Foam-extend 3.2 and CFDEM

Submitted by carboleda on Thu, 05/04/2017 - 14:35

Hi all,

I'm currently trying to set up the CFDEM(R) workbench in a personal computer with Ubuntu 16.04.1 LTS which has foam-extend-3.2 among other version of foam-extend and OpenFOAM(R).

So far, I have managed to compile LIGGGTHS(R) and CFDEM(R), following the instructions provided in:

https://www.cfdem.com/system/files/githubaccess_public.pdf

And juggling a little with the variable WM_MPLIB in the foam/foam-exten-3.2/etc/bashrc file. This variable was set to WM_MPLIB:=MPI, for the compilation of LIGGGTHS(R), and WM_MPLIB:=OPENMPI, for the compilation of CFDEM(R), depending if it worked or not during the compilation (I do not master the arcane of MPI and environment variables).

I had to comment the 13th line in file additionalLibs_3.2, namely:

#include $(CFDEM_ADD_LIBS_DIR)/additionalLibs_vtk

for the compilation of CFDEM(R).

Once the compilation completed (and correct, I guess), I proceeded to run the tutorial under tutorials/cfdemSolverPisoScalar/packedBedTemp where I had to modify the turbulence model file to comply to the format asked by foam-extend-3.2. Depending on the variable WM_MPLIB the outcome of running ./Allmake was different. For WM_MPLIB:=MPI, the log files are attached to this post. LIGGGTHS(R) seems to work while the coupling does not. My guess is that the variable WM_MPLIB should be the same while compiling LIGGGTHS(R) and CFDEM(R).

My questions:
Is it possible to do the coupling with foam-extend-3.2?
If so, is it normal to modify the WM_MPLIB variable and to comment the line in additionalLibs_3.2?

Any other information about the coupling with foam-extend-3.2 would be very helpfull.

Thanks in advance for your help,

Carlos

alice's picture

alice | Tue, 05/09/2017 - 14:07

Hello Carlos,
basically it is possible to use the coupling with foam-extend-3.2, although we did not validate the restults for our solvers. Unfortunately I am not able to open the txt file at the moment, however, some general things: first of all I'd recommend to use the same MPI (SYSTEMMPI) for the compilation of liggghts and cfdemcoupling. The latter uses the first as library... If you change this now, please make sure that also foam-extend has to be re-installed. Then, you seem to have troubles with the usage of vtk. Do you have a version installed, and if yes, which one? It is not absolutely necessary to use vtk (enables easier output for liggghts), but really nice to have. So if possible try to install it on your machine unless it is not there yes.
Cheers,
Alice