1-11340943 Documentation - Figure 4-35(a) in Linear Static Users Guide has an error See figure 4-35 titled Force, Moments, and Stresses in Plate Elements, specifically the (a) figure for Forces in the Linear Statics Users Guide PDF version. The Fxy arrow at the top of the figure is in the wrong direction. It shows pointing to the readers left and it should point to the readers right. The correct figure is shown in the Reference Manual, Figure 3-30 Forces in Shell Elements PDF version. 1-14251410 2008.0.0 (MD R3) Optimization/Design Sensitivity - RLOAD1 - Enforced Displacement, UWM 2071 (DPD2) SOL 200, analysis=mfreq, may produce incorrect results for enforced displacement specified with the RLOAD1 entry. Avoidance: This problem can be avoided by expanding output request set to include the grids with enforced motion. 1-14602401 2008.0.0 (MD R3) Submittal - SDIR must be write accessible if QUE= is specified If MSC.Nastran is submitted with QUE=NAME, then the script checks that "SDIR" is writable. In the case of QUEUES, a user may be on a headnode of a cluster, and not have write access to SDIR on the headnode, but only on the compute nodes. Avoidance: The check does NOT occur if the job is run with NODE=node_name. The check should not be there if there are queues as well. 1-15000431 Nonlinear Transient Analysis - CBUSH - Incorrect Element Forces If Field 6 of the TSTEPNL is set to AUTO, the element forces for the CBUSH element with PBUSHT will be incorrect in a SOL 129 run. Note that the AUTO and TSTEP methods have been de-documented after version 70.5. Also SOL 109 yields correct results for the test case. The element forces are correct when using the ADAPT (default) method. 1-15056901 2008.0.0 (MD R3) Distributed Memory Parallel - DOMAINSOLVER - Misleading Error Msg for ACMS w/ NUMDOM=1 Specifying NUMDOM=1 on the DOMAINSOLVER command, with ACMS, leads to an unclear error message: UFM 3200 (SEQP) LOGIC ERROR DETECTED BY SUBROUTINEC2FREE 1000 = LOCATE CODE OR VALUE. 1-19458251 Documentation - MSC.Documentation (pdf files) crash with Acrobat reader version 7 The MSC.Documentation (pdf files) can be properly browsed with Acrobat Reader version 5 and 6, but when using version 7, Acrobat reader crashes. 1-19459484 2008.0.0 (MD R3) Documentation/QRG - PCOMP - Remark 11 obsolete, kdiff is now calculated for unsym layup Remark 11 under the PCOMP Quick Reference Guide description that indiates that it is not recommended to use an unsymmetric layup in solutions that require differential stiffness is not longer applicable. 1-57215920 2008.0.0 (MD R3) SOL 400 - MATORT - UFM 2042 SOL 400 does not support MATORT and if referenced by a PSOLID entry will fail with the message: *** USER FATAL MESSAGE 2042 (M3DCONT1) MISSING MATERIAL TABLE 1 FOR ELEMENT 51585 1-58051791 2008.0.0 (MD R3) External Superelements/Fluid Response - DMIGPCH - Zero Fluid Response The fluid response of grid points that are interior to the fluid cavity and interior to fluid-structure superelement, are zero in an assembly run using DMIGPCH external superelement. Response on DISP output request in Sol 103 EXTSE creation run produces correct response using DMIGOP2 method. An assembly run also produces correct, distinct structure and fluid eigentables with DMIGPCH, so the VAFSX & AAX DMIG's are being processed, but the fluid modal solution is not calculated. 1-59206949 2008.0.0 (MD R3) Elements - PBRSECT - Arbitrary Beam coordinate system incorrect, bending stress incorrect. Arbitrary beam produce incorrect bending stress data recovery. 1-59715698 2008.0.0 (MD R3) Eigenvalue Analysis - CGAP - UFM 4016 (TA1ETD) Nastran 2005r3 and earlier releases may fail with the message: USER FATAL MESSAGE 4016 (TA1ETD) when an eigenvalue analyis contains gap elements and TEMP entries. Avoidance: Include a TEMPD entry. 1-59796301 2008.0.0 (MD R3) Complex Eigenvalues - SOL 110 with ACMS produced wrong answers with K2PP, M2PP, B2PP or TFL If K2PP, M2PP, and B2PP are used in SOL 110 along with ACMS then the run will produce wrong answers. Avoidance: Insert the following DMAP alter compile semceig alter 'dbdelete datablk=vgdr','' delete /vgdr,,,, $ compile fdrmgen alter 'return'(2),'' if ( app='ceigen' and not(mdreig) ) return $ compile spdr1 alter 'return'(2),'' endalter 1-59955051 2008.0.0 (MD R3) FRF - CONM2 - SFM 307 in FRFINFO module An FRF generation run with stress requests and CONM2 elements fails with the message: SYSTEM FATAL MESSAGE 307 in the FRFINFO module. 1-59961781 2008.0.0 (MD R3) SOL 700 - D2RAUTO is not translated. The D2RAUTO entry is not translated to the dytr.dat file and there are no warning messages. 1-60037683 2008.0.0 (MD R3) FRF - PLOTEL entries result in SFM 3200 (Logic Error 27) in FRFINFO module An FRF generation run with PLOTEL entries and a stress request for a non-existent element fails with the message: SYSTEM FATAL MESSAGE 3200 (Logic Error 27) in the FRFINFO module. Avoidance: Add a stress or force output request for a valid element 1-60037787 2008.0.0 (MD R3) FRF - PLOTEL entries result in SFM 3200 (Logic Error 12) in FRFINFO module An FRF generation run with PLOTEL entries and no stress or force output requests fails with the message SYSTEM FATAL MESSAGE 3200 (Logic Error 12) in the FRFINFO module. Avoidance: Add a stress or force output request for a valid element 1-60162658 2008.0.0 (MD R3) SOL 600 - Multiple MOMENT or FORCE entries for the same grid leads to incorrect loads For all versions of SOL600 (through mdr2 and v2007), when multiple MOMENT or FORCE entries for the SAME grid are combined using the LOAD Bulk Data Entry, there is a problem during translation that leads to incorrect loads. Avoidance: To get the correct results using current SOL 600, all moments and forces at the same grid which are applied in the same load case must be combined onto one FORCE or MOMENT entry. 1-60164831 2008.0.0 (MD R3) SOL 400 - Hybrid Damping is not implemeted into SOL400 correctly. The Hybrid Damping for the transient analysis is implemented in SOL129. It is found that this damping is not implemented correctely in SOL 400. 1-60222617 2008.0.0 (MD R3) Aeroelastic Analysis - MONITOR=ALL results in Complex Data Output A request for MONITOR point output should result in REAL/IMAGINARY output by default as the Quick Reference Guide states. However, the request: MONITOR=ALL is the same as MONITOR(PHASE)=ALL. Avoidance: To obtain the REAL/IMAGINARY output, one must specify MONITOR(REAL)=ALL. 1-60303381 2008.0.0 (MD R3) Eigenvalue Analysis - Modal analysis with ACMS and F1>0.0 may yield results outside the range If the F1 field on the EIGRL bulk data entry is greater than 0.0, ACMS will ignore the F1 entry and possibly produce eigenvalues between 0.0Hz and "F1" Hertz. Avoidance: Insert the following DMAP alter. compile invlan alter 'ufmin\<', '' $ IF ( UFMIN=0) CONTROL(3,1); Default = 0.1 It should read: Used for Relative Checking - Maximum residual force ratio (maximum allowable value of maximu residual force divided by maximum reaction force) or displacement ratio (maximum allowable value of the change in displacement increment divided by displacement increment) depending on convtyp (Real>=0) CONTROL(3,1); Default = 0.1 1-60830288 2008.0.0 (MD R3) Documentation/QRG - SOL 600 - MARCSAME Parameter According to the example on page 718 and Parameter Applicability Table on page 839, the QRG indicates that PARAM,MARCSAME must be specified in the bulk data. In fact it must be specified in the case control above the first subcase else the folloiwng warning message is issued: Severe Warning in SOL 600 translator - Case Control load ID 999 Appears in more than one subcase. For SOL 600, this may lead to wrong results. The subcases are 2 and 1 If you are sure the results will be acceptable, re-run with Case Control PARAM,MARCSAME,1 above the first subcase 1-61001551 2008.0.0 (MD R3) SOL 400 - XYPLOT - XYPLOT request fails with segmentation violation in SOL 400. If there are too many subcases or steps in SOL 400, the xyplot request in case control packet may fail with a segmentation violation. 