Protecting People and the EnvironmentUNITED STATES NUCLEAR REGULATORY COMMISSION
ACCESSION #: 9711120091
NOTE: This text document was processed from a scanned
version or an electronic submittal and has been
processed as received. Some tables, figures,
strikeouts, redlines, and enclosures may not have
been included with this submittal, or have been
omitted due to ASCII text conversion limitations. In
order to view this document in its entirety, you may
wish to use the NUDOCS microfiche in addition to the
electronic text.
The Boeing Company
P.O. Box 3707
Seattle, WA 98124-2207
November 3, 1997
G-1151-SJA-97-503
Document Control Desk
United States Nuclear Regulatory Commission
Washington, D.C. 20555
Reference: a) Boeing Letter G-1151-RSO-92-365 dated August 31,
1992; R. S. Orr to the NRC Operations Center
b) NRC Letter Docket No. 99901227 dated August 12, 1992;
L. J. Norrholm to R. S. Orr; Subject: Response to 10
CFR 21 Inquiry
Dear Sir or Madam:
In accordance with the reference correspondence and 10 CFR 21, Boeing is
sending the NRC the attached error notice(s) received from our former
software suppliers. Because of unknown current addresses, the following
former customers were not notified:
Reactor Controls, Inc.
Echo Energy Consultants, Inc.
Nuclear Applications and Systems Analysis Company (Japan)
Nuclear Power Services
Error notices have been sent to our other former customers.
Very truly yours,
Sandra J. Andrews
Nuclear Administrator
Phone: (206) 865-6248
FAX: (206) 865-4851
Mail Stop: 7A-33, or
e-mail: Sandra.Andrews@PSS.Boeing.com
Enclosure(s): ANSYS Class3 Error Reports 97-35 thru 97-49, 97-02R1, 96-
50R1, 96-38R1, ANSYS QA Notice QA-97-02, ANSYS Support
Coordinator Bulletin SCB97-02 and ANSYS Class3 Error
Summary Reports for Rev 5.3
ANSYS, Inc.
Southpointe
275 Technology Drive
Canonsburg, PA 15317
October 24, 1997
Dear Class3 Error Recipient:
Enclosed you will find ANSYS Class3 Error Reports 97-35 through 97-
49 along with 97-02 R1, 96-50 R1, 96-38 R1 and ANSYS QA Notice QA-97-02.
Also attached is ANSYS Support Coordinator Bulletin SCB97-02.
Class3 Error Report 97-02 has been revised to expand the scope of
the error to include nonlinear permeable regions. The originally issued
Class3 Error Report 97-02 erroneously stated "the energy in nonlinear
regions is correct".
Class3 Error Report 96-50 R1 has been issued to include an
additional case of curved element edges and to narrow the scope of the
error for SOLID92 and PLANE2.
Class3 Error Report 96-38 has been revised to include the
undocumenting of the SEPO field on the ASBL command at Release 5.4.
ANSYS QA Notice QA97-02 was issued to inform customers of an error
in the Microsoft Fortran compiler which may result in memory becoming
unusable.
The ANSYS Support Coordinator Bulletin (SCB97-02) contains
information pertaining to ANSYS Release 5.4. It provides information
about known, non-Class3 errors In the ANSYS program and workarounds.
I would like to remind you of the various ways that ANSYS Inc.
provides Class3 error information. Quarterly, Class3 errors are
delivered by mail to the ANSYS Support Coordinator listed on your ANSYS
license agreement. Please forward any personnel or address changes to
your local ANSYS Support Distributor (ASD) promptly to that these reports
can be delivered.
For users desiring access to Class3 errors on a more timely basis
you can be added to our email distribution list. To register for email
notification of reports, simply send an email request including your
email address, company name/address and ANSYS agreement number to
bonny.podolek@ansys.com. If you are a subscriber to email distribution,
please keep us informed of any changes in your email address by emailing
bonny.podolek@ansys.com.
Finally, Class3 Errors and QA Notices are posted on ANSYS's Internet
HomePage. The address is http://www.ansys.com. They are located in the
ANSYS Zones section of the HomePage under Customer. The username to
enter this area is "customer" and the password is "ain1fm" (the acronym
for ANSYS is number 1 for me).
For your convenience, also enclosed you will find Release 5.3 ANSYS
Class3 Error Summary Reports sorted both by error number and keyword and
a description of ANSYS Service Packs which are available on ANSYS's
Internet HomePage.
Sincerely,
William J. Brian
Quality Assurance Manager
ANSYS, Inc.
Southpointe
275 Technology Drive
Canonsburg, PA 15317
ANSYS SERVICE PACKS
In keeping with the tradition of a strong commitment to our customers,
ANSYS, Inc. has developed a strategy to provide timely corrections to
ANSYS software product releases. ANSYS Service Packs (ASPs) are created
when an important product correction is needed to provide improved ANSYS
software product functionality. Currently, ASPs are available for the
second release of ANSYS 5.3 and will be available for subsequent
releases.
Within a typical ANSYS Service Pack, users will find a description of the
product improvement and the method of providing the improvement, which
may be object file(s) relinkable into ANSYS, or they may contain non-
relinkable improvements such as a macro, a UIDL function, or a script.
This method enables delivery of an ANSYS product correction in a timely
manner without having to wait for the next ANSYS software release. In
addition, the ASP may contain input and output files for checking proper
installation of the ASP.
To access the currently available ANSYS Service Packs, users must possess
a valid Technical Enhancement and Customer Support agreement (TECS).
ANSYS, Inc. or an ANSYS Support Distributor (ASD) can provide customers
with a TECS agreement.
To obtain additional information regarding ASPs, access the ANSYS web at
http://www.ansys.com/ and click on the Software Update Manager area found
under the Service, Support link. There, users will find answers to
questions such as "What are ASPs?," or most importantly, "How can I
obtain an ASP?" Additional information located at the Software Update
Manager area includes Version Information for ANSYS 5.3 and subsequent
releases, descriptions of the currently available ASPs, instructions on
how to install and implement an ASP, and how to obtain ASP Technical
Support.
