Body force implementation without lattice effects

Lattice Boltzmann Method LBM: Theory
#21

Ok!
I see…perhaps the problem is more on the side of the BC.

Concerning this issue, I have the following remarks:

  1. Did you correct the Zou/He model to account for the presence of the body force in it?
  2. I would say that even if everything is completely right and consistent in your Zou/He model, at the end you can only guarantee the impermeability condition on the walls up to the same order of the discretization error. That is, I would say that you cannot have u=0 on the wall up to the machine accuracy, even though you have set this condition on the boundary.
  3. I don’t know if I followed your statements regarding (1) the lack of periodicity of the Zou/He formulation and (2) use of body force bc’s. Sorry, but what do you exactly mean with you that?

Regards,

Goncalo

#22

Thanks Goncalo,

  1. I don’t use both Zhou/he and Guo in the same model, I use them in separate models simply to create a pressure drop across the medium to compare the BC’s in this application
  2. Thats an interesting point which I had not considered
  3. When using a body force I treat the boundaries as periodic, so the medium is infinitely long in all directions. However with zhou/he boudary the inlet is not connected to the outlet through periodicity so the medium is effectively only infinite in two directions.

For the body force models I recover a permeability tensor which is anisotropic, and with the zhou/he models I recover a roughly isotropic permeability tensor whos magnitude is half the magnitude reported by the body force models… All tests carried out at the same Reynolds number using microCT data.

I am beginning to believe this is a characteristic of the different boundary conditions. But as ever, I am still questioning my implementation.

Regards,
Bruce

#23

Ok Bruce. Now I see your problem. Yet, unfortunately, I have no expertise in those kind of problems. Sorry :frowning:

Nevertheless, I don’t know if you are aware of this article, as well as some other articles from these same authors. They discuss issues similar to those you described before.

http://iopscience.iop.org/1742-5468/2010/11/P11026/pdf/1742-5468_2010_11_P11026.pdf

Good luck.
BTW, in case you fix your problem or identify the cause for it please post here your conclusion. Now I am curious about what it :slight_smile:

Regards,

Goncalo

#24

Thanks again Goncalo,

I was not aware of this paper, it has certainly provided some direction for my investigation. I will report back when I figure this out.

Regards,
Bruce