1-61168951 2008.0.0 (MD R3) SOL 700 - PEULER1, SET1, and EOSGAM entries not translated PEULER1, SET1, and EOSGAM entries not translated 1-61236481 2008.0.0 (MD R3) SOL 400 - PARAM,LGDISP,1 - Large rotation, poor convergence If PARAM,LGDISP,1 is requested in SOL 400, the solution may have difficulty converging if the grid point rotataion is larger than 180 degree. 1-62064821 2008.0.0 (MD R3) Dynamics - TFL command, TF entry - Inefficient Processing of Symmetric Matrices There is no checking for symmetric matrices produced by the TFL Case Control command. The program then assumes the matrices are unsymmetric and this assumption causes the selection of unsymmetric methods in complex eigenvalue, frequency response, and transient response even if the matrices are actually symmetric. 1-62178001 2008.0.0 (MD R3) Miscellaneous - User-defined PARAM keywords capability is broken The user-defined PARAM keywords capability was broken when enhancements to the "remote query" capability were added. 1-62884471 2008.0.0 (MD R3) SOL 400 - UWM 602 issued if non-unique set ID's are used in steps in SOL400 In SOL 400, if non-unique set ID's are used in multple steps, then UWM 602 will be issued and the last defined set is used. However for multiple subcases, all defined sets are used, not just the last defined one. Sets defined in steps should behave the same as sets defined in subcases. 1-62947313 2008.0.0 (MD R3) Connector Elements - SET, CWELD - Incorrect CWELD element forces Patch to patch CWELD element forces may be incorrect when data recovery is performed on a subset of the CWELD elements. 1-63212781 2008.0.0 (MD R3) Frequency Response - UMFPACK, SYSTEM(209)=16 - Wrong Answers When using the UMFPACK solver option, "system(209)=16", the FRRD1 module, i.e. frequency response, will produce wrong answers when the dynamic matrix is real and the load matrix is complex. 1-63278281 2008.0.0 (MD R3) Performance - Unsymmetric Decomp is slow on IA64 In MDR3 rnd 0 I optimized TXUYRD (-O3), the time went from over 8 hrs to less than 3 hrs. jjg 12/07/07 If an analysis is exhibiting poor performance in unsymmetric decomp (UPSD) in MSC.Nastran Version 2007r1 or MD R2.1 compared to prior releases or other systems, then a new shared object is avaible for download that should rectify the problem. The download is available at: http://www.mscsoftware.com/support/software_updates/index.cfm#nastran Another aviodance is too run with "mode=i8" This error is fixed in MSC r1a (rnd 36). I am leaving the quality open until MD is fixed. 1-63338536 2008.0.0 (MD R3) SOL 600 - "CONSTANT" gets written to marc deck which may lead to wrong answers SOL 600 may incorrectly write "CONSTANT" to the translated marc input deck. For elastic analysis with elements that offer constant dilatation option (elements 7, 10, 11, etc.), use of CONSTANT is wrong and can change the results significantly. CONSTANT should only be used in conjunction with elasto-plastic analysis. Avoidance: Specify STOP on the SOL 600 entry, edit the marc.dat file and remove the CONSTANT entryt and submit directly to marc, 1-63659021 2008.0.0 (MD R3) SOL 400 - AUTOSPC fails if there are kinematic element in the model For SOL400, If there are rigid elements with RIGID=LAGR case control command, the AUTOSPC command will not work correctly. 1-63833183 2008.0.0 (MD R3) Optimization/Design Sensitivity - Linear Gaps - SFM 3007 (SOLVE) In a SOL200, when linear gaps are defined, then if internal iterations are required the analysis aborts with the following error : ^^^ END ITERATION 1 , CHANGES 2 *** SYSTEM FATAL MESSAGE 3007 (SOLVE) ILLEGAL INPUT TO SUBROUTINE FACDRV Avoidance: SSSALTER cditer200a.alt has been developed to avoid this error. 1-63950251 2008.0.0 (MD R3) Elements - PBRSECT - Arbitrary Beam failed with intersecting segments message in the log file Arbitrary Beam problems, PBRSECT, may fail with a message indicating "intersecting segments" in the log file Avoidance: 1. The "intersecting" message typically is caused by short segments with thick varying thicknesses. If possible, keep the thickness of short segments constant. 2. For multiple cells reverse the sequence of the divider. 1-63982967 2008.0.0 (MD R3) Aeroelastic Analysis - XYPUNCH ignores XMAX in KE-Method of Flutter The XYPUNCH of the vg data in the KE-method of flutter may ignore the XMAX statement, leading to a much larger punch file. The nature of the error appears to be rather erratic between machines and releases. i 1-64426459 2008.0.0 (MD R3) SOL 600 - NLPARM Case Control Command not translated correctly SOL 600 input with multiple subcases may incorrectly increment the referenced nlparm id and not translate the entry correctly. . 1-64535990 2008.0.0 (MD R3) Optimization/Desing Sensitivity - dmp>1 and scr=no - UFM 1126 in DISUTIL for EPT SOL 200 optimization runs with dmp>1 and scr=no may fail with the message: *** USER FATAL MESSAGE 1126 (GNFIST) DMAP MODULE DISUTIL IS ATTEMPTING TO OUTPUT DATABLOCK 201 NAME = EPT , WHICH ALREADY EXISTS. USER ACTION: DELETE THE DATABLOCK, OR USE FILE EPT=OVRWRT Avoidance: Insert following DMAP alter: compile desopt alter 'delete.*scstm,pval,epta' delete /ept,geom2,geom3,geom4,mpt $ endalter 1-64550902 2008.0.0 (MD R3) SOL 600 - Marc translator error with CWELD/PWELD formulation In SOL 600, if the "PARAM,MARCWELD,1" is set, analysis with CWELDs & Type ELEMID, the shell element identification numbers of elements on patch A and B (i.e., SHIDA, SHIDB) will not be translated into the marc.dat file correctly. 1-64762002 2008.0.0 (MD R3) ADAMSMNF - Fluid in a tank gives wrong ‘Global body properties’ Current ADAMS code can not handle coupled mass formulations of which virtual mass is a subset, and will incorrect "Global Body Properties". In MD R3 (version 2008.0.0) the DMAP will issue a fatal message similar to the one issued if param,coupmass is present for virtual mass in an adams flex MNF run. 1-64783158 2008.0.0 (MD R3) Composites - PCOMP - Improve UFM 315 A PCOMP entry that specifies FT and not SB will fail with a message similar to the one below. Note: this error is first trapped in V2005.1, prior versions allow the model to run. In MD R3 (version 2008.0.0) this message will be improved by adding "SB field is required when FT field is specified." *** USER FATAL MESSAGE 315 (IFPDRV) FORMAT ERROR ON BULK DATA ENTRY PCOMP PCOMP 1000 -.06859 0.0 TSAI 70.0 0.0 +000001 ++000001 10 2.30 35 0.40 +000002 USER/PROGRAMMER INFORMATION: BULKNM = PCOMP (Name of the Bulk Data Entry) KM = 0 (0:First Line, >0: Continuation Lines) J+1 = 5 (Data Field Number, 1 through 10) MF() = 0 (Field Data Format: 0=Blank, 1=Integer, 2=Real, 3=BCD, 4=Double, -32767=End-of-Data) M() = (Field Data Value) CALLSB = IXPCOMP (Name of subroutine) 1-64932461 2008.0.0 (MD R3) Optimization/Design Sensitivity - DESGLB, DRESP2, DRESP1 - Illegal Input to dsad5 If a SOL 200 optimization analysis contains a DESGLB case control command that references DRESP2 entries that references a DRESP1 with weight as one of the arguments, the analysis may fail with a message that indicates illegal input to dsad5. Avodiance: Reference dresp1 of weight directly on DRESP2 solves the problem with 20055. However, it does not work on 2006t1. 