I have to compare two models: one made with StarCD, one made with AMESim.

In AMESim I can define for my duct the friction coefficient.

In StarCD I have roughness in wall boundary (parameter Elog or similar) and in the Turbulence model (I use k-epsilon with high Reynolds number).

How can I match them together? How can I evaluate StarCD parameters from a friction coefficient that I know?

1) AMESim is a 1D tool which you have to supply with empirical correlations for friction, heat transfer and so on 2) STAR-CD is a 3D tool with which one hopes to get rid of empirical correlations 3) So what you may compare are results of AMESim and STAR-CD with measurements or 4) results of AMESim and STAR-CD with approved empirical correlations (which of course you had implemented in AMESim beforehand)

If you have to define a friction factor to model your roughness, you can try to use the porous media facility provided from star Cd... You can found more suggestion in your alfa and beta choose in the methodology guide Example of calculation (take a look to the 2nd example)(distribuited resistence).