ANSYS Release Identifier Description
ANSYS release identifiers consist of a major release level, a minor
release level, a correction and a build date. An example of how this is
constructed is shown below:
5.3 UP030797 <--- build date (form = UPmmddyy)
^ ^
| |
| |
major release level __ __ minor release level
Major release level changes indicate that new features have been added to
the program and that some level of program architecture change and/or
file structure may have occurred. Minor release level changes also
indicate that new features have been added to the program, but files are
upwardly compatible. All known error fixes are included in both minor
and major releases. The build date corresponds to the date the program
was created. Special versions may be provided to circumvent an error and
are identified by build date. Special versions are not general releases
to all ANSYS licensees, since they typically represent error corrections
occurring only on one system, a subset of our customers who have specific
graphics devices, etc.
When a release identifier on a Class3 Error Report does not include a
build date, all build dates for the indicated release level are included.
When a release identifier for FIRST INCORRECT VERSION explicitly includes
a build date, the release level with the indicated build date as well as
all earlier build dates for that release level are affected. When a
release identifier for CORRECTED IN explicitly includes a build date, the
release level with the indicated build date as well as all subsequent
build dates contain the correction.
For example, a Class3 Error Report with "5.3 UP100396" for FIRST
INCORRECT VERSION and "5.5 UP063098" for "CORRECTED IN" would apply to
all 5.3 releases with a build date UP 100396 or later, all 5.4 releases
(regardless of build date) and all 5.5 releases with a build date earlier
than UP063098.
The ANSYS release identifier(s) shown under "corrected in" on the front
side of Class3 Error Reports indicates the first possible release that
contains the correction.
An identifier indicated under "corrected in" does not guarantee that a
general distribution of that release of ANSYS will occur. It does
indicate that the correction is known and implemented in that or any
subsequent release.
Product Applicability
The ANSYS family of component products occasionally undergoes name
changes between releases and/or changes in the functionality of derived
products (such as ANSYS-PC/LINEAR). To minimize the potential for
confusion in these areas, unless otherwise noted on the front side of the
Class3 Error Report, the error report applies to all ANSYS family
products (including standalone component products) that contain the
described feature(s) in the designated release(s).
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-35
KEYWORDS: SYSTEM CRAY
DESCRIPTION OF ERROR:
Incorrect results may be produced intermittently for any analysis
executed on CRAY T90 computer systems if the system is run with
bidirectional memory (default). According to Cray Research's Field
Notice #2389, the error may occur during "memory intensive" activity such
as ANSYS usage.
TYPICAL GUI PATH(S): N/A
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.0A Release 5.3
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Turn off bidirectional memory by inserting the following command
before the ANSYS command line.
/etc/cpu -m bdmoff
Specifically, edit the /ansys52/bin/ansys52 script, and make the
following change:
Replace: $ANSYS52_DIR/bin/ansys.e52 $*
With: /etc/cpu -m bdmoff $ANSYS52_DIR/bin/ansys.e52 $*
COMMENTS:
According to Cray Research, the likelihood of an error or an abort
to occur correlates with increased system load and memory activity.
Please see Cray Research Field Notice #2389 for a detailed description.
Cray Research Field Notices are available to all Cray customers who have
access to the CRInform system. Field Notice #2389 can also be viewed
online at the following URL:
http://crinform.cray.com/fn/field_notices/fn2389
To date, ANSYS, Inc. has not experienced or had reported to us a case of
incorrect results from this error.
AUTHOR/CORRECTOR: DATE: August 13, 1997
Lou Muccioli
REVIEWED BY QA: DATE: August 13, 1997
William J. Bryan
APPROVAL: DATE: August 13, 1997
John A. Swanson
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-36
KEYWORDS: SOLID87 PCG XG
DESCRIPTION OF ERROR:
When the preconditioned or Jacobi Conjugate Gradient solver
[EQSLV, PCG or EQSLV,JCG] is used with SOLID87, the 3-D 10-Node
Tetrahedral Thermal Solid element, automatic time-stepping based on
response eigenvalues is not invoked, which can cause larger time steps
than necessary to accurately capture transient thermal response.
TYPICAL GUI PATH(S):
Preprocessor>Element Type>Add/Edit/Delete[Add]Thermal Solid[Tet 10 node
87] Solution>New Analysis>Transient
Solution>Analysis Options>Solution Method[Full]>Equation Solver>[Jacobi
Conj Grad/Precondition CG/Iter auto select]
Solution>Load Step Opts - Time/Frequency>Time & Time Step>Automatic Time
Stepping
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Release 5.3 Release 5.4
ANSYS/Thermal Release 5.3 ANSYS/Thermal Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Use the frontal, incomplete Cholesky Conjugate Gradient or JCG out-
of-core solver [EQSLV,FRONT/ICCG/JCGOUT], or use other element types
(SOLID70 or SOLID90) instead.
COMMENTS:
AUTHOR/CORRECTOR: DATE: August 13, 1997
Subhasish Roy Choudhury
REVIEWED BY QA: DATE: August 13, 1997
William J. Bryan
APPROVAL: DATE: August 13, 1997
John A. Swanson
* Unless noted otherwise. this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-37
KEYWORDS: BIRTH AND DEATH ELEMENT THICKNESS NLGEOM LARGE
DEFORMATION
DESCRIPTION OF ERROR:
If an element that is dead (EKILL command) undergoes significant
deformation (greater than 5% strain) with large deformation turned on
(NLGEOM, ON command) and is then made alive (EALIVE command), it is
erroneously born with an updated thickness rather than its original
thickness.
Elements which are affected include SHELL43, SHELL91, SHELL93 as
well as the 2-D solids when used with plane stress:
PLANE2 KEYOPT(3)=0 or 3
PLANE13 KEYOPT(3)=0
PLANE42 KEYOPT(3)=0 or 3
PLANE82 KEYOPT(3)=0 or 3
TYPICAL GUI PATHS:
Main Menu>Solution>Other>Kill Elements
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.0 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
If thickness change is not important, use SHELL143 or SHELL63.
