[Users] Meeting Minutes

Eloisa Bentivegna bentivegna at cct.lsu.edu
Wed Jan 22 04:30:11 CST 2014


On Jan 21, 2014, at 3:23 PM, Steven R. Brandt <sbrandt at cct.lsu.edu> wrote:

> When can the Elliptic solver be called? During an evolution? Only when complete coarsest grid steps are complete?

In principle, the solver is encapsulated in a single function call that can be scheduled whenever needed. In practice, its behavior in iterations where only some refinement levels are evolved depends on when exactly this function is scheduled, and which type of multigrid cycling is selected (V-cycle, FMG, etc.). In any case, the mechanism to set which levels are included and in which order is relatively straightforward, so one could code in a customized solution with little effort.

Best,
Eloisa



More information about the Users mailing list