For the final project, we chose to cast a sword as our chess piece. The initial, unmodified file can be seen below.

The first attempt was to print it upright, but this led to massive supports being created for the hilt of the blade, which led to the blade snapping when we tried to remove the supports.

The fix for this issue was to print the piece while lying on its side, so the supports needed were significantly smaller.

For our second iteration of the design, we removed the large base included in the original file to use our own smaller base. This allowed us to save material and scale the blade up to increase the blade thickness without reaching past the size limit of the wood used in the cnc machine.

This iteration of the mold included air channels for the hilt and handle base, and a hole specifically for pouring the casting resin into. The feedback we received was to remove the air channels for the hilt, remove the pour hole, and scale the entire blade up to increase the blade thickness close to 1 mm to ensure it was strong enough. One additional change made to scale the piece to a larger size was to remake the base to be even smaller. The final 3d printed mold half used is shown below.

The next step was to create the silicone halves of the mold. The silicone halves were created by following the process in class, where the first half used the components measured by volume, while the second half needed the components measured by mass. Both halves of the mold were created from 3D printed parts, as the larger cnc machine had broken down. Cutting a mold half out of the smaller cnc machine would have resulted in a 7 hour cut.

The first batch of chess pieces were cast with white resin, with a few issues. Due to the thin blade, when the part was first taken out of the mold, it would bend under its own weight. We had to rest them on their side to ensure that the pieces did not bend while curing to their full strength. Furthermore, at the corners of the hilt, the casted piece would be incomplete indicating air bubbles and certain parts of the design, most notably the handle, would be misaligned for some pieces. As a result, we had to address both of these issues during our post-processing.

For the next batch, we wanted to reduce the time and effort spent on post-processing the pieces, so we tried cutting air channels for the hilt into the blue silicone half of the mold.

For our last batch, we wanted to mix the blue and red pigments together for a purple piece, but there was an issue with the mixed resin, and it did not cure, leading to a tacky substance that ruined our mold and was difficult to clean. One possible cause for this was the large amount of die included in the resin parts, which would have thrown off the mass measurements we were making to mix the components together in the proper quantities. The tacky substance was difficult to clean even with using acetone.

We then began post-processing the pieces by using sandpaper to remove excess resin on the base of the mold to create an even and leveled platform. Afterwards, we then used X-ACTO blades to remove imperfections and to essentially sculpture the pieces. For the newer batch, this meant sculpting and evening out the handle and hilt (sides were significantly uneven on some) as well as removing any mold-casting lines from the blade. For the older batch, the post-processing was similar with the addition of carving the corners of the hilts out as they were not fully filled out. When carving out the corners, we attempted to make a curve cut-out that was identical on both sides of the piece.

Final Pieces

Cost Estimate