Second-Sourcing Custom Silicone Molds: Transferring Tooling Between Factories Safely

Second-sourcing a custom silicone part is a process qualification, not a freight job. The crate can move in a week. The second factory still has to heat, vacuum, balance, and flash a tool it did not cut, on a press it did not try out, with a compound lot it did not validate.

Do this when you need capacity, geographic risk cover, or an exit from a shop that has become a hostage-taker. Do not do this as a surprise punishment in week two of a quality fight, with no files and a rusty shutoff. That is how you buy two scrap streams.

LSR steel, HCR compression tools, and even aluminum bridges can transfer. RTV jackets usually cannot; you recast them at the new shop from a master you still control.

Why a “simple transfer” fails

Fit to a different press. Tie-bar spacing, locator rings, cold-deck plumbing, heater connectors, vacuum pumps, clamp curves. A tool that ran on one LSR machine may be a hardware project on another. Ask for platens and connector drawings *before* the crate leaves.

Unknown wear. The sending shop’s “good” flash spec may be a land that is already rolled. The receiving shop will either trim forever or recut immediately. Photograph shutoffs and measure flash on a last-lot sample before you ship.

Missing recipe. Cure temperature, time, clamp, vacuum, injection profile, post-cure. If it lived in one technician’s head, the second source is developing a process from a used tool. That is not cheaper than a recut. It is slower.

Compound identity. Same grade code, different lot, different pigment paste, different mixer. Platinum LSR is intolerant of substitution theatre. Transfer the spec and the approved manufacturer, not “50A transparent.”

IP and politics. The first shop still has electrodes, a spare insert, and a night-shift CNC. A transfer without a no-duplicate clause and a file escrow is how you get a quiet third tool.

What must travel with the steel

Make a transfer BOM. If a line is missing, assume you will pay for it twice.

• Mold base, all inserts, slides, spares, cold-runner deck or sprue bushing, heaters, thermocouples, vacuum seals, lifting gear, nameplate photos.
• Native CAD, 2D with shrink, electrode or EDM files, DFM notes, steel list, last engineering change.
• Tryout sheet: last approved process window, not a single “standard LSR” line.
• Last FAI pack and a sealed sample set from a known-good lot (parts *and* runner if the runner still exists).
• Material spec, approved grade list, post-cure recipe, inspection plan, flash standard (photos, not adjectives).
• Maintenance log: shot count estimate, last polish, known damage.
• Packing: rust preventive, VCI, dry channels, locked slides, humidity card. A precision tool shipped as cargo arrives as a repair.

Contractually: title is yours (if it is not, you are not transferring; you are begging). No-use after ship-out. Return of files from the first shop. Receiving shop signs the same ownership, no-duplicate, and no-export clauses you should have had the first time.

Receiving: treat it like a new tool, because it is

The second factory did not cut this steel. Their first job is incoming tool inspection, not “load it and quote a piece price.”

1. Unpack against the BOM. Photograph nameplate, cavity count, shutoff, deck, connectors.
2. Clean to a known state. Degrease rust preventive in a way that is LSR-safe. Check channels for water and rust.
3. Fit-check on the intended press. Adapters and plumbing are engineering, not surprises on tryout day.
4. Dry cycle. Slides, ejectors, needles. Do not inject into a tool that is mechanically wrong.
5. First shots on the *approved* grade, window from the tryout sheet as a starting guess, not as a religion. The new press will not copy the old curve.
6. FAI against the drawing, not against the old shop’s sample vibe. Capability on flash and critical dims. If the land is tired, stop and recut. Do not “process around” a rolled shutoff for six months.
7. Lock a process window and a change-control rule. The second source is now the process owner for that site.

Expect tryout days. A transfer that is “free because the mold exists” is a lie you told purchasing. Budget receiving engineering as a line. It is still usually cheaper than a full recut *if* the tool is healthy and documented.

When to recut instead of transferring

Recut (new cavities, or a new tool at the second source from your files) when:

• The first shop will not release files or a honest wear log.
• Shutoff, slides, or deck are at end of life.
• The receiving press family cannot host the base without a science project.
• You need different cavitation or a cold deck the original never had.
• IP risk is high and you do not want the first shop’s steel traveling through a port they still influence.
• The original is aluminum you were already going to abandon for steel.

A parallel recut while the first tool keeps running is the clean dual-source: two qualified processes, two FAI packs, no hostage crate. It costs a second T_tool. It buys a real second source. Transferring a single tool between factories is dual-source only while both have it, which is never.

Sequence that keeps you out of court and off a plane

Paper first. Confirm title, pay open invoices you actually owe, issue a transfer PO to the receiving shop that includes tryout and FAI, not just freight.

Qualify the receiving shop as if it were a first source. QMS, press list, compound handling, cleanroom if medical. A crate does not fix a bad factory.

Files to you, then to them. Do not ask shop A to email shop B. You are the hub. Watermark, log, NDA.

Steel moves once, with insurance and a packing protocol. Last-lot samples in a separate carton.

Do not kill shop A’s process until shop B’s FAI is signed. Overlap is inventory and sanity.

Count inserts at both ends. If a spare cavity stays behind, you did not transfer the tool. You split it.

RFQ line to a potential second source: “Receiving inspection, press fit, tryout, and FAI on transferred LSR/HCR tool. Quote those engineering days separately from piece price. Recut option from buyer CAD as a parallel line.”

If they only quote a unit price “using customer tool,” they have not done this before.

FAQ

Can I ship an LSR tool to a new factory and expect identical parts on the first lot?

No: Different presses, heater circuits, vacuum, and compound lots shift flash and shrink. Plan a receiving FAI and a process window at the new site. Identical drawings, not identical luck.

Who pays for tryout at the receiving factory?

You should, as a stated engineering line, unless the transfer is the receiving shop’s idea to win the work and they bid tryout in. Leaving it unspoken is how tryout appears as “tool repair” later. Healthy tools still need hours. Tired tools need a recut quote, not optimism.

What if the original factory will not release CAD?

Then you do not have a portable tool. You have steel and a relationship. Use contractual file-delivery and, if the product is worth it, a parallel recut from whatever files you still hold (your STEP, your 2D). Reverse-engineering a hot LSR tool from parts is slow and inexact. Fix the contract on the *next* tool today.

Is a transferred aluminum bridge worth second-sourcing?

Rarely as a long-term dual source. Aluminum shutoffs are already a wear clock. A second shop will inherit a land you did not watch. Use aluminum to qualify geometry at one site; cut steel (maybe two insert sets) for dual-source production.

Should both factories run the same LSR grade lot?

Same *approved grade and manufacturer*, yes. Same lot is a logistics trick, not a requirement, if incoming inspection and validation allow lot changes. Do not let shop B “equivalent” a cheaper kit to hit a piece price. Write the approved list and change notice into both quality agreements.