[makaveli_lcf]

bounding problems for my experience can be only fixed for the skewed meshes can be solved using "upwind" convective scheme along with "cellLimited" gradient schemes. Ones you go for the second order in the convective schemes, you always get some unboundness of the transported scalars.

Cheers,
Alex

[rmz]

Hello Dr. VAKHRUSHEV and thank you for your reply,

for the "gradSchemes" I am using cellMDLimited Gauss linear 1.
Whould you recommend the the "cellLimited" or "cellMDLimited" ?

for the "divSchemes" I am using:
-bounded Gauss upwind for "div(phi, epsilon)" and "div(phi, k)".
-bounded Gauss limitedLinear 1 for "div (phi,U)", because I had convergence problems and this scheme helped fix the convergence.

for "snGradSchemes" I am using limited 0.333

these schemes helped fixing bounded problems, but bounded k and epsilon are still appearing after approx 800 steps.

Which is better to use for the gradSchemes "cellLimited" or "cellMDLimited"?

Thank you

[makaveli_lcf]

as far as I remember cellMDLimited makes sence only for vector fields, and it additionally make some distortions in the solution, so I stick to cellLimited