223031012
/
DAMASK_EICMD
mirror of https://github.com/achalhp/DAMASK_EICMD.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
DAMASK_EICMD/code/mesh.f90

3492 lines
124 KiB
Fortran
Raw Normal View History

added copyright text to all f90 (free) format files
2011-04-04 19:39:54 +05:30
! Copyright 2011 Max-Planck-Institut f<>r Eisenforschung GmbH
!
! This file is part of DAMASK,
! the D<>sseldorf Advanced MAterial Simulation Kit.
!
! DAMASK is free software: you can redistribute it and/or modify
! it under the terms of the GNU General Public License as published by
! the Free Software Foundation, either version 3 of the License, or
! (at your option) any later version.
!
! DAMASK is distributed in the hope that it will be useful,
! but WITHOUT ANY WARRANTY; without even the implied warranty of
! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
! GNU General Public License for more details.
!
! You should have received a copy of the GNU General Public License
! along with DAMASK. If not, see <http://www.gnu.org/licenses/>.
!
!##############################################################
added version information to all files do NOT edit text like this: $Id: constitutive_phenopowerlaw.f90 406 2009-08-31 14:13:10Z MPIE\f.roters $
2009-08-31 20:39:15 +05:30
!* $Id$
let's see whether this is the thing to do..!?!?
2007-03-20 19:25:22 +05:30
!##############################################################
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
MODULE mesh
let's see whether this is the thing to do..!?!?
2007-03-20 19:25:22 +05:30
!##############################################################
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
use prec, only: pReal,pInt
implicit none
! ---------------------------
started envIP building
2007-03-22 20:18:58 +05:30
! _Nelems : total number of elements in mesh
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
! _NcpElems : total number of CP elements in mesh
started envIP building
2007-03-22 20:18:58 +05:30
! _Nnodes : total number of nodes in mesh
fixed buggy build_ipNeighborhood and faceMatch (that sounds so easy...)
2007-04-04 14:17:34 +05:30
! _maxNnodes : max number of nodes in any CP element
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
! _maxNips : max number of IPs in any CP element
! _maxNipNeighbors : max number of IP neighbors in any CP element
fixed buggy build_ipNeighborhood and faceMatch (that sounds so easy...)
2007-04-04 14:17:34 +05:30
! _maxNsharedElems : max number of CP elements sharing a node
started envIP building
2007-03-22 20:18:58 +05:30
!
changed element type mapping to be performed right after input file parsing. mesh_element(2,e) now contains INTERNAL type index -- no subsequent mapping necessary anymore!
2008-03-25 18:22:27 +05:30
! _element : FEid, type(internal representation), material, texture, node indices
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
! _node0 : x,y,z coordinates (initially!)
! _node : x,y,z coordinates (after deformation!)
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
! _sharedElem : entryCount and list of elements containing node
started envIP building
2007-03-22 20:18:58 +05:30
!
generalized _FEasCP (string what,int ID) to use targeted arrays available what options are 'elem' and 'node'
2007-03-27 18:23:31 +05:30
! _mapFEtoCPelem : [sorted FEid, corresponding CPid]
! _mapFEtoCPnode : [sorted FEid, corresponding CPid]
started envIP building
2007-03-22 20:18:58 +05:30
!
! MISSING: these definitions should actually reside in the
! FE-solver specific part (different for MARC/ABAQUS)..!
! Hence, I suggest to prefix with "FE_"
!
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
! _Nnodes : # nodes in a specific type of element (how we use it)
! _NoriginalNodes : # nodes in a specific type of element (how it is originally defined by marc)
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
! _Nips : # IPs in a specific type of element
! _NipNeighbors : # IP neighbors in a specific type of element
! _ipNeighbor : +x,-x,+y,-y,+z,-z list of intra-element IPs and
started envIP building
2007-03-22 20:18:58 +05:30
! (negative) neighbor faces per own IP in a specific type of element
added full description of element type 2
2009-01-16 23:06:37 +05:30
! _NfaceNodes : # nodes per face in a specific type of element
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
! _nodeOnFace : list of node indices on each face of a specific type of element
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
! _maxNnodesAtIP : max number of (equivalent) nodes attached to an IP
extended linking between nodes and IPs to account for more complex situations in quadratic shape function elements.
2009-03-30 20:20:19 +05:30
! _nodesAtIP : map IP index to two node indices in a specific type of element
added full description of element type 2
2009-01-16 23:06:37 +05:30
! _NsubNodes : # subnodes required to fully define all IP volumes
started envIP building
2007-03-22 20:18:58 +05:30
! order is +x,-x,+y,-y,+z,-z but meaning strongly depends on Elemtype
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
! ---------------------------
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
integer(pInt) mesh_Nelems, mesh_NcpElems, mesh_NelemSets, mesh_maxNelemInSet
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
integer(pInt) mesh_Nmaterials
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
integer(pInt) mesh_Nnodes, mesh_maxNnodes, mesh_maxNips, mesh_maxNipNeighbors, mesh_maxNsharedElems, mesh_maxNsubNodes
added output of mesh statistics at end of input file parsing
2007-10-24 14:30:42 +05:30
integer(pInt), dimension(2) :: mesh_maxValStateVar = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=64), dimension(:), allocatable :: mesh_nameElemSet, & ! names of elementSet
mesh_nameMaterial, & ! names of material in solid section
mesh_mapMaterial ! name of elementSet for material
integer(pInt), dimension(:,:), allocatable :: mesh_mapElemSet ! list of elements in elementSet
now with first draft of nonlocal constitutive law debugging memory leak closed debugging counters corrected center of gravity stored in mesh state updated is now split into a collecting loop and an execution updateState and updateTemperature fill sequentially separate logicals and evaluate afterwards to converged added 3x3 transposition function, norm for 3x1 matrix and 33x3 matrix multiplication in math non-converged crystallite triggers materialpoint cutback (used to respond elastically) non-converged materialpoint raises terminal illness which in turn renders whole FE increment useless by means of odd stress/stiffness and thus waits for FE cutback
2009-08-11 22:01:57 +05:30
integer(pInt), dimension(:,:), allocatable, target :: mesh_mapFEtoCPelem, mesh_mapFEtoCPnode
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
integer(pInt), dimension(:,:), allocatable :: mesh_element, & ! FEid, type(internal representation), material, texture, node indices
mesh_sharedElem, & ! entryCount and list of elements containing node
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
mesh_nodeTwins ! node twins are surface nodes that lie exactly on opposite sides of the mesh (surfaces nodes with equal coordinate values in two dimensions)
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
integer(pInt), dimension(:,:,:,:), allocatable :: mesh_ipNeighborhood ! 6 or less neighboring IPs as [element_num, IP_index]
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
now with first draft of nonlocal constitutive law debugging memory leak closed debugging counters corrected center of gravity stored in mesh state updated is now split into a collecting loop and an execution updateState and updateTemperature fill sequentially separate logicals and evaluate afterwards to converged added 3x3 transposition function, norm for 3x1 matrix and 33x3 matrix multiplication in math non-converged crystallite triggers materialpoint cutback (used to respond elastically) non-converged materialpoint raises terminal illness which in turn renders whole FE increment useless by means of odd stress/stiffness and thus waits for FE cutback
2009-08-11 22:01:57 +05:30
real(pReal), dimension(:,:,:), allocatable :: mesh_subNodeCoord ! coordinates of subnodes per element
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
real(pReal), dimension(:,:), allocatable :: mesh_node0, & ! node coordinates (initially!)
mesh_node, & ! node coordinates (after deformation! ONLY FOR MARC!!!)
mesh_ipVolume ! volume associated with IP (initially!)
real(pReal), dimension(:,:,:), allocatable :: mesh_ipArea, & ! area of interface to neighboring IP (initially!)
mesh_ipCenterOfGravity ! center of gravity of IP (after deformation!)
real(pReal), dimension(:,:,:,:), allocatable :: mesh_ipAreaNormal ! area normal of interface to neighboring IP (initially!)
integer(pInt), dimension(:,:,:), allocatable :: FE_nodesAtIP ! map IP index to two node indices in a specific type of element
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
integer(pInt), dimension(:,:,:), allocatable :: FE_ipNeighbor
integer(pInt), dimension(:,:,:), allocatable :: FE_subNodeParent
integer(pInt), dimension(:,:,:,:), allocatable :: FE_subNodeOnIPFace
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
logical :: noPart ! for cases where the ABAQUS input file does not use part/assembly information
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
logical, dimension(3) :: mesh_periodicSurface ! flag indicating periodic outer surfaces (used for fluxes)
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
integer(pInt) :: hypoelasticTableStyle
integer(pInt) :: initialcondTableStyle
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: FE_Nelemtypes = 10_pInt
integer(pInt), parameter :: FE_maxNnodes = 8_pInt
integer(pInt), parameter :: FE_maxNsubNodes = 56_pInt
integer(pInt), parameter :: FE_maxNips = 27_pInt
integer(pInt), parameter :: FE_maxNipNeighbors = 6_pInt
integer(pInt), parameter :: FE_maxmaxNnodesAtIP = 8_pInt
integer(pInt), parameter :: FE_NipFaceNodes = 4_pInt
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter :: FE_Nnodes = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
int([8, & ! element 7
4, & ! element 134
4, & ! element 11
4, & ! element 27
4, & ! element 157
6, & ! element 136
8, & ! element 21
8, & ! element 117
8, & ! element 57 (c3d20r == c3d8 --> copy of 7)
3 & ! element 155, 125, 128
],pInt)
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter :: FE_NoriginalNodes = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
int([8, & ! element 7
4, & ! element 134
4, & ! element 11
8, & ! element 27
4, & ! element 157
6, & ! element 136
20,& ! element 21
8, & ! element 117
20,& ! element 57 (c3d20r == c3d8 --> copy of 7)
6 & ! element 155, 125, 128
],pInt)
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter :: FE_Nips = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
int([8, & ! element 7
1, & ! element 134
4, & ! element 11
9, & ! element 27
4, & ! element 157
6, & ! element 136
27,& ! element 21
1, & ! element 117
8, & ! element 57 (c3d20r == c3d8 --> copy of 7)
3 & ! element 155, 125, 128
],pInt)
added full description of element type 2
2009-01-16 23:06:37 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter :: FE_NipNeighbors = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
int([6, & ! element 7
4, & ! element 134
4, & ! element 11
4, & ! element 27
6, & ! element 157
6, & ! element 136
6, & ! element 21
6, & ! element 117
6, & ! element 57 (c3d20r == c3d8 --> copy of 7)
4 & ! element 155, 125, 128
],pInt)
added full description of element type 2
2009-01-16 23:06:37 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter :: FE_NsubNodes = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
int([19,& ! element 7
0, & ! element 134
5, & ! element 11
12,& ! element 27
0, & ! element 157
15,& ! element 136
56,& ! element 21
0, & ! element 117
19,& ! element 57 (c3d20r == c3d8 --> copy of 7)
4 & ! element 155, 125, 128
],pInt)
added functions -- mesh_build_subNodeCoords -- mesh_build_ipVolimes -- mesh_build_ipAreas new globals mesh_ipVolume and mesh_ipArea / mesh_ipAreaNormal contain the volume per IP, area of the interface to aach neighboring IP and the outward pointing unit normal vector of these interfaces
2009-01-16 20:59:57 +05:30
integer(pInt), dimension(FE_maxNipNeighbors,FE_Nelemtypes), parameter :: FE_NfaceNodes = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
4,4,4,4,4,4, & ! element 7
added element types 11 (4node 2D linear full) and 27 (8node 2D quadratic full)
2007-10-12 19:18:29 +05:30
3,3,3,3,0,0, & ! element 134
2,2,2,2,0,0, & ! element 11
Element type 21 (20-node hexagon with triquadric interpolation functions) is now included. 'mesh_get_nodeElemDimensions', mesh_build_elements', 'mesh_build_sharedElems' now allow for more than 14 nodes per elements which are then written in more than one line in the 'connectivity' field of the input file
2009-04-02 18:30:51 +05:30
2,2,2,2,0,0, & ! element 27
added two more element types (157 and 136)
2008-06-17 14:41:54 +05:30
3,3,3,3,0,0, & ! element 157
Element type 21 (20-node hexagon with triquadric interpolation functions) is now included. 'mesh_get_nodeElemDimensions', mesh_build_elements', 'mesh_build_sharedElems' now allow for more than 14 nodes per elements which are then written in more than one line in the 'connectivity' field of the input file
2009-04-02 18:30:51 +05:30
3,4,4,4,3,0, & ! element 136
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
4,4,4,4,4,4, & ! element 21
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
4,4,4,4,4,4, & ! element 117
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
4,4,4,4,4,4, & ! element 57 (c3d20r == c3d8 --> copy of 7)
2,2,2,0,0,0 & ! element 155, 125, 128
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNipNeighbors,FE_Nelemtypes/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter :: FE_maxNnodesAtIP = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
int([1, & ! element 7
4, & ! element 134
1, & ! element 11
2, & ! element 27
1, & ! element 157
1, & ! element 136
4, & ! element 21
8, & ! element 117
1, & ! element 57 (c3d20r == c3d8 --> copy of 7)
1 & ! element 155, 125, 128
],pInt)
added functions -- mesh_build_subNodeCoords -- mesh_build_ipVolimes -- mesh_build_ipAreas new globals mesh_ipVolume and mesh_ipArea / mesh_ipAreaNormal contain the volume per IP, area of the interface to aach neighboring IP and the outward pointing unit normal vector of these interfaces
2009-01-16 20:59:57 +05:30
integer(pInt), dimension(FE_NipFaceNodes,FE_maxNipNeighbors,FE_Nelemtypes), parameter :: FE_nodeOnFace = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
1,2,3,4 , & ! element 7
2,1,5,6 , &
3,2,6,7 , &
added two more element types (157 and 136)
2008-06-17 14:41:54 +05:30
4,3,7,8 , &
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
4,1,5,8 , &
8,7,6,5 , &
1,2,3,0 , & ! element 134
1,4,2,0 , &
2,3,4,0 , &
1,3,4,0 , &
0,0,0,0 , &
added element types 11 (4node 2D linear full) and 27 (8node 2D quadratic full)
2007-10-12 19:18:29 +05:30
0,0,0,0 , &
1,2,0,0 , & ! element 11
2,3,0,0 , &
3,4,0,0 , &
4,1,0,0 , &
0,0,0,0 , &
0,0,0,0 , &
Element type 21 (20-node hexagon with triquadric interpolation functions) is now included. 'mesh_get_nodeElemDimensions', mesh_build_elements', 'mesh_build_sharedElems' now allow for more than 14 nodes per elements which are then written in more than one line in the 'connectivity' field of the input file
2009-04-02 18:30:51 +05:30
1,2,0,0 , & ! element 27
2,3,0,0 , &
3,4,0,0 , &
4,1,0,0 , &
added element types 11 (4node 2D linear full) and 27 (8node 2D quadratic full)
2007-10-12 19:18:29 +05:30
0,0,0,0 , &
added two more element types (157 and 136)
2008-06-17 14:41:54 +05:30
0,0,0,0 , &
1,2,3,0 , & ! element 157
1,4,2,0 , &
2,3,4,0 , &
1,3,4,0 , &
0,0,0,0 , &
0,0,0,0 , &
1,2,3,0 , & ! element 136
1,4,5,2 , &
2,5,6,3 , &
1,3,6,4 , &
4,6,5,0 , &
Element type 21 (20-node hexagon with triquadric interpolation functions) is now included. 'mesh_get_nodeElemDimensions', mesh_build_elements', 'mesh_build_sharedElems' now allow for more than 14 nodes per elements which are then written in more than one line in the 'connectivity' field of the input file
2009-04-02 18:30:51 +05:30
0,0,0,0 , &
1,2,3,4 , & ! element 21
2,1,5,6 , &
3,2,6,7 , &
4,3,7,8 , &
4,1,5,8 , &
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
8,7,6,5 , &
1,2,3,4 , & ! element 117
2,1,5,6 , &
3,2,6,7 , &
4,3,7,8 , &
4,1,5,8 , &
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
8,7,6,5 , &
1,2,3,4 , & ! element 57 (c3d20r == c3d8 --> copy of 7)
2,1,5,6 , &
3,2,6,7 , &
4,3,7,8 , &
4,1,5,8 , &
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
8,7,6,5 , &
1,2,0,0 , & ! element 155,125,128
2,3,0,0 , &
3,1,0,0 , &
0,0,0,0 , &
0,0,0,0 , &
0,0,0,0 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_maxNipNeighbors,FE_Nelemtypes/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
CONTAINS
! ---------------------------
! subroutine mesh_init()
! function mesh_FEtoCPelement(FEid)
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
! function mesh_build_ipNeighorhood()
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
! ---------------------------
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
!***********************************************************
! initialization
!***********************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
subroutine mesh_init (ip,element)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
renamed interface from "mpie_interface" to "DAMASK_interface"
2011-05-11 22:31:03 +05:30
use DAMASK_interface
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
use, intrinsic :: iso_fortran_env ! to get compiler_version and compiler_options (at least for gfortran 4.6 at the moment)
added function calls in init()
2007-04-10 16:52:53 +05:30
use prec, only: pInt
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
use IO, only: IO_error,IO_open_InputFile,IO_abaqus_hasNoPart
reworked restarting for compatibility with abaqus (not yet fully working) added new orientation feature for direct simulation: component type (random) asigns random orientation to an entire grain
2011-05-28 15:14:43 +05:30
use FEsolving, only: parallelExecution, FEsolving_execElem, FEsolving_execIP, calcMode, lastMode, FEmodelGeometry
overhaul to allow for multiphase/homogenization setup.
2009-03-04 17:18:54 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: fileUnit = 222_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
integer(pInt) e,element,ip
numerics.config is now optional; if this file does not exist in the working directory, standard values from numerics.f90 are used; either way the output file tells which values were taken. added some output to the init subroutines of debug.f90, FEsolving.f90, lattice.f90, material.f90 and mesh.f90.
