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

4019 lines
149 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use prec, only: pReal, pInt
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
private
integer(pInt), public :: &
mesh_NcpElems, & ! total number of CP elements in mesh
mesh_NelemSets, &
mesh_maxNelemInSet, &
mesh_Nmaterials, &
mesh_Nnodes, & ! total number of nodes in mesh
mesh_maxNnodes, & ! max number of nodes in any CP element
mesh_maxNips, & ! max number of IPs in any CP element
mesh_maxNipNeighbors, & ! max number of IP neighbors in any CP element
mesh_maxNsharedElems, & ! max number of CP elements sharing a node
mesh_maxNsubNodes
integer(pInt), dimension(:,:), allocatable, public :: &
mesh_element, & ! FEid, type(internal representation), material, texture, node indices
mesh_sharedElem, & ! entryCount and list of elements containing node
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)
integer(pInt), dimension(:,:,:,:), allocatable, public :: &
mesh_ipNeighborhood ! 6 or less neighboring IPs as [element_num, IP_index]
real(pReal), dimension(:,:), allocatable, public :: &
mesh_ipVolume, & ! volume associated with IP (initially!)
mesh_node0, & ! node x,y,z coordinates (initially!)
mesh_node ! node x,y,z coordinates (after deformation! ONLY FOR MARC!!!)
real(pReal), dimension(:,:,:), allocatable, public :: &
mesh_ipCenterOfGravity, & ! center of gravity of IP (after deformation!)
mesh_ipArea ! area of interface to neighboring IP (initially!)
real(pReal),dimension(:,:,:,:), allocatable, public :: &
mesh_ipAreaNormal ! area normal of interface to neighboring IP (initially!)
logical, dimension(3), public :: mesh_periodicSurface ! flag indicating periodic outer surfaces (used for fluxes)
integer(pInt), private :: &
mesh_Nelems, & ! total number of elements in mesh
hypoelasticTableStyle, &
initialcondTableStyle
integer(pInt), dimension(2), private :: &
mesh_maxValStateVar = 0_pInt
character(len=64), dimension(:), allocatable, private :: &
mesh_nameElemSet, & ! names of elementSet
mesh_nameMaterial, & ! names of material in solid section
mesh_mapMaterial ! name of elementSet for material
integer(pInt), dimension(:,:), allocatable, private :: &
mesh_mapElemSet ! list of elements in elementSet
integer(pInt), dimension(:,:), allocatable, target, private :: &
mesh_mapFEtoCPelem, & ! [sorted FEid, corresponding CPid]
mesh_mapFEtoCPnode ! [sorted FEid, corresponding CPid]
real(pReal),dimension(:,:,:), allocatable, private :: &
mesh_subNodeCoord ! coordinates of subnodes per element
logical, private :: noPart ! for cases where the ABAQUS input file does not use part/assembly information
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
! Thee definitions should actually reside in the FE-solver specific part (different for MARC/ABAQUS)
started envIP building
2007-03-22 20:18:58 +05:30
! Hence, I suggest to prefix with "FE_"
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), parameter, public :: &
FE_Nelemtypes = 10_pInt, &
FE_maxNnodes = 8_pInt, &
FE_maxNsubNodes = 56_pInt, &
FE_maxNips = 27_pInt, &
FE_maxNipNeighbors = 6_pInt, &
FE_maxmaxNnodesAtIP = 8_pInt, & ! max number of (equivalent) nodes attached to an IP
FE_NipFaceNodes = 4_pInt
integer(pInt), dimension(FE_Nelemtypes), parameter, public :: FE_Nnodes = & ! nodes in a specific type of element (how we use 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
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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter, public :: FE_Nips = & ! IPs in a specific type of 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter, public :: FE_NipNeighbors = & !IP neighbors in a specific type of 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter, private :: FE_NsubNodes = & ! subnodes required to fully define all IP volumes
int([19,& ! element 7 ! order is +x,-x,+y,-y,+z,-z but meaning strongly depends on 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter, private :: FE_NoriginalNodes = & ! nodes in a specific type of element (how it is originally defined by marc)
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)
integer(pInt), dimension(FE_maxNipNeighbors,FE_Nelemtypes), parameter, private :: FE_NfaceNodes = &! nodes per face in a specific type of 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
],pInt),[FE_maxNipNeighbors,FE_Nelemtypes])
integer(pInt), dimension(FE_Nelemtypes), parameter, private :: FE_maxNnodesAtIP = & ! map IP index to two node indices in a specific type of 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), dimension(FE_NipFaceNodes,FE_maxNipNeighbors,FE_Nelemtypes), parameter, private :: &
FE_nodeOnFace = & ! List of node indices on each face of a specific type of 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
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 &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
],pInt),[FE_NipFaceNodes,FE_maxNipNeighbors,FE_Nelemtypes])
integer(pInt), dimension(:,:,:), allocatable, private :: &
FE_nodesAtIP, & ! map IP index to two node indices in a specific type of element
FE_ipNeighbor, & ! +x,-x,+y,-y,+z,-z list of intra-element IPs and(negative) neighbor faces per own IP in a specific type of element
FE_subNodeParent
integer(pInt), dimension(:,:,:,:), allocatable, private :: &
FE_subNodeOnIPFace
public :: mesh_init, &
mesh_FEasCP, &
mesh_build_subNodeCoords, &
mesh_build_ipVolumes, &
integrated subroutine for regridding into mesh.f90 and made it available for python
2012-04-10 20:45:46 +05:30
mesh_build_ipCoordinates, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
mesh_regrid, &
mesh_spectral_getResolution, &
mesh_spectral_getDimension, &
mesh_spectral_getHomogenization
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
private :: FE_mapElemtype, &
mesh_faceMatch, &
mesh_build_FEdata, &
mesh_marc_get_tableStyles, &
mesh_get_damaskOptions, &
mesh_spectral_count_nodesAndElements, &
mesh_marc_count_nodesAndElements, &
mesh_abaqus_count_nodesAndElements, &
mesh_abaqus_count_elementSets, &
mesh_abaqus_count_materials, &
mesh_spectral_count_cpElements, &
mesh_abaqus_count_cpElements, &
mesh_marc_map_elementSets, &
mesh_abaqus_map_elementSets, &
mesh_abaqus_map_materials, &
mesh_spectral_map_nodes, &
mesh_marc_map_nodes, &
mesh_abaqus_map_nodes, &
mesh_marc_map_elements, &
mesh_abaqus_map_elements, &
mesh_spectral_count_cpSizes, &
mesh_marc_count_cpSizes, &
mesh_abaqus_count_cpSizes, &
mesh_spectral_build_nodes, &
mesh_marc_build_nodes, &
mesh_abaqus_build_nodes, &
mesh_spectral_build_elements, &
mesh_marc_build_elements, &
mesh_abaqus_build_elements, &
mesh_build_ipNeighborhood, &
mesh_build_ipAreas, &
mesh_build_nodeTwins, &
mesh_tell_statistics
contains
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
!***********************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use, intrinsic :: iso_fortran_env ! to get compiler_version and compiler_options (at least for gfortran 4.6 at the moment)
use IO, only: IO_error, &
IO_open_InputFile, &
IO_abaqus_hasNoPart
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
call IO_open_inputFile(fileUnit,FEmodelGeometry) ! parse info from input file...
