

method only.
0
no output
warning and fatal messages
summary output
detailed output on cost and convergence
detailed output on orthogonalization
MAXSET
Input-integer-default=0. Vector block size for Lanczos method only.
SHFSCL
Input-real-default=0.0. Estimate of the first flexible natural frequency.

SHFSCL must be greater than 0.0. For Lanczos method only.
NORM
Input-character-default='MASS'. Method for normalizing eigenvectors. See

Option Selection on page 145 for details.
PRTSUM
Input-logical-default=TRUE. Lanczos eigenvalue summary print flag. See

Performance Diagnostics on page 93 for details.
MAXRATIO
Input-real-default= 10 5 . May be overwritten in the DMAP by: param,

maxratio, value.
143
144
6.5
Method Selection
EIGR Entry. The method selection of any method may be performed with the EIGR Bulk
Data entry using the following format:
EIGR
SID
METHOD
F1
NORM
F2
NE
ND
The SID is the set ID number corresponding to a METHOD command in the Case Control
Section. METHOD should be equal to any of the identifiers given in Solution Method
Characteristics on page 140. F1, F2 are frequency minimum and maximum values
specifying the boundaries of the users frequency range of interest. NE and ND are the
number of roots estimated and desired to be found, respectively. On the continuation
entry, the user can choose some normalization options, which are detailed in Option
Selection on page 145.
EIGRL Entry. To select the Lanczos method in details, the user should use the EIGRL Bulk
Data entry with the following format:
EIGRL
SID
F1
F2
ND
MSGLVL
MAXSET
SHFSCL
NORM
ALPH
NUMS
f1
f2
f3
f4
f5
f6
f7
f8
f9
f10
f11
f12
f13
f14
f15
The MSGLVL entry (0 through 3, default = 0) controls the amount of diagnostics output.
MAXSET specifies the maximum number of vectors in a block of the Lanczos iteration. It
is also equivalent to or may be overridden by the value of SYSTEM cell 263. The value of
SHFSCL is an estimate for the location of the first nonzero eigenvalue of the problem.
The following parameters are only used if the F1 and F2 frequency range is to be broken up
into segments. ALPH is the constant defining the modal distributions function
(Frequency Segment Option on page 146). Its default value is 1.0, which results in an
even distribution of segments. NUMS is the number of segments in the frequency range
(default = 1). f1 to f15 are segment boundaries such that F1< f1 < f2 ... < f15 < F2. f1 to f15
if not specified will be computed based on a distribution given by ALPH.
Different combinations of F1, F2, and ND specify different options in the Lanczos module
(see Frequency and Mode Options on page 145).
The main purpose of the SHFSCL is to aid the automatic shift logic in finding the
eigenvalues especially in crossing the (possibly big) gap between the computationally zero
(rigid body) modes and the finite (flexible) modes. Another use of SHFSCL is to create a
cutoff frequency for the so-called supported modes. If a mode is above SHFSCL, it is not
supported even if it is flagged by a SUPORT entry.
The NORMalization parameter for Lanczos is described in Normalization Options on
page 145.
CHAPTER 6
6.6
Option Selection
Real symmetric eigenvalue analysis offers several normalization, frequency and mode, and
performance options.
Normalization Options
The methods accessed by the EIGR entry have several normalization options. They are:
NORM = MASS
Mass normalization of eigenvectors (normalize to unit value of

generalized mass).
NORM = MAX
Maximum normalization of eigenvectors (maximum component

of vectors is unity).
NORM = POINT
A selected point normalized to unity. The point is specified by G

(grid point number) and C (components 1 through 6).
The Lanczos method (using EIGRL) has MASS or MAX normalization capability only.
The following options are valid for the Lanczos method only unless otherwise stated.
Frequency and Mode Options

At present, based on the F1, F2, and ND combinations, the following options are supported
in the Lanczos algorithm:
Table 6-3
F1
F2
ND
Option
Given
Given
Given
Lowest ND or all in range
Given
Given
Blank
All in range
Given
Blank
Given
Lowest ND in range (F1, + )
Given
Blank
Blank
Lowest one in range (F1, + )
Blank
Blank
Given
Lowest ND in (- , + )
Blank
Blank
Blank
Lowest one in (- , + )
Blank
Given
Given
Lowest ND below F2
Blank
Given
Blank
All below F2
Performance Options
Space Saver. Another option available in Lanczos method is called the space saver
option. This option is selected by setting SYSTEM (229) = 1 (default = 0) and results in
significant reduction in scratch space usage by not preserving factor matrices for later use
in the case both F1, F2 are given. However, CPU-time may increase.
145
146
Sparse Solver in Lanczos. For faster CPU execution, the Lanczos method by default
executes sparse matrix operations. The memory available for the sparse solver inside the
Lanczos module can be controlled by setting SYSTEM (146) to greater than 1. The larger the
number the larger the area reserved for the factor. Recommended values are 2, 3, 4.
SYSTEM(146) is equivalent to the FBSMEM keyword. This increased memory space is
taken from the space available for the eigenvectors; consequently, the user must find a
satisfactory compromise.
Low Level Parallel Lanczos. Low level shared memory parallel execution is supported
in the shift (decomposition) and solve steps of the Lanczos method. These methods can be
specifically selected by setting the keyword SMPARALLEL = ncpu and deselected by
setting SMPARALLEL = 7168 + ncpu, where ncpu = the number of CPUs.
I/O Reduction Options. Other performance-enhancing options which control the I/O to
CPU time ratio are described in Table 6-4:
Table 6-4 I/O Reduction Options
System
(199) =
Performance Option
l
Set memory for incore mass matrix multiply to 2 l BUFFSIZE

(default: l = 1)
0 Automatically fits the mass matrix in core if sufficient memory is

available
(193) =
0 Save
result of mass matrix multiply (default = 0)
1 Do not save
Accuracy Options. The user may reset the Lanczos iteration tolerance by:
SYSTEM(123) = k
where k is the exponent of the Lanczos tolerance criterion if it is negative, or it is the
exponent of the maximum matrix factor diagonal ratio tolerated by Lanczos if it is positive.
It is also possible to reset the pivot criterion for the decomposition inside the READ module
by SYSTEM(89) = -k, resulting in 10 k used.
Frequency Segment Option. The frequency segment option is controlled as follows.
The number of frequency segments may be specified on the EIGRL entry (NUMS) or on the
NASTRAN entry (NUMSEG). In the case both are given, NUMS is set to NUMSEG. The
intermediate frequencies may be directly given (f1 ... f15) by the user on the EIGRL entry.
It is also possible to specify ALPH on the EIGRL entry and, in this case, the intermediate
frequencies are automatically calculated by the formula shown in Table 6-5.
CHAPTER 6
Table 6-5 Frequency Segment Specification

PARAMETERS
Definition
NUMS
Number of frequency spectrum subsegments
ALPH
ALPH i
1 ----------------
100
Subsegment boundary: f i = ( f max f min ) -----------------------------------------------ALPH NUMS
1 ----------------
100
Miscellaneous Options
In the READ module, a new shift logic was introduced for the Lanczos method. If you wish
to use the old method (delivered via the REIGL module), SYSTEM(273) must be set to a
non-zero value. The default is zero.
Incompatible DAR and KAA sizes: If the DAR matrix has less rows than KAA, an
approximate size identity matrix is generated inside READ and merged into DAR to
produce a DAA matrix. This assures that the rigid body modes included will have the same
size as the flexible modes computed. This operation requires the presence of the VACOMP
data block. If VACOMP is purged, a fatal error exit is produced.
Parallel Options
Frequency domain:
dmp = n on submittal
domain solver fdmodes (numdom = n) in executive deck
The numdom value in fdmodes is equivalent to SYSTEM(197) (also known as NUMSEG).

Geometry domain:
dmp = n on submittal
domain solver gdmodes (numdom = n) in executive deck
The numdom value in gdmodes is equivalent to SYSTEM(349).

The default for both numdom values is SYSTEM(231) (also known as dmp(arallel)).
Mass Matrix Analysis Options

Indefinite Test
The mass matrix MAA is examined for all real vibration (non-buckling) eigenvalue
problems. First and foremost, MAA is checked to determine if it is indefinite. Under no
circumstances will an eigenvalue analysis proceed with an indefinite MAA. If MAA is
indefinite, UFM 4646 is printed, which includes a User Action suggesting the use of
SYSTEM(304) to perturb the mass matrix, which may render it positive semi-definite. If
SYSTEM(304) is activated, a small value is added to the diagonal of MAA, prior to the
indefinite test. Then, if the indefinite test passes, the perturbed MAA is used in the
subsequent eigenvalue analysis regardless of extraction method.
147
148
Rank Test
A rank test is performed if the indefinite test is selected. The rank of MAA (NRANK) is
determined in order to detect the presence of infinite roots. The number of eigenvectors
requested is restricted to min (N,NRANK) for the tridiagonal methods (HOU,GIV). The
rank test is controlled by SYSTEM(313), which defines a maximum ratio of MAA diagonals
to MAA-factor diagonals.
Density Control
None of the above tests is performed if it is estimated that the tests themselves would
require too much time. A test is made to determine if the density of MAA exceeds some
threshold (default 10%). If MAA exceeds the threshold, it is deemed "dense"; therefore, its
factorization might be nontrivial, and these tests are bypassed. The threshold is controlled
by SYSTEM(314). If N (the size problem) is less than or equal to 100, the density check is
bypassed.
These analyses are summarized in Table 6-6:
Table 6-6 Mass Matrix Analyses
System
Cell
Comment
Description
Default Value
303
Indefinite Test
=0 cutoff=1.0e-6
<0 cutoff=10**sys303
> bypass test
304
M Perturbation
=0 perturb=1.0e-6
<0 perturb=10**sys304
>0 bypass perturbation
313
Rank Test
=0 rank ratio=1.0e+7
>0 rank ratio=10**sys304
<0 bypass test
0
(Do the Test)
314
Density
Threshold
=0 thresh=0.01
>0 thresh=sys314/10000.
<0 do not check density
0
(10% Threshold)
0
(Do the Test)
+1
(Do not Perturb)
QL Householder Option
If sufficient memory is available, a modern (QL iteration based) version of Householder
methods (AHOU, MHOU, or HOU) is automatically used. This is also used in place of
Lanczos when the problem size is smaller than the value of the NE parameter (default =
20). Setting SYSTEM(359) = 0 will turn off the QL Householder option (default
SYSTEM(359)=1).
CHAPTER 6
6.7
Real Symmetric Eigenvalue Diagnostics

The diagnostics of the eigenvalue methods can be categorized as execution diagnostics,
numerical diagnostics, and error messages.
Execution Diagnostics
A certain level of execution diagnostics of the READ module is requested by DIAG 16. For
the Lanczos method, MSGLVL = 1, 2, or 3 in the EIGRL entry gives different levels of
execution diagnostics. These diagnostics pages are elaborate and are described in Lanczos
Diagnostics on page 152.
The following two tables are printed only when PRTSUM = TRUE (default)
Table of Shifts
The table of shifts shows the sequence of shifts taken, the Sturm counts, and the number of
modes computed at each shift for the Lanczos method. It appears as follows:
TABLE OF SHIFTS (REIGL)
SHIFT #
SHIFT
VALUE
FREQUENCY,
CYCLES
# EIGENVALUES
BELOW
# NEW EIGENVALUES
FOUND
X1
X2
X3
X4
X5
X1
The shift number
X2
The shift value in eigenvalue units
X3
The shift value in frequency units (typically Hertz)
X4
The number of modes below this shift (the Sturm count)

If X4 is FACTOR ERROR then this shift is rejected because the MAXRATIO
is too large. To override this, the user may set SYSTEM(166) = 4 and the factor
will be used despite the high MAXRATIO.
X5
The number of modes found by the algorithm at this shift
Execution Summary. The execution summary table for the Lanczos method is as follows:
EIGENVALUE ANALYSIS SUMMARY (REAL LANCZOS METHOD)
BLOCK SIZE USED
NUMBER OF DECOMPOSITIONS
NUMBER OF ROOTS FOUND
NUMBER OF SOLVES REQUIRED
TERMINATION MESSAGE:
TEXT
The TEXT of the termination message can be any of the following:
149
150
REQUIRED NUMBER OF EIGENVALUES FOUND

ALL EIGENVALUES FOUND IN RANGE
NOT ALL EIGENVALUES FOUND IN RANGE
INSUFFICIENT TIME TO FIND MORE EIGENVALUES
The run may also finish with the following message:
USER FATAL MESSAGE 5405, ERROR X OCCURRED DURING ITERATION

where X can be equal to
- 31: INSUFFICIENT WORKSPACE

- 32: QL ALGORITHM DID NOT CONVERGE
- 33: MORE EIGENVALUES FOUND THAN EXIST
- 34: FILE I/O ERROR
- 35: SVD ALGORITHM DID NOT CONVERGE
UWM 3034:
ORTHOGONALITY TEST FAILED, LARGEST TERM = X NUMBER FAILED = PAIR = X,
EPSILON = X.
This message is printed when the eigenvector accuracy is in doubt (up to a certain
numerical limit). This message is given for all methods.
SFM 3034.2:
INTERNAL FAILURE IN THE LANCZOS PROCEDURE: M-ORTHOGONAL QR
PROCEDURE FAILED TO CONVERGE. PROBABLE CAUSE: MASS MATRIX IS
INDEFINITE (MODES) OR STIFFNESS MATRIX IS INDEFINITE (BUCKLING).
Indicates that the mass/stiffness matrix is indefinite or badly scaled.
UIM 5010:
STURM SEQUENCE DATA FOR EIGENVALUE EXTRACTION TRIAL EIGENVALUE
= X, CYCLES = X, NUMBER OF EIGENVALUES BELOW THIS VALUE = X.
This information is very important in establishing the number of roots existing in certain
subregions of the frequency region.
UFM 4646:
THE MASS MATRIX IS NOT POSITIVE DEFINITE USING HOUSEHOLDER OR GIVENS
METHOD.
UFM 4645:
THE SHIFTED STIFFNESS MATRIX IS NOT POSITIVE DEFINITE IN MGIVENS OR
MHOUSEHOLDER METHOD.
UFM 4648:
THE MODAL MASS MATRIX IS NOT POSITIVE DEFINITE.
CHAPTER 6
These messages report problems from the reduction methods.

