Now and then, I have problems with plane normals for surface-related tasks, such as shell elements, surface loads, surface temperature loads, etc. I usually work around the issue by rebuilding the geometry or by using a construct-deconstruct-construct approach, for example, when dealing with the local system in shell elements.
Does anyone else experience similar problems? Is there another solution?
Depending on the geometry, I also get the same issue sometimes.
There is an ongoing research about tolerance as FEM-Design required a pretty tight value while Rhino/Grasshopper is more “relaxed”.
We do have an open case on GitHub and we hope to fix it at some point.