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
select case (FEsolver)
case ('Spectral')
call mesh_spectral_count_nodesAndElements(fileUnit)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
call mesh_spectral_count_cpElements
call mesh_spectral_map_elements
call mesh_spectral_map_nodes
call mesh_spectral_count_cpSizes
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 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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
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
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
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 function calls in init()
2007-04-10 16:52:53 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_init
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end function FE_mapElemtype
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) function mesh_FEasCP(what,myID)
added function calls in init()
2007-04-10 16:52:53 +05:30
use IO, only: IO_lc
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
implicit none
character(len=*), intent(in) :: what
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: myID
added function calls in init()
2007-04-10 16:52:53 +05:30
integer(pInt), dimension(:,:), pointer :: lookupMap
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: lower,upper,center
added function calls in init()
2007-04-10 16:52:53 +05:30
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?
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (lookupMap(1_pInt,lower) == myID) 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
mesh_FEasCP = lookupMap(2_pInt,lower)
added function calls in init()
2007-04-10 16:52:53 +05:30
return
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
elseif (lookupMap(1_pInt,upper) == myID) 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (lookupMap(1_pInt,center) < myID) then
added function calls in init()
2007-04-10 16:52:53 +05:30
lower = center
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
elseif (lookupMap(1_pInt,center) > myID) 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end function mesh_FEasCP
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)
implicit none
!*** output variables
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(out) :: matchingElem, & ! matching CP element ID
matchingFace ! matching FE face ID
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
!*** input variables
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: face, & ! FE face ID
elem ! FE elem ID
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
!*** local variables
integer(pInt), dimension(FE_NfaceNodes(face,mesh_element(2,elem))) :: &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
myFaceNodes ! global node ids on my face
integer(pInt) :: myType, &
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
candidateType, &
candidateElem, &
candidateFace, &
candidateFaceNode, &
minNsharedElems, &
NsharedElems, &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
lonelyNode = 0_pInt, &
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
i, &
n, &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
dir ! periodicity direction
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 :: element_seen
logical checkTwins
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
matchingElem = 0_pInt
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
matchingFace = 0_pInt
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
minNsharedElems = mesh_maxNsharedElems + 1_pInt ! init to worst case
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
minNsharedElems = NsharedElems ! remember min # shared elems
lonelyNode = n ! remember most lonely 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
allocate(element_seen(minNsharedElems))
element_seen = 0_pInt
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
if (all(element_seen /= candidateElem)) then ! element seen for the first time?
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
element_seen(i) = candidateElem
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_faceMatch
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
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
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
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_nodesAtIP(1:FE_maxNnodesAtIP(1),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([&
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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_maxNnodesAtIP(1),FE_Nips(1)])
FE_nodesAtIP(1:FE_maxNnodesAtIP(2),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_maxNnodesAtIP(2),FE_Nips(2)])
FE_nodesAtIP(1:FE_maxNnodesAtIP(3),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_maxNnodesAtIP(3),FE_Nips(3)])
FE_nodesAtIP(1:FE_maxNnodesAtIP(4),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_maxNnodesAtIP(4),FE_Nips(4)])
FE_nodesAtIP(1:FE_maxNnodesAtIP(5),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_maxNnodesAtIP(5),FE_Nips(5)])
FE_nodesAtIP(1:FE_maxNnodesAtIP(6),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_maxNnodesAtIP(6),FE_Nips(6)])
FE_nodesAtIP(1:FE_maxNnodesAtIP(7),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +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)/))
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_nodesAtIP(1:FE_maxNnodesAtIP(9),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_maxNnodesAtIP(9),FE_Nips(9)])
FE_nodesAtIP(1:FE_maxNnodesAtIP(10),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +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.
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(1),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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(1),FE_Nips(1)])
FE_ipNeighbor(1:FE_NipNeighbors(2),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(2),FE_Nips(2)])
FE_ipNeighbor(1:FE_NipNeighbors(3),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(3),FE_Nips(3)])
FE_ipNeighbor(1:FE_NipNeighbors(4),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(4),FE_Nips(4)])
FE_ipNeighbor(1:FE_NipNeighbors(5),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(5),FE_Nips(5)])
FE_ipNeighbor(1:FE_NipNeighbors(6),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(6),FE_Nips(6)])
FE_ipNeighbor(1:FE_NipNeighbors(7),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(7),FE_Nips(7)])
FE_ipNeighbor(1:FE_NipNeighbors(8),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(8),FE_Nips(8)])
FE_ipNeighbor(1:FE_NipNeighbors(9),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipNeighbors(9),FE_Nips(9)])
FE_ipNeighbor(1:FE_NipNeighbors(10),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +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.
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
! fill with 0.
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
! 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
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_subNodeParent(1:FE_Nips(1),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)/))
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
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
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_subNodeParent(1:FE_Nips(3),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_Nips(3),FE_NsubNodes(3)])
FE_subNodeParent(1:FE_Nips(4),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +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)/))
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_subNodeParent(1:FE_Nips(6),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_Nips(6),FE_NsubNodes(6)])
FE_subNodeParent(1:FE_Nips(7),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +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
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_subNodeParent(1:FE_Nips(9),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_Nips(9),FE_NsubNodes(9)])
FE_subNodeParent(1:FE_Nips(10),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +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
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(1),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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipFaceNodes,FE_NipNeighbors(1),FE_Nips(1)])
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(2),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipFaceNodes,FE_NipNeighbors(2),FE_Nips(2)])
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(3),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipFaceNodes,FE_NipNeighbors(3),FE_Nips(3)])
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(4),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +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)/))
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(6),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipFaceNodes,FE_NipNeighbors(6),FE_Nips(6)])
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(7),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipFaceNodes,FE_NipNeighbors(7),FE_Nips(7)])
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(8),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipFaceNodes,FE_NipNeighbors(8),FE_Nips(8)])
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(9),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipFaceNodes,FE_NipNeighbors(9),FE_Nips(9)])
FE_subNodeOnIPFace(1:FE_NipFaceNodes,1:FE_NipNeighbors(10),1: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 &
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
],pInt),[FE_NipFaceNodes,FE_NipNeighbors(10),FE_Nips(10)])
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end 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
# 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
several small corrections/polishing
2012-04-20 17:28:41 +05:30
use IO, only: &
IO_lc, &
IO_intValue, &
IO_stringValue, &
IO_stringPos
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +05:30
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
620 end subroutine mesh_marc_get_tableStyles
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
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
several small corrections/polishing
2012-04-20 17:28:41 +05:30
mesh_periodicSurface = .false.