2009-06-18 19:58:02 +05:30
openmp parallelization working again (at least for j2 and nonlocal constitutive model). In order to keep it like that, please follow these simple rules: DON'T use implicit array subscripts: example: real, dimension(3,3) :: A,B A(:,2) = B(:,1) <--- DON'T USE A(1:3,2) = B(1:3,1) <--- BETTER USE In many cases the use of explicit array subscripts is inevitable for parallelization. Additionally, it is an easy means to prevent memory leaks. Enclose all write statements with the following: !$OMP CRITICAL (write2out) <your write statement> !$OMP END CRITICAL (write2out) Whenever you change something in the code and are not sure if it affects parallelization and leads to nonconforming behavior, please ask me and/or Franz to check this.
2011-03-17 16:16:17 +05:30
!$OMP CRITICAL (write2out)
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
write(6,*)
write(6,*) '<<<+- mesh init -+>>>'
write(6,*) '$Id$'
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
#include "compilation_info.f90"
openmp parallelization working again (at least for j2 and nonlocal constitutive model). In order to keep it like that, please follow these simple rules: DON'T use implicit array subscripts: example: real, dimension(3,3) :: A,B A(:,2) = B(:,1) <--- DON'T USE A(1:3,2) = B(1:3,1) <--- BETTER USE In many cases the use of explicit array subscripts is inevitable for parallelization. Additionally, it is an easy means to prevent memory leaks. Enclose all write statements with the following: !$OMP CRITICAL (write2out) <your write statement> !$OMP END CRITICAL (write2out) Whenever you change something in the code and are not sure if it affects parallelization and leads to nonconforming behavior, please ask me and/or Franz to check this.
2011-03-17 16:16:17 +05:30
!$OMP END CRITICAL (write2out)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
reworked restarting for compatibility with abaqus (not yet fully working) added new orientation feature for direct simulation: component type (random) asigns random orientation to an entire grain
2011-05-28 15:14:43 +05:30
call mesh_build_FEdata() ! --- get properties of the different types of elements
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
call IO_open_inputFile(fileUnit,FEmodelGeometry) ! --- parse info from input file...
select case (FEsolver)
case ('Spectral')
call mesh_spectral_count_nodesAndElements(fileUnit)
call mesh_spectral_count_cpElements()
call mesh_spectral_map_elements()
call mesh_spectral_map_nodes()
call mesh_spectral_count_cpSizes()
call mesh_spectral_build_nodes(fileUnit)
call mesh_spectral_build_elements(fileUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
case ('Marc')
call mesh_marc_get_tableStyles(fileUnit)
call mesh_marc_count_nodesAndElements(fileUnit)
call mesh_marc_count_elementSets(fileUnit)
call mesh_marc_map_elementSets(fileUnit)
call mesh_marc_count_cpElements(fileUnit)
call mesh_marc_map_elements(fileUnit)
call mesh_marc_map_nodes(fileUnit)
call mesh_marc_build_nodes(fileUnit)
call mesh_marc_count_cpSizes(fileunit)
call mesh_marc_build_elements(fileUnit)
case ('Abaqus')
noPart = IO_abaqus_hasNoPart(fileUnit)
call mesh_abaqus_count_nodesAndElements(fileUnit)
call mesh_abaqus_count_elementSets(fileUnit)
call mesh_abaqus_count_materials(fileUnit)
call mesh_abaqus_map_elementSets(fileUnit)
call mesh_abaqus_map_materials(fileUnit)
call mesh_abaqus_count_cpElements(fileUnit)
call mesh_abaqus_map_elements(fileUnit)
call mesh_abaqus_map_nodes(fileUnit)
call mesh_abaqus_build_nodes(fileUnit)
call mesh_abaqus_count_cpSizes(fileunit)
call mesh_abaqus_build_elements(fileUnit)
end select
call mesh_get_damaskOptions(fileUnit)
close (fileUnit)
call mesh_build_subNodeCoords()
call mesh_build_ipCoordinates()
call mesh_build_ipVolumes()
call mesh_build_ipAreas()
call mesh_build_nodeTwins()
call mesh_build_sharedElems()
call mesh_build_ipNeighborhood()
call mesh_tell_statistics()
parallelExecution = (parallelExecution .and. (mesh_Nelems == mesh_NcpElems)) ! plus potential killer from non-local constitutive
added function calls in init()
2007-04-10 16:52:53 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FEsolving_execElem = [ 1_pInt,mesh_NcpElems]
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
allocate(FEsolving_execIP(2_pInt,mesh_NcpElems)); FEsolving_execIP = 1_pInt
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
forall (e = 1_pInt:mesh_NcpElems) FEsolving_execIP(2,e) = FE_Nips(mesh_element(2,e))
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
allocate(calcMode(mesh_maxNips,mesh_NcpElems))
calcMode = .false. ! pretend to have collected what first call is asking (F = I)
calcMode(ip,mesh_FEasCP('elem',element)) = .true. ! first ip,el needs to be already pingponged to "calc"
lastMode = .true. ! and its mode is already known...
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endsubroutine
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
fixed bug in mesh_build_sharedElems which so far counted nodes in distorted elements twice -- part of this problem has already been fixed with revision 182 but then only in getNodeElemDimensions. M mesh.f90
2008-06-17 02:19:48 +05:30
!***********************************************************
! mapping of FE element types to internal representation
!***********************************************************
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
function FE_mapElemtype(what)
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
use IO, only: IO_lc
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
fixed bug in mesh_build_sharedElems which so far counted nodes in distorted elements twice -- part of this problem has already been fixed with revision 182 but then only in getNodeElemDimensions. M mesh.f90
2008-06-17 02:19:48 +05:30
implicit none
character(len=*), intent(in) :: what
integer(pInt) FE_mapElemtype
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
select case (IO_lc(what))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case ( '7', &
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
'c3d8')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 1_pInt ! Three-dimensional Arbitrarily Distorted Brick
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case ('134', &
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
'c3d4')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 2_pInt ! Three-dimensional Four-node Tetrahedron
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case ( '11', &
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
'cpe4')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 3_pInt ! Arbitrary Quadrilateral Plane-strain
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case ( '27', &
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
'cpe8')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 4_pInt ! Plane Strain, Eight-node Distorted Quadrilateral
added two more element types (157 and 136)
2008-06-17 14:41:54 +05:30
case ('157')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 5_pInt ! Three-dimensional, Low-order, Tetrahedron, Herrmann Formulations
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case ('136', &
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
'c3d6')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 6_pInt ! Three-dimensional Arbitrarily Distorted Pentahedral
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case ( '21', &
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
'c3d20')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 7_pInt ! Three-dimensional Arbitrarily Distorted quadratic hexahedral
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case ( '117', &
'123', &
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
'c3d8r')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 8_pInt ! Three-dimensional Arbitrarily Distorted linear hexahedral with reduced integration
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
case ( '57', &
'c3d20r')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 9_pInt ! Three-dimensional Arbitrarily Distorted quad hexahedral with reduced integration
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
case ( '155', &
'125', &
'128')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 10_pInt ! Two-dimensional Plane Strain triangle (155: cubic shape function, 125/128: second order isoparametric)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case default
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
FE_mapElemtype = 0_pInt ! unknown element --> should raise an error upstream..!
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endselect
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endfunction
fixed bug in mesh_build_sharedElems which so far counted nodes in distorted elements twice -- part of this problem has already been fixed with revision 182 but then only in getNodeElemDimensions. M mesh.f90
2008-06-17 02:19:48 +05:30
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
!***********************************************************
added function calls in init()
2007-04-10 16:52:53 +05:30
! FE to CP id mapping by binary search thru lookup array
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
!
added function calls in init()
2007-04-10 16:52:53 +05:30
! valid questions are 'elem', 'node'
!***********************************************************
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
function mesh_FEasCP(what,id)
added function calls in init()
2007-04-10 16:52:53 +05:30
use prec, only: pInt
use IO, only: IO_lc
implicit none
character(len=*), intent(in) :: what
integer(pInt), intent(in) :: id
integer(pInt), dimension(:,:), pointer :: lookupMap
integer(pInt) mesh_FEasCP, lower,upper,center
mesh_FEasCP = 0_pInt
select case(IO_lc(what(1:4)))
case('elem')
lookupMap => mesh_mapFEtoCPelem
case('node')
lookupMap => mesh_mapFEtoCPnode
case default
return
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endselect
added function calls in init()
2007-04-10 16:52:53 +05:30
lower = 1_pInt
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
upper = int(size(lookupMap,2_pInt),pInt)
added function calls in init()
2007-04-10 16:52:53 +05:30
! check at bounds QUESTION is it valid to extend bounds by 1 and just do binary search w/o init check at bounds?
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (lookupMap(1_pInt,lower) == id) then
mesh_FEasCP = lookupMap(2_pInt,lower)
added function calls in init()
2007-04-10 16:52:53 +05:30
return
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
elseif (lookupMap(1_pInt,upper) == id) then
mesh_FEasCP = lookupMap(2_pInt,upper)
added function calls in init()
2007-04-10 16:52:53 +05:30
return
endif
! binary search in between bounds
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do while (upper-lower > 1_pInt)
center = (lower+upper)/2_pInt
if (lookupMap(1_pInt,center) < id) then
added function calls in init()
2007-04-10 16:52:53 +05:30
lower = center
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
elseif (lookupMap(1_pInt,center) > id) then
added function calls in init()
2007-04-10 16:52:53 +05:30
upper = center
else
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_FEasCP = lookupMap(2_pInt,center)
added function calls in init()
2007-04-10 16:52:53 +05:30
exit
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endif
enddo
added function calls in init()
2007-04-10 16:52:53 +05:30
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endfunction
added function calls in init()
2007-04-10 16:52:53 +05:30
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
!***********************************************************
added function calls in init()
2007-04-10 16:52:53 +05:30
! find face-matching element of same type
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!***********************************************************
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
subroutine mesh_faceMatch(elem, face ,matchingElem, matchingFace)
use prec, only: pInt
implicit none
!*** output variables
integer(pInt), intent(out) :: matchingElem, & ! matching CP element ID
matchingFace ! matching FE face ID
!*** input variables
integer(pInt), intent(in) :: face, & ! FE face ID
elem ! FE elem ID
!*** local variables
integer(pInt), dimension(FE_NfaceNodes(face,mesh_element(2,elem))) :: &
myFaceNodes ! global node ids on my face
integer(pInt) myType, &
candidateType, &
candidateElem, &
candidateFace, &
candidateFaceNode, &
minNsharedElems, &
NsharedElems, &
lonelyNode, &
i, &
n, &
dir ! periodicity direction
integer(pInt), dimension(:), allocatable :: element_seen
logical checkTwins
minNsharedElems = mesh_maxNsharedElems + 1_pInt ! init to worst case
matchingFace = 0_pInt
matchingElem = 0_pInt ! intialize to "no match found"
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
myType = mesh_element(2_pInt,elem) ! figure elemType
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do n = 1_pInt,FE_NfaceNodes(face,myType) ! loop over nodes on face
myFaceNodes(n) = mesh_FEasCP('node',mesh_element(4_pInt+FE_nodeOnFace(n,face,myType),elem)) ! CP id of face node
NsharedElems = mesh_sharedElem(1_pInt,myFaceNodes(n)) ! figure # shared elements for this node
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
if (NsharedElems < minNsharedElems) then
minNsharedElems = NsharedElems ! remember min # shared elems
lonelyNode = n ! remember most lonely node
endif
enddo
allocate(element_seen(minNsharedElems))
element_seen = 0_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
checkCandidate: do i = 1_pInt,minNsharedElems ! iterate over lonelyNode's shared elements
candidateElem = mesh_sharedElem(1_pInt+i,myFaceNodes(lonelyNode)) ! present candidate elem
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
if (all(element_seen /= candidateElem)) then ! element seen for the first time?
element_seen(i) = candidateElem
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
candidateType = mesh_element(2_pInt,candidateElem) ! figure elemType of candidate
checkCandidateFace: do candidateFace = 1_pInt,FE_maxNipNeighbors ! check each face of candidate
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
if (FE_NfaceNodes(candidateFace,candidateType) /= FE_NfaceNodes(face,myType) & ! incompatible face
.or. (candidateElem == elem .and. candidateFace == face)) then ! this is my face
cycle checkCandidateFace
endif
checkTwins = .false.
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do n = 1_pInt,FE_NfaceNodes(candidateFace,candidateType) ! loop through nodes on face
candidateFaceNode = mesh_FEasCP('node', mesh_element(4_pInt+FE_nodeOnFace(n,candidateFace,candidateType),candidateElem))
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
if (all(myFaceNodes /= candidateFaceNode)) then ! candidate node does not match any of my face nodes
checkTwins = .true. ! perhaps the twin nodes do match
exit
endif
enddo
if(checkTwins) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
checkCandidateFaceTwins: do dir = 1_pInt,3_pInt
do n = 1_pInt,FE_NfaceNodes(candidateFace,candidateType) ! loop through nodes on face
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
candidateFaceNode = mesh_FEasCP('node', mesh_element(4+FE_nodeOnFace(n,candidateFace,candidateType),candidateElem))
if (all(myFaceNodes /= mesh_nodeTwins(dir,candidateFaceNode))) then ! node twin does not match either
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (dir == 3_pInt) then
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
cycle checkCandidateFace
else
cycle checkCandidateFaceTwins ! try twins in next dimension
endif
endif
enddo
exit checkCandidateFaceTwins
enddo checkCandidateFaceTwins
endif
matchingFace = candidateFace
matchingElem = candidateElem
exit checkCandidate ! found my matching candidate
enddo checkCandidateFace
endif
enddo checkCandidate
deallocate(element_seen)
endsubroutine
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
!********************************************************************
! get properties of different types of finite elements
!
! assign globals:
! FE_nodesAtIP, FE_ipNeighbor, FE_subNodeParent, FE_subNodeOnIPFace
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
subroutine mesh_build_FEdata ()
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
use prec, only: pInt
implicit none
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
allocate(FE_nodesAtIP(FE_maxmaxNnodesAtIP,FE_maxNips,FE_Nelemtypes)) ; FE_nodesAtIP = 0_pInt
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
allocate(FE_ipNeighbor(FE_maxNipNeighbors,FE_maxNips,FE_Nelemtypes)) ; FE_ipNeighbor = 0_pInt
allocate(FE_subNodeParent(FE_maxNips,FE_maxNsubNodes,FE_Nelemtypes)) ; FE_subNodeParent = 0_pInt
allocate(FE_subNodeOnIPFace(FE_NipFaceNodes,FE_maxNipNeighbors,FE_maxNips,FE_Nelemtypes)) ; FE_subNodeOnIPFace = 0_pInt
! fill FE_nodesAtIP with data
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
FE_nodesAtIP(:,:FE_Nips(1),1) = & ! element 7
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
1, &
2, &
4, &
3, &
5, &
6, &
8, &
7 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(1),FE_Nips(1)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
FE_nodesAtIP(:,:FE_Nips(2),2) = & ! element 134
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
1,2,3,4 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(2),FE_Nips(2)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
FE_nodesAtIP(:,:FE_Nips(3),3) = & ! element 11
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
1, &
2, &
4, &
3 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(3),FE_Nips(3)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
FE_nodesAtIP(:,:FE_Nips(4),4) = & ! element 27
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
1,0, &
1,2, &
2,0, &
1,4, &
0,0, &
2,3, &
4,0, &
3,4, &
3,0 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(4),FE_Nips(4)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
FE_nodesAtIP(:,:FE_Nips(5),5) = & ! element 157
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
1, &
2, &
3, &
4 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(5),FE_Nips(5)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
FE_nodesAtIP(:,:FE_Nips(6),6) = & ! element 136
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
1, &
2, &
3, &
4, &
5, &
6 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(6),FE_Nips(6)/))
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
FE_nodesAtIP(:,:FE_Nips(7),7) = & ! element 21
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
1,0, 0,0, &
1,2, 0,0, &
2,0, 0,0, &
1,4, 0,0, &
1,3, 2,4, &
2,3, 0,0, &
4,0, 0,0, &
3,4, 0,0, &
3,0, 0,0, &
1,5, 0,0, &
1,6, 2,5, &
2,6, 0,0, &
1,8, 4,5, &
0,0, 0,0, &
2,7, 3,6, &
4,8, 0,0, &
3,8, 4,7, &
3,7, 0,0, &
5,0, 0,0, &
5,6, 0,0, &
6,0, 0,0, &
5,8, 0,0, &
5,7, 6,8, &
6,7, 0,0, &
8,0, 0,0, &
7,8, 0,0, &
7,0, 0,0 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(7),FE_Nips(7)/))
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
! FE_nodesAtIP(:,:FE_Nips(8),8) = & ! element 117 (c3d8r --> single IP per element, so no need for this mapping)
! reshape((/&
! 1,2,3,4,5,6,7,8 &
! /),(/FE_maxNnodesAtIP(8),FE_Nips(8)/))
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
FE_nodesAtIP(:,:FE_Nips(9),9) = & ! element 57 (c3d20r == c3d8 --> copy of 7)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
1, &
2, &
4, &
3, &
5, &
6, &
8, &
7 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(9),FE_Nips(9)/))
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
FE_nodesAtIP(:,:FE_Nips(10),10) = & ! element 155, 125, 128
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
1, &
2, &
3 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_maxNnodesAtIP(10),FE_Nips(10)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
! *** FE_ipNeighbor ***
! is a list of the neighborhood of each IP.
! It is sorted in (local) +x,-x, +y,-y, +z,-z direction.
! Positive integers denote an intra-FE IP identifier.
! Negative integers denote the interface behind which the neighboring (extra-FE) IP will be located.
