Numerical method for the k-epsilon production term
 

Hi,

Thanks for your help with my previous question on interpreting the production term in the k-epsilon equation. However, I am uncertain how to evaluate this term in a system of equations for finite element analysis in 2D.

Would I expand the production term in the k- and epsilon equations into 5 differential terms (each containing u, v, or uv), then apply a shape function to each of these variables, multiply by a weighting function, and integrate (by parts, where necessary) to produce a set of equations in integral form?

Or would I treat it as a scalar, first calculating it using values from a previous iteration (like how the turbulent eddy viscosity is evaluated, for example).

Sorry if this is more FEM than CFD for this forum. Thanks in advance for any help!

-CS