Compression Molding vs LSR Injection for Silicone Device Cases

Two factories quote the same rugged case for your POS terminal. One comes back at a higher piece price with almost no tooling line. The other wants five figures up front and then undercuts the first quote by a third per unit. Neither is wrong. You are looking at the two dominant routes for a silicone protective case — compression molding and liquid silicone rubber (LSR) injection — and the choice between silicone compression molding vs liquid silicone injection for protective cases is decided by your annual volume, your colourway count, and how much hand labour your part geometry forces onto the factory floor.

This is one of the few process decisions where the buyer, not the engineer, should hold the pen. The engineering difference is real but modest. The cost structure difference is enormous, and it moves in the opposite direction to what most first-time buyers assume.

How each process actually works on the floor

The two routes are not variations on a theme. They use different raw material, different presses, different tool construction, and different labour models.

Compression molding: preform in, flash out

Compression molding runs on high-consistency rubber (HCR), sometimes called gum stock — a dough-like uncured silicone that arrives in slabs. The factory mills in colour masterbatch and curing agent, then extrudes or cuts the compound into pre-weighed blanks. An operator lays a blank into each open cavity of a heated tool, the press closes at pressure, and the part cures in the cavity at roughly 170–200 °C. Cure time scales with the thickest section of the part, not its footprint. The press opens, an operator pulls the parts by hand, reloads, and closes again.

Because the blank is deliberately slightly overweight — you cannot risk a short fill — excess material squeezes out at the parting line. That is flash, and on compression it is not a defect, it is the design of the process. Every part comes off the press with a skirt of cured silicone that has to be removed in a second operation.

LSR injection: metered, gated, cold-runnered

LSR is a two-part, platinum-cured liquid. Pumps meter part A and part B at a fixed ratio through a static mixer, a dosing unit adds pigment, and the mixed material is injected into a hot tool through a cold runner block that keeps the silicone below its cure temperature right up to the gate. The part cures in the cavity in seconds. Tools can be built essentially flashless, parts can drop onto a conveyor unattended, and shot-to-shot repeatability is far tighter than an operator placing blanks by hand.

That capability is bought with tool complexity. An LSR tool needs vacuum sealing to evacuate air before the shot, precision venting at every flow front, hardened and highly polished steel, and the cold runner or valve gate assembly itself. It takes longer to cut and longer to tune, and every engineering change is more expensive to make.

Where the money actually goes

The honest comparison is not piece price. It is tooling plus piece price plus the labour hidden inside the piece price.

Read that table as a break-even, not a ranking. LSR wins on variable cost and loses on fixed cost. Compression wins on fixed cost and loses on variable cost. Somewhere on your volume curve the lines cross, and where they cross depends on details a spreadsheet from the sales side will not show you. A fuller breakdown of what tooling and mold costs really look like in China covers the fixed-cost half in more depth; this article is about the variable half.

Detail one: cavitation on a compression tool is limited by a human, not by the press

Buyers routinely ask why a compression tool for a case is quoted at 8 cavities when the press platen looks big enough for 24. The answer is the loading window. Every blank an operator places is already sitting on a hot cavity and beginning to cure. By the time the operator reaches cavity 24, cavity 1 has been curing under no pressure for well over a minute, and it will show up as a skin defect or a knit line. Practical cavitation on a hand-loaded compression tool is set by how many blanks one pair of hands can place before the first one starts to skin over. That is why compression cavity counts stay small, and it is the real mechanism behind the piece-price floor.

Detail two: deflash labour is the line item nobody quotes separately

There are three ways to remove flash, and they have different consequences for your part:

• Hand trimming with a blade or a picking tool. Slowest, most expensive, but the only option when the flash sits against a feature you care about — a thin bezel lip, a fine grip texture, a port cutout corner.
• Die cutting on a press. Fast and repeatable, but it needs a flat, accessible parting line and a dedicated trim die per part.
• Cryogenic deflashing — parts tumble in a chamber with liquid nitrogen until the thin flash goes brittle and media blasts it off. Cheap per part at volume. It also softens sharp edges and can dull a fine matte or micro-texture surface, because it is an abrasive tumble.

If you have specified an anti-slip micro-texture and the factory has quoted assuming cryogenic deflash, your production parts will not match your hand-finished sample. Ask which method the quote assumes, in writing, before you sign off the sample.

Detail three: put the parting line on the drawing yourself

Compression flash leaves a witness line wherever the tool splits. On a device case that line can land on the outside face beside the screen bezel — visible, catchable with a fingernail, and the first thing a retail buyer complains about — or it can land on an inner lip or a bottom edge where nobody sees it. Toolmakers place the split line for moulding convenience unless told otherwise. Write the acceptable parting-line location and the maximum allowable witness height into the drawing at RFQ stage. It costs nothing then and is expensive to move after T1.

