Merge branch 'Marc-Precision-handling' into 'development'

Marc precision handling

See merge request damask/DAMASK!607
This commit is contained in:
Daniel Otto de Mentock 2022-06-27 08:37:41 +00:00
commit de89d0fe38
19 changed files with 108 additions and 131 deletions

View File

@ -108,7 +108,7 @@ file(STRINGS "$ENV{PETSC_DIR}/$ENV{PETSC_ARCH}/lib/petsc/conf/petscvariables" PE
string(REPLACE "PETSC_FC_INCLUDES = " "" PETSC_INCLUDES "${PETSC_INCLUDES}")
message("PETSC_INCLUDES:\n${PETSC_INCLUDES}\n")
set(CMAKE_Fortran_FLAGS_${CMAKE_BUILD_TYPE} "${BUILDCMD_PRE} ${OPENMP_FLAGS} ${STANDARD_CHECK} ${OPTIMIZATION_FLAGS} ${COMPILE_FLAGS} ${PRECISION_FLAGS}")
set(CMAKE_Fortran_FLAGS_${CMAKE_BUILD_TYPE} "${BUILDCMD_PRE} ${OPENMP_FLAGS} ${STANDARD_CHECK} ${OPTIMIZATION_FLAGS} ${COMPILE_FLAGS}")
set(CMAKE_Fortran_LINK_EXECUTABLE "${BUILDCMD_PRE} ${CMAKE_Fortran_COMPILER} ${OPENMP_FLAGS} ${OPTIMIZATION_FLAGS} ${LINKER_FLAGS}")
if(CMAKE_BUILD_TYPE STREQUAL "DEBUG")

View File

@ -135,10 +135,3 @@ set (DEBUG_FLAGS "${DEBUG_FLAGS} -fsanitize=undefined")
# detect undefined behavior
# Additional options
# -fsanitize=address,leak,thread
#------------------------------------------------------------------------------------------------
# precision settings
set (PRECISION_FLAGS "${PRECISION_FLAGS} -fdefault-real-8")
# set precision to 8 bytes for standard real (=8 for pReal). Will set size of double to 16 bytes as long as -fdefault-double-8 is not set
set (PRECISION_FLAGS "${PRECISION_FLAGS} -fdefault-double-8")
# set precision to 8 bytes for double real, would be 16 bytes if -fdefault-real-8 is used

View File

@ -118,8 +118,3 @@ set (DEBUG_FLAGS "${DEBUG_FLAGS} -debug all")
# -check: Checks at runtime, where
# arg_temp_created: will cause a lot of warnings because we create a bunch of temporary arrays (performance?)
# stack:
#------------------------------------------------------------------------------------------------
# precision settings
set (PRECISION_FLAGS "${PRECISION_FLAGS} -real-size 64")
# set precision for standard real to 32 | 64 | 128 (= 4 | 8 | 16 bytes, type pReal is always 8 bytes)

View File

@ -117,8 +117,3 @@ set (DEBUG_FLAGS "${DEBUG_FLAGS} -debug all")
# -check: Checks at runtime, where
# arg_temp_created: will cause a lot of warnings because we create a bunch of temporary arrays (performance?)
# stack:
#------------------------------------------------------------------------------------------------
# precision settings
set (PRECISION_FLAGS "${PRECISION_FLAGS} -real-size 64")
# set precision for standard real to 32 | 64 | 128 (= 4 | 8 | 16 bytes, type pReal is always 8 bytes)

View File

@ -211,7 +211,7 @@ subroutine hypela2(d,g,e,de,s,t,dt,ngens,m,nn,kcus,matus,ndi,nshear,disp, &
use OMP_LIB
implicit none
integer, intent(in) :: & ! according to MSC.Marc 2012 Manual D
integer(pI64), intent(in) :: & ! according to MSC.Marc 2012 Manual D
ngens, & !< size of stress-strain law
nn, & !< integration point number
ndi, & !< number of direct components
@ -224,7 +224,7 @@ subroutine hypela2(d,g,e,de,s,t,dt,ngens,m,nn,kcus,matus,ndi,nshear,disp, &
jtype, & !< element type
ifr, & !< set to 1 if R has been calculated
ifu !< set to 1 if stretch has been calculated
integer, dimension(2), intent(in) :: & ! according to MSC.Marc 2012 Manual D
integer(pI64), dimension(2), intent(in) :: & ! according to MSC.Marc 2012 Manual D
m, & !< (1) user element number, (2) internal element number
matus, & !< (1) user material identification number, (2) internal material identification number
kcus, & !< (1) layer number, (2) internal layer number
@ -362,7 +362,7 @@ subroutine hypela2(d,g,e,de,s,t,dt,ngens,m,nn,kcus,matus,ndi,nshear,disp, &
endif
lastLovl = lovl
call materialpoint_general(computationMode,ffn,ffn1,t(1),timinc,m(1),nn,stress,ddsdde)
call materialpoint_general(computationMode,ffn,ffn1,t(1),timinc,int(m(1)),int(nn),stress,ddsdde)
d = ddsdde(1:ngens,1:ngens)
s = stress(1:ndi+nshear)
@ -383,16 +383,17 @@ subroutine flux(f,ts,n,time)
use discretization_Marc
implicit none
real(pReal), dimension(6), intent(in) :: &
real(pReal), dimension(6), intent(in) :: &
ts
integer, dimension(10), intent(in) :: &
integer(pI64), dimension(10), intent(in) :: &
n
real(pReal), intent(in) :: &
real(pReal), intent(in) :: &
time
real(pReal), dimension(2), intent(out) :: &
real(pReal), dimension(2), intent(out) :: &
f
f(1) = homogenization_f_T(discretization_Marc_FEM2DAMASK_cell(n(3),n(1)))
f(1) = homogenization_f_T(discretization_Marc_FEM2DAMASK_cell(int(n(3)),int(n(1))))
f(2) = 0.0_pReal
end subroutine flux
@ -410,7 +411,8 @@ subroutine uedinc(inc,incsub)
use discretization_Marc
implicit none
integer, intent(in) :: inc, incsub
integer(pI64), intent(in) :: inc, incsub
integer :: n, nqncomp, nqdatatype
integer, save :: inc_written
real(pReal), allocatable, dimension(:,:) :: d_n
@ -427,9 +429,9 @@ subroutine uedinc(inc,incsub)
enddo
call discretization_Marc_UpdateNodeAndIpCoords(d_n)
call materialpoint_results(inc,cptim)
call materialpoint_results(int(inc),cptim)
inc_written = inc
inc_written = int(inc)
endif
end subroutine uedinc

View File

@ -272,7 +272,7 @@ program DAMASK_grid
write(IO_STDOUT,'(2x,a,/,3(3(3x,f12.7,1x)/))',advance='no') 'R:',&
transpose(loadCases(l)%rot%asMatrix())
if (loadCases(l)%r <= 0.0) errorID = 833
if (loadCases(l)%r <= 0.0_pReal) errorID = 833
if (loadCases(l)%t < 0.0_pReal) errorID = 834
if (loadCases(l)%N < 1) errorID = 835
if (loadCases(l)%f_out < 1) errorID = 836
@ -501,7 +501,7 @@ subroutine getMaskedTensor(values,mask,tensor)
integer :: i,j
values = 0.0
values = 0.0_pReal
do i = 1,3
row => tensor%get(i)
do j = 1,3

View File

@ -151,7 +151,7 @@ subroutine VTI_readCellsSizeOrigin(cells,geomSize,origin, &
character(len=*), intent(in) :: &
fileContent
character(len=:), allocatable :: dataType, headerType
character(len=:), allocatable :: headerType
logical :: inFile, inImage, compressed
integer(pI64) :: &
startPos, endPos

View File

@ -229,9 +229,9 @@ pure function cellSurfaceArea(geomSize,cells)
real(pReal), dimension(6,1,product(cells)) :: cellSurfaceArea
cellSurfaceArea(1:2,1,:) = geomSize(2)/real(cells(2)) * geomSize(3)/real(cells(3))
cellSurfaceArea(3:4,1,:) = geomSize(3)/real(cells(3)) * geomSize(1)/real(cells(1))
cellSurfaceArea(5:6,1,:) = geomSize(1)/real(cells(1)) * geomSize(2)/real(cells(2))
cellSurfaceArea(1:2,1,:) = geomSize(2)/real(cells(2),pReal) * geomSize(3)/real(cells(3),pReal)
cellSurfaceArea(3:4,1,:) = geomSize(3)/real(cells(3),pReal) * geomSize(1)/real(cells(1),pReal)
cellSurfaceArea(5:6,1,:) = geomSize(1)/real(cells(1),pReal) * geomSize(2)/real(cells(2),pReal)
end function cellSurfaceArea

View File

@ -221,14 +221,14 @@ subroutine grid_mechanical_FEM_init
delta = geomSize/real(cells,pReal) ! grid spacing
detJ = product(delta) ! cell volume
BMat = reshape(real([-1.0_pReal/delta(1),-1.0_pReal/delta(2),-1.0_pReal/delta(3), &
1.0_pReal/delta(1),-1.0_pReal/delta(2),-1.0_pReal/delta(3), &
-1.0_pReal/delta(1), 1.0_pReal/delta(2),-1.0_pReal/delta(3), &
1.0_pReal/delta(1), 1.0_pReal/delta(2),-1.0_pReal/delta(3), &
-1.0_pReal/delta(1),-1.0_pReal/delta(2), 1.0_pReal/delta(3), &
1.0_pReal/delta(1),-1.0_pReal/delta(2), 1.0_pReal/delta(3), &
-1.0_pReal/delta(1), 1.0_pReal/delta(2), 1.0_pReal/delta(3), &
1.0_pReal/delta(1), 1.0_pReal/delta(2), 1.0_pReal/delta(3)],pReal), [3,8])/4.0_pReal ! shape function derivative matrix
BMat = reshape(real([-delta(1)**(-1),-delta(2)**(-1),-delta(3)**(-1), &
delta(1)**(-1),-delta(2)**(-1),-delta(3)**(-1), &
-delta(1)**(-1), delta(2)**(-1),-delta(3)**(-1), &
delta(1)**(-1), delta(2)**(-1),-delta(3)**(-1), &
-delta(1)**(-1),-delta(2)**(-1), delta(3)**(-1), &
delta(1)**(-1),-delta(2)**(-1), delta(3)**(-1), &
-delta(1)**(-1), delta(2)**(-1), delta(3)**(-1), &
delta(1)**(-1), delta(2)**(-1), delta(3)**(-1)],pReal), [3,8])/4.0_pReal ! shape function derivative matrix
HGMat = matmul(transpose(HGcomp),HGcomp) &
* HGCoeff*(delta(1)*delta(2) + delta(2)*delta(3) + delta(3)*delta(1))/16.0_pReal ! hourglass stabilization matrix
@ -656,7 +656,7 @@ subroutine formJacobian(da_local,x_local,Jac_pre,Jac,dummy,err_PETSc)
MatNullSpace :: matnull
PetscErrorCode :: err_PETSc
BMatFull = 0.0
BMatFull = 0.0_pReal
BMatFull(1:3,1 :8 ) = BMat
BMatFull(4:6,9 :16) = BMat
BMatFull(7:9,17:24) = BMat
@ -686,7 +686,7 @@ subroutine formJacobian(da_local,x_local,Jac_pre,Jac,dummy,err_PETSc)
enddo; enddo; enddo
row = col
ce = ce + 1
K_ele = 0.0
K_ele = 0.0_pReal
K_ele(1 :8 ,1 :8 ) = HGMat*(homogenization_dPdF(1,1,1,1,ce) + &
homogenization_dPdF(2,2,2,2,ce) + &
homogenization_dPdF(3,3,3,3,ce))/3.0_pReal

