How Thick Should an RTV Silicone Mold Wall and Bottom Be?

There is no single wall-thickness number that works for every RTV silicone mold. A 40 mm chocolate cavity, a 300 mm resin tray and a 900 mm sculpture skin can all use silicone successfully, but the structural job of the rubber is completely different in each case.

The correct design separates cavity wall, cavity-to-cavity web, bottom/base, outer border and any local reinforcement. Then it decides which loads the silicone must carry itself and which loads should be transferred into a tray, fixture or mother mold.

Stop treating every silicone section as "the wall"

A useful DFM review labels at least five zones:

1. Cavity skin - the detail surface around the product.
2. Web between cavities - carries repeated flex during multi-cavity demolding.
3. Bottom/base - controls loaded flatness and handling.
4. Outer border/flange - gives the operator something to hold and may carry keys or clamps.
5. Local reinforcement - around deep cuts, wick paths, plugs, cores or thin projections.

One nominal thickness cannot optimize all five.

What happens when the cavity wall is too thin

A thin silicone section can fail even when the raw material has good tear strength.

Typical failure mechanisms:

• local stretch concentrates at the thinnest web;
• the casting shape prints through and distorts the outside wall;
• sharp undercuts pull the wall into a notch;
• repeated peel turns a thin edge into a crack starter;
• a filled cavity bulges because the wall has insufficient section;
• keys or plug pockets punch through into neighboring geometry.

The correct fix may be local thickening, more spacing, a radius, lower demold strain or external support-not simply a harder silicone.

What happens when the mold is too thick

More silicone is not automatically safer.

Over-thickening can create:

• unnecessary weight;
• higher material cost;
• harder handling for large molds;
• slower thermal response for food or wax processes;
• reduced ability to peel around undercuts;
• a false sense of stiffness while the entire unsupported span still sags.

A very thick unsupported tray can still bow. A thin supported skin can hold geometry more accurately if the load is carried by a rigid jacket.

Bottom thickness should be designed around loaded behavior

An empty mold on a table tells you very little about how it behaves when filled.

For a large multi-cavity tray, check:

• total filled mass;
• span between supports;
• where the mold is carried;
• whether it sits in a nest or rigid tray;
• whether the casting material is hot;
• whether the base must stay flat for a depositor or scraper.

A gummy tray on a machine may need a very different base strategy from a hand-poured soap mold even if both have similar outside dimensions.

Large molds should transfer load into the mother mold

For sculptures, concrete, plaster or other heavy wet casts, the RTV layer should primarily reproduce detail and release the casting. A rigid mother mold or support shell should preserve global shape.

This changes the wall-thickness decision substantially.

Instead of pouring a massive silicone block, a better system may be:

• controlled silicone skin thickness;
• local reinforcement at tear-prone features;
• registration keys between skin and jacket;
• multi-piece FRP shell;
• bolted flanges carrying the hydrostatic load.

The support shell is not optional if the flexible skin would otherwise move under the casting weight.

Cavity spacing and wall thickness must be reviewed together

On multi-cavity molds, the gap between cavities determines how much structural silicone remains after both cavity profiles are accounted for.

A CAD layout can look spacious from the top while the side walls flare underneath and leave a very thin web at depth.

Always inspect the minimum section through the deepest region, not only the surface gap.

Wall thickness also changes demolding force

A thick wall is more resistant to bending. That helps shape retention but increases the force needed to peel it around a deep undercut.

For undercut parts, the design may deliberately use:

• a thinner flexible skin in the release zone;
• a thicker flange for handling;
• rigid support while filling;
• local plugs or split sections where the rubber should not stretch.

This is structural tuning, not a single-number specification.

How to prototype the weakest zone

The prototype should target the likely failure point:

• deepest undercut;
• narrowest cavity web;
• corner where the operator starts demolding;
• loaded edge of a large tray;
• plug/key root;
• unsupported wall of a vessel mold.

If the full mold is expensive, a local geometry coupon can sometimes validate tear and release before the whole master is committed.

What should change on the existing RUUIPON page

The current wall-thickness article already owns this search intent, so it should be rewritten in place, not duplicated. Existing fixed ranges and application-wide claims should be reframed as examples or removed unless tied to a specific design/material/test condition.

The revised page should make the central point clear: wall thickness is a load path + demold + support decision, not a universal 2.5-5 mm rule.