CFDEM®coupling- Developer Forum

Topics related to developing with CFDEM®coupling can be discussed here: discussion about implementation details, C++, MPI and debugging tools

[solved] Floating point exception error

Submitted by jtvanlew on

Hello developers,

I'm running into this error periodicly in my coupled simulations. It happens when I have many small particles in my system (curiously, just a few small particles isn't an issue). I've got about 5000 particles of radius Rp and about another 5000 of radius 0.35*Rp. The CFD runs first and its fine. Then the LIGGGHTS half of the coupling fires up and it crashes at the moment of first coupling. I just finished a similar simulation where there were maybe 6000 large particles and 2000 smaller particles -- there was no issue whatsoever.

calculating Biot number in LaEuScalarTemp.c

Submitted by jtvanlew on

Hi all,

I am trying to calculate the Biot number (Bi = h d_p / k_s) just after the heat transfer coefficient is calculated inside LaEuScalarTemp.c. I naively thought I could find the conductivity of the DEM particles with :

ks = particleCloud_.conductivity(index);

but the error comes out that


class Foam::cfdemCloud’ has no member named ‘conductivity’

Is there a way of getting the conductivity into the code so I can finish calculating the Biot number?

momentum equation

Submitted by Riccardo Maione on

Hello all,

I noticed something when I was reading the different solvers of CFDEM package, normaly the momentum equation is:

(
ddt(voidfraction, U)
+ div(phi, U)
+ divDevReff(U)
==
+ voidfraction*grad(p)
+ Source
);

so from what I have read solvers do not have the voidfraction near grad(p), is it true or I did not notice something that would make things clearer?

Thanks as always

compile the new liggghts src files to CFDEM

Submitted by kunzhang on

Hi All,

I have developped some new liggghts code under src directory of liggghts and they could be compiled within liggghts well, as doing "make fedora". Right now, I would like to compile or include these files in the CFDEM and I did "cfdemCompLIG". After that, before the new file, it shows a - ***.o. However, before the existing files, it shows r - ***.o, as shown in the attached file. Could anyone please let me know if it is the correct way to compile the new liggghts file to the CFDEM. I also attached the log_compileliggghts for your reivew.

Thanks!

Kun

LaEuScalarTemp.C

Submitted by kunzhang on

Hi All,

I am reading the file of LaEuScalarTemp.C and would like to study how to calculate the heat convection in the coupling work. I am not sure the following line is correct or not:

Rep = 2*rs*magUr/nuf;
Pr = Cp_*nuf*rho_[cellI]/lambda_;

The CFDEM version I use is CFDEMcoupling-PUBLIC-2.2.x.

I think in the Li and Mason's paper (2000, 2002)
Rep = rho_[cellI]*2*rs*magUr/nuf;
and
Pr =Cp_*nuf/lambda_;

the density items should be with particle Reynolds numbers.

Thanks!

Kun

VOF method with IB method in cfdemSolverIB

Submitted by keepfit on

Dear CFDEM developers,

i am verty interested in the resolved CFD-DEM coupling scheme (IB method). As the cfdemSolverIB solver is for single fluid phase, I wonder is it possible to extend the solver so that we can model multi-phase flow interaction with DEM particles? On the CFD side, IB method is already implemented in openFoam 1.6-ext, and a multi-phase solver interIBFoam is also available. But I guess its implementation is different from the IB method in the CFDEM scheme.

Adding new convective heat transfer correlations

Submitted by jtvanlew on

Please excuse this if the question seems below the developers here, but I'm not exactly pro on software development...

I've attempted to modify the LaEuScalarTemp file with a new correlation for the convective heat transfer coefficient. I saved the old LaEuScalarTemp file and replaced it with my modified version. I then recompiled cfdemCompCFDEM and I'm running my simulation again. So far nothing has crashed, but I'm not sure if the CFDEM code is looking at my new file.