This file contains only those errors corrected in MSC.Nastran Version 2004.0.0 on the R1 cd. Depending on what version of MSC.Nastran 2001 you currently have installed, you may also want to review the files: error2001r2.lst (version 2001.0.4) error2001r3.lst (version 2001.0.7) error2001r4.lst (version 2001.0.9) to see other errors that have been corrected versus your current installation. 2296 2004 Miscellaneous - Punched Output - Wrong Answers Punched output may contain seemingly meaningless numbers intermixed with the correct output. This will occur for element strain energy punched output (ESE(PUNCH)=n) and other output that has alpha strings intermixed with the real numbers. It can also happen with any punch if the punch thinks a certain number is real when it is really integer, or the reverse. The DEC Alpha is particularly sensitive to this in that it will crash when it occurs with no message and no "end of job" in the .f06. In the .log file you get: time: command terminated abnormally The punch file will be missing data and a large file named "core" will be in the JID directory. Avoidance: As an avoidance in Version 68, setting system(210)=1 will cure the crash. Unfortunately, it will also occasionally cau 3782 2004 DMAP - TYPE - UFM 404 In a TYPE DMAP statement, if the default value of a real or complex parameter is specified to be a letter, then the job will fail with the following misleading error message: 0*** USER FATAL MESSAGE 404 (CTYPE) 136 A NON-SINGLE QUOTE DELIMITER WAS DETECTED WHERE A SINGLE QUOTE DELIMITER IS REQUIRED The most common form of this error, which is difficult to detect, occurs when the letter "O" has been incorrectly substituted for the numeric "0". Avoidance: The message will correctly identify the line in error. Correct the offending character. 5090 2004 Documentation/QRG - PACABS, TABLED2, TABLED3, TABLED4 - UFM 6167 If the table that is used to define the resistance, reactance, or weighting function is the type TABLED2, TABLED3, or TABLED4, then the program issues "USER FATAL MESSAGE 6167 THE BULK DATA TABLED1 ENTRIES, ID= N WERE NOT SUPPLIED" Avoidance: The PACABS entry should reference TABLED1 entry instead. - note this is a documentation error. Tables referenced by TID1, TID2, TID3 can only be TABLED1- The TABLEDi i=2,3,4 imply a certian look up strategy TABLED2 and TABLED3 by shifting the curve and compressing the x-axes. TABLED4 is a polynomial expansion. The PACABS entry does not use these tables for look up. Rather it uses TABLED1 to perform a least square fit to determine a K and M value which bests fits the curve. 5556 2004 Documentation - Design Optimization User's Guide - Incomplete Data Listing Listing 7-7 in the Design Optimization User's Guide for Version 68 is missing a few lines of input data. This data is the Von Mises stress constraint data for the analytic boundary shapes example in Section 7.6 of the guide. Avoidance: Refer to the TPL deck D200AM3.dat for the correct listing. The documentation will be updated in Version 69. A new user's guide based on MSC.Nastran 2004 has been developed and is now available electronically and will soon be in hardcover. EHJ 1/2004 6404 2004 Restart - SCR=YES, DBVIEW - Datablocks Deleted If a restart run uses the DBVIEW statement in conjunction with SCR=YES, it is possible that the DBVIEWed datablocks may be deleted from the database. This is in spite of a RESTART, KEEP FMS statement. Avoidance: On restart runs, do not use SCR=YES. 7193 2004 Miscellaneous - Wrong continuation format - UFM 313 or Wrong Answers If column 1 of the continuation entry is blank, then the continuation entry is interpreted as a continuation of the last sorted bulk data entry. The error is likely to result in USER FATAL MESSAGE 313 ILLEGAL NUMBER OF WORDS ON CONTINUATION" or in rare cases possibly even wrong answers. An example follows: . Begin Bulk . . tic,100,101,1,0.,0. $ grid,1000,,0.,2.5,2.5 $ grid,101,,0.,0.,0. grid,102,,10.,0.,0. $ 2 3 4 5 6 7 8 9 10 cbar 1 1 101 102 1000 +000015 +000015 0.25 0. 0. 0.25 0. 0. $ . . enddata will be interpreted as BEGIN BULK . . CBAR 1 1 101 102 1000 +000015 . GRID 1 7340 2004 DMAP - LTU - SFM 3001 User DMAP runs submitted with scr=yes may fail with SYSTEM FATAL MESSAGE 3001. This error may result if the LTU logic deletes a datablock even though it is still needed. Avoidance: Run the job using SCR=NO or create a copy of the datablock and use the copy in the DMAP alter. 7376 2004 p-Elements - CORD3G - SFM 6333 (EQNINI) When a convective coordinate system (cord3g) is specified as a stress output coordinate system in OUTRCV, sdr2 crashes with BIOMSG: ERROR 211 HAS OCCURRED IN ROUTINE BLPK68 . Avoidance: Other than not referencing a cord3g entry, there is no avoidance. 25377 2004 Documentation - MATMOD option 25 matmod option 25 dmap dictionary article incomplete This option is designed to produce two output vectors for only unsymmetric matrices (form 1). For symmetric matrices (form 6) only the O1 vector is output. This condition can be detected by the sym output parameter, which is set to -1 for symmetric matrices, and 0 for unsymmetric matrices. 26132 2004 Optimization/Design Sensitivity - STATIC, SAERO - UFM 4574 Analysis types STATIC and SAERO, do not work together in solution 200. If they are attempted in the same run the following message is issued, "USER FATAL MESSAGE 4574, DATA RECOVERY STARTED BUT NO DISPLACEMENT VECTORS AVAILABLE" Avoidance: There is no avoidance. 29329 2004 Virtual Mass - MPRES - Missing Element Lables from Output, Missing Output Element labels for complex "PRESSURES FOR ELEMENTS ON FLUID STRUCTURE BOUNDARY" may be missing if SORT2 output is requested. Avoidance: Request SORT1 output. 29672 2004 PCOMP, STRAIN - Missing Output The case control command STRAIN=n is supposed to write ply strains to the .f06 file. This command only works if stress is also requested. Thus, there is no way to get ply strains only. Avoidance: To get only strain output, specify the STRESS=m command where SET m specifies an undefined element or use FORCE=m 29797 2004 Optimization/Design Sensitivity - DSAD, ANALYSIS=FLUTTER - SFM 3007 Flutter optimization cannot be performed with superelements unless the design model is in the residual. If attempted the run will fail and issue the following messages: ^^^ FLUTTER ANALYSIS COMPLETED. DESIGN CYCLE NUMBER= 1 *** SYSTEM FATAL MESSAGE 3007 (DSAD ) ILLEGAL INPUT TO SUBROUTINE DSAD14 0FATAL ERROR 31636 2004 Optimization/Design Sensitivity - DESSUB - UFM 6293 incorrectly issued If an attempt to place limits on a global response (such as weight) is applied using a DESSUB command, the program will incorrectly issue UFM 6293. Avoidance: Global responses (such as WEIGHT) may only be constrained using the DESGLB command. They may not be constrained by the DESSUB command. 31689 2004 Nonlinear Static Analysis - PGAP - Unnecessary Restriction for Non-Adaptive Gap The adaptive gap element requires that MU2 <= MU1; however, this restriction is unnecessarily imposed on the non-adaptive gap element. Avoidance: This limitation can be avoided by proper choice of the orientation vector. 31724 2004 Optimization/Design Sensitivity - OFREQ - Wrong Answers If sensitivity is requested in a frequency response analysis and there is an OFREQ request to limit output, the sensitivity results will be in error. Avoidance: Remove the OFREQ request from case control. In Version 2003 this error is considered fixed by simply ignoring the OFREQ Case Control command in Solution 200 during DSA phase of the job. 31832 2004 Miscellaneous - PARAM,USETPRT - Wrong Sort If there is a user defined degree of freedom set (USET, USET1), and row sorting is selected (PARAM,USETPRT,0 or 10), then the uset print tables will not be sorted in ascending order, but will be in a random order. Avoidance: Select 'Column'(val=1 or 11) or 'Row and Column'(val=2 or 12) sort on the PARAM,USETPRT entry. 31855 2004 Documentation - CGAP, PARAM,G - Unreasonably Large Damping With penalty CGAP (default), W3 damping can become excessive because of large CGAP penalty stiffness. This is a limitation that should be documented under the CGAP entry. Avoidance: When gap elements are used in nonlinear transient analysis, param,g damping should be avoided. Damping should be specified on the MATi entries and PARAM,W4 set. 32099 2004 Postprocessing - PARAM,POST,-2 - Multiple Geometry Written to OUTPUT2 File If multiple boundary conditions are specified in a normal modes analysis and param,post,-2 is set, the geometry data blocks are written to the output2 file in each subcase rather than just once. This unnecessarily increases the size of the output2 file. Avoidance: Specify PARAM,OGEOM,NO in all but the last subcase or insert the following DMAP alter: compile super3 alter 'call out2geom'(,-1) type parm,nddl,i,n,ncase $ if ( ncase<=1 ) then $ alter 'call out2geom' endif $ endalter 32725 2004 Postprocessing/Component Mode Synthesis - PARAM,FIXEDB - Missing Output If PARAM,POST is set to 0 to generate an MSC/XL database and PARAM,FIXEDB,-1 is used, the ".xdb" file will not contain the component modal results. Avoidance: There is no avoidance. Version 2003 adds a new SOL 9 to address this problem. 32852 2004 Security - FLEXlm - Can't locate Multi-processors Status: Fixed during NA_SMP upgrade of N2004. When using FLEXlm, if the number of processors specified by the PROC entry in the license file is less than the total number of processors on a machine, MSC/NASTRAN will use only one processor even if more are requested. Avoidance: Obtain a license file that allows the maximum available (or unlimited) number of processors. 33045 2004 Nonlinear Analysis - CGAP, TEMP(LOAD), ASETi - Wrong Answers at Omitted DOF Nonlinear analysis in Version 70 and 70.5 systems with multiple subcases, thermal loads, asets and gaps will give erroneous results at omitted degrees of freedom in the first printed load increment in the second subcase. If intermediate load increment results are not requested, the errors will occur in the final subcase. Avoidances: 1)Use the following alter: =========================================== COMPILE NLSTATIC $ ALTER 'SOLPRE=SOLCUR-1', 'SOLPRE=SOLCUR-1' $ ENDALTER $ =========================================== 2) Set INTOUT=YES in the NLPARM bulk data card, 3) use a version prior to 70. 4) Do not use SOL 106 with O-set in the resudual, in V2001. 33628 2004 Superelements - CSUPEXT - UWM 4155 If an image superelement has SPOINTs for component mode synthesis the the following warning message may be incorrectly encountered: USER WARNING MESSAGE 4155---GRID POINTS FOR SECONDARY SUPERELEMENT 82 ARE COLINEAR. CONFORMITY CHECKS NOT MADE. 33957 2004 Documentation/Composites - CQUAD8, PCOMP - Poor Results The bulk data description of the CQUAD8 in the Quick Reference Guide needs to be changed as follows: - Remove Figure 1 and change the title of Figure 2 to Figure 1, - The description of the MCID field needs to be changed as follows: "MCID ...................MCID coordinate system (defined by the CORDij entry or zero for the basic coordinate system) onto the surface of the element (see Remark 3). MCID is ignored ........." - Change Remark 3 as follows: "3. ........the line of constant xi." - Add the following paragraph to Remark 3: "The definition of the material coordinate system by projection is used to calculate an angle THETA. Please note that since xi changes direction throughout the element based on element shape, the material coordinate system varies similarly. Because of this an orthotropic or anisotropic material will cause the CQUAD8's stiffness to be biased by both it's shape and grid ordering. Use the QUAD4 element if a constant material coordinate system direction is desired." 35896 2004 Optimization/Design Sensitivity - PCOMP, Z0 - Unexpected Answers When optimizing ply thicknesses in Sol 200, the user must remember to recalculate Z0 if the midplane is to remain at the nodes. It is common for users to leave field 3 of the PCOMP entry (Z0) blank, and let MSC/NASTRAN calculate Z0 as -t/2. But in optimization, this field is not updated automatically, which can lead to unexpected results. Avoidance: Either:Create another design variable for Z0 (field 3 of PCOMP) and use DLINK to calculate updated Z0 or use an additional DVPREL1 to calculate z0 = 1/2 ( c0 + x1 + x2 + x3 + ... xn) where co is the thickness of invariant plies and the xi's are designed ply thicknesses) 35932 2004 Nonlinear Analysis - NLRGAP - Radial Gap without Friction, UFM 316 (IFPDRV) Using the NLRGAP does not allow input of the nonlinear forces without reference to a TABLED1 which specifies O.O for friction. Leaving the field 8 blank yields the following message: *** USER FATAL MESSAGE 316 (IFPDRV) ILLEGAL DATA ON BULK DATA ENTRY NLRGAP 20 SORTED ENTRY COUNT = 36 for the input item: SID GA GB plane tabk tabg tabu radius NLRGAP 20 90 101 XY 21 22 0.0 Avoidance: The avoidance is to specify a table for the TABU field with values of 0.0 for friction. 36646 2004 Random Analysis - CBUSH - SWM 3002, UWM 979 PSDF force plots fail for BUSH elements, but RESPONSE plots work. That is: XYPLOT ELFORCE RESPONSE.... (works) XYPLOT ELFORCE PSDF...(fails with the message below) *** SYSTEM WARNING MESSAGE 3002 (RANDOM) EOF ENCOUNTERED WHILE READING DATA BLOCK OES2 (FILE 107) IN SUBROUTINE RAND2 Avoidance: There is no avoidance. 36678 2004 Eigenvalue Analysis - EIGR, EIGRL - SFM 7340 The Lanczos method of eigenvalue extraction may give nonstandard error messages. For example a run may terminate with the following messages. *** SYSTEM FATAL MESSAGE 7340 (LNNHERR) process error reported by subroutine LNNP2CD see LNNP2CD abstract (ier = -725) *** SYSTEM FATAL MESSAGE 7340 (LNNHERR) warning reported by subroutine LNNDRVD see LNNDRVD abstract (ier = 725) Avoidance: The LNNDRVD or LNNDRVS messages meanings are given below: ier error condition. -723 insufficient workspace revealed during lanczos iteration -724 failure of tridiagonal ql iteration to converge -725 number of eigenvalues computed exceeds internal storage limitations -726 i/o error with lanczos vector file -727 failure of bidiagonal ql iteration to con 37010 2004 Optimization/Design Sensitivity - DRESP1 - SFM 4276, EC 31 When an optimization deck includes a DRESP2 entry which references the same DRESP1 ID more than once, the following system fatal message is issued from module DOM9: BIOMSG: ERROR 31 HAS OCCURRED IN ROUTINE GETSTR , FILE INDEX (FILEX) = 0 STATUS = 0 followed by: 0*** SYSTEM FATAL ERROR 4276.GETSTR ERROR CODE 31 PID= 0 Avoidance: Do not duplicate the same DRESP1 ID on a DRESP2 entry. 