RTV Vacuum Casting vs. LSR Soft Tooling Prototypes

If you're choosing a prototype path, vacuum-cast RTV parts and LSR shots are not interchangeable samples unless you only care about shape. On this site the product is the RTV mold — master, box, platinum or tin skin, vacuum, jacket — that you then dump food, wax, resin, or PU into.

RTV vacuum casting (a silicone *mold*, usually platinum or tin, pulling a vacuum around -0.095 MPa / 28–29 inHg, then casting) is a geometry printer with a rubber tool. LSR soft tooling (a printed or aluminum *injection* cavity, hot, platinum liquid silicone shot under clamp) is a small version of a molded-*part* process. Using the second to “prove” a chocolate tray is how you buy heaters you will never use. Using the first to “prove” LSR flash is how fit samples pass and production parts miss.

Two different objects called “silicone prototype”

RTV vacuum casting (this site). You make a master. You mold a rubber tool around it. You cast dumps into that tool: polyurethane, a casting silicone, chocolate, gummy, wax. Degassing and vacuum help the copy. Cycle is manual. Shrink of *platinum* RTV tools is typically ~0.1–0.3% after a full cure. The *dump* shrink is whatever you poured. Flash is a function of how you split the rubber, not of a steel shutoff. Mix 1:1 or 10:1, Shore 10A–20A skin, jacket so the sheet does not sag 0.5–1.5 mm.

LSR soft tooling (wrong object if you need a cavity). You cut or print a cavity. You heat it. You inject an LSR grade through a cold sprue or a simple cold-runner. The part is a vulcanized article. Flash is a metal shutoff problem. Shrink is the LSR’s heat-cured band (~2–3%), not the RTV tool’s.

If the prototype is a handheld mold for a photo or a dump, RTV is the product. If the prototype is a gasket that must compress in a groove at 50A after post-cure, RTV is a costume — quote a parts OEM.

Common mix-up: life-testing a poured silicone dump as LSR data

Casting a *silicone* article in an RTV mold can get you near a target Shore A. It does not get you LSR flow, knit, or post-cure shrink. Don't send that part to a life-test and call it production data.

ISO 3302-1 might class either article. It is a drawing class, not a process.

When RTV vacuum casting is the right sample — and the right product

Use it when the question is shape, assembly choreography, a photo, or a production food/resin mold:

• Undercut-heavy models you are not ready to slide-steel
• Fit checks where a 0.1–0.3 mm argument is acceptable
• Color and form reviews before anyone should cut a library master
• Very small quantities where an LSR pail and a press setup would be theatre
• Parts that will *remain* craft-cast, and trays that *are* the product (gummy, chocolate, ice)

Tin-cure RTV tools are cheaper and more forgiving of dirty printed masters. They age and shrink more. Platinum RTV tools hold size and last longer in the rack, and they inhibit on uncured resin prints. Neither chemistry is LSR.

When LSR soft tooling is the right sample

Use it when the question is the *part* process you will live with — flash a customer will measure, gate witness, post-cured shrink into a housing, a true LSR grade, a regulatory packet on *that* compound. Aluminum 1-cavity is the usual vehicle. That is not a custom RTV mold PO.

The fork that changes the buy

Ask: what decision will this sample authorize?

| Decision the sample must authorize | RTV vacuum cast / RTV mold | LSR soft tool | | --- | --- | --- | | Does the CAD assemble? | Yes | Wasteful | | Does the lip invert without tearing the *mold*? | Yes, in 10A–20A | Wrong interface | | Will LSR fill this 0.4 mm fin? | No | Yes, that is the point | | What is the flash on a 0.02 mm steel shutoff? | Irrelevant | Yes | | Can we dump pectin at 2,000 trays/month? | Yes, that is the product | Wrong machine | | Can we start a biocompatibility file on an LSR grade? | Only if you cast that compound, which is not vacuum-cast PU | Yes, if the grade is the file grade |

A hybrid that wastes money: RTV samples used to freeze CAD, then a steel LSR tool with no aluminum bridge, when the product was always a *part*. A hybrid that works for molds: print or CNC master to kill form, then production RTV skin + jacket.

What to ask when you're buying

• Is the prototype an RTV mold (name mold rubber and dump compound) or LSR in a printed/aluminum tool (name LSR grade)?
• What decision will this sample authorize?
• Platinum or tin, jacket yes/no?
• Don't treat them as equivalent samples?

