Siemens Simcenter Nastran 2019.2-1872

The following list summarizes the problems that are fixed in Simcenter Nastran 2019.2.

9430338 Running DMP on SOL 108 with FRF sets does not produce results at somefrequencies.
9366533 Incorrect data recovery for upstream Super Elements if sparse data recovery isused.
9378423 SOL 112 with RESVEC(NODYNRSP) gets wrong answers in SCN2019.1 whendegenerate eigenvalues are present.
9361350 RVEL on non-existing elements is ignored
9358943 Using SUPORT and STATSUB case control together produces SFM 6143 (UMRGE1)
9304651 Wrong MPC forces with FRF components.
9371230 Corrected a crash due to inappropriate handling of CBUSH element in SOL 200. Itoccured in case a CBUSH element supported both temperature loads on its nodesand an element response (e.g. FORCE).
9371235 Clarified error message when temperatures are missing from elements.
9367708 Added a fatal error when an invalid single quote character ‘ is in the input filename
9344047 Reformatted log file timing in Linux to be consistent with Windows
9412752 SOL 401 fails to trap the case when there is no support for kinematic hardening.
9336175 There is an issue with damping during bolt iterations which is activated byparameter BOLTDAMP on NXSTRAT card in NX Nastran 12.2 .
9286203 SOL 601 outputs extra results if the SEPDIS option on BCRRESULTS case controlcommand is requested and skip factor on TSTEP option is used
9251697 The total strain is not equal to the sum of thermal and elastic strain from theoutput values for old parabolic shells in SOL 101.
9443424 OTEMP results for models containing parabolic shells (CQUAD8 and CTRIA6) wereincorrectly reported due to bug in the results recovery code.
8413597 SOL 401 Dynamics analysis subcase fails to converge for models containing CBUSHand CBUSH1D elements with scalar damping value specified either on PBUSH orPBUSH1D card.
9395170 SOL 401 Dynamics analysis subcase ignores damping imposed through GE/W4options for LGDISP=+1 analysis when Stress stiffness is used (which is used bydefault).
9392420 For problems using BOLTSEQ,use of NINC value on TSTEP1 card instead ofBOLTSEQ card causing missed inermediate results from being output.
9350273 For type 2 bolts with NINC=1, bisection logic wasn't actually cutting back the loadafter bisection resulting in failure to solve.
9365451 In SOL 402, if the problem name contained parentheses “()”, the solution wouldfatal with a misleading error message “FATAL SYNTAX ERROR IN A LOGICALCOMMAND.”
8406458 In SOL 402, if a model contains both contacts BCTSET and gluing BGSET, thecontact parameters BCTPAR2 might not be associated to the correct contactBCTSET, depending on the order of the cards.
9398734 In SOL 402, when requesting the GPFORCES output, depending on the size of themodel, a fatal error could occur due to a mis-match in an array size. Removing theoutput request would by-pass the problem.
9410713 A nastran SOL 401 deck with a time-varying RFORCE/RFORCE1 and additionalfollower type static (non-time-varying) load, for example static pressure or FORCE1can give incorrect result and cause convergence problem.
9285115 A transient SOL 401 Thermo-mechanical multi-physics model with a time-varyingRFORCE/RFORCE1 and multi-physics pressure load can encounter convergenceproblem.
9273010 The CQUAD8 element formulation used by nastran in releases prior to NXN 1872 can give wrong results in buckling problems.
8402864 SOL 108 Random analysis scales poorly with the number of subcases
9389580 For CELAS element, when it is inactive, its initial element stiffness matrix is notscaled down. So it needs to multiply its initial element stiffness matrix by a scalefactor to fix this PR.
9356408 9356305 9356370 The software updates ELAREST data block at the beginning of a subcase, which is alimitation for multiphysics solution.
9325846 When the software encounters a non-convergence in material library, it does notstop the computing, then you will get unreasonable strain values. After the fix itprints out a fatal message for the non-convergence in material library.Formulations for shell elements in SOL 106 and SOL 401 are different. SOL 106uses engineering strain to compute stress and SOL 401 uses GREEN strain tocompute stress. when strain is small, engineering strain is similar to GREEN strain.For this PR case, it is a simple Axial tensile model, the initial length(L0) is 2000,andthe terminal length(L1) is 2400, soEngineering strain E_XX = (L1/L0)-1 = 0.2GREEN strain E_XX = 0.5*[(L1/L0)*(L1/L0)-1]= 0.22From above computation, you can see the strain is not small, so SOL 401 resultdiffers from SOL 106.
