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

2738 lines
121 KiB
Fortran
Raw Normal View History

introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!> @author Franz Roters, Max-Planck-Institut für Eisenforschung GmbH
!> @author Philip Eisenlohr, Max-Planck-Institut für Eisenforschung GmbH
!> @author Christoph Koords, Max-Planck-Institut für Eisenforschung GmbH
!> @author Martin Diehl, Max-Planck-Institut für Eisenforschung GmbH
variable attributes adjusted compiles now
2019-01-28 19:06:44 +05:30
!> @brief Sets up the mesh for the solver MSC.Marc
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
makefile modified not to link against wrong libraries when using PETSc, reading in of values in phenopowerlaw improved, HDF5 in mesh
2014-04-29 23:20:59 +05:30
module mesh
moved some more 'mesh related' functions for post processing from math.f90 to mesh.f90 f2py functions remaining in math.f90 now uses assumed size arrays in order to have simpler interfaces. This is only working with python 2.7! changed python pre- and postprocessing scripts. If you encounter any problems whith core modules, try to remove the old core.so in the lib/damask
2012-08-27 13:34:47 +05:30
use, intrinsic :: iso_c_binding
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
tMesh_marc object extends tMesh Functionality seperated for mesh_marc
2019-01-28 18:23:44 +05:30
use mesh_base
moved calculation and output of geometry related data to mesh.f90, moved output of command line related information to DAMASK_spectral_interface.f90. improved new solver structure (use with make SOLVER=NEW)
2012-08-06 18:13:05 +05:30
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
divergence_correction for basic solver variants has now 3 possibilities: 0: uncorrected, slope per sidelength (physical dimension) e = res/dim 1: corrected by sidelength, slope per unitlength e = res/1 2: corrected such that distance between FPs e = 1 alway regarding the medium length of x,y,z direction
2013-01-16 16:10:53 +05:30
integer(pInt), public, protected :: &
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
mesh_NcpElems, & !< total number of CP elements in local mesh
using arrays with new names
2018-09-24 00:23:35 +05:30
mesh_elemType, & !< Element type of the mesh (only support homogeneous meshes)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
mesh_Nnodes, & !< total number of nodes in mesh
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
mesh_Ncellnodes, & !< total number of cell nodes in mesh (including duplicates)
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
mesh_Ncells, & !< total number of cells in mesh
introducing better names allows further simplifications as we do not store max and per elem values any more for number of integration points and number of cell nodes
2018-09-23 21:07:57 +05:30
mesh_NipsPerElem, & !< number of IPs in per element
mesh_NcellnodesPerElem, & !< number of cell nodes per element
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
mesh_maxNipNeighbors, & !< max number of IP neighbors in any CP element
introducing better names allows further simplifications as we do not store max and per elem values any more for number of integration points and number of cell nodes
2018-09-23 21:07:57 +05:30
mesh_maxNsharedElems !< max number of CP elements sharing a node
!!!! BEGIN DEPRECATED !!!!!
integer(pInt), public, protected :: &
mesh_maxNips, & !< max number of IPs in any CP element
Nelems /= NcpElems only in case of Abaqus/MSC.Marc
2018-09-23 20:35:01 +05:30
mesh_maxNcellnodes !< max number of cell nodes in any CP element
introducing better names allows further simplifications as we do not store max and per elem values any more for number of integration points and number of cell nodes
2018-09-23 21:07:57 +05:30
!!!! BEGIN DEPRECATED !!!!!
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
more descriptive names
2018-09-23 22:20:54 +05:30
integer(pInt), dimension(:), allocatable, public, protected :: &
consistent (and better understandable) names
2018-10-04 09:33:48 +05:30
mesh_homogenizationAt, & !< homogenization ID of each element
mesh_microstructureAt !< microstructure ID of each element
more descriptive names
2018-09-23 22:20:54 +05:30
divergence_correction for basic solver variants has now 3 possibilities: 0: uncorrected, slope per sidelength (physical dimension) e = res/dim 1: corrected by sidelength, slope per unitlength e = res/1 2: corrected such that distance between FPs e = 1 alway regarding the medium length of x,y,z direction
2013-01-16 16:10:53 +05:30
integer(pInt), dimension(:,:), allocatable, public, protected :: &
consistent (and better understandable) names
2018-10-04 09:33:48 +05:30
mesh_CPnodeID, & !< nodes forming an element
better name
2018-09-23 23:28:43 +05:30
mesh_element, & !DEPRECATED
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
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)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
divergence_correction for basic solver variants has now 3 possibilities: 0: uncorrected, slope per sidelength (physical dimension) e = res/dim 1: corrected by sidelength, slope per unitlength e = res/1 2: corrected such that distance between FPs e = 1 alway regarding the medium length of x,y,z direction
2013-01-16 16:10:53 +05:30
integer(pInt), dimension(:,:,:,:), allocatable, public, protected :: &
third entry in mesh_ipNeighborhood stores the neighbor_index that points from each neighbor back to the central ip; needed in nonlocal model
2012-10-24 19:33:02 +05:30
mesh_ipNeighborhood !< 6 or less neighboring IPs as [element_num, IP_index, neighbor_index that points to me]
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
"unitlength" parameter, which determines the physical size of the mesh, now available as a global mesh variable "mesh_unitlength" and written to the .mesh file during init. Hence, it is available to the marc_deformedGeometry.py script via "mesh_init_postprocessing()" and "mesh_get_unitlength()", so no need for setting the scaling of the displacement vectors explicitly through an option; now displacements and nodal positions are always consistent
2013-05-07 18:36:29 +05:30
real(pReal), public, protected :: &
mesh_unitlength !< physical length of one unit in mesh
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
real(pReal), dimension(:,:), allocatable, public :: &
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_node, & !< node x,y,z coordinates (after deformation! ONLY FOR MARC!!!)
mesh_cellnode !< cell node x,y,z coordinates (after deformation! ONLY FOR MARC!!!)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
divergence_correction for basic solver variants has now 3 possibilities: 0: uncorrected, slope per sidelength (physical dimension) e = res/dim 1: corrected by sidelength, slope per unitlength e = res/1 2: corrected such that distance between FPs e = 1 alway regarding the medium length of x,y,z direction
2013-01-16 16:10:53 +05:30
real(pReal), dimension(:,:), allocatable, public, protected :: &
mesh_ipVolume, & !< volume associated with IP (initially!)
mesh_node0 !< node x,y,z coordinates (initially!)
real(pReal), dimension(:,:,:), allocatable, public, protected :: &
changed all remaining routines to fortran-fast arrays (geometry reconstruction etc.) changed all remaining routines in f2py to more clever determination of array size (requires f2py >= 2.0) enabled 3D visualize to work with odd resolution by switching to linear reconstruction PLEASE NOTE: Redefinition of routines for f2py might cause trouble -> DELETE DAMASK_ROOT/lib/damask/core.so in this case further changes: added pure statement where possible, polished, unified use of "Q" for "Quaternion" and reordered math to have similar routines together
2013-01-31 21:58:08 +05:30
mesh_ipArea !< area of interface to neighboring IP (initially!)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
divergence_correction for basic solver variants has now 3 possibilities: 0: uncorrected, slope per sidelength (physical dimension) e = res/dim 1: corrected by sidelength, slope per unitlength e = res/1 2: corrected such that distance between FPs e = 1 alway regarding the medium length of x,y,z direction
2013-01-16 16:10:53 +05:30
real(pReal), dimension(:,:,:), allocatable, public :: &
mesh_ipCoordinates !< IP x,y,z coordinates (after deformation!)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
real(pReal),dimension(:,:,:,:), allocatable, public, protected :: &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
mesh_ipAreaNormal !< area normal of interface to neighboring IP (initially!)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
divergence_correction for basic solver variants has now 3 possibilities: 0: uncorrected, slope per sidelength (physical dimension) e = res/dim 1: corrected by sidelength, slope per unitlength e = res/1 2: corrected such that distance between FPs e = 1 alway regarding the medium length of x,y,z direction
2013-01-16 16:10:53 +05:30
logical, dimension(3), public, protected :: mesh_periodicSurface !< flag indicating periodic outer surfaces (used for fluxes)
preparation for introduction of plain mode and restructuring of calc modes for CPFEM_general
2013-02-28 02:11:14 +05:30
polishing and encapsulating of Abaqus-specific local variables
2018-10-10 03:22:54 +05:30
integer(pInt), private :: &
mesh_maxNelemInSet, &
mesh_Nmaterials
grouping solver specific variables for better readability
2018-09-23 18:57:51 +05:30
integer(pInt), dimension(:,:), allocatable, private :: &
introduced preprocessor statements in mesh to hide unneeded variables as a preparation for HDF5
2013-12-28 01:33:28 +05:30
mesh_cellnodeParent !< cellnode's parent element ID, cellnode's intra-element ID
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt),dimension(:,:,:), allocatable, private :: &
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_cell !< cell connectivity for each element,ip/cell
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(:,:,:), allocatable, private :: &
FE_nodesAtIP, & !< map IP index to 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_cell, & !< list of intra-element cell node IDs that constitute the cells in a specific type of element geometry
FE_cellface !< list of intra-cell cell node IDs that constitute the cell faces of a specific type of cell
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
real(pReal), dimension(:,:,:), allocatable, private :: &
FE_cellnodeParentnodeWeights !< list of node weights for the generation of cell nodes
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(:,:,:,:), allocatable, private :: &
FE_subNodeOnIPFace
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
! These 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 :: &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_Nelemtypes = 13_pInt, &
FE_Ngeomtypes = 10_pInt, &
FE_Ncelltypes = 4_pInt, &
FE_maxNnodes = 20_pInt, &
FE_maxNips = 27_pInt, &
FE_maxNipNeighbors = 6_pInt, &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
FE_maxmaxNnodesAtIP = 8_pInt, & !< max number of (equivalent) nodes attached to an IP
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_maxNmatchingNodesPerFace = 4_pInt, &
FE_maxNfaces = 6_pInt, &
FE_maxNcellnodes = 64_pInt, &
FE_maxNcellnodesPerCell = 8_pInt, &
FE_maxNcellfaces = 6_pInt, &
FE_maxNcellnodesPerCellface = 4_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter, public :: FE_geomtype = & !< geometry type of particular element type
int([ &
1, & ! element 6 (2D 3node 1ip)
2, & ! element 125 (2D 6node 3ip)
3, & ! element 11 (2D 4node 4ip)
4, & ! element 27 (2D 8node 9ip)
- added element type 54 for macro (2d 8node 4ip reduced integration) - automatically raising error for unknown element type inside mapping function
2013-04-10 15:08:40 +05:30
3, & ! element 54 (2D 8node 4ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
5, & ! element 134 (3D 4node 1ip)
6, & ! element 157 (3D 5node 4ip)
6, & ! element 127 (3D 10node 4ip)
7, & ! element 136 (3D 6node 6ip)
8, & ! element 117 (3D 8node 1ip)
9, & ! element 7 (3D 8node 8ip)
9, & ! element 57 (3D 20node 8ip)
10 & ! element 21 (3D 20node 27ip)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of 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)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(FE_Ngeomtypes), parameter, public :: FE_celltype = & !< cell type that is used by each geometry type
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
int([ &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, & ! element 6 (2D 3node 1ip)
2, & ! element 125 (2D 6node 3ip)
2, & ! element 11 (2D 4node 4ip)
2, & ! element 27 (2D 8node 9ip)
3, & ! element 134 (3D 4node 1ip)
4, & ! element 127 (3D 10node 4ip)
4, & ! element 136 (3D 6node 6ip)
4, & ! element 117 (3D 8node 1ip)
4, & ! element 7 (3D 8node 8ip)
4 & ! element 21 (3D 20node 27ip)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of 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)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(FE_Ngeomtypes), parameter, public :: FE_dimension = & !< dimension of geometry type
introduced dedicated calculation of interface normals in IP neighborhood. (2D elements did not work so far..!) fixed definitions of some (2D) interface planes to have all normals pointing outward. removed "int, private :: i" defined for spectral compilation.
