O-Ring Groove Design: Dimensions, Tolerances and Surface Finish

Updated October 3, 2026

For machined parts, start by identifying the gland arrangement and the exact seal compound. Select a suitable manufacturer reference, calculate squeeze and available gland space across tolerances, then specify the surfaces and inspection needed to preserve those results. The example below shows the arithmetic without presenting a universal groove size.

Choose the gland arrangement before choosing dimensions

O-ring groove design depends first on whether the seal is static or moving and whether compression is axial or radial. A static face seal between two flanges needs a different housing analysis from a reciprocating rod seal. Rotary service introduces additional friction and lubrication concerns; a gland intended for slow reciprocation should not be assumed suitable for continuous rotation.

Record the fluid or gas, pressure direction, pressure cycling, minimum and maximum temperature, motion, assembly method and expected service life. For face seals, pressure direction affects which groove wall locates the ring. For radial seals, the mating bore or shaft is part of the compression calculation. Also identify any pressure reversal or vacuum requirement.

The Parker O-Ring Handbook provides application-specific design guidance. Use the relevant gland section rather than transferring a static seal recommendation to a dynamic application. A standard-size ring makes sourcing easier, but the seal supplier must still confirm that its compound and housing guidance fit the actual duty.

Define the seal size, compound and installed condition

Specify an O-ring by size system, inside diameter, cross-section, tolerance, compound and hardness designation. A color or a generic polymer family is not enough to identify a replacement seal. Two nominally similar compounds can differ in compatibility, low-temperature behavior, compression set and swelling. Confirm these properties against the medium and exposure duration.

The scope of SAE AS568F covers O-ring inside diameters, cross-sections, tolerances and dash numbers. It is a size standard, not a complete answer for groove geometry, chemical compatibility or pressure capacity. Distinguish the ring designation from the reference used to design its housing.

Installed stretch changes the ring cross-section and volume distribution. Oversized or insufficiently retained rings can also complicate assembly. Check the installed diameter in the chosen gland and use the seal manufacturer’s calculation approach when stretch is significant. Record whether the design uses a standard purchased ring or a custom seal, because the available tolerance and replacement options may change.

Calculate squeeze from the assembled gland height

For a simplified axial gland, squeeze percentage is 100 × (d − h) / d, where d is the free O-ring cross-section and h is the closed gland height. This equation describes geometry; it does not establish an acceptable squeeze range. Obtain the range for the chosen gland and compound from the seal manufacturer.

Consider a hypothetical static face gland with d = 3.50 ± 0.08 mm and h = 2.75 ± 0.05 mm. Nominal squeeze is 21.43%. Minimum squeeze combines the smallest cross-section with the largest gland height: 100 × (3.42 − 2.80) / 3.42 = 18.13%. Maximum squeeze combines 3.58 mm with 2.70 mm, giving 24.58%. These dimensions illustrate a tolerance check, not a recommended gland.

If the flanges do not close against rigid stops, h is not simply the machined groove depth. Joint clearance, gasket spacing, flange deflection and final treatment can change it. For a radial gland, derive the compression space from the actual groove and mating diameters. Name the dimensions in the stack so inspection of one part cannot be mistaken for verification of the assembled seal.

Check gland fill separately from squeeze

Squeeze and gland fill answer different questions. Squeeze concerns compression across the seal section; fill concerns how much gland space the elastomer occupies. A design can have reasonable initial squeeze yet too little remaining volume for swelling or thermal expansion. Groove width matters even when it does not directly set axial compression.

For the same hypothetical example, let groove width w = 4.50 ± 0.05 mm. A rectangular cross-section estimate is fill = 100 × (πd²/4) / (w × h). It ignores groove corner radii, installed stretch, circumference differences, swelling and thermal expansion. It is a preliminary calculation, not a substitute for installed seal and gland volume analysis.

Compare the real installed-volume result with the applicable supplier limits. Trelleborg’s O-ring catalog distinguishes application conditions and housing guidance. Do not approve a groove merely because its room-temperature nominal calculation passes; examine the temperature and fluid exposure that consume the available margin.

Separate extrusion clearance from groove width

Extrusion clearance is the opening into which pressure can force the elastomer. It is often the gap between mating components, not the width of the groove itself. Increasing groove width does not automatically solve extrusion, and reducing it can remove needed volume. Evaluate the clearance at maximum pressure, including component deflection and manufacturing tolerances.

