Hello! I’m new to grasshopper as you will soon realize, but to the point.
I have been trying for a while to create a model of a wooden frame for, lets say a dome. I have been trying to use weaverbird picture frame, and then mesh thicken to give it depth. But these two options only lets you decide thickness in procentage of edge length, which leads to different thickness on different sides.
I haven’t uploaded any gh file or anything since it’s only a polarzonohedron with the picture frame and the mesh thicken
Still upload it.
Mesh thickening just isn’t the way for this kind of thing. Offset each outline of the mesh faces inward with the faces normal plane, then start to construct the frame.
You might better like the option “1” for T: InsetType. If it doesn’t do what you are looking for, you might want to try develop a different geometric option yourself. This option works well for faces that are regular; for irregular faces it might not look balanced.
I takes a triangulated base mesh, like the icosahedron above, and thickens it.
The offset in the plane of each triangular mesh face can be altered by distance.
The “extrusion” in the averaged direction can also be changed but the distance only applies in the average extrusion direction. It’s not a straight extrusion (normal to the face), but a compromise between the face and its neighbours.
The resulting geometry are breps, not meshes!
Since you didn’t upload any geometry I wasn’t able to test it with yours. What I remember is that it only works with triangulated meshes and is a little bit finicky.
Hey, thanks for the tips. I tried to use this definition used on the icosahedron above, but I can’t seem to get it work on this project. I’m attaching the file I tried it on. But it extrudes in different directions and also i’m not quite sure how to convert the multiple srf’s into a normal mesh. It seems to use the mid points as an endpoint in the conversion. Other than that this definition is exactly what i’m looking for! Would be wildly appreciated if you could look at it and give me an direction to search in!
The torsion happens because of the anticlastic nature of the base geometry - it’s doubly curved - and the logic behind the thickening process.
The offset of the mesh face boundaries in their individual planes with a desired offset distance that remains constant while also extruding the offsetted profile curves in the planes normal direction, would result in a straight extrusion. However, in places where the curvature of your base geometry is important, the individual face extrusions would wildly intersect each other or form gaps between each other.
What I’m doing different is that for each vertex of the base mesh, all normal vectors of its connected faces are averaged and normalised. The extrusion thus isn’t done with the same face normal for each vertex of a face, but the resulting, average unit vectors of the individual vertices of a face. This means that the vectors of an individual mesh face aren’t forcibly parallel anymore, and thus torsion is introduced. However, this is probably necessary (?) to keep the same material thickness (i.e. offset distance) from the top to the bottom of each extrusion, get a homogeneous extrusion height through the system, and have all extrusions match up after being extruded.
To prevent the twisting of the beams, you’d have to make a trade-off!
I think (?) that you’d have to sacrifice the constant offset distance (i.e. material thickness) throughout the system, meaning that you could only define the starting (or ending) offset distance, but the thickness at the other end of the extrusion would be the compromise.
You can probably imagine that at places where gaps or intersections would have been formed by extruding along the face normals, substance will added or subtracted in this case, depending on the curvature.
My definition was originally for an icosahedron (cf. image above), which has synclastic curvature and the primary goal was to keep a constant material thickness to facilitate construction out of standard issue wood studs.
@diff-arch@philip.zurdosson I didn’t go through all the discussion. But yes the problem of the twist could be solved but you will have to free some other parts of the design.
You could look at Morpheus macro windows
I put many links here.
Yeah I see what you mean with the double curve, if I extrude the model in three segments there is much less torsion if any even. But then like you said, they dont fit between them. And the reason I want to to this is becouse i need to build it, so I wanna be able to get out all the angles on paper for cutting. And i’m not sure how correct they will be if the thickness changes over the extrusion? Since that adds another “angle” that wont be there in reality.
Would be wildly appreciated if you could look at it and give me an direction to search in!
