High-temperature silicone (VMQ) O-rings hold a service envelope of -60 °C to +230 °C continuous, with intermittent excursions to +250 °C. Post-cured platinum-cured grades read ≤ 25 % compression set after 22 hours at 175 °C on ASTM D395 Method B — the single durability number that separates a real high-temp seal from a compound-name claim. For OEM programs, expect $2,000-$5,000 aluminum tooling, MOQ 3,000-10,000 pcs on custom cross-sections, 7-15 day samples, and 25-35 day production out of Dongguan.
This guide is written for buyers and design engineers specifying silicone O-rings that will see steam, hot air, hot water, CIP/SIP cycles, or repeated autoclave exposure. It covers the material decision (silicone vs FVMQ vs FKM vs FFKM vs EPDM), the failure modes that show up at temperature, the post-cure protocol most factories skip, and the OEM commercial realities — tooling cost, MOQ, lead time, and the paperwork buyers should insist on. All numbers are traceable to ASTM, ISO, FDA, or USP standards cited at the bottom. ## What temperature range can a silicone O-ring actually handle?Standard VMQ silicone O-rings run -60 °C to +230 °C continuously in dry air and tolerate +250 °C intermittently. In steam or hot water the practical ceiling drops to +180 °C for long life. PVMQ (phenyl-modified) extends the low end to -100 °C for aerospace. Above +250 °C continuous, silicone hardens and cracks — switch to FKM or FFKM.
The temperature envelope is the reason silicone gets specified in the first place. Compared to the four other common O-ring elastomers, VMQ is the widest-envelope, most temperature-forgiving option: | Material | Continuous max | Intermittent max | Low-temp limit | Notes | |---|---|---|---|---| | NBR (Buna-N) | +100 °C | +120 °C | -30 °C | Cheap, oil-compatible, narrow envelope | | EPDM | +150 °C | +175 °C | -50 °C | Steam and hot water; fails on oils | | VMQ (silicone) | +230 °C | +250 °C | -60 °C | Widest envelope; fails on fuels | | FKM (Viton) | +230 °C | +250 °C | -20 °C | Fuels/oils/solvents; brittle cold | | FFKM | +327 °C | +350 °C | -25 °C | Semiconductor / aggressive chem; 20-40 × cost | Temperature rating is only half the spec. What actually matters at elevated temperature is how much of the original squeeze the ring recovers after prolonged compression — the compression-set metric covered in the next section. A material rated to +230 °C that returns 55 % compression set is not a high-temperature seal; it's a heat-tolerant part that will leak. ## How does compression set determine long-term seal life?Compression set is the permanent deformation left after prolonged squeeze at temperature — the single failure metric that predicts O-ring service life. A post-cured platinum-cured VMQ holds ≤ 25 % after 22 hours at 175 °C on ASTM D395 Method B. The same compound without post-cure reads 45-55 %. Anything above 30 % is a leak waiting for the next thermal cycle.
Compression set is measured per ASTM D395[^astm-d395] — a puck of the compound is squeezed 25 % in a fixture, held at temperature for 22 hours (Method B, constant deflection), released, and re-measured. The set percentage is how much of the original squeeze the material did not recover. For high-temperature O-rings, this number matters more than tensile strength or elongation because seals fail by permanent deformation long before they tear. Real numbers from platinum-cured 70 Shore A VMQ compression-molded at Wetop, tested at an independent lab: | Test condition (ASTM D395 Method B) | Compression set | Verdict | |---|---|---| | 22 h @ 100 °C, post-cured | 8 % | Baseline — food/pharma OK | | 22 h @ 150 °C, post-cured | 15 % | Steam / CIP OK | | 22 h @ 175 °C, post-cured | 22 % | High-temp OEM spec — passes | | 22 h @ 200 °C, post-cured | 34 % | Marginal — expect field returns | | 22 h @ 175 °C, NO post-cure | 48 % | Fail — do not ship | If a supplier will not put a compression-set number on their COA, the answer is no. The test is standard, cheap, and the only rigorous predictor of high-temperature seal life. Buyers who specify silicone O-rings for hot programs should require Method B data at the intended service temperature on every production lot — see the [silicone O-ring specifying guide](/guide/silicone-o-ring-specifying-guide/) for the full RFQ checklist. ## Why is the 4-hour post-cure at 200 °C non-negotiable?Peroxide-cured VMQ carries residual peroxide breakdown products that plasticize the crosslink network and drive high compression set. A 4-hour bake at 200 °C in a convection oven volatilizes those residues below 0.5 %, unlocks LFGB §30/§31 and FDA 21 CFR 177.2600 extraction compliance, and cuts compression set from 50 % to 25 %. Skipping post-cure is the #1 field-failure root cause.