UWM 5411:
NEGATIVE TERM ON DIAGONAL OF MASS (NORMAL MODES) OR STIFFNESS
(BUCKLING) MATRIX.
This is a Lanczos diagnostic message for information only.
Error Diagnostics
UFM 5429:
INSUFFICIENT TIME TO START LANCZOS ITERATION.
UFM 5400:
INCORRECT RELATIONSHIP BETWEEN FREQUENCY LIMITS.
This means the upper frequency limit has a lower value than the lower frequency limit.
SFM 5401:
LANCZOS METHOD IS UNABLE TO FIND ALL EIGENVALUES IN RANGE.
ACCEPTED EIGENVALUES AND ADDITIONAL ERROR MESSAGES MAY BE LISTED
ABOVE. A POSSIBLE CAUSE IS THE OCCURENCE OF HIGH MAXRATIOS. CHECK
MODEL FOR MECHANISMS IF HIGH MAXRATIOS EXIST. USER ACTION: RERUN
WITH ANOTHER METHOD OR ANOTHER SETTING ON EIGRL ENTRY.
UFM 5402:
THE PROBLEM HAS NO STIFFNESS MATRIX.
UFM 4683:
MASS (OR STIFFNESS) MATRIX NEEDED FOR EIGENVALUE ANALYSIS.
UWM 6243 (REIG):
THE DOF REQUESTED FOR POINT NORMALIZATION HAS NOT BEEN SPECIFIED
ON THE EIGR OR EIGB ENTRY.
SFM 5299:
FINITE INTERVAL ANALYSIS ERROR or
STARTING BLOCK COMPUTATION ERROR or
INSUFFICIENT STORAGE FOR LANCZOS or
FACTORIZATION ERROR ON THREE CONSECUTIVE SHIFTS or
RESTORATION OF VECTORIZATION ERROR or
IMPROPER PARAMETER SPECIFICATION FOR LANCZOS or
TRUST REGION OVERFLOW IN LANCZOS or
UNRECOVERABLE TERMINATION FROM LANCZOS ITERATION
151
152
These messages issued under 5299 are from the Lanczos method. The first one, the finite
interval analysis error, is the one most frequently encountered. This error indicates a high
matrix-factor diagonal ratio at the F1 or F2 shifts which can be caused by a modeling error
or matrix singularity.
SFM 5407:
INERTIA (STURM SEQUENCE) COUNT DISAGREES WITH THE NUMBER OF MODES
ACTUALLY COMPUTED IN AN INTERVAL.
This message flags a serious problem, i.e., spurious modes were found in the Lanczos
method.
UWM 5406:
NO CONVERGENCE IN SOLVING THE TRIDIAGONAL PROBLEM.
This message signals the abortion of the reduction methods.
UIM 5403:
BREAKDOWN OF CPU USAGE DURING LANCZOS ITERATIONS
Eigenvalue analysis can be computationally expensive and may dominate overall CPU
time. To help assess numerical performance, this message shows how much time the
primary operations (forward-backward substitution, matrix-vector multiplication, and
matrix summation and decomposition) consume during Lanczos iterations. The last entry,
LANCZOS RUN, refers to the duration of a complete set of Lanczos cycles at one shift
and contains within it all FBS and matrix multiplication times, but not the shift and factor
times. The sum of total times for SHIFT AND FACTOR and LANCZOS RUN should
then approximate the total time taken by the REIGL or LANCZOS modules.
*** USER INFORMATION MESSAGE 5403 (LNNRIGL)
BREAKDOWN OF CPU USAGE DURING LANCZOS ITERATIONS:
OPERATION
REPETITIONS
FBS (BLOCK SIZE=...)
....
MATRIX-VECTOR MULTIPLY
....
SHIFT AND FACTOR
....
LANCZOS RUN
....
AVERAGE
........
........
........
........
TOTAL
........
........
........
........
UIM 2141:
GIVENS (OR HOUSEHOLDER) TIME ESTIMATE IS X SECONDS. SPILL WILL OCCUR
FOR THIS CORE AT A PROBLEM SIZE OF X.
The reduction type methods are fairly predictable (not a characteristic of other eigenvalue
methods). The CPU time and storage estimate are given in this message.
Lanczos Diagnostics
Since the Lanczos method is the most robust and modern of the eigenvalue extraction
methods, its execution diagnostics are described here in greater detail.
CHAPTER 6
The printing of diagnostic information is controlled by the MSGLVL parameter. When left
at its default value of zero, only the Table of Shifts and the Execution Summary block are
printed in addition to the eigensolution. MSGLVL values of 1, 2, 3, or 4 yield increasingly
more detailed diagnostic information about the Lanczos operations.
The MSGLVL=1 diagnostics are organized into four major sections. Section I reports on the
original problem specification and the setting of module parameters. In this section an
interval analysis check is also shown to establish the number of eigenvalues in the range
set by the user.
Most of the detailed diagnostics are self explanatory. Some of the parameter values are
detailed below:
MODE FLAG
PROBLEM TYPE
= 1
Vibration problem
Buckling problem
= 1
Lowest ND roots in interval
Highest ND roots
All roots in interval
ND roots nearest to center frequency
The LEFT and RIGHT END POINTS are the F1, F2 values set on the EIGRL entry. The
center frequency is the center (not necessarily the arithmetic center) of the interval.
ACCURACY REQUIRED is a value automatically set by the program.
The CP TIME ALLOWED is the remainder of the time left to complete the run using the
limit set on the TIME entry.
The SHIFTING SCALE is an estimate of the smallest (in magnitude) nonzero eigenvalue in
the spectrum. This estimate can be specified by the user or automatically calculated by the
program. The INERTIA values at the specific locations are Sturm numbers.
Section II provides diagnostics on the memory usage, the setting of the working array sizes
based on the memory available, and the maximum memory allocated. The term RITZ
VECTORS means the approximate eigenvectors. A TRUST REGION is a segment of the
frequency spectrum, bounded by two shifts where all the eigenvalues are found. By the
nature of the automatic shift algorithm, there can be several of these segments.
Section III is the Lanczos run section. The text shown in this section can be repeated for each
new shift. This section also prints occasional user information messages (for example, 5010
and 4158) that report the results from the decomposition module. This section is further
expanded when the user requests additional diagnostics from the Lanczos run by setting
MSGLVL = 2 or 3.
Section IV reports the conclusion of the Lanczos run. The most frequently occurring
warning flag settings are listed below:
153
154
COMPLETION FLAG
- 99ND roots nearest to center frequency

- 26Starting block computation error
- 25Restoration of factorization error
- 24Illegal parameters error
- 23Core structure error
- 22Internal factorization error
- 21Insufficient storage
- 20Factorization error at a boundary shift
- 5Incorrect frequency range
- 2No eigenvalues found in range
0Required number of roots found
1All roots in interval found
2Not all roots in interval found
3Insufficient time to finish
4 - 7Same as 1- 3; however, the inertia count error (see
SFM 5407) occurred during the iteration
MSGLVL = 2 provides detailed information about the shift logic, the spectrum distribution,
and the cause of the termination of a Lanczos run. This TERMINATION CRITERION may
have the following values:
0Preset maximum number of Lanczos steps reached
1Cost of eigenvalue calculation increasing
2Number of needed eigenvalues was found
3Shift is too close to an eigenvalue
4Lanczos block is singular
5Running out of time
- 1Insufficient workspace
- 2QL algorithm does not converge
- 3Too many eigenvalues found
- 4File I/O error
- 5Singular value decomposition error
Finally, MSGLVL = 3 describes the Lanczos run including norms, condition numbers,
convergence criterion, shifted eigenvalues, and their residuals.
CHAPTER 6
6.8
Real Lanczos Estimates and Requirements

The time estimates for the Lanczos method are the following:
Shifting time (sec):
T d N shifts
Eq. 6-102
N steps N shifts ( 2T M + T s )
Eq. 6-103
Recursion time (sec):
Ortho-normalization time (sec):

2
2 N shift N steps
M
Eq. 6-104
2 ( N des + N steps ) N P
Eq. 6-105
Packing time (sec):
where:
N shifts = number of shifts
N steps = number of Lanczos steps
B = block size used
N des = number of modes desired
T d = decomposition time (see Decomposition Estimates and Requirements on
page 61 for details)
T s = solution time (see FBS Estimates and Requirements on page 72 for details)
T M = matrix multiply time (see MPYAD Estimates and Requirements on page 44
for details)
Disk:
N des N IPREC + D factor
Memory:
2 ( N steps B ) IPREC + 4 B N IPREC
where D factor is the factor disk space requirement.

The number of shifts in most Lanczos runs is 2, occasionally 1, sometimes 3 or more. The
number of Lanczos steps is 10 on the average.
155
156
6.9
References
Givens, W. Numerical Computation of the Characteristic Values of a Real Symmetric
Matrix. Oak Ridge National Lab., ORNL-1574, 1954.
Grimes, R. G., et al. A Shifted Block Lanczos Algorithm for Solving Sparse Symmetric
Generalized Eigenproblems. SIAM, J. Mat. Analysis Appl., 13, 1992.
Householder, A.S.; Bauer, F.L. On Certain Methods for Expanding the Characteristic
Polynomial. Numerische Mathematik, Volume 1, 1959, pp. 29-37.
Lanczos, C. An Iteration Method for the Solution of the Eigenvalue Problem of Linear
Differential and Integral Operators. Journal of the Research of the National Bureau
of Standards., Volume 45, 1950, pp. 255-282.
Lewis, J. G.; Grimes, R. G. Practical Lanczos Algorithms for Solving Structural
Engineering Eigenvalue Problems. Sparse Matrices, and Their Uses, edited by
I. E. Duff, Academic Press, London, 1981.
MacNeal, R. H.; Komzsik, L. Speeding Up the Lanczos Process. RILEM, Kosice, Slovakia,
1995.
Ortega, J. M.; Kaiser, H. F. The LR and QR Methods for Symmetric Tridiagonal Matrices.
The Computer Journal, Volume 6, No. 1, Jan. 1963, pp. 99-101.
Parlett, B. N. The Symmetric Eigenvalue Problem. Prentice Hall, Englewood Cliffs, 1980.
Smith, B. T. et al. Matrix Eigensystem Routines - EISPACK Guide. Springer Verlag, 1974.
Wilkinson, J. H. The Algebraic Eigenvalue Problem. Oxford University Press, 1965.
Wilkinson, J. H. The Calculation of the Eigenvectors of Codiagonal Matrices. The
Computer Journal, Volume 1, 1958, p. 90.
CHAPTER 6
157
158
CHAPTER
Complex Eigenvalue Analysis

Damped Models
Theory of Complex Eigenvalue Analysis
User Interface
Method Selection
Option Selection
Complex Eigenvalue Diagnostics
Complex Lanczos Estimates and Requirements
160
7.1
Damped Models
The solution of complex eigenvalue problems is important for damped models. The
solution method is applied when either viscous or structural (or both) damping is used.
The basic equation of the complex eigenvalue analysis is
Mu + Bu + Ku = 0
Eq. 7-1
where:
[ M ] = mass matrix
[ B ] = damping matrix
[ K ] = stiffness matrix
The B matrix may be purged and the M and K matrices may be real or complex,
symmetric or unsymmetric.
The eigenvalue is given by
= a + i
Eq. 7-2
The complex eigenvalue and eigenvector solution is of the form:

u = e
Eq. 7-3
CHAPTER 7
7.2
Theory of Complex Eigenvalue Analysis

Canonical Transformation to Mathematical Form
The complex eigenvalue analysis problem in general is a quadratic eigenvalue problem in
the form:
Mu + Bu + Ku = 0
Eq. 7-4
where u is the displacement vector. u is the acceleration of the grid points, i.e., the second
time derivative of u . u refers to the velocity or first time derivative. The solution of this
homogeneous system (the free, but damped vibrations) is of the form
u = e
Eq. 7-5
where is a vector of complex numbers and the eigenvalue is also complex in general.
By substituting Eq. 7-5 into Eq. 7-4 we get:
2
( M + B + K ) = 0
Eq. 7-6
In order to solve the quadratic eigenvalue problem, first a linearization transformation is

executed. This transformation converts the original quadratic problem to a linear problem
of twice the size.
It is obtained by simply rewriting Eq. 7-6 as a 2 2 block matrix equation:
M 0
0 I
B K
I 0
= 0
Eq. 7-7
where:
Eq. 7-8
This equation is now linear; however, there are shortcomings. Independently of the
eigenvalue solution method, one would need to invert both the mass and damping
matrices and an unsymmetric, indefinite matrix built from the damping and stiffness
matrices, in order to reach a solution. Although the explicit inverses are not needed, the
numerical decompositions on either of these matrices may not be very well defined.
An advancement is possible by executing a spectral transformation, i.e., introducing an
appropriate shift as:
= 0 +
Eq. 7-9
With the shift (whose appropriate selection in the complex case is rather heuristic) the
linear equation may be rewritten as:
161
162
B M 0
I
K
0 I
M 0
0 I
Eq. 7-10
Another recommended improvement is to invert the problem by introducing:

1
= --
Eq. 7-11
By substituting and reordering we get:
B M 0
I
0 I
M 0
0 I
Eq. 7-12
The latter equation is a canonical form of

x = Ax
Eq. 7-13
where:
B M 0
I
A =
0 I
M 0
Eq. 7-14
0 I
The form allows the singularity of the mass matrix. However, the zero subspaces of K , B ,
and M may not coincide. This is a much lighter and more practical restriction. This is the
formulation used in the Lanczos methods.
If M is nonsingular, then a simplified formulation of
( M
B + 0 I )
0 I
Eq. 7-15
also results in a canonical form of

Ax = x
Eq. 7-16
where:
x =
This is the formulation used in the Hessenberg methods.

