The fact that the final project required a negative mold was a blessing in disguise. Initially, when faced with the task of cutting the seal, the overlapping paths from the seal’s initial design process proved to be a bit of an annoyance. However, since our plan was to cut every black part of the seal to the same depth, any overlap would not matter. This simplified the process of preparing the SVG file to simply flipping it and scaling it up. Since the CNC cutter’s scaling issues caused every part to be 1.28 times the size of the design specification and we were aiming for a final size of about 10″ x 10″, we imported the seal into Easel and scaled it to 7.5″ x 7.5″. As we learned in the initial CNC assignment, the scaling factor also effects the bit stepover, so we scaled the bit size down in Easel as well. The smallest available mill was 1/16″ (0.0625″), so we specified a size of  0.047885″ in the program. We then chose the types of cuts for every part of the design. The letters and other design parts were specified as fill cuts, and the circles, lines, and owls were specified as “on-path”

 

 

After initially attempting to cut the entire seal mold out of MDF, we determined that the most difficult area would be the owls. Even with the seal scaled up to the maximum size allowed by the CNC cutting table, the owls have quite a few small details that consistently sheared off or shredded when cut into MDF. To solve this, we started a series of test molds focusing just on the owl. Plastic-based materials proved much more successful. We started with foam, then moved to acrylic, then wax. While they all produced good, clean results, the foam melted when exposed to the high heat of the 300Q casting material. Of the materials tested, wax worked the best, producing good results both with 300Q and plaster (though the thin plaster cracked when removed from the mold):

 

 

Moving on to the full seal, we ordered a 12″ x 12 ” block of machinable wax and cut the full seal. We were pleased with the clean result, though the block was not entirely level across its surface as delivered. This meant that the top face sat at an angle relative to the cutter, and one side of the mold was cut significantly deeper than the other side (See below). To finalize the mold, we clamped a set of acrylic “risers” to allow the casting material to form a circular base below the cast design.

Our first casting attempt in the new wax mold was in plaster. We created a clay wall on top of the risers to allow the base to be thicker and stronger, wary of the failure of a thin base in our initial testing above. The first cast came out surprisingly well, but the plaster broke in several important locations. The clarity of the edges did, however, boost our confidence going into the casting with 300Q.

The strength and relatively high flexibility of 300Q when compared to plaster made it much easier to remove from the mold. We then set out to improve the contrast on our seal, so we spray painted the entire seal and then sanded off the raised sections. Initially, we made the mistake of not using a sanding block, so the sandpaper slipped into the cracks and led to a sloppy result. We made sure to use one in the second attempt and it came out very nicely:

The process of getting to this one final product left us with this rough trail of prototypes and tests (with many more that couldn’t fit in the picture):