DAMASK_EICMD/code/DAMASK_marc.f90

409 lines
17 KiB
Fortran

! Copyright 2011 Max-Planck-Institut für Eisenforschung GmbH
!
! This file is part of DAMASK,
! the Düsseldorf Advanced MAterial Simulation Kit.
!
! DAMASK is free software: you can redistribute it and/or modify
! it under the terms of the GNU General Public License as published by
! the Free Software Foundation, either version 3 of the License, or
! (at your option) any later version.
!
! DAMASK is distributed in the hope that it will be useful,
! but WITHOUT ANY WARRANTY; without even the implied warranty of
! MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
! GNU General Public License for more details.
!
! You should have received a copy of the GNU General Public License
! along with DAMASK. If not, see <http://www.gnu.org/licenses/>.
!
!##############################################################
!* $Id$
!********************************************************************
! Material subroutine for MSC.Marc
!
! written by P. Eisenlohr,
! F. Roters,
! L. Hantcherli,
! W.A. Counts
! D.D. Tjahjanto
! C. Kords
!
! MPI fuer Eisenforschung, Duesseldorf
!
!********************************************************************
! Usage:
! - choose material as hypela2
! - set statevariable 2 to index of homogenization
! - set statevariable 3 to index of microstructure
! - make sure the file "material.config" exists in the working
! directory
! - make sure the file "numerics.config" exists in the working
! directory
! - use nonsymmetric option for solver (e.g. direct
! profile or multifrontal sparse, the latter seems
! to be faster!)
! - in case of ddm (domain decomposition)a SYMMETRIC
! solver has to be used, i.e uncheck "non-symmetric"
!********************************************************************
! Marc subroutines used:
! - hypela2
! - plotv
! - quit
!********************************************************************
! Marc common blocks included:
! - concom: lovl, ncycle, inc, incsub
! - creeps: timinc
!********************************************************************
!
#include "prec.f90"
MODULE DAMASK_interface
character(len=64), parameter :: FEsolver = 'Marc'
character(len=4), parameter :: InputFileExtension = '.dat'
character(len=4), parameter :: LogFileExtension = '.log'
CONTAINS
subroutine DAMASK_interface_init()
!$OMP CRITICAL (write2out)
write(6,*)
write(6,*) '<<<+- DAMASK_marc init -+>>>'
write(6,*) '$Id$'
write(6,*)
!$OMP END CRITICAL (write2out)
return
end subroutine
function getSolverWorkingDirectoryName()
implicit none
character(1024) getSolverWorkingDirectoryName, outName
character(len=*), parameter :: pathSep = achar(47)//achar(92) ! forward and backward slash
getSolverWorkingDirectoryName=''
outName=''
inquire(6, name=outName) ! determine outputfile
getSolverWorkingDirectoryName=outName(1:scan(outName,pathSep,back=.true.))
! write(6,*) 'getSolverWorkingDirectoryName', getSolverWorkingDirectoryName
end function
function getModelName()
use prec
implicit none
character(1024) getModelName
getModelName = getSolverJobName()
end function
function getSolverJobName()
use prec
implicit none
character(1024) getSolverJobName, outName
character(len=*), parameter :: pathSep = achar(47)//achar(92) ! forward and backward slash
integer(pInt) extPos
getSolverJobName=''
outName=''
inquire(6, name=outName) ! determine outputfile
extPos = len_trim(outName)-4
getSolverJobName=outName(scan(outName,pathSep,back=.true.)+1:extPos)
! write(6,*) 'getSolverJobName', getSolverJobName
end function
END MODULE
#include "IO.f90"
#include "numerics.f90"
#include "debug.f90"
#include "math.f90"
#include "FEsolving.f90"
#include "mesh.f90"
#include "material.f90"
#include "lattice.f90"
#include "constitutive_j2.f90"
#include "constitutive_phenopowerlaw.f90"
#include "constitutive_titanmod.f90"
#include "constitutive_dislotwin.f90"
#include "constitutive_nonlocal.f90"
#include "constitutive.f90"
#include "crystallite.f90"
#include "homogenization_isostrain.f90"
#include "homogenization_RGC.f90"
#include "homogenization.f90"
#include "CPFEM.f90"
!********************************************************************
! This is the Marc material routine
!********************************************************************
!
! ************* user subroutine for defining material behavior **************
!
!
! CAUTION : Due to calculation of the Deformation gradients, Stretch Tensors and
! Rotation tensors at previous and current states, the analysis can be
! computationally expensive. Please use the user subroutine -> hypela
! if these kinematic quantities are not needed in the constitutive model
!
!
! IMPORTANT NOTES :
!
! (1) F,R,U are only available for continuum and membrane elements (not for
! shells and beams).
!
! (2) For total Lagrangian formulation use the -> 'Elasticity,1' card(=
! total Lagrange with large disp) in the parameter section of input deck.
! For updated Lagrangian formulation use the -> 'Plasticity,3' card(=
! update+finite+large disp+constant d) in the parameter section of
! input deck.
!
! The following operation obtains U (stretch tensor) at t=n+1 :
!