2013-02-05 18:57:37 +05:30
int([ &
2, & ! element 6 (2D 3node 1ip)
2, & ! element 125 (2D 6node 3ip)
2, & ! element 11 (2D 4node 4ip)
2, & ! element 27 (2D 8node 9ip)
3, & ! element 134 (3D 4node 1ip)
3, & ! element 127 (3D 10node 4ip)
3, & ! element 136 (3D 6node 6ip)
3, & ! element 117 (3D 8node 1ip)
3, & ! element 7 (3D 8node 8ip)
3 & ! element 21 (3D 20node 27ip)
],pInt)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(FE_Nelemtypes), parameter, public :: FE_Nnodes = & !< number of nodes that constitute a specific type of element
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
int([ &
3, & ! element 6 (2D 3node 1ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
6, & ! element 125 (2D 6node 3ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
4, & ! element 11 (2D 4node 4ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
8, & ! element 27 (2D 8node 9ip)
8, & ! element 54 (2D 8node 4ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
4, & ! element 134 (3D 4node 1ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
5, & ! element 157 (3D 5node 4ip)
10, & ! element 127 (3D 10node 4ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
6, & ! element 136 (3D 6node 6ip)
8, & ! element 117 (3D 8node 1ip)
8, & ! element 7 (3D 8node 8ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
20, & ! element 57 (3D 20node 8ip)
20 & ! element 21 (3D 20node 27ip)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of 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)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(FE_Ngeomtypes), parameter, public :: FE_Nfaces = & !< number of faces of a specific type of element geometry
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
int([ &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
3, & ! element 6 (2D 3node 1ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
3, & ! element 125 (2D 6node 3ip)
4, & ! element 11 (2D 4node 4ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
4, & ! element 27 (2D 8node 9ip)
4, & ! element 134 (3D 4node 1ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
4, & ! element 127 (3D 10node 4ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
5, & ! element 136 (3D 6node 6ip)
6, & ! element 117 (3D 8node 1ip)
6, & ! element 7 (3D 8node 8ip)
6 & ! element 21 (3D 20node 27ip)
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of 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)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(FE_Ngeomtypes), parameter, private :: FE_NmatchingNodes = & !< number of nodes that are needed for face matching in a specific type of element geometry
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
int([ &
3, & ! element 6 (2D 3node 1ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
3, & ! element 125 (2D 6node 3ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
4, & ! element 11 (2D 4node 4ip)
4, & ! element 27 (2D 8node 9ip)
4, & ! element 134 (3D 4node 1ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
4, & ! element 127 (3D 10node 4ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
6, & ! element 136 (3D 6node 6ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
8, & ! element 117 (3D 8node 1ip)
8, & ! element 7 (3D 8node 8ip)
8 & ! element 21 (3D 20node 27ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
],pInt)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(FE_maxNfaces,FE_Ngeomtypes), parameter, private :: &
FE_NmatchingNodesPerFace = & !< number of matching nodes per face in a specific type of element geometry
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([ &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
2,2,2,0,0,0, & ! element 6 (2D 3node 1ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
2,2,2,0,0,0, & ! element 125 (2D 6node 3ip)
2,2,2,2,0,0, & ! element 11 (2D 4node 4ip)
2,2,2,2,0,0, & ! element 27 (2D 8node 9ip)
3,3,3,3,0,0, & ! element 134 (3D 4node 1ip)
3,3,3,3,0,0, & ! element 127 (3D 10node 4ip)
3,4,4,4,3,0, & ! element 136 (3D 6node 6ip)
4,4,4,4,4,4, & ! element 117 (3D 8node 1ip)
4,4,4,4,4,4, & ! element 7 (3D 8node 8ip)
4,4,4,4,4,4 & ! element 21 (3D 20node 27ip)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
],pInt),[FE_maxNipNeighbors,FE_Ngeomtypes])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(FE_maxNmatchingNodesPerFace,FE_maxNfaces,FE_Ngeomtypes), &
parameter, private :: FE_face = & !< List of node indices on each face of a specific type of element geometry
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of 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([&
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1,2,0,0 , & ! element 6 (2D 3node 1ip)
2,3,0,0 , &
3,1,0,0 , &
corrected mesh_build_elements (unit) so that state variables are read correctly from input file
2007-04-25 13:03:24 +05:30
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 , &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
0,0,0,0 , &
1,2,0,0 , & ! element 125 (2D 6node 3ip)
2,3,0,0 , &
3,1,0,0 , &
0,0,0,0 , &
0,0,0,0 , &
0,0,0,0 , &
1,2,0,0 , & ! element 11 (2D 4node 4ip)
added element types 11 (4node 2D linear full) and 27 (8node 2D quadratic full)
2007-10-12 19:18:29 +05:30
2,3,0,0 , &
3,4,0,0 , &
4,1,0,0 , &
0,0,0,0 , &
0,0,0,0 , &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1,2,0,0 , & ! element 27 (2D 8node 9ip)
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
2,3,0,0 , &
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 , &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1,2,3,0 , & ! element 134 (3D 4node 1ip)
added two more element types (157 and 136)
2008-06-17 14:41:54 +05:30
1,4,2,0 , &
2,3,4,0 , &
1,3,4,0 , &
0,0,0,0 , &
0,0,0,0 , &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1,2,3,0 , & ! element 127 (3D 10node 4ip)
1,4,2,0 , &
2,4,3,0 , &
1,3,4,0 , &
0,0,0,0 , &
0,0,0,0 , &
1,2,3,0 , & ! element 136 (3D 6node 6ip)
added two more element types (157 and 136)
2008-06-17 14:41:54 +05:30
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 , &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1,2,3,4 , & ! element 117 (3D 8node 1ip)
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
2,1,5,6 , &
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 , &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1,2,3,4 , & ! element 7 (3D 8node 8ip)
reworked build_ipNeighborhood added C3D8R reduced integration hexahedral element
2010-05-10 20:24:59 +05:30
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 , &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1,2,3,4 , & ! element 21 (3D 20node 27ip)
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
2,1,5,6 , &
3,2,6,7 , &
4,3,7,8 , &
4,1,5,8 , &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
8,7,6,5 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_maxNmatchingNodesPerFace,FE_maxNfaces,FE_Ngeomtypes])
integer(pInt), dimension(FE_Ngeomtypes), parameter, private :: FE_Ncellnodes = & !< number of cell nodes in a specific geometry type
int([ &
3, & ! element 6 (2D 3node 1ip)
7, & ! element 125 (2D 6node 3ip)
9, & ! element 11 (2D 4node 4ip)
fixed bug in cell generation: "FE_Ncellnodes" was not correct for elem type 17 and 136 too much memory allocated for mesh_cell
2013-04-18 23:26:14 +05:30
16, & ! element 27 (2D 8node 9ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
4, & ! element 134 (3D 4node 1ip)
15, & ! element 127 (3D 10node 4ip)
fixed bug in cell generation: "FE_Ncellnodes" was not correct for elem type 17 and 136 too much memory allocated for mesh_cell
2013-04-18 23:26:14 +05:30
21, & ! element 136 (3D 6node 6ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
8, & ! element 117 (3D 8node 1ip)
27, & ! element 7 (3D 8node 8ip)
64 & ! element 21 (3D 20node 27ip)
],pInt)
integer(pInt), dimension(FE_Ncelltypes), parameter, private :: FE_NcellnodesPerCell = & !< number of cell nodes in a specific cell type
int([ &
3, & ! (2D 3node)
4, & ! (2D 4node)
4, & ! (3D 4node)
8 & ! (3D 8node)
],pInt)
integer(pInt), dimension(FE_Ncelltypes), parameter, private :: FE_NcellnodesPerCellface = & !< number of cell nodes per cell face in a specific cell type
int([&
2, & ! (2D 3node)
2, & ! (2D 4node)
3, & ! (3D 4node)
4 & ! (3D 8node)
],pInt)
integer(pInt), dimension(FE_Ngeomtypes), parameter, public :: FE_Nips = & !< number of IPs in a specific type of element
int([ &
1, & ! element 6 (2D 3node 1ip)
3, & ! element 125 (2D 6node 3ip)
4, & ! element 11 (2D 4node 4ip)
9, & ! element 27 (2D 8node 9ip)
1, & ! element 134 (3D 4node 1ip)
4, & ! element 127 (3D 10node 4ip)
6, & ! element 136 (3D 6node 6ip)
1, & ! element 117 (3D 8node 1ip)
8, & ! element 7 (3D 8node 8ip)
27 & ! element 21 (3D 20node 27ip)
],pInt)
integer(pInt), dimension(FE_Ncelltypes), parameter, public :: FE_NipNeighbors = & !< number of ip neighbors / cell faces in a specific cell type
int([&
3, & ! (2D 3node)
4, & ! (2D 4node)
4, & ! (3D 4node)
6 & ! (3D 8node)
],pInt)
integer(pInt), dimension(FE_Ngeomtypes), parameter, private :: FE_maxNnodesAtIP = & !< maximum number of parent nodes that belong to an IP for a specific type of element
int([ &
3, & ! element 6 (2D 3node 1ip)
1, & ! element 125 (2D 6node 3ip)
1, & ! element 11 (2D 4node 4ip)
2, & ! element 27 (2D 8node 9ip)
4, & ! element 134 (3D 4node 1ip)
1, & ! element 127 (3D 10node 4ip)
1, & ! element 136 (3D 6node 6ip)
8, & ! element 117 (3D 8node 1ip)
1, & ! element 7 (3D 8node 8ip)
4 & ! element 21 (3D 20node 27ip)
],pInt)
only needed for commercial solvers
2018-09-23 19:01:19 +05:30
integer(pInt), private :: &
Nelems /= NcpElems only in case of Abaqus/MSC.Marc
2018-09-23 20:35:01 +05:30
mesh_Nelems, & !< total number of elements in mesh (including non-DAMASK elements)
not needed for spectral
2018-09-23 19:27:21 +05:30
mesh_maxNnodes, & !< max number of nodes in any CP element
only needed for commercial solvers
2018-09-23 19:01:19 +05:30
mesh_NelemSets
grouping solver specific variables for better readability
2018-09-23 18:57:51 +05:30
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]
remove non-marc specific code
2019-01-24 14:34:43 +05:30
grouping solver specific variables for better readability
2018-09-23 18:57:51 +05:30
integer(pInt), private :: &
MarcVersion, & !< Version of input file format (Marc only)
hypoelasticTableStyle, & !< Table style (Marc only)
initialcondTableStyle !< Table style (Marc only)
integer(pInt), dimension(:), allocatable, private :: &
Marc_matNumber !< array of material numbers for hypoelastic material (Marc only)
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
public :: &
mesh_init, &
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_build_cellnodes, &
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
mesh_build_ipVolumes, &
mesh_build_ipCoordinates, &
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_cellCenterCoordinates, &
consistent order of solver-specific functionality
2018-09-23 21:27:48 +05:30
mesh_FEasCP
remove non-marc specific code
2019-01-24 14:34:43 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
private :: &
consistent order of solver-specific functionality
2018-09-23 21:27:48 +05:30
mesh_get_damaskOptions, &
mesh_build_cellconnectivity, &
mesh_build_ipAreas, &
FE_mapElemtype, &
mesh_faceMatch, &
mesh_build_FEdata, &
mesh_build_nodeTwins, &
mesh_build_sharedElems, &
mesh_build_ipNeighborhood, &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
mesh_marc_get_fileFormat, &
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
mesh_marc_get_tableStyles, &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
mesh_marc_get_matNumber, &
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
mesh_marc_count_nodesAndElements, &
mesh_marc_count_elementSets, &
mesh_marc_map_elementSets, &
mesh_marc_count_cpElements, &
mesh_marc_map_Elements, &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
mesh_marc_map_nodes, &
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
mesh_marc_build_nodes, &
mesh_marc_count_cpSizes, &
consistent order of solver-specific functionality
2018-09-23 21:27:48 +05:30
mesh_marc_build_elements
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
tMesh_marc object extends tMesh Functionality seperated for mesh_marc
2019-01-28 18:23:44 +05:30
type, public, extends(tMesh) :: tMesh_marc
integer(pInt), public :: &
nElemsAll, &
maxNelemInSet, &
NelemSets,&
MarcVersion, & !< Version of input file format ToDo: Better Name?
hypoelasticTableStyle, & !< Table style
initialcondTableStyle
character(len=64), dimension(:), allocatable :: &
nameElemSet,& !< names of elementSet
mesh_nameElemSet, & !< names of elementSet
mapMaterial !< name of elementSet for material
integer(pInt), dimension(:), allocatable :: &
Marc_matNumber !< array of material numbers for hypoelastic material (Marc only)
variable attributes adjusted compiles now
2019-01-28 19:06:44 +05:30
integer(pInt) :: &
tMesh_marc object extends tMesh Functionality seperated for mesh_marc
2019-01-28 18:23:44 +05:30
mesh_Nelems, & !< total number of elements in mesh (including non-DAMASK elements)
mesh_maxNnodes, & !< max number of nodes in any CP element
mesh_NelemSets, &
mesh_maxNelemInSet
integer(pInt), dimension(:,:), allocatable :: &
mesh_mapElemSet !< list of elements in elementSet
all variables/functions were not used
2019-01-31 16:53:23 +05:30
initialize mesh and element
2019-02-01 16:54:23 +05:30
contains
procedure, pass(self) :: tMesh_marc_init
generic, public :: init => tMesh_marc_init
tMesh_marc object extends tMesh Functionality seperated for mesh_marc
2019-01-28 18:23:44 +05:30
end type tMesh_marc
type(tMesh_marc), public, protected :: theMesh
variable attributes adjusted compiles now
2019-01-28 19:06:44 +05:30
tMesh_marc object extends tMesh Functionality seperated for mesh_marc
2019-01-28 18:23:44 +05:30
contains
initialize mesh and element
2019-02-01 16:54:23 +05:30
subroutine tMesh_marc_init(self,elemType,nodes)
tMesh_marc object extends tMesh Functionality seperated for mesh_marc
2019-01-28 18:23:44 +05:30
implicit none
class(tMesh_marc) :: self
initialize mesh and element
2019-02-01 16:54:23 +05:30
real(pReal), dimension(:,:), intent(in) :: nodes
integer(pInt), intent(in) :: elemType
call self%tMesh%init('mesh',elemType,nodes)
tMesh_marc object extends tMesh Functionality seperated for mesh_marc
2019-01-28 18:23:44 +05:30
end subroutine tMesh_marc_init
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief initializes the mesh by calling all necessary private routines the mesh module
!! Order and routines strongly depend on type of solver
!--------------------------------------------------------------------------------------------------
introduced dedicated calculation of interface normals in IP neighborhood. (2D elements did not work so far..!) fixed definitions of some (2D) interface planes to have all normals pointing outward. removed "int, private :: i" defined for spectral compilation.
2013-02-05 18:57:37 +05:30
subroutine mesh_init(ip,el)
compiler_options and compiler_version supported by new Intel compiler
2018-02-02 17:06:09 +05:30
#if defined(__GFORTRAN__) || __INTEL_COMPILER >= 1800
suppress warning when compiling with gfortran
2017-10-05 20:05:34 +05:30
use, intrinsic :: iso_fortran_env, only: &
compiler_version, &
compiler_options
#endif
no need to inlcude twice
2017-02-05 04:23:22 +05:30
use DAMASK_interface
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
use IO, only: &
use statement for marc was wrong (preprocessor)
2015-04-11 10:39:15 +05:30
IO_open_InputFile, &
IO_timeStamp, &
IO_error, &
IO_write_jobFile
introduced error when element/IP/component(grain) selected for debugging does not exist
2013-10-16 18:08:00 +05:30
use debug, only: &
debug_e, &
added some debug statements
2014-12-19 00:11:02 +05:30
debug_i, &
debug_level, &
debug_mesh, &
debug_levelBasic
further improvements on implementing the plain mode
2013-04-16 22:37:27 +05:30
use numerics, only: &
"unitlength" parameter, which determines the physical size of the mesh, now available as a global mesh variable "mesh_unitlength" and written to the .mesh file during init. Hence, it is available to the marc_deformedGeometry.py script via "mesh_init_postprocessing()" and "mesh_get_unitlength()", so no need for setting the scaling of the displacement vectors explicitly through an option; now displacements and nodal positions are always consistent
2013-05-07 18:36:29 +05:30
usePingPong, &
added some debug statements
2014-12-19 00:11:02 +05:30
numerics_unitlength, &
worldrank
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
use FEsolving, only: &
took some gfortran complaints serious (unused imports, implicit castings)
2015-04-11 17:17:33 +05:30
modelName, &
syntax errors in declaration
2018-09-23 22:34:17 +05:30
calcMode, &
FEsolving_execElem, &
FEsolving_execIP
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
added function calls in init()
2007-04-10 16:52:53 +05:30
implicit none
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), parameter :: FILEUNIT = 222_pInt
calcMode not needed for spectral and FEM
2018-09-24 01:01:30 +05:30
integer(pInt), intent(in), optional :: el, ip
improved reading in of values, now only warnings in case of problematic entries in material.config divergence calculation sqrt scaling optionally introduced for basic scheme spectral solver
2013-02-08 21:25:53 +05:30
integer(pInt) :: j
added some debug statements
2014-12-19 00:11:02 +05:30
logical :: myDebug
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
removed old regridding related stuff
2016-06-29 20:29:22 +05:30
write(6,'(/,a)') ' <<<+- mesh init -+>>>'
write(6,'(a15,a)') ' Current time: ',IO_timeStamp()
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"
removed unnecsessary omp statements, removed one transpose in constitutitve
2013-01-08 16:39:20 +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
"unitlength" parameter, which determines the physical size of the mesh, now available as a global mesh variable "mesh_unitlength" and written to the .mesh file during init. Hence, it is available to the marc_deformedGeometry.py script via "mesh_init_postprocessing()" and "mesh_get_unitlength()", so no need for setting the scaling of the displacement vectors explicitly through an option; now displacements and nodal positions are always consistent
2013-05-07 18:36:29 +05:30
mesh_unitlength = numerics_unitlength ! set physical extent of a length unit in mesh
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
added some debug statements
2014-12-19 00:11:02 +05:30
myDebug = (iand(debug_level(debug_mesh),debug_levelBasic) /= 0_pInt)
"unitlength" parameter, which determines the physical size of the mesh, now available as a global mesh variable "mesh_unitlength" and written to the .mesh file during init. Hence, it is available to the marc_deformedGeometry.py script via "mesh_init_postprocessing()" and "mesh_get_unitlength()", so no need for setting the scaling of the displacement vectors explicitly through an option; now displacements and nodal positions are always consistent
2013-05-07 18:36:29 +05:30
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call IO_open_inputFile(FILEUNIT,modelName) ! parse info from input file...