37263 2004 Virtual Mass - MFLUID - PARAM,VMOPT,1 Does Not Function Properly When param, vmopt, 1 is present, the virtual mass should be added to the structural mass at the g-set, that is, before any constraints are applied. In Version 70.6 and 70.7 and V2001, setting this parameter has no effect and the solution method adds virtual mass after all constraints are applied. This may or may not affect the results Avoidance: The alter below will make vmopt, 1 take effect. compile semg list noref $ Alter 'if ( vmopt=1 and','' $ always do what user requests IF ( VMOPT=1 ) THEN $ Compile semrm list noref $ Alter 'if ( vmopt=0 or','' $ IF ( VMOPT=0 ) THEN $ Note that the original dmap would over-rule the user's request if it did not seem to "make sense". The alter always honors the vmopt request. Note that if vmopt=1 is used on medium- and large-size mode 38016 2004 Superelements - SESET failures in SEP1X using THRU range entry It has been observed that the SESET entry may cause failures when a THRU range is specified where the low GRID ID is less than the starting GRID ID of the model. Avoidance: An avoidance is to make the starting grid equal to the low the low grid identification. 38627 2004 Subcases - SPC, DEFORM - UFM 9058 In static analysis, if DEFORM and SPC entries are changed between subcases the job will fail with the message: USER FATAL MESSAGE 9058 (SUPER3) - THE SOLUTION FOR THE RESIDUAL STRUCTURE DOES NOT EXIST Avoidance: Execute the subcases as two separate runs. 38869 2004 Documentation -- DMAP Module Dictionary - APPEND 1st param cannot be 0 Both the Version 68 DMAP Module Dictionary and the current Version 70.7 online pdf documentation indicate that the 1st parameter for the APPEND module may be either a "0" or a "1" in order to create an output datablock [IN1:IN2]. However, if "0" is used then "UFM 5005 INVALID VALUE OF OPTION PARAMETER DETECTED 0" is issued, followed by "SFM 3007 MODULE = APPEND ILLEGAL INPUT TO SUBROUTINE APPEND". Avoidance: Set the 1st parameter to "1" instead of "0". 39031 2004 Documentation - DMAP Module Dictionary - MATMOD Option 5, Form 3 Example 3 under the DMAP Module Dictionary's description of MATMOD option 5 should be modified from Transform KGGBASIC coordinate system 10. MATMOD CSTM,SIL,,,,/TRANS10,/5//10 $ TRNSP TRANS10/TRANST10T $ SMPYAD TRANS10,KGGBASIC,TRANS10T,,,/KGG10/3 $ to Transform KGGBASIC to coordinate system 10. MATMOD CSTM,SIL,,,,/TRANS10,/5//10 $ TRNSP TRANS10/TRANST10T $ SMPYAD TRANST10T,KGGBASIC,TRANS10,,,/KGG10/3 $ 39109 2004 Eigenvalue Analysis - GIV, MGIV, HOU, MHOU - Poor Answers When using methods GIV, HOU, or their variants, if there are repeated roots their eigenvectors may not be orthogonal. A small test problem with 9 repeated roots produced only three orthognal vectors. Avoidance: An avoidance is to use the LAN method. 39208 2004 Plotting - XYPLOT, XYTRAN - Missing Plot Output If a non-existant grid point is requested by the XYPLOT command, the XYTRAN module makes a premature exit. i.e. when a displacement plot command cannot find the grid point, all of the acceleration and element stress/force plots could be skipped. 39375 2004 FMS - ASSIGN, scr=yes - Database Not Deleted :q vi rIn Version 70.7, if an ASSIGN statement is specified for the database, the database will not be deleted at the end of the job even if scr=yes is specified on the command line. Avoidances: There are two possible avoidances. 1. Add the following FMS statement to the input file init master(s) 2. Add a ",temp" to the end of the assign statement. 39490 2004 Nonlinear Transient Analysis - NLTRD2 - SFM 4276 UNPACK It is possible that a nonlinear transient analysis may fail with BIOMSG 33, followed by SYTEM FATAL MESSAGE 4276, indicating an UNPACK error. Avoidance: The exact cause of this problem is under investigation. One potential cause is epoints. Replacing epoints with spoints may solve the problem. 39612 2004 DMAP - ORTHOG Module Incorrectly Declares a Matrix to a Null Matrix The ORTHOG module may incorrectly declare a matrix with small terms to be null and the following message is then issued. *** SYSTEM WARNING MESSAGE 6467 (ORTHRD) A NULL COLUMN WAS ENCOUNTERED AT COLUMN NUMBER = 1 DURING ORTHOGONALIZATION 39889 2004 Distributed Memory Parallel - DMP, EMG - Job Hangs A distributed memory parallel Statics, SOL 101, job will hang if one of the tasks gets a fatal error in EMG. Avoidance: Other than correcting the input there is no avoidance. 39994 2004 Documentation/QRG - YTAXIS - UFM 676 If the XYPLOT command YTAXIS is used, the program will incorrectly issue UFM 676 indicating that it is not recognized. *** USER FATAL MESSAGE 676 (IFP1XY) YTAXIS IS NOT RECOGNIZED ON ABOVE CARD. Avoidance: Remove the command 40261 2004 Documentation/QRG - Part Superelements - PARAM,OUNIT2 is ignored The Nastran Quick Reference Guide indicates that PARAM,OUNIT2 may be specified in either the Case Control or the Bulk Data. If it is specified in the Bulk Data after a BEGIN SUPER entry it will be ignored. Avoidance: Specify the parameter in the case control.. 40263 2004 Transient Response Analysis - EPOINTS, PARAM,ADSTAT - UFM 6143 If a model has EPOINTs and PARAM,ADSTAT is specified on a pre-stressed structure in a modal transient response analysis, the run will fail with UFM 6143 indicating that incompatible matrices were attempted to be merged. Avoidance: Insert the following alter: compile super3 alter 'append.*freq0,mass' if(n1<>nruh1)then $ type parm,,i,n,ted1 $ ted1 = n1 + 1 $ matgen ,/vone2/6/ted1/1/n1 $ merge ,,f0m,,,,vone2/f0m1/1 $ $ else $ n1=nruh1 $ alter 'merge.*f0m,,,,vone' endif $ 40315 2004 Documentation - DMAP Module Dictionary - DIAGONAL The DMAP Module Dictionary description of the DIAGONAL Module's WHOLE option is incorrect for complex matrices. In case of a complex matrix, the option WHOLE gives the following result: POWER = 1.: Magnitude of the complex matrix elements POWER != 1.: The real part of the complex matrix element is raised to the power 40423 2004 nastran script - s.pmem - Memory Allocation If s.pmem is set in the nast*.ini file, then it will be ignored. Avoidance: Set s.pmem in the nastran command script: nast* ) exec $cmdtool "$@" s.pmem=1024mb 40563 2004 Documentation/QRG - AESTAT - Add Remark on Units of ROLL, PITCH and YAW Table 1 of the description of the AESTAT bulk data entry in the Quick Reference Guide states that ROLL, PITCH and YAW are urdot(R1), urdot(R2) and urdot(R3). It might be incorrectly concluded that the units are rad/s. Actually, the examples given in the Aeroelastic User's Guide as well as the more detailed description in the seminar notes show that normalized roll, pitch and yaw rates have to be input, i.e. ROLL = pb/2V, PITCH = qc/2V and YAW = rb/2V. A remark should be added which explicitly states this. 40564 2004 Documentation/QRG - AUNITS - Incomplete Description The Quick Reference Guide description of parameter AUNITS should be replaced by the clearer description from the seminar notes: "Accelerations are divided by this parameter. To convert accelerations input in g's into physical, consistent units, set AUNITS to 1/g." 40579 2004 Documentation/QRG - DVPREL1 - PSHELL Property Name for 12I/T**3 The DVPREL1 entry has the ability to reference property field id names; however, the PSHELL id field 12I/T**3 and other fields that begin with integers can not be referenced in this manner. Avoidance: Refer to the field by the field position. 40734 2004 Superelements - SEBNDRY and SECONCT with THRU - UFM 315, Wrong answers, UFM 6782 The THRU format on the SEBNDRY and SECONCT Bulk Data entries causes various failures or wrong answers: 1. THRU ranges on SEBNDRY continuation entries cause UFM 315. 2. The second THRU range on the SEBNDRY parent entry is ignored with no warning or fatal message issued and produces wrong answers. 3. Multiple SEBNDRY entries causes UFM 311. 4. THRU ranges with gaps in numbering on the SECONCT entry causes UFM 6782. See 1-11886201 for enhancement request. Avoidance: For SEBNDRY use only one THRU range. For SECONCT, specify separate THRU ranges without any gaps. 40839 2004 Documentation/QRG - RESTART, DMIG - Inefficient Restart or SFM 4252 Restarts with DMIG entries may unneccessarily re-process the entire model even though no changes to the input have occurred. The way to verify that this is happening is to run with DIAG 56 and to look in the .f04 file for "changed" and "DMIG" on the same line. If these words occur on a line output by the RESTART module, the program will re-run the entire model. Avoidance: This is currently a limitation of the sort routines used in Nastran. Other potentially problem entries include BNDFL, DMIAX, ELIST, DMIAX, ELIST, MFLUID, RADLIST, and RADMTX. If there are no changes to any BNDFL, DMIAX, DMIG, ELIST, MFLUID, RADLIST, and RADMTX entries, then insert the following alter to disable the checking for changes in these entries. compile ifpl alter 'restart.*matpool','' message //'re 40892 2004 Documentation/QRG - PBEAM The following remark should be added to the Quick Reference Guide description of the PBEAM entry: Remark 13: In nonlinear analysis the location of the 8 plastic rods is the same on the PBEAM as on the PBCOMP SYMOPT=0. 41020 2004 Optimization/Design Sensitivity - CSUPER - SFM 3007 (MTRXIN) If external superelements are used in optimization, and DESMAX has a value greater than the number of cycles required to converge, and the run is not saving the database (i.e. scr=yes), the run may fail with the following message:. SYSTEM FATAL MESSAGE 3007 (MTRXIN) ILLEGAL INPUT TO SUBROUTINE MTRXIN FATAL ERROR Avoidance: Either set DESMAX to the number of cycles required to converge or run with SCR=NO on the submittal. 41061 2004 Nonlinear Analysis - PLOADX1, PARAM,FOLLOWK - UFM The follower stiffness for PLOADX1 is incorrect. This could cause some nonlinear static problems in SOL 106 to diverge. Avoidance: Turn off the follower stiffness by specifying param, followk, no. 41379 2004 Nonlinear Transient Analysis - TSTEPNL - Poor Message, SFM 4276, EC 7777 If NDT on the TSTEPNL entry is incorrectly set to a value less than 3, the program will fail with a Floating Point Exception. Avoidance: Specify a value greater than 3 for NDT. 41397 2004 DMIG - PUNCH File - UFM 316 (IFPDRV) If MSC.Nastran is requested to generate DMIG entries, and those entries are read back into MSC.Nastran on a restart job, then the following error may occur: *** USER FATAL MESSAGE 316 (IFPDRV) ILLEGAL DATA ON BULK DATA ENTRY DMIG DMIG *KAAX 7 1 +000001 *+000001 7 1 1.833933289E+01 +000002 The error is a result from the g77 (ecgs) compiler on linux writing double precision variables with an "E" instead of with a "D" Avoidance: Edit the DMIG file and change all the E formats to D. This can be done with "g/E-/s//D-/" and "g/E+/s//D+/" from within "vi" 41437 2004 Frequency Response - PARAM,DFREQ - Missing Frequencies Parameter DFREQ is used to determine how the program will filter the loading frequencies to remove apparent duplications. This parameter is ignored in SOL 108, which may result in unintentional removal of desired loading frequencies. Avoidance: Insert the following alter in your file: compile sedfreq alter 'frlgen ','' type parm,nddl,rs,y,dfreq $ frlgen dynamics,,/frl0,frli///dfreq $ endalter 41458 2004 Utilities - RCOUT2 - INTERNAL ERROR: buf_next: bad fileposition RCOUT on the NT will not work with files larger than 2gb. This limitation has been removed in Version 2003. 125393 2004 Optimization/Design Sensitivity - EIGC - Poor Answers Any method other than complex Lanczos will not produce left-hand eigenvectors of sufficient accuracy to yield reasonable complex eigenvalue sensitivities. If another method is requested the job should issue a USER FATAL MESSAGE recommend. UFM 7629 (DOPR3) issued if method on EIGC is not CLAN or HESS. 3/28/02 gdc 1-2M26V 2004 Nonlinear Analysis - GPFORCE, NLPCI - UFM 4276 EC 3102 QOPEN In Version 2001 a request for GPFORCE output in a nonlinear run that uses arc-length methods specified on the NLPCI entry may result in USER FATAL MESSAGE 4276 EC 3102. Avoidance: Other than removing the GPFORCE request or NLPCI, there is no known avoidance. 1-2771301 2004 Postprocessing - Formatted OUPUT2 File causes excessive SWM 1204 (UNRTN) When Nastran is writing a formated op2 file for certain large models, it creates thousands of warning message: *** SYSTEM WARNING MESSAGE 1204 (UNRTN) ABSOLUTE VALUE OF SINGLE PRECISION VALUE IS LESS THAN THE MINIMUM FILTER OF 1.0E-37, AND WILL BE RESET TO THE CORRECTLY SIGNED FILTER VALUE 1-2994324 2004 DMAP - ORTHOG Module - Incorrect Matrix Trailers may lead to various fatal messages If the ORTHOG module is used to generate the Q and R output matrices for a rectangular matrix A using parameter values of IOPT=1 and ICON=0, the R output matrix, which is an upper traingular matrix, is generated with wrong trailer values. This causes a fatal error if this matrix is subsequently used in a module like MPYAD. 1-3036805 2004 Documentation - Missing Documentation for Heat Transfer in P-Elements P-Elements were extended to support heat transfer analysis in MSC.Nastran Version 70.7. The Quick Reference Guide was not issued for 70.7 and the updates have not been added to the 2001 QRG. Avoidance: If you are interested in this capability contact support and the documentation will be provided. 1-3068607 2004 Optimization/Design Sensitivity - Senstivities of PBEAMLs that include warping may be incorrec Sensitivities of warping coefficients are ignored in the design of PBEAML section types. The effect of this is that warping coefficients are unchanged from their initial value thereby creating wrong answers when these effects are significant. Avoidance: An awkward avoidance is to design equivalent PBEAM properties using DVPREL1 and DVPREL2 entries. The SA and SB points on the CBEAM entry must be grid points for warping sensitivities to be calculated correctly even when the PBEAM entry is used. Otherwise, it is suggested that if the PBEAML is used, the final PBEAML design be submitted in a new run to see if the neglected warping effects are significant. 1-3159901 2004 Distributed Memory Parallel - DMP - SFM 3001 (DISUTIL) Small models, demonstration problem v10101.dat with dmp=2 is an example, that are submitted with dmp may fail with SFM 3001 (DISUTIL) THE INPUT DATA BLOCK NAMED QGHD DOES NOT EXIST. The problem is that the partitioner is not able to partition the model into as many domains as desired. Usually this does not cause a problem and the job continues serially. Other jobs that enter the following line in subdmap sedrcvr will have the problem: "if ( prtreslt='yes' and noqg>-1) then .. ) Avoidance; Don't run dmp on small jobs. 1-3379914 2004 Optimization - PBEAML/PBARL - Incorrect Sensitivity Coefficients In an optimization/sensitivity analysis, if the model contains PBEAML or PBARL entries, then the eigenvector sensitivity coefficients will be incorrect. Avoidance: Replace the PBEAML or PBARL entry with the equivalent PBEAM or PBAR entry. 