Small Yplus and k-epsilon
 

Hi, Just for seeing if there is a difference between different RANS turbulence models for airfoil modelling, I have set up a k-epsilon run on a mesh originally constructed for SST-modelling (Ie. Yplus for the cells at the surface of the airfoil is approximately 2). When I run the K-E case on this mesh, I get a linear solver overflow, meanwhile the SST setup is solved without problems. So, I just wanted to know if the breakdown could be do to the low Yplus values? Has anyone experienced the same?

Re: Small Yplus and k-epsilon
 

What code are you running?

Re: Small Yplus and k-epsilon
 

I am using CFX-5.7.1

Re: Small Yplus and k-epsilon
 

It is probably not related to the small Yplus. You should check if the solver wrote a .res.err file. You can open this in Post and look for areas of high pressure or velocity where the solver overflowed. If the mesh is poor in this region, try remeshing it.

Sometimes the solver can tolerate poor mesh. It may be that when you run k-e, it is a little more sensitive to the bad elements.

If you don't have a res.err file, run it again and write a backup file just before the overflow.

Regards, Robin