Peroxide-cured silicone (typically dicumyl peroxide or 2,4-dichlorobenzoyl peroxide catalyst) leaves 1-3 % residual byproducts after the initial mold cure. Those byproducts include benzoic acid, chlorobenzoic acid, and cumyl alcohol — plasticizers that soften the network and cause the compound to fail extraction testing under LFGB §30/§31. The post-cure step drives them off: 1. Load molded rings on stainless-steel screens, single layer 2. Convection oven, forced air, 200 °C 3. Hold 4 hours (some specs require 8 hours for medical / pharma) 4. Ramp-down slowly to avoid thermal shock crack 5. Weigh a coupon sample before / after — verify < 0.5 % mass loss Platinum-cured VMQ (Pt-catalyzed hydrosilylation) does not require post-cure for compliance — the cure system produces no volatile byproducts. But a shorter 1-hour bake at 200 °C still lowers compression set another 3-5 percentage points, and buyers on tight sealing budgets should require it. See the [platinum-cured vs peroxide-cured silicone comparison](/guide/platinum-cured-vs-peroxide-cured-silicone/) for the full cure-system decision matrix. ## Silicone vs FKM (Viton) vs FFKM — which to choose for high-temperature service?Silicone wins on hot dry air, hot water, CIP/SIP steam, ozone, UV, and food/medical certifications. FKM wins on fuels, mineral oils, brake fluid, and aromatic solvents. FFKM wins on aggressive chemistry above +230 °C at 20-40 × the cost. If the fluid is inert and hot, spec silicone. If the fluid is a hydrocarbon, spec FKM. If it's hot and chemically aggressive, spec FFKM.
The material decision at elevated temperature is really a fluid-compatibility decision. Both silicone and FKM are rated +230 °C continuous — the temperature rating alone does not pick between them. What separates them is what the ring is sealing against: | Fluid | Silicone (VMQ) | FKM (Viton) | FFKM | |---|---|---|---| | Hot dry air +230 °C | Excellent | Excellent | Excellent | | Hot water / steam +150 °C | Excellent (peroxide) | Fair (blistering) | Excellent | | Mineral oil / hydraulic | Poor (swells 20-40 %) | Excellent | Excellent | | Gasoline / jet fuel | Poor | Excellent | Excellent | | Aromatic solvents (toluene) | Poor | Good | Excellent | | Aqueous acids (dilute) | Good | Excellent | Excellent | | Ozone / UV / weathering | Excellent | Good | Excellent | | Food / medical certs | FDA + USP VI + LFGB | Some FDA grades | Some FDA grades | | Cost per ring (indexed) | 1.0 × | 3-6 × | 20-40 × | The correct answer for most high-temp OEM programs — steam-in-place food processing, autoclaves, hot-air ducting, steam trap covers — is post-cured platinum-cured VMQ. FKM only wins when hydrocarbons enter the picture. FFKM only wins when the customer is a semiconductor fab or a chemical reactor operator with a hard operating cost per unplanned downtime. ## What AS568 sizes and custom cross-sections should OEM programs specify?AS568 dash numbers cover 90 % of catalog specs — the -100 through -400 series encode inner diameter and cross-section on a standard grid. Custom cross-sections require a new aluminum compression mold at $2,000-$5,000, amortized against 3,000-10,000 pcs MOQ. For metric-only markets use ISO 3601-1 G-series. Silicone glands need 15-25 % squeeze and 65-80 % gland fill — 5-10 % more clearance than NBR.
The SAE AS568 standard[^sae-as568] is the dash-number sizing system that every US catalog uses. Dash number encodes cross-section: - **-001 to -050:** 1/32 in (0.79 mm) cross-section - **-100 series:** 3/32 in (2.38 mm) cross-section - **-200 series:** 1/8 in (3.18 mm) cross-section - **-300 series:** 3/16 in (4.76 mm) cross-section - **-400 series:** 1/4 in (6.35 mm) cross-section Cite the RFQ as `ID × CS` (e.g., 25.0 × 3.0 mm) and reference the closest AS568 dash if catalog interchange matters. For pure-metric markets, use ISO 3601-1[^iso-3601] G-series (general purpose) or R-series (aerospace) — the ISO counterpart to AS568. Silicone gland design differs from nitrile by 5-10 % more radial clearance and gland fill. Wetop's default squeeze target is 20 % radial for static seals and 12 % for dynamic seals, gland fill 75 %. Higher fill (> 85 %) leaves nowhere for thermal expansion to go; the ring extrudes and takes permanent set. Lower fill (< 65 %) allows the ring to nutate and the seal to weep. ## What does OEM tooling cost, MOQ, and lead time actually look like?OEM tooling for compression-molded silicone O-rings runs $2,000-$5,000 for a 16-64 cavity aluminum mold, amortized against 3,000-10,000 pcs MOQ on custom cross-sections. Samples ship 7-15 days after tool sign-off. Production runs 25-35 days for compression, 35-50 days for LSR injection. Tooling is customer-owned and IP-transferable at program end.
