[rkandukuri]

Hi All,

I have a somewhat peculiar issue. I am using overset meshes to simulate a 2 bladed VAWT. When I run the case in serial and check the output files, everything appears ordinary. But when I decompose the case and run it in parallel , I see the following: the bb for one of the component meshes seems to be far outside the domain of the case.
paralle.jpg

The bb seems completely alright when I'm running the case in serial.

serial.jpg

There is nothing odd when I run checkMesh either. For the parallel case, this behavior seems to depend heavily on my decomposition strategy for the case. Different techniques influence this bb behavior differently and it has a significant impact on performance. Am I doing something wrong with my decomposition? Is there any optimal decomposition strategy for my mesh that you guys might like to propose?

Here are my case files: https://drive.google.com/file/d/1x-P...88jxLzZz8/view