When B is null, then the double-size representation can be avoided.
The problem then becomes
2
[ M + K ] = 0
Eq. 7-17
CHAPTER 7
Using the following:
= 0 +
Eq. 7-18
We can write
[ M 0 + K ] = M
Eq. 7-19
To premultiply Eq. 7-19, we use

[ M
K + 0 I ] =
Eq. 7-20
Therefore, the dynamic matrix A for the case of nonsingular M matrix is as follows:
1
M B + 0 I
I
A =
[ M K +
M 1 K
0 I
0 I ]
when B 0
Eq. 7-21
when B = 0
Because of the unsymmetric nature, the following so-called left-handed solution also
exists:
H
( M + B + K ) = 0
Eq. 7-22
For the special case of the structural damping problem, the left-handed solution of form:
H
( M + K) = 0
Eq. 7-23
also exists.
In Eq. 7-22 and Eq. 7-23, the superscript H stands for complex conjugate transpose.
The left-handed eigenvectors of the problems will only be found in the Lanczos and QZ
Hessenberg methods. The left-handed eigenvectors are useful in establishing convergence
quality.
The physical eigenvalues may easily be recovered from the back substitution of the shift
and invert operations:
1
= ---- + 0
Eq. 7-24
In order to find the relationship between the mathematical and physical eigenvectors, let
us rewrite Eq. 7-7 in the following block notation:
( M + K ) x = 0
where again
Eq. 7-25
163
164
x =
Eq. 7-26
The block matrices are simply:

K =
B K
I 0
Eq. 7-27
M =
M 0
0 I
Eq. 7-28
and
Substituting Eq. 7-24 into Eq. 7-25 and reordering yields:

[ ( K + 0 M )
M + I ] x = 0
Eq. 7-29
which is the same as Eq. 7-11 in the block notation. This proves that the right eigenvectors
are invariant under the shifted, inverted transformations, i.e., the right physical
eigenvectors are the same as their mathematical counterparts, apart from the appropriate
partitioning.
For the left-handed problem of Eq. 7-22, we rewrite Eq. 7-22 using the block notation:
H
y ( M + K ) = 0
Eq. 7-30
with a left-handed physical eigenvector of:

y
H H
= [ ]
Eq. 7-31
Substituting Eq. 7-24 again and by factoring:

H
y ( K + 0 M ) [ ( K + 0 M )
M + I ] = 0
Eq. 7-32
[ ( K + 0 M ) y ] [ M + ( K + 0 M ) ] = 0
Eq. 7-33
By grouping:
H
which is equivalent to:

H
[ ( K + 0 M ) y ] [ ( K + 0 M )
M + I ] = 0
Eq. 7-34
In this case, the mathematical problem we solve is:

H
y [ ( K + 0 M )
M + I ] = 0
Eq. 7-35
CHAPTER 7
This means that the left-handed physical eigenvectors are not invariant under the
transformation. Comparing gives the following relationship:
y
= [ ( K + 0 M ) y ]
Eq. 7-36
or expanding into the solution terms

B M 0
y =
I
Eq. 7-37
0 I
and finally
B M 0
y =
I
1, H
Eq. 7-38
0 I
The cost of this back-transformation is not very large since the factors of the dynamic
matrix are available and we need a forward-backward substitution only. The importance
of this back-transformation is rather academic since there is no physical meaning
associated with the left physical eigenvectors. On the other hand, these are the eigenvectors
output in the DMAP data block PSI, and they are not orthogonal to PHI unless properly
converted.
Dynamic Matrix Multiplication

In any eigenvalue solution method, the A dynamic matrix times a vector (block)
multiplication is the most time consuming. From Eq. 7-13, the A matrix is
A =
B M 0
I
M 0
0 I
Eq. 7-39
From its structure it is clear that the matrix does not need to be built explicitly. In the
following, the implicit execution of the dynamic matrix multiplication in both the
transpose and non-transpose case is detailed.
In the non-transpose case, any z = Ax operation in the recurrence will be identical to the
B M 0 K
z =
I
0 I
M 0
0
Eq. 7-40
solution of systems of equations. Let us consider the partitioning of the vectors according
to the A matrix partitions:
B M 0 K
I
0 I
z1
z2
M 0
x1
x2
Eq. 7-41
165
166
Developing the first row:

( B M 0 ) z1 K z2 = M x1
Eq. 7-42
Developing the second row, we have:

z1 = 0 z2 + x2
Eq. 7-43
Substituting the latter into the first row, we get:

( B M 0 ) ( 0 z2 + x2 ) K z2 = M x1
Eq. 7-44
Some reordering yields:

2
( K + 0 B + 0 M ) z2 = M x1 + ( B + 0 M ) x2
Eq. 7-45
The latter formulation has significant advantages. Besides avoiding the explicit
formulation of A , the decomposition of the 2N size problem is also avoided.
It is important that the transpose operation be executed without any matrix transpose at
all. In this case, any T = x T A operation in the recurrence will be identical to the
= x
B M 0 K
I
0 I
M 0
0
Eq. 7-46
operation. Let us introduce an intermediate vector z :
= x
B M 0 K
I
0 I
Eq. 7-47
Now partitioning these vectors according to the matrix partitions and transforming we
obtain:
x1
x2
B M 0 K
I
0 I
z1
z2
Eq. 7-48
Developing the first row:

T
x1 = ( B M 0 ) z1 + z2
Eq. 7-49
Developing the second row, we have:

T
x2 = K z1 0 z2
Eq. 7-50
Expressing z 2 from the first equation, substituting into the second and reordering yields:
CHAPTER 7
Eq. 7-51
( x2 + 0 x1 )
Eq. 7-52
x 2 + 0 x 1 = ( K + 0 B + 0 M )z 1
or in the solution form:
T
T 1
z1 = ( K + 0 B + 0 M )
The lower part of the z vector is recovered from Eq. 7-49 as:
T
z2 = x1 + ( B + 0 M ) z1
Eq. 7-53
Finally
= z
M 0
0
Eq. 7-54
Physical Solution Diagnosis

From the eigenvalue solution it will be guaranteed that the left and right mathematical
eigenvectors are bi-orthonormal:
H
Y X = I = XY
Eq. 7-55
where I is an identity matrix in essence with computational zeroes as off diagonal terms.
Based on the physical eigenvalues recovered by the shift formulae and the physical
eigenvectors, another orthogonality criterion can be formed. Using the left and right
solutions, the following equations are true for the problem:
2
( M i + B i + K ) i = 0
2
j ( M j + B j + K ) = 0
Eq. 7-56
Eq. 7-57
By appropriate pre- and postmultiplications, we get:

H
j ( M i + B i + K ) i = 0
j ( M j + B j + K ) i = 0
Eq. 7-58
Eq. 7-59
A subtraction yields:
H
j M i ( i j ) + j B i ( i j ) = 0
Eq. 7-60
Assuming j i we can shorten and obtain a mass orthogonality criterion as:

H
O 1 = j M i ( i + j ) + j B i
Eq. 7-61
167
168
The O 1 matrix has (computational) zeroes as off diagonal terms (when i j ) and nonzero
(proportional to 2 i ) diagonal terms.
H
By premultiplying Eq. 7-56 by j j , postmultiplying Eq. 7-57 by i i and subtracting

them we obtain:
H
j j ( M j + B i + K ) i j ( M j + B j + K ) i i = 0
Eq. 7-62
By expanding and cancelling we get:

H
( i + j ) j M i i j + ( j + i ) j K i = 0
Eq. 7-63
Assuming again that j i we can shorten and obtain another orthogonality condition
recommended mainly for the structural damping option as:
H
O 2 = j j M i i j K i
Eq. 7-64
This orthogonality matrix will also have zero off-diagonal terms, but nonzero
(proportional to i2 ) diagonal terms.
Hessenberg Method
The method utilized in MSC.Nastran uses the Householder reduction to upper Hessenberg
form followed by Franciss QR steps and a direct eigenvector generation scheme. To
distinguish from the QZ method, this is called the QR Hessenberg method.
Householder Transformations. An n by n matrix with the following form:
T
2vv
P = I ----------T
v v
Eq. 7-65
is introduced in Theory of Real Eigenvalue Analysis on page 101 as a Householder

transformation or reflection. These matrices are symmetric and orthogonal and capable of
zeroing out specified entries or any block of vector components. We generate a
Householder matrix for a given nonzero vector x :
x
= [ x 1, , x n ]
Eq. 7-66
so that for 1 k j n the elements from k + 1 to j of the transformed vector are zero;
i.e.,
Px = [ x 1, , x k, 0, 0, , x j + 1, , x n ]
Eq. 7-67
The procedure is as follows. We calculate

a
and build
= xk + + xj
Eq. 7-68
CHAPTER 7
= [ 0, , 0 + x k + a sign ( x k ), x k + 1, , x j, 0, , 0 ]
Eq. 7-69
It can be proven that if P is defined as in Eq. 7-65, then

Px = [ x 1, , x k 1, sign ( x k )a, 0, , 0, x j + 1, , x n ]
Eq. 7-70
which is the form described in Eq. 7-67. Similarly the matrix update:
T
PA = ( I vv )A = A v ( A v )
Eq. 7-71
It is shown that P need not be formed explicitly if v and = 2 v T v are available.

Implicit Matrix Update. For given values of A , v , and , the following algorithm in
MSC.Nastran overwrites A with PA , where
T
P = ( I vv )
For p = 1, , n
s = v ( k )A ( k, p ) + + v ( j )A ( j, p )
s = s
For i = k, , j
A ( i, p ) = A ( i, p ) s v ( i )
End loop i .
End loop p .
Note: v T = [ 0, , v , , v , 0, , 0 ] above. An analogous algorithm is used for the AP
k
j
update.
Householder Reduction to the Hessenberg Form. Let A be the general matrix on
which the transformation must be executed. Consider the following transformation:
Ar = Pr Ar 1 Pr
Eq. 7-72
where:
A0 = A
P r = I 2w w T
r t
T
wr wr
=1
The elements of w r are chosen so that A r has zeroes in the positions 1 through r 2 in the
r-th row. The configuration and partitioning of the A r 1 can be shown as:
169
170
r
Ar 1 =
nr
x x x
x x x
x x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
Hr 1
Cr 1
0
br 1 Br 1
where:
C r 1 = a submatrix of order r
H r 1 = of upper Hessenberg form
B r 1 = a square matrix order n r (part of original matrix)
b r 1 = a vector having n r components
The transformation matrix can be partitioned as follows:
r
Pr =
I 0
0 Qr
nr
0 I 2v r v r
where v r is a unit vector of order n r .

By executing the multiplication given in the right-hand side of Eq. 7-72, the following A r
is obtained:
r
Ar =
nr
Hr 1
0
Cr 1 Qr
T
cr Qr Br 1 Qr
where c r = Q r b r 1 .
If v r is chosen so that c r is null except for its first component, then H r of the order r + 1
takes the Hessenberg form.
In-Memory Algorithm. This formulation can be developed by writing P r as follows:
T
P r = I 2w r w r
Eq. 7-73
where w r is a unit vector of the order n with zeroes as the first r elements. By further
changes:
T
ur ur
P r = I ----------2
2K r
Eq. 7-74
CHAPTER 7
where:
u ir = 0, i = 1, 2, , r
u r + 1, r = a r, r + 1 S t
u ir = a ir, i = r + 2, , n
2
2K r
S r a r + 1, r S r
Because of the lack of symmetry, the pre- and postmultiplications in Eq. 7-72 must be
considered separately. The premultiplication takes the following form:
T
ur ( ur Ar 1 )
ur ur
P r A r 1 = I ----------- A r 1 = A r 1 ------------------------------- = F r
2
2
2K r
2K r
Eq. 7-75
The postmultiplication is as follows:

T
T
ur ur
( F r u r )u r
A r = F r P r = F r I ------------- = F r = ------------------------
2
2
2K r
2K r
Eq. 7-76
Because u r has zero elements in the 1 through r positions, it is seen that the
premultiplication leaves the first r rows of A r 1 unchanged, while the postmultiplication
leaves the first r columns unchanged. By introducing new intermediate vectors, the
following may be written:
T
ur Ar 1 = pr
Eq. 7-77
where p r has its first ( r 1 ) elements as zero. This results in

( u r ) T
F r = A r 1 ---------- p r
2K 2
Eq. 7-78
For the postmultiplication, the q r vector is introduced

Fr ur = qr
Eq. 7-79
which has no zero components. Finally,

ur T
A r = F r q r ----------
2K 2
r
Eq. 7-80
Two-Level Storage (Spill) Algorithm. If the execution of memory transfers is a

requirement, the following formulation is more convenient. The vector p r is defined as in
Eq. 7-77
171
172
pr = ur Ar 1
Eq. 7-81
qr = Ar 1 ur
Eq. 7-82
However, q r is as follows:
By scaling elements of u r , the following is obtained:

ur
v r = ---------2
2K r
Eq. 7-83
Finally, introducing the scalar a r as follows:

Eq. 7-84
ar = pr vr
Eq. 7-72 now can be written in the following form:
T
a r = A r 1 v r p r q r v r + ( p r v r )u r v r
T
= Ar 1 vr pr ( qr ar ur ) vr
Eq. 7-85
QR Iteration Using the Householder Matrices

It is proven (Wilkinson, 1965 ) that the general A matrix can be factorized into the product
of a unitary matrix Q and an upper triangular matrix R . The algorithm at the r-th stage is
as follows:
Ar = Qr Rr
H
Ar + 1 = Qr Ar Qr = Qr Qr Rr Qr = Rr Qr
Eq. 7-86
By including the successive iterates,

H
A r + 1 = Q r A r Q r = ( Q r Q i 1, , Q 1 A 1 ( Q 1, Q 2, , Q r ) )
Eq. 7-87
From the following form of Eq. 7-86:

( Q 1, Q 2, , Q r ) A r + 1 = A 1 ( Q 1, Q 2, , Q r )
Eq. 7-88
it is seen that A r is unitarily similar to A 1 . In general, A r tends to take the upper triangular
form. The matrix Q r can be the product of ( n 1 ) elementary unitary transformations
necessary to reduce A r to the upper triangular form R r . The transformation matrices can
be Givens rotations or Householder reflections.
In MSC.Nastran the latter Householder reflections are used. The shift of origin may be
incorporated into the single QR logic as follows:
CHAPTER 7
Ar kr I = Qr Rr
Ar + 1 = Rr Qr + kr I
Eq. 7-89
If the matrix has complex conjugate eigenvalues, the most economical way is to execute
two steps (double step). For example, the first double shift with shifts k 1 and k 2 can be
written as follows:
A1 k1 I = Q1 R1
A2 k2 I = Q2 R2
R1 Q1 + k1 I = A2
R2 Q2 k2 I = A3
Eq. 7-90
and it follows that

H
A3 = ( Q1 Q2 ) A1 Q1 Q2
Eq. 7-91
By introducing
Q =
Q1 Q2
R = R1 R2
Eq. 7-92
QR = ( A 1 k 1 I ) ( A 1 k 2 I )
Eq. 7-93
the following is obtained:
or
T
Q ( A1 k1 I ) ( A1 k2 I ) = R
Eq. 7-94
which means that Q T is the orthogonal matrix that reduces 1 k1 I ) ( A 1 k 2 I to the upper
triangular form.
Implicit QR Step. Using the Householder matrices again,
T
P r = I 2w r w r
Eq. 7-95
where w r = ( 0, 0, , 0, x, , x )
The derivation of A 3 can be produced as follows. First we create a Householder matrix
which provides the following:
P 1 x = ke 1
where:
x
= ( x , y , z , 0, , 0 )
1 1 1
x 1 = ( a 11 k 1 ) ( a 11 k 2 ) + a 12 a 21
y 1 = a 21 ( a 11 k 2 ) + ( a 22 k 1 )a 21
Eq. 7-96
173
174
z 1 = a 32 a 21
e 1 = first unit vector
Then we compute the following:
Eq. 7-97
C1 = P1 A1 P1
(See the update with Householder matrices in Theory of Real Eigenvalue Analysis on
page 101.)
The C 1 matrix no longer takes the Hessenberg form. Therefore, this matrix must be
reduced by new Householder matrices:
P n 2, , P 2, P 1, A 1, P 1, P 2, , P n 2 = A 3
Eq. 7-98
Now the question is how to formulate the P r matrices. The nonzero elements of the P r
Householder matrices are determined by x 1, y1, z 1 for P 1 and by x r, y r, z r for P 2, , P n 2 .
Convenient representation of P r (see also Theory of Real Eigenvalue Analysis on
page 101 on Householder method) for the current case is as follows:
T
2v r v r
P r = I -------------T
vr vr
T
v r = ( 0, , 0, 1, u r, v r, 0, , 0 )
yr
u r = ----------------------( xr ar )
Eq. 7-99
zr
v r = ----------------------( xr ar )
2
2
----------- = ----------------------------------- = r
T
2
2
( 1 + ur + vr )
vr vr
Then
T
Pr = I vr ( r vr )
Eq. 7-100
For the update, the algorithm described in Theory of Real Eigenvalue Analysis on
page 101 is used.
Eigenvector Computation
The Hessenberg form is convenient for generating eigenvectors when the eigenvalues are
known. The k-th eigenvector of the Hessenberg matrix H can be found by solving the
following:
( H k I )y k = 0
Eq. 7-101
CHAPTER 7
where k is the k-th eigenvalue.

