This weeks blog post brings our final assignment: molding and casting a chess piece. For this assignment the class was split up into pairs, and each pair was able to choose a chess piece to make based off of a random draw. My team consists of John Michael and myself, and the piece we were assigned was a rook.
The first part of this project dealt with finding a chess piece to fabricate. We first began by looking on thingiverse to see what chess pieces already existed, and to see if we liked any of them enough. One thing we needed to keep in mind was that our piece could not have any overhangs when laid on it’s side, because the CNC machine we will be using to make the positive of the piece cannot make any overhangs. This ended up limiting the number of thingiverse pieces we could choose because most have an ridge at the top of the piece signifying the tower wall. To account for this limitation, we ended up deciding to design our own piece. We first used solidworks to make a simple rook design, then found a small owl file on thingiverse to perch on top of the rook. We combined these pieces together in meshmixer, and this final design can be seen in figure 1.
Figure 1: The Owl Rook Design
Once we had the design completed, we used fusion 360 to set up a test cut to make sure the there would be no issues with overhangs when CNCing and to make sure that the resolution of the owls face was adequate. We ended up completing this test cut on foam, and this preliminary cut can be seen in figure 2. This test cut was a success: there were no overhang issues, and the owl had good definition. From this test we knew that our design worked, and could begin making the final CNC design for the positives.
Figure 2: Our Foam Test Cut
Making the final CNC designs required a bit more planning and effort. Since our design in not symmetrical (both sides cannot use the same CNC cut), we needed to design two cuts , and make sure that these cuts would align properly. Since our design was saved as an STL file, we needed to create the alignment pegs and holes set the placement of the piece in Fusion 360. To do this, we built a flat body behind the piece such that the rook half was centered over this body, and the alignment holes and pegs came into/out of the body. We also added a pour hole to help when using the molds. At this point we only had a design for one of the halves, so we needed to mirror the design. To do this we needed to only mirror the rook piece but not the flat body. If we were to mirror both, then the alignment holes would not line up correctly. To do this, I saved both the body and to rook as separate STL files from Fusion 360, created a new file, and imported both the body and rook. I then converted the rook to a body from mesh, used the mirror feature, then converted it back to a mesh. I then used mesh mixer to combine the back piece and the rook half for each side, and with that the designs were complete! One of these designs can be seen in Figure 3 below. At this point, I passed off the file to John Michael to prepare it to cut.
Figure 3: Our prepared design
Once I got the CNC file from Liz, we started the cutting process. When we hit start, the Carvey took forever and was clearly not at the right height. I then recruited Jeremy and Tiffany, lab techs on shift, to help me with settings. But changing the pullback height and downloading the carvey toolbox to Fusion 360 we were able to get the cut down to 45 minutes. We used a 1/8″ fishtail bit without a smoothing pass, and when we cut both positives, they came out well on the first try. However, our alignment holes did not cut out, which may be because they were too small for Fusion 360 to make a tool path so, once the positives were cut, I drilled holes for the male parts of the locking aspect of our molds using a ruler and drill press. This was not perfect, but certainly worked and we didn’t have too many issues getting our alignment correct once we started pouring. Pictures from the CNC being cut and the final positives can be seen in figures 4, 5, and 6 below.
Figure 4: Our positive being cut in the carvey!
Figure 5: One of the CNC’d positives
Figure 6: Cutting the male alignment holes with the drill press
With the positives made, we created each of the two molds, and began to make full copies of our rook design. When making preliminary pieces, we used a clamp setup pictured below in figure 7. While this setup did hold the mold together successfully, when putting the mold between the wood pieces, the mold would not align correctly, creating imperfect rook pieces. On a funnier note, when using the quick set positive material, John Michael ran out of time to pour and ended up with a pour frozen in the middle, which can be seen in Figure 8.
Figure 7: The original clamp design
Figure 8: A piece, frozen mid-pour
To account for this, we ended up holding our mold together using two pieces of wood and rubber bands. This works well, and allowed us to get the mold as aligned as possible. Through using this technique we were able to create our final pieces, which can be seen below in figure 9.
Figure 9: Our final pieces!
Cost analysis: combined Liz and I spent about 40 hours on this project, which comes out to $1,600 for labor. The carvey rental runs about $200 at a Houston makerspace, and the wood was about $8. Throw in all the chemicals needed for the mold at $240 and we come to a grand total of $2048 for two small chess pieces. Wow.
With that, both this assignment and this class have come to a close. Throughout the semester we have been able to learn a lot and work on some really cool and challenging projects. This semester has been very fun and rewarding, and we look forward to using the skills we’ve learned on all of our projects in the future!
