Project: Making a Chess Piece
By: Anyssa Castorina & Ruben Marroquin
Choosing a Design
Things we had to consider when choosing a design were size, simplicity, overhangs, and resemblance to a chess pawn.
From these things, we knew we wanted the design to have a sort of base at the bottom, as well as have the height be longer than the width. We also looked for at least one axis of symmetry to make it easier to align the molds and recreate the piece for 3D printing and CNC-ing. We wanted to make something we could be proud of and keep for afterward and we settled in a rocket ship pawn even though it was a more complex shape.
- https://www.thingiverse.com/thing:6866
- https://www.thingiverse.com/thing:2053862
After finding a suitable STL for which to base it off of, we then filled in any impossible overhangs with a solid piece that would eliminate such imperfections. We used a combination of Solidworks and Tinkercad to post-process the cad file and started the process.
Later, after casting the mold for the 3D positive, we noticed the wing seemed a bit thin and found the second shuttle chess piece that had a thicker wing. We later used this design for the CNC positive for the mold after the first mold from the CNC created a wing much too thin to survive much pressure without snapping.
Full Object Print
Before finalizing the design we were using, we 3D printed a model to make sure we were fully considering the sizing of the piece and the overhangs. After the STL file we had was cleaned up, we printed a very large version of the pawn to examine the design very closely and evaluate the feasibility of the design again.
Printing a 3D Model
Initially, to print the first file, we had to convert the design file we got from Thingiverse into an STL file so we could use 3DOS to slice and print it. After converting it into an STL, we went in and cleaned up any overhangs or awkward indents that may have caused issues while molding or casting. Next, we had to add holes and pegs to the plane the positive half sat on so that we could align the two molds and make sure they held in place while pouring the mold and while it sat to cure.
CNC-ing a Model
After cooking up the model’s STL ready for 3d printing and CNCing, we started the process of converting the STL into a G-code that the Carvey could read. Placing it in Fusion 360, we adjusted the parameters to a piece of 2×4 wood block we found and adjusted all the settings accordingly. We then exported it into Carvey’s software, easel, and began the cutting process. The First CNC took about 2 hours to complete and the second took about 90 minutes to complete. We had to redo the CNC as we failed to mirror it at first since our pawn was not radially symmetrical. While the Carvey was carving, we stopped it every ~10 minutes to vacuum the wood dust out and prevent the Carvey from getting stuck with too much sawdust.
Making a Mold
To make the mold of the chess piece, we created a temporary box around each of the fabricated positives. After this we poured a 1:1 ratio of the silicone part A and B into a cup, mixed thoroughly, then filled the temporary box to create our mold.
Casting
For casting, we used urethane to fill the mold. Because of the complicated geometry that had thin places coming far off the main part, we had to stretch the mold and pour the plastic mixture in before closing the mold and filling in the main part. Another problem we had to solve was that our pegs mated with the other pegs, and the holes mated with the other holes because we mirrored the design to get the CNC half, without considering the change needed for the hole and peg. To fix this we cut off one of the pegs, and put some glue in one of the holes to lift the bottom up, then we glued the peg into the hole where it had enough length to properly mate with the other hole. On the side that then had two flat surfaces mating, we put a needle in one of the planes and when it was time to mold, we pushed the needle through the other mold so there was a skewer between the molds, keeping them aligned and together.
1st Attempt:
For our first attempt, we used both parts of the plastic, started mixing it, and poured it after a bit of mixing. It was a successful cast, however, the cast was more translucent than regular casts. This may have been due to our mixing process or by our ratio of part A to part B of the liquid plastic.
2nd Attempt:
Our second attempt was also successful, we used rubber bands to hold the two pieces of the mold together, but found that there was a bit of liquid plastic that spilled over from the pour. Still, the two pieces lined up nicely and we were ready to experiment with using tape to hold the two pieces together.
3rd Attempt:
To experiment with the holding of the mold, we tried one mold where we did not use any mechanism to hold the two pieces together, and naturally, this failed drastically.
4th Attempt:
After the third attempt, we were now using painter’s tape to hold our molds together, this worked well and we came up with another successful print.
5th Attempt:
Now that we had the casting down, we started to experiment with different colored dyes. Here we used too much dye and the rocket became oversaturated with indigo.
6th Attempt:
In one of our attempts to use dye, we found our plastic started drying up more quickly than we had anticipated. Thus we were not able to fill the mold and decided to let the rocket rest on the puck that was leftover from one of the next pours.
7th Attempt:
Our next attempt had us changing the pour method. This time we poured with a makeshift funnel and spun around the mold to make sure the wings were well coated with plastic, and as a result, this print came out with all the wings in the best shape so far.
8th Attempt:
This next attempt was successful, with a bit of post-processing. When we lined up our mold, one of our alignment pegs became unstuck and as a result, the entire cast was slightly shifted. Some light sanding evened up the edges and the end result was successful with all the wings also well constructed.
9th Attempt:
For our tenth attempt, we tried mixing more dyes again and we realigned the alignment peg. The alignment of this one was much better, but one of our wings was not well coated.
10th Attempt:
Our tenth attempt was one of my favorites. We used a feature of the self-healing nature to mix a batch of liquid plastic with a dye added, then mixed another batch with less dye. As a result, our rocket looks two-toned, similar to the planes in the WWII era. One of my favorite pieces!
11th Attempt:
For one of our last attempts, we went for a tie-dye type aesthetic. We used the same cup to mix the plastic as previously and so there was some dye still on there. This resulted in a cool mix of green and blue and a hint of purple in the cast.
12th Attempt:
For our second to last attempt, we wanted to try out more of the blue rocket. We put a bit too much ink however at first, but it gave it a nice worn look to it. Another successful cast.
13th Attempt:
For our last attempt, we went back to the basics and came out with all that we learned. We coated all the wings, lined up the pieces nicely, and held it together with tape. As a result, we cast one final, successful cast.
Post Processing
For post-processing, the most important thing to clean up seemed to be any signs of misalignment as well as the spout from which the liquid plastic was poured. Where the mold was misaligned when we were casting pieces showed ridges along the axis of symmetry. To fix this, we carved away imperfections with a utility knife to even the piece out, and then sanded the piece to try and reduce the edges showing from the layer height of the 3D printer and the CNC.
Costs
3D Prints:
$2.08+$0.79+$0.23+$0.44+$0.24+$0.31+$0.34= $4.43
CNC Carvings:
Scrap 2”x4” = Free
Silicone Mold:
assuming about $26 for 300g of part A and part B solution, we used about 70 grams of part A and part B solution 3 times for a total of 210 grams of each solution. Then (210g/300g)*$26 = $18.20
Plastic Castings:
(about 25 grams of part A and 25 grams of part B solution) * 13 castings = 650 Grams. If 1 gal of both is equal to about $66.49 as per amazon listing, then 650 grams / 3785.41 grams * $66.59 = $11.33
