I'm thinking that it may be a better idea to create an inactive equation system that contains information about the t=0 reference mesh configuration. This may be a better solution to my problem.
On Thu, Aug 6, 2015 at 4:14 PM, namu patel <namu.pat...@gmail.com> wrote: > Hello All, > > I am trying to solve a problem that requires a moving reference mesh. The > way I'm doing this is by cloning the t=0 reference mesh and using this > cloned mesh to calculate and update the positions of the nodes of the > moving reference mesh at the current simulation time. > > I obtain expected results when I am using only a single processor, > however, depending on if I'm iterating over local or global nodes in parts > of my code, I can get funny behavior. There are parts that I have to loop > over local nodes and parts that I have to loop over global nodes to > reproduce the results that I obtained with the single processor run. (I > find this odd, especially since I don't expect the cloned mesh to be > divided amongst the processors.) Can someone please explain mesh and the > cloned mesh communication in parallel? > > Thank you, > Namu > ------------------------------------------------------------------------------ _______________________________________________ Libmesh-users mailing list Libmesh-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/libmesh-users
