[backscatter]

Hi,


It is very common for non-orthogonal meshes to fail in maintaining the bounded values of turbulent scalars (k, omega, epsilon etc) on the faces of of two non-orthogonal neighbor cells. Usually, depending upon the severity of non-orthogonality the following scenarios arise:
1) If Max. non-orthogonality < 60, the turbulence fields don't blow up usually.
2) If 70 > Max. non-orthogonality > 60, the turbulence fields will go haywire, and usage of nonOrthoCorrectors in the solver is advised to be increased .
3) If Max. non-orthogonality >> 70, then one needs to use limited discretization schemes to achieve stability or even revisit the mesh.

Now, that being said, I have an issue similar to point 1) above. My max non-orthogonality is 44 and my turbulence fields are pretty stable except for "k" which prints the bounding k warning. The avg. and max k values are quite bounded and agree with what I was expecting but I have a very small negative minimum value of k of the order of 1E-08. My simulation doesn't crash at all and the values of min k also don't change much, just oscillate between -1E-8 TO -1E-09 ish... My flow and passive scalar transport equations are pretty much converging very well.
I believe it is fine to run with that small "min k" if it is not leading to divergence of the solution. What are your thoughts on this?