COMMENTS:
AUTHOR/CORRECTOR: DATE: August 18, 1997
David L. Conover
REVIEWED BY QA: DATE: August 18, 1997
William J. Bryan
APPROVAL: DATE: August 18, 1997
John A. Swanson
* Unless noted otherwise. this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-38
KEYWORDS: COUPLING NDELE xCLEAR SUPERELEMENTS P-METHOD
DESCRIPTION OF ERROR:
If coupling is defined (CP command) and any of the nodes in the
coupled set are deleted (NDEL or xCLEAR commands, where x=V, A, L, or K)
and the nodes are subsequently recreated, then the following analyses
which rely on the node/coupling cross-referencing are incorrect:
1. Nonlinear superelement use pass (SE command) with a non-zero imposed
displacement (D command) on one of the superelement nodes that is
part of the coupled set.
2. P-method analysis using either of the 3-D elements SOLID147 or
SOLID148.
TYPICAL GUI PATH(S):
Main Menu>Preprocessor>Coupling/Ceqn
Main Menu>Preprocessor>Delete>Nodes
Main Menu>Preprocessor>Clear
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.0 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Delete and re-define the coupled sets after the nodes are recreated.
COMMENTS:
In the corrected release, coupled sets and constraint equations
which contain a node that is deleted will also be deleted.
AUTHOR/CORRECTOR: DATE: August 18, 1997
David L. Conover
REVIEWED BY QA: DATE: August 18, 1997
William J. Bryan
APPROVAL: DATE: August 18, 1997
John A. Swanson
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-39
KEYWORDS: NONLINEAR ANALYSIS NEQIT,1
DESCRIPTION OF ERROR:
For a nonlinear analysis performed with the maximum number of
equilibrium set to 1 (NEQIT,1), neither is an ERROR/WARNING message
produced nor is the run terminated. The results for such a solution as
well as any subsequent loadstep will likely be incorrect since the ESAV
file is not updated for an unconverged solution.
TYPICAL GUI PATH(S):
Main Menu>Solution>Nonlinear>Equilibrium Iter
Main Menu>Solution>Nonlinear>Criteria to Stop
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.0 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Specify more than 1 for the maximum number of equilibrium iterations
(NEQIT command) for nonlinear analysis.
COMMENTS:
The use of a small number equilibrium iterations for a nonlinear
analysis is a non-typical use of the program and is not recommended since
it will generally produce an unconverged solution.
AUTHOR/CORRECTOR: DATE: September 22, 1997
Yong-Cheng Liu
REVIEWED BY QA: DATE: September 22, 1997
William J. Bryan
APPROVAL: DATE: September 22, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
see the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the
reverse.side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-40
KEYWORDS: SELECT ALLSEL,BELOW
DESCRIPTION OF ERROR:
ALLSEL,BELOW, entity (where entity = ALL [default], VOLU, AREA, LINE
or KP) can select nodes or elements that are not appropriate in the
following cases:
1. An element that is not associated with the selected solid model
entities is erroneously selected if all of its nodes are associated with
the selected solid model entities. Typical examples are elements in
"holes" or elements (non-midside node triangle or tetrahedron) located in
corners that are formed by one or more selected bordering meshed
entities.
2. Nodes that are associated with applicable solid model entities but
are not connected to any element associated with an applicable entity are
erroneously selected. For example, ALLSEL,BELOW, AREA for an area, not
meshed with area elements, which is part of a meshed volume, will
erroneously select the nodes associated with the area.
Attributes or loads subsequently applied to selected sets of nodes
or elements can cause unintended results.
TYPICAL GUI PATH(S):
Utility menu>Select>Entities
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.0 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Check the set of selected nodes and elements produced by the above
noted cases. Use the node and element select capabilities (NSEL and
ESEL) to remove any erroneously selected nodes or elements from the
selected set.
COMMENTS:
AUTHOR/CORRECTOR: DATE: September 22, 1997
Ken Podlaszewski
REVIEWED BY QA: DATE: September 22, 1997
William J. Bryan
APPROVAL: DATE: September 22, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
see the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-41
KEYWORDS: THERMAL FLUID
DESCRIPTION OF ERROR:
Flotran analyses of thermal-fluid problems at high Peclet numbers
may contain energy imbalances and incorrect temperatures. Problems
without flow are unaffected.
The exact range of the Peclet numbers affected is problem dependent.
The user can determine whether the problem exists by examining
information in the jobname.pfl file and performing a heat balance. The
procedure for performing a heat balance is outlined in Chapter 5 of the
Flotran Analysis Guide.
TYPICAL GUI PATH(S): N/A
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.1 Release 5.5
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
The user should verify the heat balance in cases with heat
generation by looking at the difference between energy in and energy out
and comparing it to the volumetric heat generation term (if applicable).
In some cases, users specify temperatures instead of heat sources. In
such cases, the energy flow into the system is from Wall Heat Transfer.
This is checked against the energy flow in/out.
COMMENTS:
AUTHOR/CORRECTOR: DATE: September 15, 1997
Rich Lange
REVIEWED BY QA: DATE: September 15, 1997
William J. Bryan
APPROVAL: DATE: September 15, 1997
Dave Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-42
KEYWORDS: SPECTRUM PSD SPOPT,PSD
DESCRIPTION OF ERROR:
A PSD spectrum analysis [SPOPT,PSD] with closely spaced PSD input
values (less than 0.1) produces incorrect displacement, velocity and
acceleration responses. The magnitude of the error is dependent on model
frequency.
TYPICAL GUI PATH(S):
Main Menu>Solution>Analysis Option>Spectrum>PSD Analysis
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Release 5.3 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
COMMENTS:
AUTHOR/CORRECTOR: DATE: September 15, 1997
Mu-Tsang Chen
REVIEWED BY QA: DATE: September 15, 1997
William J. Bryan
APPROVAL: DATE: September 15, 1997
Dave Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-43
KEYWORDS: EMAG CIRCUIT CIRCU124 ELEM124 RESISTANCE
DESCRIPTION OF ERROR:
A zero (or unspecified) resistance value (first real constant) for
CIRCU124 with KEYOPT(1)=0, is erroneously interpreted as 1 Ohm, leading
to incorrect solutions.
TYPICAL GUI PATH(S):
Preprocessor>Element Type>Add/Edit/Delete
Preprocessor>Real Constants>Add
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Release 5.3 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Use a small value (such as 1.0e-6 Ohm) rather than zero for a
resistor with a small resistance.
