[camino123]

Hi, i just want to know what do you think about LBM updated at 2017.
I mean i've seen that the last post is dated at 2011, but now for example XFlow has been bought by Dassault Systèmes.
I'm wondering if this technology is effectively ready to be used in every time CFD applications and could overcome the classic FVM analysis, or it is not so theoretically correct and it can be used only in certain cases?

Thanks!

[camino123]

no opinions?

[arjun]

Not many people here have direct experience with LBM.

I tried classic LBM and found its usual problems that is when viscosity goes down stability goes down the hill. This is why people are trying to stabilize it with Entropic versions. Details of how big companies like Exa stabilize their solver are well hidden and not accessible to common people like us.

I spent time with Finite Volume versions of LBM and found out that it is too sensitive to dissipation errors and costwise it is not better than Finite Volume SIMPLE type solvers.

In a year or so you might see it implemented in FVUS framework (have done separately need to bring it in FVUS).

The places where I see its use is VOF type solvers where time step size is anyway restricted by courant issues so SIMPLE type methods are very costly there.

Second place where I see LBM's use is acoustic simulations and this is the area where my interest lies wrt LBM.