1-3414201 2004 Documentation/QRG - TEMP(INIT), TEMP(LOAD) Table 4-3 in Section 4A of the Quick Reference Guide indicates that TEMP(INIT) is not supported in SOL's 101, 103, and 105, 107, 108, 109, 110, 111, and 112. TEMP(INIT) may be used in SOLs 103, 107, 108, 109, 110, 111, and 112 if STATSUB is used to define the initial temperature. Likewise TEMP(LOAD) may be used in SOLs 103, 107 and 110 if STATSUB is used. 1-3422001 2004 Documentation/QRG - PARAM,INREL,-2 - Missing Documentation The documentation for PARAM,INREL,-2 is missing from the QRG. The value of param, inrel, -2 provides an inertia relief analysis in SOL 101 without the need for a SUPORTi entry. Remove all such entries. The model must have mass defined. Request TAN 4854 from your client support user documentation and example problems. 1-3437715 2004 DMAP - TABPRT - NaN in TABPRT output of ONRGY When the ONRGY datablock is printed using TABPRT using the NDDL descriptions, the final -1 is not recognized and is printed as NaN. compile sedrcvr alter 208 tabprt onrgy1//'onrgy'///1 $ endalter 1-3439701 2004 Optimization/Design Sensitivity - DESGLB, DESSUB - Wrong Answers for Superelement Optimization When a superelement optimization job includes a DESGLB (or DESSUB) command that references a constraint defined by either a volume response or weight response with ATT1=ALL, the contraint values for each SE, along with the one for the whole model are presented to the optimizer which may lead the optimizer to reach a wrong conclusion. Avoidance: There is no avoidance. 1-3507901 2004 Eigenvalue Analysis - PARAM,RESVEC - Floating Point Overflow in frrd1 Modal frequency response jobs may fail in the frrd1 module with a traceback when param,resvec,yes is specified. Traceback: Offset 0x00001680 in procedure bldpk_ Offset 0x00000790 in procedure frrd1f_ Offset 0x0000215c in procedure sqfreq_ Offset 0x0000000c in procedure freq_ Offset 0x00000124 in procedure xsem10_ Offset 0x00000164 in procedure xsemdr_ Offset 0x0000005c in procedure nastmon_ Offset 0x000009ac in procedure mainal --- End of call chain --- Avoidance: Do no specify param,resvec or request fewer eigenvalues. 1-3543501 2004 Optimization/Design Sensitivity - Multiple Superelement, Incorrect Complex Eigenvalue Sensitivities In multiple superelements, the complex eigenvalue sensitivity coefficients are incorrect for the residual superelement (SEID=0). This problem also exists in flutter sensitivity when more than one superelement is in the model. This problem can be avoided with following DMAP alter: compile ceigsens,list $ alter 'EQUIVX CDELK1/CDELK/DMRESD',''(,10) $ IF ( DMRESD<0 ) THEN $ EQUIVX CDELK1/CDELK/DMRESD $ EQUIVX CDELM1/CDELM/DMRESD $ EQUIVX CDELB1/CDELB/DMRESD $ ELSE $ ADD5 CDELK1,CDELK,,,/CDELK2 $ ADD5 CDELM1,CDELM,,,/CDELM2 $ ADD5 CDELB1,CDELB,,,/CDELB2 $ EQUIVX CDELK2/CDELK/-1 $ EQUIVX CDELM2/CDELM/-1 $ EQUIVX CDELB2/CDELB/-1 $ ENDIF $ DMRESD<0 $ In addition, flutter job may fail if upstream SEs have their own subcase. 1-3556907 2004 FLuid/Structure Analysis - FLSPOUT, PANEL, - SWM 2072, Missing Output If the FLSPOUT Case Control command selects PANELMP and GRIDMP=ALL and one or more of the PANEL Bulk Data entries lists more than one panel then SWM 2072 may be issued and acoustic output will be missing for some panels. 1-3694031 2004 Acoustic Elements - NOGO Flag Encountered, Poor Message If HEXA, PENTA, or TETRA acoustic elements fail the geometry check the job will stop after EMG and the only message output will be "NOGO FLAG ENCOUNTERED". Avoidance: Specifying GEOMCHECK STD_MSG in the Executive Section will cause the correct fatal message and the problem elements will be identified. 1-3743525 2004 Optimization/Design Sensitivity - Aeroelasticity with Temperature Loads An aeroelastic optimization task with multiple subcases and with one or more subcases having a temperature load will fail in the sensitivity phase. The type of failure is dependent on the specific design task, but there will be an indication of missing data. Avoidance: The following DMAP avoidance is available for the V70.7 and 2001 Release: compile saersens alter 'dksalx'(,-1) if ( stflg>0) then $ alter 'lauldx'(2) endif $ stflg>0 alter 'lauldx'(3) if ( ncol > 0 ) then $ alter 'lauldx'(5,2) endif $ ncol>0 compile gendv2p alter 'bcdr' PARAML CASE0//'DTI'/s,n,nskip/7//S,N,TEMPLD $ PARAML CASE0//'DTI'/s,n,nskip/6//S,N,DEFORM $ compile xsedrdr alter 'bcdr' (complete alter does not fit in allowed space. See the attachment) 1-3749805 2004 Elements - ELSUM, PCOMP - SFM 3008 (ELTPRT), Excessve Memory If ELSUM output is requested and the model has composite elements, the run may fail with in ELTPRT or ELTPT4 with a message of the form: *** SYSTEM FATAL MESSAGE 3008 (ELTPRT) INSUFFICIENT MEMORY AVAILABLE FOR SUBROUTINE PREMAT Avoidance: There is no avoidance other than providing a large amount of memory or not requesting ELSUM output. 1-3754501 2004 Nonlinear Analysis - param,fksymfac - Difficulty Converging, Follower Forces SOL 106 may diverge with symmetrized follower stiffness as the rotations becomes large. If the effect of asymmetry of the follower stiffness dominates the stiffness, but the default option for param,fksymfac symmetrizes it, then the problem may have convergence difficulty. Avoidance: Set param,fksymfac,0.0. 1-3771601 2004 Miscellaneous - NOLIN, NLRGAP - Poor Message MSC.Nastran issues an incorrect fatal error message when the NOLIN-type Bulk Data entry (including NLRGAP) references an undefined grid point ID. 1-3816002 2004 Case Control - No Warning Message Issue for Invalid Requests If a case control output request includes unsupported options, there is no warning issued and the request is simply ignored. For example if DISP(POLAR) is specified instead of DISP(PHASE) to get output as magnitude/phase, the output will be real/imaginary and no warning message will be issued. 1-3860101 2004 Complex Eigenvalue Analysis - OFREQ, OMODES - SFM 6143 (UPARTN) Using the OFREQ or OMODES feature in modal complex eigenvalue analysis results in SYSTEM FATAL MESSAGE 6143 in the UPARTN module. 1-3889601 2004 Documentation/QRG - GPFORCE - Slideline Forces ignored in the GPFORCE Calculation Remark 5 of the GPFORCE case Control description in the V2001 QRG needs to be modified to state that Slideline force contributions are ignored in the GPFORCE calculation. 1-3948211 2004 Documentation/QRG - PARAM,OMID - Missing Documentation The documentation for parameter OMID is missing from the 2001 Quick Reference Guide. The parameter is documented in the 2001 Release Guide. The documentation should read: OMID Default=NO To print or punch the results in the material coordinate system, set the parameter OMID to yes. Applicable to forces, strains, and stresses for CTRIA3, CQUAD4, CTRIA6, and CQUAD8. Other elements and outputs are not supported. This capability is not supported by pre-processors (xdb and op2 output are not changed) and grid point stress output that assumes output is in element coordinate systems. 1-3961304 2004 Dynamics - PARAM,RESVEC,YES, SOL 103 -Not Generated In a SOL 103 run, if the direct load method is used (TLOADi pointing directly to a spatial load, e.g, FORCE entry) instead of the LOADSET/LSEQ combination, then the residual vectors are not generated. Avoidance: If the residual vectors are desired in a SOL 103 run, then use the LOADSET/LSEQ combination 1-3971529 2004 Elements - CQUAD4 - UFM 4297 Incorrectly Issued for Missing MATi If the PSHELL for a valid-geometry QUAD4 references a non-existent MATi entry, then UFM 4297 (EQD4D) is issued indicating the element has an interior angle greater than 180 degrees. This message is incorrect. The UFM 4297 is followed by an appropriate UFM 2042 (EMG) that does indicate that the MATi entry is missing. This sequence of messages occurs in both Version 70.7.2 and 2001. In earlier releases (v70 and 70.5), only the appropriate 2042 message is issued. The UFM 4297 message is incorrect/misleading which will increase model debug time for the user and increase support time/cost for MSC. A small test deck demonstrating the problem is attached. 1-3975803 2004 P-Elements - CBEAM, TEMP(MAT) - Wrong Answers For the beam p-elements, the temperatures used for the temperature-dependent material properties are incorrect. Avoidance: There is no avoidance. 1-4022230 2004 Elements - CQUAD4, CTRIA3 - Misleading UFM 4297. A missing material entry may cause UFM 4297 (EQD4D) to be incorrectly issued: QUAD4 ELEMENT WITH ID = 140507 HAS AN INTERIOR ANGLE POSSIBLY GREATER THAN OR EQUAL 180 DEGREES. before issuing a correct message UFM 2042 (EMG) MISSING MATERIAL TABLE 111551 FOR ELEMENT 140507 This is misleading when a large number of UFM 4297 are issued before the correct UFM 2042. 1-4056908 2004 Distributed Memory Parallel - GDMODES, w/multiple BC's - SFM 6143 in DUMERG A sol 103 run w/ multiple subcases having different SPC sets runs in Version 70.7.2 DMP but not Version 2001. More precisely, multiple boundary condition subcases with different s-set size do not run under Geometric Domain Parallel (GDMODES). The following error results: *** SYSTEM FATAL MESSAGE 6143 (DUMERG) THE SIZES OF THE INPUT MATRICES AND THE DOF SETS ARE INCOMPATIBLE. SPECIFICALLY: The number of rows in PHA1 is not equal to the size of the COMP-set Essentially, when the F-set changes between subcases, an incompatible matrix multiplication with PHA1 is attempted. Multiple BC case works when Domainsolver FDMODES is used. 1-4082701 2004 Documentation/QRG - LOADCYN - UFM 4342 (CY1BLD) A remark needs to be added to the Quick Reference Guide LOADCYN entry that indicates that in cyclic buckling anaysis loads may only be applied to the first segment and only harmonic zero loads may be applied. 1-4082801 2004 Elements - CWELD - Incorrect sign on moment force output. ccordinate system is not published The sign of the m_y element moments of the CWELD is wrong, it must be opposite to be the same as for the CBAR. In adition, the CWELD element force item codes are not published in the Quick Reference Guide and the rules for the element coordinate system have not been published. A word document is attached to this quality that contains all the necessary information for post processing the CWELD. 1-4100101 2004 Documentation/QRG - DIAG 49 - Remove DIAG 49 from QRG DIAG 49 should be removed for the Quick Reference Guide, see error 33293. The utility f04rprt should be used to summarize the f04 Execution Summary. 1-4117910 2004 Documentation/QRG - PARAM,XYUNIT - Missing Documentation The documentation for parameter XYUNIT is missing from the 2001 Quick Reference Guide. The documentation should read: XYUNIT n XYUNIT is used in conjunction with an FMS ASSIGN statement to specify the unit number for the storage of design sensitivity data in comma separated value format for use in a spreadsheet. 1-4150825 2004 Documentation/QRG - FREQ1 - Incorrect Formula The formula in Remark 3 of the FREQ1 bulk data entry description should say: (NDF+1) not (NDF=1) 1-4168528 2004 Documentation/QRG - GPKE Not Available for Dynamic Response GPKE output is only supported for normal modes analysis. Contrary to what is indicated in Table 4-5 of the QRG, the GPKE output is not supported in SOL 108. GPKE output can be obtained for the normal modes part of the other modal dynamic solutions. 1-4195201 2004 Buckling - CWELD - Possible Wrong Answers The CWELD element does not support nonlinear analysis nor does it create a differential stiffness matrix. However, if the CWELD element is used in a SOL 105 buckling analysis, depending on the problem and machine that you are running on, the job may terminate without any message, yield incorrect results, or terminate with the following messages: MAIN: The FPU has been reset after floating point exception. MAIN: For reference, A(MAINAL) = 401380, A(/SYSTEM/) = 1915050 MAIN: A(/XNSTRN/) = 3320020 MAIN: "Floating point zero divide" (C000008E) exception encountered. MAIN: Exception occurred at address 010654CD. MAIN: Context Flags 0001003F MAIN: EAX 00000006 EBX 03 John, please verify the fix. -SJF 1-4214309 2004 Dynamic Analysis - Enforced Motion with Static Preload - Wrong Answers If enforced motion is used in a modal solution (SOL 111 or 112) with a static preload, the run will either fail indicating that no loading is defined or will run with incorrect results (if another loading is applied in addition to the enforced motion). Avoidance: Insert the following DMAP alter: compile SEMTRAN $ SOL 112 or SEMFREQ for SOL 111 alter 'umerge.*ys.*yg'(,-1) lcgen casers,slt,ett,dynamic,geom4s/cases2t/s,n,dummt/'tranresp' $ paraml cases2t//'dti'/1/4//s,n,load $ 1-4239801 2004 P-Elements - STRAIN, TEMP - Strain Output does not Include Thermal Strain The strain output for solid p-elements is the mechanical strain, which is the difference of the total and thermal strains. For a solid model undergoing free thermal expansion, the strain output would be zero, rather than the non-zero mechanical strain that is customary. The stress output is calculated from the mechanical strain and is correct. 1-4304427 2004 Restarts - Inefficient restart if Ge is changed on a PELASi entry When the Ge field of a PELAS entry is changed between a SOL 103 cold start and a SOL 111 restart, the eigenvalues are recomputed. This is inefficient as the change to damping does not affect the eigenvalue computation. Resolution: Insert the following DMAP in your file (be careful, as it disables the restart checking for element property changes) compile ifpl alter 'RESTART EPT,EPTR, /LIST','' $ 194 message //'WARNING - element property change restart checkingl/ ' has been disabled' $ endalter compile phase0, list alter 'RESTART EPTS,EPTSR','' $ line 841 message //'WARNING - element property change restart checkingl/ ' has been disabled' $ endalter 1-4332806 2004 Documentation/QRG - LOADCYH,DAREA The following sentence in Remark 5 of the QRG description of the LOADCYH entry ..FORCE2,MOMENT2,PLOAD,... should be replaced by ..FORCE2,MOMENT2,DAREA(if these entries have been converted),PLOAD,... 1-4342210 2004 Documentation/QRG - system cell 319 In the Release Guide, Section 9.8 says set system cell(119)=1 for extended error messages. This should be system cell(319)=1 instead. Release Guides are not re-issued. This error will be marked fixed when the system cell is added to the 2003 QRG. 1-4344301 2004 Documentation/IOG - Description of mattst.f and tabtst.f are reversed The descriptions of mattst.f and tabtst.f in Section 6.15 of the MSC.Nastran 2001 Installations/Operation Guide are reversed. The sample source program mattst.f reads binary output4 and not output2 and tabtst.f reads output2 and not output4. 1-4365301 2004 Frequency Response - OFREQ, SPCF, MPCF, CBUSH, frequency-dependent elements - SFM 5423 If OFREQ and MPCF (or SPCF) are requested in a run with frequency dependent elements then the run may fail with SFM 5423 in MPYAD. Avoidance: Insert the following DMAP alter: compile disprs alter ' modacc ','' MODACC CASEf,fOL,Uhf,PPf,qmgf,qsgf/ fOL0,Uhf1,PPf0,qmgf0,qsgf0/APP $ alter 'call cforce2'(1),'' call CFORCE2 USET,rgs,Uff,MSF1,BSF1,KSF1,Pgf1,fOL1, qsgf0, qdg,vspcg,/ Qpf/ APP/APP1/'spcforce'/true $ alter 'call cforce2'(3),'' call CFORCE2 USET,rgs,Unt,MSn,bSn,KSn,Pgf1,fOL1,qsgf0, qdg,vspcg,/ Qpf/ APP/APP1/'spcforce'/true $ alter 'call cforce2'(4),'' call cforce2 USET,rgm,Uhf1,MMG1,BMG1,KMG1,Pgf1, fOL1,qmgf0,,vmpcg,/ Qmpf/ APP/APP1/'mpcforce'/true $ alter 'call cforce2'(5),'' call cforce2 USET,rgm,Upf1,MMG1,BMG1,KMG1,Pgf1, fOL1,qmgf0,,vmpcg,/ Qmpf/ APP/APP1/'mpcforce'/true $ endalter 1-4371801 2004 External Superelements - EXTSEOUT - SFM 3001 (DSAP) A SOL 103 external superelement creation run using the EXTSEOUT Case Control command fails with SFM 3001 in the DSAP module if the superelement is reduced via static condensation (no QSET data provided) rather than via component mode reduction. 