my name is Simon and I’m currently doing my Phd in LBM-DEM coupling for porous media flow. Therefore I implented a modified collision operator investigated by Noble in 1998, Cook in 2001 and Han in 2017.
This operator uses a bounce back of the non-equilibrium part and it is suggested especially for porous media flow and moving solid boundaries by literature.
The implementation of the operator worked well - the fluid field is adjusted properly based on the sphere radius, but the interfacial moment exchange from fluid on solid does not work well which means: the calculated drag force excerted from the fluid on the solid is not calculated correctly.
The force should be calculated as followed
F_drag = dh^2/dt * sum(B * sum(OMG_coll * ex))
OMG_coll = modified collision operator
B = weightening function
dh = lattice grid spacing
dt = lattice time step, with dt = dh^2
ex = lattice vector for x-direction
My question therefore is: does this way of calculating the drag force directly give the physical value of force in (N)? Because the difference between numerical and analytical solution is in the order of some magnitudes.
If somebody is interested, I have no problem of sharing my code (MATLAB) and literature in order to find the problem.
Thank you very much,