[koscfd]

Dear All,

I am running a transient multiphase analysis using interFoam solver and I am trying to use all processors of my PC (6 cpus) in order to reduce the running time.
I used the parallel decomposition and so far have tried the "scotch", "simple" and "hierarchical" methods and the analysis runs much faster!
For example:
No mesh decomposition (base case): Run time= 20hours
"scotch" mesh decomposition: Run time= 5.5hours

The problem is that with the "scotch" method I get very different results from my base case. With the "simple" and "hierarchical" methods, I get results of the same magnitude as compared to my base case, but still not identical.

Do you know if there is a way to use all processors and get identical results with my base case?

Thank you in advance,
Koscfd

p.s. I am using OpenFOAM version 3.0.