This section explains the contents of the substructure matrices file (jobname.sub).
See The Standard Header for Mechanical APDL Binary Files for a description of this set. File number (Item 1) is 8.
*comdeck,fdsub
c
c *** Copyright ANSYS. All Rights Reserved.
c *** ansys, inc
c ********** description of substructure matrix file **********
character*8 SUBNM
parameter (SUBNM='sub ')
integer SUBHDLEN
parameter (SUBHDLEN=80)
LONGINT subfpL, lenSubL
integer subbk, subut
common /fdsub/ subfpL, lenSubL, subbk, subut
c write: matout
c read:
c ********** common variable descriptions ***********
co subfpL file position on file sub
co subbk block number for file sub
co subut file unit for file sub
co lenSubL length of sub file (saved for slvdta.F)
c See fddesc for documentation of how binary files are stored.
c
c ********** file format **********
c recid tells the identifier for this record. Not all records will have
c identifiers -- they are only indicated for those records whose
c record pointers are stored in the second file header.
c type tells what kind of information is stored in this record:
c i - integer
c dp - double precision
c cmp - complex
c nrec tells how many records of this description are found here
c lrec tells how long the records are (how many items are stored)
c recid type nrec lrec contents
c --- i 1 100 standard ANSYS file header (see binhed8
c for details of header contents)
c HED i 1 80 .SUB FILE HEADER (FULL MATRICES)
c
c 8, nmrow, nmatrx, nedge, numdof,
c maxn, wfmax, lenbac, nnod, kunsym, 10
c kstf, kmass, kdamp, kss, nvect,
c nWorkL, lenU1, sesort, lenlst,ptrLodL, 20
c ntrans, ptrMtx, ptrXFM, ptrHED, name1,
c name2, ptrCG, 0, name3, name4, 30
c ptrDOF, ptrDST, ptrBAC, ptrTIT, ptrNOD,
c ptrXYZ, ptrEDG, ptrGDF, thsubs, ptrPOS, 40
c ptrORG, stfmax, ptrLodH, nmodes, keydim,
c cmsMethod, name5, name6, name7, name8, 50
c nvnodes,ptrCTXM, nWorkH, 0,ptrTVAL,
c gyroDamp,kstress,nStartVN,ptrEndL,ptrEndH, 60
c ptrimsSEdat,ptrdmsSEdat,units,ptrmsSEmap, 0,
c 0, 0, 0, 0, 0, 70
c 0, 0, 0, 0, 0,
c 0, 0, 0, 0, 0 80
c HED i 1 80 .SUB FILE HEADER (SPARSE MATRICES)
c
c 9, nEqn, nmatrx, ndege, numdof,
c maxn, wfmax, lenbac, nnod, kunsym, 10
c kstf, kmass, kdamp, , nvect,
c nTermL, , , lenlst,ptrLodL, 20
c ntrans,ptrMtxL, ptrXFM, ptrHED, name1,
c name2, ptrCG, , name3, name4, 30
c ptrDOF, , ptrBAC, ptrTIT, ptrNOD,
c ptrXYZ, ptrEdg, ptrGDF, thsubs, , 40
c , stfmax,ptrLodH, , keydim,
c , name5, name6, name7, name8, 50
c ,ptrCTXM, nTermH,ptrMtxH,ptrColL,
c ptrColH,ptrCofL,ptrCofH,ptrEndL,ptrEndH, 60
c ptrimsSEdat,ptrdmsSEdat,units,ptrmsSEmap, 0,
c , , , , ,
c , , , , , 70
c , , , , ,
c , , , , 80
c each item in header is described below:
c fun08 - unit number (full sub file is 8)
c (sparse substructure file is 9)
c nmrow - number of rows in matrices (also
c number of dofs in substructure)
c nmatrx - number of matrices on file
c nedge - number of edges for outline
c numdof - number of dofs per node
c maxn - maximum node number of complete
c model presently in database
c wfmax - maximum wavefront of substruct.
c during generation pass
c lenbac - number of nodes defining
c substructure during the
c generation pass
c nnod - number of unique nodes in the
c substructure having DOFs, and
c which define this substructure
c during the use pass. Also, the
c number of nodes having master
c DOFs.
