TCO: RTV Mold Spend vs Dump Savings vs Steel Parts

If you're doing TCO on a custom RTV program, the mold pays for itself when dump savings, scrap drop, and labor drop outrun the stack before the SKU dies — and only if the shippable object is still a cavity.

TCO is not “mold quote vs dump price.” It is every dollar that exists because you chose this construction: master, jacket, recast, scrap, trim, storage, freight of the mold, leftover platinum kit, and the recast you already know you will need. Dump price is one term. If you optimize it alone, you will buy an 8-cavity monument for a mascot that lasts a season — or worse, a Class 103 LSR tool that cannot dump pectin.

Steel pays for itself on vulcanized parts when press time, scrap, and trim drop enough before the product dies. That is a different TCO on a parts OEM.

The terms, named so they can be added

Write one line each. Ranges are allowed. Blank lines are not.

T_tool. Construction: printed master vs CNC vs jacketed production RTV, cavity count, recast included, tryout, FAI. This is the investment people stare at.

T_mod. The first planned geometry change. Soft masters make this small. A finished CNC master plus art insert is medium. If CAD is not frozen, T_mod is not zero. Put a placeholder. Hardened LSR steel makes this large — if you were on the wrong object.

C_pour. Pour+set+peel time / cavities × bench hour cost × good dumps. Cavitation lives here. Set-dominated chocolate moves little with a faster robot. It moves with more pockets *or* with a jacket that does not sag 0.5–1.5 mm.

C_mat. Mass × dump grade × (1 + wasted outer-row). Jacket design can beat cavitation on this line.

C_labor. Trim, inspect, pack. Flash at the split and keys live here. A cheap glove that needs a knife on every dump can lose TCO to a more expensive jacket.

C_scrap. Start-up + running yield. Multi-cavity can raise this. Tired gloves can raise this. Wrong wait on tin can raise this.

C_setup. Per batch. Release qty determines how many times you pay it. A beautiful 10,000-dump unit price with 200-dump releases puts C_setup back on the table.

C_pm. Recast of the RTV skin, jacket refresh, UV storage. A library master assumes you will spend here. A printed throwaway assumes you will not, because the pattern is dead.

C_freight. Molds once (or twice if you dual-source). Dumps per release. Dimensional weight on bulky trays vs dense bags.

TCO over a horizon H (12 or 24 months is enough for most custom trays):

T_tool + T_mod + Σ(C_pour + C_mat + C_labor + C_scrap + C_setup + C_pm + C_freight)

Unit savings only matter as deltas on the C_ terms when you change construction.

Common mix-up: comparing T_tool of steel to the dump price of RTV

Never compare T_tool of a Class 103 LSR tool to the *unit price* of a chocolate dump. Compare TCO_H of both *objects* — and stop if the objects differ.

ISO 3302-1 does not belong in the TCO sum as “precision.” It is a drawing class if the article is rubber.

RTV vs the alternatives, as deltas

Stay on a cheap glove, no jacket. T_tool is small. C_labor and C_scrap from sag (outer row 3–6% heavy) are large. A jacketed skin wins when H × sag scrap + trim exceed the mother-mold delta.

Stay on jacketed RTV, recast as needed. T_tool medium and recurring (skins die). C_pour is a person or depositor. Duplicate trays if a torn 16 parks the SKU.

Jump to steel LSR. T_tool jumps. C_pour drops *if* you needed *parts* and yield holds. C_setup and MOQ jump. This wins only when the object is a vulcanized article, H’s part count keeps that press fed, and the product is frozen. Otherwise you bought a low C_press you never use, which is a high TCO — and you still cannot dump pectin.

A crossover sketch (no currency theatre)

Suppose (qualitatively) a jacketed 1-cavity RTV already runs the dump. An 8-cavity twin adds T_tool. You need enough dumps in H for (bench hours saved + trim hours saved + scrap avoided) > (8-cavity − 1-cavity) + extra setup if batches get larger + extra sag if the sheet is floppy.

If the dump is thick (long set), bench hours saved are the story. If the dump is flash-prone, trim and scrap are the story. If the mascot is tiny and already 20 seconds, you need volume or you are buying cavities for pride.

Plot two volumes: your *likely* H and your *sales-deck* H. If 16-cavity only wins on the deck, don't pour it. Pour 8, or a 1-cavity you can still love at the likely volume.

What to ask when you're buying

• Two constructions, same dump, same flash spec, 12-month qty *and* batch qty?
• Object named: RTV cavity vs LSR part?
• Cycle and whether set dominates?
• Trim method, scrap assumption, setup per batch, recast interval?
• T_mod placeholder if CAD is not frozen?

Worked example: aluminum-thinking vs jacketed RTV

A brand compared “steel 4-cavity Class 103” to “RTV dump price.” Object was a 12-cavity 18 mm bear, ~2,000 dumps a month, food-contact. Honest TCO: CNC master + platinum 10A–20A skin mix 1:1 vacuum degas + jacket + one recast in 12 months + trim near zero vs a sagging glove that hired knives. Steel LSR TCO was a different product. We would not have amortized 103 across a season of pectin.