This equation in detailed form is as follows:
h 11 k
h 12
h 13
h 21
h 22 k
h 23
h i, i 1 h i, i k
.
.
yk ( 1 )
h 1n
h 2n
yk ( i )
h in
= 0
Eq. 7-102
yk ( n 1 )
yk ( n )
h n, n 1 h nn k
The matrix of the system is singular, but the minor corresponding to the upper right corner
element h1n is not. Therefore, we normalize the y k vector so that the last element in it is
equal to 1.
yk ( n ) = 1
Eq. 7-103
The next to the last element in this case is given by solving the following:
h n, n 1 y k ( n 1 ) + ( h nn y k ) k ( n ) = 0
Eq. 7-104
which yields the following:

1
y k ( n 1 ) = -------------------- ( k h nn )
h n, n 1
Eq. 7-105
The other elements can be evaluated in recursive form from:

n
h i, i 1 y k ( i 1 ) + ( h i, i k ) y k ( i ) +
h i, l y k ( l ) =
Eq. 7-106
l = i+1
Then the ( i 1 ) -th element, i = n 1, n 2, , 2 is as follows:

1
y k ( i 1 ) = ------------------ ( k h i, i ) y k ( i )
h i, i 1
h i, l y k ( l )
Eq. 7-107
l = i+1
The practical implementation of the above procedure takes care of the cases where
decoupling occurs, i.e., when hi, i 1 = 0 . This process is still very unstable and an
alternative iterative solution of Eq. 7-101 exists in MSC.Nastran, which is similar to the
procedure detailed in Theory of Real Eigenvalue Analysis on page 101.
Transformation of the Vectors. If Eq. 7-72 is extended, the following is obtained:
H = P n 2 P n 1, , P 1 A 1 P 1, , P n 2
By introducing
Eq. 7-108
175
176
Z = P1 Pn 2
Eq. 7-109
the following is obtained:

T
H = Z AZ
Eq. 7-110
AZ = ZH
Eq. 7-111
or
If Eq. 7-111 is multiplied by an eigenvector of the H matrix, the following is obtained:

AZy = ZHy
Eq. 7-112
Hy = y
Eq. 7-113
because
Finally, the result is the following equation:

AZy = Zy
Eq. 7-114
which indicates that the x eigenvectors of A can be obtained by the following:

Eq. 7-115
x = Zy
This calculation is also straightforward if the Z matrix is accumulated during the

transformation from A to H . However, it is not practical to explicitly form the Z matrix
when the P r matrices are not formed. An effective implicit formulation follows.
Implicit Vector Transformation. Write Eq. 7-115 as follows:
x = P 1 P 2, , P n 2 y
Eq. 7-116
when P i represents the Householder transformation matrices. Taking advantage of their

(n 2)
(0)
matrix structure using y = y
, and x = y , write the following:
y
(r 1)
= Pr y
(r)
Eq. 7-117
and
(r 1)
= y
(r)
(r)
ur y
------------------ u r
2
2K r
Eq. 7-118
With this formulation only the u r vectors need to be accumulated during the A H
transformation and saved in the secondary storage for the vector transformations.
QZ Hessenberg Method
The QZ method is a robust algorithm for computing eigenvalues and eigenvectors of the
eigenvalue problem
CHAPTER 7
Ax = Cx
and, as such, applicable to the B = 0 case. If B 0 , the appropriate canonical
transformation is executed to linear form. There are no restrictions on A or C . For details,
see Golub and Van Loan, p. 375.
Hessenberg-Triangular Form
and Z so that the
The first stage of the QZ algorithm is to determine unitary matrices Q
H
H
matrix = Q CZ is upper triangular and = Q AZ is upper Hessenberg.

This is accomplished by first computing a unitary matrix U such that U H C is upper
triangular. Next, the matrix U H A is reduced to an upper Hessenberg matrix S by
premultiplying by a series of unitary matrices [ Q jH ] jk = 1 and postmultiplying by a series
of unitary matrices [ Z j ] H .
j=1
The matrices
Q jH
and Z j are carefully chosen so that
H AZ
S = Q jH Q jH 1 , , Q 1H U H AZ 1 Z 2, , Z j = Q
is upper Hessenberg, while
H CZ
T = Q H Q jH 1 , , Q 1H U H CZ 1 Z 2, , Z j = Q
remains upper triangular.
The QZ Step
The derivation of the QZ step is motivated by the case where T is nonsingular. If T is
nonsingular, we could form ST 1 and apply the Francis QR algorithm. Rather than
forming ST 1 explicitly, the QZ algorithm updates S and T using unitary matrices Q
and Z :
S = Q H SZ
T = Q H TZ
The matrices Q and Z are chosen so that the matrix T 1 is essentially the same as if a QR
step had been explicitly applied to ST 1 . However, since we operate with S and T rather
than ST 1 , it is not necessary for T to be invertible.
The QZ iteration continues until the matrix S converges to upper triangular form.
Eigenvalue Computation
Once S has converged to upper triangular form, the QZ algorithm will have determined
unitary matrices Q and Z so that both S = Q H AZ and T = Q H CZ are upper triangular.
Denote the diagonal entries of S by 1, 2, , Q n , and the diagonal entries of T by
1, 2, , n .
177
178
Then, for each j = 1, 2, , n , the matrix j S j T is singular. It follows that there exists a
vector uj so that j Su j = j Tu j.
Substituting for S and T and multiplying on the right by Q , we have j AZuj = j CZu.
Hence, if j 0 , we set j = j j and j = Zuj to get x j = j Cx j, as desired. However, if
j = 0 , then we have two cases to consider:
1. j = 0, j 0
2. j = 0, j = 0
In this case, j is said to be an infinite eigenvalue.

Here j is indeterminate.
The Complex Lanczos Method

The main steps of the complex Lanczos Method are: reduction to tridiagonal form, solution
of the tridiagonal problem, and eigenvector computation.
MSC.Nastran currently supports two complex Lanczos methods: the single vector and the
adaptive, block method.
The Single Vector Method

Reduction to Tridiagonal Form. The recurrence procedure introduced in Theory of
Real Eigenvalue Analysis on page 101 can also be used to reduce a general matrix to
tridiagonal form. This is the so-called biorthogonal Lanczos method. Find a nonsingular V
matrix such that:
Eq. 7-119
AV = VT
or
AV = T =
1 2
2 2 3
.
Eq. 7-120
.
Bn n
Then with U = V
, the following relation holds:

T
A U = UT
Eq. 7-121
Upon comparing columns of the right- and left-hand sides of Eq. 7-119 and Eq. 7-121, the
following is obtained:
Av j = j v j 1 + j v j + j + 1 v j + 1
T
A uj = j uj 1 + j uj + j + 1 uj + 1
Eq. 7-122
CHAPTER 7
where 1 v 0 = 1 u 0 = 0 and = 1, 2, , n 1, u and v j are complex rows and columns of

the U and V matrices.
Reordering Eq. 7-122 results in the following recurrence relations:
j + 1 v j + 1 = Av j j v j j v j 1
Eq. 7-123
j + 1 uj + 1 = A vj j uj j uj 1
There is some flexibility in choosing the scale factors of j , j . One possibility is to select
j = 1 , which results in an unsymmetric T matrix with ones on the subdiagonal. This
form is advantageous for a direct eigenvector generation scheme.
Another possibility is to select j = j, which results in a complex but symmetric T matrix.
This matrix has overwhelming advantages in the eigenvalue extraction from the
tridiagonal form, and MSC.Nastran uses this method in the implementation. Thus, the new
form of Eq. 7-123 is
j + 1 v j + 1 = Av j j v j j v j 1
T
j + 1 uj + 1 = A uj j uj j uj 1
Eq. 7-124
which is in matrix form as follows:
2
.
.
U AV = T =
Eq. 7-125
n 1
n
n
n
From the biorthonormality of vector u and v , derive the explicit form of the q coefficients.
Biorthonormality means that
0 if i j
T
ui vj =
1 if i = j
T
Eq. 7-126
The premultiplication of Eq. 7-124 by u j , u j + 1 , and v j , v j + 1 , respectively, results in the

following (using Eq. 7-126):
u
j = u i Av j or j = v j Au j
and
u
j + 1 = u j + 1 Av j or j + 1 = v j + 1 A u j
Eq. 7-127
The two versions of j and j + 1 should be the same in exact arithmetic. In the actual
implementation, the average of the two versions is used.
179
180
The algorithm can be developed as follows:

u
j = uj v j ; j = v j Auj ; j = ( j + j ) 2
v j + 1 = Av j j v j j v j 1
T
u j + 1 = A u j j uj j u j 1
T
j + 1 Av j
; j + 1 = vj + 1 A uj ; j + 1 = ( j + 1 + j + 1 ) 2
Eq. 7-128
vj + 1
vj + 1 = -----------j + 1
uj + 1
uj + 1 = -----------j + 1
where j = 1, , m < n . The algorithm starts with
biorthonormal random starting vectors for u 1 and
u0 = v 0 = 0, 1 = 0 as well as
v1 .
The breakdown of the procedure occurs when j + 1 becomes equal to zero. Then the
process is restarted with new j + 1, v j + 1 vectors. In our implementation this is done when
j + 1 < j
Eq. 7-129
where is a small number (related to the machine accuracy).

Solution of Tridiagonal Problem. The method for extracting the eigenvalues of the
complex tridiagonal form is a variant of the basic QL procedure, mentioned previously in
Theory of Real Eigenvalue Analysis on page 101. At each iteration the basic QL
procedure factors a shifted version of the current iterate as follows:
T l = QL
Eq. 7-130
where L is a lower triangular matrix, Q is orthogonal and obeys the following:

T
Q Q = I
Eq. 7-131
The next iterate T i + 1 is obtained as follows:

T 1 = QL + l
Eq. 7-132
Premultiplying by Q and postmultiplying by Q in Eq. 7-130 gives

T
T 1 = Q TQ
Eq. 7-133
By repeatedly applying Eq. 7-130 and Eq. 7-133, T i is finally converted into a diagonal
matrix whose elements are the eigenvalues of the original tridiagonal matrix; that is,
Tn =
Eq. 7-134
CHAPTER 7
Note that if the T matrix is complex, then the Q matrix is also complex.
The computation of Q is performed by a sequence of complex Givens transformation
matrices P k . Each P k is the identity matrix except for the entries P k ( i, j ) where
k i, j k + 1 . These terms are defined as follows:
P k ( k, k ) = P k ( k + 1 , k + 1 ) = C k
P k ( k + 1, k ) = P k ( k, k + 1 ) = S k
2
Eq. 7-135
Ck + Sk = 1
C k and S k are complex scalars. Each P k matrix satisfies
1
Pk = Pk = Pk
Eq. 7-136
The Q matrix is built as follows:

Q = P n 1, P n 2, , P 1
Eq. 7-137
First, P n 1 is determined by the n-th column of T with the diagonal entry shifted by .
Applying this transformation matrix to T , a new nonzero term is introduced on the
( n 2, n ) position. Subsequent values of P j are defined so that the new nonzero
introduced by P n 1 is forced up and out of the matrix, thereby preserving the complex
symmetric tridiagonal structure.
The algorithm below does not perform the explicit shifting, factorization, and
recombination of Eq. 7-130 and Eq. 7-132. Explicit shifting can result in significant loss of
accuracy. Instead, within a given iteration (generating T i + 1 from T i ), intermediate
matrices are produced as follows:
(k)
Ti
(k)
where T 0
(k 1)
= Pk Ti
P k, k = n 1, , 1
Eq. 7-138
= T , the original matrix.
Finally, the following is obtained:

(n)
(1)
Ti + 1 = Ti
Eq. 7-139
The process is repeated for i = 1, 2, , m or until the matrix T i becomes diagonal. Note
that this diagonalization occurs by successive decoupling of 1 1 and 2 2 submatrices of
T i . Decoupling occurs when an off-diagonal term is less than the product of an and the
sum of the absolute values of the two corresponding diagonal entries.
The selection of the C k, S k parameters can now be described. For k = n 1 , we must first
determine the shift that is chosen to be the one eigenvalue of the upper 2 2 submatrix
(n)
(n)
of T i closest to T i ( 1, 1 ) . Then the parameters are as follows:
181
182
(n)
T i ( n, n )
C n 1 = ------------------------------------an 1
Eq. 7-140
(n)
T i ( n 1, n )
S n 1 = -----------------------------------an 1
Eq. 7-141
where
(n)
an 1 = ( Ti
(n)
( n, n ) ) + T i ( n 1, n )
Eq. 7-142
(n)
Note again that by executing this transformation the term T i ( n 2, n ) becomes nonzero.
For the subsequent rotations k < 2 1 , the parameters are selected as follows:
(k + 1)
Ti
( k + 1, k + 2 )
C k = -------------------------------------------------------ak
(k + 1)
Ti
( k, k + 2 )
S k = --------------------------------------------ak
where
(k + 1)
( Ti
(k + 1)
( k + 1, k + 2 ) ) + ( T i
( k, k + 2
This step is reviewed in the following figure by using k = n 2 (the sub- and
superscripts of the T are ignored).
Eq. 7-143
CHAPTER 7
New term introduced

by the k =
T ( k 1, k )
n2
step
Term to be eliminated by
T ( k 1, k + 1 )
the k = n
step
T ( k, k + 2 )
T ( k, k + 1 )
New term introduced
or
by the k = n
T ( n 2, n )
T ( k + 1, k + 1 )
step
T ( k + 1, k + 2 )
or
or
T ( n 1, n 1 )
T ( n 1, n )
T ( n, n )
Figure 7-1
Chasing of Nonzero Offdiagonals.
As shown in Figure 7-1, the k-th transformation zeroes out T ( k, k + 2 ) , which was
introduced by the ( k + 1 ) th transformation, but generates a new nonzero term in
T ( k 1, k + 1 ) .
( k)
Using Eq. 7-143 the new terms of the matrix T i

(k + 1)
( 2C k ( T i
( k + 1)
( k, k + 1 ) + S k ) ) ( T i
are calculated as follows. Let

(k + 1)
( k, k ) T i
( k + 1, k
Eq. 7-144
Then
(k)
(k + 1)
T i ( k, k ) = T i
(k)
(k + 1)
T i ( k + 1, k + 1 ) = T i
(k)
( k + 1, k + 1 ) S k b k
(k + 1)
T i ( k, k + 1 ) = T i
(k)
( k, k ) S k b k
( k, k + 1 ) C k b k
Eq. 7-145
(k)
T i ( k 1, k ) = C k T i ( k 1, k )
(k)
(k)
T i ( k 1, k + 1 ) = S k T i ( k 1, k )
Finishing the calculation of Eq. 7-145, set k = k 1 and repeat from Eq. 7-144. When
Eq. 7-145 is finished for k = k , then use Eq. 7-142 and start the process from Eq. 7-144
again.
183
184
Breakdown of the Procedure. The calculation of C k and S k is performed using a

denominator with the following form:
2
a +b
Eq. 7-146
which can be very small without either a or b being small. For example, for a = 1 and
2
2
b = i, a + b = 0 . When this situation occurs, the process breaks down. The condition of
this problem can be formulated as follows. If
2
a +b
( a
+ b )
Eq. 7-147
then the procedure will terminate. in Eq. 7-147 is a small number related to the machine
accuracy.
The storage and computational efficiency of this algorithm is excellent since it needs O ( n )
2
storage and O ( n ) operations.
Error Bounds. Theoretically, the solutions of the following reduced eigenproblem
Ts i = i s i
Eq. 7-148
are approximations to the solutions of the original problem.

