The shaft is a little larger than the bore, and the stretched plastic grips it. Size the maximum interference from the hub's allowable hoop strain, take a safety factor of 1.5 to 2, then check the minimum interference after creep and at the hottest service temperature, because plastic relaxes and expands far more than metal.
A press fit holds because the shaft is slightly larger than the hole. Pushing them together stretches the hub outward, and the pressure between the two surfaces turns into friction against sliding and turning. With two metal parts that pressure stays put. With plastic it doesn't: plastic stretches far more for a given force, expands far more with heat, and slowly relaxes under any load it carries. Designing a plastic press fit means planning for all three.
A 4.0 mm hardened steel pin is pressed into a molded boss in 33% glass-reinforced nylon with a 10 mm outside diameter. How much interference can the boss take?
BASF gives the radial deformation for a rigid metal shaft in a plastic hub as δ = b × ε × [(c² − b²) ÷ (c² + b²)] × [(c² + b²) ÷ (c² − b²) + ν], where b is the pin radius, c the hub outside radius, ε the allowable strain and ν the hub's Poisson's ratio. Its own example uses 2% allowable strain and ν = 0.35 for a 33% glass-reinforced nylon grade, so use the same values.
If the pin were the same glass-filled nylon instead of steel, both parts would give, and BASF's formula for identical materials, 2 × ε × b × c² ÷ (c² + b²) for a solid shaft, allows 0.069 mm radial: more interference for the same strain. The press-fit interference calculator works the contact pressure, hub stress and holding torque from these sizes.
Engineering ToolBox lists steel at about 10.8–12.5 µm/m·°C, glass-fiber nylon at about 23 and acetal at 85–110. When the hub expands more than the pin, the interference shrinks as the part warms and grows as it cools, by (αhub − αpin) × d × ΔT.
DuPont also notes the reverse use of expansion: heating the hub or chilling the pin just before assembly reduces the press force. The thermal expansion calculator gives the size change for any pair of materials.

DuPont tested steel shafts pressed into acetal sleeves at 2% to 5% interference and hub-to-shaft diameter ratios from 1.5 to 4, plotting pull-out force against time out to 100,000 hours. The grip falls over that span. Those tests ran at room temperature; DuPont adds that higher temperature speeds up the relaxation, and that pull-out force varies with the shaft's surface finish. Its fix is mechanical: knurl or groove the metal part so the plastic flows into the grooves and keeps hold after the pressure has relaxed. BASF puts it more bluntly: allow for a large reduction of the initial clamp force.
| Need | Better choice |
|---|---|
| Locate a pin, light retention, room temperature | Plain press fit, sized as above |
| Hold torque or axial load for years, or through heat cycles | Knurled or grooved metal part, or a heat-set or ultrasonic insert |
| Metal part that must never move | Insert molding, with the plastic locked around the metal |
| Two plastic parts, permanent and sealed | A weld; see ways to join molded parts |
| Frequent removal | A screw into a boss, or a snap fit |
Calculate the maximum from the hub's allowable strain and geometry, then divide by a safety factor of 1.5 to 2, as DuPont suggests. In BASF's method, a steel pin in a 33 percent glass nylon hub with a 2.5 to 1 diameter ratio tops out near 2.5 percent of the pin diameter before the safety factor.
Yes, to some extent. Plastic under steady strain creeps and its stress relaxes, so the pressure on the shaft falls. DuPont's tests on steel shafts in acetal sleeves show pull-out force dropping over long periods, and heat speeds the loss. Knurling or grooving the metal part helps keep the grip.
The thermal expansion coefficient of unreinforced plastics may be six to eight times higher than that of most metals. A plastic hub on a steel pin loses interference when hot and gains it when cold. Check the minimum fit at the hottest temperature for holding, and the maximum fit at the coldest temperature for cracking.
Yes. When both parts are the same plastic, both deform, so the joint can take more interference for the same strain than a metal pin in a plastic hub. BASF gives a separate formula for this case. Creep and relaxation still apply, now to both parts.
Common causes are maximum interference above what the grade can take, a hub wall that is too thin, a weld line along the bore, and added strain when the part gets cold. Recheck the largest pin against the smallest bore, move the gate away from the bore, or thicken the hub.
Send the part file and the quantity you expect to run, and we'll review it before you commit to tooling. NDA first if you want one.
Send a part for reviewBook a 30-minute call