How to do space frame on patch structure

Hello guys, I am very new to Grasshopper. This time I created a patch surface for my roof skin, however when i want to apply isotrim in GS, it auto detect the untrimed surface, how can i do for this problem?

base.3dm (335.5 KB) SPACE (17.1 KB)

I’d pipe the trimmed surface through the quad remesher to get a nice base mesh and construct the space frame manually (in Grasshopper) from it.

Hey bro, im very new here, could you tell me about how to build space frame manually on organic surface?

1 Like

A couple of things:

  1. I would strongly advise to go for triangles instead of quads or higher for a vast variety of reasons related with real-life/planarity/cost and other - more or less - rational factors.
  2. This means using something the likes of MeshMachine on your BrepFace (or your Blob as a closed Brep), do a tri Mesh and then do some truss.
  3. I also would strongly advise to go for a W type of truss (seismic forces etc).
  4. Solving a truss on a “blob” (organic) type of Topology implies addressing issues far and beyond of what a novice can do (most notably without code). These are related with pragmatic clash issues between the real-life members of the truss (cones, sleeves, bolts, tubes etc) not to mention the members of the envelope (after all we do the truss in order to support something).
  5. Unless this is some artistic/decorative thingy … the argument that a novice can do some sort of AEC concept of that kind and … er … that’s all is entirely wrong and a very dangerous companion if that novice becomes a pro some day (old habits die last)

Like (red in angles means clash issues on that node):

See a static example:

Test_data_56A.3dm (1.7 MB)

Bro, soorry for this request, i really dont know about how your work process, could you send me a grasshoper file to reference?

Sorry but this kind of stuff is strictly internal to the practice.

But I could give you an advise worth solid gold - if at some time you’ll become a pro in high demand (meaning lot’s of dollars in your pockets):

PRIOR attempting to do a similar thingy (the most challenging task in parametric modelling by a huge margin [most notably: by what means can you interactively fix clash issues?]) try to master the nuts and bolts of that thingy up to 1:1 detail. Only then you would be able to deside what to do, why and how to justify the cost to the client in a holistic way. Designing an abstract AEC parametric “concept” of that type is a nothing thing.

This is called bottom-to-top design mentality. Is the norm in MCAD matters long time ago … but finally gains momentum in the AEC world (was about time I guess)

As I said many times to people in my practice: kinda designing an engine: first know what a piston is/does then do the rest.