Co-Injection Molding (2-Shot): Combining Hard Plastics (PC/ABS) and Liquid Silicone

“Co-injection” on a silicone RFQ usually means 2-shot: a rigid thermoplastic first shot, then LSR overmolded in the same cell. It does not mean a sandwich foam core like some plastic co-injection. Say 2-shot (or 2K) and name the resins, or you will get a quote for the wrong machine.

PC, ABS, and PC/ABS are the usual hard half for grips, buttons, goggles, device housings, and seals that need a frame. LSR is the soft half. The cell is a rotary or index tool with two injection units: a conventional plastic unit feeding a hot runner into a cooled plastic cavity, and an LSR unit feeding a cold deck into a heated silicone cavity. Those two thermal maps share a table. If you ignore the split, the defect is a warped PC window, an uncured LSR lip, or a bond that peels after a water soak.

Insert-molding a pre-made PC part in a second LSR tool is a legitimate cousin (transfer overmold). It is not the same cell, not the same heat history, and not the same shut-off. Pick one and RFQ it.

Two shots, two chemistries, one clock

Shot 1: PC, ABS, or PC/ABS. Melt, fill, pack, cool enough to hold shape and to survive index. Glass-filled grades change warp and bond. Flame-retardant packages and extra mold release change bond more than the data sheet admits.

Shot 2: LSR. Cold-deck feed into a hot cavity so the rubber vulcanizes against the plastic that still carries heat from shot 1. Residual heat helps the interface. A cold insert in a standalone overmold tool does not get that gift unless you preheat.

Bond is designed:

• Self-bonding LSR matched to that resin family (PC is often easier than nylon; ABS and PC/ABS are not “the PC grade”).
• Primer or plasma when the grade is not self-bonding, or when the plastic is a difficult blend.
• Mechanical interlocks: through-holes, grooves, wrap-arounds. Geometry does not care about dishwasher chemistry as much as a face bond does.

Post 050 in this library covers PC/ABS/PA bond in more depth. The 2-shot-specific failure is heat: the LSR cavity wants 160-200 °C class heat. PC will distort, stress-whiten, or lose a tight hole if the index time, insulation, and water in the plastic half are wrong. ABS can be worse on heat. Name the exact resin and HDT, not “black plastic.”

If you ignore drying of PC, the defect is splay on shot 1 that everyone blames on the LSR flash.

Shut-offs are the 2-shot tool

LSR will leak through about 0.005 mm. It will also run under a plastic edge and peel later. The steel must pinch the substrate on a designed land (in-line shut-off), not swipe along a taper you cannot maintain.

Avoid:

• OD/ID shut-offs that only crush in the slide direction.
• A 0.3 mm LSR film on a glossy PC plateau with no lock and a square step.
• Gates that impinge and wash a thin PC wall.
• Parting lines that print on a sealing lip because the rotary split was placed for steel convenience.

Crush ribs on the plastic, small and intentional, help the pinch and disappear under rubber. Flash under the substrate is a peel-start. Flash on the PC show face is a customer reject that no bond test will save.

Valve-gated LSR drops onto the overmold zone. Vestige on a keypad web is a feel defect. Open sprues on a 2-shot LSR half fight the robot and the rotary envelope.

The rotary / index problems plastic shops underestimate

• Pitch at two temperatures. The plastic half is cooler than the LSR half. Steel grows differently. Alignment of cores that pass through both shots has to be calculated at operating temperature, not on a 20 °C CAD assembly.
• Cycle hostage. The slower shot owns the cell. Thick LSR sections or slow PC pack-out both tax the robot. Thin LSR on a thick PC brick still waits for the brick.
• Contamination. Plastic hot-runner drool into an LSR cavity is a cure-inhibition and inclusion problem. LSR drool onto a PC show face is a smear. Needle valves and purge discipline are process, not accessories.
• Release. No silicone release on the bond land. Plastic release on the first-shot tool must not walk onto the overmold face. Self-bonding grades make this non-negotiable.

HCR 2-shot exists in special cells and is rare next to LSR. RTV over a PC prototype is a mockup. Neither qualifies a 2-shot LSR bond.

DFM that belongs on the CAD before the tool

• Uniform LSR thickness where you can. A blob next to a 0.4 mm membrane fights cure and flash together.
• Radii on plastic corners that would cut the rubber.
• Room for the cold deck and the hot runner in the same shoe without cooking the deck or freezing the LSR land.
• A handling surface for the robot that is not the keypad.
• Tolerance language: ISO 3302-1 on the rubber, a plastic drawing standard on the frame, and a relationship dimension for the overmold edge. Three languages. Write all three.

UL 94 on the PC/ABS, FDA 21 CFR 177.2600 or ISO 10993 on the LSR, and a peel-after-soak on the *pair* are different files. A V-0 plastic with an untested self-bonding grade is not a finished stack.

RFQ notes for 2-shot PC/ABS plus LSR

Name 2-shot LSR over PC / ABS / PC-ABS, the exact resin and color, self-bonding versus primer, interlock features, shut-off concept, and whether the cell is rotary in one tool or a two-tool insert transfer. Send the bond test (peel after water or chemical soak, not only a dry tug) and the heat limit of the plastic. If you ignore the resin’s heat, the shop will quote a beautiful cold deck and then warp your lens.

FAQ

Is 2-shot the same as insert molding a PC part in an LSR tool?

No; 2-shot keeps both injections in one cell so the plastic still carries heat into the LSR shot. Insert overmold uses a cold (or oven-preheated) part in a second tool. Bond, warp, and shut-off all change. Quote the cell you will actually run.

Why does a self-bonding LSR that worked on PC fail on PC/ABS?

The blend is a different substrate. ABS fraction, mold release, and flame-retardant packages change the interface. A grade listed for PC is not qualified on PC/ABS until you peel it after the real soak. Treat the blend as its own material.

Can the LSR half use a hot runner to save space in a rotary tool?

Not in the thermoplastic sense: LSR needs a cold deck so the feed does not vulcanize, while the plastic half uses the hot runner. Cramping both into one shoe is a layout problem. Cooking the LSR drops to “save space” seizes them.

What shut-off geometry actually keeps LSR off the PC show face?

An in-line land where steel pinches the plastic, often with a small crush rib on the substrate. Tapered swipe-by shut-offs are hard to keep tight as the two halves run at different temperatures. If LSR gets under the plastic edge, that flash becomes the peel start.

Does UL 94 on the PC/ABS cover the overmolded part?

It covers the plastic grade as tested. The assembled part, the LSR, and any primer are a different specimen. If the device needs a flame rating, test the stack. If it needs food or skin contact, file FDA 21 CFR 177.2600 or ISO 10993 on the silicone, not on the ABS datasheet.