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_ipNeighbor(:FE_NipNeighbors(1),:FE_Nips(1),1) = & ! element 7
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
2,-5, 3,-2, 5,-1, &
-3, 1, 4,-2, 6,-1, &
4,-5,-4, 1, 7,-1, &
-3, 3,-4, 2, 8,-1, &
6,-5, 7,-2,-6, 1, &
-3, 5, 8,-2,-6, 2, &
8,-5,-4, 5,-6, 3, &
-3, 7,-4, 6,-6, 4 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(1),FE_Nips(1)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_ipNeighbor(:FE_NipNeighbors(2),:FE_Nips(2),2) = & ! element 134
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
-1,-2,-3,-4 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(2),FE_Nips(2)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_ipNeighbor(:FE_NipNeighbors(3),:FE_Nips(3),3) = & ! element 11
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
2,-4, 3,-1, &
-2, 1, 4,-1, &
4,-4,-3, 1, &
-2, 3,-3, 2 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(3),FE_Nips(3)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_ipNeighbor(:FE_NipNeighbors(4),:FE_Nips(4),4) = & ! element 27
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
2,-4, 4,-1, &
3, 1, 5,-1, &
-2, 2, 6,-1, &
5,-4, 7, 1, &
6, 4, 8, 2, &
-2, 5, 9, 3, &
8,-4,-3, 4, &
9, 7,-3, 5, &
-2, 8,-3, 6 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(4),FE_Nips(4)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_ipNeighbor(:FE_NipNeighbors(5),:FE_Nips(5),5) = & ! element 157
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
2,-4, 3,-2, 4,-1, &
3,-2, 1,-3, 4,-1, &
1,-3, 2,-4, 4,-1, &
1,-3, 2,-4, 3,-2 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(5),FE_Nips(5)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_ipNeighbor(:FE_NipNeighbors(6),:FE_Nips(6),6) = & ! element 136
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
2,-4, 3,-2, 4,-1, &
-3, 1, 3,-2, 5,-1, &
2,-4,-3, 1, 6,-1, &
5,-4, 6,-2,-5, 1, &
-3, 4, 6,-2,-5, 2, &
5,-4,-3, 4,-5, 3 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(6),FE_Nips(6)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_ipNeighbor(:FE_NipNeighbors(7),:FE_Nips(7),7) = & ! element 21
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
corrected ip numbers in FE_ipNeighbor for element type 21; used to get wrong neighborhood for this element type
2010-04-06 17:15:23 +05:30
2,-5, 4,-2,10,-1, &
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
3, 1, 5,-2,11,-1, &
-3, 2, 6,-2,12,-1, &
5,-5, 7, 1,13,-1, &
6, 4, 8, 2,14,-1, &
-3, 5, 9, 3,15,-1, &
8,-5,-4, 4,16,-1, &
9, 7,-4, 5,17,-1, &
-3, 8,-4, 6,18,-1, &
corrected ip numbers in FE_ipNeighbor for element type 21; used to get wrong neighborhood for this element type
2010-04-06 17:15:23 +05:30
11,-5,13,-2,19, 1, &
12,10,14,-2,20, 2, &
-3,11,15,-2,21, 3, &
14,-5,16,10,22, 4, &
15,13,17,11,23, 5, &
-3,14,18,12,24, 6, &
17,-5,-4,13,25, 7, &
18,16,-4,14,26, 8, &
-3,17,-4,15,27, 9, &
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
20,-5,22,-2,-6,10, &
21,19,23,-2,-6,11, &
-3,20,24,-2,-6,12, &
23,-5,25,19,-6,13, &
24,22,26,20,-6,14, &
-3,23,27,21,-6,15, &
26,-5,-4,22,-6,16, &
27,25,-4,23,-6,17, &
-3,26,-4,24,-6,18 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(7),FE_Nips(7)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
FE_ipNeighbor(:FE_NipNeighbors(8),:FE_Nips(8),8) = & ! element 117
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
-3,-5,-4,-2,-6,-1 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(8),FE_Nips(8)/))
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
FE_ipNeighbor(:FE_NipNeighbors(9),:FE_Nips(9),9) = & ! element 57 (c3d20r == c3d8 --> copy of 7)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
2,-5, 3,-2, 5,-1, &
-3, 1, 4,-2, 6,-1, &
4,-5,-4, 1, 7,-1, &
-3, 3,-4, 2, 8,-1, &
6,-5, 7,-2,-6, 1, &
-3, 5, 8,-2,-6, 2, &
8,-5,-4, 5,-6, 3, &
-3, 7,-4, 6,-6, 4 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(9),FE_Nips(9)/))
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
FE_ipNeighbor(:FE_NipNeighbors(10),:FE_Nips(10),10) = & ! element 155, 125, 128
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
2,-3, 3,-1, &
-2, 1, 3,-1, &
2,-3,-2, 1 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipNeighbors(10),FE_Nips(10)/))
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
! *** FE_subNodeParent ***
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
! lists the group of nodes for which the center of gravity
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
! corresponds to the location of a each subnode.
! example: face-centered subnode with faceNodes 1,2,3,4 to be used in,
! e.g., a 8 IP grid, would be encoded:
! 1, 2, 3, 4, 0, 0, 0, 0
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeParent(:FE_Nips(1),:FE_NsubNodes(1),1) = & ! element 7
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
1, 2, 0, 0, 0, 0, 0, 0, &
2, 3, 0, 0, 0, 0, 0, 0, &
3, 4, 0, 0, 0, 0, 0, 0, &
4, 1, 0, 0, 0, 0, 0, 0, &
1, 5, 0, 0, 0, 0, 0, 0, &
2, 6, 0, 0, 0, 0, 0, 0, &
3, 7, 0, 0, 0, 0, 0, 0, &
4, 8, 0, 0, 0, 0, 0, 0, &
5, 6, 0, 0, 0, 0, 0, 0, &
6, 7, 0, 0, 0, 0, 0, 0, &
7, 8, 0, 0, 0, 0, 0, 0, &
8, 5, 0, 0, 0, 0, 0, 0, &
1, 2, 3, 4, 0, 0, 0, 0, &
1, 2, 6, 5, 0, 0, 0, 0, &
2, 3, 7, 6, 0, 0, 0, 0, &
3, 4, 8, 7, 0, 0, 0, 0, &
1, 4, 8, 5, 0, 0, 0, 0, &
5, 6, 7, 8, 0, 0, 0, 0, &
1, 2, 3, 4, 5, 6, 7, 8 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_Nips(1),FE_NsubNodes(1)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
!FE_subNodeParent(:FE_Nips(2),:FE_NsubNodes(2),2) ! element 134 has no subnodes
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeParent(:FE_Nips(3),:FE_NsubNodes(3),3) = & ! element 11
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
1, 2, 0, 0, &
2, 3, 0, 0, &
3, 4, 0, 0, &
4, 1, 0, 0, &
1, 2, 3, 4 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_Nips(3),FE_NsubNodes(3)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeParent(:FE_Nips(4),:FE_NsubNodes(4),4) = & ! element 27
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
1, 1, 2, 0, 0, 0, 0, 0, 0, &
1, 2, 2, 0, 0, 0, 0, 0, 0, &
2, 2, 3, 0, 0, 0, 0, 0, 0, &
2, 3, 3, 0, 0, 0, 0, 0, 0, &
3, 3, 4, 0, 0, 0, 0, 0, 0, &
3, 4, 4, 0, 0, 0, 0, 0, 0, &
4, 4, 1, 0, 0, 0, 0, 0, 0, &
4, 1, 1, 0, 0, 0, 0, 0, 0, &
1, 1, 1, 1, 2, 2, 4, 4, 3, &
2, 2, 2, 2, 1, 1, 3, 3, 4, &
3, 3, 3, 3, 2, 2, 4, 4, 1, &
4, 4, 4, 4, 1, 1, 3, 3, 2 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_Nips(4),FE_NsubNodes(4)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
!FE_subNodeParent(:FE_Nips(5),:FE_NsubNodes(5),5) = & ! element 157
! reshape((/&
! *still to be defined*
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
! ],pInt),(/FE_Nips(5),FE_NsubNodes(5)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeParent(:FE_Nips(6),:FE_NsubNodes(6),6) = & ! element 136
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
1, 2, 0, 0, 0, 0, &
2, 3, 0, 0, 0, 0, &
3, 1, 0, 0, 0, 0, &
1, 4, 0, 0, 0, 0, &
2, 5, 0, 0, 0, 0, &
3, 6, 0, 0, 0, 0, &
4, 5, 0, 0, 0, 0, &
5, 6, 0, 0, 0, 0, &
6, 4, 0, 0, 0, 0, &
1, 2, 3, 0, 0, 0, &
1, 2, 4, 5, 0, 0, &
2, 3, 5, 6, 0, 0, &
1, 3, 4, 6, 0, 0, &
4, 5, 6, 0, 0, 0, &
1, 2, 3, 4, 5, 6 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_Nips(6),FE_NsubNodes(6)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeParent(:FE_Nips(7),:FE_NsubNodes(7),7) = & ! element 21
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
1, 1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
1, 2, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
2, 2, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
2, 3, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
3, 3, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
3, 4, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
4, 4, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
4, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
1, 1, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
2, 2, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
3, 3, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
4, 4, 8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
1, 5, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
2, 6, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
3, 7, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
4, 8, 8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
5, 5, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
5, 6, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
6, 6, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
6, 7, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
7, 7, 8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
7, 8, 8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
8, 8, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
8, 5, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
1, 1, 1, 1, 2, 2, 4, 4, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
2, 2, 2, 2, 1, 1, 3, 3, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
3, 3, 3, 3, 2, 2, 4, 4, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
4, 4, 4, 4, 1, 1, 3, 3, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
1, 1, 1, 1, 2, 2, 5, 5, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
2, 2, 2, 2, 1, 1, 6, 6, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
2, 2, 2, 2, 3, 3, 6, 6, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
3, 3, 3, 3, 2, 2, 7, 7, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
3, 3, 3, 3, 4, 4, 7, 7, 8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
4, 4, 4, 4, 3, 3, 8, 8, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
4, 4, 4, 4, 1, 1, 8, 8, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
1, 1, 1, 1, 4, 4, 5, 5, 8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
5, 5, 5, 5, 1, 1, 6, 6, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
6, 6, 6, 6, 2, 2, 5, 5, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
6, 6, 6, 6, 2, 2, 7, 7, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
7, 7, 7, 7, 3, 3, 6, 6, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
7, 7, 7, 7, 3, 3, 8, 8, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
8, 8, 8, 8, 4, 4, 7, 7, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
8, 8, 8, 8, 4, 4, 5, 5, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
5, 5, 5, 5, 1, 1, 8, 8, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
5, 5, 5, 5, 6, 6, 8, 8, 7, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
6, 6, 6, 6, 5, 5, 7, 7, 8, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
7, 7, 7, 7, 6, 6, 8, 8, 5, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
8, 8, 8, 8, 5, 5, 7, 7, 6, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 4, 4, 4, 4, 5, 5, 5, 5, 3, 3, 6, 6, 8, 8, 7, &
2, 2, 2, 2, 2, 2, 2, 2, 1, 1, 1, 1, 3, 3, 3, 3, 6, 6, 6, 6, 4, 4, 5, 5, 7, 7, 8, &
3, 3, 3, 3, 3, 3, 3, 3, 2, 2, 2, 2, 4, 4, 4, 4, 7, 7, 7, 7, 1, 1, 6, 6, 8, 8, 5, &
4, 4, 4, 4, 4, 4, 4, 4, 1, 1, 1, 1, 3, 3, 3, 3, 8, 8, 8, 8, 2, 2, 5, 5, 7, 7, 6, &
5, 5, 5, 5, 5, 5, 5, 5, 1, 1, 1, 1, 6, 6, 6, 6, 8, 8, 8, 8, 2, 2, 4, 4, 7, 7, 3, &
6, 6, 6, 6, 6, 6, 6, 6, 2, 2, 2, 2, 5, 5, 5, 5, 7, 7, 7, 7, 1, 1, 3, 3, 8, 8, 4, &
7, 7, 7, 7, 7, 7, 7, 7, 3, 3, 3, 3, 6, 6, 6, 6, 8, 8, 8, 8, 2, 2, 4, 4, 5, 5, 1, &
8, 8, 8, 8, 8, 8, 8, 8, 4, 4, 4, 4, 5, 5, 5, 5, 7, 7, 7, 7, 1, 1, 3, 3, 6, 6, 2 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_Nips(7),FE_NsubNodes(7)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
!FE_subNodeParent(:FE_Nips(8),:FE_NsubNodes(8),8) ! element 117 has no subnodes
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
FE_subNodeParent(:FE_Nips(9),:FE_NsubNodes(9),9) = & ! element 57 (c3d20r == c3d8 --> copy of 7)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
1, 2, 0, 0, 0, 0, 0, 0, &
2, 3, 0, 0, 0, 0, 0, 0, &
3, 4, 0, 0, 0, 0, 0, 0, &
4, 1, 0, 0, 0, 0, 0, 0, &
1, 5, 0, 0, 0, 0, 0, 0, &
2, 6, 0, 0, 0, 0, 0, 0, &
3, 7, 0, 0, 0, 0, 0, 0, &
4, 8, 0, 0, 0, 0, 0, 0, &
5, 6, 0, 0, 0, 0, 0, 0, &
6, 7, 0, 0, 0, 0, 0, 0, &
7, 8, 0, 0, 0, 0, 0, 0, &
8, 5, 0, 0, 0, 0, 0, 0, &
1, 2, 3, 4, 0, 0, 0, 0, &
1, 2, 6, 5, 0, 0, 0, 0, &
2, 3, 7, 6, 0, 0, 0, 0, &
3, 4, 8, 7, 0, 0, 0, 0, &
1, 4, 8, 5, 0, 0, 0, 0, &
5, 6, 7, 8, 0, 0, 0, 0, &
1, 2, 3, 4, 5, 6, 7, 8 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_Nips(9),FE_NsubNodes(9)/))
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
FE_subNodeParent(:FE_Nips(10),:FE_NsubNodes(10),10) = & ! element 155, 125, 128
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
1, 2, 0, &
2, 3, 0, &
3, 1, 0, &
1, 2, 3 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_Nips(10),FE_NsubNodes(10)/))
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
! *** FE_subNodeOnIPFace ***
! indicates which subnodes make up the interfaces enclosing the IP volume.
! The sorting convention is such that the outward pointing normal
! follows from a right-handed traversal of the face node list.
! For two-dimensional elements, which only have lines as "interface"
! one nevertheless has to specify each interface by a closed path,
! e.g., 1,2, 2,1, assuming the line connects nodes 1 and 2.
! This will result in zero ipVolume and interfaceArea, but is not
! detrimental at the moment since non-local constitutive laws are
! currently not foreseen in 2D cases.
