How to Calculate Clamping Force Requirements for Multi-Cavity Silicone Compression Molds

Clamping force on a silicone compression press is not a thermoplastic tonnage factor, and it is not a copy of an LSR injection bar chart.

Compression molding of millable silicone (HCR) puts a preform in a hot cavity and squeezes. The press must keep the land closed while the compound flows, fills, and then expands as it heats and vulcanizes. The honest estimate is:

F = A × P × S

• A is the projected area that sees pressure, on the plane perpendicular to the ram (the shadow of every cavity, and of any runner, pot, or overflow that is still pressurized).
• P is the cavity (or land) pressure during the event you care about: fill and cure, not a catalog slogan.
• S is a safety margin for uneven load, gauge error, and the fact that the press rating is not the force at the land after friction and plate bend. Many shops use on the order of 10-25% (S ≈ 1.1-1.25). It is a margin, not a material property.

If you import “2-5 tons per square inch because that is what we use for nylon,” you will oversize a press, bruise aluminum lands, or still flash if the real pressure was elsewhere. If you import an LSR injection example that used up to about 400 bar of expansion pressure on a liquid shot, you are in the wrong process. LIM clamp is a cousin of this formula with a different P and a cold-runner area that may not see full pressure. This article is HCR compression, multi-cavity.

Projected area is a shadow, not a volume

CAD: look along the press axis. Sum the silhouette of every cavity. A tall boot with a small footprint needs less clamp than a thin, wide gasket of the same mass. That surprises people who think in grams.

Include:

• All cavities in the layout you will actually cut (8 × one cavity’s A, not “about one cavity”).
• Transfer pots and runners if this is transfer-compression and those volumes are still open to pressure.
• Overflow / flash lands that remain connected. If the land is a wide pressurized face, it is part of A. If it is a tiny break-off groove that dumps to atmosphere, do not pretend it is a second part.

Do not include the entire mold plate. Clamp on steel that is not seeing rubber pressure does not fight opening. It does bruise.

Deep cores add side pressure. Clamp force resists the opening component on the parting plane. Side load is a core-shift and plate-spread problem (different article). Do not double-count wall area into A unless that wall projects on the parting plane.

If you ignore a family of large overflow pockets still sitting on pressure, the defect is a press that looked big enough on a napkin and flashes every outer cavity.

Where P comes from (without inventing a constant)

P is measured or taken from a process the compound will actually run.

Sources that are real:

• A cavity-pressure sensor or a press force divided by a known A on a similar tool and compound.
• The specific molding pressure the rubber supplier publishes for that HCR in compression, used as a starting band, then confirmed on the press.
• Transfer: pot pressure is not cavity pressure. There is loss through the sprue. Using pot pressure as P is conservative on clamp and harsh on the tool.

Sources that are fake:

• A single “silicone is 150 kgf/cm²” rule for every durometer and every thickness.
• Thermoplastic clamp factors (tons/in² by resin family).
• Copying a Dow-style LIM example (projected part area × expansion pressure up to a few hundred bar) onto a compression biscuit.

Thick sections and high cavity temperatures raise expansion as the rubber heats. Thin gaskets may never see the peak P a thick bumper sees. Multi-cavity tools with mixed part thicknesses will see mixed P; the land still has to hold the worst cavity without opening the easy one.

Platinum HCR versus peroxide HCR changes chemistry and gas, not the formula. Vacuum compression changes whether air is in the way; it does not cancel P.

Multi-cavity layout is a clamp-quality problem

Tonnage is a scalar. The press delivers it through platens that bend.

• Balance the layout about the center of the ram. Eight cavities in a row on one side of a 200-ton press is a tilt. Outer lands open; inner lands crush.
• Cavity count versus land width. Tiny O-rings in a 64-cavity plate can have a tiny total A and still need a press that *covers the plate* and stays flat. Machine size is often a platen-size problem before it is a force problem. Tie-bar spacing and mold base footprint belong in the same sentence as F.
• Aluminum versus hardened steel lands. Extra S on a soft land is how you brinell a prototype tool and then chase flash that you created.
• Bump / vacuum sequences close at low force, then go to full force. Size F for the high-force step. Do not run the high-force step as a peening operation.

If you ignore platen deflection, the defect is flash on the corners and an over-compressed center that is out of ISO 3302-1 on thickness.

Worked structure (not a fake number)

1. Export A for one cavity from CAD (mm² or in², be consistent).
2. Multiply by cavity count. Add pressurized overflow/pot if applicable.
3. Choose P from measured or supplier process data for *this* compound and thickness class. Convert units so that F comes out in kN or tons the press understands (1 bar = 0.1 N/mm²; 1 N/mm² × 1 mm² = 1 N).
4. Multiply by S (state it: 1.15 is a choice you can defend, not a law).
5. Compare to the press rating and to the land stress (F distributed over the steel that actually touches). A huge F on a tiny land yields a pretty clamp number and a damaged tool.
6. Check layout symmetry and platen coverage.

Write the assumptions on the quote. A second shop with a different P will look cheaper or safer and is just using a different guess.

LSR injection clamp is the same algebra with a different A (often exclude a true cold-deck volume that is not in the parting pressure) and a different P (fill plus expansion of the liquid in a hot closed tool). Do not mix the P from LIM into an HCR compression RFQ.

RFQ notes so clamp arithmetic is visible

Process: HCR compression (or transfer), cavity count, CAD for projected area, compound family, whether vacuum bump is used, and any measured P from a sister tool. Ask the shop to show A, P, S, and land stress, not only “we will run it on 250 tons.” If they cannot show the arithmetic, they are matching platens by habit.

FAQ

Can I use thermoplastic tons-per-square-inch factors for silicone compression?

No. Those factors are empirically tied to plastic fill-and-pack pressures. HCR compression sees a different pressure history, including expansion during heat and cure. Use projected area times a pressure that belongs to this compound and tool, then add a stated safety margin.

Do I multiply projected area by cavity count only, or include overflow?

Include every area still sitting on pressure: cavities, transfer pots, runners, and overflow that has not dumped to atmosphere. Do not include the whole plate. Missing pressurized overflow under-sizes F. Including the entire shoe over-sizes F and can crush lands.

Is the 400 bar figure from LSR injection guides the right P for compression?

That order of pressure appears in some LIM processing notes as a cavity-expansion ceiling for liquid silicone. Compression HCR is not that process. Using 400 bar blindly on a biscuit tool will usually overstate clamp and still tell you nothing about land stress. Get P from the compression process you will run.

Why can a press with enough tons still flash a 16-cavity plate?

Tilt and bend. Force through an off-center layout or a thin plate opens the far lands. Flash is then a parallelism problem, not a missing 50 tons. Balance cavities around the ram and check land stress, not only the nameplate.

Does vacuum assist change the clamp calculation?

Vacuum removes air so the charge fills the well; it does not remove rubber pressure after close. Size F for the high-force cure step. The bump step is a lower force by design. If you stay at bump force through cure, you will flash with a perfect vacuum gauge.