E250 Fix shear wall piers picking up plates from other shear walls by Ayberkrk · Pull Request #336 · JWock82/Pynite · GitHub
[go: up one dir, main page]

Skip to content

Fix shear wall piers picking up plates from other shear walls - #336

Closed
Ayberkrk wants to merge 1 commit into
JWock82:mainfrom
Ayberkrk:fix/pier-plates-cross-wall
Closed

Ayberkrk wants to merge 1 commit into
JWock82:mainfrom
Ayberkrk:fix/pier-plates-cross-wall

Conversation

@Ayberkrk
Copy link
Copy Markdown
Contributor

ShearWall._identify_piers, _identify_coupling_beams, and the material-assignment step in generate() all loop over self.model.quads.values() — every plate in the whole model — instead of only the plates belonging to that wall. The assignment then only checks a plate's local (x, y) against a pier's bounding box, ignoring the local z coordinate entirely, so when two shear walls are parallel and placed at different origins (e.g. two floors of the same wall line, or mirrored walls), a plate from one wall can fall inside another wall's pier bounding box and get double-counted.

I reproduced the reporter's exact script: piers that should have been identical ([4,0,0] vs [4,0,12], and [4,4,0] vs [4,4,12]) reported different V/M/P, and pier.plates visibly accumulated more entries for later piers.

Fix: added ShearWall._wall_plates(), which returns only the elements from this wall's own mesh (self.model.meshes[self.name]) plus any of its flange meshes, and used it in place of self.model.quads.values() in all three places. After the fix the previously-mismatched pier pairs report identical V/M/P, and pier.plates no longer grows across piers.

Added a regression test (test_pier_plates_are_wall_specific) with two parallel walls at different origins, asserting each pier only picks up its own wall's plates. Full existing test suite (179 passed, 1 skipped) still passes.

Fixes #335

_identify_piers, _identify_coupling_beams, and the material assignment
step in generate() all iterated over self.model.quads.values(), i.e.
every plate in the entire model rather than just this wall's own
plates. When two shear walls have overlapping local x/y bounding boxes
in their own coordinate frame (e.g. parallel walls at different
origins), plates belonging to one wall were mistakenly appended to the
other wall's piers, so sum_forces() reported inflated/duplicated
shear, moment and axial values.

Added ShearWall._wall_plates(), which scopes these loops to only the
elements of this wall's own mesh (plus its flange meshes), and used it
at all three call sites. Added a regression test with two parallel
walls confirming each pier only picks up its own wall's plates.

Fixes JWock82#335
@JWock82
JWock82 commented Sep 11, 2026
Copy link
Copy Markdown
Owner

I saw this after making the fix myself. Sorry to duplicate your efforts. Try v3.1.0.

@JWock82 JWock82 closed this Sep 11, 2026
@Ayberkrk
Copy link
Copy Markdown
Contributor Author

No worries at all, glad it's already sorted in v3.1.0. Thanks for the quick fix!

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.

problem with piers sum_forces

2 participants

0