1-65037771 2008.0.0 (MD R3) Documentation/QRG - SOL 600 - PARAM, MARCOFFT The QRG documentation for PARAMETER MARCOFFT references a PARAM "MOFFOCRE" this should be "MOFFCORE". 1-65037789 2008.0.0 (MD R3) Documentation/QRG - SOL 600 - PARAM, MOFFCORE The QRG description of PARAM MOFFCORE contains a reference to "MARCOFFFT". The reference should be to "MARCOFFT" 1-65049111 2008.0.0 (MD R3) Nonlinear CASI solve fails in NLSOLV when "save/restore memory" option is turned off by user The NLSOLV module implementation of the CASI iterative solver interface includes an option to save/restore NLSOLV memory used around the call to the CASI solver. The default action is to perform the save/restore. This option can be disabled by the user. When disabled, very large problems may fail with a segmentation error in the NLSOLV module. Avoidance: Do not disable the save/restore option. 1-65063971 2008.0.0 (MD R3) ADAMSMNF - PARAM,COUPMASS,0 - Virtual MASS should not be allowed with Nastra/ADAMS interface The nastran-adams interface does not currently support a coupled virtual mass formulation. Thus, if param,coupmass,0 is specified in combination with ADAMSMNF flexbody=yes is specified a fatal message is issued. . 1-65153748 2008.0.0 (MD R3) Frequency Response - MODESELECT - Incorrect labels for PFMODE and PFPANEL The modeselect command may lead to incorrect mode id, and frequency labels in the PFMODE output. Avoidance: Use the following DMAP alter compile semfreq $ alter 'call gmax','' $ CALL GMAx, CASEg ,MATPOOLs,EQDYN ,TFPOOL ,bgpdts , ,cstms ,MAA ,GOt ,goq , GM ,USETD ,KAA ,BAA ,K4DD , PHsA ,LAMAs ,DITs ,VAFS ,PHFA , lamaf ,Aaa ,kddd ,mechg ,lammats , lammatf ,mis ,mif ,dynamic ,vgfd , rotor ,rotmat ,day ,nzyg ,goy , phba ,efmfacf ,efmfacs ,mdphiq ,mdgot , mdgoq ,mdvft ,mdmap ,mdgrdrm ,VEFS , KPP ,MPP ,BPP ,uset / , ,CMDD ,CK2DD ,CB2DD , MHH ,BHH ,KHH ,PHDH ,VPHFS , K2MG ,M2MG ,B2MG ,K2SF ,M2SF , B2SF ,k2ss ,m2ss ,b2ss ,btab , modselsv,MODELSTS,MODELSTF,vgfdt , , b2dsf ,m2dsf ,k2dsf ,k4hh ,phbh , lamasn ,lamafn / LUSETD /SOLTYP /APP /NOUE /NOA / FALSE /S,FMODE /FALSE /FALSE /FS / NOMSET /fonly /S,FMODEF/freqdep /s,xfastfr/ mdotm0 /rsonly $ alter 'call pfact','' $ CALL PFACT CASEF ,BGPDTS ,USET ,USETD ,SILS , GOT ,GOQ ,GM ,PHDH ,LAMASn , LAMAFn ,UHF ,FOL ,PHF ,VGFS , VPHFS ,KHH ,BHH ,MHH ,ABE , AGG ,mdotm ,YPF1 ,RPY ,YPF / SMPF ,SMPFD ,OSMPF1 ,AFMPF ,AFMPFD , OAFMPF1 ,ASMPF ,ASMPFD ,OASMPF1 ,APMPF , APMPFD ,OAPMPF1 ,APPF ,APPFD ,OAPPF1 , AGPF ,AGPFD ,OAGPF1 / NOASET /NOOSET /NOQSET /NOTSET /NOMSET / NOA /NOUE /APP /SOLTYP /FS / FONLY $ compile gmax $ alter 1,1 $ SUBDMAP GMAx CASES ,MATPOOL ,EQDYN ,TFPOOL ,BGPDTS , EQEXINS ,CSTMS ,MMAA ,GOT ,GOQ , GM ,USETD ,MKAA ,BAA ,K4DD , PHSA ,LAMAS ,DIT ,VASF ,PHFA , LAMAF ,AAA ,KDDD ,MECHG ,LAMMATS , LAMMATF ,MIS ,MIF ,DYNAMIC ,VGFD , ROTOR ,ROTMAT ,DAY ,NZYG ,GOY , PHBA ,EFMFACF ,EFMFACS ,MDPHIQ ,MDGOT , MDGOQ ,MDVFT ,MDMAP ,MDGRDRM ,VESF , KPP ,MPP ,BPP ,USET / KDD ,BDD ,MDD ,KDDMACC ,BDDMACC , MHH ,BHH ,KHH ,PHDH ,VPHSF , K2MG ,M2MG ,B2MG ,K2SF ,M2SF , B2SF ,K2SS ,M2SS ,B2SS ,BTAB , MODSELSV,MODELSTS,MODELSTF,VGFDT ,UNUSED3 , B2DSF ,M2DSF ,K2DSF ,K4HH ,PHBH , LAMASN ,LAMAFN / LUSETD /SOLTYP /APP /NOUE /NOA / NONLNR /FMODE /NLHEAT /CYCLIC /FS / NOMSET /FONLY /FMODEF /FREQDEP /XFASTFR / DOOTM /RSONLY $ alter 'call gmam','' $ CALL GMAM MSAA ,BSAA ,KSAA ,K4SAA ,KDSAA , PHBSA ,MFAA ,BFAA ,KFAA ,K4FAA , PHBFA ,AAA ,VASF ,M2DD ,B2DD , K2DD ,VESF ,USETD ,PHSA ,LAMAS , MIS ,LAMMATS ,EFMFACS ,PHFA ,LAMAF , MIF ,LAMMATF ,EFMFACF ,DIT ,MDPHIQ , MDGOT ,MDGOQ ,MDVFT ,MDMAP ,MDGRDRM , CASES ,DYNAMIC ,USET ,GM ,GOT , ROTOR ,ROTMAT / MHH ,BHH ,KHH ,K4HH ,PHDH , VPHSF ,BTAB ,MODSELSV,MODELSTS,MODELSTF, PHBH ,LAMASN ,LAMAFN / APP /SOLTYP /FS /FONLY /ACSYMl / UE /NOUE /DOOTM /S,FMODE /S,FMODEF/ S,XFASTFR/rsonly $ compile gmam $ alter 1,1 $ SUBDMAP GMAM MSAA ,BSAA ,KSAA ,K4SAA ,KDSAA , PHBSA ,MFAA ,BFAA ,KFAA ,K4FAA , PHBFA ,AAA ,VASF ,M2DD ,B2DD , K2DD ,VESF ,USETD ,PHSA ,LAMAS , MIS ,LAMMATS ,EFMFACS ,PHFA ,LAMAF , MIF ,LAMMATF ,EFMFACF ,DIT ,MDPHIQ , MDGOT ,MDGOQ ,MDVFT ,MDMAP ,MDGRDRM , CASES ,DYNAMIC ,USET ,GM ,GOT , ROTOR ,ROTMAT / MHH ,BHH ,KHH ,K4HH ,PHDH , VPHSF ,BTAB ,MODSELSV,MODELSTS,MODELSTF, PHBH ,LAMASN ,LAMAFN/ APP /SOLTYP /FS /FONLY /ACSYMl / UE /NOUE /DOOTM /FMODE /FMODEF / XFASTFR /rsonly $ 1-65322401 2007.1.0 SOL 600 - CBUSH using CID gets incorrect translator error If a NASTRAN input file using CBUSH elements that reference a local coordinate system (the CID filed) is run in NASTRAN v2007r1, the translator stops with the following message: ** Severe Warning in SOL 600 translator - CBUSH G0, X1, X2, X3, CID for CBUSH 1 is not fully supported. To ignore these fields, enter bulk data param,marcbush,1 This message is inappropriate, as in version 2007r1, Marc has an equivalent CBUSH that can use the local coordinate system. Setting PARAM,MARCBUSH,-1 allows the translator to translate the element correctly. The message should indicate that marc can not be run using dmp with this element is used. 1-65390757 2008.0.0 (MD R3) MDACMS - acoustics, modeselect(fluid) - UFM 1126 for PHFAHX In an mdacms acoustic run using sol 111 the following error message can occur when modeselect(fluid) requests a subset of the fluid modes: *** USER FATAL MESSAGE 1126 (GNFIST) DMAP MODULE PARTN IS ATTEMPTING TO OUTPUT DATABLOCK 201 NAME = PHFAHX , WHICH ALREADY EXISTS. Avoidance: Use PARAM,GMA,OLD or insert the following DMAP alter: compile gmam alter 1 file phfahx=ovrwrt $ 1-65434816 2008.0.0 (MD R3) SOL 600 - Beam offset for SOL 600 Beams with offsets are translated incorrectly in SOL 600. The offset is ignored and this may lead to wrong answers. 1-65662259 2008.0.0 (MD R3) Documentation/QRG - SOL 600 Continue option is incorrect The SOL 600 entry CONTINUE options 3-5 are not supported and will be removed from the QRG. 1-65900127 2008.0.0 (MD R3) [KCE:3039] Optimization/Design Sensitivity - DSAPRT does not print associated subcases for DRESP3 DSAPRT does not print associated subcases for DRESP3 1-65930441 2008.0.0 (MD R3) SOL 400 - DMAP modification fails with SFA 427 - LOPBLK table has exceeded its maximum entries When modifying the SOL 400 DMAP with addtional DO-WHILE loops, the job will fail with SFA 427 - The LOPBLK tabe has exceeded its maximum 1200 number of entries. 1-65937261 2008.0.0 (MD R3) SOL 400 - BCONTACT - Position control of rigid body is incorrect In SOL 400 if a deformable body and a rigid body are in initial contact, the position control is incorrect. 1-65983571 2008.0.0 (MD R3) Numerical Methods - DCMP - Signal 11, Segmentation Violation The combination of low available Open Core memory and large Nastran BUFFSIZE may lead to a Segmentation Violation (Signal 11) in the DCMP or READ modules during matrix factorization. This non-standard termination causes Nastran to halt with no Fatal message. Error messages generated by the operating system may appear in the log file. Avoidance: There are two potential avoidances: 1. Increase memory via the "mem=" command line keyword. 2. Reduce the BUFFSIZE. 1-66000287 2008.0.0 (MD R3) SOL 600 - PLOAD4(or any mechanical loads) & MARCIN entries - Incorrect translation When PLOAD4 or any mechanical loads are applied coupled with user specified MARCIN entries for the history data, the MARCIN entry translations will be written in the wrong location. 1-66013381 2008.0.0 (MD R3) SOL 600 - Marc's parameter option LUMP will not be applied in transient heat transfer analysis According to the MD Nastran R2.1 Quick Reference Guide, the "PARAM,MARCLUMP,0 or 1 selects whetherr the lumped specific heat matrix will be used in a transient SOL 600 heat transfer analyses. However, marc's parameter option LUMP will not be applied in transient heat transfer analysis and a consistent specific heat matrix will be used. 1-66013451 2008.0.0 (MD R3) SOL 600 - NLHEATC entry is not translated The NLHEATC entry is not translated for SOL 600 transient heat transfer analysis. it is only translated for steady state heat transfer analyisis. 1-66013500 2008.0.0 (MD R3) SOL 600 - MARCIN entries for subcases not translated in heat transfer analysis MARCIN entries for SUBCASES will not be converted into Marc input for SOL 600 heat transfer analyses 1-66026575 2008.0.0 (MD R3) Optimization - DSAD - BIOMSG: ERROR 3001 HAS OCCURRED IN ROUTINE GREAD The DSAD module attempts to identify grid id's on the dresp1 entry that do not exist. If the missiing grid is called out in the final subcase and there are no other grids invoked by this subcase, the message: BIOMSG: ERROR 3001 HAS OCCURRED IN ROUTINE GREAD , FILE INDEX = 0. wil be output from the DSAD module. Avoidance: When the above message occurs in the DSAD module, a check should be made to inusre that all the GRID id's referenced on DRESP1 entries exist. 1-66026799 2008.0.0 (MD R3) Optimization/Design Sensitivity - DRSPAN, DESGLB and DESSUB may fail in dsad module - SFM 6926 A SOL 200 job with DRSPAN and DESGLB/DESSUB that point to the same constraint set may fail in the DSAD module with SFM 6296 unable to allocate a data block. Avodiance: This is a user input error which can be avoided by removing the desglb command in the case control. The presence of drspan turns off the checking logic for usage of DRESP1 in DESGLB. The checking logic should be active for those DRESP1s that have nothing to do with DRSPAN. 1-66195125 2008.0.0 (MD R3) Dynamic Analysis - EPOINTs and DELAY entries in dynamic response, wrong answers, UWM 4193 (DPD7) If a model contains EPOINTs where at least one EPOINT has an id lower than a loaded GRID point and the model contains DELAY entries, the answers may be wrong. The f06 file contains the warning message: *** USER WARNING MESSAGE 4193 (DPD7) NO EXCITATION HAS BEEN SPECIFIED FOR ANY OF THE COMPONENTS SPECIFIED ON DELAY SET X REFERENCED ON TLOADi ENTRY X where X are integer values. Avoidance: Renumber the epoints. Ensure the EPOINT ids are always higher than the largest GRID poiint or SPOINT ids. 1-66224284 2008.0.0 (MD R3) Aeroelastic Analysis - SPLINE, SPRELAX, SPBLNDi - Incorrect Results There are two errors that affect spline processing with the spline relax feature and the spline blend feature. 1. If the blend region does not cover the entire overlap region, the splines are averaged in the unblended area when they should take their value from: f_1(x) = 1 and f_2(x) = 0 for x < -D1 f_1(x) = 0 and f_2(x) = 1 for x > D2 2. If the same spline is relaxed on two different SPRELAX entries it is possible that neither is entry is processed. 1-66226424 2008.0.0 (MD R3) Aeroelastic Analysis - No velocity input on a AERO entry leads to wrong answers or poor message In aeroelastic analysis if the user neglects to specify a velocity value on the AERO bulk data entry the analysis may fail with the message below or wrong answers may be produced. The FPU has been reset after floating point exception 1-66267651 2008.0.0 (MD R3) Optimization/Design Sensitivity - PBMSECT - UFM 7602 (DOPR1I) with arbitrary beasm cross section SOL 200 with arbitrary beam cross sections, PBMSECT, may encounter the message: USER FATAL MESSAGE 7602 (DOPR1I) 1-66309821 2008.0.0 (MD R3) Acoustic Analysis - GP5 issues UIM 6207 with null method type The GP5 module issues USER INFORMATION MESSAGE 6207 when there is no ACMODL bulk data entry defined. The message is supposed to indicate the method that will be used, but the field is blank becuase it has not been initialized prior to the message being issued. 1-66309911 2008.0.0 (MD R3) Dynamic Analysis - MCFRACTION output incorrect when MODESELECT used When the MODESELECT case control command is used to select modes to be used in the modal analysis and the MCFRACTION command is also present in the case control, the structure's natural frequency value in the MCFRACTION output may be incorrect. The value is wrong if the number of modes selected is less than the number of modes actually found in the eigenvalue analysis. The mode number is correct when it is only the natural frequncy value that is wrong. 1-66384331 2008.0.0 (MD R3) Acoustic Analysis - PFMODE case control commdand STRUCTMP keyword The PFMODE case control command requests modal participation factor output (MPF). The format of the command allows for specification of either STRUCTURE or FLUID. The STRUCTURE option produces structural MPF output using structural natural frequencies. The FLUID option produces acoustic MPF output using structural natural frequencies for acoustic structure MPF and fluid natural frequncies for acoustic fluid and panel MPF. For both cases, the STRUCTMP keyword is used to define the number of structural modes that are used during MPF calculations. When both PFMODE(STRUCTURE...) and PFMODE(FLUID...) are present in the input case control, the STRUCTMP keyword is not interpreted according to the QRG description. Rather, the value used in the structure MPF computations is the maximum of the STRUCTMP keyword values supplied in the two commands. 1-66552807 2008.0.0 (MD R3) Acoustics - CHACAB - Incorrect mass for 16 node element The mass matrix for the 16 node CHACAB acoustic barrier element is not correct. 1-66624921 2008.0.0 (MD R3) Eigenvalue Analysis - NLRSFD - UFM (DPD4) for entry that should be ignored An incorrect NLRSFD entry in eigenvalue analyis will cause the DPD4 module to be executed and a User Fatal Message to be issued. Avoidance: Correct the entry. 1-66661307 2008.0.0 (MD R3) SOL 600 - ISTRESS not honored - Wrong Answers The NASTRAN ISTRESS is not translated into the Marc input file and the job runs to completion without warning yielding results without the ISTRESS effects. 