Seal hardness, pressure, temperature and gap size interact. A harder compound or backup ring may be appropriate, but either can introduce assembly, friction or geometric constraints. Obtain an application-specific recommendation rather than selecting a pressure limit from a ring size alone. Pressure cycling and reversal may also require support on more than one side.

On a cylinder assembly, check the largest bore and smallest shaft or piston dimensions that open the relevant gap. On a face gland, check flange separation under load. If the seal can move toward an unsupported opening, include that condition in the analysis. The RFQ should identify the mating part or provide its controlling dimensions so the machinist can understand why a local tolerance matters.

Specify surface texture on each functional surface

A single roughness note applied to the whole part rarely describes the gland adequately. Identify the groove bottom, groove walls and mating sealing surface. Dynamic contact surfaces generally need a more detailed texture specification than noncontact exterior faces. Use the seal supplier’s requirements for the actual motion and material; there is no universal Ra value that proves an O-ring will seal.

Texture parameters are not interchangeable. Ra and Rz describe different characteristics; legacy RMS values should not be relabeled as Ra. Define the units, parameter, applicable standard and measurement conditions. ISO 21920-1 addresses the indication of profile surface texture. Scratches, burrs and directional tool marks that create a leakage path still need separate attention.

Avoid a blanket instruction to align machining lay with seal motion. The relevant lay and measurement direction depend on the gland and texture specification. Show the functional area and required measurement direction where necessary, then confirm stylus access in a narrow groove. For drawing examples and finishing terminology, review the CNC surface finish guide.

Design entry edges and corner radii for assembly

The edge that an O-ring crosses during installation is different from the radius at the bottom of its groove. An assembly lead-in guides the seal over a shaft or into a bore; a groove root radius affects usable gland space and tool geometry. Specify both deliberately. An uncontrolled deburring note does not define either feature.

Thread crests, ports and intersecting holes can cut a ring during assembly. Where the design cannot remove the hazard, establish a protective sleeve or assembly sequence. Apple Rubber’s machining guidance highlights the importance of seal-friendly edges and machined surfaces. Inspection should include the actual entry route, not only the groove after the seal is installed.

For CNC-machined parts, discuss cutter access, internal groove reach and achievable corner geometry during DFM review. If a bottom radius changes, recalculate gland volume instead of accepting it as a cosmetic deviation. Also specify cleanliness: a correct groove containing a sharp chip or abrasive residue can damage a seal immediately.

Put the gland requirements and final condition on the drawing

A useful drawing identifies the mating interface, groove dimensions, locating datums, edge preparation, texture requirements and seal reference. State whether dimensions apply after anodizing, plating or another treatment. Masking a groove may be feasible, but the masking boundary and any exposed substrate require review alongside the corrosion requirement.

Do not leave the gland depth referenced to a surface that will later be removed or coated without explaining the final relationship. If the mating face controls closure, it should be included in the inspection plan. The surface finishing route is part of the dimensional stack whenever it changes the sealing geometry.

Mark the characteristics that require reported results. Specify how a narrow groove width or depth will be measured, and agree on a suitable method before production. A visual inspection can detect a burr but cannot establish squeeze. Likewise, a dimensional report cannot prove compound compatibility. Keep these acceptance questions separate so each receives the appropriate evidence.

Validate the assembly and control repeat production

First-article inspection confirms the manufactured geometry against the drawing; a leak test evaluates the assembled seal under defined conditions. Use both where required. State the test medium, pressure, temperature, stabilization time, duration and permitted leakage. Include motion or pressure cycling when these determine service performance.

Inspect the removed seal for cuts, twisting, extrusion or abnormal wear after the relevant test sequence. One successful room-temperature test does not qualify all temperatures, fluids or future production lots. Changes to compound, seal source, coating, groove tooling or mating geometry should trigger a review of which evidence remains valid.

For repeat orders, retain the approved drawing revision, material and finish records, critical measurements and seal identification. Agree on inspection frequency based on risk and observed process behavior. A useful release package connects the measured groove, the purchased O-ring and the tested assembly; it does not treat any one of those records as a substitute for the others.