The Calculation, Explained
Torque is a means; preload is the end
A tightened bolt is a spring stretched by its preload F. The joint holds because the clamped faces are squeezed by F — the torque wrench is merely the most convenient way of stretching the spring. The short-form relationship used everywhere in industry is:
K, the nut factor, hides everything messy. Of the torque you apply, roughly 50% is consumed by friction under the nut face, 40% by thread friction, and only about 10% actually stretches the bolt. Change the friction — oil the threads, add anti-seize — and the same wrench setting produces a very different preload. That is why the K factor, not the wrench, is the most important entry on this page: dry steel ≈ 0.20, lightly oiled ≈ 0.15, MoS2/anti-seize ≈ 0.12. Applying a dry-table torque to lubricated threads over-tightens the bolt by a third and can yield or snap it during assembly.
Assumptions and applicability
- Scatter is inherent. Torque control typically delivers preload within ±25–35% (VDI 2230 tightening factor for wrench tightening). The preload computed here is the mean; design the joint so the minimum still clamps and the maximum still stays elastic, or move to torque-angle or hydraulic tensioning for critical joints.
- 70% of yield is the standard target for reusable, torque-controlled joints. Yield-controlled and single-use applications go higher under procedures this calculator does not replace.
- Stress area As values are for standard metric coarse threads (ISO 898-1 basis). Yield strengths are the nominal Rp0.2 of the property class.
- Torque alone does not sequence a flange: use a star/cross pattern in at least three passes (30/70/100%) so the gasket compresses evenly — the calculator gives the final value, not the procedure.
Worked example (the defaults above)
M20, class 10.9, dry (K = 0.20), target 70% of yield: F = 0.70 × 900 MPa × 245 mm² = 154 kN of clamp load per bolt (bolt stress 630 MPa, safely elastic). Torque = 0.20 × 20 mm × 154,350 N = 617 N·m (≈ 455 ft·lbf). Oil those same threads (K = 0.15) and the correct torque falls to about 463 N·m — leave the wrench at 617 and you have just specified 93% of yield by accident.
References
VDI 2230 — Systematic calculation of highly stressed bolted joints; ISO 898-1 — mechanical properties of fasteners; standard machine-design references (Shigley) for nut-factor values.
Frequently Asked Questions
What is the K factor in bolt tightening?
The K factor (nut factor) bundles all the friction in the tightening process into the short-form torque equation T = K x d x F. Only about 10% of applied torque stretches the bolt; roughly 50% is lost under the nut face and 40% in the threads. Typical values: 0.20 dry steel, 0.15 lightly oiled, 0.12 with anti-seize.
Why does lubricating a bolt change the required torque?
Because most of the torque fights friction rather than stretching the bolt. Lubricant cuts friction, so the same wrench setting produces much more preload: a dry-table torque applied to oiled threads can take a bolt from the intended 70% of yield to over 90% and yield or snap it during assembly.
How accurate is tightening by torque wrench?
Torque control typically scatters preload by plus or minus 25 to 35% (VDI 2230), because friction varies joint to joint. It is entirely adequate for most plant joints when the design allows for the scatter; critical pressure or fatigue joints use torque-angle, yield control or hydraulic tensioning instead.
Related calculators: Alignment Shim Calculator · Thermal Expansion