|Images and text copyright 2002, BNG Life Casting. Reproduction for educational purposes is permitted, provided that the content is reproduced in its entirety, including this notice. Reproduction for profit is expressly prohibited.|
This is a fairly small example of how to do a skin mold. The subject is
pretty small - easy to move around and quick to finish. As such,
it's not immediately obvious what all the advantages and disadvantages
are of this method, because when you work this small, even a really
inefficient process doesn't take too long.
The particular silicone for this demo is Polytek's. It's a platinum cured silicone, specificially, their PlatSil 73-29. The stuff on the site claims "tough, knotty tear resistance". Like h@ll. This may be better at not propagating tears than other stuff on the market, but in an absolute sense, it's pretty poor. It's really durable, as long as there are no rips, but once one gets started, look out.
This is a brief semi-scientific look at the statement above on tear strength.
Here, a very thin sample has been prepared. On one side, the silicone
has been left plain, and on the other, a swatch of Spandex has been
included. The test samples were prepared by spreading the silicone
evenly by hand with a spatula on the board, and then laying in the
cloth. No additional silicone was added; the cloth was allowed to
settle into the silicone on its own. Due to the viscosity, the
silicone is marginally thicker in the middle than at the edges.
The test setup is simple; the sample is suspended from one spring clip, and a second is attached to the bottom. The bottom clip is tied to a pitcher with a length of rope. The mass of the clip/rope/pitcher weight is 195g +/- 5g. Stress is added by pouring water from a measured volume of 1000 cc, and measurement is done by examining the level of water in the graduated cylinder after pouring, rather than measuring directly the water in the pitcher. This approach was taken as the water in the pitcher tends to splash out when the test sample fails.
This test was performed with the outer edge of the test sample (the edge
formed by the edge of the puddle) notched. This was done to eliminate
any false notches created by the cutting of the test sample from the
larger piece. This sample is the plain silicone, without cloth
reinforcement. The images above are:
This is another test on the plain silicone (no cloth). This time,
it has been pre-notched on the thicker side, near to the centerline
of the poured patch. As before, the piece starts to tear at a
stress of approximately 950g total.
Stress was increased to approximately 1700g, and the tear progressed, but turned away from the initial direction of the cut, preventing complete failure. When this piece was allowed to stabilize, a very small cut was initiated with a fresh razor blade in the already-torn region. As soon as the cut was made, the piece failed suddenly and tore completely across. The images above are:
Since it appeared at this point that cutting under stress was the
primary cause of dramatic failure, the third test was performed to
give a range of stresses between which sudden failure began.
The sample here was placed under 1200g of stress, then notched
with the razor blade. A small cut appeared, but did not progress.
Strain was then increased to 1700g. The cut widened, but did not
fail. Upon a further tiny cut, the test piece failed completely
The images above are:
A sample of the fabric-reinforced silicone was tested with the
spring-clip setup described above, but the clips were not
strong enough to hold the sample when the pitcher was filled
to capacity. The test setup was modified. In these pictures,
the test sample is held above by a screw clamp, and below in
a small vise, which is supporting a bucket. The mass of the
dry vise/bucket assembly is 2650g. As before, the sample piece
was notched to simulate normal damage due to wear and tear.
The test began with slow addition of water, to 5000g of water, for a total of roughly 7650g of stress. At this point, the fabric had just begun to yield. Upon cutting with the razor blade, the failure propagated slowly, and eventually stopped. A repeated small cut allowed the piece, now torn about halfway through, to fail completely. The images above are:
ViscosityThe viscosity on this particular formulation is pretty low, around 15,000 centipoise. Yeah, I know. What the heck is a centipoise. Well, centipoise is a measure of viscosity, and defined as the force experienced by a 1 square centimeter patch moving past another such patch at a distance of 1 centimeter. Put differently, it's a measure of shear stress, or a fluid's internal resistance to motion. Arbitrarily, water is chosen to have a viscosity of 1 centipoise, and everything else is calibrated to that. "Yes," I hear you cry, "but how do I apply that to Real Life?". Mee too. The same problem applies to the Shore hardness measurements. Let's take a moment to sort of clear things up. The scale, while useful, does not seem to me to be really linear. A couple of examples follow, which I mooched shamelessly from here. Web being what it is, though, I'd rather have a copy than trust that page will stay put. Plus, its the same as every other "common materials" viscosity list I've seen, so they probably stole it, too.
