O-ring Dimensional Specification: AS568 vs Metric vs JIS — Tolerance, Groove Design and Gland Data

Overview #

The specification error that costs procurement teams the most when sourcing O-rings from China is not selecting the wrong elastomer — it is failing to specify the dimensional standard. AS568, metric ISO, and JIS B 2401 are not interchangeable. A cross-section diameter difference of 0.1 mm between an ISO 3601 metric O-ring and its nominal AS568 equivalent is enough to cause gland underfill, reduce squeeze below the 15% minimum threshold, and produce a seal that passes visual incoming inspection but fails under pressure cycling within 200 hours of service. We have seen this exact failure mode in hydraulic manifold assemblies sourced from Chinese suppliers who defaulted to metric tooling when the drawing specified AS568.

Dimensional Standard Comparison: AS568, ISO Metric, and JIS B 2401 #

The three dominant O-ring dimensional standards — AS568 (SAE Aerospace Standard), ISO 3601 (metric), and JIS B 2401 (Japanese Industrial Standard) — define cross-section diameter (CS) and inside diameter (ID) differently, with tolerances that do not align even when nominal sizes appear close.

AS568 is the default standard for North American OEM equipment and most US-designed hydraulic and pneumatic systems. It defines sizes by a three-digit dash number (e.g., -214), with CS and ID tolerances governed by ASTM D2000 classification requirements and dimensional tables in the SAE AS568 standard itself. For a -214 O-ring, the nominal CS is 3.53 mm ± 0.10 mm and the ID is 36.17 mm ± 0.38 mm.

ISO 3601-1 defines metric O-rings in four tolerance grades: Grade N (general purpose), Grade S (precision), Grade E (extra precision), and Grade L (large diameter). For a comparable metric size (e.g., 36.0 × 3.5 mm), the ISO 3601 Grade N CS tolerance is ± 0.10 mm and ID tolerance is ± 0.36 mm — close to AS568 but not identical. The cross-section nominal of 3.5 mm vs. AS568’s 3.53 mm is a 0.03 mm difference that accumulates in tight gland designs.

JIS B 2401 uses a P-series (piston/cylinder) and G-series (general) classification. JIS P35 has a CS of 3.5 mm and ID of 34.4 mm — a different ID entirely from the AS568 -214 or ISO 36 × 3.5. Substituting JIS for AS568 without gland redesign is a dimensional mismatch, not a tolerance issue.

Most Western buyers do not realize that Chinese O-ring manufacturers default to ISO metric tooling unless AS568 is explicitly specified and verified. A drawing note that says “O-ring per AS568-214” without a dimensional inspection requirement on the purchase order will frequently result in delivery of an ISO 36 × 3.5 part — which looks identical, measures within 0.03 mm on CS, but may not meet gland squeeze requirements on the original equipment design. This is a ISO 3601 compliance gap that standard COA documentation will not reveal.

For buyers sourcing O-rings and static seals from Chinese suppliers, the first document to request is not the material COA — it is the dimensional inspection report showing actual measured CS and ID values against the specified standard’s tolerance table, with sample size and AQL level stated.

Elastomer Selection Against Operating Conditions: Critical Thresholds #

Elastomer grade selection is where most procurement teams over-rely on supplier recommendations rather than applying defined selection criteria. The correct framework is to work from four parameters in sequence: continuous service temperature, fluid compatibility, dynamic vs. static application, and compression set requirement. Every one of these has a numeric threshold that determines pass/fail at the material level.

Temperature range is the first filter. NBR (nitrile) is rated for continuous service from -40°C to +120°C in petroleum-based fluids. EPDM covers -50°C to +150°C but is incompatible with petroleum oils — a substitution error we have seen cause catastrophic seal failure in hydraulic systems where the buyer specified EPDM for temperature range without checking fluid compatibility. FKM (Viton) handles -20°C to +200°C continuous, with short-term excursions to +230°C. Silicone (VMQ) covers the widest temperature range (-60°C to +200°C) but has poor tensile strength (typically 6–9 MPa vs. NBR’s 10–20 MPa) and is unsuitable for dynamic applications.