1-4383701 2004 Aeroelastic Analysis - CBUSH - UFM 2170 In SOL 146, if a frequency-dependent element, such as a CBUSH element, is used then the job will fail with the following error messages: *** USER FATAL MESSAGE 2170 (PARTN2) BOTH THE ROW AND COLUMN PARTITIONING VECTORS ARE PURGED AND ONLY ONE MAY BE. FATAL ERROR Avoidance: Apply the folloiwng dmap avoidance. compile seaero $ v705 alter 'call.*freqrs','' $ CALL FREQRS CASES ,USETD ,DLT ,FRL ,TRL , RPH ,DIT ,EST ,CSTMS ,MPTS , BGPDTS ,GPSNTS ,GPECT ,SILS ,EDTS , KHH ,MHH ,BHH , ,QHJL , ACPT ,CSTMA ,QHHL , ,DITID , ,RGHFD ,RSGG ,RMGG ,PPF , FOL , ,vgfd , / AUHF ,QMGF ,QSGF ,PHF / SOLTYP /FREQDEP /NOA /FOURIER /FALSE / 1 /TRUE /BOV /DOMPCFFD/DOSPCFFD/ FALSE /NOUE /0 $ 1-4405020 2004 ACMS - fluid/structure, DOMAINSOLVER - SFM 3001 (FBS) When using ACMS for a fluid/structure model, the job may fail with the following fatal messages: *** SYSTEM FATAL MESSAGE 3001 (FBS) THE INPUT DATA BLOCK NAMED LLL AT POSITION 1 DOES NOT EXIST. Avoidance: Setting NUMDOM to a different value on the DOMAINSOLVER entry 1-4415801 2004 Distributed Memory Parallel - MEFFMASS and FDMODES - SFM 3008 If MEFFMASS is requested in a FDMODES DMP run in both Sol 103 (with DOMAINSOLVER) and Sol 111 (default DMP) the following error message is encountered: SYSTEM FATAL MESSAGE 3008 (EFFMAS) INSUFFICIENT MEMORY AVAILABLE FOR SUBROUTINE EFFMA4 Avoidance: Apply the following alter: compile modefsrs alter ' effmas '(,-1) if ( procid<=1 ) then $ alter ' effmas ' endif $ procid=1 alter ' effmas '(2,-1) if ( procid<=1 ) then $ alter ' effmas '(2) endif $ procid=1 endalter It tested successfully using MEFFMASS & DMP in Sol's 103 & 111. It was further tested with: 1) multiple BC in Sol 103, and 2) Resvec's in Sol 111, and worked well. It is recommended this fix is coded into next v2001 release. -- JoeM 1-4426204 2004 Miscellaneous - UFM 2042 - Missing MATi on property entry UFM 2042 up to 4 times per element if the associated property card references a missing material entry. Furthermore only the 1st 100 elements are flagged, so that problems on subsequent elements/properties are not flagged. This is an enhancement request to re-work what happens when an property entry references a non-existent material. It is recommended that: 1) The UFM identify the improper PROPERTY entry not (each) element that uses that particular property. 2) ALL properties entries that reference missing material entries be flagged, not just the first 100. Small sample deck attached that demonstrates these issues. This issues is part of the Bell Helicopter "list" reference Dennis Hilton 1-4470501 2004 Documentation/QRG - DTABLE Description needs Updating The description of the DTABLE entry in the Quick Reference Guide needs to be updated to indicate that it may be referenced by a DVMREL2 entry. 1-4507201 2004 Distributed Memory Parallel - DMP - Jobs Hang DMP jobs that encounter DMAP fatal errors in subDMAP PHASE0 may hang. DMAP level fatal errors which are handled via NOGO/EXIT in subdmap PHASE0 are not handled correctly. Avoidance: Insert the following DMAP alter: compile phase0 alter 'if ( stop0'(,4) if ( nproc>1 ) then $ param //'stsr'/1/-3 $ putsys(0,307) $ endif $ nproc>1 endalter 1-4570801 2004 Documentation/QRG - PARM,OPPHIPA The QRG documentation of the PARAM OPPHIPA should indicate that the resulting normal modes are output in the global coordinates and refer to OPPHIB for the option to have basic output. 1-4590301 2004 Axisymmetric Fluid Elements - PRESPT, RINGFL, CFLUIDi - Wrong Answers on Cray Reran the following tests in v2005t1 on nihau and updated archives(v2004,v2005): csr7377.f05 d07d2.f05 d10702.f05 e2280.f05 hydr26.f05 hydr27.f05 p3b1.f05 p3b2.f05 parcmp.f05 s32585a.f05 s32585b.f05 s34774c.f05 s34774c1.f05 s34774d.f05 s34774d1.f05 abb 1/12/04 Fluid pressure output at PRESPT points for axisymmetric fluid elements is incorrect on all short word machines. Only the Cray gives correct answers. 1-4635614 2004 Documentation - Vecplot - # of input/output datablocks In section 10.3 of the MSC.Nastran 2001 Release Guide, "Summary of DMAP Module changes from V70.7 to MSC.Nastran V2001", the VECPLOT module description for v2001 should be changed from VECPLOT XG,BGPDT,SCSTM,CSTM,CASECC,{LAMA or USET}, MEDGE,x66p,vgq,rbf/ XOUT,RESMAX,HEADCNTL,x66/ S,N,GRDPNT/COORID/IOPT/TITLE1/TITLE2/TITLE3/ ALTSHAPE/WTMASS/SEID/setnam/rbfail $ to VECPLOT XG,BGPDT,SCSTM,CSTM,CASECC,{LAMA or USET}, MEDGE,x66p,vgq/ XOUT,RESMAX,HEADCNTL,x66,gcfxx/ S,N,GRDPNT/COORID/IOPT/TITLE1/TITLE2/TITLE3/ ALTSHAPE/WTMASS/SEID/setnam/rbfail $ wher GCFxx is the ground check force. 1-4668001 2004 Documentation/QRG - INCLUDE - Long File Names MSC.Nastran 70.7 & 2001 INCLUDE statement can be longer than 72 characters, even though the QRG indicates there is a 72 character limit (Remark 2 in the INCLUDE keyword documentation) Avoidance: The MSC.Nastran Configuration and Operations Guide, Section 4.9 indicates how to use long filenames. 1-4672328 2004 Enforced Motion - SDAMP, TABDMP1 - Wrong Answers If a high modal damping level is used with the new enforced motion capability in MSC.Nastran 2001, then the answers may be incorrect. The effect on results is not significant unless the modal damping value is high. Avoidance: There are two possible avoidances: 1. Approximate the modal damping with structural damping. See quality 1-5399310 2. Put the entire structure in a superelement and apply the modal damping to the superelement. 1-4674201 2004 Documentation/QRG - PARTN Case Control Command Disabled In coupled fluid structural analysis the PARTN Case Control command has been disabled for Version 2001 but the Remark 2 under the QRG description of this command was not removed. This capability is being replaced by the FSLPOUT command. 1-4717576 2004 Elements - GEOMCHECK - UWM 9031, Poor Message If a model contains elements which cause unusable matrices, the default value of INFORM on the GEOMCHECK statement will not print a FATAL message. Therefore, the run might fail later in the process with a message which does not indicate the elements which caused the problem. In one example, a user encountered the following message with multiple elements which gave bad jacobians: ^^^ USER WARNING MESSAGE 9031 (ERRPH1) ^^^ NOGO ENCOUNTERED IN SUBDMAP SEMG Avoidance: Use the FATAL option on GEOMCHECK rather than try to run a model with bad elements. 1-4724901 2004 Documentation/QRG - TLOAD1/TLOAD2 - Missing reference to TEMPBC Remark 4 under the description of the TLOAD1 and TLOAD2 entry in the MSC.Nastran 2001 Quick Reference Guide should include the TEMPBC entry: 4. EXCITEDID may reference sets containing QHBDY, QBDYi, QVECT, QVOL and TEMPBC entries when using the heat transfer option. 1-4733501 2004 Frequency Response - SPCD, OFREQ, PARAM,DDRMM,-1 - SFM 5423 (DMPYAD) When OFREQ is used with PARAM,DDRMM,-1 to output ESE (or EKE/EDE) in an SPCD Sol 111 run, get the FATAL error: *** SYSTEM FATAL MESSAGE 5423 (DMPYAD) ATTEMPT TO MULTIPLY INCOMPATIBLE MATRICES Avoidance: Apply the following dmap avoidance: compile disprs alter 15,15 $ V2001 if( otf>0 and noypo>0 ) modacc casef,fol,ypo,,,/ fol00,ypo1,,,/app $ alter 30,30 $ V2001 if ( otf=0 and noypo>0 ) equivx ypo/ypo1/-1 $ alter 89,89 $ V2001 CALL SPMPYAD RPH1,UHF1,YPO1,V01DR/UPX/0/1 $ 1-4734006 2004 Superelements - CSUPER, EXTRN - UFM 2065 MTRXIN If DMIGDB is used when creating external superelements and they are brought into the analysis using the CSUPER entry, the program will issue the message: USER FATAL MESSAGE 2065 (MTRXIN) if SPOINTs are referenced. Avoidance: Insert the following DMAP alter: compile phase0 alter 'GP1 GEOM1I','GP1 GEOM1I' GP1 GEOM1I,geom2scr,,,,,/ XGPLS,XEQEXINS,XGPDTS,XCSTMS,XBGPDTS, XSILS,XVGFS,,/ S,N,XLUSETS/0/NOGPDT/////SEID $ endalter 1-4734222 2004 Complex Eigenvalue Analysis - EIGC, CLAN - SIM 6491, Wrong Answers A complex eigenvalue analysis that produced reasonable answers in Version 70.5 is producing wrong answers in Version 2001 using complex Lanczos after issuing the message: *** SYSTEM INFORMATION MESSAGE 6941 (CLSCLD) INVARIANT SUBSPACE DETECTED IN BLOCK LANCZOS. Avoidance: The problem is still under investigation - using SOL 110 with HESS produced the same results as the Lanczos method in Version 70.5. Setting SYSTEM(108)=2 to force the old method or setting SYSTEM(108)=64 to force full orthoganalization also avoided the problem. 1-4747001 2004 Eigenvalue Analysis - PARAM,MECHFIX - Wrong Answers If PARAM,MECHFIX,YES is set or if MSC.Nastran encounters difficulties in the eigenvalue analysis and it automatically enters the massless mechanism control but no massless mechanisms are found, then rigid body modes may be incorrectly constrained. Avoidance: See SSSALTER mmfindfix.v2001 on the SSSALTER web page for an avoidance. Another avoidance is to use param, mechfix, no, and fix any mechanism errors that appear manually. 1-4748439 2004 Miscellaneous - OUTPUT2 - Missing Modes If a user program is designed to read the element strain energy output from an OUTPUT2 (op2) file created by Version 2001, it is possible that the program might only find the first mode. Avoidance: This problem is actually due to an error fix in 2001. In the table description in the NDDL, it indicates that word 5 in the IDENT record of the element strain energy is the mode number. In Version 2001, this is correct. In previous versions, the mode number was incorrectly written in word 4. Many programs which read OUTPUT2 files are looking for the mode id in word 4 and will not find it in a v2001 OUTPUT2 file. The reading program needs to be updated. 1-4807909 2004 Modal Effective Mass - MEFFMASS - Wrong Answers for Rotational Inertia Terms The output for Modal effective mass (MEFFMASS) may have wrong values for the rotational inertia terms. Avoidance: Use the alter effmassa.v707 from the V70.7 sssalter directory. 1-4830716 2004 Superelements - PARAM,RESVEC - Residual Vectors not Calculated If a model has superelements and PARAM,RESVEC,YES is specified in the main bulk data section, it is possible that the program will not calculate residual vectors. Avoidance: Either move PARAM,RESVEC,YES into the case control above all subcases or include the following DMAP alter in your file: compile phase1dr alter 'sep2dr.*slist.*emap' $ line 80 pvt pvts,//s,n,dummy21 $ 1-4833901 2004 Distributed Memory Parallel - OMODES, OFREQ - SFM 3001 (MODACC) The presence of an OMODES or OFREQ case control request will result in a SYSTEM FATAL MESSAGE 3001 from the MODACC module, if DOMAINSOLVER GDMODES is used. This is a DMP method for SOL 103. Avoidance: Workaround is to apply the following alter: compile super3 $ alter 'EQUIVX CASEcc\/CASEDR' $ EQUIVX UL/ULS/-1 $ compile sedisp $ alter 'equivx ul\/un'(2),''(2,1) $ equivx ul1/un/nosset $ if ( nosset>-1 ) umerge uset,ys,ul1/un/'N'/'S' $ endalter 1-4837703 2004 Documentation/QRG - DRESP3 - Missing Documentation The DRESP3 (External Response) capability was developed as a funded project that was intentionally not documented in the 2001 Release Guide but can now be discussed. The design optimization capability (SOL 200) in MSC.Nastran has a powerful DRESP2 feature allowing the user to create a synthetic response. However, the types of variables that a synthetic response can use are limited to the data from MSC.Nastran. The external response feature extends the synthetic response by allowing the user to define a custom response using either in-house programs or any application programs written in Fortran, C or other languages. This enables the use of proprietary responses as the objective or a constraint in a design. The documentation for this feature is available in Solution 1-3772901 1-4854701 2004 Frequency Response - CRAC2D, CRAC3D - SFM 6984 The CRAC2D(ADUM8) and the CRAC3D(ADUM9) elements are not supposed to have any frequency (SDR2) data recovery and should simply return when running frequency response such as SOL 108. Instead they crash with SYSTEM FATAL MESSAGE 6984 (SDR2RD) INCORRECT NUMBER OF WORDS BEING REQUESTED FOR TRANSMISSION TO AND BY PHASE-2 ROUTINE. 1-4855904 2004 Documentation/QRG - Incorrect default value for SYSTEM(166) The Quick Reference Guide incorrectly states that the default value for SYSTEM cell 166 is 1. The actual default is 0. The following alter will allow you to view the default value: compile seaero alter 1 type parm,,i,n,sys166 $ sys166 =getsys(sys166,166) $ message //' the value of system cell 166 is '/sys166 $ exit $ OPTIONAL endalter $ $ 1-4866203 2004 Documentation/QRG - PARAM,OUMU is for ESE not EKE. The Quick Reference Guide incorrectly states that parameter OUMU is for element kinetic energy. It is for element strain energy. 1-4866209 2004 Eigenvalue Analysis - PARAM,POST,-1 - Extra Output If you specify PARAM,POST,-1 and PARAM,POSTEXT,NO in SOL 103, the program will still unnecessarily calculate and write ONRGY2 to the OUTPUT2 file if you have requested ESE output. Avoidance: Set PARAM,OUMU,NO and the datablock will not be calculated or written. 1-4891701 2004 Materials - MAT1 - Large Property Values may lead to overflow. Large magnitudes for the values of the material properties on the MAT1 bulk data entry can cause floating point overflows in the material property lookup routines. Avoidance: Be sure that the magnitudes of the properties are less than the SQRT of the largest magnitude number representable on the machine. 1-4966701 2004 Eigenvalue Analysis - MEFFMASS and MAX normalization - Possible Wrong Answers If the max option is used in the eigenvalue calculation, then the modal effective mass output will be off by the generalized mass term. The modal effective mass fraction, however, is correct. Avoidance: Use the default mass option or insert the DMAP alter found in the attachments. 1-5001808 2004 Miscellaneous - READ, PARAM,RESVEC - Floating Point Exception If all modes of a model are calculated and residual vectors are requested, the run will fail in the second eigenvalue solution with a floating point exception. Avoidance: This error only occurs when the static solution cannot be represented by a combination of all modes of the model. Then the residual vector contains only dof which have no mass associated with them (and therefore, an infinite eigenvalue). The avoidance is to not request residual vectors when all modes of the model are found. 