Displacement is the whole method
Silicone fills the space the part does not. That single sentence is the calculation, and every version of the mold-making problem reduces to it: work out the volume of the container up to the level the silicone will stand at, work out the volume of the part, subtract. What makes people get it wrong is that they measure the container and then guess the part, and the guess is usually far too small because an irregular object looks bigger than it is.
The measurement that removes the guess takes two minutes. Fill a container with enough water to submerge the part, mark the level, take the part out, note the level, and the difference is the part volume. A kitchen scale works better still if the part will not be harmed by water: weigh the container of water, push the part fully under while holding it on a wire, and the increase in reading in grams is the part volume in millilitres. If the part is porous, absorbent or water-soluble, bag it in thin plastic first and accept a small overstatement.
Failing that, use the bounding box method here and be honest about the fill fraction. A cast cube fills 100 percent of its box. A machined bracket might fill 60. A figure with outstretched arms fills perhaps 25 to 35, and people routinely enter 60 for one because they are picturing the box rather than the object.
Block molds and glove molds want different amounts
A block mold suspends the part in a box and fills everything around it. It is fast, it demolds cleanly on simple shapes, and it uses a great deal of silicone, because most of what you mix ends up as wall rather than as mold surface. A glove mold puts a shell of even thickness over the part and backs it with a rigid mother mold, which uses a fraction of the silicone for the same part and takes considerably more labour and skill.
| Approach | Silicone used | Where it suits |
|---|---|---|
| Block, part in a tight box | Box volume minus part | Small parts, few undercuts, one or two molds |
| Block, generous walls | Considerably more again | Deep undercuts needing a flexible wall to peel back |
| Glove, brushed shell | Roughly surface area times thickness | Large parts where silicone cost dominates |
| Glove, thickened poured shell | Between the two | Parts with a mother mold already planned |
The shell thickness on a glove mold is a genuine trade-off rather than a detail. Under about a quarter inch, a shell tears at thin sections and stretches enough to distort a casting. Much over half an inch and you have spent most of the saving that made a glove mold attractive. The figure this page produces for a glove is deliberately conservative, because it wraps the part's bounding box rather than following its surface, and running out halfway through brushing a shell wastes both the silicone already applied and the part preparation underneath it.
Wall thickness, and why the box should not be tight
A common instinct is to build the box as close to the part as possible to save silicone. Past a point that backfires. A mold wall thinner than about half an inch on a part of any size flexes when you pull a casting out, which means the mold changes shape slightly on every pull and the parts drift dimensionally. It also tears at the corners first. Half an inch of silicone all round is a reasonable floor for small work, and larger parts want proportionally more.
The top cover matters for a different reason. Silicone over the highest point of the part is what forms the mold's back face, and if that layer is thin the part prints through it and the mold sits unevenly on the bench. Give it enough that the back is flat and stiff.
Ratio, and the mistake that ruins a batch
Two-part silicones are sold at 1:1 and at 10:1 and occasionally at other ratios, and the number on the label is stated either by weight or by volume. These are not the same number. If part A and part B have different densities, a 1:1 volume ratio is not a 1:1 weight ratio, and mixing by the wrong one leaves you with an off-ratio batch that either never fully cures or cures brittle. Read which basis the data sheet specifies before the scale comes out. This page works in weight throughout, since a scale is more accurate and less messy than graduated cups, so convert a volume-stated ratio first.
Off-ratio silicone is not a problem you can fix by waiting. Where you have a genuinely ambiguous data sheet, mix a small test batch and let it run its full cure before committing to a mold you have spent a day preparing. If you are casting into the finished mold afterwards, the resin casting calculator handles the same weight-versus-volume question on the resin side, where it bites harder.
Questions people ask
How do I find the volume of a part with an odd shape?
Water displacement, and it is more accurate than any estimate you can make by eye. Put the part in a container of water, or on a wire under the surface, and read the rise. If you use a kitchen scale, the increase in grams when you push the part fully under equals the part volume in millilitres, which this page accepts directly. For a part that must stay dry, bag it in thin film and pull the film tight; the film adds a millilitre or two, which is inside the waste allowance. For a part that floats, hold it under with a thin rod rather than weighting it, so you are measuring the part and not the weight.
Should I mix silicone by weight or by volume?
By weight, with a scale that reads to a gram, unless the product specifically states a volume ratio and you have accurate graduated containers. Weight is repeatable, it accounts for whatever clings to the container, and it does not depend on how well you levelled a cup. The catch is that a ratio printed on the label may be stated on either basis, and converting between them requires knowing both component densities. A 1:1 by volume product whose components differ in density is not 1:1 by weight, and mixing it as though it were will leave part of the batch uncured. Confirm the basis from the data sheet rather than the marketing copy on the front of the container.
Why did my mold stay sticky in one spot?
On a platinum cure silicone, that is almost always cure inhibition rather than a mixing error, because a mixing error affects the whole batch evenly and inhibition is local to whatever the silicone touched. Sulphur-bearing modelling clays, latex, some adhesives and tapes, certain plasticisers in cheap plastics, and residues of uncured tin cure silicone will all stop a platinum product from curing where they contact it. Tin cure silicone is far less fussy about contamination, which is one reason people still use it for master molds over sculpting clay. If you cannot avoid the contaminant, test a coin-sized patch on the actual material first and let it run the full cure time before you commit.
How much extra should I mix?
The default here is 8 percent, which covers the film left in a mixing cup, what stays on the stir stick and spatula, and the shallow puddle you never manage to pour out. Vacuum degassing costs a little more than that because the batch expands and falls back and some of it climbs the container. Pouring into a tall narrow box wastes less than pouring into a wide shallow one. The number is not worth agonising over in one direction: silicone left in the cup is a few dollars, and a mold that is short at the top is a full remake plus another day.
Can I pour a second batch on top if I run short?
Sometimes, and it depends entirely on timing. Fresh silicone bonds to silicone that is still within its working window and has not yet skinned over, so a second batch mixed and poured immediately usually laminates without a visible line. Once the first pour has gelled, the bond is mechanical at best and the mold has a weak plane through it that will eventually split. If you know you may run short, mix the whole amount at once in a larger container rather than planning to catch up, and if you have already gelled the first layer, treat the mold as a two-piece job with a deliberate parting line rather than pretending the seam is not there.