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(1),:FE_Nips(1),1) = & ! element 7
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
9,21,27,22, & ! 1
1,13,25,12, &
12,25,27,21, &
1, 9,22,13, &
13,22,27,25, &
1,12,21, 9, &
2,10,23,14, & ! 2
9,22,27,21, &
10,21,27,23, &
2,14,22, 9, &
14,23,27,22, &
2, 9,21,10, &
11,24,27,21, & ! 3
4,12,25,16, &
4,16,24,11, &
12,21,27,25, &
16,25,27,24, &
4,11,21,12, &
3,15,23,10, & ! 4
11,21,27,24, &
3,11,24,15, &
10,23,27,21, &
15,24,27,23, &
3,10,21,11, &
17,22,27,26, & ! 5
5,20,25,13, &
20,26,27,25, &
5,13,22,17, &
5,17,26,20, &
13,25,27,22, &
6,14,23,18, & ! 6
17,26,27,22, &
18,23,27,26, &
6,17,22,14, &
6,18,26,17, &
14,22,27,23, &
19,26,27,24, & ! 7
8,16,25,20, &
8,19,24,16, &
20,25,27,26, &
8,20,26,19, &
16,24,27,25, &
7,18,23,15, & ! 8
19,24,27,26, &
7,15,24,19, &
18,26,27,23, &
7,19,26,18, &
15,23,27,24 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(1),FE_Nips(1)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(2),:FE_Nips(2),2) = & ! element 134
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
1, 1, 3, 2, & ! 1
1, 1, 2, 4, &
2, 2, 3, 4, &
1, 1, 4, 3 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(2),FE_Nips(2)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(3),:FE_Nips(3),3) = & ! element 11
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
5, 9, 9, 5 , & ! 1
1, 8, 8, 1 , &
8, 9, 9, 8 , &
1, 5, 5, 1 , &
2, 6, 6, 2 , & ! 2
5, 9, 9, 5 , &
6, 9, 9, 6 , &
2, 5, 5, 2 , &
3, 6, 6, 3 , & ! 3
7, 9, 9, 7 , &
3, 7, 7, 3 , &
6, 9, 9, 6 , &
7, 9, 9, 7 , & ! 4
4, 8, 8, 4 , &
4, 7, 7, 4 , &
8, 9, 9, 8 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(3),FE_Nips(3)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(4),:FE_Nips(4),4) = & ! element 27
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
9,17,17, 9 , & ! 1
1,16,16, 1 , &
16,17,17,16 , &
1, 9, 9, 1 , &
10,18,18,10 , & ! 2
9,17,17, 9 , &
17,18,18,17 , &
9,10,10, 9 , &
2,11,11, 2 , & ! 3
10,18,18,10 , &
11,18,18,11 , &
2,10,10, 2 , &
17,20,20,17 , & ! 4
15,16,16,15 , &
15,20,20,15 , &
16,17,17,16 , &
18,19,19,18 , & ! 5
17,20,20,17 , &
19,20,20,19 , &
17,18,18,17 , &
11,12,12,11 , & ! 6
18,19,19,18 , &
12,19,19,12 , &
11,18,18,11 , &
14,20,20,14 , & ! 7
4,15,15, 4 , &
4,14,14, 4 , &
15,20,20,15 , &
13,19,19,13 , & ! 8
14,20,20,14 , &
13,14,14,13 , &
19,20,20,19 , &
3,12,12, 3 , & ! 9
13,19,19,13 , &
3,13,13, 3 , &
12,19,19,12 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(4),FE_Nips(4)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
!FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(5),:FE_Nips(5),5) = & ! element 157
! reshape((/&
! *still to be defined*
! /),(/FE_NipFaceNodes,FE_NipNeighbors(5),FE_Nips(5)/))
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(6),:FE_Nips(6),6) = & ! element 136
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
7,16,21,17, & ! 1
1,10,19, 9, &
9,19,21,16, &
1, 7,17,10, &
10,17,21,19, &
1, 9,16, 7, &
2, 8,18,11, & ! 2
7,17,21,16, &
8,16,21,18, &
2,11,17, 7, &
11,18,21,17, &
2, 7,16, 8, &
8,18,21,16, & ! 3
3, 9,19,12, &
3,12,18, 8, &
9,16,21,19, &
12,19,21,18, &
3, 8,16, 9, &
13,17,21,20, & ! 4
4,15,19,10, &
15,20,21,19, &
4,10,17,13, &
4,13,20,15, &
10,19,21,17, &
5,11,18,14, & ! 5
13,20,21,17, &
14,18,21,20, &
5,13,17,11, &
5,14,20,13, &
11,17,21,18, &
14,20,21,18, & ! 6
6,12,19,15, &
6,14,18,12, &
15,19,21,20, &
6,15,20,14, &
12,18,21,19 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(6),FE_Nips(6)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(7),:FE_Nips(7),7) = & ! element 21
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
9,33,57,37, & ! 1
1,17,44,16, &
33,16,44,57, &
1, 9,37,17, &
17,37,57,44, &
1,16,33, 9, &
10,34,58,38, & ! 2
9,37,57,33, &
34,33,57,58, &
9,10,38,37, &
37,38,58,57, &
9,33,34,10, &
2,11,39,18, & ! 3
10,38,58,34, &
11,34,58,39, &
10, 2,18,38, &
38,18,39,58, &
10,34,11, 2, &
33,36,60,57, & ! 4
16,44,43,15, &
36,15,43,60, &
16,33,57,44, &
44,57,60,43, &
16,15,36,33, &
34,35,59,58, & ! 5
33,57,60,36, &
35,36,60,59, &
33,34,58,57, &
57,58,59,60, &
33,36,35,34, &
11,12,40,39, & ! 6
34,58,59,35, &
12,35,59,40, &
34,11,39,58, &
58,39,40,59, &
34,35,12,11, &
36,14,42,60, & ! 7
15,43,20, 4, &
14, 4,20,42, &
15,36,60,43, &
43,60,42,20, &
15, 4,14,36, &
35,13,41,59, & ! 8
36,60,42,14, &
13,14,42,41, &
36,35,59,60, &
60,59,41,42, &
36,14,13,35, &
12, 3,19,40, & ! 9
35,59,41,13, &
3,13,41,19, &
35,12,40,59, &
59,40,19,41, &
35,13, 3,12, &
37,57,61,45, & ! 10
17,21,52,44, &
57,44,52,61, &
17,37,45,21, &
21,45,61,52, &
17,44,57,37, &
38,58,62,46, & ! 11
37,45,61,57, &
58,57,61,62, &
37,38,46,45, &
45,46,62,61, &
37,57,58,38, &
18,39,47,22, & ! 12
38,46,62,58, &
39,58,62,47, &
38,18,22,46, &
46,22,47,62, &
38,58,39,18, &
57,60,64,61, & ! 13
44,52,51,43, &
60,43,51,64, &
44,57,61,52, &
52,61,64,51, &
44,43,60,57, &
58,59,63,62, & ! 14
57,61,64,60, &
59,60,64,63, &
57,58,62,61, &
61,62,63,64, &
57,60,59,58, &
39,40,48,47, & ! 15
58,62,63,59, &
40,59,63,48, &
58,39,47,62, &
62,47,48,63, &
58,59,40,39, &
60,42,50,64, & ! 16
43,51,24,20, &
42,20,24,50, &
43,60,64,51, &
51,64,50,24, &
43,20,42,60, &
59,41,49,63, & ! 17
60,64,50,42, &
41,42,50,49, &
60,59,63,64, &
64,63,49,50, &
60,42,41,59, &
40,19,23,48, & ! 18
59,63,49,41, &
19,41,49,23, &
59,40,48,63, &
63,48,23,49, &
59,41,19,40, &
45,61,53,25, & ! 19
21, 5,32,52, &
61,52,32,53, &
21,45,25, 5, &
5,25,53,32, &
21,52,61,45, &
46,62,54,26, & ! 20
45,25,53,61, &
62,61,53,54, &
45,46,26,25, &
25,26,54,53, &
45,61,62,46, &
22,47,27, 6, & ! 21
46,26,54,62, &
47,62,54,27, &
46,22, 6,26, &
26, 6,27,54, &
46,62,47,22, &
61,64,56,53, & ! 22
52,32,31,51, &
64,51,31,56, &
52,61,53,32, &
32,53,56,31, &
52,51,64,61, &
62,63,55,54, & ! 23
61,53,56,64, &
63,64,56,55, &
61,62,54,53, &
53,54,55,56, &
61,64,63,62, &
47,48,28,27, & ! 24
62,54,55,63, &
48,63,55,28, &
62,47,27,54, &
54,27,28,55, &
62,63,48,47, &
64,50,30,56, & ! 25
51,31, 8,24, &
50,24, 8,30, &
51,64,56,31, &
31,56,30, 8, &
51,24,50,64, &
63,49,29,55, & ! 26
64,56,30,50, &
49,50,30,29, &
64,63,55,56, &
56,55,29,30, &
64,50,49,63, &
48,23, 7,28, & ! 27
63,55,29,49, &
23,49,29, 7, &
63,48,28,55, &
55,28, 7,29, &
63,49,23,48 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(7),FE_Nips(7)/))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(8),:FE_Nips(8),8) = & ! element 117
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
2, 3, 7, 6, & ! 1
1, 5, 8, 4, &
3, 4, 8, 7, &
1, 2, 6, 5, &
5, 6, 7, 8, &
1, 4, 3, 2 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(8),FE_Nips(8)/))
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(9),:FE_Nips(9),9) = & ! element 57 (c3d20r == c3d8 --> copy of 7)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
1) added element type 57 (C3D20R) reduced quadratic hexahedral 2) need discussion about the present omission of the FE_nodesAtIP for type 8 --> added comment into source code
2010-08-17 04:23:24 +05:30
9,21,27,22, & ! 1
1,13,25,12, &
12,25,27,21, &
1, 9,22,13, &
13,22,27,25, &
1,12,21, 9, &
2,10,23,14, & ! 2
9,22,27,21, &
10,21,27,23, &
2,14,22, 9, &
14,23,27,22, &
2, 9,21,10, &
11,24,27,21, & ! 3
4,12,25,16, &
4,16,24,11, &
12,21,27,25, &
16,25,27,24, &
4,11,21,12, &
3,15,23,10, & ! 4
11,21,27,24, &
3,11,24,15, &
10,23,27,21, &
15,24,27,23, &
3,10,21,11, &
17,22,27,26, & ! 5
5,20,25,13, &
20,26,27,25, &
5,13,22,17, &
5,17,26,20, &
13,25,27,22, &
6,14,23,18, & ! 6
17,26,27,22, &
18,23,27,26, &
6,17,22,14, &
6,18,26,17, &
14,22,27,23, &
19,26,27,24, & ! 7
8,16,25,20, &
8,19,24,16, &
20,25,27,26, &
8,20,26,19, &
16,24,27,25, &
7,18,23,15, & ! 8
19,24,27,26, &
7,15,24,19, &
18,26,27,23, &
7,19,26,18, &
15,23,27,24 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(9),FE_Nips(9)/))
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
FE_subNodeOnIPFace(:FE_NipFaceNodes,:FE_NipNeighbors(10),:FE_Nips(10),10) = & ! element 155, 125, 128
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
reshape(int([&
added new 2D triangle elements added some sourcecode commenting on internal database formats corrected database to allow for ipVol and Area calc in 2D element cases
2011-04-06 14:05:37 +05:30
4, 7, 7, 4 , & ! 1
1, 6, 6, 1 , &
6, 7, 7, 6 , &
1, 4, 4, 1 , &
2, 5, 5, 2 , & ! 2
4, 7, 7, 4 , &
5, 7, 7, 5 , &
2, 4, 4, 2 , &
5, 7, 7, 5 , & ! 3
3, 6, 6, 3 , &
3, 5, 5, 3 , &
6, 7, 7, 6 &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
],pInt),(/FE_NipFaceNodes,FE_NipNeighbors(10),FE_Nips(10)/))
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endsubroutine
now with full functionality for element type 6 (6-node pentaeder) and with element data structure partly initialized as function variables and not parameters
2009-04-06 18:55:19 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! figure out table styles (Marc only)
added function calls in init()
2007-04-10 16:52:53 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! initialcondTableStyle, hypoelasticTableStyle
added function calls in init()
2007-04-10 16:52:53 +05:30
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_get_tableStyles (myUnit)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
use prec, only: pInt
added function calls in init()
2007-04-10 16:52:53 +05:30
use IO
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
implicit none
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 6_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
initialcondTableStyle = 0_pInt
hypoelasticTableStyle = 0_pInt
added function calls in init()
2007-04-10 16:52:53 +05:30
610 FORMAT(A300)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
added function calls in init()
2007-04-10 16:52:53 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'table' .and. myPos(1_pInt) .GT. 5) then
initialcondTableStyle = IO_intValue(line,myPos,4_pInt)
hypoelasticTableStyle = IO_intValue(line,myPos,5_pInt)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
exit
endif
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
enddo
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
620 endsubroutine
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
!********************************************************************
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
! get any additional damask options from input file
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
!
! mesh_periodicSurface
!********************************************************************
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
subroutine mesh_get_damaskOptions(myUnit)
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
use DAMASK_interface, only: FEsolver
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
use prec, only: pInt
use IO
implicit none
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), intent(in) :: myUnit
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 5_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
integer(pInt) chunk, Nchunks
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
character(len=300) line, keyword, damaskOption, v
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
mesh_periodicSurface = (/.false., .false., .false./)
610 FORMAT(A300)
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
select case (FEsolver)
case ('Spectral') ! no special keyword needed, the damask option directly goes into the header
case ('Marc')
keyword = '$damask'
case ('Abaqus')
keyword = '**damask'
end select
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
Nchunks = myPos(1)
select case (FEsolver)
case ('Marc','Abaqus')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if (IO_lc(IO_stringValue(line,myPos,1_pInt)) == keyword .and. Nchunks > 1_pInt) then ! found keyword for damask option and there is at least one more chunk to read
damaskOption = IO_lc(IO_stringValue(line,myPos,2_pInt))
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
select case(damaskOption)
case('periodic') ! damask Option that allows to specify periodic fluxes
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
do chunk = 3_pInt,Nchunks ! loop through chunks (skipping the keyword)
v = IO_lc(IO_stringValue(line,myPos,chunk)) ! chunk matches keyvalues x,y, or z?
mesh_periodicSurface(1) = mesh_periodicSurface(1) .or. v == 'x'
mesh_periodicSurface(2) = mesh_periodicSurface(2) .or. v == 'y'
mesh_periodicSurface(3) = mesh_periodicSurface(3) .or. v == 'z'
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
enddo
endselect
endif
case('Spectral')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
damaskOption = IO_lc(IO_stringValue(line,myPos,1_pInt))
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
select case(damaskOption)
case('periodic') ! damask Option that allows to specify periodic fluxes
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
do chunk = 2_pInt,Nchunks ! loop through chunks (skipping the keyword)
v = IO_lc(IO_stringValue(line,myPos,chunk)) ! chunk matches keyvalues x,y, or z?
mesh_periodicSurface(1) = mesh_periodicSurface(1) .or. v == 'x'
mesh_periodicSurface(2) = mesh_periodicSurface(2) .or. v == 'y'
mesh_periodicSurface(3) = mesh_periodicSurface(3) .or. v == 'z'
renamed subroutine "mesh_marc_get_mpieOptions" to "mesh_get_damaskOptions" which is now also used for abaqus and spectral input files. The only available option so far is "periodic", which allows for periodic fluxes in the specified direction. Examples for usage (add the following lines to the respective input file, in case of spectral file this goes into the header): - periodicity in x and z direction for marc: $damask periodic x z - periodicity in y direction for abaqus: **damask periodic y - periodicity in x and y direction for spectral: periodic y x
2012-02-03 18:07:52 +05:30
enddo
endselect
endselect
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
620 endsubroutine
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
!********************************************************************
! count overall number of nodes and elements in mesh
!
! mesh_Nelems, mesh_Nnodes
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_spectral_count_nodesAndElements (myUnit)
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 7_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
integer(pInt) a,b,c,i,j,headerLength
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
character(len=1024) line,keyword
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
mesh_Nnodes = 0_pInt
mesh_Nelems = 0_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
read(myUnit,'(a1024)') line
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos = IO_stringPos(line,2_pInt)
keyword = IO_lc(IO_StringValue(line,myPos,2_pInt))
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
if (keyword(1:4) == 'head') then
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
headerLength = IO_intValue(line,myPos,1_pInt) + 1_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
else
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
call IO_error(error_ID=842_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
endif
rewind(myUnit)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt, headerLength
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read(myUnit,'(a1024)') line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_StringValue(line,myPos,1_pInt)) == 'resolution') then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 2_pInt,6_pInt,2_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
select case (IO_lc(IO_stringValue(line,myPos,j)))
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case('a')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
a = IO_intValue(line,myPos,j+1_pInt)
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case('b')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
b = IO_intValue(line,myPos,j+1_pInt)
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
case('c')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
c = IO_intValue(line,myPos,j+1_pInt)
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
end select
enddo
changed format of resolution. instead of reading exponent of two, now the number given equals directly to the number of FPs. Allows us to use arbitrary number of FPs. example: OLD: reading 5 results in 2^5=32 FPs NEW: reading 33 results in 33 FPs
2010-10-01 16:12:15 +05:30
mesh_Nelems = a * b * c
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_Nnodes = (1_pInt + a)*(1_pInt + b)*(1_pInt + c)
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
endif
enddo
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
endsubroutine
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! count overall number of nodes and elements in mesh
generalized _FEasCP (string what,int ID) to use targeted arrays available what options are 'elem' and 'node'
2007-03-27 18:23:31 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! mesh_Nelems, mesh_Nnodes
added function calls in init()
2007-04-10 16:52:53 +05:30
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_count_nodesAndElements (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
use prec, only: pInt
added function calls in init()
2007-04-10 16:52:53 +05:30
use IO
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
implicit none
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 4_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
mesh_Nnodes = 0_pInt
mesh_Nelems = 0_pInt
added function calls in init()
2007-04-10 16:52:53 +05:30
610 FORMAT(A300)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_StringValue(line,myPos,1_pInt)) == 'sizing') then
mesh_Nelems = IO_IntValue (line,myPos,3_pInt)
mesh_Nnodes = IO_IntValue (line,myPos,4_pInt)
keyword "table" appeared more than once -- restricted interpretation to header
2007-09-28 20:26:26 +05:30
exit
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endif
enddo
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
620 endsubroutine
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
added awareness of set definitions (new function build_elemSetMapping) added new style tables for initial conditions
2007-10-23 18:39:46 +05:30
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! count overall number of nodes and elements in mesh
added awareness of set definitions (new function build_elemSetMapping) added new style tables for initial conditions
2007-10-23 18:39:46 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! mesh_Nelems, mesh_Nnodes
added awareness of set definitions (new function build_elemSetMapping) added new style tables for initial conditions
2007-10-23 18:39:46 +05:30
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_count_nodesAndElements (myUnit)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added awareness of set definitions (new function build_elemSetMapping) added new style tables for initial conditions
2007-10-23 18:39:46 +05:30
use prec, only: pInt
use IO
implicit none
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical inPart
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_Nnodes = 0_pInt
mesh_Nelems = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
if (inPart .or. noPart) then
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
select case ( IO_lc(IO_stringValue(line,myPos,1_pInt)))
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case('*node')
if( &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'output' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'print' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'file' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'response' &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
) &
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
mesh_Nnodes = mesh_Nnodes + IO_countDataLines(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case('*element')
if( &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'output' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'matrix' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'response' &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
) then
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
mesh_Nelems = mesh_Nelems + IO_countDataLines(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
endselect
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endif
enddo
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
620 if (mesh_Nnodes < 2_pInt) call IO_error(error_ID=900_pInt)
if (mesh_Nelems == 0_pInt) call IO_error(error_ID=901_pInt)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endsubroutine
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! count overall number of element sets in mesh
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! mesh_NelemSets, mesh_maxNelemInSet
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_count_elementSets (myUnit)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
use prec, only: pInt
use IO
implicit none
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_NelemSets = 0_pInt
mesh_maxNelemInSet = 0_pInt
610 FORMAT(A300)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_StringValue(line,myPos,1_pInt)) == 'define' .and. &
IO_lc(IO_StringValue(line,myPos,2_pInt)) == 'element' ) then
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_NelemSets = mesh_NelemSets + 1_pInt
mesh_maxNelemInSet = max(mesh_maxNelemInSet, &
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
IO_countContinousIntValues(myUnit))
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endif
enddo
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
620 endsubroutine
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! count overall number of element sets in mesh
added function calls in init()
2007-04-10 16:52:53 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! mesh_NelemSets, mesh_maxNelemInSet
added function calls in init()
2007-04-10 16:52:53 +05:30
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_count_elementSets (myUnit)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
use prec, only: pInt
use IO
implicit none
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical inPart
mesh_NelemSets = 0_pInt
mesh_maxNelemInSet = mesh_Nelems ! have to be conservative, since Abaqus allows for recursive definitons
added function calls in init()
2007-04-10 16:52:53 +05:30
610 FORMAT(A300)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( (inPart .or. noPart) .and. IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*elset' ) &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_NelemSets = mesh_NelemSets + 1_pInt
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
enddo
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
620 continue
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if (mesh_NelemSets == 0) call IO_error(error_ID=902_pInt)
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endsubroutine
added function calls in init()
2007-04-10 16:52:53 +05:30
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! count overall number of solid sections sets in mesh (Abaqus only)
added function calls in init()
2007-04-10 16:52:53 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! mesh_Nmaterials
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_count_materials (myUnit)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
use prec, only: pInt
use IO
implicit none
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical inPart
mesh_Nmaterials = 0_pInt
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
610 FORMAT(A300)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
if ( (inPart .or. noPart) .and. &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
IO_lc(IO_StringValue(line,myPos,1_pInt)) == '*solid' .and. &
IO_lc(IO_StringValue(line,myPos,2_pInt)) == 'section' ) &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_Nmaterials = mesh_Nmaterials + 1_pInt
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
enddo
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
620 if (mesh_Nmaterials == 0_pInt) call IO_error(error_ID=903_pInt)
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endsubroutine
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
!********************************************************************
! count overall number of cpElements in mesh
!
