Threaded Inserts: RTV Fixtures vs LSR Parts

A thread cut into 40 Shore A silicone is a suggestion. A brass or stainless insert molded into the rubber is a fastener.

Heat-staked or ultrasonic inserts that plastic designers love do not exist here: there is no thermoplastic melt pool. You either insert-mold the metal in LSR or HCR (parts OEM), or you bond/pot it in a room-temperature platinum RTV fixture (this site), or you use a clamp plate the rubber never pretends to be. Tapping the elastomer is for temporary jigs and toys. If the joint must take wrench torque or repeated assembly, design a metal thread and a rubber lock around it.

On RTV mold hardware the cousin question is different: threaded mold components — jacket bolts, locating screws, mandrel threads — must not dump platinum into a helix you still need to unscrew. Pin those threads too.

Primer versus mechanical lock on metal overmolds is covered elsewhere in this library. This article is the threaded case: bosses, weld nuts, standoffs, fittings, and the pins that keep rubber out of the helix.

Why the rubber cannot be the nut

Silicone creeps under load. A thread form in the elastomer will relax, especially in heat, oil, or a 10–30 Shore A grade. Pull-out becomes a torn helix. Torque to strip is a tear number, not a steel property. Fine pitches are worse. Coarse “threads” in hard 70A HCR can work as a cheap lamp-holder once; they are not an M3 in a device that sees service. A 15A RTV jacket with a tapped hole is a one-dump jig.

Metal inserts exist so the thread is metal-to-metal. The silicone’s job is to retain the insert (torque-out, pull-out, rotation) and to seal if that is required.

Insert geometry that actually stays

Buy or machine inserts with something for rubber to grab:

• Knurls, grooves, hex collars, undercuts, through-holes the silicone can rivet.
• A flange that the rubber cannot pull through.
• Length enough that the insert is not a washer sitting on a 0.8 mm skin.

A smooth OD with primer only is a peel waiting for the first reverse torque. Primer (silane systems, cleanliness, sometimes blast) still helps against spin in shear. Geometry does the wrench.

Materials: brass is common and machines well; stainless for corrosion and medical; aluminum if you like galling and CTE fights. Coatings and oils on catalog inserts are inhibition and bond killers. Degrease. If the insert is passivated stainless, say so; some films fight primer and poison platinum RTV.

Load path: put the metal on metal when you can. A screw that clamps a housing through the insert into a *metal* jack on the other side does not ask the silicone to carry compression. A screw that only squeezes rubber under the insert head will embed and loosen.

If you ignore the flange, the defect is an insert you can pull out with the screw still attached.

Multi-insert parts (a pad with four standoffs) need a load sequence the cell can repeat. A single missed pin in a 4-cavity HCR tool ships a blind hole that looks like a mold problem. On RTV, a missed pin in the box ships a helix full of platinum. Fixtures, magnets, or spring-loaded pins that make the insert sit before close are cheaper than a rework cell.

Tooling: keep rubber out of the thread

LSR will fill an M3 like a grease gun. RTV will too if the gap is there — slower, same helix. The tool or fixture must shut off on the insert and pin the thread.

Typical hardware:

• A locating pin or screw that occupies the female thread (or a male-threaded core if the insert is a stud).
• A land that pinches a designed face of the insert so rubber cannot run the helix. Gaps on the order of 0.005 mm are enough to flash on LSR. RTV is more forgiving on viscosity and still wrecks a thread you did not pin.
• Nests that hold height. An insert that sits 0.3 mm low becomes a flash path and a proud/sunken boss.
• For multi-cavity, poke-yoke the orientation of hex collars.

HCR compression: load inserts by hand onto pins, then the biscuit. Pins walk; operators miss a cavity. Poka-yoke belongs on that cell as much as on injection.

Heat: LSR cavities are hot. Thin aluminum inserts can grow and then sit loose after cool-down if the rubber lock was marginal. RTV fixtures are shop temperature — that CTE fight is quieter; inhibition and a missed pin are louder.

Demold: the pin must unscrew or pull without rotating the insert in green rubber. A robot that yanks a core pin out of a freshly shot 20A part can spin the insert and tear the lock. On RTV, unscrew the pin after through-cure, not at gel.

Process and quality

• No mold release on the insert OD.
• Preheat if the insert is a heat sink that would otherwise steal LIM cure from the interface (large stainless fittings). RTV: a huge cold fitting still slows through-cure; wait on a witness coupon, not the TDS cup.
• Self-bonding LSR grades that list metals: still clean, still knurl. Same for platinum RTV.
• Torque-out and pull-out tests on the *production* durometer and thickness, after post-cure if you post-cure. Day-one green rubber lies. An RTV-glued nut on a bench is not the LIM number.
• Electrical: the insert can be a live terminal. Creepage through a flash path of carbon-filled silicone is a different product. Name the grade’s conductivity.

