Combining two designs for large chess piece: Athena bust

Athena bust: https://www.thingiverse.com/thing:1567839

Pedestal: https://www.thingiverse.com/thing:2721636

For the large chess piece, we decided to combine two designs we pulled from Thingiverse: A bust of the Greek goddess Athena and a pedestal. Through lots of time spent troubleshooting and iterating different designs, we were able to create a chess piece which we were proud of.

Meshmixer → SolidWorks → Fusion 360

After finding the two files on thingiverse, we imported them into Meshmixer to begin work on them. We scaled the two parts individually to find a size for each which made sense. From there, we combined the two parts into a single solid and scaled the piece down uniformly to a size which would be appropriate for both the 3D printer and the Carvey. We also made sure to reduce the amount of triangles in the stl from a quantity of >250,000 to <25,000 in order to cope with the Carvey’s restrictions. Following this, we split the piece into two separate parts vertically: the front, more detailed face and much less detailed back. As the front piece was much more detailed than the back, we decided to have that be the piece 3D printed and have the back piece be cut with the CNC machine. Once again with the Carvey’s restrictions in mind, we did not split the pieces evenly. We made sure the back part was only ⅓ of the piece as the Carvey would not be able to cut the required depth if we made it any bigger.

After creating our two separate stl files for the front and the back parts of the Athena chess piece, we moved on to SolidWorks. We imported both files into SolidWorks separately in order to create the frame with pegs for the mold. After this step, the front part was finished and ready to be 3D printed, but the back part still needed to be converted to gcode in order to be cut by the Carvey.

To convert the back part into a gcode file, we needed to import it into Fusion 360. After closely following the tutorial shown in the Midterm assignment PDF, we were able to create two separate gcode files: the rough cut gcode and the fine cut gcode. With this, all the preparations were finished and we were finally ready to begin creating our Athena piece.

3D Printing: Reducing layer height for better detail

We used FFF 3D printing to print our front part of Athena (the side including her face). At a layer height of 0.2 mm, the print included details of the hair and facial features, however, the filament layers were clearly distinguishable and rough. We moved to SLA printing to get finer detail, but our SLA print failed twice. Returning to FFF, we reduced the layer height to 0.05 mm (¼ factor of original), which increased the print time from 4 hours to 16 hours. The result was our optimally 3D printed piece with indistinguishable layers and detailed features.

CNC: Detailed rough cut and stopped fine cut

We imported two gcodes to Easel: the rough cut and the smooth cut. Because we couldn’t use the fan for fear of the fan hitting parts of our piece, we had to vacuum the Carvey every 5-10 minutes. The rough cut resulted in a piece that was unexpectedly detailed. We switched out the bit for the fine cut, and for some reason, the toolpath made the bit carve a hole into our piece’s peg. We stopped the CNC machine to minimize damage and decided that the rough cut was detailed enough. However, we didn’t realize until the molding process that the lack of a fine cut changed the size of our piece.

Video clip of rough Cut: IMG_7010

Video clip of fine cut attempt: IMG_7011

Molding and Casting: Misalignment of CNC and 3D printed part

After setting up cardboard boxes around our 3D printed piece and CNC piece, we mixed the two parts of the silicon rubber in 1:1 by weight and filled the boxes so that the silicon was about ⅛ inch above our piece. Once the mold hardened, we combined the two molds together and poured the plastic through the hole the base of our piece made. 

In our initial plastic pieces, the CNC part and 3D printed part weren’t aligned. We realized that the two parts were different sizes due to the CNC part missing the fine cut. Trying to align the two parts better, we removed the pegs in our mold. Perfect alignment was impossible, but aligning the two molds by eye without the pegs achieved a better result.

The nose of Athena, the corner of the bust, and her back were particularly susceptible to bubbles despite us tapping the mold on the table. 

Post-Processing: Dremel to minimize misalignment

To minimize the misalignment of the two parts of Athena, we used a dremel to sand down the sides. The two parts blended better with each other and looked more unified. 

Cost Analysis

Labor: 2 people x ($7.25/hr / person) x 40 hours = $580

3D Printing Machine Time: Prusa k3 MK3S = $750. Lifetime = 800 hours (~$1/hr) x 30 hours = $30 

Carvey Machine Time: ($2,500 machine price/20,000 hr machine life) x 3 hours = $0.38

Filament: $0.81

Wood: ($10.92/16 ft) * (1 ft) = $0.68

Silicon Rubber: ($170/16 lb) * (0.67 lb) ~ $7.12

Liquid Plastic: ($132.88/15.2 lb) * (0.21 lb) ~ $1.84

 

Total Cost: $620.83 for 6 chess pieces → $103.47 for each chess piece

 

Conclusion: 

After much trial and error, and practice with Solidworks and Fusion 360 to prepare the files, we were able to use different 3D prototyping processes to create a unified chess piece of Athena that compared the abilities of CNC and 3D printing.