! call scla(un1,0.d0,itel,itel,1)
! do 3 k=1,3
! do 2 i=1,3
! do 1 j=1,3
! un1(i,j)=un1(i,j)+dsqrt(strechn1(k))*eigvn1(i,k)*eigvn1(j,k)
!1 continue
!2 continue
!3 continue
!
!********************************************************************
subroutine hypela2(&
d,& ! stress strain law to be formed
g,& ! change in stress due to temperature effects
e,& ! total elastic strain
de,& ! increment of strain
s,& ! stress - should be updated by user
t,& ! state variables (comes in at t=n, must be updated to have state variables at t=n+1)
dt,& ! increment of state variables
ngens,& ! size of stress - strain law
n,& ! element number
nn,& ! integration point number
kcus,& ! (1) layer number, (2) internal layer number
matus,& ! (1) user material identification number, (2) internal material identification number
ndi,& ! number of direct components
nshear,& ! number of shear components
disp,& ! incremental displacements
dispt,& ! displacements at t=n (at assembly, lovl=4) and displacements at t=n+1 (at stress recovery, lovl=6)
coord,& ! coordinates
ffn,& ! deformation gradient
frotn,& ! rotation tensor
strechn,& ! square of principal stretch ratios, lambda(i)
eigvn,& ! i principal direction components for j eigenvalues
ffn1,& ! deformation gradient
frotn1,& ! rotation tensor
strechn1,& ! square of principal stretch ratios, lambda(i)
eigvn1,& ! i principal direction components for j eigenvalues
ncrd,& ! number of coordinates
itel,& ! dimension of F and R, either 2 or 3
ndeg,& ! number of degrees of freedom ==> is this at correct list position ?!?
ndm,& !
nnode,& ! number of nodes per element
jtype,& ! element type
lclass,& ! element class
ifr,& ! set to 1 if R has been calculated
ifu & ! set to 1 if stretch has been calculated
)
use prec, only: pReal, &
pInt
use FEsolving, only: cycleCounter, &
theInc, &
cutBack, &
calcMode, &
lastMode, &
theTime, &
theDelta, &
lastIncConverged, &
outdatedByNewInc, &
outdatedFFN1, &
terminallyIll, &
symmetricSolver
use math, only: invnrmMandel
use debug, only: debug_info, &
debug_reset
use mesh, only: mesh_FEasCP
use CPFEM, only: CPFEM_initAll,CPFEM_general,CPFEM_init_done
!$ use OMP_LIB ! the openMP function library
!$ use numerics, only: mpieNumThreadsInt ! number of threads set by MPIE_NUMTHREADS
implicit none
! ** Start of generated type statements **
real(pReal) coord, d, de, disp, dispt, dt, e, eigvn, eigvn1, ffn, ffn1
real(pReal) frotn, frotn1, g
integer(pInt) ifr, ifu, itel, jtype, kcus, lclass, matus, n, ncrd, ndeg
integer(pInt) ndi, ndm, ngens, nn, nnode, nshear
real(pReal) s, strechn, strechn1, t
! ** End of generated type statements **
dimension e(*),de(*),t(*),dt(*),g(*),d(ngens,*),s(*), n(2),coord(ncrd,*),disp(ndeg,*),matus(2),dispt(ndeg,*),ffn(itel,*),&
frotn(itel,*),strechn(itel),eigvn(itel,*),ffn1(itel,*),frotn1(itel,*),strechn1(itel),eigvn1(itel,*),kcus(2), lclass(2)
! Marc common blocks are in fixed format so they have to be reformated to free format (f90)
! Beware of changes in newer Marc versions
include "include/concom%%MARCVERSION%%" ! concom is needed for inc, subinc, ncycle, lovl
include "include/creeps%%MARCVERSION%%" ! creeps is needed for timinc (time increment)
real(pReal), dimension(6) :: stress
real(pReal), dimension(6,6) :: ddsdde
real(pReal), dimension (3,3) :: pstress ! dummy argument for call of cpfem_general (used by mpie_spectral)
real(pReal), dimension (3,3,3,3) :: dPdF ! dummy argument for call of cpfem_general (used by mpie_spectral)
integer(pInt) computationMode, i, cp_en
! OpenMP variable
!$ integer(pInt) defaultNumThreadsInt ! default value set by Marc
!$ defaultNumThreadsInt = omp_get_num_threads() ! remember number of threads set by Marc
if (.not. CPFEM_init_done) call CPFEM_initAll(t(1),n(1),nn)
!$ call omp_set_num_threads(mpieNumThreadsInt) ! set number of threads for parallel execution set by MPIE_NUM_THREADS
if (lovl == 4) then ! Marc requires stiffness in separate call
if ( timinc < theDelta .and. theInc == inc ) then ! first after cutback
computationMode = 7 ! --> restore tangent and return it
else
computationMode = 6 ! --> just return known tangent
endif
else ! stress requested (lovl == 6)
cp_en = mesh_FEasCP('elem',n(1))
if (cptim > theTime .or. inc /= theInc) then ! reached "convergence"
terminallyIll = .false.