! mesh_NcpElems
!********************************************************************
subroutine mesh_spectral_count_cpElements ()
use prec, only: pInt
implicit none
mesh_NcpElems = mesh_Nelems
endsubroutine
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! count overall number of cpElements in mesh
added function calls in init()
2007-04-10 16:52:53 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! mesh_NcpElems
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_count_cpElements (myUnit)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
use prec, only: pInt
use IO
implicit none
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 1_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,i
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_NcpElems = 0_pInt
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
610 FORMAT(A300)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'hypoelastic') then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i=1_pInt,3_pInt+hypoelasticTableStyle ! Skip 3 or 4 lines
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
enddo
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
mesh_NcpElems = mesh_NcpElems + IO_countContinousIntValues(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
exit
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
endif
enddo
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
620 endsubroutine
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! count overall number of cpElements in mesh
added function calls in init()
2007-04-10 16:52:53 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! mesh_NcpElems
added function calls in init()
2007-04-10 16:52:53 +05:30
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_count_cpElements (myUnit)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
use prec, only: pInt
use IO
implicit none
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,i,k
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical materialFound
character (len=64) materialName,elemSetName
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_NcpElems = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
select case ( IO_lc(IO_stringValue(line,myPos,1_pInt)) )
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case('*material')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
materialName = trim(IO_extractValue(IO_lc(IO_stringValue(line,myPos,2_pInt)),'name')) ! extract name=value
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
materialFound = materialName /= '' ! valid name?
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case('*user')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if (IO_lc(IO_StringValue(line,myPos,2_pInt)) == 'material' .and. materialFound) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,mesh_Nmaterials ! look thru material names
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
if (materialName == mesh_nameMaterial(i)) then ! found one
elemSetName = mesh_mapMaterial(i) ! take corresponding elemSet
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do k = 1_pInt,mesh_NelemSets ! look thru all elemSet definitions
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
if (elemSetName == mesh_nameElemSet(k)) & ! matched?
mesh_NcpElems = mesh_NcpElems + mesh_mapElemSet(1,k) ! add those elem count
enddo
endif
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
materialFound = .false.
endif
endselect
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
enddo
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
620 if (mesh_NcpElems == 0_pInt) call IO_error(error_ID=906_pInt)
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! map element sets
added function calls in init()
2007-04-10 16:52:53 +05:30
!
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! allocate globals: mesh_nameElemSet, mesh_mapElemSet
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_map_elementSets (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 4_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt), dimension (1+2*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,elemSet
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
allocate (mesh_nameElemSet(mesh_NelemSets)) ; mesh_nameElemSet = ''
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
allocate (mesh_mapElemSet(1_pInt+mesh_maxNelemInSet,mesh_NelemSets)) ; mesh_mapElemSet = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
elemSet = 0_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=640) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if( (IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'define' ) .and. &
(IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'element' ) ) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
elemSet = elemSet+1_pInt
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_nameElemSet(elemSet) = trim(IO_stringValue(line,myPos,4_pInt))
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
mesh_mapElemSet(:,elemSet) = IO_continousIntValues(myUnit,mesh_maxNelemInSet,mesh_nameElemSet,mesh_mapElemSet,mesh_NelemSets)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
enddo
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
640 endsubroutine
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! Build element set mapping
!
! allocate globals: mesh_nameElemSet, mesh_mapElemSet
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_map_elementSets (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
integer(pInt), parameter :: maxNchunks = 4_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,elemSet,i
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical inPart
allocate (mesh_nameElemSet(mesh_NelemSets)) ; mesh_nameElemSet = ''
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
allocate (mesh_mapElemSet(1_pInt+mesh_maxNelemInSet,mesh_NelemSets)) ; mesh_mapElemSet = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
elemSet = 0_pInt
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=640) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( (inPart .or. noPart) .and. IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*elset' ) then
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
elemSet = elemSet + 1_pInt
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_nameElemSet(elemSet) = trim(IO_extractValue(IO_lc(IO_stringValue(line,myPos,2_pInt)),'elset'))
mesh_mapElemSet(:,elemSet) = IO_continousIntValues(myUnit,mesh_Nelems,mesh_nameElemSet,&
mesh_mapElemSet,elemSet-1_pInt)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
enddo
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
640 do i = 1_pInt,elemSet
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
! write(6,*)'elemSetName: ',mesh_nameElemSet(i)
! write(6,*)'elems in Elset',mesh_mapElemSet(:,i)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if (mesh_mapElemSet(1,i) == 0_pInt) call IO_error(error_ID=904_pInt,ext_msg=mesh_nameElemSet(i))
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
enddo
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
!********************************************************************
! map solid section (Abaqus only)
!
! allocate globals: mesh_nameMaterial, mesh_mapMaterial
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_map_materials (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 20_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt) myUnit,i,c
removed unused variables
2011-04-13 19:46:22 +05:30
logical inPart
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=64) elemSetName,materialName
allocate (mesh_nameMaterial(mesh_Nmaterials)) ; mesh_nameMaterial = ''
allocate (mesh_mapMaterial(mesh_Nmaterials)) ; mesh_mapMaterial = ''
610 FORMAT(A300)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
c = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
if ( (inPart .or. noPart) .and. &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
IO_lc(IO_StringValue(line,myPos,1_pInt)) == '*solid' .and. &
IO_lc(IO_StringValue(line,myPos,2_pInt)) == 'section' ) then
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
elemSetName = ''
materialName = ''
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 3_pInt,myPos(1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
if (IO_extractValue(IO_lc(IO_stringValue(line,myPos,i)),'elset') /= '') &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
elemSetName = trim(IO_extractValue(IO_lc(IO_stringValue(line,myPos,i)),'elset'))
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
if (IO_extractValue(IO_lc(IO_stringValue(line,myPos,i)),'material') /= '') &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
materialName = trim(IO_extractValue(IO_lc(IO_stringValue(line,myPos,i)),'material'))
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
if (elemSetName /= '' .and. materialName /= '') then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
c = c + 1_pInt
mesh_nameMaterial(c) = materialName ! name of material used for this section
mesh_mapMaterial(c) = elemSetName ! mapped to respective element set
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
endif
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
620 if (c==0_pInt) call IO_error(error_ID=905_pInt)
do i=1_pInt,c
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
! write(6,*)'name of materials: ',i,mesh_nameMaterial(i)
! write(6,*)'name of elemSets: ',i,mesh_mapMaterial(i)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if (mesh_nameMaterial(i)=='' .or. mesh_mapMaterial(i)=='') call IO_error(error_ID=905_pInt)
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
enddo
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
!********************************************************************
! map nodes from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPnode
!********************************************************************
subroutine mesh_spectral_map_nodes ()
use prec, only: pInt
implicit none
integer(pInt) i
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
allocate (mesh_mapFEtoCPnode(2_pInt,mesh_Nnodes)) ; mesh_mapFEtoCPnode = 0_pInt
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
forall (i = 1_pInt:mesh_Nnodes) &
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
mesh_mapFEtoCPnode(1:2,i) = i
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
endsubroutine
!********************************************************************
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
! map nodes from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPnode
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_map_nodes (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use math, only: qsort
use IO
implicit none
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
integer(pInt), parameter :: maxNchunks = 1_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
integer(pInt), dimension (mesh_Nnodes) :: node_count
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,i
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
allocate (mesh_mapFEtoCPnode(2_pInt,mesh_Nnodes)) ; mesh_mapFEtoCPnode = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
node_count = 0_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=650) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if( IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'coordinates' ) then
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=650) line ! skip crap line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,mesh_Nnodes
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=650) line
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_mapFEtoCPnode(1_pInt,i) = IO_fixedIntValue (line,[ 0_pInt,10_pInt],1_pInt)
mesh_mapFEtoCPnode(2_pInt,i) = i
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
exit
endif
enddo
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
650 call qsort(mesh_mapFEtoCPnode,1_pInt,int(size(mesh_mapFEtoCPnode,2_pInt),pInt))
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
!********************************************************************
! map nodes from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPnode
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_map_nodes (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use math, only: qsort
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt) myUnit,i,c,cpNode
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical inPart
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
allocate (mesh_mapFEtoCPnode(2_pInt,mesh_Nnodes)) ; mesh_mapFEtoCPnode = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
cpNode = 0_pInt
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=650) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
if( (inPart .or. noPart) .and. &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*node' .and. &
( IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'output' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'print' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'file' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'response' ) &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
c = IO_countDataLines(myUnit)
do i = 1_pInt,c
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
backspace(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,c
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=650) line
myPos = IO_stringPos(line,maxNchunks)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
cpNode = cpNode + 1_pInt
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_mapFEtoCPnode(1_pInt,cpNode) = IO_intValue(line,myPos,1_pInt)
mesh_mapFEtoCPnode(2_pInt,cpNode) = cpNode
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
endif
enddo
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
650 call qsort(mesh_mapFEtoCPnode,1_pInt,int(size(mesh_mapFEtoCPnode,2_pInt),pInt))
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if (int(size(mesh_mapFEtoCPnode),pInt) == 0_pInt) call IO_error(error_ID=908_pInt)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
!********************************************************************
! map elements from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPelem
!********************************************************************
subroutine mesh_spectral_map_elements ()
use prec, only: pInt
implicit none
integer(pInt) i
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
allocate (mesh_mapFEtoCPelem(2_pInt,mesh_NcpElems)) ; mesh_mapFEtoCPelem = 0_pInt
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
forall (i = 1_pInt:mesh_NcpElems) &
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
mesh_mapFEtoCPelem(1:2,i) = i
corrected some problems in building mesh
2010-07-01 20:50:39 +05:30
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
endsubroutine
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! map elements from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPelem
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_map_elements (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use math, only: qsort
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 1_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), dimension (1_pInt+mesh_NcpElems) :: contInts
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,i,cpElem
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
allocate (mesh_mapFEtoCPelem(2,mesh_NcpElems)) ; mesh_mapFEtoCPelem = 0_pInt
610 FORMAT(A300)
cpElem = 0_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=660) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if( IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'hypoelastic' ) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i=1_pInt,3_pInt+hypoelasticTableStyle ! skip three (or four if new table style!) lines
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=660) line
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
contInts = IO_continousIntValues(myUnit,mesh_NcpElems,mesh_nameElemSet,mesh_mapElemSet,mesh_NelemSets)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,contInts(1)
cpElem = cpElem+1_pInt
mesh_mapFEtoCPelem(1,cpElem) = contInts(1_pInt+i)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_mapFEtoCPelem(2,cpElem) = cpElem
enddo
endif
enddo
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
660 call qsort(mesh_mapFEtoCPelem,1_pInt,int(size(mesh_mapFEtoCPelem,2_pInt),pInt)) ! should be mesh_NcpElems
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
!********************************************************************
! map elements from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPelem
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_map_elements (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use math, only: qsort
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,i,j,k,cpElem
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical materialFound
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
character (len=64) materialName,elemSetName ! why limited to 64? ABAQUS?
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
allocate (mesh_mapFEtoCPelem(2,mesh_NcpElems)) ; mesh_mapFEtoCPelem = 0_pInt
610 FORMAT(A300)
cpElem = 0_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=660) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
select case ( IO_lc(IO_stringValue(line,myPos,1_pInt)) )
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case('*material')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
materialName = trim(IO_extractValue(IO_lc(IO_stringValue(line,myPos,2_pInt)),'name')) ! extract name=value
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
materialFound = materialName /= '' ! valid name?
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case('*user')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if (IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'material' .and. materialFound) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,mesh_Nmaterials ! look thru material names
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
if (materialName == mesh_nameMaterial(i)) then ! found one
elemSetName = mesh_mapMaterial(i) ! take corresponding elemSet
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do k = 1_pInt,mesh_NelemSets ! look thru all elemSet definitions
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
if (elemSetName == mesh_nameElemSet(k)) then ! matched?
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 1_pInt,mesh_mapElemSet(1,k)
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
cpElem = cpElem + 1_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_mapFEtoCPelem(1,cpElem) = mesh_mapElemSet(1_pInt+j,k) ! store FE id
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
mesh_mapFEtoCPelem(2,cpElem) = cpElem ! store our id
enddo
endif
enddo
endif
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
materialFound = .false.
endif
endselect
enddo
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
660 call qsort(mesh_mapFEtoCPelem,1_pInt,int(size(mesh_mapFEtoCPelem,2_pInt),pInt)) ! should be mesh_NcpElems
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if (int(size(mesh_mapFEtoCPelem),pInt) < 2_pInt) call IO_error(error_ID=907_pInt)
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
!********************************************************************
! get maximum count of nodes, IPs, IP neighbors, and subNodes
! among cpElements
!
! _maxNnodes, _maxNips, _maxNipNeighbors, _maxNsubNodes
!********************************************************************
subroutine mesh_spectral_count_cpSizes ()
use prec, only: pInt
implicit none
integer(pInt) t
t = FE_mapElemtype('C3D8R') ! fake 3D hexahedral 8 node 1 IP element
mesh_maxNnodes = FE_Nnodes(t)
mesh_maxNips = FE_Nips(t)
mesh_maxNipNeighbors = FE_NipNeighbors(t)
mesh_maxNsubNodes = FE_NsubNodes(t)
corrected some problems in building mesh
2010-07-01 20:50:39 +05:30
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
endsubroutine
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! get maximum count of nodes, IPs, IP neighbors, and subNodes
! among cpElements
!
! _maxNnodes, _maxNips, _maxNipNeighbors, _maxNsubNodes
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_count_cpSizes (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,i,t,e
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_maxNnodes = 0_pInt
mesh_maxNips = 0_pInt
mesh_maxNipNeighbors = 0_pInt
mesh_maxNsubNodes = 0_pInt
610 FORMAT(A300)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if( IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'connectivity' ) then
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line ! Garbage line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i=1_pInt,mesh_Nelems ! read all elements
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line
myPos = IO_stringPos(line,maxNchunks) ! limit to id and type
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
e = mesh_FEasCP('elem',IO_intValue(line,myPos,1_pInt))
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (e /= 0_pInt) then
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
t = FE_mapElemtype(IO_stringValue(line,myPos,2_pInt))
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_maxNnodes = max(mesh_maxNnodes,FE_Nnodes(t))
mesh_maxNips = max(mesh_maxNips,FE_Nips(t))
mesh_maxNipNeighbors = max(mesh_maxNipNeighbors,FE_NipNeighbors(t))
mesh_maxNsubNodes = max(mesh_maxNsubNodes,FE_NsubNodes(t))
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
call IO_skipChunks(myUnit,FE_NoriginalNodes(t)-(myPos(1_pInt)-2_pInt)) ! read on if FE_Nnodes exceeds node count present on current line
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
enddo
exit
endif
enddo
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
630 endsubroutine
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! get maximum count of nodes, IPs, IP neighbors, and subNodes
! among cpElements
!
! _maxNnodes, _maxNips, _maxNipNeighbors, _maxNsubNodes
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_count_cpSizes (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 2_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt) myUnit,i,c,t
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical inPart
mesh_maxNnodes = 0_pInt
mesh_maxNips = 0_pInt
mesh_maxNipNeighbors = 0_pInt
mesh_maxNsubNodes = 0_pInt
610 FORMAT(A300)
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
if( (inPart .or. noPart) .and. &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*element' .and. &
( IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'output' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'matrix' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'response' ) &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
) then
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
t = FE_mapElemtype(IO_extractValue(IO_lc(IO_stringValue(line,myPos,2_pInt)),'type')) ! remember elem type
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (t==0_pInt) call IO_error(error_ID=910_pInt,ext_msg='mesh_abaqus_count_cpSizes')
c = IO_countDataLines(myUnit)
do i = 1_pInt,c
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
backspace(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,c
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks) ! limit to 64 nodes max
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (mesh_FEasCP('elem',IO_intValue(line,myPos,1_pInt)) /= 0_pInt) then ! disregard non CP elems
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_maxNnodes = max(mesh_maxNnodes,FE_Nnodes(t))
mesh_maxNips = max(mesh_maxNips,FE_Nips(t))
mesh_maxNipNeighbors = max(mesh_maxNipNeighbors,FE_NipNeighbors(t))
mesh_maxNsubNodes = max(mesh_maxNsubNodes,FE_NsubNodes(t))
endif
enddo
endif
enddo
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
620 endsubroutine
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
!********************************************************************
! store x,y,z coordinates of all nodes in mesh
!
! allocate globals:
! _node
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_spectral_build_nodes (myUnit)
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 7_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) a,b,c,n,i,j,headerLength
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
real(pReal) x,y,z
logical gotResolution,gotDimension
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
character(len=1024) line, keyword
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
allocate ( mesh_node0 (3,mesh_Nnodes) ); mesh_node0 = 0.0_pReal
allocate ( mesh_node (3,mesh_Nnodes) ); mesh_node = 0.0_pReal
corrected some problems in building mesh
2010-07-01 20:50:39 +05:30
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
a = 1_pInt
b = 1_pInt
c = 1_pInt
x = 1.0_pReal
y = 1.0_pReal
z = 1.0_pReal
gotResolution = .false.
gotDimension = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
read(myUnit,'(a1024)') line
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos = IO_stringPos(line,2_pInt)
keyword = IO_lc(IO_StringValue(line,myPos,2_pInt))
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
if (keyword(1:4) == 'head') then
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
headerLength = IO_intValue(line,myPos,1_pInt) + 1_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
else
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
call IO_error(error_ID=842_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
endif
rewind(myUnit)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt, headerLength
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read(myUnit,'(a1024)') line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
select case ( IO_lc(IO_StringValue(line,myPos,1_pInt)) )
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
case ('dimension')
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
gotDimension = .true.
