Hey all,

I need to simulate the destruction of a high-res mesh based on the output of a physical FEM analysis. So far I managed to I managed to calculate the deformation vectors, and using it with 'point deform' to manipulate the mesh. On a macro level this gives me the precise deformation I need, however, the analyisis report doesn't include the deformation of the individual profiles and 'point deform' simulates it a very unnatural way.

Ofc there are dozen of metal bending / tearing Houdini projects out there, but my question is: Is there a way to use the 'point deform' output as a simulation input within a RDB or Vellum network? Or is this only possible if you simulate the physical event (e.g. collision)? I'm a little lost as I'm not native to Houdini.. What's the right way to solve this problem?

Thank you!

more information needed, maybe a scene or more images with your hires model + the that csv visualisation.
what inputs do you pass in and what result do you want for the output
thanks

Thank you for getting back! Find attached an example .hip + geo. It is a different mesh (from STEP vs IGS) and lower-res for the sake of bandwith but the problem stays the same. I want to get the overall transformation of the beams from the difference of the two point sets. However, I need the overall "integrity" of the beams to stay intact. Meaning that if they bend, they throw folds and wrinkles like in the photo attached. I couldn't make it work with vellum, maybe also because it's to resource intensive. The Drop Off / Weighting of the Bone Capture Proximity Node come's very close to what I think would be needed (especially together with a Tension Map like: here [www.sidefx.com]. But I can't make it work without the whole bone structure. Is it possible to construct something like this with the "skeleton" of the point?

Or is there maybe a totally different, better approach? I'm a little desperate and would v much appreciate any hints / tips!

This is the .PLY file used together with the remesher in the first post. It's based on a IGS instead of a STEP and is better suited for triangular remesh from my experience: (Dropbox) [www.dropbox.com]