Here is the uncomfortable truth about a torque spec: you are not trying to reach a torque, you are trying to reach a clamp load. Torque is just the cheapest proxy we have for it — and a leaky one. Understanding the link between the two is the difference between a joint that stays tight and one that loosens, fatigues, or crushes its gasket.
What actually holds a joint together
A bolted joint works because the fastener is stretched into tension — the preload — which clamps the members together. That clamp force is what resists separation, fights vibration loosening, and carries fatigue. The bolt is a very stiff spring; torque is simply how we wind it up.
The standard rule of thumb is to preload a reusable joint to 75% of the bolt’s proof load. That is high enough to keep the joint tight under working loads, low enough to stay elastic and allow re-use.
Target preload F = 0.75 × Proof load
At is the tensile stress area (not the nominal area — threads reduce it), and Sp is the proof strength of the grade.
The torque–tension relation
Torque and preload are tied by the short-form nut-factor equation:
- T = applied tightening torque
- K = nut factor (dimensionless) — friction lumped into one number
- F = target preload
- d = nominal bolt diameter
K is where the honesty lives. A clean, as-received steel fastener runs about K = 0.20; a lubricated one drops to roughly 0.15; a waxed or moly-coated one lower still. Most of your applied torque never becomes preload at all — it is spent grinding through friction under the head and in the threads.
Worked example — 1/2-13 Grade 5 (inch)
| Quantity | Value |
|---|---|
| Tensile stress area, At | 0.1419 in² |
| Proof strength (Grade 5), Sp | 85,000 psi |
| Proof load | 12,062 lbf |
| Target preload (75%) | 9,046 lbf |
| Nominal diameter, d | 0.5 in |
With a clean dry fastener (K = 0.20):
Now lubricate the same bolt (K = 0.15) and hold the same torque:
F = T / (K · d) = 904.6 / (0.15 × 0.5) = 12,061 lbf → 100% of proof load!
Read that again. The same 75 ft·lbf that gives a healthy 75% preload on a dry bolt drives a lubricated one to its proof load — right at the edge of yield. This is exactly why you never apply a dry torque spec to a lubed fastener. The correct lubed torque here is lower:
Worked example — M12 property class 8.8 (metric)
| Quantity | Value |
|---|---|
| Tensile stress area, At | 84.3 mm² |
| Proof strength (8.8), Sp | 580 MPa |
| Proof load | 48,894 N |
| Target preload (75%) | 36,670 N |
| Nominal diameter, d | 12 mm (0.012 m) |
That lands squarely in the published range for a dry M12 8.8 fastener — a good confidence check that the method, not just the memorized table, gives the right answer.
Why the scatter matters
Even with a calibrated wrench, torque control delivers preload with roughly ±25–30% scatter, almost all of it from K. That is why critical joints move to angle control, bolt-stretch measurement, or load-indicating washers. For everyday joints, torque is fine — as long as you match K to the real lubrication condition and use a wrench you trust. A calibrated click-type torque wrench (Tekton, CDI, or Proto in the mid range) is the single best money you can spend on joint reliability.
Special case: gasketed and flanged joints
Flanged, gasketed connections add another constraint — the gasket needs a minimum seating stress but will crush if over-clamped, and the bolt pattern has to load it evenly. That is a preload problem with a window on both ends. The Gasket Bolt Load calculator sizes the per-bolt load to seat the gasket without exceeding its crush limit, and the Friction Coefficient lookup gives defensible K values when you cannot test.
Joint stiffness — where the load actually goes
A preloaded joint has a hidden benefit that torque tables never show: as long as the clamp stays engaged, most of any external working load is carried by relieving the clamp, not by adding tension to the bolt. The split is set by the joint stiffness ratio — the bolt stiffness versus the clamped-member stiffness. A stiff joint with a relatively springy bolt sees only a small fraction of the external load reach the fastener, which is exactly why high preload is so good for fatigue life: the bolt barely feels the cyclic load until the joint starts to separate.
Lose preload and that protection vanishes — the full external load now cycles the bolt directly, and fatigue failure follows. Loosening is not just an inconvenience; it changes the load path.
Why joints loosen
- Embedment — surface asperities flatten under load in the first hours, shedding a few percent of preload. Re-torquing after break-in recovers it.
- Gasket creep / relaxation — soft members keep giving over time, which is why gasketed joints often need re-torque.
- Transverse vibration — side-to-side motion, not axial, is what actually spins nuts loose (the Junker mechanism). High preload and thread-locking counter it.
Skip the hand arithmetic
The Bolt Torque calculator computes proof load, target preload, and the correct torque for any grade, size, and lubrication condition — in both inch and metric — so your spec matches the joint you actually have.
Stop looking it up twice — run the numbers and keep the record.
This reference is free forever. When you need to actually calculate, document, and defend the result, do it in the app — then export a clean report you can hand to a reviewer.
- ✓46 engineering calculators, unlimited
- ✓Save & reload every calculation
- ✓Organize work into project folders
- ✓Branded, PE-stamped PDF reports
Free plan, no credit card. Upgrade only when it's paying for itself.