Rule of thumb · MechanicalNº 15 / 167

Torque mostly fights friction

Roughly 90% of tightening torque is lost to friction; only ~10% actually stretches the bolt.

Why it works

About half goes to under-head friction, ~40% to thread friction. That is why lubrication changes preload so dramatically at the same torque.

When it fails

The 90/10 split is a dry-steel figure, not a constant. On a plated and lubricated joint less torque goes to friction and more becomes tension, which is precisely why a wrench setting taken from a dry-thread table over-tightens a greased bolt rather than under-tightening it.

How wrong is it?

At 0.14 μ the rule says 0.1 and the exact answer is 0.1225 — 18% low. At the point most people use it the shortcut is already outside its own 10% tolerance. It only holds between 0.16 μ and 0.2 μ, which is not where it gets used.

+10%0−10%0.080.1650.25Thread and head friction (μ)

The 90/10 split is a dry-steel figure quietly presented as a law. Oil the thread and the useful share nearly doubles to 16%; let it corrode and it drops under 8%. The point of the rule survives — almost all of your effort is heating the threads — but the number moves by a factor of two with a drop of oil.

The rule against the exact answer, computed across the range. Inside the shaded band the shortcut is close enough to use; outside it, reach for the calculator.

Do it exactly

Estimate with the rule, then check it against the calculator that models it properly.

Open Bolt Torque & Preload Calculator →

More in Mechanical

← PreviousThe bolt-torque shortcut Next →Deflection limits: span/250

Where does bolt tightening torque actually go?

Roughly 90% of tightening torque is lost to friction; only ~10% actually stretches the bolt. About half goes to under-head friction, ~40% to thread friction.

Browse all 167 rules of thumb →