COMMENTS:
AUTHOR/CORRECTOR: DATE: October 14, 1997
Jian-She Wang
REVIEWED BY QA: DATE: October 14, 1997
William J. Bryan
APPROVAL: DATE: October 14, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-44
KEYWORDS: SELECT ELEMENT ESEL
DESCRIPTION OF ERROR:
Selecting or unselecting elements by element surface load (ESEL with
Item=SFE) will fail to select (or unselect) any element within the select
criteria whose TYPE number does not equal its MAT number.
For example, if an element was of type 6 and referenced material 3,
this element will not be selected even if it has an applied element
surface load that is within the specified select range. However, if that
element was of type 6 and the referenced material number was also 6, the
element would be selected.
TYPICAL GUI PATH(S):
Utility Menu>Select>Entities
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Release 5.3 Release 5.5
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
COMMENTS:
AUTHOR/CORRECTOR: DATE: October 14, 1997
Christopher W. Aiken
REVIEWED BY QA: DATE: October 14, 1997
William J. Bryan
APPROVAL: DATE: October 14, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-45
KEYWORDS: *AFUN,DEG GUI PICKING ANGULAR FUNCTION
DESCRIPTION OF ERROR:
In the GUI, when creating nodes or keypoints on the working plane by
picking (N,,P or K,,P), if degree units have been specified for angular
functions (*AFUN,DEG), then any angular functions (.e.g. SIN, COS, etc.)
entered directly in the input window to define a coordinate will
erroneously use radian units rather than degree units.
TYPICAL GUI PATH(S):
Utility Menu>Parameters>Angular Units
Preprocessor/Create/Nodes/On Working Plane+
Preprocessor/Create/Keypoints/On Working Plane+
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.1 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Use radian units for angular functions (e.g. SIN, COS, etc.) in such
cases. For cases where degree units are required for input for a
coordinate (cylindrical or spherical coordinate system) use the required
function to define a parameter, then enter the parameter name in the
input window.
COMMENTS:
The affected angular functions are SIN, COS, TAN, ASIN, ACOS, ATAN,
ATAN2, ANGLEK and ANGLEN.
AUTHOR/CORRECTOR: DATE: October 14, 1997
Paul Tallon
REVIEWED BY QA: DATE: October 14, 1997
William J. Bryan
APPROVAL: DATE: October 14, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
see the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers,
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-46
KEYWORDS: CFD FLOTRAN YPLU
DESCRIPTION OF ERROR:
In a FLOTRAN analysis, reported values of turbulence model parameter
YPLU are incorrect for cases where near-wall nodes have been placed
extremely close to the wall, in the laminar sublayer of the turbulent
boundary layer. Normally for best results, near-wall nodes should be
located outside the laminar sublayer, in the range of 30 < YPLU < 1000.
This error only affects reported YPLU values in POST1.
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.1 Release 5.4
ANSYS/FLOTRAN Rev. 5.1
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
COMMENTS:
AUTHOR/CORRECTOR: DATE: October 14, 1997
Mark Troscinski
REVIEWED BY QA: DATE: October 14, 1997
William J. Bryan
APPROVAL: DATE: October 14, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-47
KEYWORDS: CSWPLANE *GET CDWRITE ANGLES
DESCRIPTION OF ERROR:
Any local coordinate system defined by the CSWPLA command has the
following errors:
1. Values retrieved by the *GET or *VGET commands for any orientation
angles of the coordinate system will always be zero.
2. CDWRITE does not write the correct orientation angles onto File.CDB.
TYPICAL GUI PATH(S):
Utility Menu>WorkPlane>Local Coordinate Systems>Create Local CS
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.0 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Use the LOCAL or CLOCAL command to define the coordinate system
instead.
COMMENTS:
1. The CSLIST command will indicate the erroneous orientation of a
coordinate system restored from a File.CDB.
2. Any nodes that were rotated into the coordinate system in question
prior to the CDWRITE will have correct rotations on the CDWRITE file
(File.CDB).
3. A similar Class3 Error Report (#93-92) was issued for the CS and
CSKP commands.
AUTHOR/CORRECTOR: DATE: October 14, 1997
David L. Conover
REVIEWED BY QA: DATE: October 14, 1997
William J. Bryan
APPROVAL: DATE: October 14, 1997
Mark C. Imgrund
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-48
KEYWORDS: SHELL181 HYPERELASTIC
DESCRIPTION OF ERROR:
Using the five (5) parameter Mooney-Rivlin model (TB,HYPER) with
SHELL181, the finite strain shell element, produces incorrect results;
the code wrongly assumes the 3rd, 4th and 5th parameters to be equal to
zero. Both 2 and 9 parameter models produce correct results.
TYPICAL GUI PATH(S):
Main Menu>Preprocessor>Element Type>Add/Edit/Delete [Add][Shell][Hyper 4
node 181]
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Release 5.3 Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
An equivalent 5 parameter input would be to specify a 9 parameter
model with at least one of the parameters 6 through 9 being a very small
number. This will result in the 9 parameter model being invoked. The
specification of a nonzero parameter (6 through 9) should be such that it
does not adversely affect the results of the 5 parameter model. This
workaround is very limited and we do not recommend its use without
sufficient verification against established results.
COMMENTS:
AUTHOR/CORRECTOR: DATE: October 14, 1997
Sekar M. Govindarajan
REVIEWED BY QA: DATE: October 14, 1997
William J. Bryan
APPROVAL: DATE: October 14, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-49
KEYWORDS: POST1 PLESOL SHELL /GRAPHICS,FULL
SHELL ELEMENTS43, 63, 91, 93, 99, 143, 150, 162, 163, 181
DESCRIPTION OF ERROR:
In POST1, the PLESOL command will display incorrect stress, strain,
gradient and flux results for she]] elements (43, 63, 91, 93, 99, 143,
150, 162, 163, 181) when the MID or BOT location is specified by the
SHELL command and full graphics displays (/GRAPHICS,FULL) are requested.
TYPICAL GUI PATH(S):
Main Menu>General Postproc>Plot Results>Element Solu
Utility Menu>Plot>ResultsElement Solu
FIRST INCORRECT VERSION(S):* CORRECTED IN:
Release 5.4 Release 5.5
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
Use the PLNSOL command or use PowerGraphics (/GRAPHICS,POWER) to
display the results.
