[MissCFD]

I would like to better understand how turbulent viscosity is generated by k-epsilon turbulence model.

Can I say that the higher mean velocity gradient are, the more the k-epsilon model will product turbulent viscosity ?

[ghorrocks]

Yes, higher mean velocity gradients will result in more turbulent production which in general leads to more k which results in higher turbulent viscosity. But this is a gross simplification and ignores lots of important other effects.