8412889 In SOL 402, if a model contains SPCD cards referenced by LOAD case control cardsand varying from subcase to subcase, and if the user gives a lot of memory to thesolver (gigabytes), the pre-processor would incorrectly write a lot of data, resultingin a very long computation time. This has no impact on the results but it can leadthe user to think that the computation will never finish. As a workaround, the usercan reduce the amount of memory passed to the solver. If this does not solve theissue, there are 2 additional workarounds. If SPCD cards are used as fixations with a 0 value, replace SPCD by SPC/SPC1; if not, use a DLOAD case control cardreferencing the SPCD cards with a table.
9366377 SOL 402 neglected the initial velocity definitions in a DYNAMICS subcase if it wasnot the first subcase or if it was not preceded by a PRELOAD, STATICS, orDYNAMICS subcase. As a workaround, the user can change the order of thesubcases to prevent the described occurrence. It is possible to put a dummySTATICS subcase with no loads to precede the DYNAMICS subcase if necessary.
9373645 In SOL 402, in case of a restart, if SPCD, FORCE, MOMENT or RFORCE1 loads areused in the restart solution, incorrect results are obtained. There is no knownworkaround.
9390069 In SOL 402, when a global DLOAD card and several subcases including at least onewith ANALYSIS = PRELOAD are defined in the case control, the loads are incorrectlycomputed during the preload. As a workaround, the user can remove the globalDLOAD card in the case control and define a DLOAD card in each subcase withANALYSIS = STATICS or ANALYSIS = DYNAMICS.
8405620 In SOL 402, the NLCNTL2 parameters DCON and PRCS, both related to contact,were ignored. This has been correct but those parameters cannot vary fromsubcase to subcase anymore.
8412526 SOL 402 does not support temperature-dependency of the mass density. The massspecified on the MAT1 card is used and a warning message is displayed saying thatthe TABLEM1 is ignored. In previous versions, the problem can be corrected byremoving the reference to the TABLEM1 on the MATT1 card.
8414467 In SOL 402, when a TEMPD card is defined, a temperature load is applied to eachnode of the structure, excepted the internal nodes created automatically by thesolver when cutting the bolts. In the case of a restart, temperatures wereincorrectly applied to these additional nodes as well, creating incorrect thermalstrains in the bolt. As a workaround, the user can replace the TEMPD cards byequivalent TEMP cards.
8415203 In SOL 402, the TEMPD bulk cards are processed before the TEMP bulk cards. Whenboth cards are defined and are referred by a TEMP(INIT) case control card, theinitial temperature was ignored for the nodes defined in the TEMP cards. On theother hand, when no TEMPD cards are defined, the behavior is correct. As aworkaround, the user can replace the TEMPD cards by equivalent TEMP cards.
8415361 When a TEMPD bulk card has the same ID as a DTEMP bulk card, it is ignored inSOL 402. As a workaround, the user can use only DTEMP cards to define thetemperature of all the nodes.
9414690 In SOL 402, if a contact set BCTSET was associated to a contact parameters cardBCTPAR2 containing more than 9 parameters, either a fatal error would occur orthe last parameters would be ignored, depending on the case.
9427247 In a SOL 402 model containing a BOLT card referencing the global coordinatesystem (CSID=0), only the bolt axis IDIR was used but the other two directionswere randomly generated, resulting in outputting the bolt shear results inunknown axes. The global results were however correct.
9428224 In SOL 402, when a model contains several subcases with different DLOAD casecontrol cards which refer to the same SPCD cards associated to different tables from one subcase to another, incorrect results are obtained. There is no knownworkaround.
9428458 In SOL 402, if a model contains both BCTSET and BOLT cards, a fatal error “%%%E01-MCT311, software error” could occur or the wrong contact set could beactivated or deactivated in a given subcase. This was due to a conflict in thenumbering of those contact and bolt sets.
9429005 In SOL 402, when a preload subcase is followed by a dynamics subcase, the is noknown workaround.
9429970 In SOL 402, when a model contains time-unassigned loads and preload subcases,wrong results are obtained, except if the computation starts with all the preloadsubcases, which is the most common use-case. As a workaround, the user canreplace the time-unassigned loads by equivalent time-assigned loads (DLOAD –>TLOAD1 + TABLED1).