COMMENTS:
The results at she]] location TOP display correctly. TOP is the
default location if no she]] location is specified.
AUTHOR/CORRECTOR: DATE: October 14, 1997
Herb Mroz
REVIEWED BY QA: DATE: October 14, 1997
William J. Bryan
APPROVAL: DATE: October 14, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 97-02 R1
KEYWORDS: MAGNETICS SENERGY MACRO
DESCRIPTION OF ERROR:
Stored magnetic energy calculated with the SENERGY command macro
will be incorrect for the following cases:
1. Permanent magnet elements with an assigned element coordinate
system (ESYS command) that is other than the global Cartesian
system.
or 2. Three dimensional problems using the Magnetic Scalar Potential
formulation (element types SOLID96, SOLID5, SOLID98) with
current sources (element type SOURC36 or currents from coupled-
field element types SOLID5, LINK68, SOLID69 or SOLID98).
Energy from the current source (H-S field) is neglected.
or 3. Nonlinear permeable regions (defined by a B-H curve) [TB
command].
TYPICAL GUI PATH(S):
SENERGY: Main Menu > General PostProc > Elec & mag Calc
ESYS: Main Menu > Preprocessor > Attributes > Default Attribs
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.2 Release 5.4
ANSYS/EMAG Rev. 5.2
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
For Item 1: Define permanent magnet elements with the default global
Cartesian element coordinate system. Revise the coercive force
components (MP command, MGXX, MGYY, MGZZ material properties) to align
with the global Cartesian coordinate system.
MP: Main Menu > Preprocessor > Material Props > Isotropic
COMMENTS:
The integration limits for permanent magnet energy will be changed
(see Theoretical Manual, equation 19.6-18). The zero energy level is
presently defined at a value of B = 0. This will be changed in the
corrected version to a zero energy level defined at B =Br, where Br is
the residual induction of the permanent magnet. The new integration
limit level is more in line with industry standards. In addition,
absolute signs on the energy and co-energy calculations will be removed
(see equations 19.6-18, 19.6-19).
ANSYS users may obtain a corrected version of the macro for use in
ANSYS 5.3 by installing ANSYS Service Pack No. 2. For information
obtaining an ANSYS Service Pack, review the ANSYS HomePage under Service
Support (http://www.ansys.com/Enhancement).
AUTHOR/CORRECTOR: DATE: September 15, 1997
Dale Ostergaard
REVIEWED BY QA: DATE: September 15, 1997
William J. Bryan
APPROVAL: DATE: September 15, 1997
Dave Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 96-50 R1
KEYWORDS: SOLID92 PLANE2 SOLID87 CURVED EDGES
DESCRIPTION OF ERROR:
[Redline: SOLID92 and PLANE2 will produce erroneously high stress
concentrations for elements with curved edges and geometrically linear
analysis and SOLID87 may produce erroneous high fluxes across element
faces whose edges are curved if:]
[Redline: 1. curved edge lengths are of the order 1.0e-3 or less.]
[Redline: or 2. The mid node on the curved edge is located away from
the mid point location of the straight line joining
the end nodes of the curved edge by a distance of
1.0e-5 or less times the length of the straight
line.]
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Release 5.3 Release 5.4
Component Products Release 5.3 Component Products Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
1. Set the undocumented KEYOPT(8)=l for SOLID92, PLANE2, or SOLID87
before solution. This will force an alternative form of element
formulation producing the correct stresses and fluxes.
2. Use the straight-sided meshing option.
COMMENTS:
1. Straight edges give correct results.
2. This error over predicts the stress or flux.
3. Primary solution, i.e. the displacements and temperatures are
correct.
4. Unreasonably high stress concentrations or fluxes are seen when this
error occurs.
AUTHOR/CORRECTOR: DATE: September 15, 1997
Pakal Rahulkumar
REVIEWED BY QA: DATE: September 15, 1997
William J. Bryan
APPROVAL: DATE: September 15, 1997
Dave Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] CLASS3 ERROR REPORT
ERROR NO: 96-38 R1
KEYWORDS: SOLID MODELING BOOLEAN SUBTRACT ASBL
DESCRIPTION OF ERROR:
The "SEPO" field of the solid modeling boolean command "ASBL" does
not always produce separate areas. It may produce areas with separate
lines which share common keypoints.
In most cases this error is obvious. However, there may be rare
instances where the error could lead to unanticipated results if
validation of boundary conditions and loads is bypassed before solution
by the user.
FIRST INCORRECT VERSION(S):* CORRECTED IN:*
Rev. 5.1 Release 5.4
Component Products Rev. 5.1 Component Products Release 5.4
SUGGESTED USER ACTION FOR RUNNING ON UNCORRECTED VERSION:
In Release 5.3 or later, generate separate areas from unseparated
areas by using the "AGEN" command after the "ASBL" command.
COMMENTS:
[Redline: ** Starting in Revision 5.2. a warning is issued for any use o
the SEPO field in the ASBL command stating that separate coincident
keypoints may not be produced for areas divided by curves. At 5.4 the
SEPO field is no longer documented and has been removed from the ASBL
command in GUI. Any use of the undocumented SEPO field on the ASBL
command will produce the above noted warning.]
AUTHOR/CORRECTOR: DATE: October 14, 1997
Ken Podlaszewski
REVIEWED BY QA: DATE: October 14, 1997
William J. Bryan
APPROVAL: DATE: October 14, 1997
David L. Conover
* Unless noted otherwise, this report applies to all ANSYS family
products which contain the described feature in the indicated Release(s).
See the reverse side for details regarding product applicability.
Unless noted otherwise, this error report also applies to all releases
after the first incorrect one and prior to the corrected release. All
releases after "corrected in" are corrected. Manual corrections are
included in on-line documentation as appropriate. Please see the reverse
side of this sheet for additional information on ANSYS release
identifiers.