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Opened input file'; flush(6)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
call mesh_marc_get_fileFormat(FILEUNIT)
if (myDebug) write(6,'(a)') ' Got input file format'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_get_tableStyles(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Got table styles'; flush(6)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
if (MarcVersion > 12) then
call mesh_marc_get_matNumber(FILEUNIT)
if (myDebug) write(6,'(a)') ' Got hypoleastic material number'; flush(6)
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
endif
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_count_nodesAndElements(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Counted nodes/elements'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_count_elementSets(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Counted element sets'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_map_elementSets(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Mapped element sets'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_count_cpElements(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Counted CP elements'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_map_elements(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Mapped elements'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_map_nodes(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Mapped nodes'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_build_nodes(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built nodes'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_count_cpSizes(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Counted CP sizes'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_marc_build_elements(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built elements'; flush(6)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
call mesh_get_damaskOptions(FILEUNIT)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Got DAMASK options'; flush(6)
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
call mesh_build_cellconnectivity
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built cell connectivity'; flush(6)
added python script that produces vtk files with the (node based and ip based) deformed mesh from marc output file
2013-04-24 00:00:56 +05:30
mesh_cellnode = mesh_build_cellnodes(mesh_node,mesh_Ncellnodes)
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built cell nodes'; 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
call mesh_build_ipCoordinates
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built IP coordinates'; 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
call mesh_build_ipVolumes
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built IP volumes'; 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
call mesh_build_ipAreas
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built IP areas'; flush(6)
some simplifications as discussed with IBF colleagues
2015-11-26 02:25:17 +05:30
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_nodeTwins
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built node twins'; 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
call mesh_build_sharedElems
added some debug statements
2014-12-19 00:11:02 +05:30
if (myDebug) write(6,'(a)') ' Built shared elements'; 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
call mesh_build_ipNeighborhood
some simplifications as discussed with IBF colleagues
2015-11-26 02:25:17 +05:30
if (myDebug) write(6,'(a)') ' Built IP neighborhood'; flush(6)
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
introduced error when element/IP/component(grain) selected for debugging does not exist
2013-10-16 18:08:00 +05:30
if (usePingPong .and. (mesh_Nelems /= mesh_NcpElems)) &
added few comments
2014-01-17 07:08:35 +05:30
call IO_error(600_pInt) ! ping-pong must be disabled when having non-DAMASK elements
introduced error when element/IP/component(grain) selected for debugging does not exist
2013-10-16 18:08:00 +05:30
if (debug_e < 1 .or. debug_e > mesh_NcpElems) &
call IO_error(602_pInt,ext_msg='element') ! selected element does not exist
if (debug_i < 1 .or. debug_i > FE_Nips(FE_geomtype(mesh_element(2_pInt,debug_e)))) &
call IO_error(602_pInt,ext_msg='IP') ! selected element does not have requested IP
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
added few comments
2014-01-17 07:08:35 +05:30
FEsolving_execElem = [ 1_pInt,mesh_NcpElems ] ! parallel loop bounds set to comprise all DAMASK elements
leaner syntax with sourced allocation
2018-09-23 20:56:13 +05:30
allocate(FEsolving_execIP(2_pInt,mesh_NcpElems), source=1_pInt) ! parallel loop bounds set to comprise from first IP...
added few comments
2014-01-17 07:08:35 +05:30
forall (j = 1_pInt:mesh_NcpElems) FEsolving_execIP(2,j) = FE_Nips(FE_geomtype(mesh_element(2,j))) ! ...up to own IP count for each element
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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))
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
calcMode = .false. ! pretend to have collected what first call is asking (F = I)
introduced dedicated calculation of interface normals in IP neighborhood. (2D elements did not work so far..!) fixed definitions of some (2D) interface planes to have all normals pointing outward. removed "int, private :: i" defined for spectral compilation.
2013-02-05 18:57:37 +05:30
calcMode(ip,mesh_FEasCP('elem',el)) = .true. ! first ip,el needs to be already pingponged to "calc"
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
introducing better names allows further simplifications as we do not store max and per elem values any more for number of integration points and number of cell nodes
2018-09-23 21:07:57 +05:30
!!!! COMPATIBILITY HACK !!!!
! for a homogeneous mesh, all elements have the same number of IPs and and cell nodes.
! hence, xxPerElem instead of maxXX
mesh_NipsPerElem = mesh_maxNips
mesh_NcellnodesPerElem = mesh_maxNcellnodes
more descriptive names
2018-09-23 22:20:54 +05:30
! better name
consistent (and better understandable) names
2018-10-04 09:33:48 +05:30
mesh_homogenizationAt = mesh_element(3,:)
mesh_microstructureAt = mesh_element(4,:)
better name
2018-09-23 23:28:43 +05:30
mesh_CPnodeID = mesh_element(5:4+mesh_NipsPerElem,:)
introducing better names allows further simplifications as we do not store max and per elem values any more for number of integration points and number of cell nodes
2018-09-23 21:07:57 +05:30
!!!!!!!!!!!!!!!!!!!!!!!!
initialize mesh and element
2019-02-01 16:54:23 +05:30
call theMesh%init(mesh_element(2,1),mesh_node0)
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
some minor polishing on the way
2014-04-04 20:10:30 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief Gives the FE to CP ID mapping by binary search through lookup array
!! valid questions (what) 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)
allowed arbitrary grid (formerly know as resolution), i.e. any integer>1 in each dimension is possible. removed square root correction for divergence (doesn't make sense) fixed FFT debug of spectral solver
2013-04-10 15:49:16 +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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
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)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
if (lookupMap(1_pInt,lower) == myID) then ! check at bounds QUESTION is it valid to extend bounds by 1 and just do binary search w/o init check at bounds?
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of derived divergence measures and added RMS output in brackets added comments and options to the makefile
2012-02-16 00:28:38 +05:30
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
removed temperature integration and corresponding data structures and debugging options temperature is stored in crystallite, but homogeneous on one IP (not an component (grain) quantity and an input value parsed in by the BVP solver. introduced heat, a component (grain) quantity which is homogenized before returned to the heat transfer solver. went ahead with removal of dummy functions in homogenization and constitutive, this time mainly reduced function signatures to reflect actually needed quantities.
2013-10-16 18:34:59 +05:30
binarySearch: do while (upper-lower > 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
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
removed temperature integration and corresponding data structures and debugging options temperature is stored in crystallite, but homogeneous on one IP (not an component (grain) quantity and an input value parsed in by the BVP solver. introduced heat, a component (grain) quantity which is homogenized before returned to the heat transfer solver. went ahead with removal of dummy functions in homogenization and constitutive, this time mainly reduced function signatures to reflect actually needed quantities.
2013-10-16 18:34:59 +05:30
enddo binarySearch
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!> @brief Split CP elements into cells.
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!> @details Build a mapping between cells and the corresponding cell nodes ('mesh_cell').
!> Cell nodes that are also matching nodes are unique in the list of cell nodes,
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
!> all others (currently) might be stored more than once.
!> Also allocates the 'mesh_node' array.
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
subroutine mesh_build_cellconnectivity
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
implicit none
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(:), allocatable :: &
matchingNode2cellnode
integer(pInt), dimension(:,:), allocatable :: &
cellnodeParent
integer(pInt), dimension(mesh_maxNcellnodes) :: &
localCellnode2globalCellnode
introduced error when element/IP/component(grain) selected for debugging does not exist
2013-10-16 18:08:00 +05:30
integer(pInt) :: &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
e,t,g,c,n,i, &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
matchingNodeID, &
localCellnodeID
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
added some code parts for HDF5 in Fortran
2013-12-30 15:36:01 +05:30
allocate(mesh_cell(FE_maxNcellnodesPerCell,mesh_maxNips,mesh_NcpElems), source=0_pInt)
allocate(matchingNode2cellnode(mesh_Nnodes), source=0_pInt)
allocate(cellnodeParent(2_pInt,mesh_maxNcellnodes*mesh_NcpElems), source=0_pInt)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced error when element/IP/component(grain) selected for debugging does not exist
2013-10-16 18:08:00 +05:30
!--------------------------------------------------------------------------------------------------
! Count cell nodes (including duplicates) and generate cell connectivity list
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_Ncellnodes = 0_pInt
mesh_Ncells = 0_pInt
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
t = mesh_element(2_pInt,e) ! get element type
g = FE_geomtype(t) ! get geometry type
c = FE_celltype(g) ! get cell type
localCellnode2globalCellnode = 0_pInt
mesh_Ncells = mesh_Ncells + FE_Nips(g)
do i = 1_pInt,FE_Nips(g) ! loop over ips=cells in this element
do n = 1_pInt,FE_NcellnodesPerCell(c) ! loop over cell nodes in this cell
localCellnodeID = FE_cell(n,i,g)
if (localCellnodeID <= FE_NmatchingNodes(g)) then ! this cell node is a matching node
matchingNodeID = mesh_element(4_pInt+localCellnodeID,e)
if (matchingNode2cellnode(matchingNodeID) == 0_pInt) then ! if this matching node does not yet exist in the glbal cell node list ...
mesh_Ncellnodes = mesh_Ncellnodes + 1_pInt ! ... count it as cell node ...
matchingNode2cellnode(matchingNodeID) = mesh_Ncellnodes ! ... and remember its global ID
cellnodeParent(1_pInt,mesh_Ncellnodes) = e ! ... and where it belongs to
cellnodeParent(2_pInt,mesh_Ncellnodes) = localCellnodeID
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
endif
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_cell(n,i,e) = matchingNode2cellnode(matchingNodeID)
else ! this cell node is no matching node
if (localCellnode2globalCellnode(localCellnodeID) == 0_pInt) then ! if this local cell node does not yet exist in the global cell node list ...
mesh_Ncellnodes = mesh_Ncellnodes + 1_pInt ! ... count it as cell node ...
localCellnode2globalCellnode(localCellnodeID) = mesh_Ncellnodes ! ... and remember its global ID ...
cellnodeParent(1_pInt,mesh_Ncellnodes) = e ! ... and it belongs to
cellnodeParent(2_pInt,mesh_Ncellnodes) = localCellnodeID
endif
mesh_cell(n,i,e) = localCellnode2globalCellnode(localCellnodeID)
endif
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
enddo
enddo
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
enddo
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
allocate(mesh_cellnodeParent(2_pInt,mesh_Ncellnodes))
allocate(mesh_cellnode(3_pInt,mesh_Ncellnodes))
forall(n = 1_pInt:mesh_Ncellnodes)
mesh_cellnodeParent(1,n) = cellnodeParent(1,n)
mesh_cellnodeParent(2,n) = cellnodeParent(2,n)
endforall
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
end subroutine mesh_build_cellconnectivity
!--------------------------------------------------------------------------------------------------
!> @brief Calculate position of cellnodes from the given position of nodes
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!> Build list of cellnodes' coordinates.
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
!> Cellnode coordinates are calculated from a weighted sum of node coordinates.
!--------------------------------------------------------------------------------------------------
added python script that produces vtk files with the (node based and ip based) deformed mesh from marc output file
2013-04-24 00:00:56 +05:30
function mesh_build_cellnodes(nodes,Ncellnodes)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
implicit none
introduced error when element/IP/component(grain) selected for debugging does not exist
2013-10-16 18:08:00 +05:30
integer(pInt), intent(in) :: Ncellnodes !< requested number of cellnodes
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
real(pReal), dimension(3,mesh_Nnodes), intent(in) :: nodes
added python script that produces vtk files with the (node based and ip based) deformed mesh from marc output file
2013-04-24 00:00:56 +05:30
real(pReal), dimension(3,Ncellnodes) :: mesh_build_cellnodes
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
introduced error when element/IP/component(grain) selected for debugging does not exist
2013-10-16 18:08:00 +05:30
integer(pInt) :: &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
e,t,n,m, &
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
localCellnodeID
real(pReal), dimension(3) :: &
myCoords
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_build_cellnodes = 0.0_pReal
fixed bug in cell generation: "FE_Ncellnodes" was not correct for elem type 17 and 136 too much memory allocated for mesh_cell
2013-04-18 23:26:14 +05:30
!$OMP PARALLEL DO PRIVATE(e,localCellnodeID,t,myCoords)
added python script that produces vtk files with the (node based and ip based) deformed mesh from marc output file
2013-04-24 00:00:56 +05:30
do n = 1_pInt,Ncellnodes ! loop over cell nodes
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
e = mesh_cellnodeParent(1,n)
localCellnodeID = mesh_cellnodeParent(2,n)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
t = mesh_element(2,e) ! get element type
myCoords = 0.0_pReal
do m = 1_pInt,FE_Nnodes(t)
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
myCoords = myCoords + nodes(1:3,mesh_element(4_pInt+m,e)) &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
* FE_cellnodeParentnodeWeights(m,localCellnodeID,t)
enddo
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_build_cellnodes(1:3,n) = myCoords / sum(FE_cellnodeParentnodeWeights(:,localCellnodeID,t))
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
enddo
fixed bug in cell generation: "FE_Ncellnodes" was not correct for elem type 17 and 136 too much memory allocated for mesh_cell
2013-04-18 23:26:14 +05:30
!$OMP END PARALLEL DO
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
end function mesh_build_cellnodes
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief Calculates IP volume. Allocates global array 'mesh_ipVolume'
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!> @details The IP volume is calculated differently depending on the cell type.
!> 2D cells assume an element depth of one in order to calculate the volume.
!> For the hexahedral cell we subdivide the cell into subvolumes of pyramidal
!> shape with a cell face as basis and the central ip at the tip. This subvolume is
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!> calculated as an average of four tetrahedals with three corners on the cell face
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!> and one corner at the central ip.
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
subroutine mesh_build_ipVolumes
allowed arbitrary grid (formerly know as resolution), i.e. any integer>1 in each dimension is possible. removed square root correction for divergence (doesn't make sense) fixed FFT debug of spectral solver
2013-04-10 15:49:16 +05:30
use math, only: &
math_volTetrahedron, &
math_areaTriangle
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
implicit none
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt) :: e,t,g,c,i,m,f,n
real(pReal), dimension(FE_maxNcellnodesPerCellface,FE_maxNcellfaces) :: subvolume
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 $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
not needed
2019-02-01 15:52:06 +05:30
allocate(mesh_ipVolume(mesh_maxNips,mesh_NcpElems),source=0.0_pReal)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!$OMP PARALLEL DO PRIVATE(t,g,c,m,subvolume)
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
t = mesh_element(2_pInt,e) ! get element type
g = FE_geomtype(t) ! get geometry type
c = FE_celltype(g) ! get cell type
select case (c)
case (1_pInt) ! 2D 3node
forall (i = 1_pInt:FE_Nips(g)) & ! loop over ips=cells in this element
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_ipVolume(i,e) = math_areaTriangle(mesh_cellnode(1:3,mesh_cell(1,i,e)), &
mesh_cellnode(1:3,mesh_cell(2,i,e)), &
mesh_cellnode(1:3,mesh_cell(3,i,e)))
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
case (2_pInt) ! 2D 4node
forall (i = 1_pInt:FE_Nips(g)) & ! loop over ips=cells in this element
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_ipVolume(i,e) = math_areaTriangle(mesh_cellnode(1:3,mesh_cell(1,i,e)), & ! here we assume a planar shape, so division in two triangles suffices
mesh_cellnode(1:3,mesh_cell(2,i,e)), &
mesh_cellnode(1:3,mesh_cell(3,i,e))) &
+ math_areaTriangle(mesh_cellnode(1:3,mesh_cell(3,i,e)), &
mesh_cellnode(1:3,mesh_cell(4,i,e)), &
mesh_cellnode(1:3,mesh_cell(1,i,e)))
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
case (3_pInt) ! 3D 4node
forall (i = 1_pInt:FE_Nips(g)) & ! loop over ips=cells in this element
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_ipVolume(i,e) = math_volTetrahedron(mesh_cellnode(1:3,mesh_cell(1,i,e)), &
mesh_cellnode(1:3,mesh_cell(2,i,e)), &
mesh_cellnode(1:3,mesh_cell(3,i,e)), &
mesh_cellnode(1:3,mesh_cell(4,i,e)))
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
case (4_pInt) ! 3D 8node
m = FE_NcellnodesPerCellface(c)
do i = 1_pInt,FE_Nips(g) ! loop over ips=cells in this element
subvolume = 0.0_pReal
forall(f = 1_pInt:FE_NipNeighbors(c), n = 1_pInt:FE_NcellnodesPerCellface(c)) &
subvolume(n,f) = math_volTetrahedron(&
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_cellnode(1:3,mesh_cell(FE_cellface( n ,f,c),i,e)), &
mesh_cellnode(1:3,mesh_cell(FE_cellface(1+mod(n ,m),f,c),i,e)), &
mesh_cellnode(1:3,mesh_cell(FE_cellface(1+mod(n+1,m),f,c),i,e)), &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
mesh_ipCoordinates(1:3,i,e))
mesh_ipVolume(i,e) = 0.5_pReal * sum(subvolume) ! each subvolume is based on four tetrahedrons, altough the face consists of only two triangles -> averaging factor two
2d elements now have a volume corresponding to a thickness of 1; used to have zero volume now also 2d elements can be used with nonlocal model
2013-04-09 23:37:30 +05:30
enddo
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
end select
enddo
!$OMP END PARALLEL DO
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
end subroutine mesh_build_ipVolumes
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
introduced distinction between ip coordinates and cell center coordinates; the former can be globally calculated by subroutine mesh_build_ipCoordinates, the latter locally by the function mesh_cellCenterCoordinates; renamed mesh_ipCenterOfGravity to mesh_ipCordinates
2012-11-06 20:07:13 +05:30
!> @brief Calculates IP Coordinates. Allocates global array 'mesh_ipCoordinates'
! Called by all solvers in mesh_init in order to initialize the ip coordinates.
