Make LIGGGHTS taking results from OpenFOAM

Submitted by a.carmona on Tue, 05/14/2013 - 18:52

Hi all,

I've have got some results from an OpenFOAM simulation (interFaom was the solver) and now I want to couple this results to LIGGGHTS, i.e. making LIGGGHTS use this fields (U, p) to calculate DEM.

I read about the coupling method but couldn't find a way to disable OpenFOAM cicle. If I can turn off this calculus i will be able to replace the fields for the ones that i've got from interFoam.

Does anyone know about how i can modify the scheme?

Thanks in advance!

Maryam | Thu, 05/23/2013 - 22:43

This is quite similar to what I intended to do. I wanted to use the pressure and velocity fields from the openFOAM's steady state solver for a one way coupled run (flow field is not affected by the particle flow). In order to speed up the simulation I turned off the coupling by commenting out the “Pressure-velocity PISO corrector”, “PISO loop”, “Non-orthogonal pressure corrector loop”, and Ksl update sections in the cfdem_path/applications/solvers/cfdemSolverPiso.c and recompiling the solver. (rmdepall, wclean, wmake in the same directory).

I also had to disable ddtVoidfraction update in cfdemCould.c as it took too long and was unnecessary (I suppose).

Apparently it works, however I am not sure whether it affects the drag force calculations.

Maryam

cgoniva's picture

cgoniva | Wed, 06/19/2013 - 16:12

Hi,

you might switch of the calculation of the flow as Maryam described this should speed up things a lot (as you do not want to solve for the flow)!

Cheers,
Chris

pipegaldames | Tue, 08/20/2013 - 22:44

Hi, I tried to make this steps as Maryam described for my own problem, the solver is compiled perfectly, and the case run very speedy, but the results are not really good, the behaviour is very weird because the particles is moving in the inlet, but soon they are stopping very close of the insertion face.

Maryam/Chris, how is your comparation of results between your running on this way and the case complete?

Cheers,
Felipe

Maryam | Wed, 08/21/2013 - 20:58

Hi Felipe,
I can't run fully-coupled cases since my particles are too big compared to the CFD cells and fully coupled runs diverge very soon. However, my semi-coupled runs are smooth and reasonable. One reason your particles stop moving could be small drag forces. Double checking the fluid viscosity and the velocity field or testing another drag force model might help you find the cause of this problem.

Maryam

pipegaldames | Thu, 08/22/2013 - 21:52

Hi Maryam, thanks for your answer. I forgot mention that when I run semi-coupled , the wrote files of the fields (U, p , ... ) are changed and erased, for this reason the flow is empty, and DEM coupling with this empty fields and particles are not moving in CFD domain.

Felipe