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
620 end subroutine mesh_get_damaskOptions
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: myUnit
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
integer(pInt), dimension(3) :: res
res = mesh_spectral_getResolution(myUnit)
mesh_Nelems = res(1)*res(2)*res(3)
mesh_Nnodes = (1_pInt + res(1))*(1_pInt + res(2))*(1_pInt + res(3))
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_spectral_count_nodesAndElements
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_IntValue
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
620 end subroutine mesh_marc_count_nodesAndElements
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_abaqus_count_nodesAndElements(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_countDataLines, &
IO_error
added awareness of set definitions (new function build_elemSetMapping) added new style tables for initial conditions
2007-10-23 18:39:46 +05:30
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_count_nodesAndElements
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_countContinuousIntValues
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: myUnit
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
# 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, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_countContinuousIntValues(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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
620 end subroutine mesh_marc_count_elementSets
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_abaqus_count_elementSets(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_error
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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: myUnit
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_count_elementSets
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_error
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
implicit none
integer(pInt), intent(in) :: 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
integer(pInt), parameter :: maxNchunks = 2_pInt
integer(pInt), dimension (1_pInt+2_pInt*maxNchunks) :: myPos
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_count_materials
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
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_spectral_count_cpElements
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
implicit none
mesh_NcpElems = mesh_Nelems
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_spectral_count_cpElements
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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_countContinuousIntValues
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: i
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
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
mesh_NcpElems = mesh_NcpElems + IO_countContinuousIntValues(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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
620 end subroutine mesh_marc_count_cpElements
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_abaqus_count_cpElements(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_error, &
IO_extractValue
added function calls in init()
2007-04-10 16:52:53 +05:30
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: i,k
logical :: materialFound = .false.
character(len=64) ::materialName,elemSetName
# 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_count_cpElements
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_continuousIntValues
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
integer(pInt) :: elemSet = 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
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)
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))
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
mesh_mapElemSet(:,elemSet) = IO_continuousIntValues(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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
640 end subroutine mesh_marc_map_elementSets
# 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_extractValue, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_continuousIntValues, &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
IO_error
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
integer(pInt) :: elemSet = 0_pInt,i
logical :: 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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
allocate (mesh_nameElemSet(mesh_NelemSets)) ; mesh_nameElemSet = ''
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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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=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'))
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
mesh_mapElemSet(:,elemSet) = IO_continuousIntValues(myUnit,mesh_Nelems,mesh_nameElemSet,&
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_map_elementSets
# 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 solid section (Abaqus only)
!
! allocate globals: mesh_nameMaterial, mesh_mapMaterial
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_abaqus_map_materials(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_extractValue, &
IO_error
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: i,c = 0_pInt
logical :: inPart = .false.
character(len=64) :: elemSetName,materialName
# 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_nameMaterial(mesh_Nmaterials)) ; mesh_nameMaterial = ''
allocate (mesh_mapMaterial(mesh_Nmaterials)) ; mesh_mapMaterial = ''
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)
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_map_materials
# 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
!********************************************************************
! map nodes from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPnode
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_spectral_map_nodes
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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: i
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_spectral_map_nodes
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
! map nodes from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPnode
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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 math, only: qsort
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_fixedIntValue
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_marc_map_nodes
# 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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 math, only: qsort
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_countDataLines, &
IO_intValue, &
IO_error
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: i,c,cpNode = 0_pInt
logical :: 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
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)
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_map_nodes
# 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
!********************************************************************
! map elements from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPelem
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_spectral_map_elements
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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: i
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_spectral_map_elements
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
!********************************************************************
! map elements from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPelem
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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 math, only: qsort
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_continuousIntValues
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: i,cpElem = 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
allocate (mesh_mapFEtoCPelem(2,mesh_NcpElems)) ; mesh_mapFEtoCPelem = 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=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
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
contInts = IO_continuousIntValues(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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_marc_map_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
!********************************************************************
! map elements from FE id to internal (consecutive) representation
!
! allocate globals: mesh_mapFEtoCPelem
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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 math, only: qsort
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_extractValue, &
IO_error
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
integer(pInt) ::i,j,k,cpElem = 0_pInt
logical :: materialFound = .false.
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)
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
do i = 1_pInt,mesh_Nmaterials ! look thru material names
if (materialName == mesh_nameMaterial(i)) then ! found one
elemSetName = mesh_mapMaterial(i) ! take corresponding elemSet
do k = 1_pInt,mesh_NelemSets ! look thru all elemSet definitions
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
mesh_mapFEtoCPelem(1,cpElem) = mesh_mapElemSet(1_pInt+j,k) ! store FE id
mesh_mapFEtoCPelem(2,cpElem) = cpElem ! store our 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_map_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
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
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_spectral_count_cpSizes
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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: t
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_spectral_count_cpSizes
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
!********************************************************************
! get maximum count of nodes, IPs, IP neighbors, and subNodes
! among cpElements
!
! _maxNnodes, _maxNips, _maxNipNeighbors, _maxNsubNodes
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_intValue, &
IO_skipChunks
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
integer(pInt) :: 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
read (myUnit,610,END=630) line ! Garbage line
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
630 end subroutine mesh_marc_count_cpSizes
# 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_abaqus_count_cpSizes(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_extractValue ,&
IO_error, &
IO_countDataLines, &
IO_intValue
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
integer(pInt) :: i,c,t
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
620 end subroutine mesh_abaqus_count_cpSizes
# 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_spectral_build_nodes(myUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
use IO, only: &
IO_error
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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: myUnit
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
integer(pInt) :: n
integer(pInt), dimension(3) :: res = 1_pInt
real(pReal), dimension(3) :: geomdim = 1.0_pReal
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
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
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
res = mesh_spectral_getResolution(myUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
geomdim = mesh_spectral_getDimension(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
forall (n = 0_pInt:mesh_Nnodes-1_pInt)
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
mesh_node0(1,n+1_pInt) = geomdim(1) * real(mod(n,(res(1)+1_pInt) ),pReal) &
/ real(res(1),pReal)
mesh_node0(2,n+1_pInt) = geomdim(2) * real(mod(n/(res(1)+1_pInt),(res(2)+1_pInt)),pReal) &
/ real(res(2),pReal)
mesh_node0(3,n+1_pInt) = geomdim(3) * real(mod(n/(res(1)+1_pInt)/(res(2)+1_pInt),(res(3)+1_pInt)),pReal) &
/ real(res(3),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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_spectral_build_nodes
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_marc_build_nodes(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_fixedIntValue, &
IO_fixedNoEFloatValue
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
integer(pInt) :: 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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_marc_build_nodes
# 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_abaqus_build_nodes(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_floatValue, &
IO_stringPos, &
IO_error, &
IO_countDataLines, &
IO_intValue
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=300) :: line
integer(pInt) :: i,j,m,c
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
c = IO_countDataLines(myUnit) ! how many nodes are defined here?
