Theatre Design & Technology - Summer 1999 - 51

structural reinforcement, and 2) the model fitting had very little detail and did not need the fine reproduction capability of the softer, more flexible silicone rubbers.
I first secured the model fitting to a base with the legs pointed up. Non-hardening modeling clay
formed beneath the central point and around each leg covered about halfway up the filling. (See figure
3.) Once the clay was smoothed out to create a flat surface around each leg, I created a narrow shelf of
clay around each leg of the model to create a keying surface that would eventually allow the two halves of
the mold to fit together exactly and prO\ide a seal between the halves. With a lauan plywood box built so it
extended about a half inch above the highest point of the clay-covered model, the mold was ready to pour
once the joints between the clay and the box were sealed with more clay and the entire assembly was
coated with spray silicone lubricant release agent. The silicone RTV rubber was hand-mixed in a plastic
cup according to the manufacturer's directions and slowly poured over the exposed half of the model filling to the top edge of the box. I had to take some care to prevent trapping air in the rubber, in fact the
manufacturer recommends that the filled mold be placed in a vacuum chanlber to "de-air" the material.
With the thick rubber I chose for this mold, de-airing was probably a good idea, but I didn't have a handy
vacuum chamber. As it turned out, air in the rubber mixture was not a significant problem, and only
some vely fine bubbles formed against tlle model that did not significantly alter the fit of the castings into
the PVC pipes.
The plywood box came off the first half of the mold after it had cured about 24 hours at room temperature. The model fitting remained embedded in the first half of the silicone rubber mold once the
base was removed and the clay carefully peeled away. It was inlportant not to remove the model from tlle
silicone mold as that could cause misalignment in tlle second half of the process. Cleaning the model and
tlle silicone mold witll mineral spirits removed any residual clay before the lauan plywood box was reconstructed around the silicone mold, with the exposed part of the model facing the open end of the box.
(See figure 4.) Aclay plug about a half inch in diameter and a half inch long was pressed onto the top
center of the model fitting formed the sprue hole before sealing, spraying, and pouring the second half of
the mold. (See figure 5.) Removing the box, separating the two halves, removing the model, then cleaning left a mold ready for the first casting. (See figure 6.)
Before casting each piece, the inside of the mold halves got a light coating of the release agent, and
tllen the two were slotted together and secured with large rubber bands. Acarefully measured mix of the
A-B urethane resin was poured into the mold through the sprue hole and the mold was then set aside to
cure. After the resin cured, the mold was separated and the casting removed, revealing a duplicate of the
original model fitting.

MASS PRODUCTION PROCEDURES AND NOTES
I was well aware that a single mold would not allow me to cast all the fittings I needed in the time allotted
for this production. I had planned to build five molds, calculating the amount of rubber needed down to
the last few ounces in a 9.9 pound kit of RTV, but one seanl in the lauan box for one mold split apart before the rubber cured, allO\ving the rubber to ooze out of the box. Without enough rubber left to pour
another half, I ended up with only four molds. Because the mold building process took much longer than
I anticipated, I realized the total casting time based on the manufacturer's published curing rate at room
temperature (2-{) hours) would be too long. Luckily, the resin would set much faster at 150° Fso I built
an oven using four R-40 lamps \\ired to a dimmer to control the heat output. The filled molds placed in
this oven set firm enough for de-molding in less than two hours. This way, we were able to get four castings from the four molds each work day (including one evening casting session). Ultimately we managed
to C<1St more than ISO fittings in about IO days. We also replaced the rubber bands used to secure the
mold halves with string tied \\ith shoe lace knots as they better able to \\ithstand the heat of the curing
oven. (See figure .)

ASSEMBLY
Each dodecahedron was assembled from one inch diameter PVC pipes and resin. Thirty PVC pipes, cut to
length at 12 3,4" long each created a dodecahedron with about a 3'-2" diameter. The fittings pushed into
the ends of the pipe held tight with a one-inch drywall screw drilled and screwed through the pipe and
into each leg of the fitting. This process repeated until the 30 pipes and 20 fittings were assembled into a
dodecahedron. The last few pipes were difficult to shove over the fittings because of the angles and stiffTll&T

MER

9 9

Table of Contents for the Digital Edition of Theatre Design & Technology - Summer 1999