! Later on the current ip coordinates are directly prvided by the spectral solver and by Abaqus,
! so no need to use this subroutine anymore; Marc however only provides nodal displacements,
! so in this case the ip coordinates are always calculated on the basis of this subroutine.
! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
! FOR THE MOMENT THIS SUBROUTINE ACTUALLY CALCULATES THE CELL CENTER AND NOT THE IP COORDINATES,
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
! AS THE IP IS NOT (ALWAYS) LOCATED IN THE CENTER OF THE IP VOLUME.
! HAS TO BE CHANGED IN A LATER VERSION.
introduced distinction between ip coordinates and cell center coordinates; the former can be globally calculated by subroutine mesh_build_ipCoordinates, the latter locally by the function mesh_cellCenterCoordinates; renamed mesh_ipCenterOfGravity to mesh_ipCordinates
2012-11-06 20:07:13 +05:30
! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
subroutine mesh_build_ipCoordinates
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
implicit none
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt) :: e,t,g,c,i,n
real(pReal), dimension(3) :: myCoords
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 $mpie periodic z x for having periodicity in x and z direction etc. Note that this only works for regular meshes!!!
2011-02-16 21:53:08 +05:30
compiler complained
2017-04-25 16:04:14 +05:30
if (.not. allocated(mesh_ipCoordinates)) &
allocate(mesh_ipCoordinates(3,mesh_maxNips,mesh_NcpElems),source=0.0_pReal)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!$OMP PARALLEL DO PRIVATE(t,g,c,myCoords)
compiler complained
2017-04-25 16:04:14 +05:30
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
t = mesh_element(2_pInt,e) ! get element type
g = FE_geomtype(t) ! get geometry type
c = FE_celltype(g) ! get cell type
do i = 1_pInt,FE_Nips(g) ! loop over ips=cells in this element
myCoords = 0.0_pReal
do n = 1_pInt,FE_NcellnodesPerCell(c) ! loop over cell nodes in this cell
myCoords = myCoords + mesh_cellnode(1:3,mesh_cell(n,i,e))
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
enddo
compiler complained
2017-04-25 16:04:14 +05:30
mesh_ipCoordinates(1:3,i,e) = myCoords / real(FE_NcellnodesPerCell(c),pReal)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
enddo
compiler complained
2017-04-25 16:04:14 +05:30
enddo
parallelized mesh_build_subNodeCoords and mesh_build_ipCoordinates
2012-11-09 13:17:14 +05:30
!$OMP END PARALLEL DO
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
end subroutine mesh_build_ipCoordinates
added _build_ipNeighborhood and _faceMatch
2007-03-26 14:20:57 +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
introduced distinction between ip coordinates and cell center coordinates; the former can be globally calculated by subroutine mesh_build_ipCoordinates, the latter locally by the function mesh_cellCenterCoordinates; renamed mesh_ipCenterOfGravity to mesh_ipCordinates
2012-11-06 20:07:13 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief Calculates cell center coordinates.
!--------------------------------------------------------------------------------------------------
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
pure function mesh_cellCenterCoordinates(ip,el)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
moved public data res,size and homog from mesh to DAMASK_spectral_utilities (as grid and geomSize)
2013-05-08 21:22:29 +05:30
implicit none
integer(pInt), intent(in) :: el, & !< element number
ip !< integration point number
real(pReal), dimension(3) :: mesh_cellCenterCoordinates !< x,y,z coordinates of the cell center of the requested IP cell
integer(pInt) :: t,g,c,n
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
moved public data res,size and homog from mesh to DAMASK_spectral_utilities (as grid and geomSize)
2013-05-08 21:22:29 +05:30
t = mesh_element(2_pInt,el) ! get element type
g = FE_geomtype(t) ! get geometry type
c = FE_celltype(g) ! get cell type
mesh_cellCenterCoordinates = 0.0_pReal
do n = 1_pInt,FE_NcellnodesPerCell(c) ! loop over cell nodes in this cell
mesh_cellCenterCoordinates = mesh_cellCenterCoordinates + mesh_cellnode(1:3,mesh_cell(n,ip,el))
enddo
compiler complained
2017-04-25 16:04:14 +05:30
mesh_cellCenterCoordinates = mesh_cellCenterCoordinates / real(FE_NcellnodesPerCell(c),pReal)
introduced distinction between ip coordinates and cell center coordinates; the former can be globally calculated by subroutine mesh_build_ipCoordinates, the latter locally by the function mesh_cellCenterCoordinates; renamed mesh_ipCenterOfGravity to mesh_ipCordinates
2012-11-06 20:07:13 +05:30
moved public data res,size and homog from mesh to DAMASK_spectral_utilities (as grid and geomSize)
2013-05-08 21:22:29 +05:30
end function mesh_cellCenterCoordinates
introduced distinction between ip coordinates and cell center coordinates; the former can be globally calculated by subroutine mesh_build_ipCoordinates, the latter locally by the function mesh_cellCenterCoordinates; renamed mesh_ipCenterOfGravity to mesh_ipCordinates
2012-11-06 20:07:13 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!> @brief Figures out version of Marc input file format and stores ist as MarcVersion
!--------------------------------------------------------------------------------------------------
subroutine mesh_marc_get_fileFormat(fileUnit)
use IO, only: &
IO_lc, &
IO_intValue, &
IO_stringValue, &
IO_stringPos
implicit none
integer(pInt), intent(in) :: fileUnit
integer(pInt), allocatable, dimension(:) :: chunkPos
character(len=300) line
610 FORMAT(A300)
rewind(fileUnit)
do
read (fileUnit,610,END=620) line
chunkPos = IO_stringPos(line)
if ( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'version') then
MarcVersion = IO_intValue(line,chunkPos,2_pInt)
exit
endif
enddo
620 end subroutine mesh_marc_get_fileFormat
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief Figures out table styles (Marc only) and stores to 'initialcondTableStyle' and
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!! 'hypoelasticTableStyle'
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_get_tableStyles(fileUnit)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
use IO, only: &
IO_lc, &
IO_intValue, &
IO_stringValue, &
IO_stringPos
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
added function mesh_FEtoCPelement(FEid)
2007-03-21 21:48:33 +05:30
implicit none
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: fileUnit
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
initialcondTableStyle = 0_pInt
hypoelasticTableStyle = 0_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(fileUnit)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(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
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
if ( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'table' .and. chunkPos(1_pInt) > 5) then
initialcondTableStyle = IO_intValue(line,chunkPos,4_pInt)
hypoelasticTableStyle = IO_intValue(line,chunkPos,5_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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
620 end subroutine mesh_marc_get_tableStyles
Simplified 'mesh_get_nodeElemDimensions', mesh_build_elements' and 'mesh_build_sharedElems'. Corrected entries in 'FE_subNodeOnIPFace'
2009-04-03 12:33:25 +05:30
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!--------------------------------------------------------------------------------------------------
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
!> @brief Figures out material number of hypoelastic material and stores it in Marc_matNumber array
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!--------------------------------------------------------------------------------------------------
subroutine mesh_marc_get_matNumber(fileUnit)
use IO, only: &
IO_lc, &
IO_intValue, &
IO_stringValue, &
IO_stringPos
implicit none
integer(pInt), intent(in) :: fileUnit
integer(pInt), allocatable, dimension(:) :: chunkPos
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
integer(pInt) :: i, j, data_blocks
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
character(len=300) line
610 FORMAT(A300)
rewind(fileUnit)
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
data_blocks = 1_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
do
read (fileUnit,610,END=620) line
chunkPos = IO_stringPos(line)
if ( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'hypoelastic') then
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
read (fileUnit,610,END=620) line
if (len(trim(line))/=0_pInt) then
chunkPos = IO_stringPos(line)
data_blocks = IO_intValue(line,chunkPos,1_pInt)
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
endif
allocate(Marc_matNumber(data_blocks))
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
do i=1_pInt,data_blocks ! read all data blocks
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
read (fileUnit,610,END=620) line
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
chunkPos = IO_stringPos(line)
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
Marc_matNumber(i) = IO_intValue(line,chunkPos,1_pInt)
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
do j=1_pint,2_pInt + hypoelasticTableStyle ! read 2 or 3 remaining lines of data block
read (fileUnit,610,END=620) line
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
enddo
exit
endif
enddo
620 end subroutine mesh_marc_get_matNumber
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!> @brief Count overall number of nodes and elements in mesh and stores the numbers in
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!! 'mesh_Nelems' and 'mesh_Nnodes'
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_count_nodesAndElements(fileUnit)
some minor polishing on the way
2014-04-04 20:10:30 +05:30
use IO, only: &
IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_IntValue
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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
implicit none
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: fileUnit
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
polishing: removed variable names like 'unit' and 'data' that are keywords of fortran and ensured that integer and real precision matches independent of machine standard. removed cut_off parameter for damask_spectral removed outpot of 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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
610 FORMAT(A300)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(fileUnit)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
if ( IO_lc(IO_StringValue(line,chunkPos,1_pInt)) == 'sizing') &
mesh_Nelems = IO_IntValue (line,chunkPos,3_pInt)
if ( IO_lc(IO_StringValue(line,chunkPos,1_pInt)) == 'coordinates') then
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
mesh_Nnodes = IO_IntValue (line,chunkPos,2_pInt)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
exit ! assumes that "coordinates" comes later in file
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
620 end subroutine mesh_marc_count_nodesAndElements
!--------------------------------------------------------------------------------------------------
!> @brief Count overall number of element sets in mesh. Stores to 'mesh_NelemSets', and
!! 'mesh_maxNelemInSet'
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_count_elementSets(fileUnit)
some minor polishing on the way
2014-04-04 20:10:30 +05:30
use IO, only: &
IO_lc, &
IO_stringValue, &
IO_stringPos, &
IO_countContinuousIntValues
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: fileUnit
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(fileUnit)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(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
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
if ( IO_lc(IO_StringValue(line,chunkPos,1_pInt)) == 'define' .and. &
IO_lc(IO_StringValue(line,chunkPos,2_pInt)) == 'element' ) then
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
mesh_NelemSets = mesh_NelemSets + 1_pInt
mesh_maxNelemInSet = max(mesh_maxNelemInSet, &
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
IO_countContinuousIntValues(fileUnit))
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
!********************************************************************
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
! map element sets
added function calls in init()
2007-04-10 16:52:53 +05:30
!
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
! allocate globals: mesh_nameElemSet, mesh_mapElemSet
added function calls in init()
2007-04-10 16:52:53 +05:30
!********************************************************************
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_map_elementSets(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
use IO, only: IO_lc, &
IO_stringValue, &
IO_stringPos, &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
IO_continuousIntValues
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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: fileUnit
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
integer(pInt) :: elemSet = 0_pInt
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
leaner syntax with sourced allocation
2018-09-23 20:56:13 +05:30
allocate (mesh_nameElemSet(mesh_NelemSets)); mesh_nameElemSet = ''
allocate (mesh_mapElemSet(1_pInt+mesh_maxNelemInSet,mesh_NelemSets), source=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)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(fileUnit)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=640) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
if( (IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'define' ) .and. &
(IO_lc(IO_stringValue(line,chunkPos,2_pInt)) == 'element' ) ) then
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
elemSet = elemSet+1_pInt
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
mesh_nameElemSet(elemSet) = trim(IO_stringValue(line,chunkPos,4_pInt))
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
mesh_mapElemSet(:,elemSet) = &
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
IO_continuousIntValues(fileUnit,mesh_maxNelemInSet,mesh_nameElemSet,mesh_mapElemSet,mesh_NelemSets)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
640 end subroutine mesh_marc_map_elementSets
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief Count overall number of CP elements in mesh and stores them in 'mesh_NcpElems'
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_count_cpElements(fileUnit)
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, &
hot fix for so far non-supported input file format of Marc 2017
2017-10-04 20:48:06 +05:30
IO_countContinuousIntValues, &
IO_error, &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
IO_intValue, &
IO_countNumericalDataLines
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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: fileUnit
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
integer(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
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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(fileUnit)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
if (MarcVersion < 13) then ! Marc 2016 or earlier
do
read (fileUnit,610,END=620) line
chunkPos = IO_stringPos(line)
if ( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'hypoelastic') then
do i=1_pInt,3_pInt+hypoelasticTableStyle ! Skip 3 or 4 lines
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,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
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
mesh_NcpElems = mesh_NcpElems + IO_countContinuousIntValues(fileUnit) ! why not simply mesh_NcpElems = IO_countContinuousIntValues(fileUnit)? not fully correct as hypoelastic can have multiple data fields, needs update
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
exit
endif
enddo
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
else ! Marc2017 and later
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
do
read (fileUnit,610,END=620) line
chunkPos = IO_stringPos(line)
if ( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'connectivity') then
read (fileUnit,610,END=620) line
chunkPos = IO_stringPos(line)
working further on Marc2017 compatibility
2018-01-11 21:41:03 +05:30
if (any(Marc_matNumber==IO_intValue(line,chunkPos,6_pInt))) then
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
mesh_NcpElems = mesh_NcpElems + IO_countNumericalDataLines(fileUnit)
endif
endif
enddo
end if
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
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief Maps elements from FE ID to internal (consecutive) representation.
!! Allocates global array 'mesh_mapFEtoCPelem'
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_map_elements(fileUnit)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
changed misplaced private to public statement and renamed qsort to math_qsort
2013-02-01 12:56:21 +05:30
use math, only: math_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, &
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
IO_intValue, &
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_stringValue, &
IO_stringPos, &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
IO_continuousIntValues
added function calls in init()
2007-04-10 16:52:53 +05:30
implicit none
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: fileUnit
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
character(len=300) :: line, &
tmp
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
integer(pInt), dimension (1_pInt+mesh_NcpElems) :: contInts
integer(pInt) :: i,cpElem = 0_pInt
leaner syntax with sourced allocation
2018-09-23 20:56:13 +05:30
allocate (mesh_mapFEtoCPelem(2,mesh_NcpElems), source = 0_pInt)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
610 FORMAT(A300)
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
contInts = 0_pInt
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(fileUnit)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=660) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
if (MarcVersion < 13) then ! Marc 2016 or earlier
if( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'hypoelastic' ) then
do i=1_pInt,3_pInt+hypoelasticTableStyle ! skip three (or four if new table style!) lines
read (fileUnit,610,END=660) line
enddo
contInts = IO_continuousIntValues(fileUnit,mesh_NcpElems,mesh_nameElemSet,&
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
mesh_mapElemSet,mesh_NelemSets)
new Marc2017 file format finally working!
2018-02-02 19:36:13 +05:30
exit
endif
else ! Marc2017 and later
if ( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'connectivity') then
read (fileUnit,610,END=660) line
chunkPos = IO_stringPos(line)
if(any(Marc_matNumber==IO_intValue(line,chunkPos,6_pInt))) then
do
read (fileUnit,610,END=660) line
chunkPos = IO_stringPos(line)
tmp = IO_lc(IO_stringValue(line,chunkPos,1_pInt))
if (verify(trim(tmp),"0123456789")/=0) then ! found keyword
exit
else
contInts(1) = contInts(1) + 1_pInt
read (tmp,*) contInts(contInts(1)+1)
endif
enddo
endif
endif
endif
enddo
660 do i = 1_pInt,contInts(1)
cpElem = cpElem+1_pInt
mesh_mapFEtoCPelem(1,cpElem) = contInts(1_pInt+i)
mesh_mapFEtoCPelem(2,cpElem) = cpElem
enddo
call math_qsort(mesh_mapFEtoCPelem,1_pInt,int(size(mesh_mapFEtoCPelem,2_pInt),pInt)) ! should be mesh_NcpElems
major debugging and streamlining of the calculation of ipVolumes and ipAreas
2009-01-20 00:12:31 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief Maps node from FE ID to internal (consecutive) representation.