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_abaqus_build_nodes
# 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 FEid, type, mat, tex, and node list per element
!
! allocate globals:
! _element
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_spectral_build_elements(myUnit)
several small corrections/polishing
2012-04-20 17:28:41 +05:30
use IO, only: &
IO_checkAndRewind, &
IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_error, &
IO_continuousIntValues, &
IO_intValue, &
IO_countContinuousIntValues
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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
integer(pInt), dimension(3) :: res
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: e, i, j, homog = 0_pInt, headerLength = 0_pInt, maxIntCount
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), dimension(:), allocatable :: microstructures
integer(pInt), dimension(1,1) :: dummySet = 0_pInt
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character(len=65536) :: line,keyword
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=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
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
res = mesh_spectral_getResolution(myUnit)
homog = mesh_spectral_getHomogenization(myUnit)
several small corrections/polishing
2012-04-20 17:28:41 +05:30
call IO_checkAndRewind(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
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
call IO_error(error_ID=841_pInt, ext_msg='mesh_spectral_build_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
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
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)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
i = IO_countContinuousIntValues(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
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
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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!)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
microstructures = IO_continuousIntValues(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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
e = e+1_pInt ! valid element entry
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
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
mesh_element( 5,e) = e + (e-1_pInt)/res(1) + &
((e-1_pInt)/(res(1)*res(2)))*(res(1)+1_pInt) ! base node
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
mesh_element( 6,e) = mesh_element(5,e) + 1_pInt
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
mesh_element( 7,e) = mesh_element(5,e) + res(1) + 2_pInt
mesh_element( 8,e) = mesh_element(5,e) + res(1) + 1_pInt
mesh_element( 9,e) = mesh_element(5,e) +(res(1) + 1_pInt) * (res(2) + 1_pInt) ! 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
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
mesh_element(11,e) = mesh_element(9,e) + res(1) + 2_pInt
mesh_element(12,e) = mesh_element(9,e) + res(1) + 1_pInt
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
mesh_maxValStateVar(1) = max(mesh_maxValStateVar(1),mesh_element(3,e)) !needed for statistics
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(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
changed assumed arrays (:) to properly defined ones (gfortran 4.7 was complaining) and remove use statements that are not needed
2012-02-21 21:09:36 +05:30
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_spectral_build_elements
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_marc_build_elements(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_fixedNoEFloatValue, &
IO_skipChunks, &
IO_stringPos, &
IO_intValue, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_continuousIntValues
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: i,j,sv,myVal,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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
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))
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (e /= 0_pInt) then ! disregard non CP elems
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))) &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
620 rewind(myUnit) ! just in case "initial state" apears before "connectivity"
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
# 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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (initialcondTableStyle == 2_pInt) read (myUnit,610,END=620) line ! read extra line for new style
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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
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
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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
do while (scan(IO_stringValue(line,myPos,1_pInt),'+-',back=.true.)>1) ! is noEfloat value?
myVal = nint(IO_fixedNoEFloatValue(line,[0_pInt,20_pInt],1_pInt),pInt) ! state var's value
mesh_maxValStateVar(sv-1_pInt) = max(myVal,mesh_maxValStateVar(sv-1_pInt)) ! remember max val of homogenization and microstructure 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
if (initialcondTableStyle == 2_pInt) then
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
contInts = IO_continuousIntValues& ! get affected elements
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
(myUnit,mesh_Nelems,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)
fixing inconsistencies in variable assignments
2012-02-10 17:26:05 +05:30
e = mesh_FEasCP('elem',contInts(1_pInt+i))
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
mesh_element(1_pInt+sv,e) = myVal
# 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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
630 end subroutine mesh_marc_build_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
!********************************************************************
! store FEid, type, mat, tex, and node list per element
!
! allocate globals:
! _element
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_abaqus_build_elements(myUnit)
use IO, only: IO_lc, &
IO_stringValue, &
IO_skipChunks, &
IO_stringPos, &
IO_intValue, &
IO_extractValue, &
IO_floatValue, &
IO_error, &
IO_countDataLines
# 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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt), intent(in) :: 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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) :: 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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
character (len=64) :: materialName,elemSetName
character(len=300) :: 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
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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')
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
materialName = trim(IO_extractValue(IO_lc(IO_StringValue(line,myPos,2_pInt)),'name')) ! extract name=value
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
do i = 1_pInt,mesh_Nmaterials ! look thru material names
if (materialName == mesh_nameMaterial(i)) then ! found one
elemSetName = mesh_mapMaterial(i) ! take corresponding elemSet
do k = 1_pInt,mesh_NelemSets ! look thru all elemSet definitions
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))
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
mesh_element(3,e) = homog ! store homogenization
mesh_element(4,e) = micro ! store microstructure
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_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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
630 end subroutine mesh_abaqus_build_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
!********************************************************************
! get maximum count of shared elements among cpElements and
! build list of elements shared by each node in mesh
!
! _maxNsharedElems
! _sharedElem
!********************************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_build_sharedElems
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pint) e, & ! element index
t, & ! element type
node, & ! CP node index
j, & ! node index per element
myDim, & ! dimension index
nodeTwin ! node twin in the specified dimension
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 (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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
node_seen = 0_pInt ! reset node duplicates
do j = 1_pInt,FE_Nnodes(t) ! check each node of element
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
node_count(node) = node_count(node) + 1_pInt ! if FE node not yet encountered -> count it
do myDim = 1_pInt,3_pInt ! check in each dimension...
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
nodeTwin = mesh_nodeTwins(myDim,node)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (nodeTwin > 0_pInt) & ! if I am a twin of some node...
node_count(nodeTwin) = node_count(nodeTwin) + 1_pInt ! -> count me again for the twin 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
enddo
endif
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
node_seen(j) = node ! remember this node to be counted already
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
mesh_sharedElem(1,node) = mesh_sharedElem(1,node) + 1_pInt ! count for each node the connected elements
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...