1-66825861 2008.0.0 (MD R3) Optimization/Design Sensitivity - DSAD - Eigenvector Contraints, SFM GREAD A SOL 200 job with eigenvector constraints combined with other analysis types may fail in the DSAD module with an error during GREAD. Avodiance: None. If possible, run SOL 200 job with mode and eigenvector constraints only. 1-66833371 2008.0.0 (MD R3) Connect Elements - CWELD, CFAST, CQUAD4, CQUAD8 - UFM 7632 (EWLDPD) If CWELD or CFAST elements are used to connect a patch of CQUAD4 elements with CQUAD8 the following fatal message may be issued: *** USER FATAL MESSAGE 7632 (EWLDPD) FOR CWELD ELEMENT ID= 102, A CONNECTION TO ONE OF THE SHELL SURFACES CANNOT BE MADE. USER INFORMATION: THE FIRST THREE SHELL GRIDS ARE 51 53 61 *** USER FATAL MESSAGE 2026 (EMG) ELEMENT 102 GEOMETRY YIELDS UNREASONABLE MATRIX. ^^^ USER WARNING MESSAGE 9031 (ERRPH1) ^^^ NOGO ENCOUNTERED IN SUBDMAP SEMG 1-66916611 2008.0.0 (MD R3) Optimization/Design Sensitivity - ECHO NEWBULK - Incorrect punched bulk data A SOL 200 model with DVGRID and ECHO = PUNCH(NEWBULK) may contain incorrect information in the punched bulk data deck. Avoidance: The output produced using the despch parameters is correct and can be used to construct a new bulk data deck. 1-66918399 2008.0.0 (MD R3) Documentation/QRG - BCTABLE In the MDR2 QRG, the description of the BCTABLE entry indicates that it is for SOL100, it should indicate SOL 101. 1-66941791 2008.0.0 (MD R3) External Superelements and Acoustics - C-set - SFM 6144 (PARTN1PR) An external superelement creation with an acoustic cavity and C-set will result in SFM 6144 (PARTN1PR): *** SYSTEM FATAL MESSAGE 6144 (PARTN1PR) THE SIZES OF THE INPUT MATRICES AND PARTITIONING VECTORS ARE INCOMPATIBLE. SPECIFICALLY: The number of columns in KOA is not equal to the number of rows in the column partitioning vector VFT User Information: Size of input matrix and partitioning vectors: KOA : Rows= 4443394 by Cols= 2561 VFT : Rows= 562 VFO : Rows= 4443394 Avoidance: Insert the following DMAP alter: compile semr3 alter 'vfo',''(2) partn koa,vafs,vofs/,,,kotf/1 $ endalter 1-66953461 2008.0.0 (MD R3) SOL 400 - RIGID=LAGRANGE - SFM 4370 (SDCPRF) SOL 400 with RIGID=LAGRANGE may fail with SYTEM FATAL MESSAGE 4370 (SDCPRF) 1-66984621 2008.0.0 (MD R3) Connector Elements - MODGM2, CWELD, CFAST, PARAM,CHKRUN,2 does not funciton correctly An analysis with CWELD and/or CFAST elements will not be stopped after checking the connector elements when PARAM,CHKRUN,2 is set. 1-67131791 2008.0.0 (MD R3) AUTOMSET - RBE3 - SFM 4276 (BLDPRD) A simulation with RBE3’s that incorrectly define degrees of freedom as dependent and independent produces the follwing unsatisfactory fatal message if used with AUTOMSET (PARAM,AUTOMSET,YES). *** SYSTEM FATAL MESSAGE 4276 (BLDPRD) ERROR CODE 211 PID= 0 *** USER INFORMATION MESSAGE 4276 (BLDPRD) THE DIAG 44 RELATED A NASTRAN DUMP IS SENT TO LOG FILE. USER ACTION: THIS INFORMATION MAY BE USEFUL TO MSC FOR DEBUGGING YOUR PROBLEM. PLEASE CONTACT MSC CLIENT SUPPORT. 1-67273981 2008.0.0 (MD R3) Nonlinear Statics - ANALYSIS=HOT2COLD, scr=no - UFM 1126 (QQGNI) If scr=no is requested along with ANALYSIS=HOT2COLD and rigid elements or MPCs are present the run may terminate with UFM 1126 on data block QQGNI. Avoidance: Run with scr=yes and insert the following DMAP alter: compile nlstatic alter 1 file qqgni=ovrwrt $ endalter 1-67375341 2008.0.0 (MD R3) Complex Eigenvalue Analysis - HESS - Inconsistent Results A complex eigenvalue analysis using the Hess method where the mass matrix is many orders of magnitude different that the stiffness matrix may produce inconsistent results. Avoidance: Use the CLAN method as follows: EIGC 104 CLAN +EIGC +EIGC 1.0 0.0 50 50 500 or to run HESS with the following alter: compile ceigrs, list alter 65 matprn ckdxx,cbdxx,cmdxx//'long' $ add ckdxx,/ckdyy/1.e-8 $ add cbdxx,/cbdyy/1.e-8 $ add cmdxx,/cmdyy/1.e-8 $ equivx ckdyy/ckdxx/always$ equivx cbdyy/cbdxx/always$ equivx cmdyy/cmdxx/always$ 1-67457421 2008.0.0 (MD R3) Optimization/Design Sensitivity - PBRSECT/PBMSECT may fail in DOM9 a SOL 200 analysis with PBRSECT/PBMSECT entries in the design model may fail in the VKI code with intersecting boundary segments. This failure may occur as the element is designed since the segment may intersect each other as segments becoming thicker. Avoidance: Allow ample space for segments in abitrary beam elements to "grow" without intersecting. 1-67610541 2008.0.0 (MD R3) Documentation/QRG - CBAR, CBEAM OFFT field - Incorrect example The example in the QRG for both the CBEAM and CBAR shows an "O" as the 1st letter for the OFFT field. Remark 8 documents the allowed specifications. 1-67661477 2008.0.0 (MD R3) Aeroelastic Analysis - Multiple SPRELAX entries that share relaxed grids give wrong results. If there are multiple SPRELAX entries and the same grids are "relaxed" by two of more of these entries, the job runs but the results are incorrect. Avoidance: Do not attempt to relax the same aerodynamic grid from differing SPRELAX enties. 1-67677181 2008.0.0 (MD R3) Optimization/Design Sensitivity - PBMSECT, PBRSECT, DOM9 - Inefficient Processing SOL 200 with a large number of PBMSECT or PBRSECT, arbitrary beam cross section, and design variables may consume a large amount of cpu time in the DOM9 module. The time is spent updating sensitivity for nonlinear relations between design variable and properties. 1-67685315 2008.0.0 (MD R3) Documentation/QRG - OPGEOM parameter description incorrect The parameter description for OPGEOM reads, "OPGEOM < *1 prints the set definition for all degrees-of-freedom, including the aerodynamic degrees-of-freedom." This should be:, OPGEOM > -1 prints the set definition for all degrees-of-freedom, including the aerodynamic degrees-of-freedom. 1-67750169 2008.0.0 (MD R3) Elements - STRAIN, CTRIA6 - Null Strain Output If only STRAIN output is requested with PARAM,OMID,YES and CTRIA6 elements are present in the model, the STRAIN results for all the CTRIA6 elements are output as 0.0 when they should be nonzero. With PARAM,OMID,NO (default) the STRAIN results are nonzero. Avoidance: A request of STRESS output will cause the STRAIN results for the TRIA6 elements to be correct. 1-67770500 2008.0.0 (MD R3) SOL 600 - CBEAM - Orientation does not translate properly SOL 600 does not translate beams whose orientation references other than the basic coordinate system correctly. 1-67780355 2008.0.0 (MD R3) SOL 600 - Translation of MAT9 thermal expansion coefficients is incorrect. In SOL 600 the translation of MAT9's thermal expansion coefficients is incorrect. The old and new mapping is as follows: New Old α11 A1 A1 α12 A4 A2 α13 A6 A3 α22 A2 A4 α23 A5 A5 α33 A3 A6 1-67838600 2008.0.0 (MD R3) External Superelements - EXTSEOUT, MDACMS, and PARAM,FLUIDSE - SFM 6144 for MQQ External superelements, EXTSEOUT, combined with mdacms and param,fluidse will lead to the following fatal message: *** SYSTEM FATAL MESSAGE 6144 (MERGE1PR) THE SIZES OF THE INPUT MATRICES AND PARTITIONING VECTORS ARE INCOMPATIBLE. SPECIFICALLY: The number of columns in MQQ is not equal to the number of non-zeros in partitioning vector VFS The number of rows in MQQ is not equal to the number of non-zeros in partitioning vector VFS User Information: Size of input matrices and partitioning vectors: MQQ : Rows= 116 by Cols= 116 VFS : Rows= 2 no. of non-zeros= 1 no. of zeros= 1 1-67870768 2008.0.0 (MD R3) GPFORCE - SET - UWM 2354 (GPFDR), NERROR was set to -98 If a GPFORCE request is made in different subcases where each request specifies a different set, processing may be inefficient of the job may fail with the message: *** USER WARNING MESSAGE 2354 (GPFDR) MODULE IS UNABLE TO CONTINUE AND HAS BEEN TERMINATED DUE TO ERROR MESSAGE PRINTED ABOVE OR BELOW THIS MESSAGE. THIS ERROR OCCURRED IN CODE WHERE THE VARIABLE -NERROR- WAS SET = -98 *** SYSTEM WARNING MESSAGE 3002 (GPFDR) EOF ENCOUNTERED WHILE READING DATA BLOCK SCRATCH (FILE 314) IN SUBROUTINE GPFDFR 1-67922591 2008.0.0 (MD R3) Elements - CBEAM3, PBMSECT using VAM has swapped y/z axes A composite beam defined using the CBEAM3 and PBMSECT entries will swap the y/z axes. Avodiance: if possible, turn the orientation vector for CBEAM3 90 degree in YZ plan. 1-68046176 2008.0.0 (MD R3) Installation - MD Nastran R2.1 does not contain an i8 version for HP MPI MD Nastran R2.1 does not contain an i8 version for HP MPI. Critical clients have requested this system and it has been added for MD R3. 