The commercial reality that competitor pages universally omit: | Process | Tooling cost | MOQ | Sample lead | Production lead | Tolerance | Best for volume | |---|---|---|---|---|---|---| | Compression molding | $2,000-$5,000 | 3,000-10,000 | 7-15 days | 25-35 days | ±0.1 mm | 500-100,000 pcs | | Injection HCR | $3,000-$8,000 | 10,000+ | 10-18 days | 30-40 days | ±0.05 mm | 50,000-500,000 | | LSR injection | $8,000-$25,000 | 50,000+ | 15-25 days | 35-50 days | ±0.03 mm | 250,000+ per size | For most high-temperature OEM programs shipping 500-100,000 rings per year, compression molding is the right process. Multi-cavity aluminum tools (16-64 cavities) recover investment fast, and the ±0.1 mm tolerance is fine for standard AS568 glands. LSR only makes sense above ~250,000 pcs per size or when zero flash is a hard spec (medical device implants, precision fluidics). Every legitimate OEM program should require: material COA per lot, Shore A hardness reading, compression-set data at intended service temperature, post-cure log, and a tooling handover clause. For the full sourcing checklist see the [sourcing silicone factory checklist](/guide/sourcing-silicone-factory-checklist/). ## What does the OEM cost structure look like at 500 / 5,000 / 50,000 pcs?Piece price for a custom high-temperature silicone O-ring drops from ~$1.20 at 500 pcs to ~$0.35 at 5,000 pcs to ~$0.12 at 50,000 pcs, mostly through tooling amortization. Cure system (platinum vs peroxide) adds 15-25 %. Post-cure adds ~5 %. Certification packages (FDA + LFGB + USP VI COA) add $200-$500 per lot fixed.
Cost drivers in descending order for a typical 25 × 3 mm high-temperature silicone O-ring, 70 Shore A, platinum-cured, post-cured: 1. **Tooling amortization** — $3,000 tool ÷ MOQ. At 500 pcs = $6.00/pc. At 5,000 pcs = $0.60/pc. At 50,000 pcs = $0.06/pc. 2. **Base compound** — platinum-cured food-grade VMQ runs $12-$18/kg vs $9-$13/kg for peroxide-cured. A 25 × 3 mm ring uses ~1.2 g of compound. 3. **Post-cure** — oven time, energy, and yield loss add ~5 % to piece price. Not optional. 4. **Cavity yield** — 16-cavity mold at 4-minute cycle = 240 shots/hour = 3,840 rings/hour theoretical. Real-world yield 85-92 %. 5. **QC and paperwork** — Shore A / compression-set / dimensional COA per lot: $200-$500 fixed cost. Larger lots dilute this. None of the above is opaque — a legitimate factory will show a line-item breakdown when asked. If a supplier offers a rock-bottom quote without tooling amortization visible, either the tool is under-cavitied (long cycle, poor consistency), the compound is generic HCR without traceability, or post-cure is being skipped. Any of the three shows up as a compression-set failure six months into the program. ## FAQ *(Rendered from the frontmatter FAQ block above.)* ## Closing High-temperature silicone O-rings are not a commodity spec. The material name on the drawing — VMQ 70 Shore A, FDA-grade — is roughly 30 % of the actual seal life. The other 70 % is cure system, post-cure discipline, compression-set verification, and tooling that matches the volume. Wetop runs both peroxide and platinum lines out of Dongguan with per-lot COA and post-cure logs on the QMS[^iso-9001], and ships samples in 7-15 days from AS568 dash spec or a custom cross-section drawing. To scope an OEM program — material grade, tooling estimate, MOQ, lead time — [contact the engineering desk](/contact/) with your drawing (or an AS568 dash and Shore A target). We reply within one business day with a costed proposal and a signed NDA on request. [^astm-d395]: See References — ASTM D395-18 Compression Set. [^sae-as568]: See References — SAE AS568E. [^iso-3601]: See References — ISO 3601-1:2012. [^iso-9001]: See References — ISO 9001:2015 QMS Requirements.