AMI implementation: Foundation vs. ESI
 

Hey,


I happened to notice that the AMI interpolation implementation differs between the Foundation and ESI versions. In particular, features presented here https://www.openfoam.com/news/main-n...20-06/numerics raised my interest whether I should try ESI version for my AMI cases. Has anyone carried out any tests regarding pros and cons between the implementation?


I'm also interested in hearing the history related to separation of AMI implementations.

I used to have some issues developing a working AMI when I was using OF7. It took me months of trying different things with my snappyHexMeshDict when the AMI examples seemed pretty straight forward. This made me look for overset meshing as an alternative, which is only available in the ESI version. Once I started using the ESI version I tried doing AMI again and it worked instantly. I went through the whole process on both ESI and OF Foundation versions and the exact same mesh generated a working AMI with ESI but it failed in OF Foundation. This was my experience at least. But from this I prefer ESI for any AMI development.