Worked example: worst feature, one afternoon

A confectioner argued RTV vs “just shoot LSR.” Worst feature: 18 mm bear chin return. One RTV coupon in 10A–20A platinum, mix 1:1, vacuum degas, jacket; ten peels, no nick. One LSR shot in a crude aluminum cavity of the same outline did not dump pectin and did not answer peel of the *mold*. The RTV sample authorized the tray. The LSR shot was a mood board for a toy nobody asked for.

When this article is the wrong quote

If you already know the shippable object is a vulcanized LSR *part*, do not buy another RTV loop to prove flash. If you already know the object is a chocolate tray, do not buy aluminum LSR. A tiny artisan can skip vacuum. Geometry can override quantity: a full bulb still needs a core at 80 dumps.

Related reading: custom RTV molds, CNC vs printed masters, when RTV is enough, gummy trays.

Cost and calendar, as mechanisms

RTV spends on the master, the rubber, vacuum time, and skilled pouring. Piece cost of the dump stays high because a person or a depositor is the machine. Tool life is counted in casts. Good for tens, sometimes thousands of dumps with recast, then the skin tears or the flash line goes ragged.

LSR soft tooling spends on CNC, heaters, tryout on a press, and a pail of compound. First parts are slower than RTV. Each extra *part* is cheaper than a hand cast — if you needed parts.

Do not pick on “which prototype is cheaper.” Pick on which wrong decision is more expensive: a pretty PU part that green-lights a steel tool you did not need, or an LSR shot that takes two more weeks when you needed a glove yesterday.

RFQ sentence: “State whether the prototype is RTV vacuum-cast (name mold rubber and dump compound) or LSR in a printed/aluminum tool (name LSR grade, post-cure, flash spec). Do not treat them as equivalent samples.”

Send the worst feature and the dump material. We will say which sample authorizes the PO you actually want.

We would not recommend LSR soft tooling to prove a chocolate tray

We would not recommend life-testing a poured silicone dump as LSR data. We would not skip the jacket on a large RTV sample. We would not raise Shore A first to “make it more like LSR.” We would not freeze CAD on PU and then cut steel for a *part* you have not shot.

RUUIPON Mold Structure Decision: pick the object. Cavity you pour into? RTV. Vulcanized part? stop. Flat back? open-back. Deep undercuts that will not peel? two-part / core. Soft 10A–20A or wet dump? mother mold.

flowchart TD
  A[What decision does the sample authorize?] --> B{Cavity to dump into?}
  B -->|yes| C[RTV master + skin + vacuum + jacket]
  B -->|LSR part flash/shrink| D[Stop: LSR soft tool RFQ]
  C --> E{Platinum on SLA?}
  E -->|undercured| F[Wash + UV + barrier]
  E -->|ok| G[Coupon the worst lock]

Typical shop values

Caption these as typical shop values, not a named lot.

| Item | Typical shop value | | --- | --- | | Mix | Platinum 1:1 or 10:1; vacuum degas until the foam collapses | | Vacuum | About -0.095 MPa / 28–29 inHg on the chamber | | Cavity Shore | 10A–20A skin with jacket | | Platinum shrink | About 0.1–0.3%; LSR *part* contrast ~2–3% — different RFQ | | Loaded-tray edge sag | About 0.5–1.5 mm without a jacket | | Fit argument RTV can stand | About 0.1–0.3 mm on a craft dump |

FAQ

Can a vacuum-cast PU part prove LSR flash and shrink? No. RTV casting copies shape at room temperature with rubber-split flash. LSR soft tooling is a hot cavity, metal shutoff, and heat-cured platinum shrink. Fit samples that pass in PU are how production *parts* flash, shrink, or miss hardness — if parts were the product.

When is RTV vacuum casting the right sample? When the question is shape, a photo, or a production food/resin mold: undercut-heavy models, fit checks that can stand a small argument, or quantities where an LSR pail would be theatre. Tin tools forgive dirty prints and age more; platinum holds size and inhibits on uncured resin.

When must I spend on aluminum LSR instead of another RTV loop? When the sample must authorize flash a customer will measure on a *part*, whether a thin fin fills in LSR, post-cure shrink into a housing, or a regulatory packet on that LSR grade. Handheld molds and chocolate trays do not need that.

Is a silicone part poured in an RTV mold production LSR data? It can near a target Shore A. It does not give LSR flow, knit, or post-cure shrink. Do not life-test that part and call it production data.

What one-afternoon test settles the fork? Make one RTV coupon of the worst lock in 10A–20A and, only if you need *parts*, one LSR shot in the intended grade. Put the dump in the mating food line, or the part in the housing after post-cure. If they disagree, you asked the wrong sample to authorize the PO.