!! Allocates global array 'mesh_mapFEtoCPnode'
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_map_nodes(fileUnit)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
changed misplaced private to public statement and renamed qsort to math_qsort
2013-02-01 12:56:21 +05:30
use math, only: math_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, &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
IO_fixedIntValue
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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
implicit none
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: 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
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
character(len=300) line
integer(pInt), dimension (mesh_Nnodes) :: node_count
integer(pInt) :: i
typo, Abaqus/Marc did not compile
2018-09-27 23:48:37 +05:30
allocate (mesh_mapFEtoCPnode(2_pInt,mesh_Nnodes),source=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)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
node_count = 0_pInt
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(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
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=650) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
if( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'coordinates' ) then
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
read (fileUnit,610,END=650) line ! skip crap line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
do i = 1_pInt,mesh_Nnodes
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=650) line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
mesh_mapFEtoCPnode(1_pInt,i) = IO_fixedIntValue (line,[ 0_pInt,10_pInt],1_pInt)
mesh_mapFEtoCPnode(2_pInt,i) = i
enddo
exit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
changed misplaced private to public statement and renamed qsort to math_qsort
2013-02-01 12:56:21 +05:30
650 call math_qsort(mesh_mapFEtoCPnode,1_pInt,int(size(mesh_mapFEtoCPnode,2_pInt),pInt))
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief store x,y,z coordinates of all nodes in mesh.
!! Allocates global arrays 'mesh_node0' and 'mesh_node'
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_build_nodes(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
moved public data res,size and homog from mesh to DAMASK_spectral_utilities (as grid and geomSize)
2013-05-08 21:22:29 +05:30
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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: 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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
integer(pInt), dimension(5), parameter :: node_ends = int([0,10,30,50,70],pInt)
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
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
leaner syntax with sourced allocation
2018-09-23 20:56:13 +05:30
allocate ( mesh_node0 (3,mesh_Nnodes), source=0.0_pReal)
allocate ( mesh_node (3,mesh_Nnodes), source=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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(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
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=670) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
if( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'coordinates' ) then
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
read (fileUnit,610,END=670) line ! skip crap line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
do i=1_pInt,mesh_Nnodes
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=670) line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
m = mesh_FEasCP('node',IO_fixedIntValue(line,node_ends,1_pInt))
as some IO functions are no longer pure I changed some forall statements to ordinary do loops
2013-02-07 14:22:47 +05:30
do j = 1_pInt,3_pInt
"unitlength" parameter, which determines the physical size of the mesh, now available as a global mesh variable "mesh_unitlength" and written to the .mesh file during init. Hence, it is available to the marc_deformedGeometry.py script via "mesh_init_postprocessing()" and "mesh_get_unitlength()", so no need for setting the scaling of the displacement vectors explicitly through an option; now displacements and nodal positions are always consistent
2013-05-07 18:36:29 +05:30
mesh_node0(j,m) = mesh_unitlength * IO_fixedNoEFloatValue(line,node_ends,j+1_pInt)
as some IO functions are no longer pure I changed some forall statements to ordinary do loops
2013-02-07 14:22:47 +05:30
enddo
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
enddo
exit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!> @brief Gets maximum count of nodes, IPs, IP neighbors, and cellnodes among cpElements.
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!! Sets global values 'mesh_maxNnodes', 'mesh_maxNips', 'mesh_maxNipNeighbors',
In mesh_marc_build_elements use mesh_NcpElems instead of mesh_Nelems (line 2906). Fixes error when not all elements are CP elements. Some formating and comment improvements.
2014-12-15 17:21:32 +05:30
!! and 'mesh_maxNcellnodes'
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_count_cpSizes(fileUnit)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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, &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: fileUnit
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
character(len=300) :: line
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt) :: i,t,g,e,c
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
mesh_maxNnodes = 0_pInt
mesh_maxNips = 0_pInt
mesh_maxNipNeighbors = 0_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
mesh_maxNcellnodes = 0_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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
610 FORMAT(A300)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(fileUnit)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=630) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
if( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'connectivity' ) then
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
read (fileUnit,610,END=630) line ! Garbage line
do i=1_pInt,mesh_Nelems ! read all elements
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=630) line
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
chunkPos = IO_stringPos(line) ! limit to id and type
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
e = mesh_FEasCP('elem',IO_intValue(line,chunkPos,1_pInt))
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
if (e /= 0_pInt) then
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
t = FE_mapElemtype(IO_stringValue(line,chunkPos,2_pInt))
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
g = FE_geomtype(t)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
c = FE_celltype(g)
mesh_maxNnodes = max(mesh_maxNnodes,FE_Nnodes(t))
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
mesh_maxNips = max(mesh_maxNips,FE_Nips(g))
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
mesh_maxNipNeighbors = max(mesh_maxNipNeighbors,FE_NipNeighbors(c))
mesh_maxNcellnodes = max(mesh_maxNcellnodes,FE_Ncellnodes(g))
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
call IO_skipChunks(fileUnit,FE_Nnodes(t)-(chunkPos(1_pInt)-2_pInt)) ! read on if FE_Nnodes exceeds node count present on current line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
exit
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
630 end subroutine mesh_marc_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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
In mesh_marc_build_elements use mesh_NcpElems instead of mesh_Nelems (line 2906). Fixes error when not all elements are CP elements. Some formating and comment improvements.
2014-12-15 17:21:32 +05:30
!> @brief Store FEid, type, mat, tex, and node list per element.
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!! Allocates global array 'mesh_element'
!--------------------------------------------------------------------------------------------------
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
subroutine mesh_marc_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
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, &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
IO_fixedNoEFloatValue, &
IO_skipChunks, &
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_stringPos, &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
IO_intValue, &
drop support for heterogeneous meshes heterogeneous meshes are neither advisable nor typically used
2018-09-23 18:49:47 +05:30
IO_continuousIntValues, &
IO_error
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
implicit none
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: 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
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
integer(pInt), dimension(1_pInt+mesh_NcpElems) :: contInts
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt) :: i,j,t,sv,myVal,e,nNodesAlreadyRead
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
better name
2018-09-23 23:28:43 +05:30
allocate(mesh_element(4_pInt+mesh_maxNnodes,mesh_NcpElems), source=0_pInt)
drop support for heterogeneous meshes heterogeneous meshes are neither advisable nor typically used
2018-09-23 18:49:47 +05:30
mesh_elemType = -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
610 FORMAT(A300)
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(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
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
realloc lhs was not always working, fixed now and added test for new string pos function
2015-08-31 22:00:04 +05:30
chunkPos = IO_stringPos(line)
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
if( IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'connectivity' ) then
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
read (fileUnit,610,END=620) line ! garbage line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
do i = 1_pInt,mesh_Nelems
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
e = mesh_FEasCP('elem',IO_intValue(line,chunkPos,1_pInt))
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
if (e /= 0_pInt) then ! disregard non CP elems
mesh_element(1,:) only used for debug output set to -1 at the moment to indicate that it is not used. Re-implementation should be done for MSC.Marc and Abaqus only.
2018-09-23 22:10:14 +05:30
mesh_element(1,e) = -1_pInt ! DEPRECATED
drop support for heterogeneous meshes heterogeneous meshes are neither advisable nor typically used
2018-09-23 18:49:47 +05:30
t = FE_mapElemtype(IO_StringValue(line,chunkPos,2_pInt)) ! elem type
if (mesh_elemType /= t .and. mesh_elemType /= -1_pInt) &
call IO_error(191,el=t,ip=mesh_elemType)
mesh_elemType = t
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
mesh_element(2,e) = t
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
nNodesAlreadyRead = 0_pInt
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
do j = 1_pInt,chunkPos(1)-2_pInt
mesh_element(4_pInt+j,e) = mesh_FEasCP('node',IO_IntValue(line,chunkPos,j+2_pInt)) ! CP ids of nodes
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
enddo
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
nNodesAlreadyRead = chunkPos(1) - 2_pInt
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
do while(nNodesAlreadyRead < FE_Nnodes(t)) ! read on if not all nodes in one line
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
do j = 1_pInt,chunkPos(1)
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
mesh_element(4_pInt+nNodesAlreadyRead+j,e) &
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
= mesh_FEasCP('node',IO_IntValue(line,chunkPos,j)) ! CP ids of nodes
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
enddo
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
nNodesAlreadyRead = nNodesAlreadyRead + chunkPos(1)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
enddo
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
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
exit
endif
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
620 rewind(fileUnit) ! just in case "initial state" appears before "connectivity"
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
do
realloc lhs was not always working, fixed now and added test for new string pos function
2015-08-31 22:00:04 +05:30
chunkPos = IO_stringPos(line)
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
if( (IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == 'initial') .and. &
(IO_lc(IO_stringValue(line,chunkPos,2_pInt)) == 'state') ) then
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
if (initialcondTableStyle == 2_pInt) read (fileUnit,610,END=620) line ! read extra line for new style
read (fileUnit,610,END=630) line ! read line with index of state var
realloc lhs was not always working, fixed now and added test for new string pos function
2015-08-31 22:00:04 +05:30
chunkPos = IO_stringPos(line)
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
sv = IO_IntValue(line,chunkPos,1_pInt) ! figure state variable index
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
if( (sv == 2_pInt).or.(sv == 3_pInt) ) then ! only state vars 2 and 3 of interest
read (fileUnit,610,END=620) line ! read line with value of state var
realloc lhs was not always working, fixed now and added test for new string pos function
2015-08-31 22:00:04 +05:30
chunkPos = IO_stringPos(line)
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
do while (scan(IO_stringValue(line,chunkPos,1_pInt),'+-',back=.true.)>1) ! is noEfloat value?
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
myVal = nint(IO_fixedNoEFloatValue(line,[0_pInt,20_pInt],1_pInt),pInt) ! state var's value
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
if (initialcondTableStyle == 2_pInt) then
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
read (fileUnit,610,END=630) line ! read extra line
read (fileUnit,610,END=630) line ! read extra line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
endif
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
contInts = IO_continuousIntValues& ! get affected elements
In mesh_marc_build_elements use mesh_NcpElems instead of mesh_Nelems (line 2906). Fixes error when not all elements are CP elements. Some formating and comment improvements.
2014-12-15 17:21:32 +05:30
(fileUnit,mesh_NcpElems,mesh_nameElemSet,mesh_mapElemSet,mesh_NelemSets)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
do i = 1_pInt,contInts(1)
e = mesh_FEasCP('elem',contInts(1_pInt+i))
mesh_element(1_pInt+sv,e) = myVal
enddo
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
if (initialcondTableStyle == 0_pInt) read (fileUnit,610,END=620) line ! ignore IP range for old table style
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=630) line
realloc lhs was not always working, fixed now and added test for new string pos function
2015-08-31 22:00:04 +05:30
chunkPos = IO_stringPos(line)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
enddo
endif
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
else
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=630) line
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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
remove non-marc specific code
2019-01-24 14:34:43 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
remove non-marc specific code
2019-01-24 14:34:43 +05:30
!> @brief get any additional damask options from input file, sets mesh_periodicSurface
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
!--------------------------------------------------------------------------------------------------
remove non-marc specific code
2019-01-24 14:34:43 +05:30
subroutine mesh_get_damaskOptions(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
remove non-marc specific code
2019-01-24 14:34:43 +05:30
use IO, only: &
IO_lc, &
IO_stringValue, &
IO_stringPos
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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
implicit none
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
integer(pInt), intent(in) :: fileUnit
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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
remove non-marc specific code
2019-01-24 14:34:43 +05:30
integer(pInt), allocatable, dimension(:) :: chunkPos
integer(pInt) chunk, Nchunks
character(len=300) :: line, damaskOption, v
character(len=300) :: keyword
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
remove non-marc specific code
2019-01-24 14:34:43 +05:30
mesh_periodicSurface = .false.
keyword = '$damask'
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
rewind(fileUnit)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
do
added 'reset' flag for recursive function IO_read, need to reset stack when stopping exectution (will be used by constitutive and homogenization)
2013-12-11 22:19:20 +05:30
read (fileUnit,610,END=620) line
behavior of IO_stringPos was unsafe in some situations, also took the chance to rename the variables used in its context to a more meaningful name. functions in the core module are not used, including kdTree
2015-08-28 13:08:48 +05:30
chunkPos = IO_stringPos(line)
remove non-marc specific code
2019-01-24 14:34:43 +05:30
Nchunks = chunkPos(1)
all variables/functions were not used
2019-01-31 16:53:23 +05:30
if (IO_lc(IO_stringValue(line,chunkPos,1_pInt)) == keyword .and. Nchunks > 1_pInt) then ! found keyword for damask option and there is at least one more chunk to read
remove non-marc specific code
2019-01-24 14:34:43 +05:30
damaskOption = IO_lc(IO_stringValue(line,chunkPos,2_pInt))
select case(damaskOption)
case('periodic') ! damask Option that allows to specify periodic fluxes
do chunk = 3_pInt,Nchunks ! loop through chunks (skipping the keyword)
all variables/functions were not used
2019-01-31 16:53:23 +05:30
v = IO_lc(IO_stringValue(line,chunkPos,chunk)) ! chunk matches keyvalues x,y, or z?
remove non-marc specific code
2019-01-24 14:34:43 +05:30
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'
enddo
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
endselect
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var 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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
remove non-marc specific code
2019-01-24 14:34:43 +05:30
610 FORMAT(A300)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
remove non-marc specific code
2019-01-24 14:34:43 +05:30
620 end subroutine mesh_get_damaskOptions
# new interface for Abaqus # IO has some additional functionality for Abaqus parsing # ping pong scheme in FE interface now similar (and more human understandable) in both versions # mesh has better splitting of different tasks, plus operation on database whenever possible # FEsolver as new global var to indicate FEM solver type # computation mode reshuffling: 6 is now Marc special case of recycling...