1-5014405 2004 Cyclic Buckling - SOL 116 - Wrong Answers In SOL 116 (buckling with cyclic symmetry), boundary conditions may not be changed between the static preload and the buckling eigenanalysis subcases leading to wrong answers. Avoidance: Insert the following DMAP alter: COMPILE CYCBUCKL ALTER 'ISOLAPP=2' $ CALL SETQ CASERS// SEID/PEID/S,MTEMP/S,K2GG/S,M2GG/S,B2GG/S,MPC/ S,SPC/S,LOAD/S,DEFORM/S,TEMPLD/S,P2G/S,DYRD/ S,METH/S,MFLUID $ DELETE /CYCD,,,, $ CYCLIC2 GEOM4S,EQEXINS,USET/CYCD/NSEG/CTYPE $ endalter 1-5023407 2004 SSSALTERS - rigyroa alter sync=no - Wrong Answers Complex eigenvalues using the ridgyroa alter with sync=no and Lanczos causes incorrect mode shapes in v70.7 and v2001. Avoidance: Use system(108)=2 or the Hessenberg method. 1-5023428 2004 Elements - CBUSH - Unexpected Answers If the user defines a cbush entry with the same grid id for end A and end B, the cbush will be grounded. This input should result in a fatal error. 1-5023444 2004 Miscellaneous - Incorrect ELSUM results for PCOMP In V2001, using the ELSUM Case Control command with a plate element that references a PCOMP entry using the BEND option on the LAM field causes wrong large numbers in the Structural Mass field of the ELSUM printout. 1-5089207 2004 Documentation/DMAP - Error in Description of OEE datablock. In the NDDL description of datablock OEE in the DMAP Module Dictionary, there is an error in the comment notes describing the approach codes. The values for approach codes 1 thru 5 are correct, but Approach code =6 should be Transient (not Buckling part 1) the values after this should be increased by 1. IE. Approach Code Solution Type 1 Statics 2 Reigen 3 DS0 4 DS1 5 FREQ 6 TRANS 7 BKL0 8 BKL1 9 Complex Eigen 10 NonLinear Statics etc 1-5093401 2004 Optimization/Design Sensitivity - Poor Message when DRESP2 incorrectly refers to DESVAR labels Optimization run fail with no meaningful message when DRESP2 entries incorrectly refer to DESVAR labels rather than ids. The following is an example of the error message in log files: for SUNS iio: [-1] end of file lately: writing sequential formatted internal IO part of last string: | time: command terminated abnormally. for HP *** FORTRAN I/O ERROR 958: FORMAT DESCRIPTOR INCOMPATIBLE WITH CHARACTER ITEM IN I/O LIST INTERNAL FILE. LAST FORMAT: (I8,"|¦È€ There are no messages in the f04 file and f06 file. The job fails with the correct error message on NT and in version 70.7 on unix. 1-5155018 2004 Elements - PBEAML, HAT1 - Incorrect J Value The QRG description of the HAT1 section of the PBEAML and PBARL incorrectly describes dimensions as "Di" instead of "DIMi" and is missing the length of the horizontal leg. In addition, for both the HAT and the HAT1 sections, the torsional moment of inertia incorrectly uses the open section equations instead of closed section equations. This is incorrect for the HAT1 because it is shown as closed and for the HAT because hats are by definition always attached to skins. A remark needs to be added to this effect. In the unlikely event that a hat is used as open (for instance, upside down), an option needs to be added to select an open section calculation. 1-5162113 2004 Element Strain Energy - SFM 4276 EC 31 in GPFDR if ESE not in all REPCASEs If a file contains REPCASEs and ESE output is requested in some, but not all, REPCASEs and SUBCASEs, the run may fail with SFM 4276 EC 31. Avoidance: Place an ESE request in all SUBCASEs and REPCASEs. 1-5174501 2004 Elements - PBARL and PBEAML HAT1 Sections yields incorrect J value. The J calculated by MSC.Nastran for the PBARL and PBEAML HAT1 section is wrong. 1-5233709 2004 Dynamics - MEFFMASS - SFM 3007 (EFFMAS) If effective mass is requested and there is no s-set in the model, the job will fail with the following error messages: *** SYSTEM FATAL MESSAGE 3007 (EFFMAS) ILLEGAL INPUT TO SUBROUTINE PARTN Avoidance: Add and constrain a dummy grid point in the model. 1-5235204 2004 Eigenvalue Analysis - PARAM,RESVEC - Traceback in READ Mike, RESVEC has been reworked by Bella & Wilson. I suggest we remove the blanket condemnation and try to claim that all problems are solved in v2003. PJS 6/9/03 Eigenvalue jobs with residual vectors (param,resvec) may fail ungracefully on small models. In one example the job issued a traceback in the READ module. Avoidance: Do no specify residual vectors for small models. 1-5241491 2004 Elements - MAT2, PSHELL - UFM 4297 (EQD4D) If a MAT2 entry has no elastic properties and is referenced by a PSHELL entry, then a misleading message similar to the following may occur: *** USER FATAL MESSAGE 4297 (EQD4D) QUAD4 ELEMENT WITH ID = 16772653 HAS AN INTERIOR ANGLE POSSIBLY GREATER THAN OR EQUAL 180 DEGREES. Avoidance: Specify realistic material properties. The message should be changed to read: *** USER FATAL MESSAGE 4297 (EQD4D) QUAD4 ELEMENT WITH ID = 16772653 CAUSES ARE: 1) THIS ELEMENT HAS AN INTERIOR ANGLE POSSIBLY GREATER THAN OR EQUAL 180 DEGREES. 2) NO ELASTIC PROPERTIES ARE ENTERED 3) G12 IS BLANK ON THE MAT8 ENTRY 4) G33 IS BLANK ON THE MAT2 ENTRY 5) MAT2 IN NOT POSITIVE DEFINITE 1-5271102 2004 ACMS - PARAM,POST - Can not Post Process Results In ACMS for certain post-processing programs, the data block names don't match the names expected by the post-processor. This prevents these data blocks from being imported into the post-processing program. Note that Patran can read these names without any problems. Avoidance: Please contact MSC Support for A DMAP alter avoidance. 1-5284424 2004 Optimization/Design Sensitivity - PARAM,DESPCH1,-4 - Missing DRESP1 entries in punch output The punch file does not contain the DRESP1 entries when using param,despch1,-4 Avoidance: This is a documentation error. The design intent is that only mode tracking related (RTYPE=FREQ and/or EIGN) are to be printed when despch1 = + or -4 1-5328702 2004 Miscellaneous - CQUAD4, MAT1 - ELSUM Incorrect with non-structural mass When a MAT1 entry specifies nonstructural mass on a CQUAD4 element, ELSUM incorrectly calculates the element and total mass. 1-5328715 2004 Acoustic Analysis - UWM 9031 Nastran Version 70.7.2 gives a clear message (UFM 4659) that a fluid element is badly shaped, Nastran 2001 fails with ^^^ USER WARNING MESSAGE 9031 (ERRPH1) ^^^ NOGO ENCOUNTERED IN SUBDMAP SEMG without any indication of the problem. 1-5343701 2004 Optimization/Design Sensitivity - ANALYSIS=MTRANS - Enforced Motion results in SFM 5423 If SOL 200 is requesting modal transient analysis and enforced motion and the skip factor on the TSTEP entry is not unity then the run terminates with SFM 5423 in the MPYAD module. Avoidance: Insert the following DMAP alter: compile fea alter 'trlg'(2),'' TRLG CASErs,USETD,dlt,,BGPDTS,SILS,CSTMS,TRL,dit, ,,,,,,/ xppt,,,,,TOL,,yptt,ypot/////////1 $ alter 'call disprs casers,tol1,uhtds,','' call disprs casers,tol1,uhtds,,ppt1,usetd,rph,rpd, phdh,k2dd,b2dd,mdd,lll,dar, uset,rgm,mmg1,bmg1,kmg1,, rgs,msf1,bsf1,ksf1,,kss1,mss1,bss1,ypot,mechg,,,,/ tol2,uht1,pgt1,upt,qpt,qmpt,ypo1/ app/app1/soltyp/drmm/noa/noue/dospcf/dompcf/noph2/ true/0/false $ endalter 1-5361602 2004 Heat transfer - SOL 101 - SFM 3007(OFP),set NASTRAN system(210)=1 The following error message may be encountered in heat transfer analyis or in static analysis with APP HEAT: *** SYSTEM WARNING MESSAGE 6433 (ZREAD) THE DECODED C-ITEM "TCODE" OF DATA BLOCK=OEF1X , RECORD=IDENT. EXCEEDS NORMAL LIMITS "J"=**** "I"= 496. PROGRAMMER INFORMATION: PROBABLE CAUSE IS WRONG POSITION OF TABLE CODE, OR EARLIER RESET OF (C) ITEMS. *** SYSTEM WARNING MESSAGE 3030 (OFP) THE OFP MODULE CANNOT PRINT THE DATABLOCK OEF1X USER ACTION: USE THE MATPRN OR TABPT MODULE. PROGRAMMER INFORMATION: OFP LOC CODE = 510 FILE = 104 *** SYSTEM FATAL MESSAGE 3007 (OFP) ILLEGAL INPUT TO SUBROUTINE ZREAD Avoidance: This error only occurs when SYSTEM(201)=1, that is when the NEWPUNCH option is selected. This option is the default in 2001, so SYSTEM(210) must be set explicitly to 0 in 2001 to avoid the error. 1-5372208 2004 Complex Eigenvalue Analysis - EIGC - Infinite Root Aborts UWM 6137 Degrees of freedom with no stiffness may cause complex Lanczos to fail with the following message: *** USER FATAL MESSAGE 6138 (UDSFBI) INSUFFICIENT CORE FOR SPARSE FBS USER ACTION: INCREASE CORE BY 1002 WORDS Avoidance: Apply the following dmap alter: compile ceigrs list $ alter 'cead.*cphdx'(,-1) $ putsys(0,209) $ use active column solver 1-5377220 2004 Optimization/Design Sensitivity - MODTRK - UFM 6656 When running optimization with discrete design variables and mode tracking is used, the following error is obtained: *** USER FATAL MESSAGE 6656 (MODTRK) THE TRACKED SUBSET PHGREF OF MODES FROM A PREVIOUS DESIGN CYCLE CAN NOT BE FOUND. Avoidance: A dmap alter is available on request. 1-5378704 2004 Documentation/QRG - Elem ID - UFM 316 The explicit grid id and element id must be a positive integer greater than 0 and less than 100,000,000. This is documented for the grid id in the Quick Reference Guide. However, for element id, the QRG only specfies the lower bound, but does not the upper bound of 100,000,000. If an element ID of 100,000,000 or greater is used, then the following error messages will be issued and the results are unpredictable. *** USER FATAL MESSAGE 316 (IFPDRV) Note that internally, nastran can generate larger values (e.g., p-elements). Avoidance: Define the element id as (0< Integer < 100,000,000) 1-5386201 2004 Transient Heat Transfer - Models with small relative stiffness fail to converge. In the transient thermal analysis, varying temperature boundary conditions require that a value of 1.0e10 be used for the stiffness in conjuction with SLOAD entry in order to apply time-varying temperature BC. If the model has very small stiffness, then the transient solution will not converge. The 1.0e10 is defined in the code and can not be changed by the user. Avoidance: Change all the TEMPBC,TRAN into SLOAD and CELAS2 combinations. Else, see the newly implemented NASTRAN TBCMAG=1.0e3 in Nastran Version 2004 1-5386817 2004 Stress Sorting/Filering - PARAM,NUMOUT2,0 causes strange characters in the f06 output PARAM,NUMOUT2=0, used for filtering stress output according to PARAM,BIGER, may result in strange characters in the f06 file. Avoidance: Use PARAM,NUMOUT2=-1. 1-5399310 2004 Enforced Motion - SPCFORCEs - Wrong Answers Wrong SPCForces at the enforced motion dofs will result in SOLs 111, 112, and 200 if any of PARAMs G, W3, or W4 are specified. Avoidance: Insert the following DMAP alter: compile disprs alter 'call CFORCE ','' call CFORCE2 USET,rgs,yso,MSs1,BSs1, KSs1,,fOL1,,,,/ Qpy/ APP/APP /'spcforce'/true $ endalter If the DMAP option is not available then add param,ddrmm,-1 to the input file. 1-5461501 2004 DMAP - INPUTT2 - SFM 2190 if the file being read does not have a label. If you attempt to read data from a file without a label using a value of -1 or -3 for ITAPE, then the run will fail with: *** SYSTEM FATAL MESSAGE 2190 (INPTN2) ILLEGAL VALUE FOR KEY = 2, EXPECTED VALUE = 3. Either write a label on the file when creating it, or do not use -1 or -3 for ITAPE 1-5461571 2004 Restarts - PARAM,RESVEC - Potential wrong answers If residual vectors are requested on a restart run, the program will not calculate them unless the eigenvalue solution is re-solved. There will be no warning or information messages issued. If enforced motion is used (V2001+), this may result in incorrect answers if a restart is made from SOL 103 to perform the enforced motion solution. Avoidance: The following DMAP alter will force the program to re-calculate the residual vectors on a restart (inlcuding any based on the new loading). compile modefsrs $ alter 'call resvec'(,-1) type parm,,i,n,nophsa dbstatus phsa//s,n,nophsa $ if(nophsa>0)then $ copy phsa/phsa0 $ copy lama/lama0 $ delete /phsa,lamas $ endif $ endalter 1-5467701 2004 Superelement Analysis - Enforced Dynamic Motion - SFE 4276 in DPD Superelement analysis may fail in DPD with the message: SYSTEM FATAL MESSAGE 4276 (UNPACK) when enforced dynamic motion is specified. Avoidance: A workaround is available by including the following DMAP alter: compile phase1b $ alter 'noys' $ paraml pg//'trailer'/1/s,n,noys $ endalter $ 1-5500401 2004 Optimization/Design Sensitivity - RESTART - BIOMSG: ERROR 31 Section 7.11 of the V68 MSC/Nastran User's Guide for Design Sensitivity and Optimization illustrates how to use the restart feature of MSC.Nastran in SOL 200. This capablity is broken with the result that the example case does not work and will result in a message such as: BIOMSG: ERROR 31 HAS OCCURRED IN ROUTINE UNPACK Avoidance: Insert the following DMAP alter in the restart run: compile desopt alter 'call predom' DELETE /RSP2RG,CNTABRG,R1TABRG,DRSTBLG,FRQRPRG $ DELETE /R1VALRG,R2VALRG,CVALRG,wmidg,r3valrg $ DELETE /RESP3RG,rr2idrg,,, $ DELETE /DSCMG,DELW,DELV,DELB,DELF $ compile descon alter 'dbstatus' NR1OFFST = 0 $ NR2OFFST = 0 $ NCNOFFST = 0 $ 1-5504006 2004 Acoustic Analysis - Modeling Errors lead to Wrong Answers with No Warning or Fatal Messages Fluid elements that have geometry should give a fatal message if stucture points are connected to them but they do not. This includes all fluid elements except the single GRID CAABSF, the CHACAB, and the CHACBR. The CHACAB and CHACBR fluid elements have their own rules that must be followed where four structure GRIDs must be on one side of the hex and four fluid GRIDs must be on the other. Structure elements with geometry must give fatal errors when connected to fluid GRIDs. This includes all structure elements except elements such as CELASi, CDAMPi, CMASSi, CBUSH. When the above modeling error was present, different results were obtained between SOL 103 and SOL 111. 1-5533601 2004 Documentation/QRG - SYSTEM(357) - Ignore Continuation Fields SYSTEM(357) is not documented in the MSC.Nastran 2001 Quick Reference Guide. This system cell was added for a user request. When set to 1 continuation fields will be ignored and treated as blank. 1-5569401 2004 Elements - CWELD - Wrong Answers Incorrect CWELD element forces are output if any of the attached grids (the CWELD grids or the connecting shell grids) are in a global coordinate system. 1-5623817 2004 Miscellaneous - Continuation Entries The continuation fields are supposed to support a base of 36 (0-9, A-Z), which allows a maximum of 36**7 continuation entries. However, when the continuation fields are left blank, MSC.Nastran only generates the continuation field using the numbers 0-9 and it starts with 1 and goes to +9999999 and then it starts repeating. This will cause a fatal error if there are more than 9999999 continuation entries. There are two possible avoidances: 1. Set system cell 357 to 1. This is the recommended method. 2. Specify the continuation fields explicitly, then you can use all 36 numbers/characters. Start field 10 of the parent entry with +0000000 and increment the right-most column first (starting from 0,1,2,...9,A,B,...Z). Next increment the 2nd right-most column by 1 and repeat the same thing for column 1. Repeat the same process until this is done for all 7 columns. 