Based on iblank_cell this pull request creates an iblank array for nodes. The logic is as follows - If all cells shared by node are field, the node is marked field. If any cell shared by the node is hole, the node is marked hole. Everything else is marked fringe. The interface is changed to accept node iblanks when registering every patch.
I have also fixed a bug in the computation of nnode (not used anywhere currently) as part of this pull request.
Based on
iblank_cellthis pull request creates aniblankarray for nodes. The logic is as follows - If all cells shared by node are field, the node is marked field. If any cell shared by the node is hole, the node is marked hole. Everything else is marked fringe. The interface is changed to accept node iblanks when registering every patch.I have also fixed a bug in the computation of
nnode(not used anywhere currently) as part of this pull request.Edit: Following the discussion in https://github.com/jsitaraman/tioga/pull/29#discussion_r435960201 I have updated the pull request (687d927) to evaluate the node iblank based on the
donorList. Certain variables have been renamed for more clarity.Below is the
iblank_celland and correspondingiblank(averaged to cells using AMReX) for a multi-patch problem