2009-10-12 21:31:49 +05:30
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!--------------------------------------------------------------------------------------------------
remove non-marc specific code
2019-01-24 14:34:43 +05:30
!> @brief calculation of IP interface areas, allocate globals '_ipArea', and '_ipAreaNormal'
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!--------------------------------------------------------------------------------------------------
remove non-marc specific code
2019-01-24 14:34:43 +05:30
subroutine mesh_build_ipAreas
use math, only: &
math_crossproduct
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
remove non-marc specific code
2019-01-24 14:34:43 +05:30
integer(pInt) :: e,t,g,c,i,f,n,m
real(pReal), dimension (3,FE_maxNcellnodesPerCellface) :: nodePos, normals
real(pReal), dimension(3) :: normal
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
remove non-marc specific code
2019-01-24 14:34:43 +05:30
allocate(mesh_ipArea(mesh_maxNipNeighbors,mesh_maxNips,mesh_NcpElems), source=0.0_pReal)
allocate(mesh_ipAreaNormal(3_pInt,mesh_maxNipNeighbors,mesh_maxNips,mesh_NcpElems), source=0.0_pReal)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!$OMP PARALLEL DO PRIVATE(t,g,c,nodePos,normal,normals)
do e = 1_pInt,mesh_NcpElems ! loop over cpElems
t = mesh_element(2_pInt,e) ! get element type
g = FE_geomtype(t) ! get geometry type
c = FE_celltype(g) ! get cell type
select case (c)
case (1_pInt,2_pInt) ! 2D 3 or 4 node
do i = 1_pInt,FE_Nips(g) ! loop over ips=cells in this element
do f = 1_pInt,FE_NipNeighbors(c) ! loop over cell faces
forall(n = 1_pInt:FE_NcellnodesPerCellface(c)) &
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
nodePos(1:3,n) = mesh_cellnode(1:3,mesh_cell(FE_cellface(n,f,c),i,e))
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
normal(1) = nodePos(2,2) - nodePos(2,1) ! x_normal = y_connectingVector
normal(2) = -(nodePos(1,2) - nodePos(1,1)) ! y_normal = -x_connectingVector
normal(3) = 0.0_pReal
removed norm functions from math in favor of intrinsic and simplified (mostly by using existing functions, merge intrinsic and array constructors)
2016-01-10 19:04:26 +05:30
mesh_ipArea(f,i,e) = norm2(normal)
mesh_ipAreaNormal(1:3,f,i,e) = normal / norm2(normal) ! ensure unit length of area normal
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
enddo
2d elements now have a volume corresponding to a thickness of 1; used to have zero volume now also 2d elements can be used with nonlocal model
2013-04-09 23:37:30 +05:30
enddo
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
case (3_pInt) ! 3D 4node
do i = 1_pInt,FE_Nips(g) ! loop over ips=cells in this element
do f = 1_pInt,FE_NipNeighbors(c) ! loop over cell faces
forall(n = 1_pInt:FE_NcellnodesPerCellface(c)) &
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
nodePos(1:3,n) = mesh_cellnode(1:3,mesh_cell(FE_cellface(n,f,c),i,e))
removed norm functions from math in favor of intrinsic and simplified (mostly by using existing functions, merge intrinsic and array constructors)
2016-01-10 19:04:26 +05:30
normal = math_crossproduct(nodePos(1:3,2) - nodePos(1:3,1), &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
nodePos(1:3,3) - nodePos(1:3,1))
removed norm functions from math in favor of intrinsic and simplified (mostly by using existing functions, merge intrinsic and array constructors)
2016-01-10 19:04:26 +05:30
mesh_ipArea(f,i,e) = norm2(normal)
mesh_ipAreaNormal(1:3,f,i,e) = normal / norm2(normal) ! ensure unit length of area normal
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
enddo
2d elements now have a volume corresponding to a thickness of 1; used to have zero volume now also 2d elements can be used with nonlocal model
2013-04-09 23:37:30 +05:30
enddo
introduced dedicated calculation of interface normals in IP neighborhood. (2D elements did not work so far..!) fixed definitions of some (2D) interface planes to have all normals pointing outward. removed "int, private :: i" defined for spectral compilation.
2013-02-05 18:57:37 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
case (4_pInt) ! 3D 8node
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
! for this cell type we get the normal of the quadrilateral face as an average of
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
! four normals of triangular subfaces; since the face consists only of two triangles,
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
! the sum has to be divided by two; this whole prcedure tries to compensate for
! probable non-planar cell surfaces
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
m = FE_NcellnodesPerCellface(c)
do i = 1_pInt,FE_Nips(g) ! loop over ips=cells in this element
do f = 1_pInt,FE_NipNeighbors(c) ! loop over cell faces
forall(n = 1_pInt:FE_NcellnodesPerCellface(c)) &
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
nodePos(1:3,n) = mesh_cellnode(1:3,mesh_cell(FE_cellface(n,f,c),i,e))
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
forall(n = 1_pInt:FE_NcellnodesPerCellface(c)) &
normals(1:3,n) = 0.5_pReal &
removed norm functions from math in favor of intrinsic and simplified (mostly by using existing functions, merge intrinsic and array constructors)
2016-01-10 19:04:26 +05:30
* math_crossproduct(nodePos(1:3,1+mod(n ,m)) - nodePos(1:3,n), &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
nodePos(1:3,1+mod(n+1,m)) - nodePos(1:3,n))
normal = 0.5_pReal * sum(normals,2)
removed norm functions from math in favor of intrinsic and simplified (mostly by using existing functions, merge intrinsic and array constructors)
2016-01-10 19:04:26 +05:30
mesh_ipArea(f,i,e) = norm2(normal)
mesh_ipAreaNormal(1:3,f,i,e) = normal / norm2(normal)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
enddo
enddo
end select
enddo
!$OMP END PARALLEL DO
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
took some gfortran complaints serious (unused imports, implicit castings)
2015-04-11 17:17:33 +05:30
end subroutine mesh_build_ipAreas
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
remove non-marc specific code
2019-01-24 14:34:43 +05:30
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief assignment of twin nodes for each cp node, allocate globals '_nodeTwins'
!--------------------------------------------------------------------------------------------------
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
subroutine mesh_build_nodeTwins
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
implicit none
integer(pInt) dir, & ! direction of periodicity
node, &
minimumNode, &
maximumNode, &
n1, &
n2
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
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 small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
logical, dimension(mesh_Nnodes) :: unpaired
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
allocate(mesh_nodeTwins(3,mesh_Nnodes))
mesh_nodeTwins = 0_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
tolerance = 0.001_pReal * minval(mesh_ipVolume) ** 0.333_pReal
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
do dir = 1_pInt,3_pInt ! check periodicity in directions of x,y,z
if (mesh_periodicSurface(dir)) then ! only if periodicity is requested
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!*** find out which nodes sit on the surface
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** and have a minimum or maximum position in this dimension
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
minimumNodes = 0_pInt
maximumNodes = 0_pInt
minCoord = minval(mesh_node0(dir,:))
maxCoord = maxval(mesh_node0(dir,:))
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
do node = 1_pInt,mesh_Nnodes ! loop through all nodes and find surface nodes
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
if (abs(mesh_node0(dir,node) - minCoord) <= tolerance) then
minimumNodes(1) = minimumNodes(1) + 1_pInt
minimumNodes(minimumNodes(1)+1_pInt) = node
elseif (abs(mesh_node0(dir,node) - maxCoord) <= tolerance) then
maximumNodes(1) = maximumNodes(1) + 1_pInt
maximumNodes(maximumNodes(1)+1_pInt) = node
endif
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** find the corresponding node on the other side with the same position in this dimension
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
unpaired = .true.
do n1 = 1_pInt,minimumNodes(1)
minimumNode = minimumNodes(n1+1_pInt)
if (unpaired(minimumNode)) then
do n2 = 1_pInt,maximumNodes(1)
maximumNode = maximumNodes(n2+1_pInt)
distance = abs(mesh_node0(:,minimumNode) - mesh_node0(:,maximumNode))
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
if (sum(distance) - distance(dir) <= tolerance) then ! minimum possible distance (within tolerance)
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
mesh_nodeTwins(dir,minimumNode) = maximumNode
mesh_nodeTwins(dir,maximumNode) = minimumNode
Spectral solver now fully parallel (parallel IO, domain decomposition, FFTs and restart). Working but not extensively tested so please report bugs to me
2015-03-25 21:36:19 +05:30
unpaired(maximumNode) = .false. ! remember this node, we don't have to look for his partner again
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
exit
endif
enddo
endif
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
endif
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
end subroutine mesh_build_nodeTwins
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!--------------------------------------------------------------------------------------------------
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
!> @brief get maximum count of shared elements among cpElements and build list of elements shared
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!! by each node in mesh. Allocate globals '_maxNsharedElems' and '_sharedElem'
!--------------------------------------------------------------------------------------------------
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
subroutine mesh_build_sharedElems
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
implicit none
integer(pint) e, & ! element index
g, & ! element type
node, & ! CP node index
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
n, & ! node index per element
myDim, & ! dimension index
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
nodeTwin ! node twin in the specified dimension
integer(pInt), dimension (mesh_Nnodes) :: node_count
leaner syntax with sourced allocation
2018-09-23 20:56:13 +05:30
integer(pInt), dimension(:), allocatable :: node_seen
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
allocate(node_seen(maxval(FE_NmatchingNodes)))
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
node_count = 0_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
do e = 1_pInt,mesh_NcpElems
g = FE_geomtype(mesh_element(2,e)) ! get elemGeomType
node_seen = 0_pInt ! reset node duplicates
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
do n = 1_pInt,FE_NmatchingNodes(g) ! check each node of element
node = mesh_element(4+n,e)
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
if (all(node_seen /= node)) then
node_count(node) = node_count(node) + 1_pInt ! if FE node not yet encountered -> count it
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
do myDim = 1_pInt,3_pInt ! check in each dimension...
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
nodeTwin = mesh_nodeTwins(myDim,node)
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
enddo
endif
node_seen(n) = node ! remember this node to be counted already
enddo
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
mesh_maxNsharedElems = int(maxval(node_count),pInt) ! most shared node
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
leaner syntax with sourced allocation
2018-09-23 20:56:13 +05:30
allocate(mesh_sharedElem(1+mesh_maxNsharedElems,mesh_Nnodes),source=0_pInt)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
do e = 1_pInt,mesh_NcpElems
g = FE_geomtype(mesh_element(2,e)) ! get elemGeomType
node_seen = 0_pInt
do n = 1_pInt,FE_NmatchingNodes(g)
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
node = mesh_element(4_pInt+n,e)
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
if (all(node_seen /= node)) then
mesh_sharedElem(1,node) = mesh_sharedElem(1,node) + 1_pInt ! count for each node the connected elements
mesh_sharedElem(mesh_sharedElem(1,node)+1_pInt,node) = e ! store the respective element id
do myDim = 1_pInt,3_pInt ! check in each dimension...
nodeTwin = mesh_nodeTwins(myDim,node)
if (nodeTwin > 0_pInt) then ! if i am a twin of some node...
mesh_sharedElem(1,nodeTwin) = mesh_sharedElem(1,nodeTwin) + 1_pInt ! ...count me again for the twin
mesh_sharedElem(mesh_sharedElem(1,nodeTwin)+1,nodeTwin) = e ! store the respective element id
endif
enddo
endif
node_seen(n) = node
enddo
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!--------------------------------------------------------------------------------------------------
introduced preprocessor statements in mesh to hide unneeded variables as a preparation for HDF5
2013-12-28 01:33:28 +05:30
!> @brief build up of IP neighborhood, allocate globals '_ipNeighborhood'
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +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
subroutine mesh_build_ipNeighborhood
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
use math, only: &
math_mul3x3
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
implicit none
integer(pInt) :: myElem, & ! my CP element index
myIP, &
myType, & ! my element type
myFace, &
neighbor, & ! neighor index
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
neighboringIPkey, & ! positive integer indicating the neighboring IP (for intra-element) and negative integer indicating the face towards neighbor (for neighboring element)
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
candidateIP, &
neighboringType, & ! element type of neighbor
NlinkedNodes, & ! number of linked nodes
twin_of_linkedNode, & ! node twin of a specific linkedNode
NmatchingNodes, & ! number of matching nodes
dir, & ! direction of periodicity
matchingElem, & ! CP elem number of matching element
matchingFace, & ! face ID of matching element
a, anchor, &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
neighboringIP, &
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
neighboringElem, &
pointingToMe
integer(pInt), dimension(FE_maxmaxNnodesAtIP) :: &
linkedNodes = 0_pInt, &
matchingNodes
logical checkTwins
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
allocate(mesh_ipNeighborhood(3,mesh_maxNipNeighbors,mesh_maxNips,mesh_NcpElems))
mesh_ipNeighborhood = 0_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
do myElem = 1_pInt,mesh_NcpElems ! loop over cpElems
myType = FE_geomtype(mesh_element(2,myElem)) ! get elemGeomType
do myIP = 1_pInt,FE_Nips(myType) ! loop over IPs of elem
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
do neighbor = 1_pInt,FE_NipNeighbors(FE_celltype(myType)) ! loop over neighbors of IP
neighboringIPkey = FE_ipNeighbor(neighbor,myIP,myType)
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** if the key is positive, the neighbor is inside the element
!*** that means, we have already found our neighboring IP
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
if (neighboringIPkey > 0_pInt) then
mesh_ipNeighborhood(1,neighbor,myIP,myElem) = myElem
mesh_ipNeighborhood(2,neighbor,myIP,myElem) = neighboringIPkey
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** 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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
elseif (neighboringIPkey < 0_pInt) then ! neighboring element's IP
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 = FE_geomtype(mesh_element(2,matchingElem))
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** trivial solution if neighbor has only one IP
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
if (FE_Nips(neighboringType) == 1_pInt) then
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
mesh_ipNeighborhood(1,neighbor,myIP,myElem) = matchingElem
mesh_ipNeighborhood(2,neighbor,myIP,myElem) = 1_pInt
cycle
endif
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** find those nodes which build the link to the neighbor
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
NlinkedNodes = 0_pInt
linkedNodes = 0_pInt
do a = 1_pInt,FE_maxNnodesAtIP(myType) ! figure my anchor nodes on connecting face
anchor = FE_nodesAtIP(a,myIP,myType)
if (anchor /= 0_pInt) then ! valid anchor node
if (any(FE_face(:,myFace,myType) == anchor)) then ! ip anchor sits on face?
NlinkedNodes = NlinkedNodes + 1_pInt
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
linkedNodes(NlinkedNodes) = mesh_element(4_pInt+anchor,myElem) ! CP id of anchor node
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** loop through the ips of my neighbor
!*** and try to find an ip with matching nodes
!*** also try to match with node twins
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
checkCandidateIP: do candidateIP = 1_pInt,FE_Nips(neighboringType)
NmatchingNodes = 0_pInt
matchingNodes = 0_pInt
do a = 1_pInt,FE_maxNnodesAtIP(neighboringType) ! check each anchor node of that ip
anchor = FE_nodesAtIP(a,candidateIP,neighboringType)
if (anchor /= 0_pInt) then ! valid anchor node
if (any(FE_face(:,matchingFace,neighboringType) == anchor)) then ! sits on matching face?
NmatchingNodes = NmatchingNodes + 1_pInt
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
matchingNodes(NmatchingNodes) = mesh_element(4+anchor,matchingElem) ! CP id of neighbor's anchor node
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
else ! no matching, because not all nodes sit on the matching face
NmatchingNodes = 0_pInt
matchingNodes = 0_pInt
exit
endif
endif
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
if (NmatchingNodes /= NlinkedNodes) & ! this ip has wrong count of anchors on face
cycle checkCandidateIP
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** check "normal" nodes whether they match or not
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
checkTwins = .false.
do a = 1_pInt,NlinkedNodes
if (all(matchingNodes /= linkedNodes(a))) then ! this linkedNode does not match any matchingNode
checkTwins = .true.
exit ! no need to search further
endif
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** if no match found, then also check node twins
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
if(checkTwins) then
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
twin_of_linkedNode = mesh_nodeTwins(dir,linkedNodes(a))
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
endif
enddo
endif
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
!*** we found a match !!!
