Skip to content

Couple of fixes for multipatch adaptivity#891

Open
akva2 wants to merge 2 commits into
OPM:masterfrom
akva2:fix_multipatch_refinement
Open

Couple of fixes for multipatch adaptivity#891
akva2 wants to merge 2 commits into
OPM:masterfrom
akva2:fix_multipatch_refinement

Conversation

@akva2

@akva2 akva2 commented Jun 23, 2026

Copy link
Copy Markdown
Member

Two copilot identified issues;

  1. need to loop again if domain is extended
  2. improve getBoundaryCovered - use getOverlappingNodes to also grab detect nodes further "in" from the boundary, which is in particular important for higher polynomial order.

akva2 added 2 commits June 23, 2026 15:18
When extendRefinementDomain() adds new indices to the refinement set,
the propagation loop must continue to propagate those additions across
patches. Previously, changed was only updated when extracting from the
initial seed set, not when the refinement domain was extended by
neighbor conformity constraints.
getBoundaryCovered() previously relied on one-way containment checks,
which can miss boundary nodes when refined (smaller) and coarse (larger)
basis function supports overlap without strict containment.

Changed to overlap-based detection: a boundary node is considered
covered if it overlaps (via getOverlappingNodes()) with any refined node
and lies on the boundary. This is more robust for selective refinement
patterns on multi-patch domains where support hierarchies vary per edge.
@akva2 akva2 force-pushed the fix_multipatch_refinement branch from 38a84b5 to e6af6e2 Compare June 23, 2026 13:18
@akva2

akva2 commented Jun 23, 2026

Copy link
Copy Markdown
Member Author

jenkins build this with downstreams please

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant