JGonzalez/fpakc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fPsi no longer allocates memory

Browse source 
I noticed that phy2logquad had a lot of overhead. Trying to reducing it
by simplifying calls to fPsi, dPsi and such.

The function for fPsi has been made so no memory is allocated and works
under the assumption that the input array has the right size (1:numNodes)
This commit is contained in:
Jorge Gonzalez 2023-01-01 12:12:06 +01:00
commit 0db76083ec
8 changed files with 408 additions and 409 deletions
Download patch file Download diff file Expand all files Collapse all files

15
src/modules/mesh/moduleMesh.f90
View file

@ -211,11 +211,11 @@ MODULE moduleMesh
END FUNCTION getNodesVol_interface
PURE FUNCTION fPsi_interface(xi) RESULT(fPsi)
PURE SUBROUTINE fPsi_interface(xi, fPsi)
REAL(8), INTENT(in):: xi(1:3)
REAL(8), ALLOCATABLE:: fPsi(:)
REAL(8), INTENT(out):: fPsi(:)
END FUNCTION fPsi_interface
END SUBROUTINE fPsi_interface
PURE FUNCTION elemK_interface(self) RESULT(localK)
IMPORT:: meshVol
@ -496,11 +496,14 @@ MODULE moduleMesh
INTEGER:: i, nNodes
CLASS(meshNode), POINTER:: node
fPsi = self%fPsi(part%xi)
tensorS = outerProduct(part%v, part%v)
sp = part%species%n
volNodes = self%getNodes()
nNodes = SIZE(volNodes)
ALLOCATE(fPsi(1:nNodes))
CALL self%fPsi(part%xi, fPsi)
tensorS = outerProduct(part%v, part%v)
sp = part%species%n
DO i = 1, nNodes
node => mesh%nodes(volNodes(i))%obj