What gland depth and width should my O-ring groove be? Enter the O-ring cross-section diameter and pick how it seals. The calculator sizes the groove so the O-ring is squeezed the right amount for the job and still has room to deform: it returns the gland depth, the groove width, the resulting squeeze percent and the gland fill percent, and it keeps the fill below the point where the nearly incompressible rubber runs out of room.
Engineering estimate, not a quote. Final gland dimensions come from the seal maker's groove table for your exact O-ring size and compound, the sealed pressure and the clearance gap.
An O-ring does not seal because it fills the groove. It seals because it is squeezed: the gland is made a little shallower than the O-ring is thick, so the rubber is compressed across its cross-section and presses against both the groove bottom and the mating surface. That contact pressure, kept higher than the fluid pressure it holds back, is the seal. The whole job of a gland calculator is to choose two numbers β how deep the groove is and how wide β so the squeeze is in the right band and the rubber still has somewhere to go when it is compressed.
Step 1 β pick the squeeze. Squeeze is how much the cross-section is flattened, as a fraction of its free thickness. For a static seal, where nothing moves across the seal, you can squeeze hard; representative design ranges put static squeeze around 15 to 30 percent, often near 20 to 25 percent. More squeeze gives more contact force and seals better at low pressure, but it also stores more energy as compression set over time and makes the ring harder to install. For a dynamic reciprocating seal, where a rod or piston slides against the ring, high squeeze means high friction, heat and wear, so the band drops to roughly 8 to 16 percent. These bands are the representative values from published O-ring gland design standards, such as the common Parker and industry O-ring handbook tables; treat them as a starting point, not a law, and let the compound, pressure and temperature move the target.
Step 2 β set the gland depth from the squeeze. Depth is simply the free cross-section minus the amount you squeeze it: D = CS Γ (1 β s), where s is the squeeze fraction. Squeeze a 3.00 mm cross-section by 22 percent and the groove is 3.00 Γ 0.78 = 2.34 mm deep. The O-ring, 3.00 mm thick in free air, is pressed down into a 2.34 mm gap, so 0.66 mm of its height is taken up as squeeze. (If the ring is stretched over a rod, the stretch thins the cross-section slightly and reduces the real squeeze; account for it when the stretch is more than a couple of percent.)
Step 3 β set the groove width for the fill. Here is the part people get wrong. Rubber is nearly incompressible β its volume barely changes under pressure. When you squeeze the cross-section down, the rubber it displaces does not vanish; it bulges sideways. So the groove must be wider than the O-ring, with empty space for the squeezed cross-section to flow into. We measure that with gland fill: the O-ring's cross-section area divided by the groove's cross-section area. The O-ring area is A = Οβ4 Γ CSΒ² (its cross-section is a circle). The groove area is width Γ depth. Choosing a target fill fixes the width: W = A Γ· (F Γ D). Design practice keeps the worst-case fill below about 85 percent, which for a nominal design usually lands the width at roughly 1.3 to 1.5 times the cross-section diameter.
Step 4 β report squeeze and fill back. With the depth and width chosen, the calculator recomputes the actual squeeze and the actual gland fill from the geometry, so you can see both at once and check them against the bands. A groove can have a perfect squeeze and still be a bad groove if it is 95 percent full β on a hot day, or with a fluid the compound swells in, the ring will have nowhere to expand, the gland pressure will spike, and the ring can be forced out over the groove edge and nibbled off.
The calculator is checked against three cases computed by hand. Press Load the worked example to reproduce the first one.
Case 3 shows why a dynamic groove looks different: the squeeze is cut roughly in half to keep friction and wear down, and the fill is kept a little lower so the compound has extra room to swell when it warms up and soaks up fluid. The depth grows (less squeeze means a deeper groove for the same ring), and the groove is still about 1.3 times the cross-section wide.
Static radial (the O-ring sits in a groove on a bore or a piston and seals against a cylindrical wall) and static face (the ring sits in a groove on a flange and seals when two flat faces are bolted together) both tolerate high squeeze, so they are efficient, forgiving seals. The main difference is how the ring is trapped: a radial groove usually stretches the ring slightly onto a rod, or compresses it into a bore, while a face groove just captures it between two plates. For a face seal, make sure the groove is on the pressure side correctly and that the bolt load actually closes the faces onto the squeeze.
Dynamic reciprocating seals are a different animal. The ring is rubbing, so friction, heat and wear dominate the design. Squeeze comes down, surface finish on the moving part gets finer, and lubrication matters. Rotary dynamic seals (a turning shaft) are harder still and often use a different seal type entirely, because the Gow-Joule effect can make a squeezed O-ring run hotter and tighter as it heats β a reason to be conservative with squeeze on anything that moves.
Representative finish targets: static sealing faces and groove walls about 0.8β1.6 Β΅m Ra (32β63 Β΅in); a dynamic moving sealing surface finer, about 0.2β0.4 Β΅m Ra (8β16 Β΅in). Do not go mirror-smooth on a dynamic surface β a little roughness holds a lubricant film. Keep machining lay from crossing the seal line, and break the sharp groove edges with a small radius so the ring is not shaved on assembly.
