Exploring Thingiverse, we knew we wanted our chess pieces to be some sort of animal. After choosing three options, we were cleared to 3D print a hedgehog for our final project. We printed it at a scale of 62.5%, which ended up being way too small.
So, we rescaled it and printed it again.
When we came into class with our hedgehog, it didn’t clear gate one because the overhangs with the spikes would be impossible to CNC. Returning to the drawing board, we selected a pig and 3D printed it.
When we came to class with it, there was a miscommunication regarding the size. We printed it much too small, not realizing how large the chess piece should be. There was also some confusion around the pig itself, as it would be difficult to slice in half and CNC well. As demonstrated below, with slicing the pig in half, the tail would be cut off, it would be difficult to maintain features like the nostrils, and overall was likely not worth the hassle.
So, we decided to switch to a duck. We split it in half on Meshmixer before creating both halves of the mold on SolidWorks. We then 3D printed both the duck itself and the 3D half for the duck mold at the correct scale and were cleared to continue working on the final. The 3D printed half had dimensions of 5.01573 in x 3.53359 in x 1.5 in.
For gate two, we came into class with our 3D printed half and volume calculations ready in order to begin pouring the silicone mold. We created a box of indeterminate size around the 3D printed half duck mold. It took a while because we used a ton of hot glue to ensure no silicone would seep out of the bottom or sides of the box. While time consuming, it was worth it, as silicone is an expensive material. When calculating the volume, we ran into some issues because the volume of the half duck we were using was generated from the 3D printer’s volume measurements, which accounted for using only 15% infill of PLA (rather than 100%).
Realizing this was an unrealistic measurement, we estimated a larger value, mixed parts A and B, and poured the mixture in the corner, as instructed in the videos. However, it did not fully cover our duck. So, with help from the TAs, we estimated how much more we would need and added more silicone. It ended up barely covering the top of the halved duck, but it did in fact cover the entire body. Setting it out to cure, we left. When returning twenty-four hours later, we were told we should add more silicone, as it was a little sketchy, did not cover over a centimeter above the duck, and could cause issues later on. Mixing more silicone, we poured it over what had already cured and returned twenty-four hours later.
Demolding the silicone, the entire body was cured and came out cleanly. We now had a successful half of our mold.
For the CNC portion, we were easily able to maneuver the second half of the SolidWorks file and create a V-carve with paths suitable for 3D printing. However, the file was created on a computer in the OEDK computer lab, and so it was not an appropriate version for the CNC machine.
To remedy this, we had to create the file again on the computer connected to the Nomad CNC machine. After orienting and entering all our specifications for the file, we noticed that the machine was unable to change drill bits, which was necessary for our roughing and finishing passes to occur. We then had to wait for the Shapeoko Pro machine to be available and set up our file and pathways on it. After securing the wood and attempting to run the CNC the machine unfortunately failed, and we were not able to continue working on the Shapeoko as all the technicians had also left.
Thankfully, Professor Bisesti helped us transfer our file to the Carvera, as well as get our roughing and finishing paths down to under an hour. We increased our step size but also encountered several difficulties with the Carvera, including losing our file midway on the roughing path and our bit breaking towards the end of the finishing path.
Nevertheless, it all worked out, and after a lot of sanding, we were finally able to pour our CNC mold.
We mixed equal parts of A and B, then left it to dry for two days. When we returned for Thursday’s class, we were dismayed to find that a section in our CNC mold had not cured properly. Albeit the initial worry, we were indeed able to unmold the left CNC half of our duck mold and pour.
It took us a few trials in order to calculate the correct volume for our complete mold, as we used the volume given from the 3D print of our sample duck and did not take into account the infill percentage.
After a few trials, we were advised to use a measuring cylinder to obtain an accurate volume of approximately 120 mL.
After finding the correct volume, we were able to freely pour the rest of the ducks. We colored them pastel colors, two in purple and three in pink.
Post-processing them was not so difficult either. The biggest issue with our molds was that the two halves were not lined up perfectly, and a slight lip was clearly visible from the top. Additionally, the pour hole, when cured, now prevented the ducks from standing like actual chess pieces. We used exacto knives to trim and even the lip and other uneven pieces that had cured with the ducks.
Then, as a final touch, using the standing electric sander, we filed the bottom of each of the pieces to remove the proof of the pour hole.
Lastly, we cleaned up the woodshop and the table that we used, and that was the end of a rollercoaster ride of a final EDES 210 project!
- 48 oz silicone kit = $11.72 for 24 oz (about 720 mL)
- 12″ x 24″ x 1.5″ wood block shelf = $2.67 for 3.5″ x 5.5″ x 1.5″
- 60 oz polyurethane casting resin = $15.03 for 24.35 oz
- 16 colors epoxy resin dye = $1.86 (3 colors)
- 3D printer use ($10/hr) = $75 for 7.5 hours
- CNC machine rental ($10/hr) = $30 for 3 hours
- belt sander ($10/hr) = $10 for 1 hour
- exacto knives ($6/knife) = $12 for 2 knives
- labor (2 people, $7/hr each) = $280 for 20 hours total
Total cost = $438.28 to produce 5 duck chess pieces
