CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   NUMECA (http://www.cfd-online.com/Forums/numeca/)
-   -   turbulence modeling error at a stagnation point (http://www.cfd-online.com/Forums/numeca/51017-turbulence-modeling-error-stagnation-point.html)

 erdem August 8, 2006 04:27

turbulence modeling error at a stagnation point

hi

I'm studying external aerodynamics of Ahmed Body car geometry. I have used realizable k-eps, SST k-omega and RSM models. in terms of drag coefficient I got similar values with all models. but all of them calculated nearly 1.5-2 times drag at the front of the body (comparing with the original experiments. Realizable k-epsilon has the highest drag on the front.

why does this happen? would you please help me about investigating the greater pressures at the front caused by turbulence modeling? I have to write a discussion about it. this is really important. thanks for you time.

Erdem

 PROF. VIVEK. YAKKUNDI August 15, 2006 14:46

Re: turbulence modeling error at a stagnation poin

sir/madam This refers to your querry regarding turbulence modelling error at stagnation point. As you have mentioned the drag in the front has to be morein contrast to the stagnation point and other regions of the car body.This is because in the front the flow is more or less unidirectional where eddy viscosity needs to be properly modelled in order that Reynold's stress uv is rightly described.Due to wall blocking effect or viscous damping the transport normal to the wall is inhibited and therefore eddy viscosity must be decreased accordingly by an appropriate factor. Secondly the prediction of drag value in k-eps model is higher because the wall function used is based on the assumption of constant stress hypothesis( As demonstrated by Wilcox)and therefore the measurements are consistent with only constant stress. However Menter's model of shear stress transport ( k- omega) is devoid of these descrepancies and is consistent with measurements for all pressure gradients.Also RNG K-eps model is found to be ( yakhot & orszag)consistent as it contains its own low Reynold's number version and at high Reynold's numbers it gives the usual wall law. hope this reply clarifies your doubts.

 erdem August 15, 2006 15:40

Re: turbulence modeling error at a stagnation poin

Dear Mr. Yakkundi

Thank You for your kind answer clarifying my question. Is there any publication of yours regarding this subject that I can cite in my thesis report. thank you

best regards Erdem

 All times are GMT -4. The time now is 09:49.