Skip to content

cells_supporting returns an empty fault zone on a 3-D mesh #620

Description

@lmoresi

cells_supporting(name) walks each labelled point through _cells_on_edge, which is an EDGE walk: in 3-D it takes support-of-support. A conforming surface's labelled points are FACETS (height 1), so in 3-D the walk starts one level too low — support(face) is already the cells, and the second support step lands on nothing. The zone comes back all-False, silently.

The method was only ever exercised in 2-D (add_conforming_surface), where a facet IS an edge and the walk coincides with getSupport. The 3-D case first bites with the Mesh-level conforming sheet (PR #589): a per-cell viscosity assigned from the zone would silently be the background value everywhere.

Fix: dispatch on point height — a facet's support is the cells in any dimension; keep the edge walk for labelled edges (a trace chain). Test: the zone of a placed sheet equals the union of its faces' support cells (test_0862).

Underworld development team with AI support from Claude Code

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions