O-Ring Compression and Stretch Calculator

Four dimensions in, three percentages out. The page does the arithmetic that gets done on the back of a drawing and gets done wrong when the section and the gland come from different sources.

Use any consistent length unit — millimetres, inches, anything. Every output is either a percentage or a length in the unit you typed.

The starting values are placeholders chosen to show the shape of the output. They are not a recommended gland, a standard size or a typical case. Replace all four.

What each input means here

Cross-section diameter is the cord diameter of the ring in its free, unstretched, uninstalled state.

Ring inner diameter is likewise the free inner diameter, before it is put anywhere.

Groove depth is the radial distance from the groove floor to the mating surface that the ring is compressed against — the gap the section has to live in, not the depth of the cut in isolation. If your drawing gives the gland as a diameter over the groove, subtract and halve before entering it here.

Groove diameter is the diameter of the cylindrical surface the ring’s inner diameter seats on: the groove root diameter of an external, male gland. Entering a value below the ring’s free inner diameter means the ring is not being stretched onto anything, and the page will say so rather than reporting a negative stretch as though it were a design.

What it computes

Nominal squeeze is (cross-section − groove depth) ÷ cross-section, expressed as a percentage. It is the fraction of the free section that has been taken out of it by the gland. This is the figure most drawings and most conversations mean by “squeeze”.

Inner-diameter stretch is (groove diameter − free inner diameter) ÷ free inner diameter. It is the strain the ring’s inner circumference is held at once it is on the seat.

Squeeze on the stretched cord is the correction most quick calculations skip. Rubber is very nearly incompressible, so a ring that is stretched around a larger diameter has to get thinner somewhere, and it gets thinner in the section. Treating the ring as a torus of circular section and holding its volume constant gives a relationship between the cord diameter and the mean diameter, which the page solves numerically. The result is the squeeze the gland actually achieves, which is always lower than the nominal figure whenever there is any stretch at all.

That correction assumes a perfectly circular section that stays circular, a uniform stretch and no volume change. It is an idealisation, and its purpose is to show the direction and rough magnitude of an effect that is easy to forget — not to replace a measurement.

Diametral squeeze is the same compression expressed as a length rather than a percentage, computed on the stretched cord, so you can compare it directly to a tolerance.

What it deliberately does not do

It contains no material data of any kind. There is no hardness table, no temperature rating, no compatibility chart and no compression-set allowance, because those figures vary by compound, by supplier and by standard, and a number typed into a web page would be invented.

It does not compute gland fill — the proportion of the groove cross-section that the ring occupies. Fill is arguably the single most consequential gland figure after squeeze, because a groove with too little free volume cannot accommodate thermal expansion or fluid swell and will drive the ring into the clearance gap. It needs the groove width, which this page does not ask for. Work it out separately; do not assume that a gland with acceptable squeeze has acceptable fill.

It does not know your gland type. The geometry above describes a ring stretched onto a male seat. A female gland, where the ring is compressed inward into a bore groove, is a different arrangement with different signs, and the stretch output will not describe it.

It says nothing about extrusion, back-up rings, clearance gaps, surface finish, installation damage or lead-in chamfers, all of which end more seals than an out-of-band squeeze figure.

About the flags

The panel flags a squeeze or a stretch that falls outside a broad band. Those bands are wide on purpose. They are there to catch a transposed digit, a cross-section taken from one drawing and a gland from another, or an inch dimension entered next to a metric one — the errors that produce an obviously implausible result.

They are not a specification and they are not an approval. What is acceptable depends on whether the seal is static or dynamic, on the pressure, on the compound, on the temperature range, on the gland type and on whichever standard or supplier design guide the part is drawn to. A figure inside the band can be wrong for your application and a figure outside it can be entirely correct. Where a real limit exists, it belongs to that document, not to this one.

Where the numbers become approximate

Everything on this page is exact geometry applied to nominal dimensions. Real parts arrive with tolerances, and squeeze is a difference of two dimensions, so its tolerance is proportionally much wider than the tolerance on either input. A gland and a ring that both sit comfortably within tolerance can combine to give a squeeze well outside the value the drawing implies.

The useful exercise is therefore to run the calculation three times: once at nominal, once with the section at its minimum and the gland at its deepest, and once at the opposite extreme. If the squeeze stays sensible across all three, the design has margin. If it does not, the fix is a tolerance, not a compound.