c kunsym - unsymmetric matrix key
c = 0 - symmetric
c = 1 - unsymmetric
c kstf - stiffness matrix present key
c = 0 - matrix is not on file
c = 1 - matrix is on file
c kmass - mass matrix present key
c = 0 - matrix is not on file
c = 1 - matrix is on file
c =-1 - Lumped mass vestor (Sparse only)
c kdamp - damping matrix present key
c = 0 - matrix is not on file
c = 1 - matrix is on file
c kss - stress stiffening matrx present
c = 0 - matrix is not on file
c = 1 - matrix is on file
c nvect - number of load vectors
c (at least 1 is required)
c nWorkL,H - BCS workspace length (only for
c bacsub)
c nTermL,H - Number of terms in sparse matrix
c lenU1 - length of intermediate transformation
c vector
c sesort - DOF set sort key
c = 0 - numbers are not sorted
c = 1 - numbers are sorted in
c ascending order
c lenlst - maximum length of DOF set for
c this substructure (maxn*numdof)
c ptrLod - pointer to the start of the load
c vectors (see also ptrLodh)
c ntrans - transformed key
c = 0 - substructure has not been
c transformed
c > 0 - substructure copied
c from another substructure,
c via either SESSYM or SETRAN
c ptrMtxL,H - pointer to the start of the
c substructure matrices (iDiagL for
c sparse matrices)
c ptrXFM - pointer to the substructure
c transformations
c ptrHED - pointer to the SUB file header
c name1 - first four characters of the
c substructure file name, in
c integer form
c name2 - second four characters of the
c substructure file name, in
c integer form
c name3 - third four characters of the
c substructure file name, in
c integer form
c name4 - fourth four characters of the
c substructure file name, in
c integer form
c ptrDOF - pointer to the DOF/node list
c ptrDST - pointer to the local DOF set
c ptrBAC - pointer to the nodes comprising
c the substructure
c ptrTIT - pointer to the title
c ptrNOD - pointer to the unique nodes
c defining the substructure
c ptrXYZ - pointer to the coordinates of the
c unique nodes
c ptrEDG - pointer to the substructure edges
c ptrGDF - pointer to the global DOF set
c ptrCG - pointer to the element mass information
c thsubs - element type key
c = 0 - structural
c = 1 - 1st order non-structural
c (generally from thermal)
c = 2 - 2nd order non-structural
c ptrPOS - pointer to the sorted substructure
c DOF set to the original
c ptrORG - pointer to the DOF set of the model
c during the generation pass
c stfmax - maximum diagonal stiffness term
c (packed into an integer)
c ptrLodh- High 32 bits of 64 bit pointer
c nmodes - number of modes used to generate
c CMS s.e.
c keydim - dimensionality key
c = 1 - axisymmetric
c = 2 - 2-D
c = 3 - 3-D
c cmsMethod - component mode synthesis method
c name5 - fifth four characters of the
c substructure file name, in integer
c form
c name6 - sixth four characters of the
c substructure file name, in integer
c form
c name7 - seventh four characters of the
c substructure file name, in integer
c form
c name8 - eighth four characters of the
c substructure file name, in integer
c form
c nvnodes - number of virtual nodes that contain
c the modal coordinates
c gyroDamp - gyroscopic damping matrix key
c = 0 - damping matrix is
c not skew-symmetric if present
c = 1 - damping matrix is
c skew-symmetric, due to
c gyroscopic effect
c kStress - = 1 if modal element results are
c on the .cms file
c nStartVN - starting number of the virtual nodes
c if defined with CMSOPT command
c ptrCTXM - coordinate transformation
c ptrColL,H - pointer to the iCol sparse matrix
c array
c ptrCofL,H - pointer to the of the
c sparse matrix Sk(1:nTerm),
c Sm(1:nTermL),Sc(1:nTermL),
c Ss(1:nTermL) Each matrix is a
c single large record
c ptrEndL,H - Next location after end of file
c ptrimsSEdat - pointer to integer record for stage data
c ptrdmsSEdat - pointer to double precision records for stage data
c ptrmsSEmap - pointer to map of modes for stage data
c units - unit system used
c =-1 - no /UNITS specification
c = 0 - user defined units
c = 1 - SI
c = 2 - CSG
c = 3 - U.S. Customary, using feet
c = 4 - U.S. Customary, using inches
c = 5 - MKS
c = 6 - MPA
c = 7 - uMKS
c note: name1/2/3/4/5/6/7/8 are the
c inexc4 representation of the
c 32 character filename.
c name1/2/5/6/7/8 will be "0"
c for pre rev 5.2 files - cwa
c Note: If ntrans > 0, records from position ptrDOF to ptrGDF will be
c identical to the data for the copied substructure.
