Checkmate: Taking “Make a chess piece” literally

Introduction

We have now arrived at the end of the course, and with it our final project! This time we were encouraged to work with other partners so Santi, Baris and I teamed up for it. This project was simpler than the midterm, and taking into account final presentations and projects from other courses, we decided not to complicate the chess piece too much and make something we know will have fewer challenges for fabricating it.

Possible pieces to make

We decided to go with designs from the World Chess Championship (WCC) and settled on  the Bishop as it was the biggest piece that didn’t have complex overhangs when laid horizontally, which would be tricky or even outright impossible with the CNC machines at the OEDK. First we reduced the resolution of the STL file by reducing the triangle vertices on MeshMixer so that we could work on it on SolidWorks, then once in SW we used one of the planes to cut the part through the middle and start working on the bases for the positives of the molds.

3D Printing

First version of the base, with the recommended small pegs

With the base STL ready, we tried going with our 3D printed positive for the mold; here we found out that the minimum required size for the pegs and holes might be too small and that printing the piece vertically may leave some artifacting from removing the skirt on the underside of the print. We proceeded to change the SolidWorks parts accordingly with pegs of 10mm diameter and 3mm height. With this final version we then printed our final version of the 3D print positive and updated the model of the CNC positive.

CNC positive

For the CNC segment of the project, we used the section adjacent to our bishop to create an STL file, facilitating the construction of the toolpath for our CNC cut. We transferred our file to one of the OEDK computers and uploaded it into VCarve Pro software. In VCarve, we input the dimensions of the wood selected for our CNC mold. After uploading the STL file, we placed it in the left corner of the wood. We then accessed the toolpath tab in the software and started adjusting the rough machining toolpath. For tool selection, we downloaded the tool file from Canvas, choosing the 0.25-inch end mill. This procedure was replicated for the finishing toolpath, which resulted in an end product that allowed us to visualize and verify the accuracy of the bishop’s cut in the wood. After confirming the precision, we transferred the toolpath from VCarve to the CNC machine’s cutting software. 

We commenced by attaching our wood piece to a larger one, ensuring both were firmly clamped to the machine, preventing movement on any axis. Next, we executed the job zero process by placing the probe in our wood’s corner and guiding the machine towards it to set the machine’s zero position. After completing the probing, we switched from the probing tool to the 0.25-inch tool and initiated the cut. Our first attempt was unsuccessful; the machine struggled to cut the wood smoothly, producing unusually loud noises. Following Dr. Wettergreen’s advice, we replaced the wood with a more suitable piece of wood for the CNC machine. After verifying the accuracy of the toolpath values, we probed and re-ran the cut, which resolved the smooth-cutting issue. However, we encountered another problem: the holes in our design weren’t cut through. We re-examined the toolpath but couldn’t identify any missed steps that would explain the incomplete cuts. Deciding to reattempt the cut, we increased the speed by 200% since most of the design was already cut, and we only needed to finish some details. Ultimately, our final cut was successful, and our design was fully executed as expected, although the reason for the initial failure to cut the holes remains unclear.

 

Molding 

We initiated our project by preparing negative molds from silicone. The process for creating the positive counterparts involved both CNC and 3D printing techniques, which proved similar in execution. Mindful of the cost of silicone, we aimed for precision in our measurements. For instance, the chess piece, approximately 1 oz in volume, required a mold of about 0.5 oz. Calculating the total volume of the mold at 266 cm³ (roughly 9 oz) was a meticulous process, involving the dimensions of a cardboard dam encasing the positive piece. Subtracting the volume of the chess piece from the total mold size, we determined the exact quantity of silicone mix required – 8.5 oz.

For the silicone mold kit, a 1:1 mix ratio of ingredients A and B was essential. We mixed them carefully, avoiding air bubbles, for 1 minute and 30 seconds. The mix was then poured over the positive shape and left to set at room temperature overnight.

Once cured, we dismantled the cardboard dams and extracted the molds. This procedure was replicated for both the CNC machined and 3D printed pieces, resulting in two molds for each side of the chess piece.

Casting

Casting was an enjoyable stage. We aligned the silicone mold halves, securing them with rubber bands and flanking them with wooden planks for even support. The volume of the chess piece was precisely 0.7 oz, dictating equal parts of 0.35 oz for both Part A and Part B of the plastic resin. After a thorough mix for 1 minute and 30 seconds, we poured the resin into the mold. The exothermic reaction between the resin parts was fascinating, yielding warm pieces upon extraction. Our main challenge was air bubbles trapped in the bishop’s neck, causing defects. After several trials, we discovered a solution: stirring the resin within the mold to expel the air. Regrettably, this realization came late, leaving only two pieces bubble-free in their necks.

Post Processing

After casting we removed the excess material that accumulated in the conjunction sites of the molds using exoknifes. After removing the main defects we sanded the pieces using 100 and 200 grids.

We are mostly happy how the pieces turned out. They look professional and are able to show main differences created in the surface between CNC machining and 3D printing. Overall, it was a valuable and fun learning experience.

 

Cost Analysis

Cost Price Amount Source Quantity Unit TOTAL
White PLA 1kg spool  $      18.99 1000 Amazon.com 272.5 g $5.17
3D Printer Technician (3D print setup and post processing)  $      21.00 per hour Ziprecruiter.com 1 hours $21.00
3D Printer time  $        0.21 per hour Prusa3d.com 2.5 hours $0.53
1″x3″x8ft. Pine  $        2.54 1 Lowes.com 2 logs $5.08
CNC operator  $      24.00 per hour Ziprecruiter.com 1.5 hours $36.00
CNC machine energy  $        0.28 per hour circlemwoodworking.com 1.5 hours $0.42
Silicone Molding Kit  $      29.99 46 Amazon.com 17 us oz  $   11.08
 Plastic Casting Resing  $      57.99 128 Specialtyresin.com 8 us oz  $      3.62
Prototyping Engineer  $      34.00 per hour Ziprecruiter.com 1 hours  $   34.00
Quality Control  $      18.00 per hour Ziprecruiter.com 1.5 hours $27.00
Facility & tooling cost  $    250.00 per month txrx.org 6 hours $2.08
NET TOTAL $145.99
Print Friendly, PDF & Email