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 2_pInt,6_pInt,2_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
select case (IO_lc(IO_stringValue(line,myPos,j)))
case('x')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
x = IO_floatValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
case('y')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
y = IO_floatValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
case('z')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
z = IO_floatValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
end select
enddo
case ('resolution')
gotResolution = .true.
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 2_pInt,6_pInt,2_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
select case (IO_lc(IO_stringValue(line,myPos,j)))
case('a')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
a = 1_pInt + IO_intValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
case('b')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
b = 1_pInt + IO_intValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
case('c')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
c = 1_pInt + IO_intValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
end select
enddo
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
end select
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
enddo
! --- sanity checks ---
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
if ((.not. gotDimension) .or. (.not. gotResolution)) call IO_error(error_ID=842_pInt)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if ((a < 1_pInt) .or. (b < 1_pInt) .or. (c < 0_pInt)) call IO_error(error_ID=843_pInt) ! 1_pInt is already added
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
if ((x <= 0.0_pReal) .or. (y <= 0.0_pReal) .or. (z <= 0.0_pReal)) call IO_error(error_ID=844_pInt)
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
forall (n = 0_pInt:mesh_Nnodes-1_pInt)
mesh_node0(1,n+1_pInt) = x * real(mod(n,a),pReal) / real(a-1_pInt,pReal)
mesh_node0(2,n+1_pInt) = y * real(mod(n/a,b),pReal) / real(b-1_pInt,pReal)
mesh_node0(3,n+1_pInt) = z * real(mod(n/a/b,c),pReal) / real(c-1_pInt,pReal)
corrected node coordinates and sorting for spectral method
2012-01-17 22:23:56 +05:30
end forall
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
mesh_node = mesh_node0 !why?
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
endsubroutine
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! store x,y,z coordinates of all nodes in mesh
!
! allocate globals:
! _node
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_build_nodes (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), dimension(5), parameter :: node_ends = int([0,10,30,50,70],pInt)
integer(pInt), parameter :: maxNchunks = 1_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,i,j,m
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
allocate ( mesh_node0 (3,mesh_Nnodes) ); mesh_node0 = 0.0_pReal
allocate ( mesh_node (3,mesh_Nnodes) ); mesh_node = 0.0_pReal
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=670) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if( IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'coordinates' ) then
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=670) line ! skip crap line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i=1_pInt,mesh_Nnodes
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=670) line
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
m = mesh_FEasCP('node',IO_fixedIntValue(line,node_ends,1_pInt))
forall (j = 1_pInt:3_pInt) mesh_node0(j,m) = IO_fixedNoEFloatValue(line,node_ends,j+1_pInt)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
exit
endif
enddo
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
670 mesh_node = mesh_node0
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
!********************************************************************
! store x,y,z coordinates of all nodes in mesh
!
! allocate globals:
! _node
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_build_nodes (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 4_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt) myUnit,i,j,m,c
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical inPart
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
allocate ( mesh_node0 (3,mesh_Nnodes) ); mesh_node0 = 0.0_pReal
allocate ( mesh_node (3,mesh_Nnodes) ); mesh_node = 0.0_pReal
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=670) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
if( (inPart .or. noPart) .and. &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*node' .and. &
( IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'output' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'print' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'file' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'response' ) &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
c = IO_countDataLines(myUnit) ! how many nodes are defined here?
do i = 1_pInt,c
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
backspace(myUnit) ! rewind to first entry
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,c
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=670) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
m = mesh_FEasCP('node',IO_intValue(line,myPos,1_pInt))
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
forall (j=1_pInt:3_pInt) mesh_node0(j,m) = IO_floatValue(line,myPos,j+1_pInt)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
endif
enddo
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
670 if (int(size(mesh_node0,2_pInt),pInt) /= mesh_Nnodes) call IO_error(error_ID=909_pInt)
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
mesh_node = mesh_node0
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endsubroutine
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
!********************************************************************
! store FEid, type, mat, tex, and node list per element
!
! allocate globals:
! _element
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_spectral_build_elements (myUnit)
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 7_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
integer(pInt) a,b,c,e,i,j,homog,headerLength,maxIntCount
integer(pInt), dimension(:), allocatable :: microstructures
integer(pInt), dimension(1,1) :: dummySet = 0_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
character(len=65536) line,keyword
character(len=64), dimension(1) :: dummyName = ''
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
a = 1_pInt
b = 1_pInt
c = 1_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
read(myUnit,'(a65536)') line
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos = IO_stringPos(line,2_pInt)
keyword = IO_lc(IO_StringValue(line,myPos,2_pInt))
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
if (keyword(1:4) == 'head') then
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
headerLength = IO_intValue(line,myPos,1_pInt) + 1_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
else
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
call IO_error(error_ID=842_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
endif
rewind(myUnit)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt, headerLength
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
read(myUnit,'(a65536)') line
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
select case ( IO_lc(IO_StringValue(line,myPos,1_pInt)) )
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
case ('resolution')
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 2_pInt,6_pInt,2_pInt
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
select case (IO_lc(IO_stringValue(line,myPos,j)))
case('a')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
a = 1_pInt + IO_intValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
case('b')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
b = 1_pInt + IO_intValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
case('c')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
c = 1_pInt + IO_intValue(line,myPos,j+1_pInt)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
end select
enddo
case ('homogenization')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
homog = IO_intValue(line,myPos,2_pInt)
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
end select
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
enddo
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
maxIntCount = 0_pInt
i = 1_pInt
do while (i > 0_pInt)
i = IO_countContinousIntValues(myUnit)
maxIntCount = max(maxIntCount, i)
enddo
rewind (myUnit)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i=1_pInt,headerLength ! skip header
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
read(myUnit,'(a65536)') line
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
100 allocate (mesh_element (4_pInt+mesh_maxNnodes,mesh_NcpElems)) ; mesh_element = 0_pInt
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
allocate (microstructures (1_pInt+maxIntCount)) ; microstructures = 2_pInt
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
e = 0_pInt
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
do while (e < mesh_NcpElems .and. microstructures(1) > 0_pInt) ! fill expected number of elements, stop at end of data (or blank line!)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
microstructures = IO_continousIntValues(myUnit,maxIntCount,dummyName,dummySet,0_pInt) ! get affected elements
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,microstructures(1_pInt)
e = e+1_pInt ! valid element entry
corrected node coordinates and sorting for spectral method
2012-01-17 22:23:56 +05:30
mesh_element( 1,e) = e ! FE id
mesh_element( 2,e) = FE_mapElemtype('C3D8R') ! elem type
mesh_element( 3,e) = homog ! homogenization
mesh_element( 4,e) = microstructures(1_pInt+i) ! microstructure
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_element( 5,e) = e + (e-1_pInt)/(a-1_pInt) + ((e-1_pInt)/((a-1_pInt)*(b-1_pInt)))*a ! base node
mesh_element( 6,e) = mesh_element(5,e) + 1_pInt
mesh_element( 7,e) = mesh_element(5,e) + a + 1_pInt
corrected node coordinates and sorting for spectral method
2012-01-17 22:23:56 +05:30
mesh_element( 8,e) = mesh_element(5,e) + a
mesh_element( 9,e) = mesh_element(5,e) + a * b ! second floor base node
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_element(10,e) = mesh_element(9,e) + 1_pInt
mesh_element(11,e) = mesh_element(9,e) + a + 1_pInt
corrected node coordinates and sorting for spectral method
2012-01-17 22:23:56 +05:30
mesh_element(12,e) = mesh_element(9,e) + a
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
mesh_maxValStateVar(1) = max(mesh_maxValStateVar(1),mesh_element(3,e)) !needed for statistics
mesh_maxValStateVar(2) = max(mesh_maxValStateVar(2),mesh_element(4,e))
enddo
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
enddo
corrected some problems in building mesh
2010-07-01 20:50:39 +05:30
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
110 deallocate(microstructures)
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
if (e /= mesh_NcpElems) call IO_error(880_pInt,e)
parsing of spectral geom file now accepts two-dimensional microstructure definition (plus range indications by 'x to y') For instance: 1 2 3 4 5 6 7 8 9 10 11 12 ...
2012-01-12 22:31:24 +05:30
endsubroutine
new errors regarding Spectral method input file parsing mesh setup from Spectral input file $Id$ in FEsolving activated
2010-05-06 22:10:47 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! store FEid, type, mat, tex, and node list per element
!
! allocate globals:
! _element
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_marc_build_elements (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 66_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
character(len=300) line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), dimension(1_pInt+mesh_NcpElems) :: contInts
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
integer(pInt) myUnit,i,j,sv,val,e
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
allocate (mesh_element (4_pInt+mesh_maxNnodes,mesh_NcpElems)) ; mesh_element = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos(1:1+2*1) = IO_stringPos(line,1_pInt)
if( IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'connectivity' ) then
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line ! Garbage line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,mesh_Nelems
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks) ! limit to 64 nodes max (plus ID, type)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
e = mesh_FEasCP('elem',IO_intValue(line,myPos,1_pInt))
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (e /= 0_pInt) then ! disregard non CP elems
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_element(1,e) = IO_IntValue (line,myPos,1_pInt) ! FE id
mesh_element(2,e) = FE_mapElemtype(IO_StringValue(line,myPos,2_pInt)) ! elem type
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
forall (j = 1_pInt:FE_Nnodes(mesh_element(2,e))) &
mesh_element(j+4_pInt,e) = IO_IntValue(line,myPos,j+2_pInt) ! copy FE ids of nodes
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
call IO_skipChunks(myUnit,FE_NoriginalNodes(mesh_element(2_pInt,e))-(myPos(1_pInt)-2_pInt)) ! read on if FE_Nnodes exceeds node count present on current line
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
enddo
exit
endif
enddo
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
620 rewind(myUnit) ! just in case "initial state" apears before "connectivity"
read (myUnit,610,END=620) line
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos(1:1+2*2) = IO_stringPos(line,2_pInt)
if( (IO_lc(IO_stringValue(line,myPos,1_pInt)) == 'initial') .and. &
(IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'state') ) then
if (initialcondTableStyle == 2_pInt) read (myUnit,610,END=620) line ! read extra line for new style
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line ! read line with index of state var
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos(1:1+2*1) = IO_stringPos(line,1_pInt)
sv = IO_IntValue(line,myPos,1_pInt) ! figure state variable index
if( (sv == 2_pInt).or.(sv == 3_pInt) ) then ! only state vars 2 and 3 of interest
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line ! read line with value of state var
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos(1:1+2*1) = IO_stringPos(line,1_pInt)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do while (scan(IO_stringValue(line,myPos,1_pInt),'+-',back=.true.)>1) ! is noEfloat value?
val = nint(IO_fixedNoEFloatValue(line,[0_pInt,20_pInt],1_pInt),pInt) ! state var's value
mesh_maxValStateVar(sv-1_pInt) = max(val,mesh_maxValStateVar(sv-1_pInt)) ! remember max val of homogenization and microstructure index
if (initialcondTableStyle == 2_pInt) then
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line ! read extra line
read (myUnit,610,END=630) line ! read extra line
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
contInts = IO_continousIntValues(myUnit,mesh_Nelems,mesh_nameElemSet,mesh_mapElemSet,mesh_NelemSets) ! get affected elements
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,contInts(1)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
e = mesh_FEasCP('elem',contInts(1_pInt+i))
mesh_element(1_pInt+sv,e) = val
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (initialcondTableStyle == 0_pInt) read (myUnit,610,END=620) line ! ignore IP range for old table style
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos(1:1+2*1) = IO_stringPos(line,1_pInt)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
endif
else
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
enddo
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
630 endsubroutine
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! store FEid, type, mat, tex, and node list per element
!
! allocate globals:
! _element
!********************************************************************
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
subroutine mesh_abaqus_build_elements (myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
use prec, only: pInt
use IO
implicit none
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: maxNchunks = 65_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt) myUnit,i,j,k,c,e,t,homog,micro
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
logical inPart,materialFound
character (len=64) materialName,elemSetName
character(len=300) line
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
allocate (mesh_element (4_pInt+mesh_maxNnodes,mesh_NcpElems)) ; mesh_element = 0_pInt
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
610 FORMAT(A300)
inPart = .false.
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
rewind(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos(1:1+2*2) = IO_stringPos(line,2_pInt)
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*part' ) inPart = .true.
if ( IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*end' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) == 'part' ) inPart = .false.
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
adopted improvements done by Arun Prakash. mesh: elemType identification based on lower case Abaqus now reports more errors IO: new function to inquire whether inputfile contains "parts" new function to assemble multiply included inputfile into a flat one awareness of range generation in element numbers error reporting
2010-07-13 15:56:07 +05:30
if( (inPart .or. noPart) .and. &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
IO_lc(IO_stringValue(line,myPos,1_pInt)) == '*element' .and. &
( IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'output' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'matrix' .and. &
IO_lc(IO_stringValue(line,myPos,2_pInt)) /= 'response' ) &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
) then
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
t = FE_mapElemtype(IO_extractValue(IO_lc(IO_stringValue(line,myPos,2_pInt)),'type')) ! remember elem type
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (t==0_pInt) call IO_error(error_ID=910_pInt,ext_msg='mesh_abaqus_build_elements')
c = IO_countDataLines(myUnit)
do i = 1_pInt,c
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
backspace(myUnit)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,c
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=620) line
myPos = IO_stringPos(line,maxNchunks) ! limit to 64 nodes max
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
e = mesh_FEasCP('elem',IO_intValue(line,myPos,1_pInt))
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (e /= 0_pInt) then ! disregard non CP elems
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_element(1,e) = IO_intValue(line,myPos,1_pInt) ! FE id
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_element(2,e) = t ! elem type
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
forall (j=1_pInt:FE_Nnodes(t)) &
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_element(4_pInt+j,e) = IO_intValue(line,myPos,1_pInt+j) ! copy FE ids of nodes to position 5:
call IO_skipChunks(myUnit,FE_NoriginalNodes(t)-(myPos(1_pInt)-1_pInt)) ! read on (even multiple lines) if FE_NoriginalNodes exceeds required node count
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
endif
enddo
endif
enddo
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
620 rewind(myUnit) ! just in case "*material" definitions apear before "*element"
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
materialFound = .false.
do
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line
myPos = IO_stringPos(line,maxNchunks)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
select case ( IO_lc(IO_StringValue(line,myPos,1_pInt)))
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case('*material')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
materialName = trim(IO_extractValue(IO_lc(IO_StringValue(line,myPos,2_pInt)),'name')) ! extract name=value
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
materialFound = materialName /= '' ! valid name?
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
case('*user')
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
if ( IO_lc(IO_StringValue(line,myPos,2_pInt)) == 'material' .and. &
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
materialFound ) then
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
read (myUnit,610,END=630) line ! read homogenization and microstructure
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
myPos(1:1+2*2) = IO_stringPos(line,2_pInt)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
homog = nint(IO_floatValue(line,myPos,1_pInt),pInt)
micro = nint(IO_floatValue(line,myPos,2_pInt),pInt)
do i = 1_pInt,mesh_Nmaterials ! look thru material names
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
if (materialName == mesh_nameMaterial(i)) then ! found one
elemSetName = mesh_mapMaterial(i) ! take corresponding elemSet
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do k = 1_pInt,mesh_NelemSets ! look thru all elemSet definitions
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
if (elemSetName == mesh_nameElemSet(k)) then ! matched?
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 1_pInt,mesh_mapElemSet(1,k)
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
e = mesh_FEasCP('elem',mesh_mapElemSet(1+j,k))
mesh_element(3,e) = homog ! store homogenization
mesh_element(4,e) = micro ! store microstructure
mesh_maxValStateVar(1) = max(mesh_maxValStateVar(1),homog)
mesh_maxValStateVar(2) = max(mesh_maxValStateVar(2),micro)
enddo
endif
enddo
endif
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
enddo
materialFound = .false.
endif
endselect
enddo
did a lot of polishing: - removed unnecessary "return" before end of subroutine or function: - changed undetermined array length (:) to (1:3) To prevent problems with some code analysing tools: - "3D oneliner loops" (with ";) only for "do" and "enddo" at the same time - removed line continuation in OMP statements made the makefile more flexible, removed heap-arrays switch
2011-09-13 21:24:06 +05:30
630 endsubroutine
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!********************************************************************
! get maximum count of shared elements among cpElements and
! build list of elements shared by each node in mesh
!
! _maxNsharedElems
! _sharedElem
!********************************************************************
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
subroutine mesh_build_sharedElems()
use prec, only: pInt
implicit none
integer(pint) e, & ! element index
t, & ! element type
node, & ! CP node index
j, & ! node index per element
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
myDim, & ! dimension index
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
nodeTwin ! node twin in the specified dimension
integer(pInt), dimension (mesh_Nnodes) :: node_count
integer(pInt), dimension (:), allocatable :: node_seen
allocate(node_seen(maxval(FE_Nnodes)))
mpie_cpfem_abaqus_std.f: first call is with kinc==1 mesh: allow multiple element sets with identical material
2011-03-23 21:50:12 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
node_count = 0_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
t = mesh_element(2,e) ! get element type
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
node_seen = 0_pInt ! reset node duplicates
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 1_pInt,FE_Nnodes(t) ! check each node of element
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
node = mesh_FEasCP('node',mesh_element(4+j,e)) ! translate to internal (consecutive) numbering
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
if (all(node_seen /= node)) then
node_count(node) = node_count(node) + 1_pInt ! if FE node not yet encountered -> count it
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do myDim = 1_pInt,3_pInt ! check in each dimension...
nodeTwin = mesh_nodeTwins(myDim,node)
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
if (nodeTwin > 0_pInt) & ! if I am a twin of some node...