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
nodeTwin = mesh_nodeTwins(myDim,node)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (nodeTwin > 0_pInt) then ! if i am a twin of some node...
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
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_build_sharedElems
# 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
!***********************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_build_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
implicit none
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
integer(pInt) myElem, & ! my CP element 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
myIP, &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
myType, & ! my 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
myFace, &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
neighbor, & ! neighor index
neighboringIPkey, & ! positive integer indicating the neighboring IP (for intra-element) and negative integer indicating the face towards neighbor (for neighboring element)
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
candidateIP, &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
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) :: &
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
linkedNodes = 0_pInt, &
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
matchingNodes
logical checkTwins
allocate(mesh_ipNeighborhood(2,mesh_maxNipNeighbors,mesh_maxNips,mesh_NcpElems))
mesh_ipNeighborhood = 0_pInt
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
do myElem = 1_pInt,mesh_NcpElems ! loop over cpElems
myType = mesh_element(2,myElem) ! get elemType
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_build_ipNeighborhood
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_build_subNodeCoords
added full description of element type 2
2009-01-16 23:06:37 +05:30
implicit none
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
integer(pInt) e,t,n,p,Nparents
# 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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
Nparents = count(FE_subNodeParent(1_pInt:FE_Nips(t),n,t) > 0_pInt)
do p = 1_pInt,Nparents ! loop through present parent nodes
mesh_subNodeCoord(1:3,FE_Nnodes(t)+n,e) = mesh_subNodeCoord(1:3,FE_Nnodes(t)+n,e) &
+ 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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
mesh_subNodeCoord(1:3,n+FE_Nnodes(t),e) = mesh_subNodeCoord(1:3,n+FE_Nnodes(t),e)/real(Nparents,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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_build_subNodeCoords
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
!***********************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_build_ipCoordinates
added full description of element type 2
2009-01-16 23:06:37 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use prec, only: tol_gravityNodePos
added full description of element type 2
2009-01-16 23:06:37 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
implicit none
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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
if (.not. allocated(mesh_ipCenterOfGravity)) allocate(mesh_ipCenterOfGravity(3,mesh_maxNips,mesh_NcpElems))
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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
gravityNode = .false. ! reset flagList
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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
do j = 1_pInt,mesh_maxNnodes+mesh_maxNsubNodes-1_pInt ! walk through entire flagList except last
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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_build_ipCoordinates
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 volume
!
! allocate globals
! _ipVolume
!***********************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_build_ipVolumes
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
use math, only: math_volTetrahedron
implicit none
added full description of element type 2
2009-01-16 23:06:37 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
do f = 1_pInt,FE_NipNeighbors(t) ! loop over interfaces of IP and add tetrahedra which connect to CoG
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) &
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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
volume(j,n) = math_volTetrahedron(nPos(:,n), & ! calc volume of respective tetrahedron to CoG
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_build_ipVolumes
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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_build_ipAreas
added full description of element type 2
2009-01-16 23:06:37 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
use math, only: math_vectorproduct
added full description of element type 2
2009-01-16 23:06:37 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
implicit none
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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +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)
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
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))
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +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) &
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +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 new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_build_ipAreas
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
!***********************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_build_nodeTwins
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
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
subnodeparent check is now running only over actual number (not max) of parent nodes...
2012-04-17 14:49:44 +05:30
if (mesh_periodicSurface(dir)) then ! only if periodicity is requested
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 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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_build_nodeTwins
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 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
!
!***********************************************************
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
subroutine mesh_tell_statistics
use math, only: math_range
use IO, only: IO_error
use debug, only: debug_what, &
debug_mesh, &
debug_levelBasic, &
debug_levelExtensive, &
debug_levelSelective, &
debug_e, &
debug_i
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
implicit none
integer(pInt), dimension (:,:), allocatable :: mesh_HomogMicro
character(len=64) :: myFmt
integer(pInt) :: i,e,n,f,t, myDebug
myDebug = debug_what(debug_mesh)
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +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 $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
!$OMP CRITICAL (write2out)
if (iand(myDebug,debug_levelBasic) /= 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
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,*)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
write (myFmt,'(a,i32.32,a)') '(9x,a2,1x,',mesh_maxValStateVar(2),'(i8))'
write (6,myFmt) '+-',math_range(mesh_maxValStateVar(2))
write (myFmt,'(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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
write (6,myFmt) 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)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
endif
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelExtensive) /= 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
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,*)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
