Liquid Silicone Rubber (LSR) Design Considerations

Liquid silicone rubber (LSR) occupies a peculiar niche in injection molding: a cold barrel and a hot mold. The material cures in thirty seconds, making almost all experience with ABS or PP directly inapplicable. Treat LSR like a thermoplastic, and you're likely to scrap a USD 25,000–60,000 mold due to flash, sticking, or parting line blow-outs before you get a truly sealed first article. The truly critical design decisions for LSR happen in CAD, not on the mold shop floor. Hardness selection, wall thickness uniformity, parting line placement, under-cut geometry, and the trade-offs between cold runners and conventional gates—these are all locked in before the first cavity is cut. This article will dissect these decisions one by one, explaining why some parts can be consistently mass-produced with a 40-second cycle time and 99% yield, while others constantly suffer from sticking, flashing, and tearing during ejection.

What LSR is from a process perspective

LSR is a two-component, platinum-catalyzed silicone supplied as pumpable A and B liquids. They meet in a static mixer, are metered into a cold barrel at approximately 15–25 °C, and then injected into a hot mold at 160–200 °C, completing cross-linking within 10–30 seconds. The part ejected is the finished elastomer—no cooling stage, no regrind, and once the runner cures in the wrong place, there's no second chance. The cold barrel, hot mold—this inversion is the most important mental model. It determines gate placement, vent design, cold slugs, and the thermal isolation between the mold and the injection unit. Ignore this, and you'll either get premature curing in the barrel or uncured liquid seeping from the parting line.

Decide hardness before drawing the part

Hardness determines wall thickness, draft angles, undercut tolerance, and flash behavior. A 30 Shore A gasket and a 70 Shore A grommet do not share the same design rules. Lock in the hardness with your application engineer before freezing the CAD geometry—every downstream dimension depends on it.

Wall thickness, uniformity, and cure time traps

LSR cures from the outside in. A 1.0 mm wall thickness at 180 °C cures in about 12 seconds; a thick section of 4.0 mm in the same mold might not be fully cured after 45 seconds, while the adjacent thin wall has already over-cured and become brittle. The solution is not to extend the cycle time, but to make the wall thickness uniform. Areas exceeding 3 mm should be cored out, and the overall thickness ratio of the part should be kept within 2:1. All transitions between thick and thin sections must have ample radii. The minimum wall thickness for LSR can be quite thin: with 40 Shore A and appropriate venting and gate placement, 0.25 mm is moldable. For parts with wall thicknesses below 0.4 mm, the gate must be located within 15 mm of the thin wall, otherwise, it will cure before filling is complete.

Parting lines, shut-off surfaces, and flash issues

Uncured LSR has a viscosity of about 100–1000 Pa·s and will seep into any gap greater than approximately 5 microns. Shut-off surfaces must meet at a positive angle (at least 3–5°), not knife-edge to knife-edge. The parting line must be placed where any trace of flash will not affect part function. A common practice for medical valves is to place the parting line on a non-sealing rib, never crossing a sealing surface.

Cold runners vs. conventional runners

LSR is thermoset; any material remaining in the runner that cures becomes scrap and cannot be reground. Cold runner systems keep the runner below the curing temperature, delivering the material directly to the gate, completely eliminating runner waste. Mold construction costs approximately 20–35% more, but for products with an annual shipment of 500,000 units, the material savings can pay back the investment within four to eight months.

Undercuts, overmolding, and demolding

Soft LSR can tolerate undercuts that would require side-actions for rigid polymers. A part with 40 Shore A, 2 mm wall thickness, and an undercut depth of 0.8 mm can be successfully demolded with a straight pull core, provided the draft angles and radii are correct. At 70 Shore A or higher, the same geometry would tear. Overmolding onto a rigid substrate—common for soft-touch grips in medical applications—requires the substrate to withstand 180 °C for 30 seconds. In practice, this usually means PPSU, PEI, PBT, or high-temperature nylon, not ABS or PP. Demolding strategies rarely rely on a single method: a typical mold will combine generous draft angles (at least 2–5°), air blasts, stripper plates, and manual removal if necessary. Mold release agents are only used as a last resort—they can migrate into the part and potentially jeopardize biocompatibility certifications.

Three applications, three vastly different design windows

A handheld surgical grip where hardness was locked in before mold design

A surgical instrument manufacturer needed a grip: soft enough for a four-hour surgery, yet hard enough to transmit torque to an internal steel shaft. The first prototype, molded in 30 Shore A, failed to transmit torque at 0.8 N·m. The second version, in 70 Shore A, could transmit torque but caused surgeon hand fatigue in less than 90 minutes of simulated testing. The critical design move: the team switched to a 50 Shore A outer wall with 2 mm thick 80 Shore A internal ribs, completed in a single two-shot injection. The soft feel came from the outer layer, the torque transmission from the internal ribs. Mold costs increased from USD 42,000 for the single-shot version to USD 78,000 for the two-shot version, but it saved the cost of a third prototyping round (USD 28,000, eight weeks) and met the original clinical timeline. Mass production runs on a 120-ton machine with a 52-second cycle, hard material through a cold runner, soft material through a conventional sub-gate—a reminder that two-shot LSR molds are essentially two process windows sharing a thermal boundary.

An automotive connector seal that survived salt spray testing after rework

A Tier 1 supplier's 55 Shore A LSR seal failed after 240 hours in salt spray testing, while the specification required 480 hours. The root cause was a 0.15 mm flash ring at the parting line, allowing electrolyte to penetrate beneath the sealing surface. Shifting the parting line 1.5 mm to a non-functional rib and tightening the shut-off angle from 2° to 5° required eight weeks of mold rework, but ultimately passed 720 hours with a safety margin.

An A-1 optical grade LED headlight lens

Optical grade LSR, 60 Shore A, molded into a 4 mm thick lens with A-1 polished mold surfaces, combined with a vacuum venting sequence that pulled the cavity pressure down to 50 mbar before each shot. Initial yield was only 62% due to micro-bubbles. After adding 0.05 mm of compression packing during the final filling stage and reducing the cure temperature from 180 °C to 165 °C to slow down outer layer curing, the yield increased to 94%. Micro-bubble defects in optical LSR are almost always traceable to venting issues, not the material itself.

Matching LSR material grades to applications

Dos and Don'ts of LSR design

Common mistakes in first LSR projects