August 20, 2013, 07:20
|
Correlation between mesh quality and solution/convergence
|
#1
|
Senior Member
Josef Runsten
Join Date: Jul 2010
Location: Gothenburg, Sweden
Posts: 188
Rep Power: 23
|
Figured I would give this a go in the forums. I am on the lookout for examples of how mesh quality affects the performance and accuracy of ones calculation.
We all know about the different mesh metrics (skewness, aspect ratio, volume change etc), but does anyone have cases where their significance are actually shown? For example running the same case with two different meshes, where one diverges and one converges, mainly due to one of these metrics?
Ideally the cases should be simple enough to perform in a few minutes, since I want to be able to showcase the limitations to other people. Solver doesn't really matter, fluent or CFX is fine. I'd rather use ansys meshing for meshing.
So what do you think? Is it possible to show, or are the solvers nowadays to good at handling bad meshes that is hidden? Again, both accuracy and solver efficiency is of interest, as long as the effects can be clearly shown.
|
|
|