RTV Prototype vs Production Tray: What Each Tool Proves

If you're buying a custom RTV silicone mold, a prototype is a tool whose job is to prove something cheaply: shape, peel, or pocket fill. A mass-production tray's job is to hold dump weight, jacket flatness, and a recast path at a quantity band. They aren't two sizes of the same object. A poured platinum cavity can prototype a figurine and teach you nothing about LIM flash. An aluminum HCR two-plate can be the production tool for the life of a gasket — that's a molded *part*, a different WYD-style quote. Steel isn't a moral upgrade from RTV.

If you treat prototype and production as a mandatory ladder (print, RTV, aluminum, steel, 32 cavities), you'll buy metal you don't run and skip the jacket you actually needed.

What the prototype is supposed to prove

Write the proof on the PO.

Shape and peel. Does the lip seat. Does the mascot look right. Does the tray nest. Does the rubber peel without tearing a 10A–20A lip. A printed master plus a 1–2 cavity RTV block, or even a one-piece glove, can answer that. Chemistry match to a future LSR *part* is optional and usually the wrong ask.

Dump and fill-weight. Does every pocket take the same pectin or chocolate. Does the jacket stay flat. Then the prototype already needs production-family RTV (platinum vs tin, Shore of the skin) and a rigid support, even if cavity count is still one or two.

Process scars on a molded *part*. Gate vestige, knit, flash class. Only a tool in the production *part* process shows those — HCR compression or LSR with a cold-runner. An RTV prototype will not. Don't file RTV peel as LIM flash.

If the prototype is supposed to prove all three, it isn't a prototype. It's a first production tool with extra anxiety, or it's two RFQs glued together. Mixing "cheap RTV" with "must match LSR flash" is how prototype programs overrun and still ship the wrong lesson.

Soft tools, aluminum, and steel aren't a ladder you must climb

RTV soft mold (this site). Fast geometry. Good for castings, food-mold masters, chocolate and gummy trays, and look-like parts. Wrong as a stand-in for LSR injection or HCR compression data. Tin-cure versus platinum-cure still matters inside this box: tin keeps shrinking. Production RTV is the same chemistry with more pockets, a stiffer jacket, and a recast clause.

Aluminum cavity. Common HCR production metal. Also a common LSR prototype tool when the shop will inject into aluminum. Enough for many *part* programs forever. Wear and shut-off life depend on duty, not on the word prototype. Different RFQ.

Steel cavity. Chosen for shut-off life, texture durability, cavity count, and LSR volume. Not required because a slide deck said "production grade tooling." Different RFQ.

Cavity count is a separate axis. One RTV cavity is still a prototype if the mascot is changing. Sixteen RTV pockets are a production bet if the drawing is frozen and the dump is the process you'll keep.

Printed or high-temperature polymer cavities sit off this ladder. Some shops use them for a handful of LSR or HCR shots. Treat them as disposable process witnesses. They aren't a mass-production plan, and they aren't RTV mold-making either.

Chemistry match versus shape match

This is the split that burns money.

Shape match: the envelope and the logo look right. Buyers show it at a fair. RTV or even a painted print can do it.

Chemistry match for a *flexible mold you will sell*: platinum or tin RTV, Shore of the skin, food-contact path (FDA 21 CFR 177.2600, LFGB), jacket material. The prototype *is* the RTV article. Stop borrowing LSR language.

Chemistry match for a *vulcanized part you will sell*: platinum LSR on a LIM cell, or millable HCR on a press. Poured RTV at the same Shore A is a different polymer network. Don't file RTV tensile data as LSR design verification. Don't file HCR compression-set as LSR compression-set because both say 50A. Shore A is a durometer. It isn't a process.

When a prototype tool should be thrown away

Throw it away (or warehouse it as a history piece) when:

• The production process is different (RTV tray to LSR *part*, or HCR to LSR).
• Tin-cure was a short-run copy and you need a platinum library next season.
• Shrink was wrong on the *casting* and the master can't be recut economically.
• Cavity count and jacket registration can't scale.
• The insert was a compromise (sticky SLA master, no jacket) that will balloon at volume.

Keep it and run it when the process is still RTV, the drawing froze, and output is enough. Many chocolate and gummy trays never "graduate" to steel. Forcing an LSR recut to satisfy a procurement template isn't engineering — it's a new object.

Handoff documents that prevent a second tool

If you do need a second RTV tool (more pockets, new logo, recast), hand over:

• Frozen CAD with revision and shrink note (cast-article compensation, not an LSR 2–3% class number).
• Platinum or tin, mix, Shore of the skin, jacket material.
• First-article fill-weights and the misses you already accepted.
• Parting, keys, logo depth, master finish.
• What the prototype *failed* to prove, so the production tray doesn't repeat a cosmetic argument.
• Who keeps the unsanded master.

A photo album of samples isn't a handoff. Neither is "same as prototype but in steel." Steel will copy a bad parting just as faithfully as a poured jacket — and it's a different factory.

Ownership should be in the same pack. If the prototype master is yours, say whether it transfers, stays as spare, or is scrapped. If the shop retained it, a production recast at another factory starts from CAD only. That's survivable when the CAD is complete. It isn't survivable when the only record of the split was a pourer's memory.