ANSYS [Registered copy] QA NOTICE
NOTICE NO: QA97-02
SUBJECT: MEMORY LEAK PC NT 95
DESCRIPTION:
Due to an error in the Microsoft Fortran compiler used for the ANSYS
versions and releases noted below, each time a Fortran formatted read or
write is performed by the ANSYS program a small amount of memory is made
unusable. This unusable memory accumulates and is not freed up until the
ANSYS program is exited. This memory leak does not affect analysis
results and is not noticeable in most cases. However in some analyses,
the small memory leaks can accumulate to a point where system problems
occur due to a lack of available free memory. For cases where
accumulated memory losses cause problems, symptoms may include:
- the system complains that it is running out of resources and
suggests the user shut down some applications
- ANSYS appears to hang, or aborts
- other running applications may be affected
This compiler bug only manifests itself when a very large amount of
ASCII reads/writes are being performed by the ANSYS program. Reading or
writing of binary files (db, emat, rst, etc.) does not appear to produce
leaks. For most analyses, the number of ASCII reads or writes is not
significant enough to make the memory problem noticeable. However, the
looping capability provided in the ANSYS APDL macro language can cause a
large number of applicable file reads to occur and may result in a
significant amount of memory becoming unusable.
AFFECTED VERSIONS: ANSYS Revisions 5.2 - 5.4 for Windows 95
ANSYS Releases 5.3 - 5.4 for Windows NT
(Note: ANSYS for DEC Alpha is NOT affected)
SUGGESTED USER ACTION:
For cases where the amount of memory made unusable causes a problem,
the following practices may reduce the amount of memory affected:
- In APDL macros, replace all *GO,:label and
*IF,var,oper,var,place,num type branching with *DO ... *ENDDO
loops
- run the analysis using several separate ANSYS sessions (i.e.
run pre-processing, solution load steps and post-processing
each in a separate ANSYS session)
- exit and restart ANSYS between CDWRITEs and CDREADs of very
large models
Page 2 of 2 ANSYS QA NOTICE
NOTICE NO: QA97-02
SUGGESTED USER ACTION: (continued)
- after IGES import of very large models; save, exit, restart
ANSYS, and resume the model from the binary .db file
COMMENTS:
There is currently no information available as to when a correction
for this compiler error will be available.
AUTHOR: DATE: August 18, 1997
Rick Reynolds
REVIEWED BY QA: DATE: August 18, 1997
William J. Bryan
APPROVAL: DATE: August 18, 1997
John A. Swanson
SCB97-02
ANSYS Support Coordinator Bulletin
This Support Coordinator Bulletin contains information pertaining to
ANSYS Release 5.4. It provides information about known, non-Class3
errors in the ANSYS program and information regarding workarounds.
SYSTEM
KEYWORDS: IBM RS590 RUBY or GTX1000 GRAPHICS BOARD
DESCRIPTION: On the IBM RS590 machine running under AIX 4.1.4 with the
RUBY or GTX1000 Graphics Board installed, dynamic rotations in X11 mode
with the triad present will leave spurious dots on the screen.
AFFECTED VERSIONS: Release 5.3, Release 5.4 and possibly Revisions 5.1 &
5.2
WORKAROUND: Currently none. IBM is looking into the problem and when
they have developed a patch or fix we will place this information on our
web site and make the patch available to the affected users.
KEYWORDS: HP IBM R/S6000 ANS_ADMIN
DESCRIPTION: If you are uninstalling the ANSYS 5.4 product on the HIP or
IBM R/S6000 platform and using the GUI version of ANS_ADMIN, the process
may hang.
AFFECTED VERSIONS: Release 5.4
WORKAROUND: If you encounter this situation, you may kill the ANS-ADMIN
processes from another window and continue. It is recommended that you
use the non-GUI version of ANS-ADMIN to uninstall the product(s).
KEYWORDS: IBM R/S6000 AIX 4.1.5 LARGE MEMORY INSTALLATION GUIDES
DESCRIPTION: The 2 PTFs for the AIX 4.1.5 operating system listed on page
1-14 of the ANSYS 5.4 Installation Guide are not sufficient to allow a
large memory request (e.g., -m of 1800 or more).
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: It is required that AIX 4.1.5 GOLD be installed. An upgrade
to AIX 4.1.5 via the maintenance package and then application of the
latest (September 1997) fixes will upgrade the O.S. to the 4.1.5.X level
but will not correct the problem. You will not get the native AIX 4.1.5
software enhancements nor will you get native AIX 4.1.5 hardware support.
The error that existed for memory allocation in AIX 4.1.4 was corrected
as a native AIX 4.1.5 software enhancement. Only by installing the AIX
4.1.5 OS from the install media (migrate, preservation or complete) will
this Hardware/Software support get installed and correct the swap space
allocation error.
SCB97-02 Page 1 of 8
KEYWORDS: SGI R8000/R10000 w/MULTIPLE PROCESSORS BLOCK LANCZOS
SOLVER
SPARSE SOLVER
DESCRIPTION: On the SGI R8000/R10000 machine running a 64 bit operating
system, a modal analysis using the Block Lanczos or Sparse solver will
use the maximum number of processors that are available even though
NPROC=1 (/config,nproc, 1) or NUM_PROC=1 is set in the config.ans file.
AFFECTED VERSIONS: Release 5.3 Release 5.4
WORKAROUND: Set the system environment variable "MP_SET_NUMTHREADS" to a
value of one before running the analysis.
KEYWORDS: PC WIN95 NT4.0
DESCRIPTION: The side menus in the GUI may appear to be slow on certain
PC platforms.
AFFECTED VERSIONS: Release 5.4
WORKAROUND: None
KEYWORDS: NT3.51 NT4.0 MATERIAL LIBRARIES
DESCRIPTION: The material library shipped with ANSYS 5.4 has an extension
on the files that ANSYS will not recognize when a material is requested
from the library
AFFECTED VERSIONS: Release 5.4
WORKAROUND: Globally change the material library file extensions to 3
characters or less.
GRAPHICS / GUI
KEYWORDS: GUI TSHAP CONTACT ELEMENTS
DESCRIPTION: When contact elements are present and TSHAP is set to the
CIRC option, the plot within the GUI will be off center and to the right
of the screen
AFFECTED VERSIONS: Release 5.4
WORKAROUND: Use the PAN - ZOOM - ROTATE menu to position the model on the
screen.
