CNC vs 3D-Printed Masters for Low-Volume RTV Molds

If you're making a low-volume custom RTV mold, aluminum CNC is the default *master*. A 3D-printed master is a fast exception with a finish tax, an inhibition tax, and a durability limit. It is not a cheaper aluminum tool, and it is not a printed LSR injection cavity.

LSR is a platinum thermoset shot into a *hot* cavity. That fork is a parts RFQ. This article is the RTV analog: you pour Shore 10A–20A platinum or tin against a master, mix 1:1 or 10:1, vacuum degas, often with a jacket. Any master that cannot hold finish, size, and chemistry will lie to you about peel, shrink (~0.1–0.3% platinum), and cosmetics.

What a printed master is allowed to prove

Printed masters for RTV exist. Shops use SLA/DLP resin, sometimes sintered metal, sometimes a printed jacket. The honest job is: peel study, split location, undercut release, “does this lip even invert,” and a handful of dumps in the production compound.

What it is not allowed to prove, except by luck:

• A stable fill-weight over hundreds of dumps if the print creeps or the jacket is the print
• Wear life on a textured or polished cosmetic face
• Multi-cavity sag balance on a floppy printed sheet
• Any claim about production depositor cycle time

Platinum RTV inhibits on undercured resin prints. Full UV post-cure, wash, and a barrier coat are part of tooling. Tin is more forgiving of dirty masters and keeps shrinking in storage. Don't pick tin so the print “works.”

Printed *metal* improves hardness versus photopolymer and still loses to cut aluminum on polish and the first ding. Treat it as a fast master, not as a Class 104 analog.

Common mix-up: quoting a printed LSR insert as an RTV mold

A printed polymer *cavity* for hot LSR sags, creeps, and flashes. Heat is the limiter. That is not this product. If the PDF has heaters, clamp, and a cold-runner, you are on the wrong quote. Ask for a master + RTV skin.

ISO 3302-1 does not decide CNC vs print. It is a drawing class if the article is rubber.

What CNC aluminum (or machined resin) is for

A CNC aluminum master is bridge and production RTV. You cut real fillets. You can polish. You can recast the skin against the same metal for years if you don't sand it “to clean it up.” You can add a steel core on a hollow.

Limits, said plainly:

• Aluminum dings. A dropped wrench is a witness mark on every later dump.
• A metal master is colder than a resin print; chocolate set can shift. Coupon it.
• Lifetime of the *skin* is still dumps plus recast. The master is the library.

If the program is a few hundred dumps, design almost frozen, CNC or a fully finished print can be the whole answer. If the program is tens of thousands of dumps, the master should be CNC (or metal) so recasts stay honest.

The three limits that decide, in shop order

Check in this order. Stop at the first fail for print.

Inhibition and finish. Will platinum see a fully washed, UV-post-cured, barrier-coated print? Layer lines are texture whether you specified one or not. If the dump is cosmetic food, do not ship voxels.

Size hold. SLA that is still post-curing is not frozen. Platinum will not shrink you back to CAD. CNC holds the window.

Abrasion and recast. Every peel is a wear event on the *skin*, but a soft print used as a production master dents. Budget the print as a disposable pattern or do not use it for the library master.

If all three are acceptable for a dozen to a few score dumps, print. Otherwise CNC.

What to ask when you're buying

• Is the quote a printed *master* plus RTV skin, or a printed LSR *cavity*?
• Full UV post-cure, wash, barrier coat on resin?
• Who keeps the master for recast — metal or print?
• Jacket: CNC/printed shell, or none?
• Flash spec and fill-weight applied to the RTV dumps, not to a hot LSR shot?

Worked example: 12-cavity bear, print then CNC

A 12-cavity 18 mm gummy, 4.0 g, about 2,000 dumps a month. Printed master to kill the chin peel (ten demolds, no white mark). Then CNC aluminum master for the library, platinum 10A–20A skin, mix 1:1, vacuum degas, rigid jacket. We would not have left the SLA as the recast pattern after a logo freeze. A parallel printed LSR insert does not dump pectin.

When this article is the wrong quote

If you actually need low-volume LSR injection, aluminum CNC *cavities* with heaters are the default bridge — different RFQ, different heat tax. A one-off artisan tray from a physical chocolate can skip both. Geometry can override quantity: a deep core still wants metal at 80 dumps.

Related reading: custom RTV molds, turnaround times, converting STEP, gummy trays.

A practical split that shops actually use

1. Print a master (or a core) to kill geometry risk. Barrier-coat if platinum.
2. Cut CNC aluminum for the library master once peel and logo freeze.
3. Recast the RTV skin as it wears. Don't jump to steel *part* tools unless you need molded parts.

Skipping step 2 because the print “looked fine” is how you recast against a dented pattern.

Refuse: a multi-cavity printed *LSR* tool quoted as this mold; a printed master quoted as “production” with a food fill-weight and no barrier coat; CNC with no jacket plan on a large wet dump.

RFQ sentence: “Quote a CNC master plus platinum RTV skin with jacket; optionally quote a printed master for geometry trials only, with inhibition control stated. Do not quote a printed LSR injection cavity as this mold.”

Send the STL and dump material. We will mark print vs CNC and whether the print may stay as the library master.

We would not recommend a printed LSR cavity as an RTV mold

We would not recommend leaving an undercured SLA in platinum. We would not skip the jacket on a large printed sheet. We would not raise Shore A first to hide voxel noise. We would not cut Class 103 steel “while we are here” if you still dump chocolate.

RUUIPON Mold Structure Decision: master material after peel is proven. Flat back? open-back. Deep undercuts that will not peel? two-part / core — print or CNC does not remove a lock. Soft 10A–20A or wet dump? mother mold, even if the master was cheap.

flowchart TD
  A[Low-volume RTV?] --> B{Printed or CNC master?}
  B -->|print| C{Platinum + full post-cure + barrier?}
  C -->|no| D[Inhibit or own the voxels]
  C -->|yes| E[Geometry trials; recast library?]
  B -->|CNC| F[Library master; recast skins]
  A --> G{Quote is printed LSR cavity?}
  G -->|yes| H[Stop: different part RFQ]

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%; print error is usually larger | | Loaded-tray edge sag | About 0.5–1.5 mm without a jacket | | Print role | Tens of dumps to freeze peel; CNC for the library | | 12-cavity 18 mm bear | ~2,000 dumps/month after freeze is a common band |

FAQ

What is a printed RTV master actually allowed to prove? Peel, split, undercut release, whether a lip inverts, and a handful of dumps in the production compound. It is not allowed to prove a stable fill-weight over hundreds of dumps if the print creeps, texture life, or production depositor time.

Why does platinum RTV fail on many resin prints? Inhibition on undercured SLA, plus layer lines that become texture. Full UV post-cure, wash, and a barrier coat are tooling. Tin forgives dirty masters and then keeps shrinking.

Does printed metal replace CNC aluminum for the library master? It improves hardness versus photopolymer and still loses to cut aluminum on polish and the first ding. Treat sintered masters as fast patterns, not as a Class 104 analog.

Should I print a multi-cavity RTV sheet to get more samples? You can for geometry. Do not stack sag problems on a consumable jacket. Stay at 1–2 cavities until fill-weight is real, then copy onto a rigid mother mold.

What RFQ split keeps print and CNC from blending? Quote CNC master plus RTV skin with jacket. Optionally quote a printed master for geometry trials only, with inhibition control stated. Apply fill-weight and food-contact to the RTV dumps, not to a printed LSR insert.