We were assigned to make a 3d printed joint that connected 2-4 wood rods which were about 20 mm in diameter. I decided to make one with the shape of a methane molecule. This way the bottom three beams can support the top one very effectively. This shape is also very modular-structure-able, so it's also something to consider if I wanted to make a larger structure.

Step 1: 3d Modeling

The first step is to model the structure using rhino.

I looked up the molecular structure of the methane molecule and it was a shape of a Regular tetrahedron, with each vertices (the H atom) having a molecular bond with the center of mass (the C atom). So, I started by making a regular tetrahedron in rhino, which there had to be an easier way of doing it than counting the height and matching it with different views......

Anyway, I did successfully make the shape and connected the vertices and the center, then I scaled it to different sizes (2x, 5x, 10x) and did a bunch of pipe commands on them, with a diameter of 24 mm (2mm for the wall thickness) and different amounts of division alone the path.

I ended up liking the 10x scale with 1 division along the path best (as shown in p3, they are respectively 2, 3, 1 division(s) along the curve) because I liked the shape in the center. (Looking back I probably should have done a no cap version so that the shape towards the ends is more rounded.)

Step 2: 3D Modeling Continued

Then I converted the subD shape into nurbs, and kept editing it.

I placed a bunch of surfaces perpendicular to the center construction lines and used them to cut the pipes shorter (each surface had a 150 mm distance from the center of the shape).

I filleted the edges and began trying to punch out the holes where the rods goes in.

Here comes the problem: because of the imprecise nature of the subD algorithm, the center of each pipe (on the 3d shape) was not where the construction line intersects with it -- it had about several mm of offset. So when I made the rods referencing the intersection point, it does not align with the pipe either. So I had to manually move each one of the 4 rods to position using a lot of helping lines and even setting up a custom view perpendicular to one of the pipes. There was no way I could make it 100% accurate but it was good enough for the project.

In the end, I was successful punching the holes in the pipes with boolean operation and went ahead to printing.

multi joint.stl
Download View in 3D

Step 3: Printing the Stuff!

Unfortunately the ultimaker was the only machine that can fit my project in, so I lowered the infill density a bit so that it didn't take 2 days to print, but it did still take 24 hours and change to finish the job.

Step 4: Final Product

After 24 hours I took it out and removed the support both on the bottom and inside the pipes (there wasn't a way to lay it down so that there was not support inside the pipe, and if so I would have to cut the shape in half and print two halfs independently and gluing them together, which I didn't want to do since it affects the structural integrity).

The residual support inside the pipes ended up helping the wooden rods stay inside nice and rigid which was a good thing.

One thing left that i still want to do was some how cleaning up the bottom the of the piece which looks a bit messy after the support was removed. I don't know if sanding would help but I might try that.