HardnessAnother measurement you'll see about molding rubbers is the hardness. This is nearly always on the "Shore A" scale, which is for softer plastics and rubbers. (Harder stuff is on the Shore "D" scale, which basically covers the range from very hard rubber to moderately soft plastics (Nylon, for instance))
The measurement is made by pressing a probe into the material under a known force, and measuring how far the probe got. That setup is known as a durometer, and so you also occasionally see "durometer" used as if it were a scale of units, the implication usually being that they mean the Shore A scale.
There's no direct physical meaning (other than that) behind the scale, which runs from 0 (complete penetration) to 100 (nearly none). A good treatment of the subject is available. To a greater or lesser degree, hardness trades off fairly evenly for flexibility. Thus if you want to mold something with significant undercuts, you'd use a softer rubber to allow that, but you'd be trading off for stiffness and strength, and thus mold durability in service.
This is a cast of Waynette, and she'll be our model for this demonstration. Notice that she's sitting on a brick to keep her up off of the newspaper. This is fairly important for a couple of reasons:
|Here, the backs of the pieces. Joel, on the left, has been finished a bit more thoroughly than Waynette. This will make for a somewhat easier job of cleanup when casts come out of that mold. Waynette's hair is already irregular at the back, though, and rather than introduce smooth curves, I'm just going to leave it like this.|
A brief, and probably pointless tutorial on how to use an analog
balance. Pointless, because I cannot recommend going and getting
an analog balance. Digital scales are cheaper, faster, and just
as good. I have one for plaster, but it reads in increments
too coarse to use for measuring the small amounts of silicone
for this project.
Leftmost, the pitcher is on the balance, and empty. The tare weight (the black cylinder in back) is used to account for the weight of the pitcher without using the weights that indicate the number of grams on the balance.
The second picture is after pouring the silicone base. Before pouring, the 100g weight was moved to the 100g position. Next, the base was poured into the pitcher until the balance went over 100g. The picture shows the actual amount as just about 119g.
The silicone in use here wants 10 parts of base to 1 part of catalyst. 119 grams of base need just under 12 grams of cat. To get this, I set the balance for 10 additional grams, and pour the catalyst. When the balance starts to move, I stop pouring. The scale takes a bit to settle. When it does, I set the scale for the additional 2 grams, and finish pouring very slowly. The silicone wants a very exact mix ratio to set properly, and this is what it takes with an analog balance.
|After mixing the silicone, lots of bubbles have been introduced. This is pretty out of focus, but you get the idea.|
To remove the bubbles, I put the mix in a vacuum chamber, and wait
until these big bubbles stop showing up. It'll keep bubbling
once the pressure gets down to the boiling point of one of the
components, but it's much more gentle.
The usual way of it is that the mix just rises as one piece, like bread. Just like a head on a glass of soda, though, the liquid is busy running back down between and around the bubbles. When the weight of the remaining liquid is low, and the skin on the bubbles is too thin, the mix all falls at once, rolling over on itself as it falls. A little more time under vacuum is in order to take care of air that was at the bottom of the mix before and didn't really rise much.
You can just see at the right where this mix has already risen up the sides. It's back down in this picture, but not quite ready to come out.
You probably don't actually need a vacuum chamber for this method of making a mold. It's just insurance. As long as you purchase a silicone that mixes up to a relatively thin consistency (thin enough that bubbles rise and burst on their own from an unthickened mix), you're fine. You may need to put on more than one thin coat before using chemicals to thicken the silicone for later layers, but that's about all the adjustment you should need.