To estimate the accuracy of the solution, we use
i i m + 1 si ( m )
Eq. 7-149
Eq. 7-149 is a generalization of such an error bound in the current symmetric Lanczos code.
The error bound for the original eigenvalues of Eq. 7-2 can be found as follows:
m + 1 si ( m )
i i
----------------- ------------------------------------i
i
pi 0
m + 1 si m
for B 0 ---------------------- 1 ------------------------------i
pi 0
Eq. 7-150
pi 0
m + 1 si ( m )
for B = 0 ----------------------- 1 ------------------------------------2
i
pi 0
Eigenvector Computation. To calculate the eigenvectors of the tridiagonal form, an
inverse power iteration procedure is used. First, a random right-hand side is generated for
the inverse iteration. Then an LU decomposition of the tridiagonal matrix is performed by
Gaussian elimination with partial pivoting. After a back substitution pass, the convergence
of the approximate eigenvector is checked by its norm. If the norm is greater than one, then
the eigenvector is accepted. This norm shows sufficient growth, considering that the
CHAPTER 7
procedure began with a random right-hand side vector with elements less than one.
Otherwise, the eigenvector is normalized, and the process is repeated with this vector as
the right-hand side. The iteration is repeated up to three times.
Practice indicates that most of the time one iteration pass is sufficient. The computation is
very stable, especially with partial pivoting. The process can be summarized as follows.
1. Decomposition
P ( T i I ) = LU
Eq. 7-151
where:
T = the tridiagonal matrix
i = an eigenvalue
P = permutation matrix (row interchanges)
L, U = factors
2. Iteration
LU u 2 = u 1
Eq. 7-152
where:
u 1 = random starting vector
u 2 = approximate eigenvector
If u 2 < 1 and the iteration count is less than 3, then
u2
u 1 = -----------u2
Eq. 7-153
and the iteration continues; otherwise, u2 is the eigenvector.

3. Converting the Solution
The conversion of eigenvalues is performed by the following equations:
p i = i + 0, B 0
2
p i = i + 0, B = 0
The eigenvector conversion requires the computation of the following:
x i = Vs i , for the right vectors
or
y i = Us i , for the left vectors
where:
Eq. 7-154
185
186
s i = eigenvector of Eq. 7-162 corresponding to

U = [ u 1, , u m ]
V = [ v 1, , v m ]
m = number of roots found by the procedure at the particular shift
4. Initialization
The Lanczos recurrence is started with orthonormalized random vectors. Any
numerical procedure can be used that generates long enough pseudorandom
number sequences for the random number generation. Suppose the random
vectors are as follows:
u 1, v 1
Orthonormalization can be performed in the following form:
u1
u 1 = --------------------------12
T
( v1 u1 )
Eq. 7-155
v1
v 1 = --------------------------12
T
( u1 v1 )
These vectors are now suitable to start the Lanczos procedures.
5. Outer Orthogonalization
It is well known that the repeated occurrence of the eigenvectors can be prevented
by assuring orthogonality between the already accepted eigenvectors and the
Lanczos vectors. This outer orthogonality can be maintained using the following
iteration process (i.e., the modified Gram-Schmidt):
c
j
vi
j
vi
( x k v i )x k
k = 1
for j = 1, 2, , l
T
until
x k v i < for max k = 1, , c

This process should be executed similarly with the ui vectors.
The respective orthogonalization formula is as follows:
c
j
ui
j
ui
( y k u i )y k
k = 1
In the above formulae, xk, y k are already accepted eigenvectors and

k = 1, 2, , c .
Eq. 7-156
CHAPTER 7
6. Inner Orthogonalization
When the Lanczos procedure is carried out in practice using finite precision
arithmetic, then the strict biorthogonality of the u, v sequences is lost. The
biorthogonality can be maintained if uj + 1 is reorthogonalized with respect to
v 1, , v j + 1 , and v j + 1 is reothogonalized with respect to u 1, , u j + 1 using the
following formulae:
j
vj + 1 = vj + 1
ui vj + 1 vi
i = 1
j
uj + 1 = uj + 1
Eq. 7-157
T
vi uj + 1 ui
i = 1
For numerical stability, the subtraction is performed immediately after a term is

calculated, and the sum is not accumulated.
Inner orthogonality monitoring is performed by the following:
u
j =
v
j =
uj + 1 vj
Eq. 7-158
vj + 1 uj
No orthogonalization is necessary when

u
max ( j , j ) <
If necessary, the orthogonalization is performed against all previous vectors. If
the orthogonalizations are kept in the core, this fact implies a limit on m .
7. Shift Strategy
The default logic begins with a shift at (0.1, 1.0). The Lanczos recurrence process
is performed until breakdown. Then the eigensolutions of the reduced problem
are evaluated as approximations. If the user desires, then the shifts are executed
at user-specified locations as shown.
187
188
Im
( 1 1 )
( 2 2 )
( )
( k k )
Re
Default Shift
Figure 7-2
Recommended Shift Points for Complex Lanczos
The shifts are used until the required number of eigenvalues are found.
Unfortunately, there is no analogue to the Sturm sequence theory used for the real
eigensolution. As a result, there is no assurance that no gaps exist between the
eigenvalues found.
The Adaptive, Block Lanczos Method

The solution of the mathematical eigenproblem in its canonical form shown in Eq. 7-13 will
be more efficiently accomplished with the block Lanczos method.
The block Lanczos method (see Bai, et al., 1996) generates two sets of biorthonormal blocks
of vectors P j and Q j such that:
H
Pi Qj = I
Eq. 7-159
when i = j ( i, j n ) and zero otherwise. Note that we are using superscript H to denote
the complex conjugate transpose. These vector sets reduce the A system matrix to T j block
tridiagonal matrix form:
H
T j = P j AQ j
Eq. 7-160
P j = P 1 P 2 , , P j
Eq. 7-161
where the
and
CHAPTER 7
Q j = Q 1 Q 2, , Q j
Eq. 7-162
matrices are the collections of the Lanczos blocks. The structure of the tridiagonal matrix is:
Tj =
A1 B2
C2 A2
. .
. .
. .
Bj
Eq. 7-163
Cj Bj
The block Lanczos process is manifested in the following three term recurrence matrix
equations:
H
Bj + 1 Pj + 1 = Pj A Aj Pj Cj Pj 1
Eq. 7-164
Q j + 1 C j + 1 = AQ j Q j A j Q j 1 B j
Eq. 7-165
and
Note that in both of these equations the transpose of the system matrix A is avoided.
In order to find the mathematical eigenvalues and eigenvectors we solve the block
tridiagonal eigenvalue problems posed as:
H
w T j = w
Eq. 7-166
and
T j z = z
Eq. 7-167
where in above equations the size of the reduced tridiagonal matrix is j times p , assuming
a fixed block size p for now. The eigenvalues of the tridiagonal problem (the so-called
Ritz values) are approximations of the eigenvalues of the mathematical problem stated
in Eq. 7-13. The approximations to the eigenvectors of the original problem are calculated
from the eigenvectors of the tridiagonal problem (Ritz vectors) by:
y = Pj w
Eq. 7-168
x = Qj z
Eq. 7-169
and
where j, Q j are the matrices containing the first j Lanczos blocks of vectors and w, z are
the left and right eigenvectors of the tridiagonal problem. Finally, x, y are the right and left
approximated eigenvectors of the mathematical problem.
189
190
A beautiful aspect of the Lanczos method (exploited also in the READ module) is that the
error norm of the original problem may be calculated from the tridiagonal solution,
without calculating the eigenvectors. Let us introduce a rectangular matrix E j having an
identity matrix as bottom square. Using this, a residual vector for the left-handed solution
is:
s
= y A y
= ( w E j )B j + 1 P j + 1
Eq. 7-170
which means that only the bottom p (if the current block size is p ) terms of the new Ritz
vector w are required due to the structure of E j . Similarly for the right-handed vectors:
H
r = Ax x = Q j + 1 C j + 1 ( E j z )
Eq. 7-171
An easy acceptance criterion (an extension of the one used in the real case) may be based
on the norm of the above residual vectors as:
min ( s
H
2
r 2 ) con
Eq. 7-172
where the con value to accept convergence is either user given or related to an
automatically calculated machine epsilon.
Based on a detailed error analysis of these quantities, we modify this criterion by
considering a gap:
gap ( , T j ) = min i
Eq. 7-173
where i . With this the recommended criterion is

H
min ( s
s 2 r 2
- con
2 r 2 ) -------------------------------gap ( , T j )
Eq. 7-174
The . 2 denotes the Euclidean norm.

In the above, we assumed that the Lanczos blocks have a uniform size of p . It is possible to
generalize this to allow for the j-th iteration to have pj variable block size. Such flexibility
may be advantageous in the case of clustered eigenvalues or to avoid the breakdown of the
Lanczos process.
Let us assume at the ( j + 1 ) -st iteration, the block size is increased by k and the ( j + 1 ) -st
Lanczos vectors are augmented as:
[ Pj + 1 Pj + 1 ]
Eq. 7-175
[ Qj + 1 Qj + 1 ]
Eq. 7-176
and
CHAPTER 7
where the * vectors are the yet undefined augmentations. It is easy to see that appropriate
augmentations will maintain the validity of the three member recurrence of Eq. 7-164 and
Eq. 7-165 as follows:
H
[ Bj + 1 0 ]
Pj + 1
= Pj A Aj Pj Cj Pj 1
H
Pj + 1
Eq. 7-177
and
[ Qj + 1 Qj + 1 ]
Cj + 1
0
= AQ j Q j A j Q j 1 B j
Eq. 7-178
The Lanczos process can therefore formally continue with:

Bj + 1 = [ Bj + 1 0 ]
Cj + 1 =
Cj + 1
0
Eq. 7-179
Eq. 7-180
and
Pj + 1 = [ Pj + 1 Pj + 1 ]
Eq. 7-181
[ Qj + 1 Qj + 1 ]
Eq. 7-182
The conditions of successful continuation with augmented blocks are the orthogonality
requirements of:
H
P j + 1 Qj = 0
Eq. 7-183
and
H
Pj Qj + 1 = 0
Eq. 7-184
It is of course necessary that the newly created, augmented pair of Lanczos vector blocks
inner product
H
Pj Qj + 1
is not singular, since its decomposition will be needed by the algorithm. Specifically, we
need the smallest singular values of the inner product matrix to be larger than a certain
small number. A possible choice for the augmentations is to have k pairs of random vectors
and orthogonalize them against the earlier vectors by using a modified Gram-Schmidt
procedure. The orthogonalization may be repeated several times to assure that the smallest
value is above the threshold.
191
192
The most prevalent usage of the block size adaption is to cover existing clusters of
eigenvalues. In the real block implementation, we were not able to change the block size
on the fly (adapt). Therefore, an estimate of the largest possible cluster was needed a priori.
For typical structural applications, the default block size of 7 (anticipating 6 as the highest
multiplicity) was used.
The multiplicity of a cluster may be found on the fly with the help of the Ritz values. The
number of expected multiplicities in a cluster is the number of elements of the set
satisfying:
i k clu max ( i , k )
Eq. 7-185
where clu is the user specified cluster threshold. The order of the largest cluster of the Ritz
values is calculated every time a convergence test is made and the block size is
appropriately adjusted. This procedure is not very expensive, since it is done on the
tridiagonal problem.
Preventing Breakdown. It is easy to see that the Lanczos process breaks down in some
circumstances. These are:
Either R j or S j or both are rank deficient.

H
Neither are rank deficient, but R j S j is rank deficient.

The breakdown of the first kind prevents the execution of the QR decomposition of the jth blocks. This is fairly easy to overcome by an orthogonalization procedure also applied
in the current complex Lanczos implementation.
Specifically, if S j is deficient, then we restart the Lanczos process with a random Q j + 1
made orthogonal to all previous left Lanczos vectors P j as:
H
Pj Qj + 1 = 0
Eq. 7-186
If Sj is just nearly rank deficient (detected by the QR decomposition of j = Q j + 1 C j + 1),

then we reorthogonalize this Q j + 1 to the previous left Lanczos vectors, as shown in the
above equation.
Rank deficiency (full or near) of R j is treated similarly with respect to the right Lanczos
vectors.
The breakdown of the second type (a serious breakdown) manifests itself in the singular
H
value decomposition of P j + 1 Q j + 1 . In this type of breakdown, some or all of the singular
values are zero, as follows:
H
H
Pj + 1 Qj + 1 = Uj 0 Vj
0 0
Eq. 7-187
where is nonsingular if it exists. Using the augmentation techniques shown earlier, this
problem may also be overcome. First, calculate and partition as follows:
CHAPTER 7
Pj + 1 Uj =
P( 1 ) P( 2 )
Eq. 7-188
Qj + 1 Vj =
Q( 1 ) Q( 2 )
Eq. 7-189
and
where the number of columns in the second partition is the number of zero singular values.
Create the following projector matrix:
H
j = Qj Pj
Eq. 7-190
Bai et al., 1996 proves that the vectors of:

P( 2 ) = ( I j )
Q( 2 )
Eq. 7-191
P( 2 )
Eq. 7-192
and
Q( 2 ) = ( I j )
in the following augmentation:

Pj + 1 =
P( 1 ) P( 2 ) P( 2 )
Eq. 7-193
Qj + 1 =
Q( 1 ) Q( 2 ) Q( 2 )
Eq. 7-194
and
will result in an always nonsingular P j + 1 Q j + 1 product.