1-5668101 2004 Aeroelastic Optimization - Multiple Boundary Conditions lead to GREAD Error in DSAL Aeroelastic optimization with multiple boundary conditions may lead to a GREAD error in the DSAL module. Avoidance: Apply the following DMAP alter: compile respsen list alter 'aug1x' delta=true $ compile pslgdv alter 'caseaa'(2),'caseaa'(2) SSG1 SLT,BGPDVP,CSTMS,medge,ESTDV2F,MPTX,ETTDV,EDT,, caseau,DIT,,,,gpsnts,cstm0,,/ aAG ,PTELMa,/ LUSETS/0/1/'ALL'/altshape $ alter 'caseaa'(3),'caseaa'(3) SSG1 SLT,BGPDVb,CSTMS,,ESTDV2B,MPTXB,ETTDV,EDT,, CASeau,DIT,,,,gpsnts,cstm0,,/ aAGB ,PTELMAB,/LUSETS/0/1/'ALL' $ alter 'pmg'(4),'pmg'(5) dbstatus pma//s,n,nopma$ IF ( nopma<0 ) then $ compile fea alter 'fluidse0'(2) type parm,,i,n, nsolga = 0 alter 'saersens'(,-1) nsolga = nsolga + nsol $ alter 'psload'(6),'psload'(6) IF ( FEAOPT='SENSITIVITY' AND NOT ITOL( NOSEDV ) AND ( ITOL(-SAERODSA) OR ITOL(-STATDSA) ) AND PSLOAD AND ( SAERCC AND NSOLGA>0 ) ) THEN $ 1-5676909 2004 Loads - TEMPD - Duplicate TEMPD ID causes UFM 7522 If an input file has multiple TEMPD entries with the same id, the program will issue UFM 7522, incorrectly indicating that there are conflicts with DEFORM entries. Avoidance: If you get this error message, check all TEMPi, DEFORM, and TEMPD entries for unique ids. 1-5684501 2004 Optimization/Design Sensitivity - DRESP1, ATTB - UFM 5628 (DSAD) If a DRESP1 entry is defined with a function name in the ATTB field it is considered as an implied DRESP2. If an implied DRESP2 selects an element related response such as frstre or frforc, then the job will failed in the DSAD module with UFM 5628. 1-5687106 2004 External Superelements - DMIGOP2 - Previous Versions not Upward Compatible If an external superelement created using DMIGOP2 in a previous version is used in V2001, the run will fail in various ways. Avoidance: If possible, re-create the external superelement using V2001. If not, contact MSC with this error number and request the DMAP avoidance. 1-5704209 2004 Complex Eigenvalues - EIGC - SFM 3002 in Complex Lanczos If complex lanczos method is used and the EIGC entry does not have ND0 specified and has no continuation lines, the program will fail in the CEAD module with: BIOMSG: ERROR 3002 HAS OCCURRED IN ROUTINE FREAD Avoidance: Provide ND0 and/or continuation lines. 1-5784601 2004 Random Analysis - RANDPS with Cross Spectrum - Possible Wrong Answers, SWM 3002 (RANDOM) If a random analysis with a cross spectrum, requests random output for a non-existing grid point, then the answers may be incorrect. The following warning messages are issued: *** SYSTEM WARNING MESSAGE 3002 (RANDOM) EOF ENCOUNTERED WHILE READING DATA BLOCK OUGV2 (FILE 104) IN SUBROUTINE RAND2 *** USER WARNING MESSAGE 979 (XYTRAN) AN XY-OUTPUT REQUEST FOR POINT OR ELEMENT ID xxxx Avoidance: Remove the output request for the non-existing grid point 1-5784622 2004 Nonlinear Analysis - CRAC3D, TEMP(LOAD) - SFM 3003 In a nonlinear analysis is running on an NT and the model contains CRAC3D elements and temperature loads, then the job may fail with the following messages. *** SYSTEM FATAL MESSAGE 3003 (TA1) ATTEMPT TO READ PAST THE END OF A LOGICAL RECORD IN DATA BLOCK EST (FILE 201) IN SUBROUTINE TA1NLE 0FATAL ERROR *** SYSTEM INFORMATION MESSAGE 1156 (XCLNUP) GINO FILE 201 WAS LEFT OPEN AT PROGRAM TERMINATION. 1-5809604 2004 Elements - CWELD and PARAM,POST,-2 - Wrong answers If a CWELD element generates GRID points for the connection, the locations of these new GRID points will be written into the OUTPUT2 file incorrectly if PARAM,POST,-2 is used. Avoidance: Insert the following DMAP alter: compile phase0 alter 'geom1qx' call dbstore geom1qx,,,,//0/0/' '/0 $ compile super3 alter 'gp1','' call dbfetch /geom1qx,,,,/0/0/0/0/0 $ GP1 GEOM1qx,GEOM2,,,,,/ GPL,EQEXIN,GPDT,CSTM,BGPDT,SIL,,,/0/0/0 $ alter 'call.*out2g.*geom2'(2,),'' CALL OUT2GEOM, CSTMs,GPL,GPDT,EPT,MPT,GEOM2s,GEOM3s,GEOM4s, GEOM1s,DIT,DYNAMICS,BGPDTs,CASECC,MATPOOL, AXIC//OUNIT1X/S,NP $ $ endalter 1-5843101 2004 Optimization/Design Sensitivity - UFM 6499 (DOM9D) When an optimization task has more active and violated constraints than the number of design variables, the optimization module may produce the error message: USER FATAL MESSAGE 6499 (DOM9D) or may have difficulty converging. Avoidance: Set METHOD to 3 (SQP) or 2 (SLP) on the DOPTPRM entry. *** USER FATAL MESSAGE 6499 (DOM9D) THE DOT OPTIMIZATION ALGORITHM HAS ENCOUNTERED AN ERROR THAT MAKES FURTHER PROGRESS IMPOSSIBLE. SEE MESSAGES ABOVE. 1-5862001 2004 Nonlinear Transient Analysis - PLOADX1 - Wrong Answers Nonlinear transient analyis should treat the PLOADX1 as a follower force but does not. This may lead to wrong answers. 1-5872004 2004 Nonlinear Analysis - PARAM,LGDISP - CBEAM with Offsets may not produce UFM 6174. A nonlinear analysis with BEAM elements with offsets should produce the following fatal message: *** USER FATAL MESSAGE 6174 (TA1NLE) BEAM ELEMENTS WITH OFFSETS ARE NOT SUPPORTED IN DIFFERENTIAL STIFFNESS OR NONLINEAR ANALYSIS. REMODEL WITHOUT USING BEAMS WITH OFFSET. However, if only some of the BEAMS in the model have offsets the message may not be issued and poor answers may be produced. 1-5872801 2004 Optimization/Design Sensitivity - STATSUB - UFM 6468 (GETCOL) USER FATAL MESSAGE 6468 may occur in an optimization analysis when multiple buckling subcases with separate STASUB requests are employed. Avoidance: The following 2001 DMAP should avoid the problem: compile respsen alter 'casebuck','dsnokd'(3) DSVG1 KDICTDSF,KELMDSF,BGPDTS,SILS,CSTMS,KDICT, KELM,BUDG,,LBTAB,DSPT1/ EGKBK///2/0 $ IF ( cdifx='yes' ) THEN $ DSVG1 KDICTDSB,KELMDSB,bgpdts,sils,cstms,kdict, kelm,budg,,lbtab,dspt1/ EGKBKB///2/0 $ ADD5 egkbk,EGKBKB,,,/EGKBKC/.5/-.5 $ EQUIVX EGKBKC/EGKBK/-1 $ ENDIF $ cdifx='yes' if ( dsnokd<>0.0 ) then $ dbview casebuck=casers (where isolapp=2) $ IF ( SEID=0 ) THEN $ PARAML CASEBUCK//'DTI'/1/3 //S,N,SPC $ PARAML CASEBUCK//'DTI'/1/2 //S,N,MPC $ PARAML CASEBUCK//'DTI'/1/255//S,N,SUPORT $ PARAML CASEBUCK//'DTI'/1/5 //S,N,BMETH $ PARAML CASEbuck//'DTI'/1/256//S,N,STATSUB $ ENDIF $ SEID=0 GETCOL CASEbuck,CASE0//1/S,N,COLNO $ PARAML CASESX//'DTI'/1/7//S,N,TSET $ EMG EST,CSTMS,MPTS,DIT,,UG,ETT,EDT,,,BGPDTS, GPSNTS,,,,/KDELM,KDDICT,,,,/ 1/0/0//1//Tset/-1//////COLNO $ 1-5884008 2004 Optimization/Design Sensitivity - Poor Convergence An optimization job using the modified feasible method may encounter divergent behavior when it starts with a feasible design and degrades to an infeasible design. Avoidance: It may help to scale the objective to unit. 1-5884017 2004 Nonlinear Transiet Analysis - TSTEPNL - SFM 4276 (NLTRD2) Nonlinear transient analysis may fail in the NLTRD2 module with SYSTEM FATAL MESSAGE 4276 (as shown below) if multiple load cases exist and NO (the time step interval for output) is set greater than 1 on the TSTEPNL entry. FATAL 4276 ERROR CODE 1044 (BIOMSG: ..... 1-5884081 2004 Nonlinear Transient Analysis - SFM 4276 EC 1044 If output is not requested for all time steps in a nonlinear transient analysis the job may fail with: SYSTEM FATAL MESSAGE 4276 ERROR CODE 1044 Avoidance: Request output for all time steps. 1-5896512 2004 Element Strain Energy - ESE(PUNCH) - Extra Line added to puch output In Version 2001, if ESE(PUNCH) is requested, an additional line has been added to the output. This line identifies the mode number and eigenvalue for the subsequent output. There are two known problems with this additional line: 1) This additional line might cause some post-processors to fail to read the punch file. 2) It is possible that the printed eigenvalue may be incorrect Avoidance: 1) Currently there is no way to obtain the old format, so the avoidance is to edit the file or write a program to strip out the additional lines. 2) Verify the eigenvalues from the .f06 file 1-5905907 2004 Superelements - EXTSEOUT(OTM) - Missing Output A SOL 111 external superelement creation run using the EXTSEOUT(OTM) feature followed by a SOL 111 assembly run with six dynamic load cases gives results for only the first five subcases for the external superelements in the assembly run. The error only occurs when there is only one forcing frequency. Avoidance: Specify more than one forcing frequency. 1-5905910 2004 Superelements - EXTSEOUT - Missing Output A SOL 111 external superelement creation run using the EXTSEOUT feature followed by a SOL 111 assembly run with two FREQ requests in the Case Control does not produce any output for the external superelements in the assembly run. Avoidance: Insert the following DMAP alter: compile super3 alter 'norec','' paraml casers//'trai'/1/s,n,norec//s,n,nocasers $ if ( nocasers<0 ) paraml casers1//'trai'/1/s,n,norec $ compile semfreq alter 'dloop=dloop','' paraml casef//'trai'/1/s,n,norec $ dloop=dloop+norec $ endalter 1-5927813 2004 Miscellaneous - Large Element IDs - UFM 316 In MSC.Nastran 2001, if an element ID is equal to or greater than 100001001, then the following fatal message will be issued *** USER FATAL MESSAGE 316 (IFPDRV) ILLEGAL DATA ON BULK DATA ENTRY CBAR ON SORTED ENTRY COUNT 2 USER INFORMATION: ELEMENT ID xxxxxxxxx EXCEEDS MAXIMUM VALUE ALLOWED 100001001. The error is that the avoidance is not given in the message. Avoidance: If such a large value must be uses, then insert the following NASTRAN statement at the top of the input file NASTRAN SYSTEM(182)=nnn where nnn is a number larger than the largest element id needed. See quality 5447 for related limits. 1-5963032 2004 Documentation/QRG - ACCELERATION - Output is available for SOL 103 for Response Spectrum Analysis In the MSC.Nastran Quick Reference Guide, Remark 2 on the ACCELERATION case control command incorrectly states that ACCELERATION output is only available in Frequency response and Transient response. Table 4-5 in the QRG correctly indicates that acceleration output is available in SOL 103. However, acceleration output is only available when the response spectrum method is used. 1-6030001 2004 Nonlinear Analsyis - Incorrect Intermediate Composite Ply Stresses for Thermal Loads When composite ply stresses are requested for SOL 106, the contribution for the thermal strain may be calculated incorrectly for intermediate load steps. It is correct for the final load step. Avoidance: There is no avoidance. 1-6047016 2004 Documentation/QRG - GPFORCE \The following line should be added to the end of Remarks 2 and 3 of the MSC.Nastran Quick Reference Guide description of the GPFORCE entry. See remark 4 for further details. Furthermore, CQUAD4 should be added to the list of elements supported for the output of grid point forces aligned with the edge of the element. 1-6047029 2004 Documentation/QRG - Strain Failure Indices Direction Code not Documented When using the strain failure criteria for composite elements, the -1, -2, or -12 output follows the failure indices printout. The intepretations of these values are not documented anywhere. They can be intepreted as follows: -1 - failure in the fiber direction -2 - failure in the matrix direction -12 - failure in the inplane shear. 1-6077806 2004 Complex Eigenvalue Analysis - EIGC, NDJ - Poor Message, BIOMSG Error 3002 As stated in the Quick Reference Guide, the NDJ field must not be left blank on the EIGC entry when using the complex Lanczos method If the NDJ field is left blank on the EIGC entry, then the job will fail with the following unclear messages: BIOMSG: ERROR 3002 HAS OCCURRED IN ROUTINE FREAD , FILE INDEX = 0. STATUS = 0 Avoidance: Specify the appropriate NDJ field. 1-6077837 2004 Documentation - WEIGHTCHECK - punch option not supported The WEIGHTCHECK command does not support the punch option using the standard case control command WEIGHTCHECK(punch...)=yes as illustrated in the Quick Reference Guide. Avoidance: If the WEIGHTCHECK output is desired to be send to the punch file, request it in the print format and include the following alter. $ compile semg1 $ alter 2 $ type parm,,i,n,sys2 $ sys2=getsys(sys2,2) $ putsys(7,2) $ alter 'return'(,-1) $ putsys(sys2,2) $ $ compile semrm $ alter 'call.*semr2'(,-1) $ type parm,,i,n,sys2 $ sys2=getsys(sys2,2) $ putsys(7,2) $ alter 'call.*semr2' $ putsys(sys2,2) $ $ 1-6166948 2004 Documentation/QRG - PGAP - Descripion of Kb field is wrong The QRG description of the PGAP's Kb field should have the following default: "Default = 10e-14*KA" instead of: "Default = MU1*KA" 1-6181949 2004 Nonlinear Analysis - PCOMP, Unsymmetric Ply Thickness, Negative Factor Diagonal QUAD4 elements using offsets fail to converge when using SOL4 for large displacements. The same model with QUAD4 elements runs correctly in SOL 106, also TRIA3 elements with offsets run correctly in SOL 4. 1-6229106 2004 Frequency Response/Random Analysis - STRAIN(PSDF) - Wrong Answers for Strain PSDF. When there are both XYPLOT requests for STRAIN PSDF and STRESS PSDF, the strain results are incorrect and in fact are identical to the stress results. Avoidance: Remove the XYPLOT commands for STRESS PSDF output. 1-6242809 2004 Frequency Response - SPCD - Incorrect SPCFORCE plot results with damping If modal damping is present in a modal frequency response analysis, then the SPCFORCES generated by an XYPLOT request will be wrong unless there is a corresponding SPCFORCE case control request. See also quality 1-4672328. 1-6242834 2004 Dynamic Analysis - MPCFORCE, enforced motion - Possible Incorrect mpcforce output In a modal response analysis (SOLs 111, 112), if the new enforced motion method (introduced in V2001) is used, then the MPCForce output at grid points which are dependent upon the enforced motion dof will be incorrect. Avoidance: Either use param,ddrmm,-1 (not recommended for performance reasons), the direct solutions (SOLs 108,109), or apply the following alter for V2001: $v2001 avoidance for error 1-6242834 compile disprs alter 'call cforce2 USET,rgm,Uhf1' type db mmg,bmg,kmg $ if ( noypo>0 ) then $ call cforce2 USET,rgm,ygo,MMG,BMG,KMG,,fOL1,,, vmpcg,/ Qmpfy/ APP/APP/'mpcforce'/true $ call xadd5 qmpfy,qmpf,,,/qmpfx/cu/cu/cz/cz/cz $ equivx qmpfx/qmpf/-1 $ endif $ noypo>0 endalter keywords: spcd ddrmm mpcforce 1-6273303 2004 Transient Analysis - TRD1 - Wrong Answers Transient analysis SOL 109 may, in rare cases, produce wrong answers in MSC.Nastran Versions 70.7 and 2001. This can only happen when the following message appears: UIM 5222 (TRD1) COUPLED SOLUTION ALGORITHM USED. The model that encountered this error had TF entries, which created a "BDD" matrix with null columns. Avoidance: If you suspect wrong answers you can apply the following dmap avoidance or execute Version 70.5 to see if the answers change. COMPILE SEDTRAN $ Avoid error in TRD1 coupled integration, v70.5 or v2001 ALTER 'TRD1 *CASET'(,-1) $ PARAML BDD//'NULL'////S,N,NULLB $ IF ( NULLB > -1 ) THEN $ MATMOD BDD,,,,,/VBD,IBDD/12/S,N,NCOLB/1 $ IF ( NCOLB > 0 ) THEN $ MESSAGE //' NULL COLUMNS FOUND IN MATRIX BDD' $ ADD BDD,IBDD/BIDD//(1.