2D Poiseuille flow at the boundary

Lattice Boltzmann Method LBM: Theory
#1

I have implemented the 2D cylinder Poiseuille flow with the Lattice Boltzmann method, and I am using “on-grid bounce back” boundary conditions.

However, my velocity profile is strange. There is a finite velocity at the boundary, but I was under the impression that the whole point of the bounce-back BC was to ensure no-slip. My velocity profile looks like this.

Does this behavior at the boundary imply an error in my implementation?

I’d be happy to post my code, if anyone is interested in taking a look.

Thanks in advance,
Niles.

#2

Dear Niles,

It is a well-known phenomenon that bounceback boundaries don’t have a velocity equal to zero exactly at the boundary. In fact, the velocity will be zero somewhere in between the bounceback node and the adjacent fluid node. See e.g. the book by Wolf-Gladrow about this.

Kind regards,

Elrohir