View File

@ -198,7 +198,7 @@ subroutine spectral_utilities_init()
CHKERRQ(err_PETSc)
cells1Red = cells(1)/2 + 1
wgt = 1.0/real(product(cells),pReal)
wgt = real(product(cells),pReal)**(-1)
num%memory_efficient = num_grid%get_asInt('memory_efficient', defaultVal=1) > 0 ! ToDo: should be logical in YAML file
num%divergence_correction = num_grid%get_asInt('divergence_correction', defaultVal=2)
@ -265,7 +265,7 @@ subroutine spectral_utilities_init()
N = fftw_mpi_local_size_many(3,[cellsFFTW(3),cellsFFTW(2),cellsFFTW(1)/2_C_INTPTR_T+1_C_INTPTR_T],&
tensorSize,FFTW_MPI_DEFAULT_BLOCK,PETSC_COMM_WORLD,z,devNull)
if (z /= cells3) error stop 'domain decomposition mismatch (tensor)'
if (int(z) /= cells3) error stop 'domain decomposition mismatch (tensor)'
tensorField = fftw_alloc_complex(N)
call c_f_pointer(tensorField,tensorField_real, &
[3_C_INTPTR_T,3_C_INTPTR_T,2_C_INTPTR_T*(cellsFFTW(1)/2_C_INTPTR_T+1_C_INTPTR_T),cellsFFTW(2),z])
@ -274,7 +274,7 @@ subroutine spectral_utilities_init()
N = fftw_mpi_local_size_many(3,[cellsFFTW(3),cellsFFTW(2),cellsFFTW(1)/2_C_INTPTR_T+1_C_INTPTR_T],&
vectorSize,FFTW_MPI_DEFAULT_BLOCK,PETSC_COMM_WORLD,z,devNull)
if (z /= cells3) error stop 'domain decomposition mismatch (vector)'
if (int(z) /= cells3) error stop 'domain decomposition mismatch (vector)'
vectorField = fftw_alloc_complex(N)
call c_f_pointer(vectorField,vectorField_real, &
[3_C_INTPTR_T,2_C_INTPTR_T*(cellsFFTW(1)/2_C_INTPTR_T+1_C_INTPTR_T),cellsFFTW(2),z])
@ -283,7 +283,7 @@ subroutine spectral_utilities_init()
N = fftw_mpi_local_size_3d(cellsFFTW(3),cellsFFTW(2),cellsFFTW(1)/2_C_INTPTR_T+1_C_INTPTR_T,&
PETSC_COMM_WORLD,z,devNull)
if (z /= cells3) error stop 'domain decomposition mismatch (scalar)'
if (int(z) /= cells3) error stop 'domain decomposition mismatch (scalar)'
scalarField = fftw_alloc_complex(N)
call c_f_pointer(scalarField,scalarField_real, &
[2_C_INTPTR_T*(cellsFFTW(1)/2_C_INTPTR_T+1_C_INTPTR_T),cellsFFTW(2),z])
@ -385,17 +385,17 @@ subroutine utilities_updateGamma(C)
xiDyad_cmplx(l,m) = conjg(-xi1st(l,i,j,k-cells3Offset))*xi1st(m,i,j,k-cells3Offset)
end do
do concurrent(l = 1:3, m = 1:3)
temp33_cmplx(l,m) = sum(cmplx(C_ref(l,1:3,m,1:3),0.0_pReal)*xiDyad_cmplx)
temp33_cmplx(l,m) = sum(cmplx(C_ref(l,1:3,m,1:3),0.0_pReal,pReal)*xiDyad_cmplx)
end do
#else
forall(l = 1:3, m = 1:3) &
xiDyad_cmplx(l,m) = conjg(-xi1st(l,i,j,k-cells3Offset))*xi1st(m,i,j,k-cells3Offset)
forall(l = 1:3, m = 1:3) &
temp33_cmplx(l,m) = sum(cmplx(C_ref(l,1:3,m,1:3),0.0_pReal)*xiDyad_cmplx)
temp33_cmplx(l,m) = sum(cmplx(C_ref(l,1:3,m,1:3),0.0_pReal,pReal)*xiDyad_cmplx)
#endif
A(1:3,1:3) = temp33_cmplx%re; A(4:6,4:6) = temp33_cmplx%re
A(1:3,4:6) = temp33_cmplx%im; A(4:6,1:3) = -temp33_cmplx%im
if (abs(math_det33(A(1:3,1:3))) > 1e-16) then
if (abs(math_det33(A(1:3,1:3))) > 1.e-16_pReal) then
call math_invert(A_inv, err, A)
temp33_cmplx = cmplx(A_inv(1:3,1:3),A_inv(1:3,4:6),pReal)
#ifndef __INTEL_COMPILER
@ -518,17 +518,17 @@ subroutine utilities_fourierGammaConvolution(fieldAim)
xiDyad_cmplx(l,m) = conjg(-xi1st(l,i,j,k))*xi1st(m,i,j,k)
end do
do concurrent(l = 1:3, m = 1:3)
temp33_cmplx(l,m) = sum(cmplx(C_ref(l,1:3,m,1:3),0.0_pReal)*xiDyad_cmplx)
temp33_cmplx(l,m) = sum(cmplx(C_ref(l,1:3,m,1:3),0.0_pReal,pReal)*xiDyad_cmplx)
end do
#else
forall(l = 1:3, m = 1:3) &
xiDyad_cmplx(l,m) = conjg(-xi1st(l,i,j,k))*xi1st(m,i,j,k)
forall(l = 1:3, m = 1:3) &
temp33_cmplx(l,m) = sum(cmplx(C_ref(l,1:3,m,1:3),0.0_pReal)*xiDyad_cmplx)
temp33_cmplx(l,m) = sum(cmplx(C_ref(l,1:3,m,1:3),0.0_pReal,pReal)*xiDyad_cmplx)
#endif
A(1:3,1:3) = temp33_cmplx%re; A(4:6,4:6) = temp33_cmplx%re
A(1:3,4:6) = temp33_cmplx%im; A(4:6,1:3) = -temp33_cmplx%im
if (abs(math_det33(A(1:3,1:3))) > 1e-16) then
if (abs(math_det33(A(1:3,1:3))) > 1.e-16_pReal) then
call math_invert(A_inv, err, A)
temp33_cmplx = cmplx(A_inv(1:3,1:3),A_inv(1:3,4:6),pReal)
#ifndef __INTEL_COMPILER
@ -587,8 +587,8 @@ subroutine utilities_fourierGreenConvolution(D_ref, mu_ref, Delta_t)
!$OMP PARALLEL DO PRIVATE(GreenOp_hat)
do k = 1, cells3; do j = 1, cells(2) ;do i = 1, cells1Red
GreenOp_hat = cmplx(1.0_pReal,0.0_pReal,pReal) &
/ (cmplx(mu_ref,0.0_pReal,pReal) + cmplx(Delta_t,0.0_pReal) &
* sum(conjg(xi1st(1:3,i,j,k))* matmul(cmplx(D_ref,0.0_pReal),xi1st(1:3,i,j,k))))
/ (cmplx(mu_ref,0.0_pReal,pReal) + cmplx(Delta_t,0.0_pReal,pReal) &
* sum(conjg(xi1st(1:3,i,j,k))* matmul(cmplx(D_ref,0.0_pReal,pReal),xi1st(1:3,i,j,k))))
scalarField_fourier(i,j,k) = scalarField_fourier(i,j,k)*GreenOp_hat
enddo; enddo; enddo
!$OMP END PARALLEL DO
@ -608,7 +608,7 @@ real(pReal) function utilities_divergenceRMS()
print'(/,1x,a)', '... calculating divergence ................................................'
flush(IO_STDOUT)
rescaledGeom = cmplx(geomSize/scaledGeomSize,0.0_pReal)
rescaledGeom = cmplx(geomSize/scaledGeomSize,0.0_pReal,pReal)
!--------------------------------------------------------------------------------------------------
! calculating RMS divergence criterion in Fourier space
@ -652,7 +652,7 @@ real(pReal) function utilities_curlRMS()
print'(/,1x,a)', '... calculating curl ......................................................'
flush(IO_STDOUT)
rescaledGeom = cmplx(geomSize/scaledGeomSize,0.0_pReal)
rescaledGeom = cmplx(geomSize/scaledGeomSize,0.0_pReal,pReal)
!--------------------------------------------------------------------------------------------------
! calculating max curl criterion in Fourier space