-30,0.) $ EQUIVX BIDD/BDD/ALWAYS $ ENDIF $ ENDIF $ 1-6377906 2004 Optimization/Design Sensitivity - DRESP2 - Job with many dresp2 entries failed in xqaros or xqfnc7. An optimization job may produce wrong sensitivity coefficients or fail in an unexpected routine such as complaining that negative values are being used in a SQRT function when the following conditions are met: 1) The job requests printing sensitivity coefficients and 2) A DRESP2 response is defined that also references to another DRESP2 3) or a DRESP3 response is defined that also references to a DRESP2 response 4) and one design variable is referenced by more than one DVXREL1 entries, Avoidance 1: Remove DSAPRT or OPTEXIT=+(-)4 request and replace it with IPRINT = 7 on the DOPTPRM entry. The gradients printed from IPRINT=7 are for constraints and have been scaled by the factor, SIGN(1/BOUND) where BOUND is either the lower or upper bounds specified on a DCONSTR and SIGN is + for the upper bound and SIGN is - for the lower bound. Avoidance 2: Create dummy type 2 properties by referring to dummy design variables whose lower and upper bounds are identical. The number of dummy type 2 properties plus number of independent design variables must be greater than or equal to the number of type 1 properties in the task. A SOL 200 job with dresp2 entries (around 540 entries) failed differently in two machines: On linux: it failed in xqfnc7 (dom6/dom61/dr2dx/xqaros) complaining that negative values were being used in a sqrt function. On AIX, it failed in xqaros indicating a floating point overflow. /net/shere/shere/users/shz/csr/fd/icw1/test_01.dat (...) 1-6410501 2004 Optimization/Design Sensitivities - DVGRID, DVSHAP, DVBSHAP, BNDGRID - Zero Sensitivies When an optimization job meets the following conditions, the sensitivities with respect to property variables may be zero: 1) the design model includes shape variables (dvgrid, dvshap, dvbshap or bndgrid) and 2) the design model includes property variables (DViRELj) and 3) the analysis model includes rigid elements (R-type elements) and the grid points associated with the r-ype elements are being designed by shape variables, The problem may be avoided with the following dmap alter: Alter 1: compile respsen alter 'dum'(,1),'dum'(2) type parm,,n,i,ndvs,ndvp ndvs = 3 ndvp = ndvtot - ndvs matgen ,/colpart/6/ndvtot/15/3 $ matprn colpart // $ partn dum,colpart,/,,a12dum,/1 $ partn umy,colpart,/a11umy,,,/1 merge a11umy,,a12dum,,colpart,/mdum/1 $ delete /ugy,,,, $ umerge uset, uny, mdum / ugy / 'G'/ 'N'/ 'M' $ As long as the property variables proceed the shape variables, this will work with any problem by adjusting ndvs, the number of shape variables in the problem. 1-6420213 2004 Optimization/Design Sensitivity - PARAM,INREL,-2 - SFM 6424 In an optimization run, if the automatic inertia relief feature (PARAM,INREL,-2) is used, then the job may fail with the following error messages: SYSTEM FATAL MESSAGE 6424 (DFNSB) THE NUMBER OF ROWS IN THE LOAD VECTOR AND THE FACTOR MATRIX ARE INCOMPATIBLE Avoidance: There are 2 possible avoidances: 1. Use the manual inertia relief (param,inrel,-1) or 2. Contact client support and obtain the alter ir200a.v2001 from the internal SSSALTER web site. 1-6442501 2004 Heat Transfer Analysis - UFM 4016 (GETEMP) THERE ARE NO TEMPERATURES FOR RAD Heat transfer analysis may fail with the following message: *** USER FATAL MESSAGE 4016 (GETEMP) THERE ARE NO TEMPERATURES FOR RADBC ELEMENT xx IN SET 1 0FATAL ERROR Avoidance: Do not define radiation to space on the backside of the plate elements. Go back in PATRAN and change the bottom to top. In the entries below, the CHBDYG IDs should be positive instead of negative $ Radiation Boundary Conditions of Load Set : core-rad-bottom RADBC 1204 1. 0 -100109 RADBC 1204 1. 0 -100110 1-6467801 2004 Dynamics - Superelements and Enforced motion - Wrong answers When enforced motion is specified on an exterior point of a superelement then wrong results are obtained in the superelement if the default DDRMM method of dynamic data recovery is used. Avoidance: Specify PARAM,DDRMM,-1. 1-6505241 2004 Superelements - DMIG, INREL - Wrong Answers If inertia relief analysis is requested for a superelement model and K2GG and/or M2GG Case Control commands are requested for a given superelement then the results for that superelement and all of its upstream superelements will be incorrect. Avoidance: Apply the following DMAP alter: compile sedisp alter 'dbview crj0','' dbview crj0=crj where (seid=0 and peid=0 and wildcard) $ endalter 1-6510414 2004 Modal Frequency Response - SPCD - Poor Results at Low Frequencies If enforced motion is used in SOL 111, it is possible that the SPCFORCE output might be incorrect at low frequencies. Avoidance: Use SOL 108 instead. 1-6518631 2004 Elements - CHEXA, MATTi - Wrong Answers The 8 node chexa element, modified in v70.7, incorrectly uses a MATTi for temperature dependent material properties. Avoidance; 1) use nastran system(309)=1 2) use 20 noded chexa 3) use 10 nodes ctetra 1-6561701 2004 Eigenvalue - ND field on EIGRL entry - SFM 3002 If on the EIGRL Bulk Data entry, ND is greater than problem (matrix) size, then the run may fail with SFM 3002 in the READ module (subroutine reigl3). Avoidance: Set ND to less than or equal to the problem size. 1-6599317 2004 Documentation - MATHP, TREF and TEMP(LOAD) - Wrong Answers TEMP(MATE) is not supported for hyperelastic elements in SOL 101. If SOL 101 is used with hyperelatic element (PLSOLID and MATHP) and the fields for AV and TREF are input on the MATHP entry and TEMP(BOTH) is requested in the case control, the results will be zero. If TEMP(BOTH) is changed to TEMP(LOAD) then plausible wrong answers are obtained. Avoidance: Place a TEMP(INIT) above the subcase level pointing to a TEMPD bulkdata entry with the TREF value specified. Put TEMP(LOAD) within a subcase. 1-6609014 2004 Acoustic Analysis - PARAM, GFL ignored, Wrong Answers In an uncoupled acoustic analysis, PARAM, GFL is ignored in SOL 108. Re-assigned to Dave B because its faster than him telling me what to type in - Paresh 1-6614001 2004 Elements - CTRIA3, MAT2, MAT8 - UFM 9050 The CTRIA3 element with orthotropic or anisotropic materials and a material coordinate system defined by the projection of a coordinate system will often have an incorrect stiffness matrix depending on how far the material coordinate system is rotated to the element coordinate system. This may lead to the following fatal message: USER FATAL MESSAGE 9050 (SEKRRS) RUN TERMINATED DUE TO EXCESSIVE PIVOT RATIOS IN MATRIX KLL. Avoidance: There is no avoidance. 1-6646806 2004 Optimization/Design Sensitivity - DRESP3 being ignored and not used When a DRESP3 entry in a SOL 200 job references to a combination of DRESP2 and any other arguments except DRESP1, the expected to be retained DRESP3 response is missing from the sensitivity calculation or optimization. Avoidance: In V2001 R2 (2001.0.4) or R3 (2001.0.7) add a dummy DRESP1 argument to each DRESP3 1-6697301 2004 Eigenvalue Analysis - PARAM,RESVEC,YES and enforced motion - Wrong answers A SOL 103 deck with PARAM,RESVEC,YES and employing enforced motion data via RLOADi and SPCD Bulk Data entries does not generate any augmented modes due to residual vectors. Avoidance: Apply the following DMAP alter: COMPILE SEMODES ALTER 'CALL SETQ','' $ CALL SETQ CASESX//SEID/PEID/S,MTEMP/S,K2GG/ S,M2GG/S,B2GG/S,MPC/S,SPC/S,LOAD/ S,DEFORM/S,TEMPLD/S,P2G/S,DYRD/ S,METH/S,MFLUID $ 1-6733944 2004 Superelements - SEP1X - SFM 6802 If more than 50 grids are at the same location in the residual structure then the message below is issued. If 50 or fewer grids occur at the same location then the message is not issued. *** SYSTEM FATAL MESSAGE 6802 (SEPSUP) SUPERELEMENT 0 HAS FAILED TO AUTOMATICALLY DERIVE A SEARCH KEY. USER ACTION: REVIEW MODEL FOR REASONABLE GEOMETRY AND SELECT SEARCH KEY VIA OVER-RIDE PARAMETER AND/OR REVIEW GRID POINT LOCATIONS FOR REASONABLE GEOMETRY 1-6875107 2004 Documentation/IOG - Incorrect INCLUDE Examples The INCLUDE example on page 105 of the MSC.Nastran IOG is incorrect. The example include '/proj' '/dept123' '/sect456' '/joe/flange.bdf' should be include '/proj /dept123 /sect456 /joe/flange.bdf' 1-6889219 2004 Superelements - DMIGOP2 type superelements no longer work in main bulk data, CSUPER In V2001, if an external DMIGOP2 type superelement is included as a main bulk data superelement, then the job aborts with SFM 4252 because the upstream XAAV is missing. See activity for why. Avoidance: Insert the following DMAP alter or use part superelements or use a different type of external superelement ECHOOFF $ COMPILE EXTIN NOLIST $ ALTER 'DBSTATUS XSOP2DIR' $ IF (NOOP2DIR<=0 AND SEID>0 AND PEID=0) THEN $ INPUTT2 /XSOP2DIR,,,,/0/INUNIT $ DBSTATUS XSOP2DIR//S,N,NOOP2DIR $ ENDIF $ ENDALTER $ ECHOON $ 1-6970301 2004 Postprocessing - PARAM,POST,-2 - Incorrect Displacements for Modal Response (SOL 111, 112) Analysis The eigenvectors rather than the response vectors are incorrectly output to the op2 file if post=-2 in modal dynamic response analysis. Avoidance: Two avoidances exist: 1. Add Param,DDRMM,-1 (not recommended), OR 2. Add PARAM,OUGCORD,GLOBAL 3. If running SOL 111, Insert following DMAP alter (tested with V2001.0.7 of MSC.Nastran). Note: ALTER ONLY WORKS for CASE of DISP=ALL compile sedrcvr alter 'sdr2.*bougv1','' $ delete /ugb,,,, $ drmh1 ougv1,,,/tougv1,mougv1,,,,,,/ $ $matprn mougv1/ $ IF ( NOCSTMS>0 ) THEN $ MPYAD TRANSG,mougv1,/UGB $ ELSE $ EQUIVX mougv1/UGB/-1 $ ENDIF $ SDR2 CASEDR,CSTMS,,,EQEXINS,,,OL1,BGPDTN,,, UGB,,,OINT,PELSETS,VIEWTB,,,,,,/ ,,BOUGV1,,,,/APP///////ACOUT/PREFDB/-1./-1./ ADPTINDX/ADPTEXIT/FALSE $ alter 'output2.*bougv1' $ $message //' ' $ $message //'Printed displacements in basic system' $message //' ' $ $ofp bougv1/ $ ENDALTER 1-6991176 2004 Aeroelastic Analysis - CAERO1, SPLINE1 - FPS Spline may lead to SFM 4276, Floating Point Exception If a SPLINE1 with the FPS option is used to connect a CAERO1 entry that only has one spanwise box, then a floating point exception may occur in the GI module. This error will also occur if the spline connects to a chordwise strip of elements on a CAERO1 that is one spanwise box wide. Avoidance: Split the CAERO1 entry so that it has two span wise boxes, or use the IPS option on the SPLINE1. 1-6997601 2004 Elements - CWELD - UFM 2007 occurs when CWELD and fluid elements are present USER FATAL MESSAGE 2007 occurs in MODGM2 when CWELD elements and fluid elements are present. The CWELD generates grids GA anf GB which are considered undefined. 1-6998602 2004 Acoustic Analysis - PARAM, FLUIDSE and SOL 103 - SFM 3001 If PARAM,FLUIDSE,>0 is specified in a SOL 103 run and differential stiffness or residual vectors are requested, then the run will fail with: *** SYSTEM FATAL MESSAGE 3001 (MTRXIN) THE INPUT DATA BLOCK IN POSITION 3 DOES NOT EXIST. Avoidance: Insert the following DMAP alter: compile phase0 alter 'endif \$ not(sebulk)' equivx eqexin/eqexinq/-1 $ call dbstore eqexinq,,,,//0/0/' '/0 $ compile aspline alter 'makvgac','' call dbfetch /eqexinq,,,,/0/0/0/0/0 $ call makvgac eqexinq,eqexins,uset0,drg/ vgxac,vgac/ seid/nsenqset/luset $ endalter 1-7186605 2004 Optimization/Design Sensitivity - Superelements, DRESP2 - UFM 7139 (DOPR3N) there is some problem in dopr3j in handling missing dresp2. ufm 7139 should and is now issued by dopr3x. - David David, Shenghua has a deck (see attachement), we could have the fatal error message like below: *** USER FATAL MESSAGE 7139 (DOPR3N) ON DRESP2 entry ID = 23, DRESP2 ID = 231 DOES NOT REFERENCE AN EXISTING DRESP2 ENTRY. However, it appears that you removed the 7139 UFM when you worked on QD1-7495101 (DRESP2 reference implied DRESP2). I guess that you removed the UFM due to possilbe implied DRESP2. Thus, we need change the 7139. I have DOPR3n to move UFM7139 back (stirling/users/xmy/vv). But you may change the contents of UFM7139. Xiaoming 06/17/03 When an optimization job with superelements references a nested DRESP2 entry, the job fails with the following message although the responses are defined correctly: *** USER FATAL MESSAGE 7139 (DOPR3N) ON DRESP2 entry ID = 22, DRESP2 ID = 2221 DOES NOT REFERENCE AN EXISTING DRESP2 ENTRY. 1-7241319 2004 Optimization/Design Sensitivity - Misleading Message When an optimisation analysis (sol 200) with analysis=mfreq stops with a hard convergence the message issued may be "no feasible solutions are found". Avoidance: There is no avoidance. 1-7264401 2004 sol 200 - ufm 5629 (dsad1c) - FRSTRE - ATTB=AVG When setting ATTB to AVG on a DRESP1 entry with RTYPE=FRSTRE the following error message will appear: USER FATAL MESSAGE 5629 (DSAD1C) THERE ARE DRESP1 ENTRIES FOR RESPONSE TYPE = FRSTRE BUT THERE ARE NO CORRESPONDING OUTPUT RESPONSES. USER ACTION: CHECK THAT THE ELEMENT ID'S LISTED ON THE DRESP1 ENTRIES EXIST. Avoidance: There is no avoidance other than to build a DRESP2 response for the average. 1-7421606 2004 DBC - PARAM,POST - UFM 1250(BLKRD), SFM 6059(DIOMSG) MSC.Nastran may fail in the DBC module when a large number of SPCD enties with different set ids exist in the model. The fatal messsages encountered are: USER FATAL MESSAGE 1250 (BLKRD) Read block number beyond end-of-file. SYSTEM FATAL MESSAGE 6059 (DIOMSG) Avoidance: Use PARAM,LOADU,0 to avoid this problem. 1-7452105 2004 Optimization - rmsdisp, rmsvelo, rmsacce - SFM 3001 Mike Reymond has imported the fix for tier 1 system. - David Chou 7/7/03 this limitation has been removed in v2004. As of now 7/2/03, param,enfpath,old must be inserted to ensure a successful run. Mike Reymond will provide a fix for enfpath,new. - David Chou In an optimization run, if random response is requested, and the model contains superelements, then the job will fatal with one of the following error messages: SYSTEM FATAL MESSAGE 3001 (DSARMG) THE INPUT DATA BLOCK NAMED DRDUTB AT POSITION 2 DOES NOT EXIST or NO DRESPI ENTRIES EXIST FOR DESOBJ ID= xx Avoidance: The only avoidance is to not use superelements. 