Velocity in multicomponent flow (shan chen)

[czb]

I have slightly modified example in showCases/multiComponent3d/rayleighTaylor3D.cpp to display velocity field in x, y and z direction by adding:

imageWriter.writeScaledPpm(createFileName("ux", iT, 6), *computeVelocityComponent(heavyFluid, slice, 0));

The resulting velocities seem to be correct only at the boundary between the two fluids. In the body of the fluids the velocities seem to be almost zero.

How can I get correct velocities in the bodies of the fluids?

Motivation: I would like to use non-Newtonian Carreau fluid together with Shan-Chen multicomponent scheme. The Carreau computes computeRhoBarJPiNeq. So I just want to be sure that I get correct PiNeq when using Shan-Chen with the non-Newtonian fluid.

EDIT: Or can you please confirm that Shan-Chen produces correct velocity fields in the fluid and it is just me that understands that wrongly?

[czb]

I have found the following which answers my question:

Examination of the velocity fields in SCMP simulations reveals high velocities to phase boundaries. These are non-physical. They present at equilibrium and there appears to be no mass exchange associated them. Some papers that focus on or mention these velocities Wagner (2003), Lishchuk et al. (2003), Nourgaliev et al. (2003), Raiskinmäki et al. (2000).