View File

@ -601,9 +601,9 @@ module function RGC_updateState(P,F,avgF,dt,dPdF,ce) result(doneAndHappy)
! calculate the stress and penalty due to volume discrepancy
vPen = 0.0_pReal
do i = 1,nGrain
vPen(:,:,i) = -1.0_pReal/real(nGrain,pReal)*num%volDiscrMod*num%volDiscrPow/num%maxVolDiscr* &
sign((abs(vDiscrep)/num%maxVolDiscr)**(num%volDiscrPow - 1.0),vDiscrep)* &
gVol(i)*transpose(math_inv33(fDef(:,:,i)))
vPen(:,:,i) = -real(nGrain,pReal)**(-1)*num%volDiscrMod*num%volDiscrPow/num%maxVolDiscr &
* sign((abs(vDiscrep)/num%maxVolDiscr)**(num%volDiscrPow - 1.0_pReal),vDiscrep) &
* gVol(i)*transpose(math_inv33(fDef(:,:,i)))
end do
end subroutine volumePenalty

View File

@ -2008,7 +2008,7 @@ subroutine buildTransformationSystem(Q,S,Ntrans,cOverA,a_cF,a_cI)
],pReal),shape(CFTOHP_SYSTEMTRANS))
real(pReal), dimension(4,cF_Ntrans), parameter :: &
CFTOCI_SYSTEMTRANS = reshape([&
CFTOCI_SYSTEMTRANS = real(reshape([&
0.0, 1.0, 0.0, 10.26, & ! Pitsch OR (Ma & Hartmaier 2014, Table 3)
0.0,-1.0, 0.0, 10.26, &
0.0, 0.0, 1.0, 10.26, &
@ -2021,7 +2021,7 @@ subroutine buildTransformationSystem(Q,S,Ntrans,cOverA,a_cF,a_cI)
-1.0, 0.0, 0.0, 10.26, &
0.0, 1.0, 0.0, 10.26, &
0.0,-1.0, 0.0, 10.26 &
],shape(CFTOCI_SYSTEMTRANS))
],shape(CFTOCI_SYSTEMTRANS)),pReal)
integer, dimension(9,cF_Ntrans), parameter :: &
CFTOCI_BAINVARIANT = reshape( [&
@ -2040,7 +2040,7 @@ subroutine buildTransformationSystem(Q,S,Ntrans,cOverA,a_cF,a_cI)
],shape(CFTOCI_BAINVARIANT))
real(pReal), dimension(4,cF_Ntrans), parameter :: &
CFTOCI_BAINROT = reshape([&
CFTOCI_BAINROT = real(reshape([&
1.0, 0.0, 0.0, 45.0, & ! Rotate cF austensite to bain variant
1.0, 0.0, 0.0, 45.0, &
1.0, 0.0, 0.0, 45.0, &
@ -2053,7 +2053,7 @@ subroutine buildTransformationSystem(Q,S,Ntrans,cOverA,a_cF,a_cI)
0.0, 0.0, 1.0, 45.0, &
0.0, 0.0, 1.0, 45.0, &
0.0, 0.0, 1.0, 45.0 &
],shape(CFTOCI_BAINROT))
],shape(CFTOCI_BAINROT)),pReal)
if (present(a_cI) .and. present(a_cF)) then
do i = 1,sum(Ntrans)

View File

@ -1110,7 +1110,7 @@ pure function math_rotationalPart(F) result(R)
C = matmul(transpose(F),F)
I_C = math_invariantsSym33(C)
I_F = [math_trace33(F), 0.5*(math_trace33(F)**2 - math_trace33(matmul(F,F)))]
I_F = [math_trace33(F), 0.5_pReal*(math_trace33(F)**2 - math_trace33(matmul(F,F)))]
x = math_clip(I_C(1)**2 -3.0_pReal*I_C(2),0.0_pReal)**(3.0_pReal/2.0_pReal)
if (dNeq0(x)) then
@ -1386,7 +1386,7 @@ subroutine selfTest()
call random_number(v3_3)
call random_number(v3_4)
if (dNeq(abs(dot_product(math_cross(v3_1-v3_4,v3_2-v3_4),v3_3-v3_4))/6.0, &
if (dNeq(abs(dot_product(math_cross(v3_1-v3_4,v3_2-v3_4),v3_3-v3_4))/6.0_pReal, &
math_volTetrahedron(v3_1,v3_2,v3_3,v3_4),tol=1.0e-12_pReal)) &
error stop 'math_volTetrahedron'

View File

@ -131,7 +131,7 @@ subroutine FEM_utilities_init
call PetscOptionsInsertString(PETSC_NULL_OPTIONS,trim(petsc_optionsOrder),err_PETSc)
CHKERRQ(err_PETSc)
wgt = 1.0/real(mesh_maxNips*mesh_NcpElemsGlobal,pReal)
wgt = real(mesh_maxNips*mesh_NcpElemsGlobal,pReal)**(-1)
end subroutine FEM_utilities_init

View File

@ -247,7 +247,7 @@ subroutine mesh_FEM_build_ipCoordinates(dimPlex,qPoints)
mesh_ipCoordinates(dirI,qPt,cell+1) = pV0(dirI)
do dirJ = 1, dimPlex
mesh_ipCoordinates(dirI,qPt,cell+1) = mesh_ipCoordinates(dirI,qPt,cell+1) + &
pCellJ((dirI-1)*dimPlex+dirJ)*(qPoints(qOffset+dirJ) + 1.0)
pCellJ((dirI-1)*dimPlex+dirJ)*(qPoints(qOffset+dirJ) + 1.0_pReal)
enddo
enddo
qOffset = qOffset + dimPlex

