For this project to be a chess piece based on a surface that would be both unproblematic and would lend itself fairly well to the way in which we were meant to cast in halves. After looking at thingiverse, I found a low poly knight that I really liked.

First, I used this STL to generate the solidworks files. I did this by importing the original knight STL and converting the mesh into a solid feature. From here, I drew in some reference geometry that I planned to split the halves of the mold with. This included the splitting plane and the geometry for the “pegs” that would be used to square the pieces to each other. Then, I created a stock to split from for the size of the total volume of the intended mold halves. Now, I individually saved this document into two different versions: the left and right molds. In both files, I did a split to get rid of the area of the mold that would be filled with silicon, making sure to leave the peg on in one file while taking it out in the other. Next, I shortened the back side so that the 3d prints of the molds would be a little less thick and take less time. Finally, I did a square extrude for the height of the mold to place the walls. And then, I had 2 good solidworks files of my molds! (some files below)

Knight  | KnightLeftMold | KnightRight

Next, I went to 3d printing the molds, this was pretty easy, as I got a clean print on both halves on the first try. I also printed a to scale version of the intended product to reference how the casted part came out relative to a print. Here are how they came out:

From here, we poured the silicon into the mold forms. This was done by filling the molds up with water first, then measuring this amount out and using the same volume to mix my silicon. The mix was 2 part and 1 : 1. After mixing well, I simply left the casts to sit overnight.

Finally, it was time to pour the chess pieces themselves. First, I rubber-banded the two halves together so that the material would not leak. Then, I mixed the two part together (1 : 1) in the little shot glasses that we were given. This was a really cool step, as the process was both exothermic and let you see the clear fluid turn white when it was starting to solidify. With the first pour, I discovered that I had a massive bubble where in the snout of the horse head. I fixed this in all the next prints by making sure to pour up to about where the fluid filled the head, down to the neck. I then tilted the whole mold to fill that lower snout, and shook the mold a little bit to allow the bubbles to let loose. This worked really well on all the rest of the pieces.

After I had a few good pieces, I decided to mix it up a little. On one, I slightly overpoured into the shot glass, I decided that it would be interesting to see how it turned out, so I poured it into the mold and let it sit for a while. This piece took a little longer to cure, but came out more clear and hazy. It was also much more flexible when I demolded it, and did not appear to harden much more. I then decided to do that same mix on the next experiment, in which I used a printed printed mold that I also made previously to test. Here, I applied some mold release to the PLA mold and rubber banded the silicon it. This mold sealed much better than the previous molds and deformed a lot less with the rubber bands. After leaving the cast to cure, I started to demold it when the cast had a similar consistency to the previous experiment and it was beginning to cool down. Here, the cast demolded fairly easily from the silicon, but stuck a lot more to the PLA mold. This caused the edges of the cast’s PLA side to deform a bit. After this, I post processed the parts a little by taking a razor blade to all of the little bubbles and extraneous material at the mold edges. With this, I had all the necessary pieces.

Finally, I finished the CAM file for a half section mold. To do this, I imported the STL I used for printing into Fusion360. I followed along with the doc (restarting Fusion360 the necessary few times) and successfully set up the cam paths and the nc file, finally post-processing using the easel post-processor. With this export, it seems I was done.

Costs:

• Labor: $$115.00
  • 2 hr CAD/CAM @ $30/hr = $60.00
  • 1/2 hr Printer Setup @ $15/hr = $7.50
  • 1/2 hr Mold Maker @ $15/hr = $7.50
  • 2 hr Casting @ $15/hr = $30.00
• Materials: $ 8.67
  • PLA: (from 3dprinterOS)
    • Right mold(Ultimaker): $0.91
    • Left mold(Prusa): $0.57
  • 2 Part Silicon:
    • ~50 ml / 0.013 gal @ $150 / gal = $1.98
  • 2 Part Plastic:
    • 9 parts @ 30 ml /part @ $70 / gal = $5.21
• Machine Time: $105.00
  • 3D printing (Prusa MK3): 2.5 hr @ $15/hr = $37.50
  • 3D printing (Ultimaker): 4.5 hr @ $15/hr = $67.50
• Total Cost: $228.67

 

This was a perfect final project to end this class; while there might have some bumps along the way (and long printer queues), I really enjoyed the whole process of building this piece. This is one of my favorite classes that I’ve taken at Rice so far; while, at times, it is difficult, frustrating, and draining (like that midterm project), it is one of the most rewarding classes I have taken. I came in as someone with a lot more prototyping experience than most everyone else at Rice, but even so, I’ve learned so much.