The point of this first coat is to get the entire surface of
the piece covered with a thin layer of silicone. This ensures
that you don't have any air bubbles right at the surface of
the mold. The thinness of the layer ensures that you can
see what you're doing, and that you can see any bubbles in
the silicone and pop them. This first batch of silicone is not
thickened at all, which also really helps to ensure that there
are (all together now) no bubbles.
With this particular silicone (and silicones in general), there is no mold release required. The manufacturer suggests that you should seal the surface of plaster using vaseline. While you can, it doesn't seem to make a difference for TufStone. The answer may be different for more porous plaster, though. Note, however, that you cannot leave lots of vaseline. As mentioned in several other places, platinum silicone is finicky about curing, and too much Vaseline is one of the things it won't cure over. If you just leave the piece for a day or two or three, the Vaseline will wick into the plaster, and you're fine. Another alternative is to melt paraffin (mineral wax to you Brits), and brush this on. It helps a lot if the plaster piece is quite warm, but still comfortable to hold. If you get any ridges of wax, you can fix them with a hair dryer, and blot up with a paper towel.
As mentioned in the balance tutorial, this batch has 131 grams of silicone, and that turns out to be just enough to do three heads about this size.
Applying the silicone is pretty straightforward. I use an acid brush because it's disposable and has stiff bristles. This is a help in really pushing the silicone into crevices.
First shot is just starting the brush-on. Dip the brush into the silicone, rather than pouring on silicone and brushing out. If you pour on when it's this thin, you'll just get too much on, and it'll run off and be wasted. I generally start with the nose, mouth, and eyes to make sure I've really worked the silicone down into the important areas of detail.
Second picture is just after finishing the brush on. It's a bit uneven, but you can already see drips forming at the bottom.
Third shot is after a few more minutes. The silicone has evened itself out, and dripped off a bit more. You can control the wastage a bit better than this if you try, but at that point, the estimate you make in guessing at the batch size starts to become a significant component of the error.
It takes 8-12 hours for this coat to go from fluid, to gummy, to sticky, to tacky, to set. One advantage of platinum silicones is that you can accelerate the cure time with heat. Tin cured silicones generally don't set any faster with heat. If you go this route, a couple of things are worth being careful of.
This is to show how platinim-cured silicone does and does not
cure, depending on what it touches. On the left, a drip onto
newsprint. Next, that same drip peeled up and flipped over.
It takes the imprint of the paper (and some of the ink), but
doesn't leave any residue on the paper beyond a bit of a grease
The image on the far right is of another drip placed on a pseudo-latex glove. Peeling up the drip reveals a layer of uncured goo. This is a double whammy when it happens on a real piece. Not only do you have a section of mold that didn't take, but you also have quite a job getting the goo off so you can apply more silicone.
When dealing with uncured silicone on a plaster piece, generally, I cut very lightly around the uncured patch, leaving a fairly wide margin, and peel up the cured ring. This lets me get at the goo with paper towels without pulling up and displacing the good parts. This is generally good enough, and a second layer will go on and cure without trouble. This is only good for mystery cure failures, and little spots that you caused by touching the mold with gloves. Spots caused by remains of tin-cured silicones just won't cure until you get the tin-cured stuff completely out of that crevice.
Note that when you do this, it is very important not to pull up the
edges of the good layer. If you do, when you come back to brush on
the next layer, some of the silicone will force itself under the
bottom layer, and lift it. Since the silicone is pretty thick, it
won't completely flatten back out, and you get a really odd looking
defect on pieces cast from that mold. On the left is a different mold
where this happened, and on the right is what happens to plaster made
from this mold.
This is a general rule. I know it's tempting. I know it's fun. But don't pull up the rubber from the mold until the very end of the process. Especially tempting is picking at the drips, but this will peel it up, too, and if it's thin, it'll tear. The big reason not to pick at it is that if you do, it gets hard to make the mother mold. When the plaster goes on, it's heavy, and it sticks to the silicone a bit. If the silicone is loose, it'll pull away and hang down if the piece curves under at all. This will distort the mother mold.