1-68288009 2008.0.0 (MD R3) Acoustic, Fluid Analysis - CETRA - PFLUID TET4 element Produces Wrong Answer The lack of a coupled mass matrix for the fluid formulation of the four noded CTETRA element can result in wrong answers and spurious modes. Setting NASTRAN,SYSTEM(446)=1 will result in the old answers. 1-68343941 2008.0.0 (MD R3) Topology Optimization - Stress constrainst lead to Illegal input to DSAH31 A message that indicates illegal input to DSAH31 will be issued in a SOL 200 topology optimization run when stress constraints are included. Avoidance: Remove the stress constraint from the problem definition 1-68477767 2008.0.0 (MD R3) Parallel Processing - DMP - DFMFR*: MAXIMUM ESTIMATED FRONT SIZE MAY BE EXCEEDED. When dmp processing is selected with MDMOES the ordering method must be Extreme (BEND). However, an error in the code may allow METIS to be chosen this may lead to poor performance and the following message:  *** SYSTEM INFORMATION MESSAGE 6916 (MREORDR)      DECOMP ORDERING METHOD CHOSEN: DEFAULT, ORDERING METHOD USED: METIS  DFMFR*: MAXIMUM ESTIMATED FRONT SIZE MAY BE EXCEEDED.  DFMFR*: MAXIMUM ESTIMATED FRONT SIZE MAY BE EXCEEDED.  DFMFR*: MAXIMUM ESTIMATED FRONT SIZE MAY BE EXCEEDED.  DFMFR*: MAXIMUM ESTIMATED FRONT SIZE MAY BE EXCEEDED. 1-68554289 2008.0.0 (MD R3) SOL 600 - CHEXA20 - *** error - alias set up on element type 150 which is not on sizing card” SOL 600 may translate the CHEXA20 element incorrectly. The translated alias entry may point to elment type 20 instead of 21. This may lead to the error message: *** error - alias set up on element type 150 which is not on sizing card” 1-68655761 2008.0.0 (MD R3) Aeroelastic Analysis - SPLINE7 - Splining nearly coincident points produces bad spline matrix Splining two coincident points with a SPLINE7 produces good results. However, if one of the points is moved by a very small amount, the resulting spline matrix is bad. 1-68672091 2008.0.0 (MD R3) SOL600 - QRG MATEP - Missing documentation for addrad method The QRG description of the MATEP "addrad" method was inadvertantly removed from the QRG. The addrad method should be described as: addrad for additive decomposition using the radial return process (SOL 600 only) 1-68799191 2008.0.0 (MD R3) Optimization/Design Sensitivity - DRESP2, CEIG, LAMA, EiGN, FLUT - Zero Sensitivity Optimization, SOL 200, with superelements where DRESP2 entries reference a global response via a DRESP1 entry, such as CEIG/LAMA/EIGN/FLUT, may fail due to zero sensitivity on violated constraints. 1-68928401 2008.0.0 (MD R3) DMP - PFMODE, PFPANEL, PFGRID - Extra output The modal participation commands PFMODE, PFPANEL, and PFGRID will result in extra output for the slave processes. Avoidance: The following alter can be used to avoid the problem: compile feafreq $ alter 'call pfact'(1,-1) $ if ( procid<=1 ) then $ alter 'IF ( NOAFPM=1 ) CALL AFPMDRCV' $ endif $ compile sedfreq $ alter 'call pfact'(1,-1) $ if ( procid<=1 ) then $ alter 'IF ( NOAFPM=1 ) CALL AFPMDRCV' $ endif $ alter 'IF ( NOAFPM=1 ) CALL AFPMEND','' $ IF ( NOAFPM=1 and procid<=1 ) CALL AFPMEND PVT,AFPMLIST $ compile semfreq $ alter 'call pfact'(1,-1) $ if ( procid<=1 ) then $ alter 'IF ( NOAFPM=1 ) CALL AFPMDRCV' $ endif $ alter 'IF ( NOAFPM=1 ) CALL AFPMEND','' $ IF ( NOAFPM=1 and procid<=1 ) CALL AFPMEND PVT,AFPMLIST $ 1-69091628 2008.0.0 (MD R3) Fluid-structure component modes with BNDFREEi or CSETi, andPARAM,SESDMAP and QSETi or PARAM,AUTOQSET If component modes and PARAM,SESDAMP,YES is requested on a fluid-structure then the run will terminate in subDMAP cmpmode with: SFM 3001 MPYAD SOL103 datablock 201 Avoidance: Insert the following DMAP alter: Compile cmpmode Alter 'IF ( NZBK4ZZ','' Paraml /mphvz/'name'////s,n,nomphvz $ IF ( NZBK4ZZ>0 and nomphvz>-1 ) then $ alter 'mpyad mvv,phivz' merge mphvz,,,,,vvfx/mphvz1/1 $ equivx mphvz1/mphvz/-1 $ 1-69113343 2008.0.0 (MD R3) File Management Section - INIT in RC(F) - UFM723(RDINIT) If INIT statement spans two or more lines in RC(F) file, MSC.Nastran 2007r1 and MD R2 will read only the first line and the folloiwng fatal message will be issued: INIT SCRATCH LOGICAL=(SCR1(2000MB),SCR2(2000MB),SCR3(2000MB), NASTRAN REAL=902037504 $(PROGRAM DEFAULT) *** USER FATAL MESSAGE 723 (RDINIT) LOGICAL= EXPRESSION HAS AN UNBALANCED NUMBER OF PARENTHESES Avoidance: Enter the INIT lines directly in the input deck. 1-69161608 2008.0.0 (MD R3) DMAP - INPUTT4, sparse, ASCII format - terminates without message When inputting an OUTPUT4 file created using the sparse format (IUNIT<0) and ASCII format (ASSIGN ... FORMATTED) the run may terminate in INPUTT4 without message. Specifically it occurs if the matrix has a string word length such that when it is multipied by the word size (e.g, 1 word for real single, 2 for real double and complex single, and 4 for complex double) exceeds 99,999,999. Avoidance: There is no known avoidance other than to not use the sparse format or switch to the binary format. 1-69175341 2008.0.0 (MD R3) Composites - Temperature-Dependent composites may cause wrong result or infinite loop in SDRCOMP Temperature-Dependent composite materials will cause either wrong result or an infinite loop in SDRCOMP Avoidance: Use param,nocomps,-1 1-69398011 2008.0.0 (MD R3) Optimization/Design Sensitivity - Complex Eigenvalue with Static Displacement Constraints Fails A SOL 200 job with complex eigenvalues and static displacement constraints may fail if a semi-analytical method is utilized for sensitivity computation. AVoidance: Use the adjoint load method which is the default. Specify 'param,autoadj,no' to turn on semi-analytical method. 1-69423731 2008.0.0 (MD R3) Postprocessing - PARAM,POST,0 - SFM 6062 (DBC), Poor Message The fatal message below indicates that the run failed while writig to the xdb file a primary index overflow. The xdb file may be corrupt. *** SYSTEM FATAL MESSAGE 6062 (DBC) **** DIOMSG ERROR MESSAGE 6 FROM SUBROUTINE WRTLST OVERFLOW OF DICTIONARY PRIMARY INDEX FOR DATA BASE UNIT 1. Avoidnace: . The avoidance is to rerun the job and include the following assign statement in the File Management Section of the input fie... ASSIGN DBC='YourFileName.xdb',RECL=32768,UNIT=40,DELETE 1-69539261 2008.0.0 (MD R3) Dynamics - auto-suport - wrong answers If the stiffness matrix is slightly ill-conditioned and non-singular the auto-suport feature may mistakenly assume the model is free-free and include Lagrange multipliers to support the structure and produce wrong answers. Avoidance: Insert PARAM,AUTOSPRT,NO the following DMAP alter: compile autosprt alter 'diagonal dk\/dk1',''(,4) type parm,,i,n,irow,xovr2 $ PARAML KXX1//'TRAILER'/1/S,N,NOXSET $ matmod dk,,,,,/srtlst,srtbool/35/1//2 $ xovr2 = noxset/2 $ paraml srtlst//'dmi'/1/xovr2/s,n,medrow $ irow = int(medrow) $ paraml dk//'dmi'/1/irow/s,n,kavsq $ kfilt=kavsq*kzero $ endalter 1-69578457 2008.0.0 (MD R3) Composites - CQUADR with PCOMP not picking up ply direction correctly leads to bad transverse shear CQUADR with PCOMP is not picking up the ply direction correctly and this may lead to incorrect transverse shear stiffness. Avoidance: Uset the CQUAD4. 