Bolted joint & torque

A bolt is a spring you stretch on purpose

The point of tightening is not to make the parts touch. It is to store enough tension in the fastener that the joint never sees a load cycle. Target preload is set as a fraction of proof strength:

Fᵢ = (%) · Sₚ · Aₛ

Sixty-five to seventy-five per cent of proof is the usual band for a reusable joint. Higher is possible with tension control; higher with a torque wrench is optimism.

The torque equation, and why it is weak

T = K · Fᵢ · d

K is the nut factor: about 0.20 dry, 0.15 lightly lubricated, 0.12 waxed. It is not a material property. It bundles thread friction, under-head friction and surface condition into one number, and it is the reason torque control is worth roughly ±25 % on the preload you actually get.

Something like ninety per cent of the torque you apply is spent on friction. Ten per cent becomes tension. Change the lubricant and you have changed the joint, whether or not anyone updated the torque spec.

What the bolt sees

Fᵇ = Fᵢ + Φ·Fₑₓₜ
Fᶜₓₐₘₚ = Fᵢ − (1−Φ)·Fₑₓₜ

Φ is the share of external load that lands on the bolt rather than unloading the joint. For a stiff steel joint with a stiff fastener it is typically 0.15 to 0.30. A soft gasket pushes it up sharply, which is exactly why gasketed flanges relax.

The number to watch is the residual clamp. When it reaches zero the joint separates, and from then on the bolt takes the full external load directly — which is how a joint that was fine in static analysis fails in fatigue.