Compression set is the parameter that determines long-term sealing performance and is the one most often missing from Chinese supplier COAs. Per ASTM D395 Method B, acceptable compression set for a static seal elastomer is ≤ 25% after 70 hours at the rated service temperature. For dynamic seals, the threshold tightens to ≤ 15%. FKM at 175°C/70h typically measures 12–18% compression set; NBR at 100°C/70h typically measures 18–28%. When we qualify Chinese O-ring suppliers, we reject any batch where compression set exceeds 25% on static-grade material or 15% on dynamic-grade material — regardless of what the hardness reading shows.

Shore A hardness for O-ring applications is standardized at 70 Shore A for general-purpose static seals, with 60 Shore A used for low-pressure or low-temperature applications and 80–90 Shore A for high-pressure dynamic applications (above 10 MPa system pressure). Per ASTM D2000, hardness tolerance is ± 5 Shore A points. In our supplier qualification program, we reject batches where Shore A deviates more than ± 3 points from the specified grade — tighter than the ASTM allowance — because lot-to-lot hardness variation beyond ± 3 points correlates with inconsistent compound mixing, which also affects compression set.

Most procurement teams focus on unit price when sourcing O-rings from China. The variable that actually drives total cost is rejection rate at incoming inspection — and that is determined by dimensional tolerance class and compression set consistency, not by price per piece. A Grade S ISO 3601 O-ring costs 15–30% more than Grade N from the same Chinese supplier. If your incoming inspection AQL 2.5 rejection rate drops from 3% to 0.4%, the Grade S premium pays back within two production batches.

For related sealing components used in rotary and dynamic applications, see oil seals and rotary seals for shaft seal selection criteria that apply the same elastomer qualification logic.

Gland Design Parameters and Groove Tolerances #

Gland design is where dimensional standard selection becomes a physical engineering constraint. The groove dimensions must be matched to the O-ring standard — not just the O-ring size. Using AS568 gland dimensions with an ISO metric O-ring, or vice versa, changes the diametral squeeze and the void fill percentage, both of which determine whether the seal functions.

For a static face seal application using a 70 Shore A O-ring, the target diametral squeeze is 15–25% of the O-ring cross-section diameter. For a -214 AS568 O-ring (CS = 3.53 mm), this means a groove depth of 2.65–2.99 mm. For the ISO 36 × 3.5 equivalent (CS = 3.50 mm), the same squeeze percentage requires a groove depth of 2.63–2.98 mm — a 0.02 mm difference that is within machining tolerance but confirms the standards are not dimensionally identical.

Groove width (for static applications) should provide 50–70% void fill at operating temperature. For a 3.53 mm CS O-ring, groove width is typically 4.8–5.3 mm. Groove surface finish should be Ra ≤ 1.6 µm for static seals and Ra ≤ 0.4 µm for dynamic seals per ISO 3601-2 gland design recommendations.

In our qualification program, we have seen suppliers pass initial sample approval with correctly dimensioned O-rings and then deliver out-of-spec material at production volume. The trigger is almost always a raw material substitution at the compounder level — a change in carbon black loading or plasticizer content that shifts Shore A by 4–6 points and compresses the compression set performance outside the acceptable window. A standard COA showing hardness within ± 5 Shore A will not catch this without incoming compression set spot-testing per ASTM D395 Method B.

The English technical content available for O-ring gland design from Chinese suppliers is almost entirely absent. Chinese manufacturers publish dimensional tables but rarely publish gland design data in English, and almost never publish lot-to-lot compression set data. That gap is precisely why buyers who rely on supplier-provided documentation alone end up with seals that pass incoming inspection and fail in service.

Practical Guidance for Buyers #

When sourcing O-rings from Chinese suppliers, the first specification to request is not the material COA — it is the dimensional inspection report showing actual measured CS and ID values against the tolerance table of the specified standard (AS568, ISO 3601 Grade N or S, or JIS B 2401), with sample size and AQL level stated. Most buyers ask for hardness first. Hardness is the easiest parameter to pass on a COA and the least predictive of sealing performance.

