Hello experts! A warm hello from Austria. I have a question for you. How would you model a reception desk from planar triangles. I would like to start with a cuboid. I have to do the design and there can’t be warped surfaces. Thank you for your help! A beginner
after the first days of the course - Rhino is my favorite animal !!! 
well triangles are always planar.
what do you mean by warped surfaces ?
any (hand drawn) design sketch? of the desk you want to do ?
kind regards -tom
Hy Tom_P! 
Thank you for your answer. In the construction I am not allowed to have deformed surfaces. All faces must be absolutely planar.
What is the easiest way for me to model something like this?
Kind regards Ben 
Hi,
I’d start drawing the bottom and top defining polylines in the top view, and move the top one to the required height in perspective or one of the side views.
Then you could Loft both straight to get the sides, Explode, and see which surfaces are non-planar.
CurvatureAnalysis can help identifying curved surfaces.
Next you could start replacing non-planar surfaces by substituting non-planar ones with triangular faces with SrfPt. As @Tom_P has already pointed out, tris are always planar.
You can use the existing surfaces as guides and references, but make sure to delete them after your done.
Join the surfaces and create the top and bottom ones with Cap, making the object a closed polysurface.
Et voilà!
i think @diff-arch shows the main steps.
two more notes
(1) for the initial curves:
(2) using _srfPt
(A) clicking 3 Points will result in a trimmed surface.
(B) clicking 3 Points - and the last point a seccond time will result in a untrimmed surface with two coincident points.
screenshots shows (bottom) initial triangular surface with control points on - and untrimmed (top row) surfaces.
I recommend approach (B) if you want to finetune the 3D-Design or have additional steps in modifying the surface.
Workflow B might - in worse cases - result in some shapes like this:
hope this helps - kind regards -tom
I’d suggest modelling this sort of triangulated object with a meshes. The advantage of a mesh with triangular faces is that you can drag the points and the neighboring faces follow. Not the case when modelling with nurbs. In the end you can still convert everything to nurbs / surfaces but for formfinding, I’d say it’s easier with a mesh.
Beat me to it.
This was the first thing that came to my mind.
thank you very much for your great help!
If this is a design piece for production, I’d surely go for polysurface/solid modelling.
Also the mesh modelling tools in Rhino are far inferior to the n.u.r.b.s. ones, and a nightmare if you want to introduce Boolean operations, fillets, chamfers, or similar manipulations.
I suggested using meshes for formfinding and then convert to nurbs.
All climbing wall designs I create are meshes first and only later on converted to surfaces like you describe for panel thickness and detail work…
nurbs fight!
Good points on both side, which will inevitably come down to the project and desired workflow. A large triangulated façade in the conceptual phase is going to be benefited by working in a mesh, the fabrication and detail would go towards solids & surfaces.
Adding agreement here: start with triangulated mesh like Martin suggests, then do all your detailing to a Nurbs model that you can obtain from the triangulated mesh using MeshToNurbs.
You can even play with grasshopper a bit here. Maybe try the tri-mesh component, that will mesh into triangle any sort of blob/mesh/brep.
You could also hook your desired form mesh into a grasshopper input for MeshToNurbs and then add all the detailing booleans parametrically. And when you see that you need to change the base form, you do that at the original tri-mesh level.
G