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
mesh_ipNeighborhood(1,neighbor,myIP,myElem) = matchingElem
mesh_ipNeighborhood(2,neighbor,myIP,myElem) = candidateIP
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
exit checkCandidateIP
fixed small bug when calculating mesh_Ncells
2013-04-19 18:11:06 +05:30
enddo checkCandidateIP
endif ! end of valid external matching
endif ! end of internal/external matching
enddo
enddo
enddo
do myElem = 1_pInt,mesh_NcpElems ! loop over cpElems
myType = FE_geomtype(mesh_element(2,myElem)) ! get elemGeomType
do myIP = 1_pInt,FE_Nips(myType) ! loop over IPs of elem
do neighbor = 1_pInt,FE_NipNeighbors(FE_celltype(myType)) ! loop over neighbors of IP
neighboringElem = mesh_ipNeighborhood(1,neighbor,myIP,myElem)
neighboringIP = mesh_ipNeighborhood(2,neighbor,myIP,myElem)
if (neighboringElem > 0_pInt .and. neighboringIP > 0_pInt) then ! if neighbor exists ...
neighboringType = FE_geomtype(mesh_element(2,neighboringElem))
do pointingToMe = 1_pInt,FE_NipNeighbors(FE_celltype(neighboringType)) ! find neighboring index that points from my neighbor to myself
if ( myElem == mesh_ipNeighborhood(1,pointingToMe,neighboringIP,neighboringElem) &
.and. myIP == mesh_ipNeighborhood(2,pointingToMe,neighboringIP,neighboringElem)) then ! possible candidate
if (math_mul3x3(mesh_ipAreaNormal(1:3,neighbor,myIP,myElem),&
mesh_ipAreaNormal(1:3,pointingToMe,neighboringIP,neighboringElem)) < 0.0_pReal) then ! area normals have opposite orientation (we have to check that because of special case for single element with two ips and periodicity. In this case the neighbor is identical in two different directions.)
mesh_ipNeighborhood(3,neighbor,myIP,myElem) = pointingToMe ! found match
exit ! so no need to search further
endif
endif
enddo
endif
enddo
enddo
enddo
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +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
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
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief mapping of FE element types to internal representation
!--------------------------------------------------------------------------------------------------
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
integer(pInt) function FE_mapElemtype(what)
- added element type 54 for macro (2d 8node 4ip reduced integration) - automatically raising error for unknown element type inside mapping function
2013-04-10 15:08:40 +05:30
use IO, only: IO_lc, IO_error
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
implicit none
character(len=*), intent(in) :: what
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
select case (IO_lc(what))
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
case ( '6')
FE_mapElemtype = 1_pInt ! Two-dimensional Plane Strain Triangle
case ( '155', &
'125', &
'128')
FE_mapElemtype = 2_pInt ! Two-dimensional Plane Strain triangle (155: cubic shape function, 125/128: second order isoparametric)
define functions where needed only use solver specific element names
2019-01-31 15:59:56 +05:30
case ( '11')
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
FE_mapElemtype = 3_pInt ! Arbitrary Quadrilateral Plane-strain
define functions where needed only use solver specific element names
2019-01-31 15:59:56 +05:30
case ( '27')
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
FE_mapElemtype = 4_pInt ! Plane Strain, Eight-node Distorted Quadrilateral
- added element type 54 for macro (2d 8node 4ip reduced integration) - automatically raising error for unknown element type inside mapping function
2013-04-10 15:08:40 +05:30
case ( '54')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 5_pInt ! Plane Strain, Eight-node Distorted Quadrilateral with reduced integration
define functions where needed only use solver specific element names
2019-01-31 15:59:56 +05:30
case ( '134')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 6_pInt ! Three-dimensional Four-node Tetrahedron
Added thermo-mechanical element type for ABAQUS
2014-10-23 13:47:37 +05:30
case ( '157')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 7_pInt ! Three-dimensional, Low-order, Tetrahedron, Herrmann Formulations
Added thermo-mechanical element type for ABAQUS
2014-10-23 13:47:37 +05:30
case ( '127')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 8_pInt ! Three-dimensional Ten-node Tetrahedron
define functions where needed only use solver specific element names
2019-01-31 15:59:56 +05:30
case ( '136')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 9_pInt ! Three-dimensional Arbitrarily Distorted Pentahedral
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
case ( '117', &
define functions where needed only use solver specific element names
2019-01-31 15:59:56 +05:30
'123')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 10_pInt ! Three-dimensional Arbitrarily Distorted linear hexahedral with reduced integration
define functions where needed only use solver specific element names
2019-01-31 15:59:56 +05:30
case ( '7')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 11_pInt ! Three-dimensional Arbitrarily Distorted Brick
define functions where needed only use solver specific element names
2019-01-31 15:59:56 +05:30
case ( '57')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 12_pInt ! Three-dimensional Arbitrarily Distorted quad hexahedral with reduced integration
define functions where needed only use solver specific element names
2019-01-31 15:59:56 +05:30
case ( '21')
fixed bug that was introduced in the element type mapping in version 2305
2013-04-10 18:52:02 +05:30
FE_mapElemtype = 13_pInt ! Three-dimensional Arbitrarily Distorted quadratic hexahedral
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
case default
- added element type 54 for macro (2d 8node 4ip reduced integration) - automatically raising error for unknown element type inside mapping function
2013-04-10 15:08:40 +05:30
call IO_error(error_ID=190_pInt,ext_msg=IO_lc(what))
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
end select
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
end function FE_mapElemtype
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
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief find face-matching element of same type
!--------------------------------------------------------------------------------------------------
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
subroutine mesh_faceMatch(elem, face ,matchingElem, matchingFace)
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
implicit none
integer(pInt), intent(out) :: matchingElem, & ! matching CP element ID
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
matchingFace ! matching face ID
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
integer(pInt), intent(in) :: face, & ! face ID
elem ! CP elem ID
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
integer(pInt), dimension(FE_NmatchingNodesPerFace(face,FE_geomtype(mesh_element(2,elem)))) :: &
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
myFaceNodes ! global node ids on my face
integer(pInt) :: myType, &
candidateType, &
candidateElem, &
candidateFace, &
candidateFaceNode, &
minNsharedElems, &
NsharedElems, &
lonelyNode = 0_pInt, &
i, &
n, &
dir ! periodicity direction
integer(pInt), dimension(:), allocatable :: element_seen
logical checkTwins
improved regridding, especially interfacing to python (also for nearest neighbor search)
2012-04-24 22:32:27 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
matchingElem = 0_pInt
matchingFace = 0_pInt
minNsharedElems = mesh_maxNsharedElems + 1_pInt ! init to worst case
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
myType = FE_geomtype(mesh_element(2_pInt,elem)) ! figure elemGeomType
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
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
do n = 1_pInt,FE_NmatchingNodesPerFace(face,myType) ! loop over nodes on face
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
myFaceNodes(n) = mesh_element(4_pInt+FE_face(n,face,myType),elem) ! CP id of face node
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
NsharedElems = mesh_sharedElem(1_pInt,myFaceNodes(n)) ! figure # shared elements for this node
if (NsharedElems < minNsharedElems) then
minNsharedElems = NsharedElems ! remember min # shared elems
lonelyNode = n ! remember most lonely node
endif
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
allocate(element_seen(minNsharedElems))
element_seen = 0_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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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?
element_seen(i) = candidateElem
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
candidateType = FE_geomtype(mesh_element(2_pInt,candidateElem)) ! figure elemGeomType of candidate
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
checkCandidateFace: do candidateFace = 1_pInt,FE_maxNipNeighbors ! check each face of candidate
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
if (FE_NmatchingNodesPerFace(candidateFace,candidateType) &
/= FE_NmatchingNodesPerFace(face,myType) & ! incompatible face
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
.or. (candidateElem == elem .and. candidateFace == face)) then ! this is my face
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
cycle checkCandidateFace
endif
checkTwins = .false.
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
do n = 1_pInt,FE_NmatchingNodesPerFace(candidateFace,candidateType) ! loop through nodes on face
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
candidateFaceNode = mesh_element(4_pInt+FE_face(n,candidateFace,candidateType),candidateElem)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
exit
endif
enddo
if(checkTwins) then
checkCandidateFaceTwins: do dir = 1_pInt,3_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
do n = 1_pInt,FE_NmatchingNodesPerFace(candidateFace,candidateType) ! loop through nodes on face
calculation of cellnode positions now available in python via damask.core first call damask.core.mesh.mesh_init_postprocessing(meshfilename) to initialize all necessary mesh variables then damask.core.mesh.mesh_build_cellnodes(nodes) calculates the cellnode positions for a given list of node positions the meshfile that is needed for the init is created automatically by mesh_init in DAMASK !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-22 00:18:59 +05:30
candidateFaceNode = mesh_element(4+FE_face(n,candidateFace,candidateType),candidateElem)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
if (all(myFaceNodes /= mesh_nodeTwins(dir,candidateFaceNode))) then ! node twin does not match either
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
if (dir == 3_pInt) then
cycle checkCandidateFace
else
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
cycle checkCandidateFaceTwins ! try twins in next dimension
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
endif
endif
enddo
exit checkCandidateFaceTwins
enddo checkCandidateFaceTwins
endif
matchingFace = candidateFace
matchingElem = candidateElem
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
exit checkCandidate ! found my matching candidate
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
enddo checkCandidateFace
endif
enddo checkCandidate
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
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
end subroutine mesh_faceMatch
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
integrated subroutine for regridding into mesh.f90 and made it available for python
2012-04-10 20:45:46 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!--------------------------------------------------------------------------------------------------
!> @brief get properties of different types of finite elements
!> @details assign globals: FE_nodesAtIP, FE_ipNeighbor, FE_cellnodeParentnodeWeights, FE_subNodeOnIPFace
!--------------------------------------------------------------------------------------------------
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
subroutine mesh_build_FEdata
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
implicit none
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
integer(pInt) :: me
leaner syntax with sourced allocation
2018-09-23 20:56:13 +05:30
allocate(FE_nodesAtIP(FE_maxmaxNnodesAtIP,FE_maxNips,FE_Ngeomtypes), source=0_pInt)
allocate(FE_ipNeighbor(FE_maxNipNeighbors,FE_maxNips,FE_Ngeomtypes), source=0_pInt)
allocate(FE_cell(FE_maxNcellnodesPerCell,FE_maxNips,FE_Ngeomtypes), source=0_pInt)
allocate(FE_cellnodeParentnodeWeights(FE_maxNnodes,FE_maxNcellnodes,FE_Nelemtypes), source=0.0_pReal)
allocate(FE_cellface(FE_maxNcellnodesPerCellface,FE_maxNcellfaces,FE_Ncelltypes), source=0_pInt)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
!*** fill FE_nodesAtIP with data ***
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = 0_pInt
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 6 (2D 3node 1ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1,2,3 &
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
several small corrections/polishing
2012-04-20 17:28:41 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 125 (2D 6node 3ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
1, &
2, &
3 &
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
fixed bug concerning file handling.
2012-04-12 00:16:36 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 11 (2D 4node 4ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
1, &
2, &
4, &
3 &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
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
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 27 (2D 8node 9ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
1,0, &
1,2, &
2,0, &
1,4, &
0,0, &
2,3, &
4,0, &
3,4, &
3,0 &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
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
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 134 (3D 4node 1ip)
reshape(int([&
1,2,3,4 &
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 127 (3D 10node 4ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
1, &
2, &
3, &
4 &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
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
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 136 (3D 6node 6ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
1, &
2, &
3, &
4, &
5, &
6 &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
several small corrections/polishing
2012-04-20 17:28:41 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 117 (3D 8node 1ip)
reshape(int([&
1,2,3,4,5,6,7,8 &
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 7 (3D 8node 8ip)
reshape(int([&
1, &
2, &
4, &
3, &
5, &
6, &
8, &
7 &
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
me = me + 1_pInt
FE_nodesAtIP(1:FE_maxNnodesAtIP(me),1:FE_Nips(me),me) = & ! element 21 (3D 20node 27ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
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 &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
],pInt),[FE_maxNnodesAtIP(me),FE_Nips(me)])
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
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +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.
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = 0_pInt
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 6 (2D 3node 1ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
-2,-3,-1 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 125 (2D 6node 3ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
2,-3, 3,-1, &
-2, 1, 3,-1, &
2,-3,-2, 1 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 11 (2D 4node 4ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
2,-4, 3,-1, &
-2, 1, 4,-1, &
4,-4,-3, 1, &
-2, 3,-3, 2 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 27 (2D 8node 9ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
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 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 134 (3D 4node 1ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([&
-1,-2,-3,-4 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 127 (3D 10node 4ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
2,-4, 3,-2, 4,-1, &
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
-2, 1, 3,-2, 4,-1, &
2,-4,-3, 1, 4,-1, &
2,-4, 3,-2,-3, 1 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 136 (3D 6node 6ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
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 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 117 (3D 8node 1ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([&
-3,-5,-4,-2,-6,-1 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 7 (3D 8node 8ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([&
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 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_ipNeighbor(1:FE_NipNeighbors(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 21 (3D 20node 27ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
2,-5, 4,-2,10,-1, &
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, &
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, &
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 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pInt),[FE_NipNeighbors(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
! *** FE_cell ***
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = 0_pInt
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 6 (2D 3node 1ip)
reshape(int([&
1,2,3 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 125 (2D 6node 3ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 4, 7, 6, &
2, 5, 7, 4, &
3, 6, 7, 5 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 11 (2D 4node 4ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 5, 9, 8, &
5, 2, 6, 9, &
8, 9, 7, 4, &
9, 6, 3, 7 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 27 (2D 8node 9ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 5,13,12, &
5, 6,14,13, &
6, 2, 7,14, &
12,13,16,11, &
13,14,15,16, &
14, 7, 8,15, &
11,16,10, 4, &
16,15, 9,10, &
15, 8, 3, 9 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 134 (3D 4node 1ip)
reshape(int([&
1, 2, 3, 4 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 127 (3D 10node 4ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 5,11, 7, 8,12,15,14, &
5, 2, 6,11,12, 9,13,15, &
7,11, 6, 3,14,15,13,10, &
8,12,15, 4, 4, 9,13,10 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 136 (3D 6node 6ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 7,16, 9,10,17,21,19, &
7, 2, 8,16,17,11,18,21, &
9,16, 8, 3,19,21,18,12, &
10,17,21,19, 4,13,20,15, &
17,11,18,21,13, 5,14,20, &
19,21,18,12,15,20,14, 6 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 117 (3D 8node 1ip)
reshape(int([&
1, 2, 3, 4, 5, 6, 7, 8 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 7 (3D 8node 8ip)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 9,21,12,13,22,27,25, &
9, 2,10,21,22,14,23,27, &
12,21,11, 4,25,27,24,16, &
21,10, 3,11,27,23,15,24, &
13,22,27,25, 5,17,26,20, &
22,14,23,27,17, 6,18,26, &
25,27,24,16,20,26,19, 8, &
27,23,15,24,26,18, 7,19 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cell(1:FE_NcellnodesPerCell(FE_celltype(me)),1:FE_Nips(me),me) = & ! element 21 (3D 20node 27ip)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 9,33,16,17,37,57,44, &
9,10,34,33,37,38,58,57, &
10, 2,11,34,38,18,39,58, &
16,33,36,15,44,57,60,43, &
33,34,35,36,57,58,59,60, &
34,11,12,35,58,39,40,59, &
15,36,14, 4,43,60,42,20, &
36,35,13,14,60,59,41,42, &
35,12, 3,13,59,40,19,41, &
17,37,57,44,21,45,61,52, &
37,38,58,57,45,46,62,61, &
38,18,39,58,46,22,47,62, &
44,57,60,43,52,61,64,51, &
57,58,59,60,61,62,63,64, &
58,39,40,59,62,47,48,63, &
43,60,42,20,51,64,50,24, &
60,59,41,42,64,63,49,50, &
59,40,19,41,63,48,23,49, &
21,45,61,52, 5,25,53,32, &
45,46,62,61,25,26,54,53, &
46,22,47,62,26, 6,27,54, &
52,61,64,51,32,53,56,31, &
61,62,63,64,53,54,55,56, &
62,47,48,63,54,27,28,55, &
51,64,50,24,31,56,30, 8, &
64,63,49,50,56,55,29,30, &
63,48,23,49,55,28, 7,29 &
],pInt),[FE_NcellnodesPerCell(FE_celltype(me)),FE_Nips(me)])
! *** FE_cellnodeParentnodeWeights ***
! center of gravity of the weighted nodes gives the position of the cell node.
! fill with 0.