This is the second coat. The point of this coat is to build
up a slightly thicker layer, again for bubble isolation. In layers
above this one, we don't care too much about tiny bubbles, as
they won't print through this layer. As with the first batch,
the second batch is vacuumed to eliminate bubbles. Some bubbles
inevitably get reintroduced by brushing, but this is just thin
enough that 99% of them rise and burst. The few that remain
are at the surface, and thin-skinned enough to easily pop
once cured. This mix is thick enough that bubbles
don't immedately rise, though, so not all of them are obvious
when you walk away from the brush-on.
This layer could equally well be built up of a number of thinner layers of silicone at the same consistency as the first layer, but that's not necessary. Since the first layer has assured us that there are no bubbles at all on the surface, we can mix the silicone up a bit more thickly. This is the same silicone as before, but a pigment has been added. The pigment is from PRS, and is used for coloring prosthetics. The cost is about $20 for a bottle, which should last you a long time. This was left over from producing a pair of feet in silicone for a client. When I was doing experiments for that project, I found out that the pigment thickened this brand of silicone. I have no idea if this is generally so, or if I just got lucky.
In addition to thickening the silicone, the pigment also has the effect of shortening the working time of the silicone mix. Certainly, on a silicone with a 12 hour cure time normally, this isn't a big deal. With Polytek's Gel-10, which sets in minutes, it is. You can counteract this tendency with their proprietary retarder, or thinning fluid, or both.
In addition to thickening the silicone, it also gives a really obvious clue about where the second coat is and isn't. You can sort of see this just by shine (cured silicone goes matte), but with the color, this is much easier.
The first picture illustrates how much thicker this stuff is than the regular silicone. This ribbon was poured directly out of the pitcher. The second picture illustrates how far this one ribbon goes when brushed out. The third shot is of the head completely covered by the second coat. You can see that there is much less silicone dripping off this time.
Note that we're also starting to see a trade off. The silicone is more viscous, and so it covers more thickly. On the other hand, the surface is a bit more ropy and un-even.
In addition to retarder and thinning fluid, Polytek also sells
a thixotroic agent. This thickents the silicone so that it will
stay put in thick layers. The more you add, the thicker it gets.
Only 1%-2% is needed.
Since we haven't yet done anything about putting a bit of a lip on the back of the mold, now is a good time. When putting plaster into the mold, this lip gives you a nice defined edge for the piece, as well as helping to shape the raw plaster to the right form when doing by-hand layup.
The second picture is just a dollop from a spatula to indicate how thick this mix really is.
Spreading the batch around with a spatula, the point of this coat is to fill in undercuts, uneven places, and especially to smooth out the eyes, nose, and lips for the last coat. Note that even with careful work, there are unavoidable ridges from the spatula.
This mix sets up fairly quickly. By the time all three heads had been worked on, the batch was just about stiff enough for smoothing with an acid brush. The brush is used back and forth over the whole surface to merge ridges back down.
Except in the areas in the center of the face, these coats are still fairly thin - from 1/16 to 1/8 of an inch thick. This is somewhat sturdy, but is extremely prone to tearing if any small cut starts. To help with tear resistance, we're going to embed some fibers. What you don't want to do is embed cloth that doesn't give. If the silicone is being pulled hard enough to stretch, but the reinforcement isn't giving at all, the rubber will start to tear around and through the rubber. Not all at once, mind you, but this does reduce the life of the mold a bit.
The first shot shows the same head, with another thin coat of unthickened silicone.
In the second shot, the fabric being used is shown. This is Spandex - pretty cheap per yard, comparatively speaking. Experiment has shown that this kind of silicone is OK with setting up around this material. I start by just laying a big chunk of fabric down doing my best to get all the way out to the edge.
Third shot is of the swatch wrapped around to the other side.
Foruth picture illustrates the difference between newly applied cloth and earlier layers. Use the acid brush to stipple the cloth down into and through the silicone already on the piece.
