What to make?!
Katy and Kaira here! When tasked with this assignment, we knew we wanted to go the animal route for our chess piece, but struggled through a couple of options before landing on our final piece. Our first choice was from the woodland animal chess set, where we wanted to produce the squirrel queen. After realizing that no halved configuration would work on the CNC mill due to the overhangs, we had to either add a base to fill in under the squirrel belly or choose another piece. A lot of the other animals we found had the same issue, until we found the polar bear files. The options were so good we had to test print a couple on the Prusa FDM 3d printer to see what would show up the best and what sizing we wanted. We did patissier polar bear, milk and cookie polar bear, candy cane bear, and snow globe bear. One of the issues we had with scaling the bears was that the body was so rotund they became very wobbly at larger sizes with the items the bears were holding. In the end, milk and cookie polar bear was the winner, and we found out we could plug the small gap between the milk cup and arm with clay when we went to create the mold which made our decision easy!
Forming the polar bear- part one
We next began slicing the model on Autodesk Fusion then SolidWorks to separate the part to be CNC milled and the part to be 3D printed. We scaled the part down to the size we wanted then sliced in “half” before adding a base plate, channel (for air bubbles), and pegs/holes to hold the mold together when it came time to cast.
Our first 3D print came out well when we did a layer height of 0.1 and an infill of 75, but when we went to mill the other half, we had a couple kinks we needed to work out. One of the pegs was a bit too close to the polar bear, which made it difficult for the tool to get in between. This was in part a limitation of the size bits we had available and that we were creating a recessed piece. We had also put the file right in the center of our block, but had hoped to use the feet to pour in the plaster, so we wanted to move the file to the edge. Furthermore, we decided to cut back on the number of pegs, tolerance the holes to be slightly larger than the pegs, and remove the channel on the milled part since having it on the 3d printed piece would be enough to remove the air bubbles.
We milled the piece again with the same settings (rough with ⅛ flat end mill, finish with ⅛ ball nose mill, and finish with 1/16 ball nose mill) and successfully finished our part! Now onto molding.
Forming the polar bear- part two
Before starting molding we needed to secure the edges so no molding material would spill out when we went to pour. To do this we hot glued cardboard along all of the edges. This was a bit more intensive for the 3D printed half that had no walls, but for the milled piece we just added one wall by the feet and then thin walls to raise the other three sides a bit. We managed to seal things well with the hot glue gun so there were no leaks.
For molding we mixed 1:1 by weight of the acetate liquids – one green and one yellow. We used a scale to accurately measure our equivalent weights then mixed them together before pouring the liquid into each positive half (from part one), making sure none of the polar bear was visible. The mold had to set for 4-5 hrs, so we came back the next day to remove it. The mold on the 3D printed half came off very easily, especially once the cardboard collars were removed. The mold on the milled piece was a different story, probably partly due to the removable cardboard not extending all the way to the bottom of the mold, and the greater stick of the mold to the wood than the plastic.
In typical fashion, the mold was highly detailed and it was clear which half came from the milled part and which came from the 3D printed part by the layering lines. The milled mold half had layering visible at the sides seam of the bear due to the slope limiting the bit access. The 3D printed mold half had the layering visible at the belly, which makes sense with the z-axis layering upward to the top of the piece. We cleaned up the molds a little where the lines weren’t as clean as we wanted, like around the feet where some liquid had gotten in between the bear feet and the cardboard, using an x-acto knife. Once it was up to our standards, it was time for casting and keeping our fingers crossed that our polar bear chess pieces would turn out!!
Forming the polar bears- part three
For casting, we used liquid polyurethane, which was also in two parts but was combined with equivalent volume. Since the mixing of the liquids is a chemical (exothermic) reaction, we needed enough volume of each to cause that reaction while not wasting too much with our small molds. We secured our molds together with popsicle sticks and rubber bands then poured in from the feet our first cast.
We forgot to tap this one down at all while the casting material was filling, so we noticed when we removed it 10 mins later that the liquid hadn’t gotten into the small space where the cookie was supposed to be. While this was clearly a defect to us, we were encouraged to see that there wasn’t anything else wrong with the part per say, and gave us a part to test ‘flash’ trimming methods on. It was determined that the deburring tool would be too large for the minimal flash on the seam where the mold halves came together, so we did half with the x-acto knife and half using a metal file. We felt the x-acto gave a cleaner cut and better allowed us to cut away the desired material and not accidentally file away surrounding areas, so that was what we used on the following pieces.
The next casted polar bears turned out with their proper cookies, and tapping the liquid right after pouring helped to bring bubbles to the surface and give time for the fluid to settle. This also gave us the ability to fill in the feet a little extra when some of the fluid went through our leakage channel and the overall level dropped. We assumed the channel was the cause of the feet having a “caved-in” quality to them, however, after many times of overfilling the mold, we came to understand that this was just part of the reaction and we could easily sand down the feet to make them all identical. The nose was a slight defect, being such a small part that it always filled with an air bubble before the fluid could filter in, even when we swirled the mold around to ensure all the sides were coated in the midst of pouring. There were also some visible air bubbles around the paw which could be seen upon very close inspection, but nothing too major. One of the pieces successfully got a nose, but the other bears were stuck with their cute cupped-hole noses. After 6 pieces we started running into some issues with the cast not setting. First, we had a mold that leaked significantly with major defects around the feet after attempting to top it off, probably too late in the chemical reaction. The mold must not have been fully dried after that one, because the next one leaked, then took so long to set that we gave up and took it out in its half-set state with a gooey exterior. #fail
We tried to dab dry the mold with a paper towel but realized that it should probably air dry as well, so we left it for the night and resolved to finish in the morning. The next morning we were met with the same leaking problem, attributed to the popsicle sticks not applying an even pressure to the mold halves. At least this time the piece set properly, although was a bit too short as some of its volume had leaked out. We transitioned to using blocks of wood on either side of the mold and it was smooth sailing from there on out. We had created a good system of x-acto-ing the finished pieces while the next was setting, so we just had to touch up a couple of polar bears before our project was complete!
In all, it was a neat project, working through the steps of editing the part in SolidWorks, 3D printing and milling the positives, molding the negatives, and then casting the final chess pieces themselves.
Cost: Materials: Wood blocks- $3.48 for 2 blocks from stock 2x4 3D prints- $3.10 Molding material- $9.99 for 9.6oz Casting material- $19.99 for 16oz @~2oz per piece x 11 pieces Labor: $7.5/hr @ 10hr x 2 people = $150 Overhead/Machinery: $5/hr @ 10 hrs = $50 Total = $236.56