cycleCounter = -1 ! first calc step increments this to cycle = 0
if (inc == 0) then ! >> start of analysis <<
lastIncConverged = .false. ! no Jacobian backup
outdatedByNewInc = .false. ! no aging of state
lastMode = .false. ! pretend last step was collection
calcMode = .false. ! pretend last step was collection
!$OMP CRITICAL (write2out)
write (6,'(a,i6,x,i2)') '<< HYPELA2 >> start of analysis..! ',n(1),nn
call flush(6)
!$OMP END CRITICAL (write2out)
else if (inc - theInc > 1) then ! >> restart of broken analysis <<
lastIncConverged = .false. ! no Jacobian backup
outdatedByNewInc = .false. ! no aging of state
lastMode = .true. ! pretend last step was calculation
calcMode = .true. ! pretend last step was calculation
!$OMP CRITICAL (write2out)
write (6,'(a,i6,x,i2)') '<< HYPELA2 >> restart of analysis..! ',n(1),nn
call flush(6)
!$OMP END CRITICAL (write2out)
else ! >> just the next inc <<
lastIncConverged = .true. ! request Jacobian backup
outdatedByNewInc = .true. ! request aging of state
lastMode = .true. ! assure last step was calculation
calcMode = .true. ! assure last step was calculation
!$OMP CRITICAL (write2out)
write (6,'(a,i6,x,i2)') '<< HYPELA2 >> new increment..! ',n(1),nn
call flush(6)
!$OMP END CRITICAL (write2out)
endif
else if ( timinc < theDelta ) then ! >> cutBack <<
terminallyIll = .false.
cycleCounter = -1 ! first calc step increments this to cycle = 0
calcMode = .true. ! pretend last step was calculation
!$OMP CRITICAL (write2out)
write(6,'(a,i6,x,i2)') '<< HYPELA2 >> cutback detected..! ',n(1),nn
call flush(6)
!$OMP END CRITICAL (write2out)
endif ! convergence treatment end
calcMode(nn,cp_en) = .not. calcMode(nn,cp_en) ! ping pong (calc <--> collect)
if ( calcMode(nn,cp_en) ) then ! now --- CALC ---
if ( lastMode /= calcMode(nn,cp_en) ) then ! first after ping pong
call debug_reset() ! resets debugging
outdatedFFN1 = .false.
cycleCounter = cycleCounter + 1_pInt
endif
if ( outdatedByNewInc ) then
outdatedByNewInc = .false. ! reset flag
computationMode = 1 ! calc and age results
else
computationMode = 2 ! plain calc
endif
else ! now --- COLLECT ---
if ( lastMode /= calcMode(nn,cp_en) .and. &
.not. terminallyIll ) then
call debug_info() ! first after ping pong reports (meaningful) debugging
endif
if ( lastIncConverged ) then
lastIncConverged = .false. ! reset flag
computationMode = 4 ! collect and backup Jacobian after convergence
else
computationMode = 3 ! plain collect
endif
endif
theTime = cptim ! record current starting time
theDelta = timinc ! record current time increment
theInc = inc ! record current increment number
lastMode = calcMode(nn,cp_en) ! record calculationMode
endif
call CPFEM_general(computationMode,ffn,ffn1,t(1),timinc,n(1),nn,stress,ddsdde, pstress, dPdF)
! Mandel: 11, 22, 33, SQRT(2)*12, SQRT(2)*23, SQRT(2)*13
! Marc: 11, 22, 33, 12, 23, 13
! Marc: 11, 22, 33, 12
forall(i=1:ngens) d(1:ngens,i) = invnrmMandel(i)*ddsdde(1:ngens,i)*invnrmMandel(1:ngens)
s(1:ngens) = stress(1:ngens)*invnrmMandel(1:ngens)
if(symmetricSolver) d(1:ngens,1:ngens) = 0.5_pReal*(d(1:ngens,1:ngens)+transpose(d(1:ngens,1:ngens)))
!$ call omp_set_num_threads(defaultNumThreadsInt) ! reset number of threads to stored default value
end subroutine
!********************************************************************
! This routine sets user defined output variables for Marc
!********************************************************************
!
! select a variable contour plotting (user subroutine).
!
!********************************************************************
subroutine plotv(&
v,& ! variable
s,& ! stress array
sp,& ! stresses in preferred direction
etot,& ! total strain (generalized)
eplas,& ! total plastic strain
ecreep,& ! total creep strain
t,& ! current temperature
m,& ! element number
nn,& ! integration point number
layer,& ! layer number
ndi,& ! number of direct stress components
nshear,& ! number of shear stress components
jpltcd & ! user variable index
)
use prec, only: pReal,pInt
use mesh, only: mesh_FEasCP
use IO, only: IO_error
use homogenization, only: materialpoint_results,materialpoint_sizeResults
implicit none
real(pReal) s(*),etot(*),eplas(*),ecreep(*),sp(*)
real(pReal) v, t(*)
integer(pInt) m, nn, layer, ndi, nshear, jpltcd
if (jpltcd > materialpoint_sizeResults) call IO_error(667,jpltcd) ! complain about out of bounds error
v = materialpoint_results(jpltcd,nn,mesh_FEasCP('elem', m))
return
end subroutine