When this article is the wrong quote

If you already run LSR *parts* on aluminum and want steel, fill C_labor and C_scrap for the tool you have — that TCO belongs on a parts OEM. A tiny artisan can skip nine-term algebra. Geometry can override quantity: one deep core can dominate T_tool at 80 dumps.

Related reading: custom RTV molds, hidden costs, when to leave RTV, gummy trays.

Ways TCO lies

• Amortizing steel across a lifetime the product does not have — and the object is a tray.
• Ignoring T_mod because “we’ll only change the logo” (you will change the logo; recast the insert).
• Counting 8-cavity yield as 1-cavity yield.
• Forgetting leftover food-contact platinum kits on small batches.
• Treating freight of a bulky jacket as zero because it is “NRE.”
• Using a unit price that assumed one giant lot while you will kanban weekly.

What to demand on the quote so TCO is computable

Two constructions, same dump, same flash spec, 12-month dumps *and* batch qty:

• Mold (master, cavities, jacket, recast)
• Pour+set time and whether set dominates
• Dump waste per piece
• Trim method
• Scrap assumption
• Setup per batch
• Stated recast interval

Then you can add. If they only give unit price at 10,000, you cannot do TCO. You can only do hope.

The test: freeze H, freeze batch size, fill the nine terms for 1-cavity jacketed RTV and 8-cavity jacketed RTV. The lower sum wins. If you cannot fill C_labor and C_scrap, you are not ready for more cavities; you are ready for a better tryout on the tray you have. If the competing quote is steel LSR, you are not ready for TCO — you are ready to name the object.

Send likely H, batch size, and dump material. We will fill the RTV terms and refuse to blend a 103 into the sum.

We would not recommend amortizing Class 103 across a seasonal gummy

We would not recommend comparing steel mold price to RTV dump price. We would not ignore T_mod. Raising Shore A first is the wrong fix for high C_labor; fix the split. We would not jump to 8-cavity to decorate a TCO slide while the jacket still sags.

RUUIPON Mold Structure Decision: TCO follows structure. Flat back? open-back is cheaper to recast. Deep undercuts that will not peel? two-part / core in T_tool. Soft 10A–20A or wet dump? mother mold in T_tool, not as a surprise in C_scrap.

flowchart TD
  A[TCO horizon H] --> B{Object is a cavity?}
  B -->|no, LSR part| C[Stop: parts TCO]
  B -->|yes| D[Fill nine RTV terms]
  D --> E{CAD frozen?}
  E -->|no| F[Keep T_mod large; delay extra cavities]
  E -->|yes| G[1-cavity vs 8-cavity at real batch]
  G --> H{Jacket flat?}
  H -->|no| I[Fix C_scrap before adding pockets]

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 | | Cavity Shore | 10A–20A skin with jacket | | Platinum shrink | About 0.1–0.3% | | Loaded-tray edge sag | About 0.5–1.5 mm without a jacket — a C_scrap driver | | Outer-row vs center if tray sags | Edge often 3–6% heavy | | Horizon | 12 or 24 months is enough for most custom trays |

FAQ

Why not compare steel mold price to RTV dump price? TCO is construction plus the first planned geometry change plus pour, material, labor, scrap, setup per batch, recast, and freight over a horizon. Dump price is one term. Optimizing it alone buys an 8-cavity monument for a seasonal SKU, or a 103 that cannot dump food. Compare TCO of both constructions at the same dump and flash spec — and only if both are cavities.

What is T_mod and why isn't it zero before freeze? It is the first planned CAD change. Printed masters make it small; a finished CNC master with art is medium; hardened steel *part* tools make it large. "We'll only change the logo" is still a placeholder recast. Ignoring it is a classic TCO lie.

When does jumping to 8-cavity RTV lose TCO? When the horizon's dump count cannot keep that depositor fed, batches get larger, yield drops versus 1-cavity, or the mascot is unfrozen. You bought a low pour cost you never use, which is a high TCO. Jacketed 8-cavity wins only if bench hours, trim, and scrap avoided beat the tray delta inside the horizon.

How does weekly kanban lie about a pretty 10,000-dump unit price? Setup returns every release. Leftover platinum kits on small batches, freight of a bulky jacket treated as free NRE, and amortizing steel across a lifetime the product does not have also lie. Plot likely volume and sales-deck volume; if 16-cavity only wins on the deck, don't pour it.

What missing numbers mean I am not ready for more cavities yet? If you cannot fill trim labor and scrap for the tray you already run, you need a better tryout, not a pocket jump. Demand set time (and whether set dominates), dump waste, trim method, scrap assumption, setup per batch, and recast interval on both constructions. If the competing quote is steel LSR, name the object first.