SCB97-02 Page 2 of 8
CDWRITE / IGES
KEYWORDS: SOLID MODELING CDWRITE CDREAD IGES
DESCRIPTION: CDWRITE with the options; ALL, COMB or SOLID uses IGES
transfer to write out modeling information. CDREAD with the same options
uses the same IGES transfer. Since the IGES transfer does not Guarantee
topological consistency, files may be produced that cannot be read back
into ANSYS when the above fields are used.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: Use the SAVE and RESUME commands to save data, or use the DB
option with the CDWRITE/CDREAD command.
KEYWORDS: SOLID MODELING IGES
DESCRIPTION: There may be IGES files created by the IGESOUT command in
ANSYS that cannot be read back into ANSYS with the IGESIN command.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: Use the SAVE and RESUME commands to save data, or use the DB
option with the CDWRITE/CDREAD command.
KEYWORDS: CDWRITE CDREAD
DESCRIPTION: A file created with CDWRITE in the 5.4 Beta release will not
read in the ANSYS 5.4 released version.
AFFECTED VERSIONS: Release 5.4
WORKAROUND: Issuing a SAVE at BETA 5.4 and a RESUME at 5.4 released will
allow transfer of data between the two releases.
ELEMENTS
KEYWORDS: SHELL99 NONLINEAR BUCKLING
DESCRIPTION: In a nonlinear buckling problem SHELL99 may not converge.
SHELL99 can be used with most nonlinear problems (NLGEOM,ON) but is
unpredictable in nonlinear buckling problems.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: Use SHELL91 for nonlinear buckling problems..
SCB97-02 Page 3 of 8
KEYWORDS: CIRCU124 COUPLING COEFFICENT MUTUAL INDUCTOR
DESCRIPTION: When using a coupling coefficient (k) of 1 for a CIRCU124
element defined as a mutual inductor will produce huge pivot ratios and
apparently bad results.
AFFECTED VERSIONS: Release 5.3
WORKAROUND: Use a value of .999 for 1.0. In Release 5.4, an input value
of k greater than or equal to 1.0 is redefined as having a value of .999.
KEYWORDS: BEAM4 PIPE16 MODAL GYROSCOPIC MODOPT
DESCRIPTION: When running a modal analysis using a BEAM4 / PIPE 16 model
with the gyroscopic moment term turned on, (nonzero SPIN velocity input
through real constant 11 in BEAM4, or Real Constant 13 in PIPE 16) the
eigenvalue frequencies computed by the damped eigensolver may be
sensitive to the value of the 'shift' used, e.g, MODOPT,DAMP,10,shift.
This is due to the fact that the element matrix coefficients may become
poorly conditioned, the ratio of maximum to minimum coefficient becoming
too large. This is especially true when the SPIN velocity is large.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: The user can verify that his model is not becoming affected
by this by verifying that the frequencies remain unchanged with two
different 'shift' values.
KEYWORDS: CONTA171 CONTA172 CONTA173 CONTA174
DESCRIPTION: When using CONTA171, CONTA172, CONTA173 or CONTA174 elements
with an uncompressed element list, incorrect or erroneous results may be
produced. An uncompressed element list is due from elements being
deleted and/or from remeshing which can produce unused element numbers in
the list of elements.
AFFECTED VERSIONS: Release 5.4
WORKAROUND: Issue a NUMCMP,ELEM prior to solution. This will compress
the element list and correct the problem.
KEYWORDS: CONTA171 CONTA172 CONTA173 CONTA174 PARALELL/MULTIPLE
PROCESSORS
DESCRIPTION: If running on a machine with multiple processors active
(setting of num_proc greater than 2) and using the contact elements
CONTA171, CONTA172, CONTA173, or CONTA174, erroneous results may be
produced. An uncompressed element list is due from elements being
deleted and/or from remeshing which can produce unused element numbers in
the list of elements.
AFFECTED VERSIONS: Release 5.4
WORKAROUND: Set the number of processors to 1 (num_proc = 1).
SCB97-02 Page 4 of 8
SOLID MODELING / MESHING
KEYWORDS: SOLID MODELING BOOLEAN AFILLT
DESCRIPTION: There are occasions where the AFILLT command will not cut
the areas properly when sharing a common line and using a small radius.
AFFECTED VERSIONS: Revision 5.2, Release 5.3, Release 5.4
WORKAROUND: It is possible in some cases to define a line and drag it
along the path desired in order to form a fillet.
KEYWORDS: ELEMENT SHAPE CHECKING ESEL CHECK
DESCRIPTION: The command "CHECK,ESEL,WARN" should select all elements
that produce shape warnings or errors. Quadrilateral elements
degenerated into triangles, or brick elements degenerated into wedges,
pyramids, or tetrahedra, produce warnings about the degeneracies.
However, only the element with the highest element number with those
degeneracies will be selected by CHECK,ESEL,WARN in the absence of other
flaws in those element(s).
This error also affects the 5.4 GUI paths Preprocessor>-Meshing- Plot Bad
Elms and Sel Bad Elems.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: Treat the selection or plotting of 1 or 2 degenerate elements
as notification that the model could contain an arbitrary number of such
elements.
KEYWORDS: SOLID MODELING xREFINE VIMP
DESCRIPTION: The xREFINE family of commands (KREF, LREF, AREF, NREF,
EREF) and the VIMP command have been shown to store an incorrect value in
the database if any of the meshed lines has only one element division on
it. This erroneous value can potentially impact solid modeling or
meshing operations further in the analysis. This value may be incorrect
when the following conditions are present:
1. When using the xREFINE commands on an area or tet mesh of
linear (non mid-side noded) elements and the mesh on any of the
lines on the areas or volumes affected by the refinement has
only one element division on it.
or 2. When using the VIMP command on linear tet elements, with the
CHGBND argument set to 1 (i.e., boundary modification is
allowed) and any line on the volume has only one element
division on it. Note that the problem will not occur if the
line with the one element division is attached to another
meshed volume, or if the line has been assigned the 1 division
by the LESIZE command.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND:
For the REFINE or VIMP commands;
- use quadratic elements, or
- on areas / volumes specify more than 1 division on all of the
lines.
For the VIMP command:
- set the CHGBND field to a value of 0 (do not allow boundary
modification), or
- use the LESIZE command to specify 1 division instead of the
ESIZE, SMRTSIZE or other sizing commands to automatically set
this value to 1.