View File

@ -230,14 +230,14 @@ subroutine FEM_mechanical_init(fieldBC)
CHKERRQ(err_PETSc)
call SNESSetConvergenceTest(mechanical_snes,FEM_mechanical_converged,PETSC_NULL_VEC,PETSC_NULL_FUNCTION,err_PETSc)
CHKERRQ(err_PETSc)
call SNESSetTolerances(mechanical_snes,1.0,0.0,0.0,num%itmax,num%itmax,err_PETSc)
call SNESSetTolerances(mechanical_snes,1.0_pReal,0.0_pReal,0.0_pReal,num%itmax,num%itmax,err_PETSc)
CHKERRQ(err_PETSc)
call SNESSetFromOptions(mechanical_snes,err_PETSc); CHKERRQ(err_PETSc)
!--------------------------------------------------------------------------------------------------
! init fields
call VecSet(solution ,0.0,err_PETSc); CHKERRQ(err_PETSc)
call VecSet(solution_rate ,0.0,err_PETSc); CHKERRQ(err_PETSc)
call VecSet(solution ,0.0_pReal,err_PETSc); CHKERRQ(err_PETSc)
call VecSet(solution_rate,0.0_pReal,err_PETSc); CHKERRQ(err_PETSc)
allocate(x_scal(cellDof))
allocate(nodalWeightsP(1))
allocate(nodalPointsP(dimPlex))
@ -338,11 +338,10 @@ subroutine FEM_mechanical_formResidual(dm_local,xx_local,f_local,dummy,err_PETSc
PetscInt :: cellStart, cellEnd, cell, field, face, &
qPt, basis, comp, cidx, &
numFields, &
bcSize,m
bcSize,m,i
PetscReal :: detFAvg, detJ
PetscReal, dimension(dimPlex*dimPlex,cellDof) :: BMat
IS :: bcPoints
IS :: bcPoints
allocate(pV0(dimPlex))
@ -358,7 +357,7 @@ subroutine FEM_mechanical_formResidual(dm_local,xx_local,f_local,dummy,err_PETSc
CHKERRQ(err_PETSc)
call DMGetLocalVector(dm_local,x_local,err_PETSc)
CHKERRQ(err_PETSc)
call VecWAXPY(x_local,1.0,xx_local,solution_local,err_PETSc)
call VecWAXPY(x_local,1.0_pReal,xx_local,solution_local,err_PETSc)
CHKERRQ(err_PETSc)
do field = 1, dimPlex; do face = 1, mesh_Nboundaries
if (params%fieldBC%componentBC(field)%Mask(face)) then
@ -386,23 +385,23 @@ subroutine FEM_mechanical_formResidual(dm_local,xx_local,f_local,dummy,err_PETSc
IcellJMat = reshape(pInvcellJ,shape=[dimPlex,dimPlex])
do qPt = 0, nQuadrature-1
m = cell*nQuadrature + qPt+1
BMat = 0.0
BMat = 0.0_pReal
do basis = 0, nBasis-1
do comp = 0, dimPlex-1
cidx = basis*dimPlex+comp
BMat(comp*dimPlex+1:(comp+1)*dimPlex,basis*dimPlex+comp+1) = &
matmul(IcellJMat,basisFieldDer((((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp )*dimPlex+1: &
(((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp+1)*dimPlex))
i = ((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp
BMat(comp*dimPlex+1_pPETSCINT:(comp+1_pPETSCINT)*dimPlex,basis*dimPlex+comp+1_pPETSCINT) = &
matmul(IcellJMat,basisFieldDer(i*dimPlex+1_pPETSCINT:(i+1_pPETSCINT)*dimPlex))
enddo
enddo
homogenization_F(1:dimPlex,1:dimPlex,m) = reshape(matmul(BMat,x_scal),shape=[dimPlex,dimPlex], order=[2,1])
enddo
if (num%BBarStabilisation) then
detFAvg = math_det33(sum(homogenization_F(1:3,1:3,cell*nQuadrature+1:(cell+1)*nQuadrature),dim=3)/real(nQuadrature))
detFAvg = math_det33(sum(homogenization_F(1:3,1:3,cell*nQuadrature+1:(cell+1)*nQuadrature),dim=3)/real(nQuadrature,pReal))
do qPt = 0, nQuadrature-1
m = cell*nQuadrature + qPt+1
homogenization_F(1:dimPlex,1:dimPlex,m) = homogenization_F(1:dimPlex,1:dimPlex,m) &
* (detFAvg/math_det33(homogenization_F(1:3,1:3,m)))**(1.0/real(dimPlex))
* (detFAvg/math_det33(homogenization_F(1:3,1:3,m)))**(1.0_pReal/real(dimPlex,pReal))
enddo
endif