write(6,'(a8,1x,a5,1x,2(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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelSelective) /= 0_pInt .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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelSelective) /= 0_pInt .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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
write(6,'(i8,1x,i5,2(1x,i15),1x,i20,3(1x,f12.8))') e,i,f,n,FE_subNodeOnIPFace(n,f,i,t),&
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)
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
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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelSelective) /= 0_pInt .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))
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelSelective) /= 0_pInt .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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelSelective) /= 0_pInt .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))
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelSelective) /= 0_pInt .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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelSelective) /= 0_pInt .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
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
if (iand(myDebug,debug_levelSelective) /= 0_pInt .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 $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
enddo
enddo
enddo
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
endif
!$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
deallocate(mesh_HomogMicro)
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end subroutine mesh_tell_statistics
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
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
function mesh_regrid(resNewInput) !use new_res=[0.0,0.0,0.0] for automatic determination of new grid
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
use prec, only: &
pInt, &
pReal
use DAMASK_interface, only: &
getSolverJobName
use IO, only: &
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
IO_read_jobBinaryFile ,&
IO_write_jobBinaryFile
use math, only: &
math_nearestNeighborSearch, &
reworked neares neigbor search subroutine, now also works for voronoi tesselation (general nn search in periodic cell)
2012-05-04 18:37:37 +05:30
deformed_FFT, &
math_mul33x3
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
integer(pInt), dimension(3) :: mesh_regrid
integer(pInt), dimension(3), optional, intent(in) :: resNewInput
integer(pInt):: maxsize, i, j, k, ielem, Npoints, NpointsNew, spatialDim
integer(pInt), dimension(3) :: res, resNew
integer(pInt), dimension(:), allocatable :: indices
real(pReal), dimension(3) :: geomdim
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
real(pReal), dimension(3,3) :: Favg
reworked neares neigbor search subroutine, now also works for voronoi tesselation (general nn search in periodic cell)
2012-05-04 18:37:37 +05:30
real(pReal), dimension(:,:), allocatable :: &
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
coordinatesNew, &
coordinatesLinear
real(pReal), dimension(:,:,:), allocatable :: &
material_phase, material_phaseNew
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
real(pReal), dimension(:,:,:,:,:), allocatable :: &
F, FNew, &
Fp, FpNew, &
Lp, LpNew, &
dcsdE, dcsdENew
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
real(pReal), dimension (:,:,:,:,:,:,:), allocatable :: &
dPdF, dPdFNew
real(pReal), dimension (:,:,:,:), allocatable :: &
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
coordinates, &
Tstar, TstarNew, &
stateHomog, &
stateConst
integer(pInt), dimension(:,:), allocatable :: &
sizeStateConst, sizeStateHomog
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
res = mesh_spectral_getResolution()
geomdim = mesh_spectral_getDimension()
Npoints = res(1)*res(2)*res(3)
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
reworked neares neigbor search subroutine, now also works for voronoi tesselation (general nn search in periodic cell)
2012-05-04 18:37:37 +05:30
allocate(F(res(1),res(2),res(3),3,3))
integrated subroutine for regridding into mesh.f90 and made it available for python
2012-04-10 20:45:46 +05:30
call IO_read_jobBinaryFile(777,'convergedSpectralDefgrad',trim(getSolverJobName()),size(F))
read (777,rec=1) F
close (777)
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
! ----Calculate deformed configuration and average--------
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
do i= 1_pInt,3_pInt; do j = 1_pInt,3_pInt
Favg(i,j) = sum(F(1:res(1),1:res(2),1:res(3),i,j)) / Npoints
enddo; enddo
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
allocate(coordinates(res(1),res(2),res(3),3))
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
call deformed_fft(res,geomdim,Favg,1.0_pReal,F,coordinates)
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
deallocate(F)
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
! ----Store coordinates into a linear list--------
allocate(coordinatesLinear(3,Npoints))
ielem = 0_pInt
do k=1_pInt,res(3); do j=1_pInt, res(2); do i=1_pInt, res(1)
ielem = ielem + 1_pInt
coordinatesLinear(1:3,ielem) = coordinates(i,j,k,1:3)
enddo; enddo; enddo
deallocate(coordinates)
! ----For 2D /3D case----------------------------------
if (res(3)== 1_pInt) then
spatialDim = 2_pInt
else
spatialDim = 3_pInt
endif
! ----determine/calculate new resolution--------------
if (.not. present(resNewInput)) then
corrected automatic determination of resolution
2012-05-08 20:42:43 +05:30
resNew = res
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
else
corrected automatic determination of resolution
2012-05-08 20:42:43 +05:30
if(all(resNewInput==0.0_pReal)) then
mesh_regrid =[ 0,0,0]
return !no automatic determination right now
else
resNew = resNewInput
endif
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
endif
NpointsNew = resNew(1)*resNew(2)*resNew(3)
! ----Calculate regular new coordinates------------
allocate(coordinatesNew(3,NpointsNew))
reworked neares neigbor search subroutine, now also works for voronoi tesselation (general nn search in periodic cell)
2012-05-04 18:37:37 +05:30
ielem = 0_pInt
do k=1_pInt,resNew(3); do j=1_pInt, resNew(2); do i=1_pInt, resNew(1)
ielem = ielem + 1_pInt
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
coordinatesNew(1:3,ielem) = math_mul33x3(Favg, geomdim/real(resNew,pReal)*real([i,j,k],pReal) &
reworked neares neigbor search subroutine, now also works for voronoi tesselation (general nn search in periodic cell)
2012-05-04 18:37:37 +05:30
- geomdim/real(2_pInt*resNew,pReal))
enddo; enddo; enddo
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
!----- Nearest neighbour search -----------------------
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
allocate(indices(Npoints))
call math_nearestNeighborSearch(spatialDim, Favg, geomdim, NpointsNew, Npoints, &
coordinatesNew, coordinatesLinear, indices)
deallocate(coordinatesNew)
indices = indices / 3_pInt**spatialDim +1_pInt
reworked neares neigbor search subroutine, now also works for voronoi tesselation (general nn search in periodic cell)
2012-05-04 18:37:37 +05:30
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
!---------------------------------------------------------------------------
allocate(material_phase (1,1, Npoints))
allocate(material_phaseNew (1,1, NpointsNew))
call IO_read_jobBinaryFile(777,'recordedPhase',trim(getSolverJobName()),size(material_phase))
read (777,rec=1) material_phase
close (777)
do i = 1, NpointsNew
material_phaseNew(1,1,i) = material_phase(1,1,indices(i))
enddo
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
call IO_write_jobBinaryFile(777,'recordedPhase',size(material_phaseNew))
write (777,rec=1) material_phaseNew
close (777)
