BimorphNodes v2.1 Released with Ultra-Efficient Clash Detection and Geometry Intersection Nodes

8 posts were split to a new topic: BimorphNodes Curve.IntersectAll tolerance exception curve length too small

For anyone who might be looking to get the intersection points from the result of the Element.IntersectsElement clash test, I’ve added a video showing one possible workflow here:

1 Like

Anyone have a suggestion for the best way to run the clash detection between elements from a linked model that was moved into a new position in my model? I use the Springs.Collector.LinkedInstanceElements node to get elements from a linked model to clash with elements in my model, but I had to move the model into position in my project. The Springs node outputs the elements and the transform of the link, but the Element Intersect doesn’t work properly because the linked elements aren’t based off the same origin.

image

Hi @brunelli.b

Yes: wait for Monday when BimorphNodes v2.2 will be released!

The v2.2 Element.IntersectsElement, Element.IntersectsSolid and the BoundingBox nodes include new infrastructure to handle the limitation you’ve observed (I explained it here). The result?..

  1. Supports an unlimited number of link instances (i.e. if you have elements from many different links, it doesn’t matter, the v2.2 node’s can handle as many as you throw at them, regardless of any transformations made to the link instance(s)

  2. Maintains ultra-efficiency - the problem posed by elements from links is relatively straight-forward to solve, where it gets complicated is minimising the impacts on efficiency, more so with it being the fundamental ambition of the intersection BimorphNodes. Hence, I’ve implemented a smart logic-tree which selects the most efficient intersection process to ensure there are near-negligible compromises on speed-of-execution

  3. And on that subject, the entire codebase has been micro-optimised and built on the newer Revit APIs (2017+) resulting in a 15-25% speed increase! It means the wall vs duct test (4.5M possible clash tests) we conducted in the OP is now executing in under 7 seconds, and the worst-case scenario (due to the amount of processes that have to be performed) specifically, transformed link element vs transformed link element, is only marginally slower (I’ve built backwards compatibility for 2015-2016 too)

I’ll announce the release on the forum - it would be great to get your feedback once they are released.

4 Likes