It is possible to extend this procedure to the case of near-breakdown when the singular
values may not be exact zeroes, but smaller than desired. In this case, it is recommended to
increase the block size for those singular values that are below a specified threshold.
Finally, one may only use random vectors instead of the projection matrix.
Maintaining Biorthonormality. The maintenance of the biorthonormality of the P j and
Q j vectors is the cornerstone of the Lanczos algorithm. Local biorthonormality, that is,
maintaining the condition between the consecutive Lanczos vectors, is fairly simple by
executing the following steps:
H
Eq. 7-195
Eq. 7-196
Rj = Rj Pj ( Qj R j )
Sj = Sj Qj ( Sj Rj )
193
194
These steps use data only available in memory, therefore they are cost effective even when
executed repeatedly. Unfortunately, this method does not ensure that multiple
eigenvectors will not reoccur. This may be prevented by a full reorthonormalization
scheme using a modified Gram-Schmidt process. It is implemented that way in the single
vector complex Lanczos method of MSC.Nastran. A measure of the orthogonality of the
current Lanczos vectors, with respect to the already accepted eigenvector, is:
H
PH
Qj Pj + 1 1
j Qj + 1 1
d j + 1 = max ---------------------------------------, ---------------------------------------

P j 1 P j + 1 1 Q j 1 P j + 1 1
Eq. 7-197
where . 1 is the matrix column norm. This quantity is usually compared to a machine
computational accuracy indicator as:
dj + 1
mac
Eq. 7-198
where mac is the automatically calculated machine epsilon. The appropriateness of the
choice of the square root was proven in the real Lanczos method (called partial
orthogonality there) and was already successfully employed in the READ module.
That measure, however, is very expensive, both in CPU and I/O regards. Specifically, the
numerator requires the retrieval of the P j and Q j vector blocks from secondary storage and
a multiplication by them. A method of maintaining partial orthogonality with a limited
access of the P j and Q j matrices, uses
Xj + 1
Yj + 1
1
inf
d j + 1 = max --------------------------------------- , ---------------------------------------
Pj 1 Qj + 1
Qj 1 Pj + 1
1
Eq. 7-199
where . inf is now the matrix row norm. The norms of the denominator may be updated
in every iteration step without retrieving the eigenvectors. This method is calculating the
numerator terms of
Xj + 1 = Pj Qj + 1
Eq. 7-200
Yj + 1 = Pj + 1 Qj
Eq. 7-201
and
utilizing the fact that these also satisfy the three term recurrence equations perturbed by
rounding errors as follows:
Xj + 1 = Tj
Xj
0
Xj
0
Aj
Xj 1
l
W1
Bj +
0
l
Bj + 1 W2
Eq. 7-202
CHAPTER 7
where
l
Eq. 7-203
Eq. 7-204
W1 = Pj Qj + 1
and
W2 = Pj + 1 Qj
The superscript l refers to the left side. Similarly, to the right side:
Yj + 1 =
r +
Xj 0 Tj Aj Yj 0 Cj Y
j 1 W1
0
r
W2
Cj + 1
Eq. 7-205
where
r
Eq. 7-206
Eq. 7-207
W1 = Pj Qj 1
and
W2 = Pj + 1 Qj
with superscript r referring to the right side. After these steps, the matrices are perturbed
to simulate the round-off effect as:
1
X j + 1 = ( X j + 1 + F j )C j + 1
Eq. 7-208
and
1
Yj + 1 = Bj + 1 ( Yj + 1 + Fj )
Eq. 7-209
where F j is a random matrix having a norm of mac . If the test

dj + 1
mac
Eq. 7-210
fails, then a retroactive modified Gram-Schmidt procedure (similar to the local one above)
is needed. The cost of this procedure is O ( n ) , the dominant cost being the inner products
of the Lanczos blocks producing the W vector blocks. This assumes that the block structure
of T j is taken into consideration.
Mathematical Algorithm. The simplified mathematical algorithm not containing all the
procedures developed in the preceding sections follows:
195
196
1. Initialization
H
a. Choose starting vector blocks P 1, Q 1 such that P 1 , Q 1 = I

H H
b. Calculate R 1 = ( P 1 A ) and S1 = AQ 1
2. Iteration for j = 1, 2, ,
a. Compute recurrence:
H
= Pj Sj
= R j Pj A j
= Sj Q j A j
b. QR decompose:
H
j+1
j
= P j + 1 Bj +
= Qj + 1 Cj + 1
c. SV decompose:
H
j Qj + 1
= Uj j Vj
d. Compute recurrence:
1
j+1
= B j + 1 U j
j
j+1
= j
12
Vj Cj +
1
+1
= Pj + 1 Uj j
j+1
= Qj + 1 Vj j
e. Start recurrence:
H
+1
= ( P j + 1 A Cj + 1 Pj )
+1
= AQ j + 1 Q j B j +
Figure 7-3
Block Lanczos Logic
Singular Value Decomposition (SVD)

Since a crucial element of the algorithm in Figure 7-3 is the singular value decomposition,
(c), this section gives more detail.
Given any n x matrix A , ( n m ) there exist unitary matrices U and V such that
A = UV H
CHAPTER 7
where has the same order as A , and has the form

=
and is a diagonal matrix with real diagonal entries 1 2 m 0 . The diagonal

entries of are the singular values of A , and are mathematically defined to be the positive
square roots of the eigenvalues of A H A .
The singular value decomposition of A is computed without forming A H A . The first step
is to compute unitary matrices U 1 and V 1 such that
B = U 1H AV 1
is bidiagonal.
The main step of the SVD algorithm implicitly performs the Francis QR iteration on the
matrix B H B .
be the matrix consisting of the
It is also possible to define an "economy" size SVD. Let U
first m columns of U . Then
V H
A = U
is an alternate form of the SVD.
The method is also available directly to the user as shown in the next section.
197
198
7.3

Available methods in MSC.Nastran are the Hessenberg methods and the complex Lanczos
methods. The Hessenberg method is a reduction method, while the Lanczos method is an
iterative method. The characteristics of these methods are:
Method
Type
Identifier
Application
Restriction
Hessenberg
Reduction
HESS
All roots, few

vectors
M 1 0
QZ Hessenberg
Reduction
QZHESS
All roots, few

vectors
None
Complex
Lanczos
Iterative
CLAN
Few roots
[ K ] + s [ B ] + s [ M ] 0
SVD
Reduction
SVD
Singular value
and/or
vectors
B, M must be purged
CHAPTER 7
7.4
User Interface
CEAD
KXX,BXX,MXX,DYNAMIC,CASECC,VDXC,VDXR/
CPHX,CLAMA,OCEIG,LCPHX,CLAMMAT/
S,N,NEIGV/UNUSED2/SID/METH/EPS/ND1/ALPHAJ/OMEGAJ/
MAXBLK/IBLK/KSTEP $
Input Data Blocks:

KXX
Stiffness matrix.
BXX
Viscous damping matrix.
MXX
Mass matrix.
DYNAMIC Table of Bulk Data entry images related to dynamics.

CASECC
Table of Case Control command images.
VDXC
Partitioning vector with 1.0 at rows corresponding to null columns in K, B,

and M.
VDXR
Partitioning vector with 1.0 at rows corresponding to null rows in K, B, and

M.
Output Data Blocks:

CPHX
Complex eigenvector matrix.
CLAMA
Complex eigenvalue summary table.
OCEIG
Complex eigenvalue extraction report.
LCPHX
Left-handed complex eigenvector matrix.
CLAMMAT Diagonal matrix with complex eigenvalues on the diagonal.

Parameters:
NEIGV
Output-integer-no default. NEIGV indicates the number of eigenvalues

found. If none were found, NEIGV is set to -1.
UNUSED2
Input-integer-default=1. Unused.
SID
Input-integer-default=0. Alternate set identification number.

If SID=0, the set identification number is obtained from the CMETHOD
command in CASECC and used to select the EIGC entry in DYNAMIC.
If SID>0, then the CMETHOD command is ignored and the EIGC entry is
selected by this parameter value. Applicable for all methods
If SID<0, then both the CMETHOD command and all EIGC entries are
ignored and the subsequent parameter values (E, ND1, etc.) will be used to
control the eigenvalue extraction. Applicable for single vector Lanczos,
block Lanczos, QZ Hessenberg, QR Hessenberg, and SVD (Singular Value
Decomposition).
199
200
METH
Input-character-default='CLAN'. If SID<0, then METH specifies the method

of eigenvalue extraction
CLAN
Complex Lanczos (block or single vector),
HESS
QZ Hessenberg or QR Hessenberg,
SVD
Singular Value Decomposition.
EPS
Input-real-default=1.E-5. Used only when SID<0.
ND1
Input-integer-default=0. The number of desired eigenvectors. Used only

when SID<0.
ALPHAJ
Input-real-default=0.0. Real part of shift point for the Lanczos method. Used
only when SID<0.
OMEGAJ
Input-real-default=0.0. Imaginary part of shift point for the Lanczos method.

Used only when SID<0.
MAXBLK
Input-real-default=0.0. Maximum block size. Used only when SID<0.
IBLK
Input-real-default=0.0. Initial block size. Used only when SID<0.
KSTEP
Input-real-default=0.0. Frequency of solve. Used only when SID<0.
CHAPTER 7
7.5
Method Selection
Complex Eigenvalue Extraction Data
EIGC
Defines data needed to perform complex eigenvalue analysis.

Format:
1
EIGC
SID
METHOD
NORM
ND0
10
The following continuation is repeated for each desired search region. (J = 1 to n, where n
is the number of search regions.)
ALPHAAJ
OMEGAAJ
MBLKSZ
IBLKSZ
KSTEPS
NDJ
Examples:
EIGC
EIGC
14
15
CLAN
+5.6
-5.5
HESS
Field
Contents
SID
Set identification number. (Unique Integer > 0) Referred to by

CMETHOD = SID.
METHOD
Method of complex eigenvalue extraction. (Character: "HESS" or "CLAN")
NORM
Method for normalizing eigenvectors. See Option Selection on page 204.
ND0,J
Number of eigenvalues desired.
Grid or scalar point identification number. Required if and only if

NORM = "POINT". (Integer > 0)
Component number. Required if and only if NORM="POINT" and G is a

geometric grid point. (0 < Integer < 6)
Convergence criterion. (Real > 0.0. Default values are: 10-15 for
METHOD = "HESS", E is machine dependent for METHOD = "CLAN".)
MBLKSZ
Maximum block size. (Default = 7, Real > 0.0) Block Lanczos only.
IBLKSZ
Initial block size. (Default = 2, Real > 0.0) Block Lanczos only.
KSTEPS
Frequency of solve. (Default = 5, Integer > 0) Block Lanczos only.
201
202
Comments:
1. The EIGC entry may or may not require continuations as noted below.
For the "HESS" method, continuations are not required; and their contents
are ignored when present, except for ND0. However, it is recommended
that continuations are not used.
For the "CLAN" method when the continuation entry is not used a shift is
calculated automatically. When a shift is input on the first continuation
entry it is used as the initial shift. Only one shift is used. Data on other
continuation entries is ignored.
2. The MBLKSZ and IBKLSZ parameters are integers in concept, but must be input
at real numbers (that is, with a decimal point.) They represent maximum sizes,
and may be reduced internally for small size problems.
Alternate EIGC Bulk Data Entry

The following alternate format is valid for all existing methods except inverse power:
1
EIGC
SID
METHOD
NORM
EPS
ND1
10
KEYWORD1=<value> KEYWORD2=<value> KEYWORD3=<value>
where KEYWORD may be any of the parameters from the original entry except SID, as well
as:
NDj
Number of desired roots at shift j. (Integer > 0)
SHIFTRj
The real part of shift j. (Real)
SHIFTIj
The imaginary part of shift j. (Real)
KSTEPSj
Block tridiagonal solution frequency at shift j; (only block Lanczos).

