Cord diameter falls as 1/√(1+stretch). Fit a ring with 5% stretch and it arrives 2.4% thinner than the groove was cut for — so the squeeze you get is always under the squeeze you asked for.
Rubber is very nearly incompressible, so the cord’s volume is fixed. Stretching the ring round a larger circle makes it longer, and the only place the volume can come from is the cross-section: area goes as 1/(1+s) and diameter as 1/√(1+s). The groove, meanwhile, was machined for the nominal cord and does not move. On a 3.53 mm cord cut for 20% squeeze, a 3% stretch delivers 18.8% and a 10% stretch delivers 16.1% — and small rings need proportionally more stretch for the same step in diameter, so the tightest seals lose the most. It is also why installed stretch is held to about 5%: past that you are trading away cord and living on rubber under permanent tension.
It is the LEAST the cord thins, not the most — a real ring does not stretch uniformly and Poisson’s ratio is a shade under a half, both of which take more off rather than less, so treat the number as an upper bound. And it runs the other way in a bore groove, where the ring is squeezed into a smaller circle and the cord thickens: there the limit is buckling, around 3%, not lost squeeze.
Estimate with the rule, then check it against the calculator that models it properly.
Open O-Ring Calculator →Cord diameter falls as 1/√(1+stretch). Fit a ring with 5% stretch and it arrives 2.4% thinner than the groove was cut for — so the squeeze you get is always under the squeeze you asked for. Rubber is very nearly incompressible, so the cord’s volume is fixed.