deallocate(material_phase)
deallocate(material_phaseNew)
!---------------------------------------------------------------------------
allocate(F (3,3,1,1, Npoints))
allocate(FNew (3,3,1,1, NpointsNew))
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
call IO_read_jobBinaryFile(777,'convergedF',trim(getSolverJobName()),size(F))
read (777,rec=1) F
close (777)
do i = 1, NpointsNew
FNew(1:3,1:3,1,1,i) = F(1:3,1:3,1,1,indices(i))
enddo
call IO_write_jobBinaryFile(777,'convergedF',size(FNew))
write (777,rec=1) FNew
close (777)
deallocate(F)
deallocate(FNew)
!---------------------------------------------------------------------
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
allocate(Fp (3,3,1,1,Npoints))
allocate(FpNew (3,3,1,1,NpointsNew))
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
call IO_read_jobBinaryFile(777,'convergedFp',trim(getSolverJobName()),size(Fp))
read (777,rec=1) Fp
close (777)
do i = 1, NpointsNew
FpNew(1:3,1:3,1,1,i) = Fp(1:3,1:3,1,1,indices(i))
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
enddo
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
call IO_write_jobBinaryFile(777,'convergedFp',size(FpNew))
write (777,rec=1) FpNew
close (777)
deallocate(Fp)
deallocate(FpNew)
!------------------------------------------------------------------------
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
allocate(Lp (3,3,1,1,Npoints))
allocate(LpNew (3,3,1,1,NpointsNew))
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
call IO_read_jobBinaryFile(777,'convergedLp',trim(getSolverJobName()),size(Lp))
read (777,rec=1) Lp
close (777)
do i = 1, NpointsNew
LpNew(1:3,1:3,1,1,i) = Lp(1:3,1:3,1,1,indices(i))
enddo
call IO_write_jobBinaryFile(777,'convergedLp',size(LpNew))
write (777,rec=1) LpNew
close (777)
deallocate(Lp)
deallocate(LpNew)
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
!----------------------------------------------------------------------------
allocate(dcsdE (6,6,1,1,Npoints))
allocate(dcsdENew (6,6,1,1,NpointsNew))
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
call IO_read_jobBinaryFile(777,'convergeddcsdE',trim(getSolverJobName()),size(dcsdE))
read (777,rec=1) dcsdE
close (777)
do i = 1, NpointsNew
dcsdENew(1:6,1:6,1,1,i) = dcsdE(1:6,1:6,1,1,indices(i))
enddo
call IO_write_jobBinaryFile(777,'convergeddcsdE',size(dcsdENew))
write (777,rec=1) dcsdENew
close (777)
deallocate(dcsdE)
deallocate(dcsdENew)
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
!---------------------------------------------------------------------------
allocate(dPdF (3,3,3,3,1,1,Npoints))
allocate(dPdFNew (3,3,3,3,1,1,NpointsNew))
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
call IO_read_jobBinaryFile(777,'convergeddPdF',trim(getSolverJobName()),size(dPdF))
read (777,rec=1) dPdF
close (777)
do i = 1, NpointsNew
dPdFNew(1:3,1:3,1:3,1:3,1,1,i) = dPdF(1:3,1:3,1:3,1:3,1,1,indices(i))
enddo
call IO_write_jobBinaryFile(777,'convergeddPdF',size(dPdFNew))
write (777,rec=1) dPdFNew
close (777)
deallocate(dPdF)
deallocate(dPdFNew)
!---------------------------------------------------------------------------
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
allocate(Tstar (6,1,1,Npoints))
allocate(TstarNew (6,1,1,NpointsNew))
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
call IO_read_jobBinaryFile(777,'convergedTstar',trim(getSolverJobName()),size(Tstar))
read (777,rec=1) Tstar
close (777)
do i = 1, NpointsNew
TstarNew(1:6,1,1,i) = Tstar(1:6,1,1,indices(i))
enddo
call IO_write_jobBinaryFile(777,'convergedTstar',size(TstarNew))
write (777,rec=1) TstarNew
close (777)
deallocate(Tstar)
deallocate(TstarNew)
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
reworked neares neigbor search subroutine, now also works for voronoi tesselation (general nn search in periodic cell)
2012-05-04 18:37:37 +05:30
!----------------------------------------------------------------------------
regridding is now working, changed the subroutine into a function changed order of arrays in nearest neighbor search to make it fortran fast constitutive.f90 and homogenization.f90 write state size out during initialization setup/setup_processing.py is using byterecl to be compatible with binary files written out by solver
2012-05-08 20:27:06 +05:30
allocate(sizeStateConst(1,Npoints))
call IO_read_jobBinaryFile(777,'sizeStateConst',trim(getSolverJobName()),size(sizeStateConst))
read (777,rec=1) sizeStateConst
close (777)
maxsize = maxval(sizeStateConst)
allocate(StateConst (1,1,Npoints,maxsize))
call IO_read_jobBinaryFile(777,'convergedStateConst',trim(getSolverJobName()))
k = 0_pInt
do i =1, Npoints
do j = 1,sizeStateConst(1,i)
k = k+1_pInt
read(777,rec=k) StateConst(1,1,i,j)
enddo
enddo
close(777)
call IO_write_jobBinaryFile(777,'convergedStateConst')
k = 0_pInt
do i = 1,NpointsNew
do j = 1,sizeStateConst(1,i)
k=k+1_pInt
write(777,rec=k) StateConst(1,1,indices(i),j)
enddo
enddo
close (777)
deallocate(sizeStateConst)
deallocate(StateConst)
!----------------------------------------------------------------------------
allocate(sizeStateHomog(1,Npoints))
call IO_read_jobBinaryFile(777,'sizeStateHomog',trim(getSolverJobName()),size(sizeStateHomog))
read (777,rec=1) sizeStateHomog
close (777)
maxsize = maxval(sizeStateHomog)
allocate(stateHomog (1,1,Npoints,maxsize))
call IO_read_jobBinaryFile(777,'convergedStateHomog',trim(getSolverJobName()))
k = 0_pInt
do i =1, Npoints
do j = 1,sizeStateHomog(1,i)
k = k+1_pInt
read(777,rec=k) stateHomog(1,1,i,j)
enddo
enddo
close(777)
call IO_write_jobBinaryFile(777,'convergedStateHomog')
k = 0_pInt
do i = 1,NpointsNew
do j = 1,sizeStateHomog(1,i)
k=k+1_pInt
write(777,rec=k) stateHomog(1,1,indices(i),j)
enddo
enddo
close (777)
deallocate(sizeStateHomog)
deallocate(stateHomog)
deallocate(indices)
mesh_regrid = resNew
end function mesh_regrid
integrated subroutine for regridding into mesh.f90 and made it available for python
2012-04-10 20:45:46 +05:30
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
function mesh_spectral_getDimension(fileUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
use IO, only: &
several small corrections/polishing
2012-04-20 17:28:41 +05:30
IO_checkAndRewind, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_open_file, &
IO_stringPos, &
IO_lc, &
IO_stringValue, &
IO_intValue, &
IO_floatValue, &
IO_error
use DAMASK_interface, only: &
getModelName, &
InputFileExtension
implicit none
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
integer(pInt), dimension(1_pInt + 7_pInt*2_pInt) :: positions ! for a,b c + 3 values + keyword
integer(pInt), intent(in), optional :: fileUnit
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
integer(pInt) :: headerLength = 0_pInt
real(pReal), dimension(3) :: mesh_spectral_getDimension
character(len=1024) :: line, &
keyword
integer(pInt) :: i, j
logical :: gotDimension = .false.
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
integer(pInt) :: myUnit
if(.not. present(fileUnit)) then
myUnit = 289_pInt
call IO_open_file(myUnit,trim(getModelName())//InputFileExtension)
else
myUnit = fileUnit
endif
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
several small corrections/polishing
2012-04-20 17:28:41 +05:30
call IO_checkAndRewind(myUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
read(myUnit,'(a1024)') line
positions = IO_stringPos(line,2_pInt)
keyword = IO_lc(IO_StringValue(line,positions,2_pInt))
if (keyword(1:4) == 'head') then
headerLength = IO_intValue(line,positions,1_pInt) + 1_pInt
else
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
call IO_error(error_ID=841_pInt, ext_msg='mesh_spectral_getDimension')
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
endif
rewind(myUnit)
do i = 1_pInt, headerLength
read(myUnit,'(a1024)') line
positions = IO_stringPos(line,7_pInt)
select case ( IO_lc(IO_StringValue(line,positions,1)) )
case ('dimension')
gotDimension = .true.
