Lavender | HTM(A)A

Molding and Casting

Design

During Make Something Big week, I used sheet material to generate objects with solid volume. In an attempt to do the inverse of this, this week I decided to use the casting process to produce a surface-like object. I wanted the cast material to have some memory for the coiled shape, giving the lanyard some tendency to contract when not loaded with any weight (keys).

3d Model of Mold Positive: the geometry unrwaps into a single closed loop of material.

Milling the Mold Positive

To do this, I coiled a lanyard-length line into the bounding box of the cast, leaving a minimum spacing to accommodate the width of the bits in the milling process. When de-molded, the geometry should produce a single closed loop that I can clip to my keys and wear around my neck. The entire mold has a draft angle applied, since I was worried about removing the thin material from the mold at the end.

The mold layout fits the length of the lanyard into the bounding box of the wax.

The coiled design uses the near-minimum spacing between the edges to accommodate the width of the 1/8in bits.

Because of the draft angle, there was some discussion about that the best tradeoff between milling time, surface finish, and geometric resolution. Initially, the final tool-path (which maintained the steep draft angle nicely) was extremely dense, resulting in an extremely long milling time.

Too-dense scallop toolpath for milling the draft angle into the sides of the extruded surface

As an alternative, it would have been possible to mill the geometry with a tapered bit; however, the angle of the taper was about 2 degrees greater than what I had originally modeled. To use this method, I would have had to add more space between the surface folds in my geometry, and the tape would be disproportionately thick on one side. Eventually, it made sense to settle for a courser surface finish to save time and maintain the original model geometry.

One possible solution: use a tapered end mill. However, this would have required a change to my original geometry, and the three-degree draft angle would have diminished the tapiness of the cast

Finally, it seemed to be a better decision to decrease the stepover of the original scallop toolpath, which resulted in a somewhat more reasonable milling time.

In MasterCam, the milling process used three tool paths:

Rough cut: 1/4in flat end mill
Contour: 1/8in flat end mill
Scallop: 1/8in ball end mill

First rough cut, starting on a contour cut on the inner edge of the shape

Third toolpath completed with draft angle

Casting the Mold Negative

Since I didn’t leave any edges around the boundary of the wax positive, I built a chipboard box around the wax, using a strip of adhesive-backed foam around the edge to seal the seam between the board and the wax to prevent leaking. I also fully taped around the bottom of the mold, just in case any oomoo wanted to leak out of the edge. I also applied mold release before pouring the oomoo.

Edge lining and chipboard bounding box for casting the oomoo

I taped all the way around the box to prevent leaks

I was worried about demolding, so I used mold release

Demold successful, but with a few air bubbles.

Casting the Object

I decided to use a relatively hard polyurethane rubber (Smooth-On PMC 746) for the final cast so that the result would be a flexible tape. The pigment is supposed to be green, but I added too much so it looks black.

...but easy to peel away when it was partially cured

Demolding