fvSchemes for Skewed and nonOrthogonal Mesh?
 

I have prepared a Mesh for a UAV using snappyHEX. the mesh passed the check mesh but highest non orthogonality is around 85 but average is pretty less and max skewness is 11. I wish to know what schemes I should be using to obtain acceptable results with this kind of mesh. I am using the default schemes settings in fv schemes

85 seems pretty high so you should first try to decrease it in the mesh if you can.

Otherwise, you can try to set the laplacianSchemes to linear corrected or to linear limited 0.5

For more info on numerical schemes available, check out http://www.openfoam.org/docs/user/fvSchemes.php

You need "linear uncorrected" not "corrected" for the laplacian if the mesh is bad. Also, start with the safest (upwind) for all convective parts and "Euler" for ddt. "Gauss linear" for all gradients.

fv-Schemes in general
 

Hey everyone

I try to find an explantation to the fv-schemes.
I know how to use the main schemes. But I like to understand all schemes from this list.
Does anyone know where to find information about these.
Information could be anything like, the mathematical explanation the fv-schemes ( still I don´t understand c++) or descriptions about when to use which scheme.

The schemes I want to know about are listed in this document:

https://docs.google.com/document/d/1...it?usp=sharing

Tomorrow I´m going to ask someone to help me with C++, but I would be happy for each kind of information :-)

Best regards Tobi

Why when the mesh is bad Gauss uncorrected is needed?

Hello Everyone,

I think this problem is discussed many times on forum. The best answer for the same is answered in the following post:

http://www.cfd-online.com/Forums/ope...implefoam.html

For more information refer the weblink:

http://www.openfoam.org/docs/user/fvSchemes.php

-
Best Luck!

But you need to pay attention in that link. They are talking about two different things. The first answer is about non-orthogonal correction of simple algorithm, which is indeed not needed for accuracy, if the solver converges.
Your question was about a correction of the laplace operator. This is something different. Just keep in mind that uncorrected is bounded and corrected is unbounded.