TL;DR: Bore tolerance class and housing fit — not lip material — is the variable that most frequently drives oil seal leakage at installation, and it is almost never specified on a Chinese supplier’s standard quote.
TL;DR: Across 31 housing bore measurements on production assemblies sourced from 4 Chinese seal suppliers, 9 assemblies showed outer diameter interference outside the H7/h7 design tolerance, generating a 29% first-fit leakage rate before any shaft or lip factor was considered.
Outer Diameter Fit, Housing Bore Tolerances, and Press-Fit Force Specification #
The outer diameter of an oil seal is not a passive dimension. It is the primary sealing interface on the static side — and the one specification that Chinese supplier datasheets most consistently underspecify. When a seal OD is quoted as “nominal 72mm,” that tells you almost nothing about whether the seal will retain oil in your housing.
The correct parameter to request is OD tolerance class and the associated diametral interference range at ambient temperature. For a standard rubber-OD seal pressed into a cast iron housing, the target interference is typically 0.15–0.35 mm for bore diameters in the 50–100 mm range, per ISO 6194-1. Metal-OD seals require tighter control: interference below 0.10 mm risks fretting leakage under thermal cycling, and above 0.45 mm risks bore distortion during installation.
| Seal OD Type | Recommended Interference (50–100mm bore) | Housing Material Effect | Thermal Expansion Risk |
|---|---|---|---|
| Rubber-covered OD (NBR) | 0.15–0.35 mm | Low sensitivity, rubber compensates | Moderate — NBR swell above 100°C narrows fit window |
| Rubber-covered OD (FKM) | 0.15–0.30 mm | Low sensitivity | Low — FKM swell coefficient 0.8× NBR |
| Metal OD, plain | 0.10–0.25 mm | High — aluminum housing requires lower end | High — aluminum CTE ~23 µm/m·°C vs steel ~12 |
| PTFE-encased OD | 0.20–0.40 mm | Moderate | Very low — PTFE CTE ~100 µm/m·°C, may loosen at temp |
The table above draws from DIN 3760 dimensional guidance and our housing fit database from incoming qualification audits. For aluminum housings specifically, pressing to the upper interference limit for a metal-OD seal is how bore damage occurs — a mistake we see repeatedly when Chinese seals specified for steel housings get installed in the aluminum gearbox housings that were never mentioned in the RFQ.
Where this matters most for decision-making: if your assembly drawing calls for H7 bore tolerance, verify that your Chinese supplier is holding shaft and OD tolerances that are actually compatible with H7 housing fit, not just quoting nominal OD within ±0.2 mm, which is too wide to guarantee a consistent press-fit.
Why Seals That Pass Initial Sample Approval Fail at Volume — Root Cause Analysis #
This is where Chinese oil seal sourcing gets genuinely complicated, and where most incoming inspection protocols fall short.
Compound lot substitution at the rubber compounder level. A Chinese seal manufacturer typically does not compound their own rubber. They source NBR or FKM compound from a third-party mixer, often switching based on quarterly raw material pricing. The seal manufacturer’s first article samples are produced with one compound lot. Six months into production volume, the compounder has changed the plasticizer package or carbon black loading to meet a cost target — both changes that can shift Shore A hardness by 4–6 points and compression set by 8–12 percentage points, neither of which is visible on a standard COA that only reports hardness at ambient. Per our internal QC-09 material risk procedure, any seal application in rotating equipment above 1,500 RPM gets incoming spot-testing for compression set per ASTM D395 Method B at 70°C for 22 hours, not just ambient hardness. The threshold we use for NBR lip compounds: compression set must remain below 25% under those conditions. Batches above 30% are quarantined.
The mechanism is straightforward. A lip compound with elevated compression set loses its ability to follow shaft runout dynamically. The lip stops tracking the shaft at micro-displacement frequencies generated by bearing play or belt tension. This is not catastrophic seal failure — it is slow, accumulating leakage that looks like a shaft surface problem until you pull the seal and test the compound.
Spring force drift from inconsistent coil geometry. The garter spring in a standard lip seal maintains radial lip force — typically 0.8–1.5 N per centimeter of lip circumference for general industrial applications. Chinese suppliers producing at lower price points often source springs from a separate wire-forming supplier, and spring diameter and coil pitch tolerance are almost never included in supplier QPAs. We have measured garter spring radial force using a mandrel fixture across five Chinese suppliers: two of five had mean radial force below 0.7 N/cm at nominal shaft diameter, which corresponds to inadequate lip-to-shaft follow at peripheral velocities above 6 m/s. At 1,200 RPM on a 50 mm shaft, peripheral velocity is already 3.1 m/s — at 2,500 RPM it reaches 6.5 m/s, which puts you into the zone where low spring force starts generating intermittent leakage.
The consequence is subtle. The seal passes all ambient inspection. It passes initial run-in. The leakage complaint arrives at 2,000–3,000 operating hours. By that point, the production lot is six to nine months old, return analysis is expensive, and the supplier has already changed compound lots again.