FDA 21 CFR 177.2600 and ISO 10993 apply to the rubber (and to what the insert is made of, if it contacts). A medical-grade LSR around a mystery zinc insert is an incomplete file.

RUUIPON Mold Structure Decision: Inserts in a dump or a prototype boot? Nest them in the box; pin the helix. Deep undercuts plus inserts? → two-part so you can load pins. Soft Shore? → mother mold so the nest height does not walk.

Common mix-up: primer-only on a smooth bushing

A lot of RFQs call a smooth OD plus primer wrench-proof. In the shop that usually means spin-out after heat aging — or after the first reverse torque on a 20A RTV pot.

What to ask when you're buying

• Insert drawing: knurl, flange, material, coating, thread?
• Pin shut-off stack vs the leak you will actually run?
• Torque-out and pull-out after the thermal history you will ship?
• Load path: metal-to-metal clamp, or squeezing rubber under the head?
• Process: LSR insert mold, HCR compression, or RTV fixture only?

Worked example

You're insert-molding four M3 brass standoffs in a 50A LSR pad, 4-cavity steel, 12,000 / yr — that is a parts OEM: occupy each helix with a pin, shut off on a flange, knurl the OD. For 30 prototypes of the same pad, pot the knurled inserts in a platinum RTV fixture at room temp, pins in the helices. Do not publish the RTV torque number as the LIM number. Tapping 40A is a suggestion that creeps.

We would not recommend tapping 20–40A as a service fastener

We would not recommend a smooth OD with primer only, release on the insert, or T1 torque on green rubber as the ship test. We would also not swap brass for stainless after you qualified pull-out, or leave jacket-bolt threads unpinned in a pour box.

Typical shop values

Reference values only. Confirm material-specific limits against the supplier’s technical data and agree project acceptance criteria before production.

| Item | Typical shop value | | --- | --- | | Example geometry | M3 brass, 50A pad, pin + flange shut-off | | LSR flash leak | About 0.005 mm — LSR fills an M3 like a grease gun | | LSR shrink | About 2–3% — test after post-cure | | HCR shrink | About 1.5–2.5% | | Platinum RTV shrink | About 0.1–0.3% — still pin the helix | | ISO 3302-1 | M2 common on rubber; boss height is nest + shrink |

Copy this into an insert-thread RFQ

• Process: LSR insert mold, HCR compression, or RTV fixture
• Insert drawing, thread, pin supplied by whom
• Allowed flash on the helix (ideally none); rubber thickness around insert
• Torque and pull tests after full thermal history
• Seal function if the screw must gasket

Send the insert solid and the pad section. We'll quote pins and knurls, not a tapped 40A hole. Custom RTV molds.

Related: flex PCB encapsulation, 2-shot PC/ABS, overmolding onto glass.

FAQ

Can I tap threads directly into molded silicone instead of using an insert? Only for low-duty, low-cycle uses in harder compounds, and even then the thread will creep. For machine screws, assemblies, and service parts, insert-mold or pot metal. Tapping 20–40 Shore A is how you strip a hole on the second assembly.

How do I keep LSR or RTV out of an M3 thread? Occupy the thread with a pin or sacrificial screw and shut off on a designed insert face. LSR flashes through about 0.005 mm. A loose pin plus a dirty helix is a filled thread. Spec the insert tolerance and the pin together. Same pin discipline in a pour box.

Is primer enough to stop an insert from spinning? Primer helps shear on a clean OD. A wrench is a peel and a torsion on a smooth cylinder. Add knurls, hex, grooves, or a through-rivet of rubber. Primer plus a smooth bushing is a spin-out after heat aging.

Should inserts be brass or stainless in medical silicone? Stainless is the usual corrosion and documentation choice. Brass is common in industrial and electrical. Either must be degreased and called out in the biocompatibility and materials file if there is patient or fluid contact. Do not swap alloys after you qualified pull-out and bond.

Why did torque-out pass at T1 and fail after post-cure? The rubber shrank and the network finished. An insert that was jammed in green LSR can sit looser, or the lock can tear at a flash nick, after the oven. Test pull and torque after the full thermal history you will ship, on the production thickness, not on a warm first shot. RTV fixtures skip that oven; still wait for through-cure before you publish torque.