! example: face-centered cell node with face nodes 1,2,5,6 to be used in,
! e.g., an 8 node element, would be encoded:
! 1, 1, 0, 0, 1, 1, 0, 0
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = 0_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
me = me + 1_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 6 (2D 3node 1ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
reshape(real([&
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
1, 0, 0, &
0, 1, 0, &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
0, 0, 1 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
me = me + 1_pInt
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 125 (2D 6node 3ip)
reshape(real([&
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
1, 0, 0, 0, 0, 0, &
0, 1, 0, 0, 0, 0, &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
0, 0, 1, 0, 0, 0, &
0, 0, 0, 1, 0, 0, &
0, 0, 0, 0, 1, 0, &
0, 0, 0, 0, 0, 1, &
1, 1, 1, 2, 2, 2 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 11 (2D 4node 4ip)
reshape(real([&
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
1, 0, 0, 0, &
0, 1, 0, 0, &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
0, 0, 1, 0, &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
0, 0, 0, 1, &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 1, 0, 0, &
0, 1, 1, 0, &
0, 0, 1, 1, &
1, 0, 0, 1, &
1, 1, 1, 1 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 27 (2D 8node 9ip)
reshape(real([&
1, 0, 0, 0, 0, 0, 0, 0, &
0, 1, 0, 0, 0, 0, 0, 0, &
0, 0, 1, 0, 0, 0, 0, 0, &
0, 0, 0, 1, 0, 0, 0, 0, &
1, 0, 0, 0, 2, 0, 0, 0, &
0, 1, 0, 0, 2, 0, 0, 0, &
0, 1, 0, 0, 0, 2, 0, 0, &
0, 0, 1, 0, 0, 2, 0, 0, &
0, 0, 1, 0, 0, 0, 2, 0, &
0, 0, 0, 1, 0, 0, 2, 0, &
0, 0, 0, 1, 0, 0, 0, 2, &
1, 0, 0, 0, 0, 0, 0, 2, &
4, 1, 1, 1, 8, 2, 2, 8, &
1, 4, 1, 1, 8, 8, 2, 2, &
1, 1, 4, 1, 2, 8, 8, 2, &
1, 1, 1, 4, 2, 2, 8, 8 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 54 (2D 8node 4ip)
reshape(real([&
1, 0, 0, 0, 0, 0, 0, 0, &
0, 1, 0, 0, 0, 0, 0, 0, &
0, 0, 1, 0, 0, 0, 0, 0, &
0, 0, 0, 1, 0, 0, 0, 0, &
0, 0, 0, 0, 1, 0, 0, 0, &
0, 0, 0, 0, 0, 1, 0, 0, &
0, 0, 0, 0, 0, 0, 1, 0, &
0, 0, 0, 0, 0, 0, 0, 1, &
1, 1, 1, 1, 2, 2, 2, 2 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 134 (3D 4node 1ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
reshape(real([&
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
1, 0, 0, 0, &
0, 1, 0, 0, &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
0, 0, 1, 0, &
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
0, 0, 0, 1 &
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 157 (3D 5node 4ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
reshape(real([&
1, 0, 0, 0, 0, &
0, 1, 0, 0, 0, &
0, 0, 1, 0, 0, &
0, 0, 0, 1, 0, &
1, 1, 0, 0, 0, &
0, 1, 1, 0, 0, &
1, 0, 1, 0, 0, &
1, 0, 0, 1, 0, &
0, 1, 0, 1, 0, &
0, 0, 1, 1, 0, &
1, 1, 1, 0, 0, &
1, 1, 0, 1, 0, &
0, 1, 1, 1, 0, &
1, 0, 1, 1, 0, &
0, 0, 0, 0, 1 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 127 (3D 10node 4ip)
reshape(real([&
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, &
0, 1, 0, 0, 0, 0, 0, 0, 0, 0, &
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, &
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, &
0, 0, 0, 0, 1, 0, 0, 0, 0, 0, &
0, 0, 0, 0, 0, 1, 0, 0, 0, 0, &
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, &
0, 0, 0, 0, 0, 0, 0, 1, 0, 0, &
0, 0, 0, 0, 0, 0, 0, 0, 1, 0, &
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, &
1, 1, 1, 0, 2, 2, 2, 0, 0, 0, &
1, 1, 0, 1, 2, 0, 0, 2, 2, 0, &
0, 1, 1, 1, 0, 2, 0, 0, 2, 2, &
1, 0, 1, 1, 0, 0, 2, 2, 0, 2, &
3, 3, 3, 3, 4, 4, 4, 4, 4, 4 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
me = me + 1_pInt
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 136 (3D 6node 6ip)
reshape(real([&
1, 0, 0, 0, 0, 0, &
0, 1, 0, 0, 0, 0, &
0, 0, 1, 0, 0, 0, &
0, 0, 0, 1, 0, 0, &
0, 0, 0, 0, 1, 0, &
0, 0, 0, 0, 0, 1, &
1, 1, 0, 0, 0, 0, &
0, 1, 1, 0, 0, 0, &
1, 0, 1, 0, 0, 0, &
1, 0, 0, 1, 0, 0, &
0, 1, 0, 0, 1, 0, &
0, 0, 1, 0, 0, 1, &
0, 0, 0, 1, 1, 0, &
0, 0, 0, 0, 1, 1, &
0, 0, 0, 1, 0, 1, &
1, 1, 1, 0, 0, 0, &
1, 1, 0, 1, 1, 0, &
0, 1, 1, 0, 1, 1, &
1, 0, 1, 1, 0, 1, &
0, 0, 0, 1, 1, 1, &
1, 1, 1, 1, 1, 1 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
me = me + 1_pInt
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 117 (3D 8node 1ip)
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
reshape(real([&
1, 0, 0, 0, 0, 0, 0, 0, &
0, 1, 0, 0, 0, 0, 0, 0, &
0, 0, 1, 0, 0, 0, 0, 0, &
0, 0, 0, 1, 0, 0, 0, 0, &
0, 0, 0, 0, 1, 0, 0, 0, &
0, 0, 0, 0, 0, 1, 0, 0, &
0, 0, 0, 0, 0, 0, 1, 0, &
0, 0, 0, 0, 0, 0, 0, 1 &
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
me = me + 1_pInt
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 7 (3D 8node 8ip)
reshape(real([&
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
1, 0, 0, 0, 0, 0, 0, 0, & !
0, 1, 0, 0, 0, 0, 0, 0, & !
0, 0, 1, 0, 0, 0, 0, 0, & !
0, 0, 0, 1, 0, 0, 0, 0, & !
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
0, 0, 0, 0, 1, 0, 0, 0, & ! 5
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
0, 0, 0, 0, 0, 1, 0, 0, & !
0, 0, 0, 0, 0, 0, 1, 0, & !
0, 0, 0, 0, 0, 0, 0, 1, & !
1, 1, 0, 0, 0, 0, 0, 0, & !
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
0, 1, 1, 0, 0, 0, 0, 0, & ! 10
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
0, 0, 1, 1, 0, 0, 0, 0, & !
1, 0, 0, 1, 0, 0, 0, 0, & !
1, 0, 0, 0, 1, 0, 0, 0, & !
0, 1, 0, 0, 0, 1, 0, 0, & !
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
0, 0, 1, 0, 0, 0, 1, 0, & ! 15
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
0, 0, 0, 1, 0, 0, 0, 1, & !
0, 0, 0, 0, 1, 1, 0, 0, & !
0, 0, 0, 0, 0, 1, 1, 0, & !
0, 0, 0, 0, 0, 0, 1, 1, & !
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
0, 0, 0, 0, 1, 0, 0, 1, & ! 20
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
1, 1, 1, 1, 0, 0, 0, 0, & !
1, 1, 0, 0, 1, 1, 0, 0, & !
0, 1, 1, 0, 0, 1, 1, 0, & !
0, 0, 1, 1, 0, 0, 1, 1, & !
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1, 0, 0, 1, 1, 0, 0, 1, & ! 25
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
0, 0, 0, 0, 1, 1, 1, 1, & !
1, 1, 1, 1, 1, 1, 1, 1 & !
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
me = me + 1_pInt
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 57 (3D 20node 8ip)
reshape(real([&
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & ! 5
0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & ! 10
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 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, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, & ! 15
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, & !
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, & ! 20
1, 1, 1, 1, 0, 0, 0, 0, 2, 2, 2, 2, 0, 0, 0, 0, 0, 0, 0, 0, & !
1, 1, 0, 0, 1, 1, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, 2, 2, 0, 0, & !
0, 1, 1, 0, 0, 1, 1, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, 2, 2, 0, & !
0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 2, 0, 0, 0, 2, 0, 0, 0, 2, 2, & !
1, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 2, 0, 0, 0, 2, 2, 0, 0, 2, & ! 25
0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 2, 2, 2, 2, 0, 0, 0, 0, & !
3, 3, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4 & !
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
FE_cellnodeParentnodeWeights(1:FE_Nnodes(me),1:FE_Ncellnodes(FE_geomtype(me)),me) = & ! element 21 (3D 20node 27ip)
reshape(real([&
changes towards supporting new Marc2017 input file format still not working
2018-01-10 21:43:25 +05:30
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & ! 5
0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 1, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & ! 10
0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 1, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, & ! 15
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, & !
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, & !
0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, & !
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, & !
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, & ! 20
0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, & !
0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, & !
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, & !
0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, & !
0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, & ! 25
0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, & !
0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, & ! 30
0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, & !
0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, & !
4, 1, 1, 1, 0, 0, 0, 0, 8, 2, 2, 8, 0, 0, 0, 0, 0, 0, 0, 0, & !
1, 4, 1, 1, 0, 0, 0, 0, 8, 8, 2, 2, 0, 0, 0, 0, 0, 0, 0, 0, & !
1, 1, 4, 1, 0, 0, 0, 0, 2, 8, 8, 2, 0, 0, 0, 0, 0, 0, 0, 0, & ! 35
1, 1, 1, 4, 0, 0, 0, 0, 2, 2, 8, 8, 0, 0, 0, 0, 0, 0, 0, 0, & !
4, 1, 0, 0, 1, 1, 0, 0, 8, 0, 0, 0, 2, 0, 0, 0, 8, 2, 0, 0, & !
1, 4, 0, 0, 1, 1, 0, 0, 8, 0, 0, 0, 2, 0, 0, 0, 2, 8, 0, 0, & !
0, 4, 1, 0, 0, 1, 1, 0, 0, 8, 0, 0, 0, 2, 0, 0, 0, 8, 2, 0, & !
0, 1, 4, 0, 0, 1, 1, 0, 0, 8, 0, 0, 0, 2, 0, 0, 0, 2, 8, 0, & ! 40
0, 0, 4, 1, 0, 0, 1, 1, 0, 0, 8, 0, 0, 0, 2, 0, 0, 0, 8, 2, & !
0, 0, 1, 4, 0, 0, 1, 1, 0, 0, 8, 0, 0, 0, 2, 0, 0, 0, 2, 8, & !
1, 0, 0, 4, 1, 0, 0, 1, 0, 0, 0, 8, 0, 0, 0, 2, 2, 0, 0, 8, & !
4, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 8, 0, 0, 0, 2, 8, 0, 0, 2, & !
1, 1, 0, 0, 4, 1, 0, 0, 2, 0, 0, 0, 8, 0, 0, 0, 8, 2, 0, 0, & ! 45
1, 1, 0, 0, 1, 4, 0, 0, 2, 0, 0, 0, 8, 0, 0, 0, 2, 8, 0, 0, & !
0, 1, 1, 0, 0, 4, 1, 0, 0, 2, 0, 0, 0, 8, 0, 0, 0, 8, 2, 0, & !
0, 1, 1, 0, 0, 1, 4, 0, 0, 2, 0, 0, 0, 8, 0, 0, 0, 2, 8, 0, & !
0, 0, 1, 1, 0, 0, 4, 1, 0, 0, 2, 0, 0, 0, 8, 0, 0, 0, 8, 2, & !
0, 0, 1, 1, 0, 0, 1, 4, 0, 0, 2, 0, 0, 0, 8, 0, 0, 0, 2, 8, & ! 50
1, 0, 0, 1, 1, 0, 0, 4, 0, 0, 0, 2, 0, 0, 0, 8, 2, 0, 0, 8, & !
1, 0, 0, 1, 4, 0, 0, 1, 0, 0, 0, 2, 0, 0, 0, 8, 8, 0, 0, 2, & !
0, 0, 0, 0, 4, 1, 1, 1, 0, 0, 0, 0, 8, 2, 2, 8, 0, 0, 0, 0, & !
0, 0, 0, 0, 1, 4, 1, 1, 0, 0, 0, 0, 8, 8, 2, 2, 0, 0, 0, 0, & !
0, 0, 0, 0, 1, 1, 4, 1, 0, 0, 0, 0, 2, 8, 8, 2, 0, 0, 0, 0, & ! 55
0, 0, 0, 0, 1, 1, 1, 4, 0, 0, 0, 0, 2, 2, 8, 8, 0, 0, 0, 0, & !
24, 8, 4, 8, 8, 4, 3, 4, 32,12,12,32, 12, 4, 4,12, 32,12, 4,12, & !
8,24, 8, 4, 4, 8, 4, 3, 32,32,12,12, 12,12, 4, 4, 12,32,12, 4, & !
4, 8,24, 8, 3, 4, 8, 4, 12,32,32,12, 4,12,12, 4, 4,12,32,12, & !
8, 4, 8,24, 4, 3, 4, 8, 12,12,32,32, 4, 4,12,12, 12, 4,12,32, & ! 60
8, 4, 3, 4, 24, 8, 4, 8, 12, 4, 4,12, 32,12,12,32, 32,12, 4,12, & !
4, 8, 4, 3, 8,24, 8, 4, 12,12, 4, 4, 32,32,12,12, 12,32,12, 4, & !
3, 4, 8, 4, 4, 8,24, 8, 4,12,12, 4, 12,32,32,12, 4,12,32,12, & !
4, 3, 4, 8, 8, 4, 8,24, 4, 4,12,12, 12,12,32,32, 12, 4,12,32 & !
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
],pReal),[FE_Nnodes(me),FE_Ncellnodes(FE_geomtype(me))])
! *** FE_cellface ***
me = 0_pInt
me = me + 1_pInt
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
FE_cellface(1:FE_NcellnodesPerCellface(me),1:FE_NipNeighbors(me),me) = & ! 2D 3node, VTK_TRIANGLE (5)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
2,3, &
3,1, &
1,2 &
],pInt),[FE_NcellnodesPerCellface(me),FE_NipNeighbors(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
FE_cellface(1:FE_NcellnodesPerCellface(me),1:FE_NipNeighbors(me),me) = & ! 2D 4node, VTK_QUAD (9)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
2,3, &
4,1, &
3,4, &
1,2 &
],pInt),[FE_NcellnodesPerCellface(me),FE_NipNeighbors(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
FE_cellface(1:FE_NcellnodesPerCellface(me),1:FE_NipNeighbors(me),me) = & ! 3D 4node, VTK_TETRA (10)
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
1,3,2, &
1,2,4, &
2,3,4, &
1,4,3 &
],pInt),[FE_NcellnodesPerCellface(me),FE_NipNeighbors(me)])
switched element library to geomType based. saves to copy same geometry description for different elements that are essentially similar regarding the IP number but differ in total node count. introduced quadratic tetrahedron (Marc element 127 -- element 157 might also work, but did not perform well in fully elastic calc so far)
2012-11-16 04:15:20 +05:30
me = me + 1_pInt
introduced output of initial geometry (vtk file) to mesh
2013-04-18 22:10:49 +05:30
FE_cellface(1:FE_NcellnodesPerCellface(me),1:FE_NipNeighbors(me),me) = & ! 3D 8node, VTK_HEXAHEDRON (12)
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
reshape(int([&
Major restructuring of mesh.f90 due to introduction of "cells". A cell is the volume (area in 2D elements) that surrounds an ip. It is described by up to 8 cell nodes and is used to calculate ip volume, area, and face normals. It can also be used to generate a mesh for e.g. paraview that contains ip output as cell centered data. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! For marc simulations, run ./code/setup/setup_code.sh !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2013-04-15 13:43:20 +05:30
2,3,7,6, &
4,1,5,8, &
3,4,8,7, &
1,2,6,5, &
5,6,7,8, &
1,4,3,2 &
],pInt),[FE_NcellnodesPerCellface(me),FE_NipNeighbors(me)])
introduced preprocessor identifiers Marc, Abaqus, and Spectral to enable conditional compilation. This allows deleted dummy functions that are used by one solver only. Mainly affected modules are IO and mesh. Most of the changes in mesh result from reordering the functions when grouping them depending on their solver. Further advantage is that FE solver do not need FFTW and kdtree2 anymore. The include files for these two libraries moved to DAMASKROO/lib now as I figured out how to use a include path in the Makefile. Put all the files I got when testing compilation with abaqus in a folder which to become the abaqus compilation test.
2012-06-15 21:40:21 +05:30
end subroutine mesh_build_FEdata
fixed reading/writing of integer arrays with function for real arrays
2012-08-16 20:25:23 +05:30
end module mesh