@ -425,22 +424,22 @@ subroutine FEM_mechanical_formResidual(dm_local,xx_local,f_local,dummy,err_PETSc
call DMPlexComputeCellGeometryAffineFEM(dm_local,cell,pV0,pCellJ,pInvcellJ,detJ,err_PETSc)
CHKERRQ(err_PETSc)
IcellJMat = reshape(pInvcellJ,shape=[dimPlex,dimPlex])
f_scal = 0.0
f_scal = 0.0_pReal
do qPt = 0, nQuadrature-1
m = cell*nQuadrature + qPt+1
BMat = 0.0
BMat = 0.0_pReal
do basis = 0, nBasis-1
do comp = 0, dimPlex-1
cidx = basis*dimPlex+comp
BMat(comp*dimPlex+1:(comp+1)*dimPlex,basis*dimPlex+comp+1) = &
matmul(IcellJMat,basisFieldDer((((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp )*dimPlex+1: &
(((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp+1)*dimPlex))
i = ((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp
BMat(comp*dimPlex+1_pPETSCINT:(comp+1_pPETSCINT)*dimPlex,basis*dimPlex+comp+1_pPETSCINT) = &
matmul(IcellJMat,basisFieldDer(i*dimPlex+1_pPETSCINT:(i+1_pPETSCINT)*dimPlex))
enddo
enddo
f_scal = f_scal + &
matmul(transpose(BMat), &
reshape(transpose(homogenization_P(1:dimPlex,1:dimPlex,m)), &
shape=[dimPlex*dimPlex]))*qWeights(qPt+1)
f_scal = f_scal &
+ matmul(transpose(BMat), &
reshape(transpose(homogenization_P(1:dimPlex,1:dimPlex,m)), &
shape=[dimPlex*dimPlex]))*qWeights(qPt+1_pPETSCINT)
enddo
f_scal = f_scal*abs(detJ)
pf_scal => f_scal
@ -465,28 +464,25 @@ subroutine FEM_mechanical_formJacobian(dm_local,xx_local,Jac_pre,Jac,dummy,err_P
PetscObject, intent(in) :: dummy
PetscErrorCode :: err_PETSc
PetscDS :: prob
Vec :: x_local, xx_local
PetscSection :: section, gSection
PetscDS :: prob
Vec :: x_local, xx_local
PetscSection :: section, gSection
PetscReal, dimension(1, cellDof) :: MatB
PetscReal, dimension(dimPlex**2,cellDof) :: BMat, BMatAvg, MatA
PetscReal, dimension(3,3) :: F, FAvg, FInv
PetscReal :: detJ
PetscReal, dimension(:), pointer :: basisField, basisFieldDer, &
PetscReal, dimension(3,3) :: F, FAvg, FInv
PetscReal :: detJ
PetscReal, dimension(:), pointer :: basisField, basisFieldDer, &
pV0, pCellJ, pInvcellJ
PetscScalar, dimension(:), pointer :: pK_e, x_scal
PetscScalar,dimension(cellDOF,cellDOF), target :: K_e
PetscScalar,dimension(cellDOF,cellDOF) :: K_eA , &
K_eB
PetscScalar,dimension(cellDOF,cellDOF) :: K_eA, K_eB
PetscInt :: cellStart, cellEnd, cell, field, face, &
qPt, basis, comp, cidx,bcSize, m
IS :: bcPoints
PetscInt :: cellStart, cellEnd, cell, field, face, &
qPt, basis, comp, cidx,bcSize, m, i
IS :: bcPoints
allocate(pV0(dimPlex))
@ -530,30 +526,29 @@ subroutine FEM_mechanical_formJacobian(dm_local,xx_local,Jac_pre,Jac,dummy,err_P
CHKERRQ(err_PETSc)
call DMPlexComputeCellGeometryAffineFEM(dm_local,cell,pV0,pCellJ,pInvcellJ,detJ,err_PETSc)
CHKERRQ(err_PETSc)
K_eA = 0.0
K_eB = 0.0
MatB = 0.0
FAvg = 0.0
BMatAvg = 0.0
K_eA = 0.0_pReal
K_eB = 0.0_pReal
MatB = 0.0_pReal
FAvg = 0.0_pReal
BMatAvg = 0.0_pReal
do qPt = 0, nQuadrature-1
m = cell*nQuadrature + qPt + 1
BMat = 0.0
BMat = 0.0_pReal
do basis = 0, nBasis-1
do comp = 0, dimPlex-1
cidx = basis*dimPlex+comp
BMat(comp*dimPlex+1:(comp+1)*dimPlex,basis*dimPlex+comp+1) = &
matmul( reshape(pInvcellJ, shape = [dimPlex,dimPlex]),&
basisFieldDer((((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp )*dimPlex+1: &
(((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp+1)*dimPlex))
i = ((qPt*nBasis + basis)*dimPlex + comp)*dimPlex+comp
