incEquilibrium computes f^eq_inc – t_i, where t_i are the lattice weights. That is a model for incombressible flows. The units are the some as for the the standard LBM f^eq. Please have a look in the book of Krüger. There is a section un unit conversation.
In the line 353 of dynamics.hh, the pressure is calculated as p = rho / invCs2, which is calculated as I expect.
However, in the line 450 of util.h, the lattice pressure is calculated by p = (rho-1) * Cs2.
The difference between rho and rho-1 comes from the difference between pressure and lattice pressure?
Then, is there any reason that the lattice pressure is defined by using the factor of rho-1 instead of rho?
No, the difference come from that in OpenLB the distribution is stored as f_OpenLB = f – f^eq. Therefore, also the collision operation is presented in a different way.
Your equation f_OpenLB = f – f^eq means Cell[iPop] = f_iPop – t_iPop? or it is something else?
In openLB, I have found two expressions to calculate pressure.
1) p = rho / invCs2 which can be reduced to p = rho * Cs2
2) p = (rho-1) * Cs2.
Do you mean 1) is the usual pressure and 2) is the special pressure only for the codes of openLB?
Your equation f_OpenLB = f – f^eq means Cell[iPop] = f_iPop – t_iPop? or it is something else?
A: That is right!
In openLB, I have found two expressions to calculate pressure.
1) p = rho / invCs2 which can be reduced to p = rho * Cs2
2) p = (rho-1) * Cs2.
Do you mean 1) is the usual pressure and 2) is the special pressure only for the codes of openLB?
A: No, that depends on the actall model.
I have found the functions “incBgkCollision” and “constRhoBgkCollision”. I suppose that the former is related to the incEquilibrium [Zou et al., 1995]. Is this correct?
The first is right if I remember well. The latter is just a projection to a mean pressure=0 (rho=1). It is quite natual to do, so I have no idea where it comes form. I think, I have a little not on that in my thesis.