Hi evereybody

I am interested in studying pressure distribution on different objects. Is there anyone that know which turbulence model, differencing scheme and solution algorithm is best to use? The K-E model seemes to overpredict the pressure at the stagnation point a lot.

I am also interested in studying the frequency of the von Karman's vortex street in transient analysis. Is there anyone that know which turbulence model, differencing scheme and solution algorithm is best to use in this case?

Regards

Lennart

It seems that the best combination is: cubic k-e turbulence model (especially the lower Reynolds number version if you can afford the mesh density near wall) and MARS scheme for convection term discretisation. The solution algorithm should not afttect converged results. The v2f model is very good - but you have to pay for it since it's an external model from Standford.

CD should be able to send you its case of flow pass the square cylinder with von Karman vortex. It seems K-e model is good enough to do it with above 95% blending between Centeral and upwind schemes. But you have to have reasonable mesh density around the obstacle to resolve the vertex. Also set your upstream boundary (inlet?) far enough from the obstacle with proper profile of boundary condition.

Hi John

Thank you for your answer.

I have tested different turbulence models (heigh Reynolds only) and MARS. I have compared the Cp-factor at the stagnation point. As I think you know

Cp = (P-Pref) / (½·Rho·U·U)

Theoretically it should be 1.0 at the stagnation point. My little test givs that the RNG turbulence model gives the best result. It was also much easier to get STAR to converge with RNG than "Cubic Heigh Reynolds". Wth the standard K-E model I got the highest (=worst) values.

Have you compared the pressure calculated with STAR with measurements?