Auto Adaptation + Periodic BCs

[samwicks]

Hi all.

I was playing with the grid auto-adaption on a favourite test study of mine which uses periodic boundary conditions. When enabled, the grid will happily refine itself until it gives the error that meshes on the two periodic faces are outside the tolerence for matching up.

Is there a way to prevent this?

[james.neville]

Hi Samwicks,

Thanks for your question. Currently, there is no provision to ensure that the periodic boundaries maintain a certain tolerance relative to each other. However, this also applies to simulations not utilizing the new mesh adaptation feature.

It's necessary that the periodic faces have similar mesh detail and there are couple methods to achieve this. I'd recommend experimenting with a few adaptation settings to see what works. Enabling or disabling "allow coarsening" may help. Also, you might try increasing the adaptation resolution factor or lowering the adaptation growth rate. Increasing resolution factor (default 0.75) will lessen the amount of refinement which occurs during each additional cycle. Think of this parameter as a scale factor: higher numbers will result in a reduction refinement similar to the size adjustment slider in the mesh generation dialog. Lowering the growth rate (default 1.3) will propogate changes due to refinement at a slower rate to the neighboring regions. A lower growth rate will cause the mesh to look more uniform, but this can come at a large mesh count penalty. I'd start with quarter point increments to the advanced parameters above. You'll find that there is a tremendous amount of mesh control available with just a few adjustments.

Hope this helps!

James