Stretch and swell both eat into your margins. Installing an O-ring over a rod stretches its inner diameter; keep that under about 5 percent, because stretch thins the cross-section and quietly reduces the squeeze you designed for. Chemical and thermal swell runs the other way β the compound absorbs fluid or heats up and grows β which is exactly why the gland fill must have headroom. If you know the service fluid swells your compound by several percent, design the nominal fill lower so the worst case still clears 85 percent.
At higher pressures the O-ring is pushed hard against the low-pressure side of the groove and tries to extrude into the tiny clearance gap between the two sealed parts. Past a representative threshold near 10 MPa (about 1500 psi) β lower with a soft compound or a large gap β a backup ring is added on the low-pressure side to close that gap, like a washer that the O-ring cannot squeeze past. A backup ring takes groove width, so the gland must be made wider to fit it; this calculator sizes the O-ring portion of the gland, not the backup, so add the backup ring's width to the groove when you need one. Harder compounds, smaller diametral clearances and backups all raise the pressure a seal survives before it extrudes.
| Method | Use it when | Limit |
|---|---|---|
| This calculator | First-pass gland sizing from the O-ring cross-section; sanity-checking a groove on a drawing; explaining squeeze and fill | Rectangular groove, round cross-section; does not size backup rings, clearance gaps or stretch correction |
| Seal maker's groove table | Any final design with a named O-ring size and compound | Tied to that catalogue's sizes; still your job to pick static vs dynamic and check swell |
| Published gland design standard | Setting the squeeze, fill, gap and finish bands you design to | General ranges; the compound and service move them |
| FEA / seal simulation | Critical, high-pressure, high-cycle or extreme-temperature seals | Needs good compound data and takes real effort |
Sources for the method. The squeeze, gland-fill, surface-finish, stretch and backup-ring ranges used here are the representative values published in O-ring gland design standards, for example the common Parker and industry O-ring handbook tables. They are hedged on purpose: the right numbers for a seal depend on the compound, the hardness, the sealed pressure, the temperature and the fluid, and the seal maker's groove table for your exact ring governs. The geometry β depth from squeeze, area from the circular cross-section, width from the fill β is exact; the target bands are guidance.
Reviewed by the JCH tooling team Β· updated November 2026
Set the depth so the O-ring is squeezed the right amount: depth = cross-section Γ (1 β squeeze fraction). For a static seal a common target is 15β30% squeeze, so a 3.00 mm cross-section groove is about 2.1β2.55 mm deep; at 22% it is 3.00 Γ 0.78 = 2.34 mm. Dynamic reciprocating seals use 8β16% squeeze, so a 5.33 mm cross-section runs about 4.5β4.9 mm deep. These follow published O-ring gland design standards such as the common Parker and industry handbook ranges; the exact value depends on the compound, the pressure and the finish.
Wider than the ring, because squeezing the cross-section pushes rubber sideways and the groove must never be completely full. A common rule is a width of about 1.3β1.5 Γ the cross-section, chosen so the gland fill (ring area Γ· groove area) stays near 70β85%. For a 3.00 mm cross-section at 22% squeeze and 75% fill the groove is about 4.0 mm wide. Add width, not fill, to allow for swell.
Gland fill is the O-ring cross-section area divided by the groove area (width Γ depth). Rubber is nearly incompressible, so when squeezed it flows sideways and must have empty space to flow into. If the groove is too full, the ring has nowhere to go when it is compressed or when it swells from heat or fluid; the gland pressure spikes, the ring can be forced out of the groove, and friction and extrusion rise. Designers keep the worst case below about 85%, which usually means a nominal fill near 70β80%.
Squeeze = (cross-section β gland depth) Γ· cross-section. Static seals are typically squeezed 15β30%; more squeeze seals better at low pressure but raises compression set and installation force. Dynamic reciprocating seals use 8β16% so friction and wear stay manageable. Too little leaks, too much takes a set and wears. It is a band, not one number, and the compound and pressure move it.
Representative values: static grooves and faces about 0.8β1.6 Β΅m Ra (32β63 Β΅in); a dynamic moving sealing surface finer, about 0.2β0.4 Β΅m Ra (8β16 Β΅in) β smooth enough to limit wear but not so smooth it starves the contact of lubricant. Keep machining lay from crossing the seal line and break the groove edges.
When pressure is high enough to extrude the ring into the clearance gap. As a representative threshold, backups are common above roughly 10 MPa (about 1500 psi), with the exact limit set by the gap, the compound hardness and the temperature. The backup goes on the low-pressure side and closes the gap, and the groove must be widened to fit it.
For a bore or rod seal the inner diameter is usually stretched a little so the ring stays seated, but keep it under about 5% (often 2β5%). Stretch thins the cross-section and reduces the actual squeeze, so large stretch must be accounted for; over-stretching also ages the ring faster and pulls the cross-section out of round.
Not sure whether your groove is too full, or how much squeeze a moving seal should run? Ask our AI assistant. It answers from this website, and our tooling team follows up if you want.
AI assistant. Answers can be wrong; nothing it says is a quote or a promise.
This calculator gives an engineering estimate. Our tooling team checks the groove depth, width, fill, finish and clearance against your O-ring size, compound and pressure, and proposes backup rings or a different seal where it matters β while every change is still a model edit.
Book a 30-min call Send a drawing or sample