Silicone Mold Tolerance & Shrinkage: What Buyers Should Specify
Silicone mold tolerance and shrinkage for chocolate, resin, and wax. Specify dimensions, draft, and material offsets so casts match design intent.
Kyler Yang · Founder6 min read
On this page
A mold can be beautiful and still fail QC if the cast measures 2.3 mm short on a critical snap-fit lip. Silicone mold tolerance and shrinkage are two linked topics: silicone tooling has modest thermal expansion, but your cast material shrinks as it cools or cures - sometimes more than the mold itself. Buyers who specify only "make it 50 mm" without material context get surprises. This guide shows what to put on your RFQ so RUUIPON prototypes (3-7 days, MOQ 50, quote ~24 hrs) hit dimension first pass.
Quick Answer: What Tolerance Should You Expect?
Typical finished cast tolerance from custom RTV silicone tooling: ±0.2-0.5 mm on features under 100 mm for chocolate and wax; ±0.1-0.3 mm achievable on rigid resin post-cure if master was CNC-accurate. Shrinkage is not a single constant - specify target finished part size after your material fully cures, not just cavity size.
Two Shrink Sources: Mold Silicone vs Cast Material
| Source | Magnitude | When it matters |
|---|---|---|
| Platinum silicone cure shrinkage | ~0.1-0.3% | Precision resin masters |
| Silicone thermal expansion | Variable with pour temp | Hot wax injection |
| Chocolate cooling contraction | ~0.5-1.5% ( formulation dependent ) | Bar weights for retail |
| Epoxy/resin cure shrinkage | ~0.5-3% | Thick pours, inlay pockets |
| Wax solidification | ~1-2% | Pillar height, container fit |
We scale masters using offsets you approve after prototype pours. See prototype to production.
Tolerance Bands by Feature Type
- Flat logo emboss depth: ±0.1 mm - visually sensitive.
- Overall length/width: ±0.3-0.5 mm - acceptable in most food SKUs.
- Hole / socket diameter: specify minimum not nominal - resin shrinks inward.
- Wall thickness of cast part: driven by your process, not mold alone.
Compare to steel injection (±0.05 mm) - do not specify injection tolerances on RTV without prototype proof.
Spec Sheet Template for Your RFQ
- Finished part dimensions (after cure/cool) - mm
- Material brand / type ( couverture %, epoxy brand, wax blend )
- Pour temperature range
- Critical vs reference dimensions on drawing
- Acceptable deviation per feature
- Master source ( sample part, STL, CAD )
- Number of prototype iterations included in quote
Email template to [email protected]. Include photos of caliper checks on existing casts if revamping a failed tool.
By Material: Food, Resin, Candle, Soap
| Application | Shrink note | RUUIPON product page |
|---|---|---|
| Chocolate | Cool + fat crystal - weigh bars not just measure once | Chocolate molds |
| Gummy | Moisture loss after demold | Gummy molds |
| Resin | Exothermic cure shrink - post-cure 24h before measure | Resin molds |
| Candle wax | Strong thermal shrink - wick hole offset | Candle molds |
| Soap | Water evaporation over days | Soap molds |
Food programs: Food & Confectionery. Home fragrance: Candle & Soap.
Measuring & QC Protocol
- Measure at consistent temperature (20-23°C lab).
- For chocolate, temper and cool on production rack before caliper.
- Record three pulls per cavity; average.
- Flag outliers - may indicate flash or incomplete fill, not shrink.
- Sign off digital revision before MOQ 50 production.
Multi-cavity layout tolerances: layout guide.
Document shrink as a living number. First prototype pull establishes baseline; second pull after 72-hour post-cure (resin) or overnight cool (chocolate) confirms stability. Do not approve MOQ 50 production on single pull measurement unless geometry is non-critical.
CAD designers exporting STL for masters should include shrink allowance notes in metadata - "finished bar 100 mm, chocolate couverture 70% cocoa, target after cool 100 mm." Factory engineers apply offset outward on master, not inward on hope.
Assembly tolerance stacks when molds feed downstream processes - wick tabs, clamshell inserts, cardboard sleeves. Specify which interface dimension is master-controlled vs derived.
International teams benefit from metric-only drawings. Mixing inch CAD with mm factory invites 25.4× errors on small features - catastrophic on snap-fit wax melt inserts.
QC photography: caliper at feature, mold ID tag visible, date stamp. Builds audit trail for OEM accounts and simplifies WeChat revision threads with supplier.
Shrink varies by fill temperature - hot-poured wax into cold mold vs pre-warmed mold changes external dimensions on same tool. Standardize pour SOP before blaming tool.
Include tolerance table as appendix on purchase order: feature name, nominal mm, +/- band, measurement method. Factory QC signs against that table at first article.
For snap-fit lids on ice sphere molds or storage containers, specify whether tolerance applies to silicone cast or assembled PP lid - different materials, different shrink.
Finite element simulation is overkill for most RTV buyers - prototype iteration at MOQ 50 beats simulation license cost for single SKU launches.
Weight-specified products ( legal-for-trade chocolate ) care about volume via dimensions - caliper plus scale cross-check catches shrink drift scale alone misses.
Tooling revision letter should state which dimensions changed and by how much - prevents "approved" email ambiguity when only logo depth moved 0.2 mm.
Assembly fixtures for multi-cavity demold sometimes need tolerance looser than part spec - distinguish fixture CAD from product CAD in file naming.
Coordinate shrink discussion with RTV vs compression process choice - cast RTV trays and press-made silicone parts do not share the same shrink model even before your cast material applies.
GD&T on drawings optional for craft buyers; plain-language table of five critical dimensions sufficient for MOQ 50 RFQ if each row has measurement tool specified (caliper vs micrometer).
Third-party inspection before balance payment: hire QC agent in Shenzhen for first OEM PO - verifies tolerance table on site before molds fly.
Shrink compensation on multi-cavity trays must be consistent cavity-to-cavity - if cavity 3 measures short, check fill level not only tool dimension; partial fills mimic shrink error.
Digital caliper export to CSV for trend chart - slope on logo width over ten pulls predicts when to reorder before out-of-spec customer complaint.
Reference ISO 3302 for rubber shrink in discussions with engineering-savvy buyers - RTV silicone differs but conversation signals competence.
Temperature-compensated measurement: measure cold chocolate bar at 18°C not warm from cooling tunnel - 2°C difference moves apparent shrink on thin features.
Digital thread: store approved PDF drawing revision with PO number in cloud folder - five years later reorder references correct rev without archaeology.
Communicate nominal vs limit dimensions to marketing - they should not promise "exactly 50.0 mm" on consumer page when spec is 50.0 ±0.4 mm.
Assembly test fixtures: if mold feeds mechanical assembly, ship one fixture with first MOQ 50 for factory-side pre-ship pull test.
Frequently Asked Questions
Can you guarantee ±0.1 mm on chocolate?
Not without material-specific pilot data. Chocolate formulas vary too much.
Should I oversize the master in CAD?
Only with agreed shrink factor from prototype - blind oversize often overcorrects.
Does shore hardness affect shrink?
Minimal on cavity size; flexibility affects demold distortion during measure. See Shore guide.
Two-part mold alignment affect tolerance?
Yes - registration keys critical. See two-part vs one-part.
WhatsApp with caliper photos?
+86 135 3006 2969
Prototype includes dimensional report?
Available on request for OEM accounts.
RTV cast molds vs compression silicone - which for tolerance?
RTV vs compression comparison
Ready to order custom silicone molds?
Free quote for OEM/ODM projects. MOQ 20 · Prototype 3-7 days · FDA/LFGB available.