BMat(comp*dimPlex+1_pPETSCINT:(comp+1_pPETSCINT)*dimPlex,basis*dimPlex+comp+1_pPETSCINT) = &
matmul(reshape(pInvcellJ,[dimPlex,dimPlex]),basisFieldDer(i*dimPlex+1_pPETSCINT:(i+1_pPETSCINT)*dimPlex))
enddo
enddo
MatA = matmul(reshape(reshape(homogenization_dPdF(1:dimPlex,1:dimPlex,1:dimPlex,1:dimPlex,m), &
shape=[dimPlex,dimPlex,dimPlex,dimPlex], order=[2,1,4,3]), &
shape=[dimPlex*dimPlex,dimPlex*dimPlex]),BMat)*qWeights(qPt+1)
shape=[dimPlex*dimPlex,dimPlex*dimPlex]),BMat)*qWeights(qPt+1_pPETSCINT)
if (num%BBarStabilisation) then
F(1:dimPlex,1:dimPlex) = reshape(matmul(BMat,x_scal),shape=[dimPlex,dimPlex])
FInv = math_inv33(F)
K_eA = K_eA + matmul(transpose(BMat),MatA)*math_det33(FInv)**(1.0/real(dimPlex))
K_eA = K_eA + matmul(transpose(BMat),MatA)*math_det33(FInv)**(1.0_pReal/real(dimPlex,pReal))
K_eB = K_eB - &
matmul(transpose(matmul(reshape(homogenization_F(1:dimPlex,1:dimPlex,m),shape=[dimPlex**2,1_pPETSCINT]), &
matmul(reshape(FInv(1:dimPlex,1:dimPlex), &
@ -568,10 +563,10 @@ subroutine FEM_mechanical_formJacobian(dm_local,xx_local,Jac_pre,Jac,dummy,err_P
enddo
if (num%BBarStabilisation) then
FInv = math_inv33(FAvg)
K_e = K_eA*math_det33(FAvg/real(nQuadrature))**(1.0/real(dimPlex)) + &
K_e = K_eA*math_det33(FAvg/real(nQuadrature,pReal))**(1.0_pReal/real(dimPlex,pReal)) + &
(matmul(matmul(transpose(BMatAvg), &
reshape(FInv(1:dimPlex,1:dimPlex),shape=[dimPlex**2,1_pPETSCINT],order=[2,1])),MatB) + &
K_eB)/real(dimPlex)
K_eB)/real(dimPlex,pReal)
else
K_e = K_eA
endif
@ -641,7 +636,7 @@ subroutine FEM_mechanical_forward(guess,timeinc,timeinc_old,fieldBC)
CHKERRQ(err_PETSc)
call DMGlobalToLocalEnd(dm_local,solution,INSERT_VALUES,x_local,err_PETSc)
CHKERRQ(err_PETSc)
call VecAXPY(solution_local,1.0,x_local,err_PETSc); CHKERRQ(err_PETSc)
call VecAXPY(solution_local,1.0_pReal,x_local,err_PETSc); CHKERRQ(err_PETSc)
do field = 1, dimPlex; do face = 1, mesh_Nboundaries
if (fieldBC%componentBC(field)%Mask(face)) then
call DMGetStratumSize(dm_local,'Face Sets',mesh_boundaries(face),bcSize,err_PETSc)
@ -659,7 +654,7 @@ subroutine FEM_mechanical_forward(guess,timeinc,timeinc_old,fieldBC)
!--------------------------------------------------------------------------------------------------
! update rate and forward last inc
call VecCopy(solution,solution_rate,err_PETSc); CHKERRQ(err_PETSc)
call VecScale(solution_rate,1.0/timeinc_old,err_PETSc); CHKERRQ(err_PETSc)
call VecScale(solution_rate,timeinc_old**(-1),err_PETSc); CHKERRQ(err_PETSc)
endif
call VecCopy(solution_rate,solution,err_PETSc); CHKERRQ(err_PETSc)
call VecScale(solution,timeinc,err_PETSc); CHKERRQ(err_PETSc)
@ -685,9 +680,8 @@ subroutine FEM_mechanical_converged(snes_local,PETScIter,xnorm,snorm,fnorm,reaso
call SNESConvergedDefault(snes_local,PETScIter,xnorm,snorm,fnorm/divTol,reason,dummy,err_PETSc)
CHKERRQ(err_PETSc)
if (terminallyIll) reason = SNES_DIVERGED_FUNCTION_DOMAIN
print'(/,1x,a,a,i0,a,i0,f0.3)', trim(incInfo), &
' @ Iteration ',PETScIter,' mechanical residual norm = ', &
int(fnorm/divTol),fnorm/divTol-int(fnorm/divTol) ! ToDo: int casting?
print'(/,1x,a,a,i0,a,f0.3)', trim(incInfo), &
' @ Iteration ',PETScIter,' mechanical residual norm = ',fnorm/divTol
print'(/,1x,a,/,2(3(2x,f12.4,1x)/),3(2x,f12.4,1x))', &
'Piola--Kirchhoff stress / MPa =',transpose(P_av)*1.e-6_pReal
flush(IO_STDOUT)

