[raz]

Hi all,

I am simulating a turbulent channel flow over a backwards-facing step. I have chosen to use RANS + standard k-w turbulence model (with low-Re corrections).

My flow conditions are those from:

Driver, D. & Seegmiller, H., 1985. Features of a reattaching turbulent shear layer in divergent channel flow. AIAA Journal, 23(2), pp. 163-171.


with an inlet pipe length of 1m, outlet pipe length of 5.5m and a step height of 1.27cm.


When I increase my mesh density such that y+ = 1, the SIMPLEC algorithm generates a converged solution only after about 18k iterations whereas the Coupled algorithm only after 3k iterations. In both cases I am using QUICK interpolation schemes. I have compared the two solutions and they are identical.

I know that once low-Re corrections are enabled a converged solution is obtained harder as the equations for k-w are integrated in the boundary layer as well but what I don't understand is why SIMPLEC is slower than the Coupled algorithm. I am simulating an incompressible flow so it shouldn't be the case. I am suspecting that it may be due to the choice in turbulence modelling but I am not sure.

Does anyone know any other potential causes for this?