9433593 Model contained SPCD cards and 2 subcases (statics + modal). It was exportedfrom Simcenter as SOL 401, then converted to SOL 402. Therefore, it containedcompanion SPC cards next to SPCD. This would not be the case for modelsexported directly as SOL402. This is now allowed for 402 as well
9435310 In SOL 402, if a model contains both BCTSET and BOLT cards, a fatal error “%%%E01-MCT311, software error” could occur or the wrong contact set could beactivated or deactivated in a given subcase. This was due to a conflict in thenumbering of those contact and bolt sets.
9436569 In SOL 402, if the modal subcase was generating a fatal error, for example if noeigen value was found in the requested interval, the solution would still finish witha successful return value. It was up to the user to look for the error messages inthe F06 file. This issue has been fixed now.
9436864 In SOL 402, if a model contains a PSOLCZ card without any thickness specified, thesolver was using a zero-thickness, generating a fatal error “%%%E02-ASP147 ,ELEMENT xxx Zero thickness”. This has been changed to use 1.0 as a defaultthickness.
9438311 With SOL 402, if a simulation contained MATDMG EUD, the simulations crasheddue to issue for Jacobian strain which was corrupted and prevent strainmeasurement transformation. The workaround was to use MATDMG UD. Now thisissue has been fixed.
9438345 In SOL 402, if the user requested the SPCFORCES for a MODAL (or FOURIER orCYCMODES) subcase, a crash would occur. Since this output request was notsupported for those subcase types, the user could simply remove it to by-pass theissue. This has now been fixed and the output request is ignored.
9461974 In SOL 402, if the model contained an RFORCE card referenced by a DLOAD casecontrol card and the grid point identification number through which the rotationvector acted was not located at (0,0,0), the forces were not computed correctly.
9464334 In SOL 402, with MATCRP TYPE=301 and NLCNTL2 CRINFAC=1.0, the solver couldcrash because during iteration of material law integration, some trial von Misesstress became negative. The workaround was to decrease MATCRP THRESH toimprove convergence or to change NLCNTL2 CRINFAC from 1.0 to 0.5. This hasnow been fixed.
9378522 Inaccurate load approximation in RDMODES leads to wrong results.
8401892 A hang in SOL 111 analysis using DMP when more than 16 processors are used.
9324380 Rigid modes with negative eigenvalues caused Lanczos to crash
9414728 The performance of grid contribution computaton is slow if the number of gridsrequested is very large.
9439161 Nastran issues fatal message in the ATV data recovery after solution if the requiredoutput freqnency is not in the range of ATV matrices.
9386762 For frequency dependent vibroacoustic analyses that specify SDAMP on thestructure, the SDAMP(structure) is also applied to the acoustic part.
9094749 ATV computation performace worsen with DMP due to PTMIC DMP collection
8410452 There was a problem in the software for the CBUSH1D and CBUSH elements in SOL401. Specifically, they were always processed on the assumption that 'PARAM,LGDISP' was set to 'on'. Additionally, 'PARAM, MATNL' affected whether or not theCBUSH1D elements used non-linear spring data (if defined).
9358257 When SMP is turned on, the results of GRID POINT FORCE BALANCE is wrong forSOL401 with 3D solid elements. When SMP is turned off, the results of GRID POINTFORCE BALANCE is correct. After the bug fix, correct results for GRID POINT FORCEBALANCE is obtained regardless of the SMP setting.
9295772 SOL 101, 401, and 402 does not require SB when FT is specified on the PCOMPcard.
8412521 In SOL 402, a wrong value of the Young modulus or Poisson ratio could be used insome very specific cases. In particular, this was the case for large strain analysis(using Cauchy stress/stress measure), with material non-linearities(PARAM,MATNL) turned on and a MATS1 card referencing a TABLES1 in total strainwithout thermal dependency but a Young modulus having a thermal dependencyor having a Poisson coefficient with thermal dependency.
9303135 In SOL 402, when requesting the stresses for modal subcases would result in azero output. This was due to the change in stress/strain measure performed in theprevious release.
9391634 In SOL 402, if the problem name contained a dot “.” (beside for the file extension),e.g., “name.1.dat”, the solution would fatal with a misleading error message“FATAL ERROR: INPUT OP2 FILE "….sf1" NOT FOUND”
8396518 SDAMPING(structure) is not applied correctly in the context of SOL 108 Femaostrong coupling vibroacoustic scenario.