View File

@ -593,7 +593,7 @@ function integrateStateFPI(F_0,F,subFp0,subFi0,subState0,Delta_t,ph,en) result(b
iteration: do NiterationState = 1, num%nState
dotState_last(1:sizeDotState,2) = merge(dotState_last(1:sizeDotState,1),0.0, nIterationState > 1)
dotState_last(1:sizeDotState,2) = merge(dotState_last(1:sizeDotState,1),0.0_pReal, nIterationState > 1)
dotState_last(1:sizeDotState,1) = dotState
broken = integrateStress(F,subFp0,subFi0,Delta_t,ph,en)
@ -756,7 +756,7 @@ function integrateStateRK4(F_0,F,subFp0,subFi0,subState0,Delta_t,ph,en) result(b
real(pReal), dimension(3), parameter :: &
C = [0.5_pReal, 0.5_pReal, 1.0_pReal]
real(pReal), dimension(4), parameter :: &
B = [1.0_pReal/6.0_pReal, 1.0_pReal/3.0_pReal, 1.0_pReal/3.0_pReal, 1.0_pReal/6.0_pReal]
B = [6.0_pReal, 3.0_pReal, 3.0_pReal, 6.0_pReal]**(-1)
broken = integrateStateRK(F_0,F,subFp0,subFi0,subState0,Delta_t,ph,en,A,B,C)
@ -1263,8 +1263,6 @@ module subroutine mechanical_restartRead(groupHandle,ph)
integer(HID_T), intent(in) :: groupHandle
integer, intent(in) :: ph
integer :: en
call HDF5_read(plasticState(ph)%state0,groupHandle,'omega_plastic')
call HDF5_read(phase_mechanical_S0(ph)%data,groupHandle,'S')

View File

@ -912,7 +912,7 @@ pure subroutine kinetics_tw(Mp,T,abs_dot_gamma_sl,ph,en,&
real(pReal), dimension(param(ph)%sum_N_tw), optional, intent(out) :: &
ddot_gamma_dtau_tw
real :: &
real(pReal) :: &
tau, tau_r, tau_hat, &
dot_N_0, &
x0, V, &
@ -988,7 +988,7 @@ pure subroutine kinetics_tr(Mp,T,abs_dot_gamma_sl,ph,en,&
real(pReal), dimension(param(ph)%sum_N_tr), optional, intent(out) :: &
ddot_gamma_dtau_tr
real :: &
real(pReal) :: &
tau, tau_r, tau_hat, &
dot_N_0, &
x0, V, &

View File

@ -801,7 +801,7 @@ subroutine selfTest()
real(pReal), dimension(3,3) :: om, t33
real(pReal), dimension(3,3,3,3) :: t3333
real(pReal), dimension(6,6) :: C
real :: A,B
real(pReal) :: A,B
integer :: i