I have made a relatively simple definition specifically for profile extrusions in architecture.
The only problem is that once the profiles are extruded, I can not seem to close the breps.
I have tried cap, cap ex and I have tried patching the open crvs. I have even tried playing with the tollerances.
The problem is that I need the breps to be closed in order for them to work in unreal engine. So perhaps turning the open breps into meshes might be an option?
Anyhow its a handy script if anyone wants it.
Please see attached .gh file with curves embeded.
Would very much appreciate any help.
try to avoid those sweep1-issues - see post above.
check curve quality (duplicated points ? tiny segments…) / planarity. post a minimum .gh definition that shows the problem
(just 4 components: crv, crv, sweep1, cap )
try to keep the rail curve as simple as possible - still showing the problem
post a screenshot - so others see at least the complex profile / the geometric issue you re having.
I truly regret the time I spent on this. Fail and waste of time. I won’t even describe all the things I tried.
The section curve was poorly formed so I filtered out extremely short segments of the curve, changed my unit tolerance to 0.01 and rejoined the remaining fragments to make a new section curve - internalized.
But all efforts resulted in either an “Invalid Brep” after Cap Holes or multiple sections from the Swp1(or a Python version of that) when multiple sections (PFrames) were used. How silly
Awesome you got it to work!
My apologies for the profile curves; they originated from CAD blocks I downloaded from an Architectural Mouldings vendor’s website. I simplified the curves rebuilt it manually and ensured they were planar.
I truly appreciate your addition of an automated process for cleaning it up.
Ive been a struggling for me for some time with this in grasshopper. When I finally decided to tackle it in Rhino, I also eventually managed to make it work with “Sweep 2”. But doing it in grsshopper is much better!
Using “Perp Frame” makes life much easier as I often have trouble aligning with the object plane.
Thank you for investing your time in this, @Joseph_Oster. And @Tom_P, thank you for your input. I believe you’re correct; a single curve isn’t sufficient for a rail unless it’s supplemented with more information, so your point makes sense.
Regarding keeping the rail curve simple, the definition mostly functions when the rail is uncomplicated. The challenges seem to arise in complex turns. I appreciate your advice about sending only the relevant parts of the definition; I’ll make sure to do that next time. Thanks, guys!
