[ViLaks]

Hi All,

I would like to know what effect the minimum edge length has in the mesh quality. Here is my problem. My mesh skewness was 0.92 and number of cells was 40L with minimum edge length of the order of 10^-04 mm. I had convergence problems and my problem did not converge at all. After refining the geometry further, my minimum edge length became of the order 10^-01 mm and my mesh had skewness of 0.90 with 42L cells. On running with similar case setup, the solution converged.

Is there a relation between minimum edge length and mesh quality?
Is it so that, for example, a 0.88 skewness mesh withmin edge length 0.5 mm better than a 0.88 skewed mesh min edge length of order 10^-05 mm?

Thanks and REgards
Vignesh

[LuckyTran]

Skewness can usually be traced back to the topology of your mesh and your overall geometry.

Imagine a small wedge with an angle of say 3°. With a certain minimum edge length, you split this wedge into two parts of 1.5° which has some skewness. Now you lower the minimum edge length and are now able to fit 3 cells into this little corner; so you now have 3 cells with 1° angles and now you have a more skewed mesh! This example shows how packing a bunch of smaller cells into a little corner does not improve skewness. But the real problem is the wedge, the geometry! The fact that you have a little wedge with an angle of 3° in your domain is ultimately the cause of the skewness in this example. In this example, the geometry limits the skewness no matter how fine or coarse you make the mesh, you can't have a better skewness angle than 3° and everything you do can only make it worse.

[ViLaks]

Dear Lucky Tran,

Thanks for your precise explanation!!!!

With Regards
Vignesh