1-69604951 2008.0.0 (MD R3) SOL 400 - VCCT - Incorrect defaults documented for GCII and GCIII field leads to UFM 315 (IFPDRV) Incorrect defaults for various fields of the VCCT entry may lead to the message: *** USER FATAL MESSAGE 315 (IFPDRV) FORMAT ERROR ON BULK DATA ENTRY VCCT The following discrepancies are noted: IDCR: QRG indicates default=1, currently these is no default Code will be change to default to 1. INCM: QRG indicates default=2, default=1 Documentation will be corrected. TIME: QRG has no default & SOL-600 only, current default=0.0 QRG will be corrected. GC-II: not supported GC-III: not supported TABGC-II: not supported TABGC-III: not supported Code will be changed to support: GC-II: added with default=0.0 (QRG changed accordingly) GC-III: added with default=0.0 (QRG changed accordingly) TABGC-II: added with default=0 TABGC-III: added with default=0 1-69900403 2008.0.0 (MD R3) Optimization/Design Sensitvity - Static and Complex Eigenvalue Subcases may lead to fatal message A SOL 200 job with static and mceig subcases may fail with a message indicating a missing CPHP, comple eigenvectors Avoidance: Apply the following dmap alter compile ceigsens $ alter 'paraml.*cases','' $ paraml casem//'dti'/1/5//s,n,meth $ 1-69947571 2008.0.0 (MD R3) Restarts - DBLOCATE /DBDICT job crashes with Fortran IO error A DBLOCATE job fails with the following system mesage on AIX system: "1525-002 The I/O statement on the internal file cannot be completed because the end of the internal file was reached. The program will stop." Other system may fail with differerent system message. Avoidance: re-run the job with more memory. 1-70170479 2008.0.0 (MD R3) Optimization/Design Sensitivity - DFREQ - Wrong sensitivities for multiple boundary conditions SOL 200 with DFREQ and multiple boundary condition static subcases may produce incorrect sensitivity coefficient. Avoidance: Inclue the following DMAP alters compile pslgdv $ alter 'dbview.*case0','' $ DBVIEW CASE0=casdsn WHERE (SEID=0 ) $ $ compile respsen $ alter 'dbview.*case0','' $ DBVIEW CASE0=casdsn WHERE (SEID=0 ) $ 1-70251161 2008.0.0 (MD R3) DMAP - MATMOD Option 12 has problems with Complex Matrices For complex matrices, if a column's first entry is of the form ( 0.0, i ) then this column is mistakenly identified by MATMOD Option 12 as a null column when all other entries in the column are 0.0. Avoidance: Use MATMOD option 34 to divide the complex matrix into two real matrices and search these two matrices for null columns. 1-70311681 2008.0.0 (MD R3) Static Analysis - MONTPNT1 - Results with multiple boundary conditions can be in error If a sol 101 run is made that includes monpnt1 entries and multiple boundary conditions, only the a-set from the final subcase is used in reducing the loads to the a-set with the result that the monpnt1 results can be in error for the preceding subcases. 1-70386662 2008.0.0 (MD R3) Optimization/Design Sensitivity - PLOAD1 or PLOAD4 - UFM 2268 or UFM 3001 in subdmap UPDATE A SOL 200 model with both SAERO and STATIC subcases and PLOAD1 or PLOAD4 entries, will fail with the messages: *** USER FATAL MESSAGE 2286 (SSGSLB) ELEMENT 869689 REFERENCED ON PLOAD1 2 NOT FOUND *** SYSTEM FATAL MESSAGE 3001 (SSG1) THE INPUT DATA BLOCK IN POSITION 5 DOES NOT EXIST. USER INFORMATION: THIS ERROR IS CAUSED BY ONE OF THE FOLLOWING: 1. THE DATA BLOCK IS UNSPECIFIED ON THE DMAP MODULE. 2. THE DATA BLOCK IS SPECIFIED ON THE DMAP MODULE AND ON THE SUBDMAP STATEMENT BUT NOT ON THE CORRESPONDING CALL STATEMENT. USER ACTION: 1. IF YOU ARE EXECUTING AN MSC.SOFTWARE SUPPLIED SOLUTION SEQUENCE AND NOT USING THE ALTER EXECUTIVE CONTROL STATEMENT, THEN CHECK FOR BULK DATA AND/OR CASE CONTROL INPUT ERRORS. IF NO ERRORS CAN BE FOUND, THEN CONTACT MSC.SOFTWARE CLIENT SUPPORT. 2. IF YOU ARE EXECUTING A DMAP PROGRAM NOT SUPPLIED BY MSC.SOFTWARE, THEN FOR DEBUGGING PURPOSES INSERT STATEMENT DIAGON(20) BEFORE THE MODULE SHOWN ABOVE. PROGRAMMER INFORMATION: THE FIST NUMBER IS 105 AND SUBROUTINE IS GOPEN Avoidance: Apply the following dmap avoidance. compile update alter ' ssg1 ','' paraml cases//'dti'/1/394//s,n,nsm $ TA1 ,,ECTS,EPTs,BGPDTS,SILS,ETT,CSTM0,,,,,/ EST0,,GEI,GPECT0,,,ditid,,nsmest/ LUSETS/-1/S,N,NOSIMP/1/S,N,NOGENL/0//////nsm $ SSG1 SLT,BGPDTS,CSTMS,medge,EST0,MPTS,ETT,EDT,,CASESX, DIT,,DEQATN,DEQIND,gpsnts,cstm0,scstm,geom4s,,,,,,/ pg,,,,/luset/1//'allids '///seid $$ 1-70446681 2008.0.0 (MD R3) SOL 600 - MATEP - Ryield field is ignore In SOL 600, the Ryield field of MATEP entry is ignored. The user can not select yield criteria. 1-70704881 2008.0.0 (MD R3) Splines - SPLINEX - UFM 2330 (GIMAPX) If SPLINEX is the only spline type in the input file, then it is possible that the SET1 entries from the independent element component be added to the SET1 record of the SPLINE datablock in MKSPLINE. These SET1 entries are already processed in the MAKECOMP module and their data are already in the SPLINE datablock on the AECOMP record. The GI module processing of SPLINEX does not need these SET1 entries. However, the GI module also performs error checking on the SET1 entries and assumes that the data are all grid IDs. Having element IDs in the SET1 list can lead to the following fatal error: *** USER FATAL MESSAGE 2330 (GIMAPX) USER INFORMATION: SET1 OR SPLINE3 ENTRY NO. 100001 REFERENCES EXTERNAL ID NO. 8000001 WHICH DOES NOT EXIST. Avoidance: The workaround is to create a dummy grid with the same ID as the 'missing' element. 1-70821471 2008.0.0 (MD R3) Optimization/Design Sensitivity - DRESP3 - SFM 3007 Illegal input to DOM10L with DRESP3 and P2CR set If an attempt is made to limit the print of the most critcal constraints and their associated responses using the P2CR parameter on the DOPTPRM entry and there are DRESP3 responses, the following error will be output: *** SYSTEM FATAL MESSAGE 3007 (DOM10) ILLEGAL INPUT TO SUBROUTINE DOM10L User Avoidance: Remove the P2Cr request or modify the P2 value so that responses are not reported. Suppressing the print as suggested above does not affect the optimzation task. 1-70999911 2008.0.0 (MD R3) Elements - NSML and NSML1 area calculations sensitive to element order - Wrong Answers NSML and NSML1 area calculations have been provend to be sensitive to element order and this can lead to wrong answers. 1-71075451 2008.0.0 (MD R3) Splines - SPRELAX can give wrong answers The SPRELAX entry is used to smooth a spline based on an adjacent spline. Wrong answers may result when this is used to perform relaxation based on multiple adjacent splines User Avoidance: Do not attempt to perform the relaxation based on multiple adjacent splines. 1-71200197 2008.0.0 (MD R3) SOL 600 - NSTATE specification ignored, Incorrect number of states in post file. SOL 600 in Version 2007.0.0 does not process the AUTO STEP entry correctly and all calculated states may be written to the post file even though NSTATE is specified. 