Three dimensions describe every O-ring: inner diameter (ID, often written d1), cross-section diameter (CS, often written d2 or W), and outer diameter (OD). Only two of them are independent — the third always follows from:
OD = ID + 2 × CS
Measure two, calculate the third, and use the calculated value as a cross-check.
Tools that work
A caliper is useful for preliminary identification, but display resolution is not measurement accuracy. Rubber deforms under contact force. Choose a measuring method suitable for the specified tolerance, check the instrument, and keep the ring relaxed. A manufacturer-approved gauge or non-contact method may be needed for an acceptance measurement.
Two habits improve every measurement:
- Take at least three readings around the ring and average them — mould parting lines and slight ovality are normal.
- Use the lightest possible contact force. If the caliper visibly dents the rubber, the reading is low.
Measuring inner diameter
Lay the ring flat, without stretching it, and measure across the inside of the ring at its widest point. Do not hang the ring on the caliper jaws — its own weight stretches it. For small rings, slide the ring onto a cone gauge until it stops and read the scale at the contact line.
If the ring is used and has taken a compression set, the ID may have grown slightly; treat used-ring measurements as approximate and prefer identifying the original groove dimensions.
Measuring cross-section
Close the jaws gently on the cord at four points around the ring and average. On a used ring, measure away from the flattened contact faces if possible. A cross-section that measures noticeably oval usually indicates compression set rather than a non-standard size.
Measuring outer diameter
OD is the least reliable direct measurement, because the ring flexes away from the jaws. Prefer calculating OD = ID + 2 × CS. When only the OD is measurable (for example, the ring sits in a blind groove), measure OD and CS, then calculate ID = OD − 2 × CS.
Matching to a standard size
With ID and CS in hand, compare them against the common standard families:
- AS568 (inch-based, used worldwide): cross-sections of 1.78 / 2.62 / 3.53 / 5.33 / 6.99 mm.
- ISO 3601 and GB/T 3452.1: consult the applicable size table and edition; do not assume every series or rounded metric dimension is equivalent to AS568.
Nominal cross-sections of 3.53 mm and 3.55 mm differ by only 0.02 mm. A worn or compressed sample may not distinguish them reliably. Compare both ID and CS, the permitted tolerances and the exact designation in the AS568 and ISO 3601 references. A nearest-size result is a candidate, not a replacement approval.
Worked example: turn measurements into a design check
Suppose an undamaged, relaxed ring measures approximately ID = 20.00 mm and CS = 2.50 mm. Its calculated OD is 20.00 + 2 × 2.50 = 25.00 mm. Enter both readings into the O-ring size calculator and compare the candidate ID and CS separately. Do not round the measurements to the first catalogue entry before comparing them.
If an independent OD reading is 25.4 mm, the 0.4 mm disagreement is a reason to repeat the measurements and check ovality or deformation; it does not by itself identify another standard size. Retain the spread of readings rather than hiding it in an average.
After identifying a candidate, use the radial static calculator for a piston or rod gland, or the face-seal calculator for an axial groove. Use supplier-confirmed nominal dimensions and tolerances for release. Measurement establishes a likely size; groove geometry and service conditions establish suitability.
Measurement reference: Apple Rubber’s guide to measuring ID and cross-section. The numerical example above is an independent illustration, not a manufacturer’s specification.
Common mistakes
- Stretching the ring while measuring the ID (reads high).
- Squeezing the cord with the caliper (reads low).
- Measuring a swollen ring after fluid exposure and treating the result as the original size.
- Assuming OD measurements are authoritative — derive OD instead.
Standard size data is for preliminary engineering selection only. Final purchasing and design should be verified against the applicable standard edition, supplier data sheets and actual operating conditions.