Summing up – pure frustration. I had that perfect little Dynamo script ready for the client where we still wait for money…, another story.
That is the model:
The Dynamo graph:
So the client takes it and uses it on one of his real live projects – fail, epic…
My Revit session gave up after consuming 32 Gb of RAM while testing. Bummer.
NOTE: The architectural model was an IFC import and most of the walls disintegrated into faces. That means whatever geometric test we will be doing will be tough I n terms of computing. Basically we’ll need to take the cross product of all MEP against all Architecture.
THIS IS THE PERFECT DYNAMO QA OPP…
So – what are we going to doi about that?
First gut-driven feeling – parsing Revit clash reports:
Not so happy either – false negatives. That is not good. Lost more than 50% clashes on some ‘Geometry cannot..” – warning. I mean really – #Autodesk – you swamp us with stuff and we still have a 1990ies “Geometry SAT fail” thing going on? Really?
So my second thinking was – alternative strategy: I turned onto #Dynaworks – Dynamo package.
So first – run a clash detective in Navisworks; resulting in 14.000+ clashes
Tried to feed it into this graph:
So – kudos to Adam Sheater (find him here. for his package for Dynamo) we can parse a clash report from Navisworks and put stuff back into Revit.
Result – immense computing time – stable – well at least it did something. Trying to visualize the bounding boxes of the clashes Navisworks found was then again – interesting…
At this moment – there is no practical solution. Potentially break it all down into smaller pieces. And it really helps that all walls in the IFC are on one single level (banging my head)
Still trying to make sense here – how are we going to handle the massive amount of clashes?
#gotta call the guy… not a pretty message….
Yep – stuff for the days to come and just not a #HappyEnd today…