Thin-Wall Silicone Molding: Pushing Manufacturing Limits Down to 0.2mm Thickness

0.2 mm is an edge, not a default wall. Some LSR grades will fill a short membrane at that thickness in a hot, vacuumed, well-gated steel tool. Many will not. A postcard-sized 0.2 mm panel from a single gate is a different request from a 0.2 mm web next to a 1 mm feeder.

An earlier shop note in this library used about 0.25 mm as a common LSR floor for short spans. 0.2 mm is the next millimetre-tenth people ask for on diaphragms, slit valves, keypad webs, and speaker gaskets. It is process-limited: grade, flow length, cavity heat, vacuum, and tool. Treat it as a capability trial on the real geometry, not as a catalog promise.

HCR compression does not live here. Millable rubber is the wrong viscosity. Platinum or tin RTV will copy a 0.2 mm flash and will struggle to *cast* a 0.2 mm designed wall unless you have a core, vacuum, and a very cooperative mix. This article is LSR injection.

Fill is a race against scorch

LSR is a liquid that shear-thins, then it starts to crosslink in the hot cavity. A thin section has little thermal mass. It heats through quickly. That is good for cycle time and bad if the front skins before the last millimetre fills.

What actually moves the 0.2 mm line:

• Flow length / thickness. A 5 mm run at 0.2 mm is not the same as a 40 mm run. Gate into a thicker feeder and let the membrane be the last, short branch when you can. Starving a membrane that sits *upstream* of a boss is a classic short.
• Viscosity of the grade. 10-30 Shore A LSR often runs thinner than a 70 Shore A filled grade. Color paste and self-bonding packages change viscosity. The Shore number is not the flow number; the rheology curve is.
• Cavity temperature. Hotter fills easier *and* scorches earlier. Thin walls want a window, not a slogan of “turn up the heat.”
• Injection rate and vacuum. LSR will trap air in a 0.2 mm dead end and call it a short. Vacuum plus vents measured in microns, not thermoplastic depths, are how you tell air from scorch.
• Gate size and vestige. A gate big enough to fill a membrane may leave a bump you cannot live with on a diaphragm. Valve gates help. They do not repeal physics.

If you ignore last-fill air, the defect is a lacy edge that looks like the grade “cannot do 0.2 mm” when the grade never saw a clear path.

Cure, tear, and the wall that survives demold

Filled is not finished.

A 0.2 mm LSR wall can cure in a few seconds and then tear on a pin, a cup, or a human thumbnail. Draft, polish versus texture, and ejection method are part of the thickness spec. Mirror cavities vacuum-lock thin membranes. Heavy texture on a 0.2 mm wall is a perforation pattern.

Demold while green and you stretch the membrane out of tolerance (ISO 3302-1 will not save you). Demold too late and some grades stick harder. Robotic cells need a membrane that leaves on a stripper or air without inverting. That is often a 0.3-0.5 mm conversation even when a 0.2 mm coupon filled on a plaque.

Use is the third limit. A 0.2 mm diaphragm that fills and ejects may still crease, pin-hole, or fail a burst test. Tear strength (ASTM D624) and elongation (ASTM D412) belong in the grade pick. A pretty 0.2 mm web in a 10 Shore A gel is not a 0.2 mm valve in 50 Shore A.

Tooling that makes 0.2 mm less of a stunt

• Steel, not a printed cavity. Aluminum tryouts can prove fill on a short coupon and then bruise on the shut-off.
• Uniform steel temperature at the membrane. A cold corner is a short. A hot spot is a scorch line.
• Vacuum. Non-negotiable on blind membranes.
• Flash control. The same low viscosity that fills 0.2 mm will leak 0.005 mm at the land. A membrane next to a leaking land is a flash farm.
• Measurement. Contact probes dent 0.2 mm rubber. Optical methods with a written edge rule, on parts that have cooled and, if specified, post-cured.

Do not specify 0.2 mm on a whole housing to “save material.” Save 0.2 mm for the feature that needs it, and feed it.

What to put on the drawing

A 0.2 mm wall that sits next to a shut-off is also a flash problem. The same low viscosity that filled the membrane will leak the land. Keep the thin feature away from the split when you can, or budget a flashless land next to it. A membrane that fills and then wears a film on every shot is not a thin-wall success.

Write the thin wall as a local callout with flow context: thickness, span, gate side, and whether vacuum is assumed. Put a realistic general wall (often 0.5 mm and up for LSR production) on the rest. If 0.2 mm is a hard product need, budget a fill study and a tear study, not only a mill quote.

Multi-cavity thin-wall tools fail in a specific way: the easy cavities fill and flash while the long-flow cavity shorts. Balance the cold runner mechanically (orifice and land), not by turning down a heater zone on the good cavities. A 0.2 mm family mold is a fill-balance problem first. If one cavity needs a unique gate, it may not belong in that family.

FDA 21 CFR 177.2600 and ISO 10993 do not create thin walls. They constrain which grades you may even try. Start with a grade that is legal *and* runny enough, then cut steel.

FAQ

Is 0.2 mm a guaranteed minimum wall for all LSR parts?

No. It is a local limit that depends on grade rheology, flow length, cavity heat, vacuum, and gating. Short webs next to a feeder can make it; large panels usually cannot. Quote 0.2 mm as a trial on the real span, not as a catalog wall.

Why did a 0.2 mm plaque fill but the part shorted?

The plaque was not your flow length, gate, or last-fill trap. Air in a dead-end membrane looks like a short. So does scorch halfway down a long 0.2 mm run. Copy the part’s path in the study, including vacuum, or the plaque is a souvenir.

Can HCR compression mold a 0.2 mm wall?

Not as a production default. Millable silicone does not flow like LSR. Thin HCR is a special, usually thicker than this. If the product must be 0.2 mm, you are in LSR injection territory, or you are redesigning.

Will a 0.2 mm membrane hold ISO 3302-1 tolerances?

Thickness tolerance on a 0.2 mm wall is a large *percentage* even when the millimetre number looks fine. Tooling, flash, and probe indent all move it. Agree optical measurement and a local tolerance, and do not expect a blanket class to make the membrane magic.

Does higher cavity temperature always help thin-wall fill?

It lowers viscosity and shortens the time before scorch. Past a point the front skins and you get a worse short plus a brittle edge. Map a window. Vacuum and a shorter flow path usually buy more than another 10 °C.