I am interested in varying the closure coefficients of a two-equation turbulence model for a given CFD simulation in order to see the effect (or lack of effect) from the corresponding coefficients. Is there a standard way to do this? Most coefficients are rather small ~0.01 to ~10, so randomly varying them, which would result in very large numbers as coefficient inputs, would necessitate the collection of a huge amount of data in order to see the effects on the flow field. Any help on this topic would be greatly appreciated.

Thanks in advance.

I would suggest something along the lines that was done with the Spalart-Allmaras model (AIAA-92-0439). Apply this approach to whatever you wish to do with the model in the end and built from the simplest case up.

-- Jarmo

First of all, you may try to solve your two equtaions models manually in very simple case :

Decay of homogeneous turbulence (U=0; k(t=0)>0; Eps(t=0)>0)

Growing of homogeneous turbulence under constant shear stress (dU/Dy >0; k(t=0)>0; Eps(t=0)>0)

Asymptotic boundary layer between two flat plane in the log low region (Production(k)=Eps), Du/Dy is known, uv is known...

Then you may see what is the effect of each constant and try to guess what should be their best variation domain.

As an example, forn the k-eps model :

Decay -> Ceps_2 (k propto t^(x), x depends on Ceps_2)

Homogeneus shear stress -> Ceps2-Ceps1; C_nu (uv/k(t->infinity) is known ...)

Boundary Layer : C_nu, Sigma_eps (see Patel for Sigma_eps)