do j = 2_pInt,6_pInt,2_pInt
select case (IO_lc(IO_stringValue(line,positions,j)))
case('x')
mesh_spectral_getDimension(1) = IO_floatValue(line,positions,j+1_pInt)
case('y')
mesh_spectral_getDimension(2) = IO_floatValue(line,positions,j+1_pInt)
case('z')
mesh_spectral_getDimension(3) = IO_floatValue(line,positions,j+1_pInt)
end select
enddo
end select
enddo
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
if(.not. present(fileUnit)) close(myUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
if (.not. gotDimension) &
call IO_error(error_ID = 845_pInt, ext_msg='dimension')
if (any(mesh_spectral_getDimension<=0.0_pReal)) &
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
call IO_error(error_ID = 844_pInt, ext_msg='mesh_spectral_getDimension')
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
end function mesh_spectral_getDimension
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
function mesh_spectral_getResolution(fileUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
use IO, only: &
several small corrections/polishing
2012-04-20 17:28:41 +05:30
IO_checkAndRewind, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_open_file, &
IO_stringPos, &
IO_lc, &
IO_stringValue, &
IO_intValue, &
IO_floatValue, &
IO_error
use DAMASK_interface, only: &
getModelName, &
InputFileExtension
implicit none
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
integer(pInt), dimension(1_pInt + 7_pInt*2_pInt) :: positions ! for a,b c + 3 values + keyword
integer(pInt), intent(in), optional :: fileUnit
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
integer(pInt) :: headerLength = 0_pInt
integer(pInt), dimension(3) :: mesh_spectral_getResolution
character(len=1024) :: line, &
keyword
integer(pInt) :: i, j
logical :: gotResolution = .false.
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
integer(pInt) :: myUnit
if(.not. present(fileUnit)) then
myUnit = 289_pInt
call IO_open_file(myUnit,trim(getModelName())//InputFileExtension)
else
myUnit = fileUnit
endif
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
several small corrections/polishing
2012-04-20 17:28:41 +05:30
call IO_checkAndRewind(myUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
read(myUnit,'(a1024)') line
positions = IO_stringPos(line,2_pInt)
keyword = IO_lc(IO_StringValue(line,positions,2_pInt))
if (keyword(1:4) == 'head') then
headerLength = IO_intValue(line,positions,1_pInt) + 1_pInt
else
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
call IO_error(error_ID=841_pInt, ext_msg='mesh_spectral_getResolution')
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
endif
rewind(myUnit)
do i = 1_pInt, headerLength
read(myUnit,'(a1024)') line
positions = IO_stringPos(line,7_pInt)
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
select case ( IO_lc(IO_StringValue(line,positions,1_pInt)) )
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
case ('resolution')
gotResolution = .true.
do j = 2_pInt,6_pInt,2_pInt
select case (IO_lc(IO_stringValue(line,positions,j)))
case('a')
mesh_spectral_getResolution(1) = IO_intValue(line,positions,j+1_pInt)
case('b')
mesh_spectral_getResolution(2) = IO_intValue(line,positions,j+1_pInt)
case('c')
mesh_spectral_getResolution(3) = IO_intValue(line,positions,j+1_pInt)
end select
enddo
end select
enddo
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
if(.not. present(fileUnit)) close(myUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
if (.not. gotResolution) &
call IO_error(error_ID = 845_pInt, ext_msg='resolution')
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
if((mod(mesh_spectral_getResolution(1),2_pInt)/=0_pInt .or. & ! must be a even number
mesh_spectral_getResolution(1) < 2_pInt .or. & ! and larger than 1
mod(mesh_spectral_getResolution(2),2_pInt)/=0_pInt .or. & ! -"-
mesh_spectral_getResolution(2) < 2_pInt .or. & ! -"-
(mod(mesh_spectral_getResolution(3),2_pInt)/=0_pInt .and. &
mesh_spectral_getResolution(3)/= 1_pInt)) .or. & ! third res might be 1
mesh_spectral_getResolution(3) < 1_pInt) &
call IO_error(error_ID = 843_pInt, ext_msg='mesh_spectral_getResolution')
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
end function mesh_spectral_getResolution
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
function mesh_spectral_getHomogenization(fileUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
use IO, only: &
several small corrections/polishing
2012-04-20 17:28:41 +05:30
IO_checkAndRewind, &
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
IO_open_file, &
IO_stringPos, &
IO_lc, &
IO_stringValue, &
IO_intValue, &
IO_error
use DAMASK_interface, only: &
getModelName, &
InputFileExtension
implicit none
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
integer(pInt), dimension(1_pInt + 7_pInt*2_pInt) :: positions ! for a, b, c + 3 values + keyword
integer(pInt), intent(in), optional :: fileUnit
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
integer(pInt) :: headerLength = 0_pInt
integer(pInt) :: mesh_spectral_getHomogenization
character(len=1024) :: line, &
keyword
integer(pInt) :: i, j
logical :: gotHomogenization = .false.
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
integer(pInt) :: myUnit
if(.not. present(fileUnit)) then
myUnit = 289_pInt
call IO_open_file(myUnit,trim(getModelName())//InputFileExtension)
else
myUnit = fileUnit
endif
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
several small corrections/polishing
2012-04-20 17:28:41 +05:30
call IO_checkAndRewind(myUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
read(myUnit,'(a1024)') line
positions = IO_stringPos(line,2_pInt)
keyword = IO_lc(IO_StringValue(line,positions,2_pInt))
if (keyword(1:4) == 'head') then
headerLength = IO_intValue(line,positions,1_pInt) + 1_pInt
else
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
call IO_error(error_ID=841_pInt, ext_msg='mesh_spectral_getHomogenization')
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
endif
rewind(myUnit)
do i = 1_pInt, headerLength
read(myUnit,'(a1024)') line
positions = IO_stringPos(line,7_pInt)
select case ( IO_lc(IO_StringValue(line,positions,1)) )
case ('homogenization')
gotHomogenization = .true.
mesh_spectral_getHomogenization = IO_intValue(line,positions,2_pInt)
end select
enddo
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
if(.not. present(fileUnit)) close(myUnit)
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
if (.not. gotHomogenization ) &
call IO_error(error_ID = 845_pInt, ext_msg='homogenization')
if (mesh_spectral_getHomogenization<1_pInt) &
corrected bug causing wrong element construction when using spectral solver. Also improved error messages for invalid resolution, dimension, and homogenization.
2012-05-21 14:36:02 +05:30
call IO_error(error_ID = 842_pInt, ext_msg='mesh_spectral_getHomogenization')
merged code for python and spectral solver interfacing (shared most of it anyway). put functionality for getting header information (dimension, homogenization, resolution) in functions in mesh.f90
2012-04-11 22:58:08 +05:30
end function mesh_spectral_getHomogenization
integrated subroutine for regridding into mesh.f90 and made it available for python
2012-04-10 20:45:46 +05:30
added new, flexible debugging scheme. now all modules have their own debug specification. compiles and runs, I hope nothing is broken did a lot of polishing
2012-03-09 01:55:28 +05:30
end module mesh