(Integer > 0)
MBLKSZj
Maximum block size at shift j; (only block Lanczos). (Integer > 0)
IBLKSZj
Initial block size at shift j; (only block Lanczos). (Integer > 0)
Note:
In the parameters above, the value of j ranges from 1 to 10.
Examples:
EIGC
CLAN
EPS=1.E-12 ND1=12 SHIFTR1=0. SHIFTI1=2.4E2
EIGC
HESS
ND1=10
CHAPTER 7
EIGC
CLAN
shiftr1=0.0, shifti1=20., nd1=5, iblksz1=2, mblksz1=5
Remarks:
1. The first field of the keyword-driven continuation entry must be blank.
2. If any of the parameters METHOD, NORM, G, C, EPS, or ND1 are specified on
the continuation entry, the corresponding field on the original entry must be
blank.
3. A maximum of 10 shifts may be specified.
203
204
7.6
Option Selection
Complex eigenvalue analysis in MSC.Nastran supports general damping and
normalization options as well as specific Hessenberg and Lanczos options.
Damping Options
The presence of the [ B ] matrix indicates the viscous damping option when the following
equation is solved:
2
[ M + B + K ]u = 0
Eq. 7-211
This problem is transformed into a linear problem that is twice the size of the original
matrices and provides eigenvalue solutions in complex conjugate pairs.
When the [ B ] matrix is not present and damping is introduced via imaginary stiffness
terms (structural damping), then the following problem is solved:
2
[ M + K ]u + 0
Eq. 7-212
In this case the roots are not complex conjugate pairs. The mixed case of having both
viscous and structural damping is also possible.
Normalization Options
The default normalization (and the only method available for the Hessenberg and Lanczos
methods) is MAX. This option normalizes the component (of the eigenvector with the
largest magnitude) to one as the real part and zero as the imaginary part.
The POINT normalization option uses G for a grid and C for a component to set the
component to a value of (1.0, 0.0). This option is not currently available for the complex
eigenvalue methods.
Hessenberg and Lanczos Options

Hessenberg Spill Option. The spill option of the Hessenberg method (with QR) is less
robust than the default (with QZ), but the latter has no spill. Since this option requires a
much smaller amount of memory to run a problem, larger problems can be solved on a
given computer.
Single Vector Lanczos Option. The adaptive block option of the Lanczos method is very
robust and efficient even for large, direct complex eigenvalue analysis jobs and is the
default. The single vector option must be forced.
These options are selected via SYSTEM(108).
CHAPTER 7
SYSTEM(108)
Bit
Decimal
EIGC Entry
Selection
HESS
QZ Hessenberg without spill (default)
HESS
QR Hessenberg with spill
CLAN
Single vector complex Lanczos
CLAN
Adaptive, block complex Lanczos (default)
CLAN
Debug output for both complex Lanczos
256
HESS
Force LR Hess (aka old Hessenberg

without spill)
10
512
HESS
Force QZ Hess
Since the cell is binary, appropriate combinations are also valid. For example,
SYSTEM(108) = 12 is a proper setting for the block method with debug output.
Internal Block Lanczos Options. There are several internal detailed options available
for the block Lanczos method (also in SYSTEM(108)) as shown in the following table:
Option
Action
16
Turn off block size reduction in block Lanczos
32
Turn off block size augmentation in block Lanczos
64
Enforce full orthogonality in block Lanczos
128 Turn off initial vector preprocessing in block Lanczos

1024
Override defaults for small problems in block Lanczos
2048
Output XORTH matrix in place of ROOTS matrix
4096
Turn off block FBS for complex matrices in block Lanczos
8192
Turn off symmetric decomposition in block Lanczos
16384 Turn off real reduction phase (always use complex arithmetic)
32768 Force spill of Lanczos vectors (testing purposes only)
65536 Old semi-algebraic sorting criterion
131072
Force spill during eigenvector computation (testing purposes only)
262144
Turn off autoshift logic
SVD Option. The singular value decomposition of a K matrix is produced if B and M are
purged.
205
206
The SVD method is provided for DMAP applications. If used in SOLs 107 or 110, and mass
or damping terms are present, a user fatal exit is taken. The SVD operation decomposes the
input stiffness matrix K into the factors U, , and V as described in Solution Method
Characteristics on page 198. The ND1 value has a meaning for the SVD functions which
differs from eigensolution.
ND1
Output
>0
All vectors of U and V are output.
=0
U and V are returned in a purged state, that is, only

the singular value matrix is computed..
<0
is returned as a square matrix whose number of

columns is equal to the minimum number of rows or
columns of the input matrix. U and V are truncated
to be commensurate with . This is a method to
reduce the costs of solving very rectangular input
matrices by providing a partial solution for the most
interesting vectors.
CHAPTER 7
7.7
Complex Eigenvalue Diagnostics

Hessenberg Diagnostics
The Hessenberg method has no internal diagnostics when the no spill option is used. The
spill option has the following diagnostics:
NEW HESSENBERG ROUTINE

TIME OF TRANSFORMATION TO HESS FORM: X (REDUCTION TIME)
STARTING QR ITERATION WITH NINC = X
The NINC value is the number of columns that are in memory.
INFINITY NORM = X
This is the maximum term (in magnitude) in the Hessenberg matrix.
FINAL EPSILON =
The E convergence criterion (from the EIGC entry) is adjusted for the problem.
TIME OF FINDING EIGENVALUES=X (QR ITERATION TIME)

VECTOR GENERATION WITH NINC=X
The number of columns of the Hessenberg matrix held in core.
TIME OF VECTOR GENERATION=X

VECTOR REVERSE TRANSFORMATION WITH NINC=X
The number of eigenvectors held in core is NINC.
TIME OF VECTOR TRANSFORMATION = X (MULTIPLICATION BY

HOUSEHOLDER VECTORS).
Hessenberg diagnostics contain both numerical and performance information. The
performance is low if NINC N at any phase where N is the problem size.
The adjusted convergence criterion is
=
N H E
Eq. 7-213
where H is the infinity norm of the Hessenberg matrix and E is the user-given (or
default) convergence criterion.
Complex Lanczos Internal Diagnostics

Here, the newer block method diagnostics is shown.
207
208
Complex Lanczos Diagnostics DIAG 12. The two levels of internal diagnostics of
complex Lanczos are requested via DIAG 12 and SYSTEM(108) = 8. The structure of the
DIAG 12 diagnostics is as follows:
*** USER INFORMATION MESSAGE 6361 - COMPLEX LANCZOS DIAGNOSTICS
THIS DIAGNOSTICS IS REQUESTED BY DIAG 12.
INITIAL PROBLEM SPECIFICATION
DEGREES OF FREEDOM =
XXX
ACCURACY
REQUIRED
XXX
REQUESTED MODES
XXX
NUMBER OF SHIFTS
XXX
DAMPING MODE FLAG =
XXX
SIZE OF WORKSPACE =
XXX
BLOCK SIZE
XXX
STEP SIZE
XXX
The accuracy required is an echo of E on the EIGC entry or the default if E is not used.
6
Default = 10 . The number of shifts equals to the number of the continuation entries. The
damping mode flag is 0 when damping matrix is present; otherwise, it is 1.
CURRENT SHIFT IS AT X,Y
CURRENT BLOCK SIZE IS X
NUMBER OF MODES REQUIRED AT THIS SHIFT IS XX
The most important parts of the debugging diagnostics (requested by SYSTEM(108) = 8)
are as follows:
MATHEMATICAL EIGENVALUE
ESTIMATED ACCURACY
EIGENVALUE #
REAL
IMAGINARY
LEFT
RIGHT
....
...
....
....
....
....
...
....
....
....
This message could appear any number of times. It indicates the end of an internal Lanczos
process.
At the end of the Lanczos run, the following table is printed:
MATHEMATICAL EIGENVALUES
i ( real ) i ( imag )
i = 1, , NR j
Eq. 7-214
Mathematical Solution. This table contains the shifted solutions where NR j is the
reduced size at shift j .
Besides the diagnostics detailed above, additional information is printed for debugging
purposes, such as the dynamic matrix calculation and the eigenvectors of the tridiagonal
form. These outputs are not explained here. This part of the diagnostics may be extensively
long, therefore the user should not use SYSTEM(108) = 8 on large problems.
CHAPTER 7
The acceptance of these approximate roots is documented in the following table, again
requested by DIAG12:
MATHEMATICAL SOLUTION
DIRECT RESIDUALS
REAL
IMAGINARY
LEFT
RIGHT
i ( real )
i ( imag )
yi ( A i I )
yi ( A i I )
When the error value is less than the E convergence criterion set by the user on the EIGC
entry, then the i-th approximate eigenvalue is accepted as a physical eigenvalue. These
accepted eigenvalues are printed in the following table:
PHYSICAL EIGENVALUES
k ( real ) k ( imag )
k = 1, , NF j
Eq. 7-215
where NF j is the number of acceptable roots found at shift j .

The final acceptance is documented by the state equation summary table as follows:
PHYSICAL SOLUTION
STATE EQUATION RESIDUALS
REAL
IMAGINARY
LEFT
i ( real )
i ( imag )
i ( M i + B i + K )
RIGHT
2
( M i + B i + K ) i
Any rejected roots are printed in a similar table preceded by the following message:
THE FOLLOWING ROOTS HAD AN UNACCEPTABLY LARGE
STATE EQUATION RESIDUAL
The accepted physical eigenvalues are also printed in the regular complex eigenvalue
summary output of the CEAD module: CLAMA. Finally, an eigenvalue summary table is
always printed as follows:
EIGENVALUE ANALYSIS SUMMARY (COMPLEX LANCZOS METHOD)
NUMBER OF MODES FOUND
NUMBER OF SHIFTS USED
NUMBER OF DECOMPOSITIONS
NUMBER OF VECTORS IN CORE
Complex Lanczos Messages and Errors

To monitor the process (especially in the case of multiple shifts), the complex Lanczos
method may issue the following messages:
209
210
UWM 5451:
NO ROOTS FOUND AT THIS SHIFT.
UIM 5453:
FEWER ROOTS THAN REQUIRED HAVE BEEN FOUND.
UWM 5452:
NO ROOTS ACCEPTED AT THIS SHIFT.
UIM 5445:
NO ROOTS FOUND AT ALL.
UIM 5444:
ALL ROOTS HAVE BEEN FOUND.
The following message may be repeated up to three times:
UIM 5443:
DYNAMIC MATRIX IS SINGULAR AT THE SHIFT OF X, Y.
The program attempts to perturb the shift point (up to three times) to obtain an acceptable
decomposition.
SWM 6938,*:
BREAKDOWN IN BLOCK LANCZOS METHOD.
This message is given on the various (*) breakdown conditions of the Lanczos process
along with a recommendation.
UFM 5446:
COMPLEX LANCZOS NEEDS X MORE WORDS.
This message could come from various places.
The user must provide more memory.
SIM 6941:
INVARIANT SUBSPACE DETECTED IN BLOCK LANCZOS.
This message indicates the need for augmenting the current block.
SFM 6939.*:
UNABLE TO READ EIGENVECTORS FROM SCRATCH FILE.
USER ACTION: CLEAN UP DEVICE.
These I/O related messages should not occur normally. If they do occur, the user should
clean up the disk, verify the database allocations, etc. (* = 0, 1 or 2)
CHAPTER 7
Block CLAN Performance Analysis
** USER INFORMATION MESSAGE 5403 (CLASD*)
*
BREAKDOWN OF CPU USAGE DURING COMPLEX LANCZOS ITERATIONS:
OPERATION
REPETITIO
NS
SHIFT AND FACTOR
TIMES (SEC): TOTAL

AVERAGE
.8
3.9
MATRIX-VECTOR MULTIPLY & FBS
908
.1
93.2
REORTHOGONALIZATION
574
.1
38.3
74
.5
34.5
13.8
69.2
SOLVE BLOCK TRIDIAGONAL PROBLEM

EIGENVECTORS AND RESIDUALS
** SYSTEM INFORMATION MESSAGE 6940 (CLRRDD)

*
SPILL OCCURRED WHEN CALCULATING LANCZOS
VECTORS.
X OUT OF A TOTAL OF Y LANCZOS VECTORS HAVE BEEN STORED OUT OF
CORE.
USER ACTION: TO PREVENT SPILL, INCREASE OPEN CORE SIZE BY AT LEAST Z
WORDS
** SYSTEM INFORMATION MESSAGE 6940 (CLRVRD)

*
SPILL OCCURRED WHEN CALCULATING PHYSICAL EIGENVECTORS.
USER ACTION: TO PREVENT SPILL, INCREASE MAXIMUM BLOCK SIZE BY AT
LEAST Z
Orthogonality Analysis
Two additional orthogonality criteria (see theory in Theory of Complex Eigenvalue
Analysis on page 161) that can be used for solution testing purposes are the O 1, O 2
matrices.
The matrices created by these criteria should be diagonal. The off-diagonal terms are
x
supposed to be computational zeroes ( 10 ) . The absolute magnitude of the off-diagonal
terms is a good indicator of the correctness of the solution.
211
212
The user can use the following DMAP to calculate these matrices in connection with the
CEAD module:
SMPYAD
LCPHX,K,CPHX,,,/PTKP/3////1 $
MPYAD
LCPHX,EIGENV,/PL $
MPYAD
LCPHX,EIGENV,/PTL $
SMPYAD
PTL,M,PL,,,PTKP/O1/3//-1//1 $
SMPYAD
LCPHX,B,CPHX,,,/PTBP/3////1 $
SMPYAD
LCPHX,M,PL,,,PTBP/SUM2/3////1 $
SMPYAD
PTL,M,PHI,,,SUM2/O2/3//+1//1 $
The EIGENV matrix is a matrix containing the eigenvalues on the diagonal. This matrix can
be created as a conversion of the CLAMA table by
LAMX
,,CLAMA/EIGENV/-2 $
The user can filter out the small terms by using:

MATMOD
O1,O2,,,,/ORTHO1F,ORTHO2F/
2////1.-6 $
6
The filtered matrices should only contain off-diagonal terms greater than 10 . Terms
6
greater than 10 point to those eigenpairs that do not satisfy the orthogonality criterion.
The DMAP statement
MATPRN
ORTHO1F,ORTHO2F// $
can be used to print out the matrices.
CHAPTER 7
7.8
Complex Lanczos Estimates and Requirements

The time estimates for single vector complex Lanczos are detailed below.
Shifting time (sec) is
Td
Eq. 7-216
4 N steps ( N C M + T s )
Eq. 7-217
Recursion time (sec) is
Normalization time (sec) is

2
( 2 IPREC )N des N M
Eq. 7-218
( 2 IPREC )N des N P
Eq. 7-219
Packing time (sec) is
where:
N des = number of modes desired
N steps = number of Lanczos steps
T d = decomposition time (see Decomposition Estimates and Requirements on
page 61 for details)
T s = solution time (see Decomposition Estimates and Requirements on page 61
for details)
C = average front size
Memory:
IPREC) ( 6 MBLKSZ + 8 N MBLKSZ + 8 MBLKSZ MSZ
where:
MBLKSZ = maximum block size
MSZT = maximum size of Tj matrix
N = problem size
The rest of the available memory is used for temporary storage of accepted eigenvectors to
reduce the I/O cost of outer orthogonalization.
213
214
7.9
References
Bai, Z., et al. ABLE: An Adaptive Block Lanczos Method for Non-Hermitian Eigenvalue
Problems. Department of Mathematics, University of Kentucky, 1996.
Cullum, J. K.; Willoughby, R. A. Lanczos Algorithms for Large Symmetric Eigenvalue
Computations. Birkhuser, 1985.
Cullum, J.; Willoughby, R. A. Large Scale Eigenvalue Problems. North-Holland, 1986.
Komzsik, L. Implicit Computational Solution of Generalized Quadratic Eigenvalue
Problems. Journal of Finite Element Analysis and Design, 2000.
Kowalski, T. Extracting a Few Eigenpairs of Symmetric Indefinite Matrix Pairs. Ph.D.
Thesis, University of Kentucky, 2000.
Glossary of Terms
216
AVG
Average.
BUFFER
An area of memory reserved for data transfer between secondary

storage and memory.
BUFFPOOL
A poll of buffers used by the executive system.
BUFFSIZE
Buffersize in words (in machine precision).
Average front size.
CEAD
Complex eigenvalue analysis module.
cell
An element of the SYSTEM common block of MSC.Nastran.
CLAN
Complex Lanczos method identifier.
DECOMP
Matrix decomposition functional module.
Dense
A matrix or vector is dense when it has few or no zero elements.
DIAGs
Diagnostic flags of MSC.Nastran.
DOF(s)
Degree(s)-of-freedom.
EXECUTIVE
The portion of MSC.Nastran controlling the execution of the

program.
FACTOR
Triangular matrix, which is the result of DECOMP.
FBS
Forward-backward substitution functional module.
GINO
General input-output system.
GIV
Givens method identifier.
HESS
Hessenberg method identifier.
HOU
Householder method identifier.
ID
Identification.
IPREC
Machine precision: 1 for short-word machines.

