lagrangian particle tracking fail in parallel
I'm trying to implement a new hybrid lagrangian particle/FV solver for pdf method. But I'm experiencing a problem with parallization.
It seems that the some particle can not get through the processor patch and the the solver traped in a dead loop. (the return value of trackToFace() is zero.)
Did anyone have the same experience?
I finally figured out what's wrong. When creating your new particle class be careful about how you write the particleIO.H. I find some weired particle private variable values after particle across the processor boundary. If you write the IO wrong then the code will read wrong memory location and the code will fail at parallel
|All times are GMT -4. The time now is 21:35.|