Fourth shot shows the difference between a newly-applied piece of cloth and that which has been stippled down.
Continue laying on cloth until the whole surface is covered. A bit of overlap is nice, but it's OK if you don't cover the whole surface. The major point is to ensure that there's fabric out to all the edges to stop rips fast. Secondarily, the cloth keeps the silicone from running off, and so you can build up significant thickness without adding pigment or thixotropic agent.
Once all the cloth is on, brush over a bit more silicone to seal the surface, and to ensure that the plaster from the mother mold can't grab into the cloth reinforcement.
Note that you shouldn't reverse the order of the above two steps. Around noses and eye sockets, there is enough spring in the cloth that it will stretch back, and "tent" itself over low spots. This leaves bubbles inside the mold, wich will either allow the face of the mold to sag in towards the bubble, or bulge the mold out when the air inside the bubble is heated due to the curing reaction of the plaster being cast. The solution to this is to fill in low spots with the thixotropic mixture first, before the cloth layer.
Now that we've got a finished silicone mold, we need a way to
keep it from flopping around while we pour plaster into it.
To do that, we'll put a rigid mold over the top of the silicone
mold. This is known as a "mother mold". The etymology is
a bit vague, but if you need to just remember the term, you
can think of the mother mold as cradling the actual mold.
There are all sorts of ways to make a mother mold, but I prefer
a fiberglass mold. USG makes
a plaster called FGR-95, which stands for FiberGlass-Reinforced.
This stuff is incredibly strong for its weight, and thin. Two layers is plenty (probably gross overkill), and it comes out to under 1/4 inch thick. The glass is simple to work with; it cuts with scissors, and is even easier to cut once it's wet. Yes, the plaster gets all over the scissors, but just wait for it to set, and bang them on something hard and it mostly all just pops off. The only downside, and it's significant, is the glass splinters. If you're like me, and you get those d@mn slivers everywhere, you'll want to double-glove. It pretty much eliminates them, and I've gotten used to the occasional splinter.
Basically, you just buy the fiberglass mat, and smear the mixed plaster into it. Be sure you get fiberglass that's intended for this - stuff that's meant for use with plastic resin has a different binder to hold the strands together. Once the glass gets wet, the water-based binder lets the individual strands move around, and you can get quite a bit of stretch out of the mat. Really odd shapes, or really round shapes, like heads, take a bit of smoothing out of wrinkles. Alternatively, you can just cut at the wrinkles, and lay it down overlapped.
In the pictures above, we see two layers of fiberglass precut. Next picture, pre-wetting the outside of the silicone mold with plaster, and lastly, wetting what will be the inside of the mat with plaster before laying it on. I've put some pigment in the plaster so it'll be obvious what's plaster and what's the mold.
|Laying up is simplicity itself. Just get the glass wet and push it into place. On the left is two layers of glass just after finishing the process. On the right, just after trimming the edges, and adding a bolt. More on the bolt later.|
This is the other side of the piece, after the plaster has set.
The second picture is just after breaking the loose plaster off
of the left side of the face. Since the fiberglass is so very
inflexible, it's important to be really aware of the shape of
the piece. If the mother mold goes even a bit more than
halfway around a round shape, or is trapping any undercut areas,
you've just sealed the plaster original into a nearly indestructible
Yes, if you miscalculate, you can get the shell off. When it's only just set, a sharp putty knife can cut it, or lift and break small areas without damaging the silicone. If you've waited more than a few hours, then your best bet is a hacksaw blade for ceramics. This is the kind that has some grains of carbide glued to the edge, rather than the kind with teeth. Both will cut through the shell, but the grit kind seems to do much less damage to the silicone underneath.
This is the first half of the mother mold pulled off of the silicone.