9424124 Slow-down in femao postprocessing for displacements in Simcenter Nastran 2019.1
9424467 NXN2019.1 requires more memory in DMP than in NXN12
9442859 Contact model with TEMP loads at the grids causes a crash in TEMPFRIC module
9400955 Peformance degradation in FOCOST module for large model with 5-8 millioncontact elements
9377201 SOL 401 BCRPARA does not support type RIGID but is not trapped as an error
9361342 Trap input when duplicate element Ids of different element types are present inpresent in glue or contact
9292288 SOL 200 model with different BGSETs in different SUBCASES causes UFM 5629(DSAD1C)
8406114 HDF5 Output - ELNAME Label for GPFORCE output is incomplete in hdf files
9291179 HDF5 errors in Nastran log file and crash/undefined behavior when trying to writeEPT to HDF5 output.
8415256 In the linear solution sequences, when a bolt (etype=1) is present in the modeland this bolt has been preloaded in subcase 'n', any subcase that follows subcase'n' that does not contain the bolt but only has a service load will see issues withglue stiffness not being applied correctly (assuming there is glue anywhere in themodel).
9414809 In the linear solution sequences, BOLTRESULTS case control card is not supported.Instead of ignoring the card during runtime and allowing the solution to continue,a FATAL is issued. This behavior is to be changed and the solution will be allowedto proceed.
9403440 In models where number of instances of bolts exceeds 1000 (number of subcasesx number of bolts), internal BGSET ID created for bolts can clash with internalBGADD IDs, thereby producing a FATAL error since BGSET and BGADD IDs shouldnot be the same. This has been fixed in 2019.2. Workaround is to add anotherdecimal place to the BGADD ID created by the user
9454857 In SOL 402, the number of threads used during the computation was not printedin the F06 file if this number was set using the OMP_NUM_THREADS environmentvariable.
8405181 Nastran error message for checksum failure in a deck that was modifed outside ofSimcenter was not very descriptive.
9325528 In SOL 101, when iterative solver is used, the glue forces were not included in theSPC force calculation and so SPC force result was not giving correct value.
9317116 Nastran was crashing when it was running out of memory inside the iterativesolver. Memory checks are introduced so that Nastran exits gracefully in case ofinsufficient memory.
9280351 During the Analysis phase of MUMPS solver, a memory allocation error happensinside Metis. Unfortunately, MUMPS code fails to catch this allocation error andNastran hangs downstream.
9435925 Subcase with bolt (boltld in subcase) followed by subcase without bolt (the latteris where issue is noticed)
9388230 ASSIGN SC_H5 fails in case path includes spaces
9366003 PCOMP laminate element ply results incorrect from random analysis…
9362151 NX Nastran 2019.1 fails in MODCON for some large models.
9359892 Fatal error in FEMAO analysis for enforced motion from large file from SPCDFILE.
9327491 DISP(PLOT)=NONE still writes data to OP2 file
8405863 In SOL402, if the model contained a contact set including the contact parameterBCTPAR2 PDEPTH, this parameter was not correctly taken into account and thesolver was using the default value of 5 times the size of the largest element face
8410691 The error message shown below missing from the Nastran help documentation isadded back in 2019.2. USER FATAL MESSAGE 4663, IMPLAUSIBLE STRESS-STRAIN CURVE FOR ELEMENT-ID= ****The yield stress is found to be negative, which is implausible. This could occur ifthe stress-strain curve is defined for or extrapolated to the negative value of thestress. Check the table for stress-strain curve.
9448625 Different performance in SPDC and DISOFPM based on selected frequency. A casewith higher frequencies solves slower than a case with lower frequencies.
9371535 Nastran crashed in restart when running with SMP.
9370672 Nastran doesn't allow more processes than the number of frequencies in SOL 108and SOL 111 analyses.
9414690 In SOL402, if a contact set BCTSET was associated to a contact parameters cardBCTPAR2 containing more than 9 parameters, either a fatal error would occur orthe last parameters would be ignored, depending on the case.
9435310 In SOL 402, if a model contains both BCTSET and BOLT cards, a fatal error “%%%E01-MCT311, software error” could occur or the wrong contact set could beactivated or deactivated in a given subcase. This was due to a conflict in thenumbering of those contact and bolt sets.
9398734 In SOL 402, when requesting the GPFORCES output, depending on the size of themodel, a fatal error could occur due to a mis-match in an array size. Removing theoutput request would by-pass the problem.
