Importing the step file 2527918_01_P_1_10_E31.stp I encountered an infinite loop when the triangulation tries to simplify a NurbsSurface.
The problem occurs because the face 6138_NurbsSurface_FullSurface.zip is very small and there is a singularity exactly at the max parameter.
Gemini suggested three possible solutions:
- "Adaptive tolerance": compute tolerance based on the size of the domain, and add safetyCounter to break if the loop gets trapped trying to find a missing point.
I also added an out ok parameter to know then not all the reguired points were found.
- "Offsetting Near Singularities": Instead of deleting a point when it hits a singularity, you nudge it slightly away (just outside the tolerance zone). This allows you to evaluate infinitely close to the pole without triggering a crash or division-by-zero, keeping your total count intact.
- "Gap Interpolation": standard sampling to collect all safe, non-singular points, if we didn't find the number of required points, add middle points of the largest gaps until the required number of points is reached.
I Implemented 1, because is the one that doesn't change to much the actual concept.
@SOFAgh what do you think?
Importing the step file 2527918_01_P_1_10_E31.stp I encountered an infinite loop when the triangulation tries to simplify a NurbsSurface.
The problem occurs because the face 6138_NurbsSurface_FullSurface.zip is very small and there is a singularity exactly at the max parameter.
Gemini suggested three possible solutions:
I also added an out ok parameter to know then not all the reguired points were found.
I Implemented 1, because is the one that doesn't change to much the actual concept.
@SOFAgh what do you think?