The most common sourcing mistake is accepting a “metric equivalent” substitution when AS568 is specified. A 0.03 mm CS difference between AS568 -214 and ISO 36 × 3.5 sounds marginal. In a gland designed for AS568 with a 15% minimum squeeze requirement, it can reduce actual squeeze to 14.1% — below threshold — and produce a seal that weeps under pressure cycling within 200 hours.

Before committing to volume order, require three consecutive batch COAs showing compression set per ASTM D395 Method B at the rated service temperature, with results ≤ 25% for static applications or ≤ 15% for dynamic applications. Require dimensional inspection data at ISO 3601 Grade S tolerance (CS ± 0.07 mm, ID ± 0.25 mm) if your gland design has less than 20% squeeze margin. A supplier who cannot provide three consecutive batch compression set results is not qualified for volume supply, regardless of price.

Frequently Asked Questions #

Q1: What is the most critical dimensional parameter to verify when sourcing O-rings from China?
A: Cross-section diameter (CS) measured against the specified standard’s tolerance table — not Shore A hardness. A 0.03 mm CS deviation from AS568 nominal can reduce gland squeeze below the 15% minimum threshold without triggering a hardness failure.

Q2: Can I substitute an ISO 3601 metric O-ring for an AS568 size without redesigning the gland?
A: Not reliably. For the -214 / 36 × 3.5 pair, the CS difference is 0.03 mm and the ID difference is 0.17 mm. Whether this is acceptable depends on your gland’s squeeze margin. If your design has less than 20% squeeze, request Grade S tolerance per ISO 3601 (CS ± 0.07 mm) and verify actual measured dimensions before approving the substitution.

Q3: Why do O-rings that pass incoming inspection fail in service within 200 hours?
A: This is where most sourcing decisions go wrong. The threshold is compression set ≤ 25% after 70h at rated temperature per ASTM D395 Method B. Incoming inspection typically checks hardness and dimensions — neither catches a compound with degraded compression set performance caused by raw material substitution at the compounder level.

Q4: What certifications and test documentation should I require before approving a Chinese O-ring supplier for volume orders?
A: Require three consecutive batch COAs showing: Shore A hardness per ASTM D2000 (± 5 Shore A, or ± 3 Shore A if you apply tighter incoming criteria), compression set per ASTM D395 Method B at rated service temperature, and dimensional inspection data with sample size and AQL level stated against the specified standard (ISO 3601, AS568, or JIS). A supplier who cannot provide all three for three consecutive batches is not ready for volume qualification.

Q5: Is a higher Shore A hardness always better for high-pressure applications?
A: No. 80–90 Shore A is appropriate above 10 MPa system pressure to resist extrusion. Below 10 MPa, 70 Shore A provides better conformance to surface finish irregularities and lower breakout friction in dynamic applications. Specifying 90 Shore A for a 5 MPa static application is a common over-specification that increases leak risk, not reduces it.

What to Specify in Your O-Ring Purchase Order — Checklist

A complete O-ring purchase order specification must include all of the following to prevent substitution, dimensional mismatch, and material non-conformance:

• Dimensional standard and size designation: State AS568 dash number, ISO 3601 size (ID × CS in mm) with tolerance grade (N, S, or E), or JIS B 2401 series and number — never “metric equivalent acceptable”
• Elastomer grade: Specify compound type (NBR, FKM, EPDM, HNBR, VMQ) and ASTM D2000 line callout (e.g., M2BG714 for 70 Shore A NBR)
• Shore A hardness: Nominal value ± 5 Shore A per ASTM D2000, or ± 3 Shore A if tighter incoming inspection applies
• Compression set requirement: ≤ 25% (static) or ≤ 15% (dynamic) after 70h at rated service temperature per ASTM D395 Method B — required on COA
• Operating conditions: Continuous service temperature range, fluid media, system pressure (MPa), and whether application is static or dynamic
• Dimensional inspection requirement: Actual measured CS and ID values on COA, sample size and AQL level stated, tolerance per specified standard
• Lot traceability: Batch number, compound lot number, and compounder identity on COA

Published by sinoraw.com Technical Team | Request a sourcing consultation