1-72340381 Acoustics - MECHFIX - Wrong answers or UFM 6144 If masslass mechanisms are detected in a fluid-structure model and the constraint modes are printed with PARAM,MECHPRT,YES will be wrong and the SPCForces will be wrong in versions prior to MDR3. But all other results are unaffected by this error. In MDR3 and V2008, the run will fail with UFM 6144 for matrix MECHG. Avoidance: Insert the following DMAP alter: compile moders alter ' sdr1 '(,-1) dbstatus vafs//s,n,novafs $ if ( novafs>0 ) then $ MERGE mecha,,,,,VAFS/mecha1/1 $ equivx mecha1/mecha/-1 $ endif $ novafs>0 1-72344531 2008.0.0 (MD R3) Licensing - MSC_LICENSE_DISCRETE - Double Feature Checkout, UFM 3060 Setting the variable MSC_LICENSE_DISCRETE is necessary for correct license processing when two license servers, one with concurrent and one with Masterkey or EALU tokens, is used. An error exists in license processing when the MSC_LICENSE_DISCRETE variable is set that may result in a double checkout of the same feature. The error only occurs when the features on the first server are concurrent, i.e. not Masterkey or EALU, and the license check for the feature exists in multiple places in the code. Although all products make use of MSC_LICENSE_DISCRETE, only nastran tends to check for the same feature in multiple locations, so it is much more likely to encounter the problem. A second problem may result in a failed license checkout when a license feature does not exist at all on the first server but does on the second. Avoidance: If unsetting MSC_LICENSE_DISCRETE or pointing to the token based server first is not practical, then the system may be corrected by updating the lapi.so file in the delivery with a corrected version available at: http://www.mscsoftware.com/support/software_updates/lapi.cfm http://www.mscsoftware.com/support/software_updates/. Please note that due to quality 1-72344531, this problem is not resolved on SGI Irix in Verison 2008. 1-72589921 Licensing - Lapi, Discrete - SGI Irix will not work with Lapi 7.3.0 and above MD Nastran R3 on SGI Irix has reached the operating system limit of the dso table size and this prevents upgrading the Lapi FLEXlm licensing layer to 7.3 or above. Until this is resolved the DISCRETE double check out problem, quality 1-72344531, can not be addressed on this platform. 1-73000881 Documentation/QRG - NLADAPT SOL 400) - Incorrect field definition The QRG descriptionn of the NLAPAPT entry incorrectly references a DMAP field in addition to the correct field description which is IDAMP. This reference should be removed. 1-73052213 2008.0.1 SOL 700 - RFORCE requires grid (G) reference for rotation axis (requirement is not documented) RFORCE (angular velocity/acceleration load) has an optional grid (G) entry to define the center of rotation. In SOL 700 if the grid (G) entry is not entered on the RFORCE a fatal message occurs in the .dytr_prep.out file. This grid option is not required in standard MSC/MD Nastran solution sequences but is required in SOL 700; however this requirement is not documented. Avoidance: Create a grid at the center of rotation and reference this grid on the RFORCE entry. 1-73626721 2008.0.1 Parallel Processing - PARALLEL, SYSTEM(107) - Vendor implementation of subroutine TFORK is missing Shared memory parallel processing does not work on Windows 32 MD R3 Nastran (Version 2008.0.0). If attempted (setting parallel (system(107) keyword on the nastran or submittal line) the job will fail with the message: system fatal: Vendor implementation of subroutine TFORK is missing Avoidance: The MD R3 delivery will be corrected with Version 2008.0.1. 1-73816550 Documentation/QRG - PBRSECT - Composites are not supported The QRG documentation for the PBRSECT entry incorrectly indicates that it supports composites and can reference Core and/or Layer. This capability is only supported with PBMSECT and if attempted with PBRSECT will lead to an error in IFP indicating illegal format. 1-73873851 2008.0.1 Miscellaneous - f04 file missing delta cpu information The MD R3 Verson 2008.0.0 f04 file (Execution Summary) on Windows 32 indicates 0.0 for DEL I/0, CPU and DEL CPU. The MD R3 delivery will be corrected in 2008.0.1. 1-74141249 Licensing - MD R3 Windows 32, Linux x8664, prog= - UFM 3060 (SCA Service Loader) The MD R3 (2008.0.0) Windows 32 and Linux x8664 SCA enabled builds can not be used with vertical application licensing such as SimXpert or SimDesigner unless a full MD Nastran License also exists. If attempted the run will fail with the message: *** USER FATAL MESSAGE 3060 (SCA Service Loader) - AUTHORIZATION ERROR *** UNABLE TO OBTAIN LICENSES FOR SCA.MDSolver.System.PrintServer PLEASE REFER TO THE LOG FILE FOR ADDITIONAL DETAILS 1-74205731 Performance - Windows Vista 64, Server 2008 Vista 64 and Server 2008 64 have removed a system tuning parameter that controls memory usage. Without this option, when memory usage approaches 100 percent, I/O intensive applications, especially those such as Nastran that generate very large data files overwhelm the system cache, to the point that there is no physical memory left and the machine will appear to hang. Avoidance: The work-around is to tell the various I/O intensive applications to bypass the use of system caching for their very large files. This is done by specifying the FILE_FLAG_NO_BUFFERING option in the FlagsAndAttributes argument specified for CreateFile when the file is opened. For MSC.Nastran and MD Nastran, the files for which this bypass is necessary are the SCRATCH and SCR300 files. In nastran this is accomplished by specifying RAW=YES as a SYS keyword for the appropriate files. Note that RAW=YES also requires that BUFFIO be requested instead of the default of MAPIO=YES.) Unfortunately, through a documentation oversight, the use of RAW=YES for Windows is not documented in the IOG. This will be corrected. These SYS options may also be specified by using the SYSFIELD command line or RC File keyword option. Example of these specification on the nastran command is nastran jid MEM=4GB SYSFIELD=SCR*(RAW=YES,BUFFIO=YES,WNUM=32,WSIZE=16MB) 1-74563841 SOL 400 - Fails in in SDR2 after divergence When a divergent condition is detected, SOL 400 always packs the final iteration resuts for output and post-processing before leaving the NLSOLV module. However, in some cases, such as when a job fails because of material divergence, it may leave NLSOLV without saving all required data. In this case, the job will fail later in the SDR2 module. . 1-74623771 MDR3 - BCTABLE - Segmentation Error in SOL700 ... SOL 700 may fail with a segmentation error in dytran-lsdyna or may hang on other systems. Avoidance: Modifying the BCTABLE from 3 pairs instead of 6 allowed the job to run. 1-74935310 2008.0.0 (MD R3) Optimization/Design Sensitivity - SFM 3007 (DOM9), ILLEGAL INPUT TO SUBROUTINE ADS9P4D A model with a large number of design variables and constraints may fail with the message: *** SYSTEM FATAL MESSAGE 3007 (DOM9) ILLEGAL INPUT TO SUBROUTINE ADS9P4D Avoidance: Use the DOT optimizer instead. This can be selected by specifying OPTCOD DOT on the DOPTPRM entry. 1-74941232 2008.0.1 SOL 700 - DYPARAM,VELMAX,,NO/YES does not work as documented In SOL 700 the options YES/NO on the DYPARAM entry are not working in MD R3. Only the value for VELMAX will work properly. 1-75553921 SOL 700 - DYPARAM,LSDYNA,BIBARY,D3PLOT is not effective. DYPARAM,LSDYNA,BIBARY,D3PLOT to control output time interval isn't effective in SOL 700. Avoidance: Use PARAM,DYDTOUT.