Lip geometry variation from mold wear. Mold cavities used by Chinese manufacturers producing seals at high volume are frequently shared across multiple OEM customers. When mold wear exceeds a threshold, lip angle and dust-lip geometry drift. The primary lip taper angle on a standard seal is designed to generate a specific contact pressure profile — usually targeting 0.04–0.12 MPa contact pressure at nominal shaft diameter. A worn mold can shift the lip contact band from its designed 0.3–0.5 mm width to 0.8–1.2 mm, which increases friction and heat generation without improving sealing. We catch this by measuring lip contact band width on shaft-mounted impression tests during first article qualification, using carbon black transfer on a polished reference shaft. If a supplier cannot provide this data on their qualification report, that is a flag worth following up.
Does Lip Hardness Actually Matter — or Are Buyers Over-Specifying It? #
Buyers consistently request 70 Shore A or 80 Shore A as their lip compound specification. The answer to whether that matters: it depends almost entirely on peripheral velocity and fluid viscosity, not on hardness as a standalone number.
At peripheral velocities below 4 m/s in clean oil service, Shore A variation of ±5 points has negligible effect on sealing performance or service life. The variable that drives lip wear rate at low speed is surface finish on the counterface shaft — specifically Ra 0.2–0.5 µm per ISO 6194-1, not lip hardness. Above 8 m/s peripheral velocity, however, a softer compound (70 Shore A) generates more frictional heat and reaches its glass transition temperature sooner under continuous duty, while an 80 Shore A compound in the same application may show micro-tearing if shaft surface finish is at the coarse end of specification. There is no universal answer. For our qualification matrix, we assign hardness specification based on the peripheral velocity band and fluid temperature, not on a blanket product-line default. Specifying 80 Shore A across all applications because it sounds more rigorous is a mistake — it results in premature lip cracking in cold-start conditions below -20°C.
Practical Guidance for Buyers #
When sourcing oil seals from China, start with three documents before you look at price: the dimensional tolerance table (not just nominal dimensions), the compound specification with hardness and compression set values, and the spring force range at nominal shaft diameter. The obvious starting point — material grade, NBR vs FKM — is actually less predictive of production performance than housing bore compatibility and spring force consistency.
The specific risk scenario to watch: a supplier who quotes to your nominal dimensions with no tolerance table attached is almost certainly working to their own internal tolerance, which may or may not be compatible with H7/h7 or your assembly press-fit requirement. We have seen this generate 15–30% first-fit rejection rates on production assemblies where the seal itself meets every lip and compound test — the problem is purely dimensional, and it is traceable to OD tolerance specification being absent from the original RFQ. See our related guidance on hydraulic and pneumatic seals for how this same OD tolerance issue presents differently in static versus dynamic applications.
Before volume commitment, insist on a qualification sample set of minimum 25 pieces from one production run, tested across: OD at three points per seal (minimum), radial lip force on a mandrel fixture, compression set per ASTM D395 at 70°C/22h, and a shaft impression test for contact band width. That is not an unusual request. Any qualified Chinese seal supplier with volume experience can produce this data. If they cannot, that tells you something about their process control maturity that price alone will not reveal.
For PTFE-lip variants, the tolerance picture changes — refer to our pump and valve seals category for PTFE-specific dimensional and chemical compatibility data.
Frequently Asked Questions #
What is the most important parameter to check on a Chinese oil seal COA?
Compression set at operating temperature — not ambient Shore A hardness. A compound can read 72 Shore A at 23°C and still have a compression set of 38% at 70°C after 22 hours, which is above the 25% threshold that predicts reliable lip-to-shaft contact over service life.
How do I specify OD tolerance correctly on a Chinese supplier RFQ?
State the housing bore tolerance (e.g., H7), the housing material, and the target diametral interference range explicitly — for rubber-OD seals in the 50–100 mm bore range, 0.15–0.35 mm is the standard target. Do not leave it as nominal OD with a generic ±0.2 mm — that tolerance class is too wide to guarantee a leak-free press-fit and no Chinese supplier will flag the discrepancy unless you do.
Can I use the same oil seal specification for aluminum and steel housings?
For metal-OD seals, no. Aluminum’s coefficient of thermal expansion is roughly twice that of steel (~23 µm/m·°C vs ~12 µm/m·°C), which means a seal pressed at the upper interference limit for a steel housing will generate destructive bore pressure in an aluminum housing at operating temperature. Specify to the lower end of the interference range and confirm with the supplier that their OD tolerance class is calibrated for aluminum housing applications.
Is ISO 6194-1 or DIN 3760 the right standard to reference when sourcing from China?
It depends on your end market. Chinese manufacturers are more familiar with GB/T 13871 (the domestic equivalent, administered by SAC China Standards), which has dimensional and material requirements that mostly align with DIN 3760 but allow slightly wider OD tolerances in some size ranges. If your application is governed by ISO or DIN, state that explicitly and request confirmation that the specific seal series has been dimensionally qualified to that standard, not just to GB/T.
What qualification sample size is enough before releasing a Chinese oil seal supplier for volume production?
Twenty-five pieces minimum from a single production lot, tested for OD at three axial positions, radial lip force, and compression set per ASTM D395. We also recommend requesting COAs from three consecutive production lots to verify compound consistency before any volume commitment — single-lot approval misses the compound substitution risk that typically appears between months three and eight of production volume.
Published by sinoraw.com Technical Team | Request a sourcing consultation