On the left, it's just as it came off. On the right, a wood rasp
was used to trim down the ragged edge. Once this is done, I also
use a propane torch. Run this slowly along the edge, and wash it
over the surface, and all the little glass fibers that are sticking
up out of the main mass (and are thus un-insulated) just wither
and melt in the flame. When the piece is still wet, most of the
excess heat just goes into driving off the water, and so the plaster
isn't much damaged by this treatment. Do it dry, and the plaster
will ping alarmingly, but nothing worse, in my experience.
The plastic tubing has been taken off of the bolt, and the area right around it has been sanded down to mostly-flat.
Before, I mentioned a bolt, and that we'd get back to it. The reason
for the bolt is fairly obvious: if you've got to make your mold in
sections, it'd be nice to have a way to hold it together. This is
even more relevant for this kind of mother mold. Since it doesn't
go all the way around, you can't just wrap everything in mold
straps to hold it together. My solution is this bolt arrangement.
It's just a carriage bolt and a washer. The plastic tubing is actually two tubes, one inside the other. This slides over the the bolt, and keeps the plaster from gumming up the threads. For this size mold, I'm only going to use one bolt. The irregularities in the mother mold do most of the work of keeping the halves registered together correctly; the bolt just keeps things from slipping apart.
Placing the bolt is very simple; you just lay on the bulk of the mold. Next, build a little ring out of the scraps cut off when the edge is trimmed. This hollow accepts the head of the bolt. Place the bolt, and then wet a couple of small squares of fiberglass. When it's wet, it's easy to poke a hole through the middle with a finger, or with the tubing on the bolt. Just push these down over the washer to hold it in place, and smooth everything out.
|Here, we're ready to put on the second half. Experience says that just Vaseline is not sufficient as a mold release to keep the halves apart. This picture shows some pallet wrap used as a separation barrier. Cling film for food works just as well; this is what I had on hand. Note that it gets tucked under the first half, rather than just laid on top. This keeps the plastic from slipping around, and lets the plaster for the second half make a perfect fit with the mold underneath.|
The second half of the mother mold goes on just like the first.
Back to white plaster for the contrast. Note also that the
tubing is back on the bolt. This continues to protect the threads
on the bolt, and it also anchors the plastic wrap.
Before, I mentioned that there are two layers of tubing, and you may have wondered whether this was overkill. It is, from the point of view of protecting the threads. The reason for the second layer of tubing is to make a bigger hole in the second half of the mold. The point here is that if you've got multiple bolts holding two pieces together, unless you're very careful (or lucky), the bolts are not going to point in exactly the same direction. If the holes you have to fit them through are small, it's hard to get that second piece back on. A bigger hole makes for easier fitting.
On some areas of some molds, though, you'll just have to break down and elongate the holes, either with a file, or with a hacksaw blade. Here, the two halves are finished. The second half has been pulled off, sanded smooth, and torched to eliminate wild hairs. A washer and a nut are added to hold the halves of the mother mold together.
Now, finally, you can peel the silicone off of the plaster master.
Work around the edges gently first. Since it's quite thin, the
mold is happy to reverse itself as it comes off. This inversion
also stretches out the features around the nose and mouth, which
makes it even easier to get these unmolded without tearing anything.
This is also the way to get the mold off of a fresh cast. Once the silicone is off, just leave it everted. It makes it much easier to fit back onto the original plaster master for storage.
I happen to store the master inside the mold. With this silicone, it shouldn't be necessary to do this to keep the mold formed correctly, but it does just about halve the storage requirements, and I'm cramped for space.
Finished! The halves of the mother mold are reassembled, and the
silicone mold has been placed inside. Next to it is the original
plaster that was taken from the alginate mold of the model.
This shot shows just how thin that first layer of silicone is; you can see the color of the second layer where the plaster original came to a sharp point and the first coat ran off of the comparatively steeper slope there.
|Lastly, a fresh cast taken from the new mold. You can tell which is the original and which is the copy; the original has yellowed a bit from absorbing something or other from the silicone.|
|Images and text copyright 2002, BNG Life Casting. Reproduction for educational purposes is permitted, provided that the content is reproduced in its entirety, including this notice. Reproduction for profit is expressly prohibited.|