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
node_count(nodeTwin) = node_count(nodeTwin) + 1_pInt ! -> count me again for the twin node
enddo
endif
node_seen(j) = node ! remember this node to be counted already
enddo
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_maxNsharedElems = int(maxval(node_count),pInt) ! most shared node
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
allocate(mesh_sharedElem(1+mesh_maxNsharedElems,mesh_Nnodes))
mesh_sharedElem = 0_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
t = mesh_element(2,e)
node_seen = 0_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 1_pInt,FE_Nnodes(t)
node = mesh_FEasCP('node',mesh_element(4_pInt+j,e))
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
if (all(node_seen /= node)) then
mesh_sharedElem(1,node) = mesh_sharedElem(1,node) + 1_pInt ! count for each node the connected elements
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_sharedElem(mesh_sharedElem(1,node)+1_pInt,node) = e ! store the respective element id
do myDim = 1_pInt,3_pInt ! check in each dimension...
nodeTwin = mesh_nodeTwins(myDim,node)
if (nodeTwin > 0_pInt) then ! if i am a twin of some node...
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
mesh_sharedElem(1,nodeTwin) = mesh_sharedElem(1,nodeTwin) + 1_pInt ! ...count me again for the twin
mesh_sharedElem(mesh_sharedElem(1,nodeTwin)+1,nodeTwin) = e ! store the respective element id
endif
enddo
endif
node_seen(j) = node
enddo
enddo
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
deallocate(node_seen)
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
endsubroutine
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
!***********************************************************
! build up of IP neighborhood
!
! allocate globals
! _ipNeighborhood
!***********************************************************
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
subroutine mesh_build_ipNeighborhood()
use prec, only: pInt
implicit none
integer(pInt) myElem, & ! my CP element index
myIP, &
myType, & ! my element type
myFace, &
neighbor, & ! neighor index
neighboringIPkey, & ! positive integer indicating the neighboring IP (for intra-element) and negative integer indicating the face towards neighbor (for neighboring element)
candidateIP, &
neighboringType, & ! element type of neighbor
NlinkedNodes, & ! number of linked nodes
twin_of_linkedNode, & ! node twin of a specific linkedNode
NmatchingNodes, & ! number of matching nodes
dir, & ! direction of periodicity
matchingElem, & ! CP elem number of matching element
matchingFace, & ! face ID of matching element
removed unused variables
2011-04-13 19:46:22 +05:30
a, anchor
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
integer(pInt), dimension(FE_maxmaxNnodesAtIP) :: &
linkedNodes, &
matchingNodes
logical checkTwins
allocate(mesh_ipNeighborhood(2,mesh_maxNipNeighbors,mesh_maxNips,mesh_NcpElems))
mesh_ipNeighborhood = 0_pInt
linkedNodes = 0_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do myElem = 1_pInt,mesh_NcpElems ! loop over cpElems
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
myType = mesh_element(2,myElem) ! get elemType
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do myIP = 1_pInt,FE_Nips(myType) ! loop over IPs of elem
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do neighbor = 1_pInt,FE_NipNeighbors(myType) ! loop over neighbors of IP
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
neighboringIPkey = FE_ipNeighbor(neighbor,myIP,myType)
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
!*** if the key is positive, the neighbor is inside the element
!*** that means, we have already found our neighboring IP
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (neighboringIPkey > 0_pInt) then
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
mesh_ipNeighborhood(1,neighbor,myIP,myElem) = myElem
mesh_ipNeighborhood(2,neighbor,myIP,myElem) = neighboringIPkey
!*** if the key is negative, the neighbor resides in a neighboring element
!*** that means, we have to look through the face indicated by the key and see which element is behind that face
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
elseif (neighboringIPkey < 0_pInt) then ! neighboring element's IP
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
myFace = -neighboringIPkey
call mesh_faceMatch(myElem, myFace, matchingElem, matchingFace) ! get face and CP elem id of face match
if (matchingElem > 0_pInt) then ! found match?
neighboringType = mesh_element(2,matchingElem)
!*** trivial solution if neighbor has only one IP
if (FE_Nips(neighboringType) == 1_pInt) then
mesh_ipNeighborhood(1,neighbor,myIP,myElem) = matchingElem
mesh_ipNeighborhood(2,neighbor,myIP,myElem) = 1_pInt
cycle
endif
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
!*** find those nodes which build the link to the neighbor
NlinkedNodes = 0_pInt
linkedNodes = 0_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do a = 1_pInt,FE_maxNnodesAtIP(myType) ! figure my anchor nodes on connecting face
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
anchor = FE_nodesAtIP(a,myIP,myType)
if (anchor /= 0_pInt) then ! valid anchor node
if (any(FE_nodeOnFace(:,myFace,myType) == anchor)) then ! ip anchor sits on face?
NlinkedNodes = NlinkedNodes + 1_pInt
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
linkedNodes(NlinkedNodes) = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_FEasCP('node',mesh_element(4_pInt+anchor,myElem)) ! CP id of anchor node
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
else ! something went wrong with the linkage, since not all anchors sit on my face
NlinkedNodes = 0_pInt
linkedNodes = 0_pInt
exit
endif
endif
enddo
!*** loop through the ips of my neighbor
!*** and try to find an ip with matching nodes
!*** also try to match with node twins
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
checkCandidateIP: do candidateIP = 1_pInt,FE_Nips(neighboringType)
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
NmatchingNodes = 0_pInt
matchingNodes = 0_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do a = 1_pInt,FE_maxNnodesAtIP(neighboringType) ! check each anchor node of that ip
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
anchor = FE_nodesAtIP(a,candidateIP,neighboringType)
if (anchor /= 0_pInt) then ! valid anchor node
if (any(FE_nodeOnFace(:,matchingFace,neighboringType) == anchor)) then ! sits on matching face?
NmatchingNodes = NmatchingNodes + 1_pInt
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
matchingNodes(NmatchingNodes) = &
mesh_FEasCP('node',mesh_element(4+anchor,matchingElem)) ! CP id of neighbor's anchor node
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
else ! no matching, because not all nodes sit on the matching face
NmatchingNodes = 0_pInt
matchingNodes = 0_pInt
exit
endif
endif
enddo
if (NmatchingNodes /= NlinkedNodes) & ! this ip has wrong count of anchors on face
cycle checkCandidateIP
!*** check "normal" nodes whether they match or not
checkTwins = .false.
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do a = 1_pInt,NlinkedNodes
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
if (all(matchingNodes /= linkedNodes(a))) then ! this linkedNode does not match any matchingNode
checkTwins = .true.
exit ! no need to search further
endif
enddo
!*** if no match found, then also check node twins
if(checkTwins) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
dir = int(maxloc(abs(mesh_ipAreaNormal(1:3,neighbor,myIP,myElem)),1),pInt) ! check for twins only in direction of the surface normal
do a = 1_pInt,NlinkedNodes
check for twin nodes only in direction of the surface normal. this solves a problem with models of single element thickness and periodic fluxes.
2011-02-24 14:56:30 +05:30
twin_of_linkedNode = mesh_nodeTwins(dir,linkedNodes(a))
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
if (twin_of_linkedNode == 0_pInt .or. & ! twin of linkedNode does not exist...
all(matchingNodes /= twin_of_linkedNode)) then ! ... or it does not match any matchingNode
cycle checkCandidateIP ! ... then check next candidateIP
check for twin nodes only in direction of the surface normal. this solves a problem with models of single element thickness and periodic fluxes.
2011-02-24 14:56:30 +05:30
endif
enddo
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
endif
!*** we found a match !!!
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
mesh_ipNeighborhood(1,neighbor,myIP,myElem) = matchingElem
mesh_ipNeighborhood(2,neighbor,myIP,myElem) = candidateIP
check for twin nodes only in direction of the surface normal. this solves a problem with models of single element thickness and periodic fluxes.
2011-02-24 14:56:30 +05:30
exit checkCandidateIP
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo checkCandidateIP
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
endif ! end of valid external matching
endif ! end of internal/external matching
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
enddo
enddo
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
endsubroutine
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added full description of element type 2
2009-01-16 23:06:37 +05:30
!***********************************************************
! assignment of coordinates for subnodes in each cp element
!
! allocate globals
! _subNodeCoord
!***********************************************************
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
subroutine mesh_build_subNodeCoords()
added full description of element type 2
2009-01-16 23:06:37 +05:30
use prec, only: pInt,pReal
implicit none
integer(pInt) e,t,n,p
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
if (.not. allocated(mesh_subNodeCoord)) then
allocate(mesh_subNodeCoord(3,mesh_maxNnodes+mesh_maxNsubNodes,mesh_NcpElems))
endif
mesh_subNodeCoord = 0.0_pReal
added full description of element type 2
2009-01-16 23:06:37 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
added full description of element type 2
2009-01-16 23:06:37 +05:30
t = mesh_element(2,e) ! get elemType
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do n = 1_pInt,FE_Nnodes(t)
mesh_subNodeCoord(1:3,n,e) = mesh_node(1:3,mesh_FEasCP('node',mesh_element(4_pInt+n,e))) ! loop over nodes of this element type
added full description of element type 2
2009-01-16 23:06:37 +05:30
enddo
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do n = 1_pInt,FE_NsubNodes(t) ! now for the true subnodes
do p = 1_pInt,FE_Nips(t) ! loop through possible parent nodes
if (FE_subNodeParent(p,n,t) > 0_pInt) & ! valid parent node
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
mesh_subNodeCoord(1:3,FE_Nnodes(t)+n,e) = mesh_subNodeCoord(1:3,FE_Nnodes(t)+n,e) &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
+ mesh_node(1:3,mesh_FEasCP('node',mesh_element(4_pInt+FE_subNodeParent(p,n,t),e))) ! add up parents
Element type 21 (20-node hexagon with triquadric interpolation functions) is now included. 'mesh_get_nodeElemDimensions', mesh_build_elements', 'mesh_build_sharedElems' now allow for more than 14 nodes per elements which are then written in more than one line in the 'connectivity' field of the input file
2009-04-02 18:30:51 +05:30
enddo
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
mesh_subNodeCoord(1:3,n+FE_Nnodes(t),e) = mesh_subNodeCoord(1:3,n+FE_Nnodes(t),e) &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
/real(count(FE_subNodeParent(:,n,t) > 0_pInt),pReal)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
enddo
added full description of element type 2
2009-01-16 23:06:37 +05:30
enddo
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endsubroutine
added full description of element type 2
2009-01-16 23:06:37 +05:30
!***********************************************************
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
! calculation of IP coordinates
added full description of element type 2
2009-01-16 23:06:37 +05:30
!
! allocate globals
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
! _ipCenterOfGravity
added full description of element type 2
2009-01-16 23:06:37 +05:30
!***********************************************************
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
subroutine mesh_build_ipCoordinates()
added full description of element type 2
2009-01-16 23:06:37 +05:30
now deleted old _single files
2011-02-25 18:19:18 +05:30
use prec, only: pInt, tol_gravityNodePos
added full description of element type 2
2009-01-16 23:06:37 +05:30
implicit none
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
integer(pInt) e,f,t,i,j,k,n
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
logical, dimension(mesh_maxNnodes+mesh_maxNsubNodes) :: gravityNode ! flagList to find subnodes determining center of grav
added full description of element type 2
2009-01-16 23:06:37 +05:30
real(pReal), dimension(3,mesh_maxNnodes+mesh_maxNsubNodes) :: gravityNodePos ! coordinates of subnodes determining center of grav
real(pReal), dimension(3) :: centerOfGravity
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
if (.not. allocated(mesh_ipCenterOfGravity)) then
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
allocate(mesh_ipCenterOfGravity(3,mesh_maxNips,mesh_NcpElems))
endif
now with first draft of nonlocal constitutive law debugging memory leak closed debugging counters corrected center of gravity stored in mesh state updated is now split into a collecting loop and an execution updateState and updateTemperature fill sequentially separate logicals and evaluate afterwards to converged added 3x3 transposition function, norm for 3x1 matrix and 33x3 matrix multiplication in math non-converged crystallite triggers materialpoint cutback (used to respond elastically) non-converged materialpoint raises terminal illness which in turn renders whole FE increment useless by means of odd stress/stiffness and thus waits for FE cutback
2009-08-11 22:01:57 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
t = mesh_element(2,e) ! get elemType
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,FE_Nips(t) ! loop over IPs of elem
corrected small mistake in calculation for ip center of gravity / ip coordinates; used to ignore subnodes that had coordinates (0.0, 0.0, 0.0)
2009-12-22 18:53:20 +05:30
gravityNode = .false. ! reset flagList
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
gravityNodePos = 0.0_pReal ! reset coordinates
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do f = 1_pInt,FE_NipNeighbors(t) ! loop over interfaces of IP
do n = 1_pInt,FE_NipFaceNodes ! loop over nodes on interface
corrected small mistake in calculation for ip center of gravity / ip coordinates; used to ignore subnodes that had coordinates (0.0, 0.0, 0.0)
2009-12-22 18:53:20 +05:30
gravityNode(FE_subNodeOnIPFace(n,f,i,t)) = .true.
added full description of element type 2
2009-01-16 23:06:37 +05:30
gravityNodePos(:,FE_subNodeOnIPFace(n,f,i,t)) = mesh_subNodeCoord(:,FE_subNodeOnIPFace(n,f,i,t),e)
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
enddo
enddo
corrected small mistake in calculation for ip center of gravity / ip coordinates; used to ignore subnodes that had coordinates (0.0, 0.0, 0.0)
2009-12-22 18:53:20 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do j = 1_pInt,mesh_maxNnodes+mesh_maxNsubNodes-1_pInt ! walk through entire flagList except last
corrected small mistake in calculation for ip center of gravity / ip coordinates; used to ignore subnodes that had coordinates (0.0, 0.0, 0.0)
2009-12-22 18:53:20 +05:30
if (gravityNode(j)) then ! valid node index
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do k = j+1_pInt,mesh_maxNnodes+mesh_maxNsubNodes ! walk through remainder of list
now deleted old _single files
2011-02-25 18:19:18 +05:30
if (gravityNode(k) .and. all(abs(gravityNodePos(:,j) - gravityNodePos(:,k)) < tol_gravityNodePos)) then ! found duplicate
corrected small mistake in calculation for ip center of gravity / ip coordinates; used to ignore subnodes that had coordinates (0.0, 0.0, 0.0)
2009-12-22 18:53:20 +05:30
gravityNode(j) = .false. ! delete first instance
Element type 21 (20-node hexagon with triquadric interpolation functions) is now included. 'mesh_get_nodeElemDimensions', mesh_build_elements', 'mesh_build_sharedElems' now allow for more than 14 nodes per elements which are then written in more than one line in the 'connectivity' field of the input file
2009-04-02 18:30:51 +05:30
gravityNodePos(:,j) = 0.0_pReal
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
exit ! continue with next suspect
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endif
enddo
endif
enddo
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
centerOfGravity = sum(gravityNodePos,2)/real(count(gravityNode),pReal)
now with first draft of nonlocal constitutive law debugging memory leak closed debugging counters corrected center of gravity stored in mesh state updated is now split into a collecting loop and an execution updateState and updateTemperature fill sequentially separate logicals and evaluate afterwards to converged added 3x3 transposition function, norm for 3x1 matrix and 33x3 matrix multiplication in math non-converged crystallite triggers materialpoint cutback (used to respond elastically) non-converged materialpoint raises terminal illness which in turn renders whole FE increment useless by means of odd stress/stiffness and thus waits for FE cutback
2009-08-11 22:01:57 +05:30
mesh_ipCenterOfGravity(:,i,e) = centerOfGravity
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
enddo
enddo
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
endsubroutine
!***********************************************************
! calculation of IP volume
!
! allocate globals
! _ipVolume
!***********************************************************
subroutine mesh_build_ipVolumes()
use prec, only: pInt, tol_gravityNodePos
use math, only: math_volTetrahedron
implicit none
added full description of element type 2
2009-01-16 23:06:37 +05:30
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
integer(pInt) e,f,t,i,j,n
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: Ntriangles = FE_NipFaceNodes-2_pInt ! each interface is made up of this many triangles
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
real(pReal), dimension(3,FE_NipFaceNodes) :: nPos ! coordinates of nodes on IP face
real(pReal), dimension(Ntriangles,FE_NipFaceNodes) :: volume ! volumes of possible tetrahedra
if (.not. allocated(mesh_ipVolume)) then
allocate(mesh_ipVolume(mesh_maxNips,mesh_NcpElems))
endif
mesh_ipVolume = 0.0_pReal
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
t = mesh_element(2_pInt,e) ! get elemType
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,FE_Nips(t) ! loop over IPs of elem
do f = 1_pInt,FE_NipNeighbors(t) ! loop over interfaces of IP and add tetrahedra which connect to CoG
forall (n = 1_pInt:FE_NipFaceNodes) &
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
nPos(:,n) = mesh_subNodeCoord(:,FE_subNodeOnIPFace(n,f,i,t),e)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
forall (n = 1_pInt:FE_NipFaceNodes, j = 1_pInt:Ntriangles) & ! start at each interface node and build valid triangles to cover interface
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
volume(j,n) = math_volTetrahedron(nPos(:,n), & ! calc volume of respective tetrahedron to CoG
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
nPos(:,1_pInt+mod(n-1_pInt +j ,FE_NipFaceNodes)), & ! start at offset j
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
nPos(:,1_pInt+mod(n-1_pInt +j+1_pInt,FE_NipFaceNodes)), & ! and take j's neighbor
Calculation of current ip volumes now working. Crystallite output also reflects current grain volume, not reference volume. However, this is only available for Marc. Abaqus and spectral method still return the reference ip volume. The ip coordinates though are correctly updated for all solver types.
2011-08-10 22:07:17 +05:30
mesh_ipCenterOfGravity(:,i,e))
mesh_ipVolume(i,e) = mesh_ipVolume(i,e) + sum(volume) ! add contribution from this interface
enddo
mesh_ipVolume(i,e) = mesh_ipVolume(i,e) / FE_NipFaceNodes ! renormalize with interfaceNodeNum due to loop over them
enddo
enddo
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endsubroutine
added full description of element type 2
2009-01-16 23:06:37 +05:30
!***********************************************************
overhaul to allow for multiphase/homogenization setup.
2009-03-04 17:18:54 +05:30
! calculation of IP interface areas
added full description of element type 2
2009-01-16 23:06:37 +05:30
!
! allocate globals
overhaul to allow for multiphase/homogenization setup.