c XFM dp 1 125 Substructure transformations (5*25 double
c precisions). This record has meaning only
c if ntrans > 0. You can define up to five
c levels of transformations, with 25 variables
c in each level. Up to the first seven
c variables are used as follows:
c
c If the substructure was transferred (via the
c SETRAN command):
c 1st variable - 1.0
c 2nd variable - nodal increment
c 3rd variable - reference number of
c coordinate system where substructure will
c be transferred
c 4th variable - reference number of
c coordinate system where substructure is
c presently defined
c 5th variable - x coordinate increment
c 6th variable - y coordinate increment
c 7th variable - z coordinate increment
c If the substructure used symmetry (via the
c SESYMM command):
c 1st variable - 2.0
c 2nd variable - nodal increment
c 3rd variable - number of coordinate
c component to be used in operation
c = 1 - x coordinate
c = 2 - y coordinate
c = 3 - z coordinate
c 4th variable - reference number of
c coordinate system to be used for symmetry
c operation
c CTXM dp 1 250 Substructure transformations
c DOF i 1 numdof Degrees of freedom per node (Global)
c (curdof(i),i=1,numdof)
c DOF reference numbers are:
c UX = 1, UY = 2, UZ = 3, ROTX= 4, ROTY= 5, ROTZ= 6, AX = 7, AY = 8
c AZ = 9, VX =10, VY =11, VZ =12, GFV1=13, GFV2=14, GFV3=15, WARP=16
c CONC=17, HDSP=18, PRES=19, TEMP=20, VOLT=21, MAG =22, ENKE=23, ENDS=24
c EMF =25, CURR=26, SP01=27, SP02=28, SP03=29, SP04=30, SP05=31, SP06=32
c DST i 1 nmrow This record contains degrees of freedom for
c this substructure of the unique nodes, as
c used with this substructure, in ascending
c order. This index is calculated as
c (N-1)*numdof+DOF, where N is the node number
c and DOF is the DOF reference number given
c above
c (lsort(i),i=1,nmrow)
c POS i 1 nmrow This record stores the positions of the
c local DOF set in relation to the generated
c DOF set. (lposit(i),i=1,nmrow)
c ORG i 1 nmrow DOF set of the model as defined during the
c generation pass. This index is calculated as
c (N-1)*NUMDOF+DOF, where N is the position
c number of the node in the nodal equivalence
c table and DOF is the DOF reference number
c given above
c (lorig(i),i=1,nmrow)
c BAC i 1 lenbac This group describes nodes that defined the
c substructure during the generation pass of
c the analysis. Nodal data is stored in arrays
c equal to the number of used or referenced
c nodes. This table equates the number used
c for storage to the actual node number.
c (Back(i),i=1,lenbac)
c TIT i 1 20 Substructure title (converted to integers -
c see inexc4)
c NOD i 1 nnod This record describes unique nodes defining
c the substructure for the use pass of the
c analysis. These are also the nodes having
c master degrees of freedom.
c (node(i),i=1,nnod)
c XYZ dp nnod 6 This record describes the coordinates of a
c unique node, in the order X, Y, Z, THXY,
c THYZ, and THZX. Nodal order corresponds to
c that of the node list given above
c (xyzang(j,i),j=1,6)
c EDG dp nedge 6 This record contains beginning and ending
c locations (X1,Y1,Z1,X2,Y2,Z2 coordinates) of
c a straight line comprising an edge of the
c substructure.
c GDF LONG 1 nmrow This record describes global degrees of
c freedom of the unique nodes in ascending
c order, as used during the analysis use pass.
c This index is calculated as (N-1)*32+DOF,
c where N is the node number and DOF is the
c DOF reference number given above
c (l(i),i=1,nmrow) (sorted)
c (Made LONGINT in rev 14.0)
c CG dp 1 49 Are stored consecutively:
c - Totmass: total mass
c - CGx,CGy,CGz: center of mass (X,Y,Z) with lumped calculation
c - mm11,mm22,mm33,mm12,mm23,mm13:
c components of upper triangular inertia matrix (rotational mass) about origin
c - MASSTR: translational mass matrix (3*3, row by row) about origin
c - MASSROSP: inertia matrix (3*3) about point defined with SPOINT command
c - MASSTRRO: coupled translational/rotational mass matrix (3*3) about origin
c - CG'x,CG'y,CG'z: center of mass (X,Y,Z) with precise calculation
c - MASSROCG': inertia matrix (3*3) about center of mass with precise calculation
c TVAL dp 1 31 current time value corresponds to each load step
c (substructuring analysis only).
c lrec=31 is the default,
c but lrec=VAL if /CONFIG,NUMSUBLV,VAL has been issued
c IMSSE i 1 17 integer record for stage data (see multiStage.inc):
c msNsector,msHindex,msNumoff,
c msCECYC(i)(i=1,2),msNodElm(i)(i=1,12)
c DMSSE dp 1 6 double precision record for stage data (see multiStage.inc):
c msBox(i)(i=1,6)
c MSSEMAP i 1 nmodes integer record for stage map of paired modes,
c each mode is associated with base or duplicate sector:
c mssemap(i)= 1 ith mode is associated to the base sector
c mssemap(i)=-1 ith mode is associated to the dupl sector
c The substructure matrices are written at this position in the file. One row
c of each matrix is written to the file at a time. i.e. the first row of each
c matrix is written, then the second row of each matrix, etc. this pattern
c continues until all nmrow rows of each matrix have been written to the file.
c MAT dp 1 nmrow Row of the stiffness matrix, if nmatrx > 0.
c (ak(i,j),i=1,nmrow)
c --- dp 1 nmrow Row of the mass matrix, if nmatrx > 1.
c (am(i,j),i=1,nmrow)
c --- dp 1 nmrow Row of the damping matrix, if nmatrx > 2.
c (ac(i,j),i=1,nmrow)
c --- dp 1 nmrow Row of the stress stiffening matrix, if
c nmatrx > 3.
c (gs(i,j),i=1,nmrow)
c LOD dp nvect nmrow This record contains the load vectors.
c (f(i),i=1,nmrow)