Common mix-up: treating RTV as an LSR qualification mule

A lot of RFQs treat a platinum glove mold as "the prototype" for a future LIM part. In the shop that usually means you approved shape and then discovered flash, shrink, and compound don't transfer. If production is vulcanized parts, prototype in the production family on a parts OEM. If production is a tray you dump into, prototype in RTV and stop climbing a metal ladder.

What to ask when you're buying

• What must this first tool prove: shape/peel, dump weight, or *part* flash?
• Platinum or tin, Shore of the skin, jacket yes/no?
• Cavity count now versus after freeze?
• Who keeps the unsanded master for the recast?
• If we later need HCR or LSR *parts*, is that a separate RFQ (it should be)?

Worked example: 4-cavity proto, then 16-cavity tray

A gummy brand needed a 28 mm bear, 4.0 g target, food-contact, about 8,000 dumps a month once the mascot froze. First tool was a 4-cavity platinum RTV sheet, 10A–20A skin, rigid jacket, so peel and fill-weight could change without recasting sixteen pockets. After two witness dumps the face froze; the production tray copied the same master at 16 cavities. A parallel path that jumped to LSR steel "so we can scale" would have been a different object that doesn't deposit pectin.

When this split does not apply

If you actually need thousands of molded silicone *parts*, don't buy an RTV prototype as a fake LIM mule. Quote HCR or LSR on a parts OEM. A one-off art casting in tin-cure doesn't need a production recast clause. Geometry can override quantity: one undercut-heavy 80 mm figurine at 30 resin casts still needs a multi-piece production-style jacket.

Related reading: what a custom RTV mold is, how to cut development cost, sampling, custom RTV molds.

Prototype vs production lines to copy onto the PO

Copy this. A price without these lines isn't comparable.

• Object: poured RTV cavity (not HCR, not LSR part tool)
• Proof this tool must pass: shape / peel / fill-weight (pick)
• Platinum or tin; Shore of skin; jacket material
• Cavity count now; cavity count after freeze
• Who keeps the unsanded master; recast of the skin included or extra
• Explicit ban: do not treat this as LIM flash data

Send the pocket STEP, what the first tool must prove, and monthly dumps after freeze. We'll say whether you need a 1–2 cavity peel check or a production tray with a recast clause.

We would not recommend treating RTV as an LSR qualification

We would not recommend filing a platinum RTV peel as LIM flash data. We would not jump to a 16-cavity production tray while the mascot is still changing. We would not raise Shore A to "make it production," and we would not cut LSR steel "so we can scale later" while you still need a cavity to dump into.

RUUIPON Mold Structure Decision: first tool proves shape and peel. Production RTV adds jacket stiffness, cavity count, and a recast clause. Flat back? open-back. Deep undercuts that won't peel? two-part / plug / core. Soft 10A–20A or wet dump? mother mold. If you later need vulcanized parts, that's a new HCR or LSR RFQ.

flowchart TD
  A[What must this tool prove?] --> B{Shape and peel only?}
  B -->|yes| C[1-2 cavity RTV prototype]
  B -->|dumps weight recast| D[Production RTV tray plus jacket]
  A --> E{Need molded silicone parts?}
  E -->|yes| F[Stop: different WYD-style part RFQ]
  C --> G{CAD frozen?}
  G -->|no| C
  G -->|yes| D

Typical shop values

Caption: Typical shop values (not a named lot).

| Item | Typical shop value | | --- | --- | | Prototype RTV | 1–2 cavities, 10A–20A platinum, mix 1:1 or 10:1 | | Production tray | Same chemistry; more pockets; rigid jacket | | Platinum shrink | About 0.1–0.3% after full cure; tin keeps moving — bad as a library prototype | | Loaded-tray edge sag without a jacket | About 0.5–1.5 mm | | Contrast LSR / HCR *part* shrink | LSR about 2–3%; HCR about 1.5–2.5% on a different tool | | 4-cavity proto then 16-cavity tray | Common once dump weight and peel are frozen |

FAQ

Can I use a platinum RTV prototype to qualify an LSR consumer part? You can use it to look at shape. You cannot use it to qualify LSR flash, shrink, or compound. Run LIM samples in the production grade on a parts OEM before you freeze a steel family tool.

Is aluminum only for prototypes in silicone? No. Aluminum is a normal HCR compression production metal. It's also used for short-run LSR. Steel is a duty choice, not a graduation requirement. Neither metal is this RTV product.

When should I jump straight to a multi-cavity production tray? When the drawing is frozen, the dump is named, and the quantity band makes extra pockets cheaper than labor. If any of those is false, extra cavities are extra recasts.

Does a prototype mold need the same food-contact file as production? If you will eat from the prototype or file it as the production compound, yes, use the named path (FDA 21 CFR 177.2600, LFGB). If the prototype is a look-like RTV block that will never ship, don't pretend it carries the file.

What's the cheapest honest prototype for a future chocolate tray? Usually a 1–2 cavity platinum RTV skin against a finished master, with the jacket you intend to keep. A poured "gasket" in RTV is cheaper still and answers the wrong question if you actually needed HCR parts.