SCB97-02 Page 5 of 8
SOLVER / SOLUTION
KEYWORDS: NONLINEAR BUCKLING STRESS STIFFENING
DESCRIPTION: Prior to Release 5.3, ANSYS would bisect and eventually
terminate when a negative main diagonal was encountered. In beam and
shell models, with stress stiffening turned on (SSTIF,ON), this was
frequently an indication that the buckling limit had been exceeded.
Beginning with Release 5.3, ANSYS solves for the negative main diagonal
without bisecting due to the negative main diagonal. Instead ANSYS gives
a WARNING message that small equation pivot terms have been detected by
the solver.
As a result of this new functionality, if compressive loads are applied
quickly and produce negative stiffness values for the lateral degrees of
freedom the solution will appear to skip over the instability points in
loading. The attained solution may follow a different post-buckling path
than expected.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: If you are interested in post-buckling behavior, use arc-
length (ARCLEN,ON) in addition to stress-stiffening (SSTIF,ON) and large
deflection (NLGEOM,ON). If the arc-length method does not support your
load type, then small load steps must be used without auto-time stepping
(AUTOTS,OFF). Also, the best form of perturbation for this type of
problem is a small geometric imperfection and not a lateral force. If
you are interested in obtaining buckling values and mode shapes then
eigenvalue buckling may be used.
KEYWORDS: PCG THERMAL EXPANSION
DESCRIPTION: Convergence in a PCG solution (EQSLV,PCG) is performed by
checking if the force computed from this iteration's displacements
matches the applied loads within the tolerance:
[symbols omitted]
where [symbol omitted] is the convergence tolerance (EQSLV,PCG,toler) and
[symbols omitted] is the norm of the applied force.
If the norm of the applied load [Symbols omitted] is large, then the
stress or reaction error will also be large. Large norms occur when a
body (e.g. gravity or thermal) loads are applied.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: When the load is large, use smaller tolerance in PCG in order
to have less solution errors of reaction or stress or use an alternative
solver.
KEYWORDS: PCG SOLVER MODAL POWERDYNAMICS
DESCRIPTION: The automatic shifting strategy used in Powerdynamics
(MODOPT,SUBSP and EQSLV,PCG) may cause the eigensolution to fail due to
"negative diagonals" encountered..
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: Use Block Lanczos or Subspace with the Frontal solver in
place of the PCG solver.
SCB97-02 Page 6 of 8
KEYWORDS: PCG NEWTON-RAPHSON FILE.ABT
DESCRIPTION: When using the PCG solver in a nonlinear problem, there may
be cases where the file.abt will not abort the job as documented.
AFFECTED VERSIONS: Release 5.4
WORKAROUND: Use a solver other than the PCG solver. Note that we
document that you should not use the PCG solver for difficult nonlinear
jobs.
KEYWORDS: BLOCK LANCZOS SOLVER EIGENVECTOR EXTRACTION
DESCRIPTION: When extracting a very large number of modes using the Block
Lanczos Solver, the extracted modes beyond a certain number will all have
0.0 displacements on the results (.rst) file. At Release 5.3 the maximum
number of modes that can requested in a single run is 718. At Release
5.4, the maximum number is 913.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: If you wish to extract a number of modes greater than the
known limits, you must perform two analyses. For example, to extract
1000 modes you would first extract and expand 500 modes using the command
MODOPT,LANB,500. For the second analysis you would issue the command
MODOPT,LANB,500,FREQB, where FREQB would be the eigenvalue of 500th mode
of the first analysis. This would shift the frequencies of interest in
the 2nd analysis to the point that the first analysis finished extracting
frequencies.
KEYWORDS: AUX12 RADIATION AXISYMMETRIC FORM FACTORS
DESCRIPTION: The form factors computed in an axisymmetric hidden
radiation procedure in AUX12 may be incorrect when the elements involved
have a very small or non-existent sight line and the mesh is coarse.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: For both hidden and non-hidden methods, the finer the mesh of
the radiating surface elements, the more accurate are the form factors.
However, when hidden method is used, it is particularly important to have
a fairly refined mesh in order to obtain the same level of accuracy in
view factor computation as the non-hidden method. Even though increasing
number of rays used (controlled by NZONE argument of VTYPE command) helps
in accuracy, for a coarse mesh, increasing NZONE to even its maximum
limit might not yield accurate solution for the hidden method.
KEYWORDS: PCG SOLVER RUNSTAT
DESCRIPTION: The RUNSTAT when used with the PCG solver will produce
estimates of memory and file size that are greater than what is actually
needed for the job to run.
AFFECTED VERSIONS: Release 5.4
WORKAROUND: None
SCB97-02 Page 7 of 8
LS/DYNA EXPLICIT DYNAMICS
KEYWORDS: POST PROCESSING NUMVAR
DESCRIPTION: The default number of variables defined with the NUMVAR
command is 30 for an explicit dynamics analysis. It is documented as
only defining 10 by default.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: N/A
KEYWORDS: EDLOAD EDCURVE /MENU
DESCRIPTION: The GUI will not allow input of a 2D array for the EDCURVE
or EDLOAD command.
AFFECTED VERSIONS: Release 5.3, Release 5.4
WORKAROUND: You must use the command input window to input the required
information for the 2D array with EDCURVE or EDLOAD.
FLOTRAN
KEYWORDS: FLUID141 FLUID142 COMPRESSIBLE FLOW
DESCRIPTION: Simulation of transient compressible, laminar, thermal
problems involving moving shock waves in an enclosed region can produce
source terms in the energy equation that lead to continued increase of
the total temperature.
AFFECTED VERSIONS: Revision 5.1, Revision 5.2, Release 5.3, Release 5.4
WORKAROUND: None
Author: Date: October 21, 1997
Michael F. Michalski
Reviewed by QA: Date: October 21, 1997
William J. Bryan
Approval: Date: October 21, 1997
David L. Conover
SCB97-02 Page 8 of 8
Attachment "ANSYS 5.3 CLASS3 ERROR SUMMARY BY REPORT NUMBER" 2 pages,
omitted.
Attachment "ANSYS 5.3 CLASS3 ERROR SUMMARY; KEYWORD SORT OF ALL
REVISION
5.3 CLASS3 ERRORS, Date of this report: 10/16/97", 6 pages, omitted.
*** END OF DOCUMENT ***
