1. Preparing the model

We used Fusion 360 to develop the file that we would create. First, we found a low-poly design on Thingiverse that seemed interesting, so we prepared the file. As this was low-poly, we did not have to import it into MeshMixer, so we simply scaled it to the proper size, and cut the file in half. Next, we used the built-in post processing techniques of Fusion 360 to generate g-code that could cut our piece out.

2. CNC Machining the wood

The next step was to cut out our design in wood. However, it did not turn out very well. We learned that parallel cuts going along the circumference of the piece perform better than ones axially. Additionally, because of the low-poly design, we had bad undercuts and an uneven design. It was difficult to see the finished result and just looked sloppy, so we would scrap this design.

3. Fixing the problem

Our next design improved in 2 major ways – first, the design changed to a more standard chess piece. This gave us a more uniform shape and overall just looked better. Second, we used wax instead of wood to cut our piece. The wax was a much smoother finish, and it was easy to sand down. It was also a larger depth than the wood, so we could make a full scale piece. Additionally, the base of the model was right up against the bounding box, so the base would act as a pour spout.

 

4. Creating the negative mold

With this CNC piece, our next step was to create half of the mold that we would use for our final piece. We tried three separate attempts of molding materials, with varying degrees of success. The first one on the left was Silicon 30 mold, which did not perform well at all. It was incredibly hard, and left bumps and air pockets. It also took a day to fully cure. It did not preserve the detail of the model as well as we hoped, so we aimed for other options. The next one we looked at was Oomoo, which was the middle mold. This one had better detail and was more flexible, however it degraded somewhat easily. This means we were left with molds with holes and cracks, and not an ideal design. Lastly, we used Mold Star, which finally achieved everything we had hoped. It was flexible, maintained the details well, and did not fall apart. It also had a very quick cure time, so we could make multiple quickly. We used Mold Star for all future negative molds we created.

5. Beginning to mold

Next up, we were finally able to mold! We used liquid plastic like our fingers, which cured in a manner of minutes. Our first attempt was to create two halves and glue them together. While this actually worked decently, the pieces ended up being a strange shape, since if the mold was not totally flush with the Mold Star, it wouldn’t be semicircular. We also tried to simply estimate the two halves together, however, some misalignment formed. Fortunately, Gigi had the idea of using the glued piece that was properly aligned to form a registration key in our mold.

6. Keys to success

We created the registration keys by placing the glued queen in the Mold Star, and using a drill press to create holes for the registration key. We then inserted dowels of the proper dimensions in one of them, which forces them to stay aligned. This actually worked surprisingly well!

7. Success!

We are left with two beautiful queens, and the mold available to make many more. This was after fairly extensive post processing to eliminate the effects from bleed, and cleaning as the piece had gotten dirty after sanding.

8. Conclusion

Overall, I learned a lot from this, as I knew nothing about molding and casting coming in, and little about CNC machining. I am now much more comfortable with this process, and feel like I would be able to apply it on future projects. I also enjoyed the applicable nature of this assignment, which would help the OEDK to create a chess set in the near future.