[jiejie]

Dear Foamers

I found a very strange problem when I tried to reconstructParMesh with the GGI results (a rotating propeller). The simulation was run with 128 cores

1st run: the case was run from t = 0 to t = 1. I used the reconstructParMesh to reconstruct the data without any problem. The data was good under Paraview

2nd run: due to the limited hard drive space on the cluster, I removed everything except for t =1 in the processor0 to processor127. the case was run from t = 1 to t = 2 without any problem. Again, I used the reconstructParMesh to reconstruct my data without error. However, the mesh geometry was completely wrong (looks like the node connectivity was lost, points are not connected correctly)

Did anyone have the same problem before?

Was this because I removed the 0 directory from each of the processor1-processor127 directories?

Should I change the startTime in the controDict as I realise that reconstructParMesh tends to create a polyMesh directory at time - startTime in the controlDict before it does the reconstruction processor?

I really appreciate if anyone used the reconstructParMesh.

Jie