1-7461013 2004 Documentation/QRG - INCLUDE Command May be Nested Remark 1 of the INCLUDE Case Control Statement in Section 4A of the QRG incorrectly states that INCLUDE statements may NOT be nested. It should say INCLUDE statements MAY be nested. Section 5 for the bulk data needs correction as well. 1-7495101 2004 Optimization/Design Sensitivity - DRESP2 - UFM 7139, Failed to locate implied DRESP2 IDs If an implied DRESP2, references a DRESP1 with a function name as input for the ATTB field, then the job may fail with UFM 7139. *** USER FATAL MESSAGE 7139 (DOPR3N) ON DRESP2 entry ID = 10, DRESP2 ID = 5 DOES NOT REFERENCE AN EXISTING DRESP2 ENTRY. Avoidance : Renumber the DRESP1 IDs such that the DRESP1 with function name in the ATTB field has the the lowest ID among all DRESP1s. 1-7627701 2004 Eigenvalue Analysis - residual structure CMR - massless mechanism - SFM 6143 Residual structure component mode reduction will fail with SYSTEM FATAL MESSAGE 6143 if a massless mechanism exists in the a-set. Avoidance: Apply the following dmap alter compile moders alter 'segoa'(,-1),''(,1) $ got is actually goa SDR1 USET,,MECHA,,,GOt,GM,,,,/ MECHg,,/1/'STATICS' $ endalter 1-7656709 2004 Optimization/Design Sensitivity - DSAL - SFM 4276 An optimization job may fail in the DSAL module with the following error message: *** SYSTEM FATAL MESSAGE 4276 (GREAD) ERROR CODE 3001 PID= 0 Avoidance: Increase the memory space using the mem= option on the nastran submit command so that module dsvg1 will only need a single pass to perform the triple matrix product. The amount of memory required can be found from SYSTEM INFORMATION MESSAGE 7502 printed in the f04 file. 1-7734708 2004 Distrubuted Memory Parallel - LAM 6.3.1 cannot build with RedHat Linux MSC.Nastran V70.7 and 2001 require LAM 6.3.1 to run distributed memory parallel. This version of LAM will NOT compile with RedHat 7.2. To avoid the need to compile LAM, the binaries may be downloaded from ftp://ftp.msclinux.com/contrib/jjg/nastran This is not an issue in MSC.Nastran 2003. For installation: 1) cd /tmp 2) download usr_local_lam631.tgz 3) cd /usr/local 4) tar xvzf /tmp/usr_local_lam631.tgz 5) rm /tmp/usr_local_lam631.tgz 6) modify /etc/csh.cshrc and /etc/profile to have /usr/local/lam-6.3.1/bin in your path 1-7782902 2004 Optimization/Design Sensitivity - Null Design Sensitivity Coefficients When an optimization task includes the design variable whose initial value is zero, the sensitivity coefficients with respect to that design variable become null. Avoidance: Do not start a design with zero design values. 1-7785001 2004 Miscellaneous - GROUNDCHECK(SET=A) - Missing Output No output is printed for GROUNDCHECK(SET=A). Avoidance: Insert the following dmap alter: compile sekr alter 'setbit','' PARAML CASES//'DTI'//324//S,N,SETBIT $ endalter 1-7793712 2004 Aeroelastic Analysis - MPCFORCE - Missing Output Aeroelastic analysis (SOL 146) in Version 2001 does not honor the MPCFORCE output request. The case control command does not yield any MPCFORCE output. Version 70.7 works correctly. Avoidance: Apply the following dmap alter: compile seaero alter 'call fdrmgen','' call fdrmgen uset,vgfd,cases,usetd,gmd,gm,goa,phdh, msf,m2sf,bsf,b2sf,ksf,k4sf,k2sf, MMG,M2MG,BMG,B2MG,KMG,K4MG,K2MG,,, MSS,M2Ss,BSS,B2Ss,KSS,K4Ss,K2Ss,,ys, ects,bgpdts,sils,xy 1-7820401 2004 Rigid Elements - GP4 - Missing UFM 1201 A SOL 103 job may fail without printing any messages in the f06 or log files when illegal rigid element connections exist but there is no SPC request. Avoidance: Add a dummy spc request. If illegal rigid element connections exist, UFM 1201 will be output. 1-7939303 2004 Rigid Elements/Multipoint Constraints - MPCAX - UFM 316 MPCAX entry may result in UFM 316. Avoidance: Add NASTRAN SYSTEM(357)=1. 1-7979705 2004 Transient Analsyis - PRELOAD, MPCF - SFM 5423 A linear or nonlinear transient analysis that has a static preload subcase and requests mpcforces will fail with SYSTEM FATAL MESSAGE 5423. Avoidance: Insert the following DMAP alter: compile super3 alter 'paraml qg','' paraml qg//'trailer'/1/s,n,ncqq//s,n,nqg $ if ( nqg<0 ) paraml qmg//'trailer'/1/s,n,ncqq $ endalter 1-7980301 2004 System Limits - ECHO=SORT - Limited to 26 entries. There is an undocumented limit that the number of bulk data entry names included with the ECHO=SORT case control command cannot exceed 26. Items exceeding that number are ignored. Avoidance: Limit the list to to the 26 most critical names or do not use this option on the ECHO=SORT command. 1-8002452 2004 DMAP - TABPRT and USET option - UWM 3059 The TABPRT DMAP module with the USET table print option does not allow DOF sets with more than 2 characters to be output. For example given: DEFUSET,K,LTMA USET1 LTMA 123 111 And this TABPRT command: tabprt uset, bgpdts//'USET'/10///'LTMA'/ $ Results in this warning message and no USET table output: *** USER WARNING MESSAGE 3059 (TABPRT) SET IDENTIFIER LT DOES NOT EXIST. ERROR DETECTED IN SUBROUTINE *** USER WARNING MESSAGE 3059 (TABPRT) SET IDENTIFIER MA DOES NOT EXIST. ERROR DETECTED IN SUBROUTINE Prior to V70.7, the user could get output for LTMA set by entering the USET bit string for the K set in this example. This was broken in V70.7. Resolution: Printout of all DOF sets will include the 4 character user sets as follows: tabprt uset, bgpdts//'USET'/10/-1/// $ 1-8010801 2004 Acoustic optimization - DRESP1, PRES - Poor Answers Frequency response acoustic optimization convergence problems may occour if response type PRES is used on a DRESP1 entry. The problem may converge to a poor solution. If PRES is replaced by FRDISP the analysis yields better results. PRES and FRDISP should give identical results. Avoidance: Use FRDISP instead of PRES. 1-8076115 2004 Postprocessing - DBC - UFM 1250 (BLKRD) MSC.Nastran Linux 2001 will issue the following message if an xdb file larger than 2gigs is created. *** USER FATAL MESSAGE 1250 (BLKRD) STATUS = 6007, FILX = 40, LOGNAME = DBC , NSBUF3 = 16384 FILE = TEST.xdb BLKNBR = 1129465448 ERROR MESSAGE IS -- Read block number beyond End-of-file. 0FATAL ERROR Avoidance: This problems has been corrected in Version 2001.0.9 available on the Nastran 2001 R4 CD. 1-8126205 2004 Data Recovery - STRESS(CORNER), PCOMP - UWM 6418, UFM 3005 If Stress(corner) is requested it should be ignored for composite elements. Instead the run fails with the following messages: *** USER WARNING MESSAGE 6418 (SDR2A) THE BILIN OPTION FOR STRESSES OVERRIDES ALL OTHER OPTIONS IN STRAIN AND FORCE COMMANDS. *** USER FATAL MESSAGE 3005 (SDR2) ATTEMPT TO OPERATE ON THE SINGULAR MATRIX (NONE) IN SUBROUTINE SQD43D Verified fix in version 2003r2 on Atlantic. -- SSP. File is on /scratch/ssp/errorfix. 1-8130019 2004 Distributed Memory Parallel/Eigenvalues - Hangs if Massless Mechanism Exist If an eigenvalue problem is run with DMP and the model contains massless mechanisms, the job may hang. After the 1st eigensolution stops with FATAL factor errors, the solution properly goes into the massless mechanism check, detects and constrains the mechanism degrees of freedom. The run starts the 2nd eigensolution in the master cpu but doesn't restart the slaves and the job hangs. Avoidance(s): 1) Run in serial. 2) Run with PARAM,MECHFIX,YES in dmp (as opposed to default MECHFIX,AUTO). 1-8192502 2004 DMAP - VEC, UPARTN, MATGPR, etc. - UFM 2120, 2121, 2122, 2123, UWM 3059 The use of "+" and "-" to form new dof sets (e.g.; VEC USET/VGX/'G-A'/'M' $) as may be specified in the set name parameter inputs in the MATGPR, VEC, UPARTN, UMERGE, and UMERGE1 modules will cause various fatal errors like UWM 3059, UFM 2120, UFM 2121, UFM 2122, and UFM 2123. Also, leading blanks in the set name parameters will cause similar fatal errors. Also, MATGPR does not allow for set equations longer than 4 characters. Avoidance: There is no avoidance. 1-8227901 2004 Optimization/Design Sensitivity - DRESP2 - UFM 7147 An optimization job fails with the UFM 7147 complaining that the objective function referernces a non-scalar number. Although the job appears to be defining a correct scalar quantity, the quantity is defined by a nested DRESP2 entry that further refers to a DRESP1 entry with the MAX function. 1-8275506 2004 Optimization/Design Sensitivity - MFREQ with FREQ4 failed in DSAH - SFM 3002 SOL 200 with ANALYIS=MFREQ with FREQ4 failed in the DSAH module with the following message: *** SYSTEM FATAL MESSAGE 3002 (DSAH) EOF ENCOUNTERED WHILE READING DATA BLOCK DFFDNF (FILE 118) IN SUBROUTINE Prior to the above failure the following messages were issued in every cycle from the second on: *** USER WARNING MESSAGE 6644 (FRPROD) SOME OR ALL COMPUTED EXCITATION FREQUENCIES ARE OUTSIDE THE RANGE F1 AND F2 FOR FREQ4 BULK DATA ENTRY WITH ID = 502 THESE COMPUTED EXCITATION FREQUENCIES ARE IGNORED. Avoidance: Use a different FREQi entry to specify the frequencies of interest but not that the error may also occur with a FREQ entry with only one frequency specified. 1-8310201 2004 Distributed Memory Parallel/ACMS - EDE and EKE - Missing Output The EDE and EKE output are not supported in ACMS. Avoidance: There is no avoidance. 1-8315703 2004 GPSTRESS - STRESS(CORNER) - Wrong Answers Inaccurate grid point stresses may be output for quad4 elements in nonlinear analysis (SOL 106) when stress(corner) is requested. Avoidance: There is no avoidance other than not requesting corner stress output. 1-8328301 2004 Optimization/Design Sensitivity - MODTRACK, PARAM,OPTEXIT - Misleading Message See notes activitity dated 7/15/2003 If the user has requested mode tracking along with PARAM OPTEXIT <4, the DOM12 module will be called and a message like: RUN TERMINATED DUE TO MAXIMUM NUMBER OF DESIGN CYCLES = 10. will be produced even though no design cycles have been executed. Avoidance: Apply the following dmap alter: compile exitopt alter 'dom12','dom12' IF ( (optexit<>3) and (optexit>3 or cnvflg>0 or descycl1=DESMAX OR numddv>0 or dsprint or dsunform or dsexport or modetrak>0 or badmesh) ) DOM12 ,,xval,,,propof,optprmg,his,destab,geom1,coordo,edom,mtrak, ept,geom2,mpt,epttabf,dvptabf,xvalp,geom1p,,,,,pcompt/ ,,dbcopt,/ descycl1///CNVFLG/3/OPTEXIT/desmax/mdtrkflg/ despch/despch1/modetrak/eignfreq/dsaprt/protyp/badmesh/ xyunit $ 1-8383704 2004 Random Response - RANDOM - Wrong answers in punch file In a random response run if the PSD is output to the output2 file the values may be incorrect. The output from the XYPLOT is correct. 1-8424704 2004 Parallel ACMS - MPCF, SPCF, OLOAD - UFM 5423 (SADD5) Parallel ACMS may fail if all of the following conditions exist: 1. DMP>1 2. MPCFORCEs are requested. 3. Loads are applied to dependent DOFs which are also interior to a domain (superelement) or: 1. DMP>1 2. SPCFORCEs are requested. 3. Loads are applied to constrained DOFs which are also interior to a domain (superelement) If neither MPCF nor SPCF is requested but OLOAD is requested and loads are applied to a dependent or constrained DOFs which are also interior to a domain, then the OLOAD output will be incomplete for the slaves' forcing frequencies. Avoidance: Apply the following dmap avoidance. compile semfreq alter 'call super3'(2,-1) equivx frl0/frl/-1 $ endalter 1-8448301 2004 Optimization/Design Sensitivity - CQUADP, CTRIA3 unrecognizable Entries in Punch File A p-element optimization run may produce user unrecognizable Bulk Data entries in the punch file for all types of shell elements such as CQUAD4P, CTRIA3P etc. These entries will lead to a fatal message if they are subsequently read back in as bulk data. Avoidance: Remove these entries before reading back in. 1-8448304 2004 Miscellaneous - IFP, PBEAML - Format error in punch file A pch file containing updated PBEAML entries written in the large field format may fail when read back into MSC.Nastran due to format error. Avoidance: Add a blank line to the end of the PBEAML entry if the number of the existing lines is an odd number. 1-8516805 2004 Optimization/Design Sensitivity - PCOMP - Incorrect composite element matrices in SOL 200 If there are differences in the design model with respect to the analysis model and the differences arise from DVCRELi entries and from DVPRELi entries that point to PCOMP's, the composite updates will be ignored for the initial analysis. The result will be that the analysis, sensitivities and first redesign will be in error. It's possible that subsequent design iterations, which have the correct properties, will be able to recover, but it is also possible that the initial errors will be such that recovery is not possible. Avoidance: Apply the following alter: compile desopt alter 'modgeom2'(4), 'geom2nn'(3,1) alter 'modgeom2'(5,-1), 'modgeom2'(5) IF ( MODGEOM2 ) THEN $ DESITER=1 $ INCREMENT QUALIFIER IF GEOM2 IS UPDATED $ EXCEPT WITH MULTI-MASTER METHOD EQUIVX GEOM2NN/GEOM2/ALWAYS $ else $ DBEQUIV DATABLK=GEOM2, WHERE (DESITER=0 AND WILDCARD) , CONVERT (DESITER=1) OVRWRT $ ENDIF $ MODGEOM2 1-8564404 2004 Documentation/QRG - UFM 6632 (PAABSF) occurs when small value used within TZIMID The following fatal message may occur when the value for TZIMID on the PAABSF entry is less than the machine's numerical zero. *** USER FATAL MESSAGE 6632 (EASFD) THE NORM OF IMPEDANCE IS TOO SMALL Avoidance: Specify a larger value for TZIMID. 1-8581001 2004 Optimization/Design Sensitivity - PBUSH - Failed Convergence If a designed PBUSH is connected to a GRID that has a displacement coordinate system specified (CD on field 7 of the GRID bulk data entry), the senstivity with respect to the PBUSH properties are likely to be in error with the result that the optimization will not be able to make progress and may terminate with various messages indicating failed convergence or maximum number of design cycles. Avoidance: Remove the CD from the GRID entry or replace the PBUSH with an equivalent CROD or CELAS entry. 1-8587906 2004 Optimization/Design Sensitivity - RANDOM - SFM 3007 (DOPRAN) In Sol 200, if there are more than four frequency response subcases being combined for random response, then the following error message may occur: SYSTEM FATAL MESSAGE 3007 (DOPRAN) ILLEGAL INPUT TO SUBROUTINE DOPRN2 Avoidance: Until the error is fixed, limit random response in SOL 200 to four or less combined subcases. 1-8629707 2004 Elements - QUAD4, TRIA3 - Wrong Principal Strains computed in DDRMM Principal strains are computed incorrectly in DDRMM and are off by a factor of 2. Avoidance: Add param,ddrmm,-1 to the input. 1-8680201 2004 Case Control - SET - SFM 6143 (DUPART), Poor Message The following case control will result in SFM 6143: SET 1 = 100 SUBCASE 1 SET 1 = 101 DISP = 1 if grid point 100 does not exist. Avoidance: Remove the superfluous SET definition. 1-8751074 2004 Elements - CBEAM, PBEAML - Z-Section produces incorrect warping coefficient The calculation for the warping coefficient of a Z-section beam (PBEAML) is wrong. 1-8781636 2004 Large File - SGI - UFM 1250 On the SGI, if NFS mount is used for filesystems that are larger than 2 gig, then the job may fail with the following messages: USER FATAL MESSAGE 1250 (BIORD) Specified record is beyond maximum for this filesystem. 1-8783605 2004 Elements - CHEXA, TEMP(LOAD), PSOLID - Incorrect stresses When 8 noded CHEXA elements have different element coordinate systems and default values for the PSOLID entry are used along with TEMP(LOAD), incorrect stresses are output at corner/gauss points. Avoidance: Set NASTRAN SYSTEM(309)=1 to use pre 70.7 CHEXA or set IN=2 on the PSOLID entry 1-8803801 2004 Optimization/Design Sensitivity - APRCOD=2 OR 3 - Poor Sensitivities An error has been discovered in the calculation of the gradient of a response when using the mixed method or the convex method (APRCOD=2 or 3) with inverse design variables. The "exact" sensitivity should be multiplied by (x0/x)**2. Instead it is multiplied by x0/x. The approximate responses are calculated correctly, so the error only influences the determination of the search direction once the design variables have moved from their starting values. Avoidance: APRCOD=1 will not have this error