The material differences that follow the process

HCR and LSR are not the same silicone with different viscosity. Cure chemistry differs, and so do the downstream requirements.

Peroxide-cured HCR typically needs a post-bake — hours in an oven at elevated temperature — to drive off cure by-products. Skip it and you keep residual odour, worse compression set (the case relaxes and loses grip over months), and less stable dimensions as the part continues to shrink. Platinum-cured LSR is an addition cure with no by-products, so post-cure is often optional and driven by volatile limits rather than by basic part performance.

Silicone also shrinks substantially — commonly in the low single-digit percentages, varying by grade, filler loading and durometer — so cavities are cut oversize to a shrink factor chosen for one specific compound. Change the compound mid-programme, or change durometer, and the shrink factor changes with it. A case that fit the device at 50 Shore A can be measurably tighter or looser at 60 Shore A out of the same steel. If you are switching grade after tooling, ask for a fresh dimensional report, not a verbal assurance.

For overmolded constructions the difference is sharper still. Self-adhesive LSR grades chemically bond to certain thermoplastic substrates in a two-shot cycle. Compression-molded HCR over a plastic insert generally relies on primer, mechanical interlock, or both. If your case is a hard PC or ABS frame with a silicone bumper, this is the single most important question to resolve early.

Reading a factory floor: what an injection line does and does not prove

Here is the trap. Many silicone case factories also run a plastic injection line — WJM Silicone in Longgang, Shenzhen is a straightforward example, with silicone compression molding on one side of a 12,000 m² floor and plastic injection molding on Haitian machines on the other, feeding in-house surface treatment, printing, assembly and inspection. That is a genuinely useful combination if your case is silicone over a PC or ABS frame, or if you want the hard and soft parts of the same programme under one roof.

It is not, by itself, evidence of LSR capability. A thermoplastic injection press has a plasticising screw, a heated barrel and a cooled tool. An LSR press has a pumping and metering station for two liquid components, a static mixer, a cold runner or cooled sprue, and a heated tool. The building blocks point in opposite thermal directions. A factory can own dozens of thermoplastic presses and have no LSR dosing equipment at all.

So when a supplier answers "yes, injection is available" to an LSR question, treat that as an ambiguous answer rather than a confirmation. WJM's published capability list names silicone compression molding and plastic injection molding; whether LSR injection exists in-house, is subcontracted, or is not offered is a question to put to the factory directly rather than an assumption to build a programme on. The same discipline applies to any supplier you shortlist — and it is one of the underlying reasons quotes for the same spec come back so far apart.

Choosing, in the order that actually matters

Work through this in sequence. Stop at the first line that decides it.

1. Substrate bonding. Silicone bonded to a rigid plastic frame in one cycle points toward self-adhesive LSR two-shot. Silicone-only or primer-bonded assembly leaves both routes open.
2. Wall thickness and section. Thin walls and fine detail favour LSR. Thick corner bumpers and heavy sections favour compression, because injection gains less when cure time is dominated by section thickness anyway.
3. Colourway count. Six retail colourways in modest quantities each is a compression story. One black case in continuous production is an LSR story.
4. Annual volume and programme life. A one-year accessory programme rarely amortises an LSR tool. A multi-year fleet contract usually does.
5. Cosmetic tolerance. If a parting-line witness anywhere on the visible face is unacceptable, price the flashless route honestly rather than assuming deflash will hide it.
6. Engineering-change risk. If the device is not frozen, cheap-to-modify compression tooling is a real risk hedge.

What to ask the supplier next

Send this as a numbered list and ask for written answers before you accept any quote comparison.

1. Is this part quoted as compression molding or LSR injection? If LSR — is the LSR dosing and cold-runner equipment in your plant, or subcontracted?
2. How many compression presses, at what tonnage and platen size, and how many cavities are you quoting on this tool?
3. What is the quoted cure time per shot, and what section thickness drives it?
4. Which deflashing method does the quote assume — hand trim, die cut, or cryogenic tumble?
5. Where does the parting line fall on this part, and what witness height is acceptable at incoming inspection?
6. Do you post-bake this compound? At what temperature and for how long, and in what oven capacity?
7. What shrink factor was the cavity cut to, and which exact compound and durometer does that factor assume?
8. If the durometer changes after T1, what is the dimensional consequence and who pays for the correction?
9. For an overmolded build: self-adhesive grade, primer, or mechanical interlock — and what bond-strength evidence can you provide from a trial?
10. What is the cost and lead-time delta between the two routes at 5,000, 25,000 and 100,000 units per year?

Ask the same ten questions of every factory on your shortlist. The ones who answer with numbers rather than adjectives are the ones worth sampling with.