9436864 In SOL 402, if a model contains a PSOLCZ card without any thickness specified, thesolver was using a zero-thickness, generating a fatal error “%%%E02-ASP147 ,ELEMENT xxx Zero thickness”. This has been changed to use 1.0 as a defaultthickness.
8412889 In SOL 402, if a model contains SPCD cards referenced by LOAD case control cardsand varying from subcase to subcase, and if the user gives a lot of memory to thesolver (gigabytes), the pre-processor would incorrectly write a lot of data, resultingin a very long computation time. This has no impact on the results but it can leadthe user to think that the computation will never finish. As a workaround, the usercan reduce the amount of memory passed to the solver. If this does not solve theissue, there are 2 additional workarounds. If SPCD cards are used as fixations witha 0 value, replace SPCD by SPC/SPC1; if not, use a DLOAD case control cardreferencing the SPCD cards with a table.
8414467 In SOL 402, when a TEMPD card is defined, a temperature load is applied to eachnode of the structure, excepted the internal nodes created automatically by thesolver when cutting the bolts. In the case of a restart, temperatures wereincorrectly applied to these additional nodes as well, creating incorrect thermalstrains in the bolt. As a workaround, the user can replace the TEMPD cards byequivalent TEMP cards.
8415361 When a TEMPD bulk card has the same ID as a DTEMP bulk card, it is ignored inSOL 402. As a workaround, the user can use only DTEMP cards to define thetemperature of all the nodes.
9373645 In SOL 402, in case of a restart, if SPCD, FORCE, MOMENT or RFORCE1 loads areused in the restart solution, incorrect results are obtained. There is no knownworkaround.
9429005 In SOL 402, when a preload subcase is followed by a dynamics subcase, the initialconditions (when defined) are not taken into account, which causes wrong results.There is no known workaround.
8407645 In SOL 402, the NLCNTL2 parameters DCON and PRCS, both related to contact,were ignored. This has been correct but those parameters cannot vary fromsubcase to subcase anymore.
9438345 In SOL 402, if the user requested the SPCFORCES for a MODAL (or FOURIER orCYCMODES) subcase, a crash would occur. Since this output request was notsupported for those subcase types, the user could simply remove it to by-pass theissue. This has now been fixed and the output request is ignored.
9464334 In SOL 402, with MATCRP TYPE=301 and NLCNTL2 CRINFAC=1.0, the solver couldcrash because during iteration of material law integration, some trial von Misesstress became negative. The workaround was to decrease MATCRP THRESH toimprove convergence or to change NLCNTL2 CRINFAC from 1.0 to 0.5. This hasnow been fixed.
8412526 SOL 402 does not support temperature-dependency of the mass density. The massspecified on the MAT1 card is used and a warning message is displayed saying thatthe TABLEM1 is ignored. In previous versions, the problem can be corrected byremoving the reference to the TABLEM1 on the MATT1 card.
9303135 In SOL 402, when requesting the stresses for modal subcases would result in azero output. This was due to the change in stress/strain measure performed in theprevious release.
9436569 In SOL 402, if the modal subcase was generating a fatal error, for example if noeigen value was found in the requested interval, the solution would still finish witha successful return value. It was up to the user to look for the error messages inthe F06 file. This issue has been fixed now.
9356421 There are 2 node types in an inactive element:Floating nodes: These are the nodes that will be removed because all elementsattached to these nodes will be removed. Node results such as acceleration,displacement should be removed.Remaining nodes: These nodes remain in the structure, because only someattached elements are removed, not all. Node results such as acceleration,displacement still need to be printed out.For issue 1, although the element on the face is inactive, the nodes of this elementare remaining nodes.For issue 2, only the mid node lying on the edge of the cube with two visible facesis floating node. If you change the 3D mesh from parabolic tetrahedron to lineartetrahedron elements, then there are not floating nodes.
9437014 To be consistent with SOL 402, for MATCZ bulk data entry, when Y0S > GIC, thesoftware issues a FATAL message in SOL 401. The description in QRG documentwas updated for Y0S of MATCZ

The following tables list the supported 64-bit operating system requirements to run Simcenter Nastran 2019.2. The product was tested using these operating system levels.

INTEL Windows (64-bit)
- Win Server 2012 R2
- Win Server 2016
- Windows 10 Professional and Enterprise