2D Ahmed body turbulence model
 

I have been simulating an Ahmed body in a steady stream. I have obtained wildly different results from using different turbulence models. All simulations have the y+ range of 0<y+<~1.


1) k-epsilon Realizable - Enhanced wall treatment:
The flow remains attached over the top of the body

2) k-omega SST - Low Re correction:
The flow separates at the curved front surface of the body.


Question: Which of these is "better"?

Attachment 74766

Attachment 74767

your loyal servant,

Jerg

Compare against exp. data
 

Ahmed body is a rather well studied profile, so, there is a lot of literature available at various Re numbers. I suppose it would be better to compare against those. I have no experience with Ahmed body, so, I cannot comment on the accuracy of the predicted flow profile but validation requires some theoretical or experimental data or an experienced peer.