2009-03-04 17:18:54 +05:30
! _ipArea, _ipAreaNormal
added full description of element type 2
2009-01-16 23:06:37 +05:30
!***********************************************************
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
subroutine mesh_build_ipAreas()
added full description of element type 2
2009-01-16 23:06:37 +05:30
use prec, only: pInt,pReal
use math
implicit none
removed some unused variables correct some spelling errors in material.config adopted marc2008 interface routine removed sequential interface routines as they are not needed any more
2009-03-05 19:48:50 +05:30
integer(pInt) e,f,t,i,j,n
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
integer(pInt), parameter :: Ntriangles = FE_NipFaceNodes-2_pInt ! each interface is made up of this many triangles
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
real(pReal), dimension (3,FE_NipFaceNodes) :: nPos ! coordinates of nodes on IP face
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
real(pReal), dimension(3,Ntriangles,FE_NipFaceNodes) :: normal
real(pReal), dimension(Ntriangles,FE_NipFaceNodes) :: area
added full description of element type 2
2009-01-16 23:06:37 +05:30
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
allocate(mesh_ipArea(mesh_maxNipNeighbors,mesh_maxNips,mesh_NcpElems)) ; mesh_ipArea = 0.0_pReal
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
allocate(mesh_ipAreaNormal(3_pInt,mesh_maxNipNeighbors,mesh_maxNips,mesh_NcpElems)) ; mesh_ipAreaNormal = 0.0_pReal
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
t = mesh_element(2,e) ! get elemType
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,FE_Nips(t) ! loop over IPs of elem
do f = 1_pInt,FE_NipNeighbors(t) ! loop over interfaces of IP
forall (n = 1_pInt:FE_NipFaceNodes) nPos(:,n) = mesh_subNodeCoord(:,FE_subNodeOnIPFace(n,f,i,t),e)
forall (n = 1_pInt:FE_NipFaceNodes, j = 1_pInt:Ntriangles) ! start at each interface node and build valid triangles to cover interface
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
normal(:,j,n) = math_vectorproduct(nPos(:,1_pInt+mod(n+j-1_pInt,FE_NipFaceNodes)) - nPos(:,n), & ! calc their normal vectors
nPos(:,1_pInt+mod(n+j-0_pInt,FE_NipFaceNodes)) - nPos(:,n))
removed all math functions only for double precision by the more flexible counterpart, e.g. "dsqrt --> sqrt", "dsin --> sin". Should not cause any harm, as long as "implicit none" is used. Now it is possible to compile a single precision spectral solver/crystal plasticity by replacing mesh.f90 and prec.f90 with mesh_single.f90 and prec_single.f90. For the spectral method, just call "make precision=single" instead of "make". Use "make clean" evertime you switch precision
2011-02-25 14:55:53 +05:30
area(j,n) = sqrt(sum(normal(:,j,n)*normal(:,j,n))) ! and area
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
end forall
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
forall (n = 1_pInt:FE_NipFaceNodes, j = 1_pInt:Ntriangles, area(j,n) > 0.0_pReal) &
added 'spectralPictureMode' for reading in geometries with 1to1 mapping (e.g. EBSD pattern) polishing !incremental update, wait for commit of damask_spectral.f90 before checking out
2011-10-18 14:50:29 +05:30
normal(1:3,j,n) = normal(1:3,j,n) / area(j,n) ! make myUnit normal
now with first draft of nonlocal constitutive law debugging memory leak closed debugging counters corrected center of gravity stored in mesh state updated is now split into a collecting loop and an execution updateState and updateTemperature fill sequentially separate logicals and evaluate afterwards to converged added 3x3 transposition function, norm for 3x1 matrix and 33x3 matrix multiplication in math non-converged crystallite triggers materialpoint cutback (used to respond elastically) non-converged materialpoint raises terminal illness which in turn renders whole FE increment useless by means of odd stress/stiffness and thus waits for FE cutback
2009-08-11 22:01:57 +05:30
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_ipArea(f,i,e) = sum(area) / (FE_NipFaceNodes*2.0_pReal) ! area of parallelograms instead of triangles
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_ipAreaNormal(:,f,i,e) = sum(sum(normal,3),2_pInt)/& ! average of all valid normals
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
real(count(area > 0.0_pReal),pReal)
added full description of element type 2
2009-01-16 23:06:37 +05:30
enddo
enddo
enddo
added function calls in init()
2007-04-10 16:52:53 +05:30
added a new module called numerics.f90 which reads in all numerical "parameters" from the file numerics.config (also being added). From now on this file has to be located in the working directory of the FEM-model and has to contain all necessary parameters.
2009-06-15 18:41:21 +05:30
endsubroutine
added full description of element type 2
2009-01-16 23:06:37 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
!***********************************************************
! assignment of twin nodes for each cp node
!
! allocate globals
! _nodeTwins
!***********************************************************
subroutine mesh_build_nodeTwins()
use prec, only: pInt, pReal
implicit none
integer(pInt) dir, & ! direction of periodicity
node, &
minimumNode, &
maximumNode, &
n1, &
n2
integer(pInt), dimension(mesh_Nnodes+1) :: minimumNodes, maximumNodes ! list of surface nodes (minimum and maximum coordinate value) with first entry giving the number of nodes
real(pReal) minCoord, maxCoord, & ! extreme positions in one dimension
tolerance ! tolerance below which positions are assumed identical
real(pReal), dimension(3) :: distance ! distance between two nodes in all three coordinates
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
logical, dimension(mesh_Nnodes) :: unpaired
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
allocate(mesh_nodeTwins(3,mesh_Nnodes))
mesh_nodeTwins = 0_pInt
tolerance = 0.001_pReal * minval(mesh_ipVolume) ** 0.333_pReal
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do dir = 1_pInt,3_pInt ! check periodicity in directions of x,y,z
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
if (mesh_periodicSurface(dir)) then ! only if periodicity is requested
!*** find out which nodes sit on the surface
!*** and have a minimum or maximum position in this dimension
minimumNodes = 0_pInt
maximumNodes = 0_pInt
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
minCoord = minval(mesh_node0(dir,:))
maxCoord = maxval(mesh_node0(dir,:))
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do node = 1_pInt,mesh_Nnodes ! loop through all nodes and find surface nodes
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
if (abs(mesh_node0(dir,node) - minCoord) <= tolerance) then
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
minimumNodes(1) = minimumNodes(1) + 1_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
minimumNodes(minimumNodes(1)+1_pInt) = node
ip coordinates are now updated every cycle; this is needed for the nonlocal internal stress fields * Marc: node displacements are added to initial node coordinates (mesh_node0) to get current node positions (mesh_node), then ip coordinates are deduced * Abaqus: ip coordinates are directly updated, no update of node coordinates! * Spectral: for the moment no update of either ip or node coordinates! passing only dummy values with initial ip coordinates
2011-05-24 21:27:59 +05:30
elseif (abs(mesh_node0(dir,node) - maxCoord) <= tolerance) then
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
maximumNodes(1) = maximumNodes(1) + 1_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
maximumNodes(maximumNodes(1)+1_pInt) = node
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
endif
enddo
!*** find the corresponding node on the other side with the same position in this dimension
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
unpaired = .true.
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do n1 = 1_pInt,minimumNodes(1)
minimumNode = minimumNodes(n1+1_pInt)
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
if (unpaired(minimumNode)) then
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do n2 = 1_pInt,maximumNodes(1)
maximumNode = maximumNodes(n2+1_pInt)
fixed bug (segmentation fault) occurring for non-renumbered meshes: ipNeighborhood building did use FE IDs of twin nodes to address memory -- dangerous..!
2011-07-31 21:12:59 +05:30
distance = abs(mesh_node0(:,minimumNode) - mesh_node0(:,maximumNode))
if (sum(distance) - distance(dir) <= tolerance) then ! minimum possible distance (within tolerance)
mesh_nodeTwins(dir,minimumNode) = maximumNode
mesh_nodeTwins(dir,maximumNode) = minimumNode
unpaired(maximumNode) = .false. ! remember this node, we don't have to look for his partner again
exit
endif
enddo
endif
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
endif
enddo
endsubroutine
added output of mesh statistics at end of input file parsing
2007-10-24 14:30:42 +05:30
!***********************************************************
! write statistics regarding input file parsing
! to the output file
!
!***********************************************************
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
subroutine mesh_tell_statistics()
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
use prec, only: pInt
use math, only: math_range
use IO, only: IO_error
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
use debug, only: debug_verbosity, &
debug_e, &
debug_i, &
debug_selectiveDebugger
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
implicit none
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
integer(pInt), dimension (:,:), allocatable :: mesh_HomogMicro
character(len=64) fmt
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
integer(pInt) i,e,n,f,t
fixed bug in mesh_build_sharedElems which so far counted nodes in distorted elements twice -- part of this problem has already been fixed with revision 182 but then only in getNodeElemDimensions. M mesh.f90
2008-06-17 02:19:48 +05:30
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
if (mesh_maxValStateVar(1) < 1_pInt) call IO_error(error_ID=170_pInt) ! no homogenization specified
if (mesh_maxValStateVar(2) < 1_pInt) call IO_error(error_ID=180_pInt) ! no microstructure specified
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
allocate (mesh_HomogMicro(mesh_maxValStateVar(1),mesh_maxValStateVar(2))); mesh_HomogMicro = 0_pInt
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
if (mesh_element(3,e) < 1_pInt) call IO_error(error_ID=170_pInt,e=e) ! no homogenization specified
if (mesh_element(4,e) < 1_pInt) call IO_error(error_ID=180_pInt,e=e) ! no microstructure specified
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
mesh_HomogMicro(mesh_element(3,e),mesh_element(4,e)) = &
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
mesh_HomogMicro(mesh_element(3,e),mesh_element(4,e)) + 1_pInt ! count combinations of homogenization and microstructure
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
corrected bug in debug: stressloop info was erroneous mesh init now showing much more ip statistics in verbose mode
2010-09-07 14:36:02 +05:30
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
if (debug_verbosity > 0_pInt) then
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
!$OMP CRITICAL (write2out)
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,*) 'Input Parser: STATISTICS'
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,*) mesh_Nelems, ' : total number of elements in mesh'
write (6,*) mesh_NcpElems, ' : total number of CP elements in mesh'
write (6,*) mesh_Nnodes, ' : total number of nodes in mesh'
write (6,*) mesh_maxNnodes, ' : max number of nodes in any CP element'
write (6,*) mesh_maxNips, ' : max number of IPs in any CP element'
write (6,*) mesh_maxNipNeighbors, ' : max number of IP neighbors in any CP element'
write (6,*) mesh_maxNsubNodes, ' : max number of (additional) subnodes in any CP element'
write (6,*) mesh_maxNsharedElems, ' : max number of CP elements sharing a node'
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,*) 'Input Parser: HOMOGENIZATION/MICROSTRUCTURE'
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,*) mesh_maxValStateVar(1), ' : maximum homogenization index'
write (6,*) mesh_maxValStateVar(2), ' : maximum microstructure index'
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (fmt,'(a,i32.32,a)') '(9x,a2,1x,',mesh_maxValStateVar(2),'(i8))'
write (6,fmt) '+-',math_range(mesh_maxValStateVar(2))
write (fmt,'(a,i32.32,a)') '(i8,1x,a2,1x,',mesh_maxValStateVar(2),'(i8))'
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i=1_pInt,mesh_maxValStateVar(1) ! loop over all (possibly assigned) homogenizations
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,fmt) i,'| ',mesh_HomogMicro(i,:) ! loop over all (possibly assigned) microstructures
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
enddo
write(6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,*) 'Input Parser: ADDITIONAL MPIE OPTIONS'
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
write(6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,*) 'periodic surface : ', mesh_periodicSurface
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
write(6,*)
call flush(6)
!$OMP END CRITICAL (write2out)
endif
if (debug_verbosity > 1) then
corrected bug in debug: stressloop info was erroneous mesh init now showing much more ip statistics in verbose mode
2010-09-07 14:36:02 +05:30
!$OMP CRITICAL (write2out)
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,*) 'Input Parser: SUBNODE COORDINATES'
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,'(a8,1x,a5,1x,a15,1x,a15,1x,a20,3(1x,a12))') 'elem','IP','IP neighbor','IPFaceNodes','subNodeOnIPFace','x','y','z'
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_selectiveDebugger .and. debug_e /= e) cycle
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
t = mesh_element(2,e) ! get elemType
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,FE_Nips(t) ! loop over IPs of elem
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_selectiveDebugger .and. debug_i /= i) cycle
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do f = 1_pInt,FE_NipNeighbors(t) ! loop over interfaces of IP
do n = 1_pInt,FE_NipFaceNodes ! loop over nodes on interface
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,'(i8,1x,i5,1x,i15,1x,i15,1x,i20,3(1x,f12.8))') e,i,f,n,FE_subNodeOnIPFace(n,f,i,t),&
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
mesh_subNodeCoord(1,FE_subNodeOnIPFace(n,f,i,t),e),&
mesh_subNodeCoord(2,FE_subNodeOnIPFace(n,f,i,t),e),&
mesh_subNodeCoord(3,FE_subNodeOnIPFace(n,f,i,t),e)
enddo
enddo
enddo
enddo
write(6,*)
write(6,*) 'Input Parser: IP COORDINATES'
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,'(a8,1x,a5,3(1x,a12))') 'elem','IP','x','y','z'
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_selectiveDebugger .and. debug_e /= e) cycle
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,FE_Nips(mesh_element(2,e))
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_selectiveDebugger .and. debug_i /= i) cycle
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,'(i8,1x,i5,3(1x,f12.8))') e, i, mesh_ipCenterOfGravity(:,i,e)
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
enddo
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,*) 'Input Parser: ELEMENT VOLUME'
corrected bug in debug: stressloop info was erroneous mesh init now showing much more ip statistics in verbose mode
2010-09-07 14:36:02 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,'(a13,1x,e15.8)') 'total volume', sum(mesh_ipVolume)
corrected bug in debug: stressloop info was erroneous mesh init now showing much more ip statistics in verbose mode
2010-09-07 14:36:02 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,'(a8,1x,a5,1x,a15,1x,a5,1x,a15,1x,a16)') 'elem','IP','volume','face','area','-- normal --'
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_selectiveDebugger .and. debug_e /= e) cycle
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,FE_Nips(mesh_element(2,e))
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_selectiveDebugger .and. debug_i /= i) cycle
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,'(i8,1x,i5,1x,e15.8)') e,i,mesh_IPvolume(i,e)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do f = 1_pInt,FE_NipNeighbors(mesh_element(2,e))
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,'(i33,1x,e15.8,1x,3(f6.3,1x))') f,mesh_ipArea(f,i,e),mesh_ipAreaNormal(:,f,i,e)
corrected bug in debug: stressloop info was erroneous mesh init now showing much more ip statistics in verbose mode
2010-09-07 14:36:02 +05:30
enddo
enddo
enddo
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,*) 'Input Parser: NODE TWINS'
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
write (6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,'(a6,3(3x,a6))') ' node','twin_x','twin_y','twin_z'
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do n = 1_pInt,mesh_Nnodes ! loop over cpNodes
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_e <= mesh_NcpElems) then
if (any(mesh_element(5:,debug_e) == n)) then
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,'(i6,3(3x,i6))') n, mesh_nodeTwins(1:3,n)
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
endif
endif
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
write(6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,*) 'Input Parser: IP NEIGHBORHOOD'
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
write(6,*)
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write(6,'(a8,1x,a10,1x,a10,1x,a3,1x,a13,1x,a13)') 'elem','IP','neighbor','','elemNeighbor','ipNeighbor'
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_selectiveDebugger .and. debug_e /= e) cycle
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
t = mesh_element(2,e) ! get elemType
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do i = 1_pInt,FE_Nips(t) ! loop over IPs of elem
debugging output is now controlled by the "verbosity" parameter in the debug.config ranging from 0 (=almost no output) to 8 (=very detailed output) 0 : only version infos and all from "hypela2"/"umat" 1 : basic outputs from "CPFEM.f90", basic output from initialization routines, debug_info 2 : extensive outputs from "CPFEM.f90", extensive output from initialization routines 3 : basic outputs from "homogenization.f90" 4 : extensive outputs from "homogenization.f90" 5 : basic outputs from "crystallite.f90" 6 : extensive outputs from "crystallite.f90" 7 : basic outputs from the constitutive files 8 : extensive outputs from the constitutive files If verbosity is equal to zero, all counters in debug are not set during calculation (e.g. debug_StressLoopDistribution or debug_cumDotStateTicks). This might speed up parallel calculation, because all these need critical statements which extremely slow down parallel computation.
2011-03-21 16:01:17 +05:30
if (debug_selectiveDebugger .and. debug_i /= i) cycle
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
do n = 1_pInt,FE_NipNeighbors(t) ! loop over neighbors of IP
made code standard conform to Fortran 2008 (ignoring warning concerning comments beyond character 132). Basically, changing "x" format specifier to "Nx" ("1x") plus removing $ format specifier added compiler switches for gfortran and ifort to check for standard conformity old gnu compilers <4.4 are not longer supported because they don't provide the c binding for fftw
2012-02-01 00:48:55 +05:30
write (6,'(i8,1x,i10,1x,i10,1x,a3,1x,i13,1x,i13)') e,i,n,'-->',mesh_ipNeighborhood(1,n,i,e),mesh_ipNeighborhood(2,n,i,e)
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
enddo
enddo
1) added distribution of leapfrog breaks 2) lattice_symmetryType is now a function (former lookup array was buggy) 3) stricter check of state var values (>0!) and memory deallocation done
2010-09-30 13:01:53 +05:30
!$OMP END CRITICAL (write2out)
Now able to have real periodicity for fluxes (fluxes leaving the model on one side enter on the other side). To enable this feature one has to add the following somewhere in the marc input file: $mpie periodic x y z for having periodicity in all directions $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
endif
deallocate(mesh_HomogMicro)
endsubroutine
reordered (and partly redistributed) error message identifiers, deleted those which are not in use anymore. all constitutive as well numerics now raises an error, if an unknown keyword is found in the respective config file
2012-02-13 23:11:27 +05:30
END MODULE mesh