2 for long-word machine.
Kernel
Internal numerical and I/O routines used heavily by functional

modules.
Keyword
A word specific to a particular function or operation in

MSC.Nastran.
LHS
Left-hand side.
M value, unit numerical kernel time in msec.
MAXRATIO
A diagnostics parameter to show ill-conditioning.
MEM
Memory area reserved for memory files.
MPC
Multipoint constraint.
MPYAD
Matrix multiply and add functional module.
Glossary of Terms
Problem size.
NZ
Nonzero words in matrix.
P value, unit data packing time using columns.
PARALLEL
Keyword to specify multiple CPU execution.
Pi
Pi value, unit data packing time using terms.
Ps
Ps value, unit data packing time using strings.
RAM
Random access memory area used by the executive system.
READ
Real eigenvalue analysis module.
RHS
Right-hand side.
RMS
Root mean squared.
SEQP
Sequencing module.
SOLVIT
Iterative equation solution module.
Sparse
A matrix or vector is sparse when it has many zero elements.
SPARSE
Keyword to specify indexed kernel usage.
String
A sequence of consecutive nonzero terms in a matrix column.
STRL
Average string length.
STURM Number
Number of negative terms on the factor diagonal.
Trailer
An information record following (trailing) a matrix, which contains

the main characteristics.
Matrix density.
217
218
MSC.Nastran Numerical Methods Users Group
Bibliography
Babikov, P. & Babikova, M. An Improved Version of Iterative Solvers for Positive Definite
Symmetric Real and Non-Hermitian Symmetric Complex Problems. ASTE, JA-A81,
INTECO 1995.
Bai, Z., et al. ABLE: An Adaptive Block Lanczos Method for Non-Hermitian Eigenvalue
Problems. Department of Mathematics, University of Kentucky, 1996.
Brown, John. Price/Performance Analysis of MSC/NASTRAN. Proc. of the Sixteenth MSC Eur.
Users Conf., Paper No. 17, September, 1989.
Bunch, J. R.; Parlett, B. N. Direct Methods for Solving Symmetric Indefinite Systems of Linear
Equations. Society for Industrial and Applied Mathematics Journal of Numerical
Analysis, Volume 8, 1971.
Caldwell, Steve P.; Wang, B.P. An Improved Approximate Method for Computing Eigenvector
Derivatives in MSC/NASTRAN. 1992 MSC World Users Conf. Proc., Vol. I, Paper No. 22,
May, 1992.
Chatelin, F. Eigenvalues of Matrices. Wiley, 1993.
Chan, T.; Wan, W. L. Analysis of Projection Methods for Solving Linear Systems with Multiple
Right Hand Sides. CAM Report #26, UCLA, 1994.
Chiang, K. N.; Komzsik, L. The Effect of a Lagrange Multiplier Approach in MSC/NASTRAN on
Large Scale Parallel Applications. Comp. Systems in Engineering, Vol 4, #4-6, 1993.
Conca, J.M.G. Computational Assessment of Numerical Solutions. ISNA 92, Praque, 1992
Cullum, J. K.; Willoughby, R. A. Lanczos Algorithms for Large Symmetric Eigenvalue
Computations. Birkhuser, 1985.
220
Cullum, J.; Willoughby, R. A. Large Scale Eigenvalue Problems. North-Holland, 1986.

Duff, I. S.; Reid, J. K. The Multifrontal Solution of Indefinite Sparse Symmetric Linear
Systems. Harwell Report CSS122, England, 1982.
Efrat, I.; Tismenetsky, M. Parallel iterative solvers for oil reservoir models. IBM J. Res.
Dev. 30 (2), 1986.
Francis, J. G. F. The QR Transformation, A Unitary Analogue to the LR Transformation.
The Computer Journal, Volume 4, No. 3, Oct. 1961, and No. 4, Jan. 1962.
George, A.; Liu, J. W. Computer Solutions of Large Sparse Positive Definite Systems.
Prentice Hall, 1981.
Givens, W. Numerical Computation of the Characteristic Values of a Real Symmetric
Matrix. Oak Ridge National Lab., ORNL-1574, 1954.
Goehlich, D.; Komzsik, L. Decomposition of Finite Element Matrices on Parallel
Computers. Proc. of the ASME Int. Computers in Engineering Conf., 1987.
Goehlich, D.; Fulton, R.; Komzsik, L. Application of a Parallel Equation Solver to Static
FEM Problems. Int. J. for Computers and Structures, 1989.
Gockel, M. A. Handbook for Dynamic Analysis. The MacNeal-Schwendler Corp., Los
Angeles, 1983.
Grimes, R. G.; Lewis, J. G.; Simon, H. D.; Komzsik, L.; Scott, D. S. Shifted Block Lanczos
Algorithm in MSC/NASTRAN. MSC/NASTRAN Users Conf. Proc. Paper No. 12,
March, 1985.
Grimes, R. G., et al. A Shifted Block Lanczos Algorithm for Solving Sparse Symmetric
Generalized Eigenproblems. SIAM, J. Mat. Analysis Appl., 13, 1992.
Hageman & Young. Applied Iterative Methods. Academic Press, 1981.
Hendrickson, B., Rothberg, E. Improving the Runtime and Quality of Nested Dissection
Ordering. Silicon Graphics, Inc., Mountain View, CA, April 11, 1996.
Karypis, G., Kumar, V. METIS, A Software Package for Partitioning Unstructured
Graphs, Partitioning Meshes, and Computing Fill-Reducing Orderings of Sparse
Matrices, Version 3.0.3, University of Minnesota, Department of Computer
Sciences/Army HPC Research Center, Minneapolis, MN, November 5, 1997.
(http://www.cs.umn.edu/~karypis)
Komzsik, L. Implicit Computational Solution of Generalized Quadratic Eigenvalue
Problems. Journal of Finite Element Analysis and Design, 2000.
Bibliography
Komzsik, L. New Features of the Lanczos Module in Version 67 of MSC/NASTRAN. Proc.

of the 18th MSC Eur. Users Conf., Paper No. 13, June, 1991, Rome.
Komzsik, L. Optimization of Finite Element Software Systems on Supercomputers.
Supercomputing in Engineering Analysis, ed. Hojjat Adeli, Marcel-Dekker, 1991.
Komzsik, L. Parallel Processing in Finite Element Analysis. Finite Element News,
England, June, 1986.
Komzsik, L. Parallel Static Solution in Finite Element Analysis. The MSC 1987 World
Users Conf. Proc., Vol. I, Paper No. 17, March, 1987.
Komzsik, L; Rose, T. Substructuring in MSC/NASTRAN for Large Scale Parallel
Applications. Computing Systems in Engineering, Vol 2, #1, 1991.
Kowalski, T. Extracting a Few Eigenpairs of Symmetric Indefinite Matrix Pairs. Ph.D.
Thesis, University of Kentucky, 2000.
Lanczos, C. An Iteration Method for the Solution of the Eigenvalue Problem of Linear
Differential and Integral Operators. Journal of the Research of the National Bureau
of Standards., Volume 45, 1950, pp. 255-282.
Lewis, J. G.; Grimes, R. G. Practical Lanczos Algorithms for Solving Structural
Engineering Eigenvalue Problems. Sparse Matrices, and Their Uses, edited by
I. E. Duff, Academic Press, London, 1981.
Levy, R.; Wall, S. Savings in NASTRAN Decomposition Time by Sequencing to Reduce
Active Columns. NASTRAN: Users Exper., pp. 627-632, September, 1971, (NASA
TM X-2378).
MacNeal, R. H.; Komzsik, L. Speeding Up the Lanczos Process. RILEM, Kosice, Slovakia,
1995.
MacNeal, R. H. The NASTRAN Theoretical Manual. The MacNeal-Schwendler Corp., Los
Angeles, 1972.
Manteuffel, T. A. An Incomplete Factorization Technique for Positve Definite Linear
Systems, Math. of Computation, Vol 34, #150, 1980.
McCormick, C.W. Review of NASTRAN Development Relative to Efficiency of Execution.
NASTRAN: Users Experience, pp. 7-28. September, 1973. (NASA TM X-2893)
Mera, A. MSC/NASTRANs Numerical Efficiency for Large Problems on CYBER Versus
Cray Computer. Proc. of the MSC/NASTRAN Eur. Users Conf., June, 1983.
Newmark, N. M. A Method of Computation for Structural Dynamics. Proceedings of the
Americal Society of Civil Engineers, 1959.
Ortega, J. M.; Kaiser, H. F. The LR and QR Methods for Symmetric Tridiagonal Matrices.
The Computer Journal, Volume 6, No. 1, Jan. 1963, pp. 99-101.
Parlett, B. N. The Symmetric Eigenvalue Problem. Prentice Hall, Englewood Cliffs, 1980.
221
222
Petesch, D.; Deuermeyer, D.; Clifford, G. Effects of Large Memory on MSC/NASTRAN

Performance. Proc. of the 18th MSC Eur. Users Conf., Paper No. 12, June, 1991.
Pissanetzsky, S. Sparse Matrix Technology. Academic Press, 1984.
Poschmann, P.; Komzsik, L. Iterative Solution Technique for Finite Element Applications.
Journal of Finite Element Analysis and Design, 19, 1993.
Poschmann, P.; Komzsik, L., Sharapov, I. Preconditioning Techniques for Indefinite Linear
Systems. Journal of Finite Element Analysis and Design, 21, 1997.
Rothberg, E. Ordering Sparse Matrices Using Approximate Minimum Local Fill, Silicon
Graphics, Inc., Mountain View, CA, April 11, 1996.
Saad, Y. Numerical Methods for Large Eigenvalue Problems. Halsted Press, 1992.
Shamsian, S.; Komzsik, L. Sparse Matrix Methods in MSC/NASTRAN. The MSC 1990
World Users Conf. Proc., Vol. II, Paper No. 64, March, 1990.
Sinkiewicz, J. E. Numerical Stability of Fine Mesh Torus Models. Proc. of the
MSC/NASTRAN Users Conf., March, 1979.
Wilkinson, J. H. The Calculation of the Eigenvectors of Codiagonal Matrices. The
Computer Journal, Volume 1, 1958, p. 90.
223
I N D E X
MSC.Nastran
Numerical
Methods Users
Guide
A
accuracy
required, 153
analysis
performance, 18
analytic
data, 15
B
backward
substitution, 64
buckling, 100
BUFFPOOL, 3
BUFFSIZE, 3
C
canonical
form, 100, 102
cells
system, 3
constants
timing, 12
criterion
convergence
Hessenberg, 207
termination
Lanczos, 154
D
damping
options, 204
density, 34
deselection
MPYAD method, 38
DIAG
flags, 6
DISTRIBUTED PARALLEL, 3
E
EIGC Bulk Data entry
specification of, 201
eigenvalues
complex, 160
real, 100
EIGRL
entry, 144
empirical
data, 15
estimates, 44, 61, 155, 213
executive
system, 17
F
filtered
matrix, 212
forward
substitution, 64
G
Givens, 140
H
Hessenberg
method, 198
HICORE, 3
Householder
method, 140
I
identifiers
method, 34
ill-conditioning, 57
incompatible
matrices, 42
Index
224 INDEX
IORATE, 3
iterative, 100
K
kernel
functions, 10
L
Lanczos, 100, 152, 198
left-handed, 64
loop
inner
outer, 13
outer, 13
M
matrix
condition, 57
decomposition, 46
multiplication, 22
trailers, 7
MAXRATIO
parameter, 58
memory estimates, 17
MPYAD
module, 36
N
negative
terms on factor diagonal, 58
normalization, 145, 204
O
orthogonality
test, 211
Index
P
PARALLEL, 3
performance
analysis, 18
pivot
threshold, 56
R
REAL, 3
reduction
method, 100
RITZ
vectors, 153
S
selection
MPYAD method, 39
shifting scale, 153
singularity, 57
SMPYAD
module, 36
space
saver, 145
SPARSE, 3
sparse
decomposition, 58
spill
algorithm, 204
storage
requirements, 17
STURM
number, 58
system cells, 3
T
THRESH, 56
trailer
matrix, 7
U
USPARSE, 3
V
vector
kernels, 10

MSC - Nastran Numerical Methods User's Guide PDF

Uploaded by

Copyright:

Available Formats

MSC - Nastran Numerical Methods User's Guide PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MSC - Nastran Numerical Methods User's Guide PDF

Uploaded by

Copyright:

Available Formats

What is the document about?

What is the document about?

What numerical methods are discussed?

What numerical methods are discussed?

MSC.

by Dr. Louis Komzsik

About this Book, 12

List of MSC.Nastran Books, 14

Permission to Copy and Distribute MSC Documentation, 20

Diagnostic (DIAG) Flags, 6

Theory of Matrix Multiplication, 23

Parallel Multiply Method, 33

DMAP User Interface, 36

MPYAD Estimates and Requirements, 44

Decomposition Estimates and Requirements, 61

Theory of Forward-Backward Substitution, 65

FBS Estimates and Requirements, 72

Iterative Solver Solutions, 76

Theory of the Conjugate Gradient Method, 77

Iterative Method Selection, 87

Iterative Solution Diagnostics, 92

Iterative Solver Estimates and Requirements, 95

Real Eigenvalue Problems, 100

Theory of Real Eigenvalue Analysis, 101

Solution Method Characteristics, 140

DMAP User Interface, 141

Method Selection, 144

Option Selection, 145

Real Symmetric Eigenvalue Diagnostics, 149

Real Lanczos Estimates and Requirements, 155

Damped Models, 160

Theory of Complex Eigenvalue Analysis, 161

Solution Method Characteristics, 198

User Interface, 199

Method Selection, 201

Option Selection, 204

Complex Eigenvalue Diagnostics, 207

Complex Lanczos Internal Diagnostics, 207

Complex Lanczos Estimates and Requirements, 213

MSC.Nastran Numerical Methods Users Guide, 223

MSC.Nastran Numerical Methods Users Guide

About this Book

About this Book

Utility Tools and Functions on page 1

Using This Guide

List of MSC.Nastran Books

Resolution of installation problems

Gouda, The Netherlands

Permission to Copy and Distribute MSC Documentation

All copyright notices must be included on all copies.

Please complete and mail to MSC for approval:

Please do not write below this line.

MSC.Nastran Numerical Methods Users Guide

Utility Tools and Functions

Execution Cells. System cells related to execution types are

NASTRAN SYSTEM (CELL) = value

PUTSYS (value, CELL)

Enable SPMPYAD T and NT

Deselect SMPMYAD T and NT