By Alex Dumas © 2010

For my Swift project I wanted to use as most of the original wiggits (model kit parts) used to detail the studio model. The main problem I was facing though was that I was building a much smaller replica at 1/87th, than the studio Swift model which is estimated to be 1/32th. In other words, I was building a replica three times smaller than the original subject.

Flying-Sub hatch reduced to 30% of its initial size (after 7 cycles...)

The Shrink-Ray!

The solution was either to find the equivalent parts from smaller scale kits - assuming they have the same breakdown and details - or to reduce them to the needed scale;  Meaning reducing them 30% of their size, or a 70% reduction rate if you prefer the other way around.

The way to achieve this is by making a mold that shrinks. How to do that? Well, simply by adding a solvent to the silicon mix that will evaporate while curing. I use Smooth-On Mold Max 30 silicon and their technical support suggested mineral spirit as a compatible solvent (I use odourless mineral spirit). They didn't have any ratios to suggest so I tried a few to come up with this formula:

Silicon mix + 50% of mineral spirit (by weight) = 15% reduction (85% of subject size)

The solvent will cause the silicon mix to be a lot more fluid (less viscous) so be sure to have a well sealed form, otherwise silicon will leak before hardening. The curing time is also much longer. The mold will start solidified within the normal curing time BUT will shrink slowly over a few weeks period. Remove the mold from the pattern the day after to allow it to shrink freely afterward. You can jump start the shrinking process by puting the mold in the oven at 100F for 4 to 6 hours; You will smell the solvent evaporates so I really suggest you get odourless mineral spirit...

Some math

Reducing parts to 85% of their original size was obviously not enough for my needs so I had to do a few reducing cycles to get down to 30%:

Initial part size = 100%
After 1 cycle = 85%
After 2 cycles = 72%
After 3 cycles = 61%
After 4 cycles = 52%
And so on to 30%...

I have experimented with higher solvent ratios to found out that too much solvent seems to result in mould deformations (warping) once cured. Therefore, 50% of solvent in the mix of silicon will keep all details without deforming them.

The use of solvents may corrode your pattern so be sure to make a test first. So far, I have tried styrene, aluminium, resin and primed patterns and I had no problem.

Even though I haven't tried this technique with two parts molds yet, logic dictates that the same approach should work. Let me know if you try :o)

Reducing the Ultraprobe! (August 10, 2010)

I have used my reducing technique on James Small's command module pattern (to go on a studio scale Ultraprobe) and it worked very well. Although, the moulds were more massive and therefore took forever to fully cured (at least 2 months!) I adjusted the formula to get approximately 13% of reduction in one cycle. Here a few pictures of the result.

Return Home

Alex Dumas
scihighmodels.com