Bolt Torque Calculator
Torque specs for SAE and metric bolts from proof load — and what lubrication does to the answer, which is more than the grade does.
What's on the threads changes the answer more than the spec does
The same torque, on the same bolt, at every thread condition. The right-hand column is the one that breaks fasteners: it is the clamp force you would actually get if you applied the dry table figure to threads in that state.
Torque barely measures the thing you care about
You are trying to stretch a bolt by a known amount. What you are actually doing is overcoming friction and hoping. Only about 10 to 15% of the torque you apply becomes bolt tension — roughly half is consumed under the head or nut bearing face and most of the rest inside the threads. A torque wrench is a friction meter that infers tension, and even used carefully it scatters the resulting preload by something like a quarter to a third.
That is why lubrication changes the answer more than the specification does. The relationship is T = K × D × F, and K is a friction term: about 0.20 for plain dry steel and about 0.10 with moly grease. So the same torque on greased threads produces twice the preload it produces dry. Anti-seize is roughly half as dramatic and still enough to matter. "Torque to 75 ft-lb" is an incomplete instruction unless it also says what is on the threads, and applying a dry table figure to a fastener you have just anti-seized is one of the commonest ways to snap a bolt while believing you followed the book.
- Angle control bypasses friction almost entirely. Once a joint is snug, further rotation is geometry rather than friction, which is why cylinder heads, connecting rods and structural steel are specified by angle or by measured stretch. Note that the structural turn-of-nut method uses much larger angles than pure bolt elongation, because it also compresses the joint and deliberately runs the bolt past yield — the two numbers are not comparable.
- Stronger is not automatically better. Grade 8 and class 12.9 are less ductile than Grade 5 and 8.8, so they tolerate shock, bending and misalignment less well. In a joint that flexes, the harder fastener is the one that snaps.
- A big Grade 2 bolt is weaker than it looks. Proof strength drops from 55 to 33 ksi above ¾ inch, so torque goes down as the bolt gets bigger across that boundary. Class 4.6 has the same problem in metric.
- Stainless galls. The threads can cold-weld partway down and seize solid, at which point the fastener will neither tighten nor come out. Anti-seize is close to mandatory on it — which also means the dry figures never applied to it in the first place.
- Reused fasteners have already changed. They have yielded slightly at the thread roots and their surface friction is different. Backing the target off to about 65% of proof is cheaper than working out later which bolts were reused.
- Tensile stress area is not the major diameter. The load-carrying area sits between the major and minor diameters, and using the major overstates a bolt's capacity by around 25%. It is also why a fine thread takes more preload than a coarse one of the same nominal size.
How to use
- Pick the bolt size and its grade or class.
- Set the thread condition — dry, oiled, anti-seized or plated.
- Choose a preload target, or leave it at the standard 75% of proof.
- Read the torque, and check the spread table for what the dry figure would do.
Frequently asked questions
Why does lubrication change the torque so much?
Because torque is mostly fighting friction rather than stretching the bolt. The relationship is T equals K times D times F, where K is a friction term: roughly 0.20 for plain dry steel and about 0.10 with moly grease. That means the same torque on greased threads produces about twice the preload it produces dry. Anti-seize is around half as dramatic and still enough to matter. Torque to 75 foot-pounds is an incomplete instruction unless it also says what is on the threads.
How much of my torque actually becomes clamp force?
Only about 10 to 15 per cent. Roughly half is consumed by friction under the head or nut bearing face and most of the rest inside the threads, so a torque wrench is really a friction meter that infers the tension you care about. Even carefully applied, torque control scatters the resulting preload by something like a quarter to a third, which is why critical joints are not specified by torque at all.
What should I use instead of torque on a critical joint?
Angle control after snug, or a direct stretch measurement. Once a joint is snug, further rotation is geometry rather than friction, so angle largely sidesteps the variable that makes torque unreliable. That is why cylinder heads, connecting rods and structural steel are specified that way. Measuring bolt elongation with a micrometer is more accurate still where both ends are accessible, and load-indicating washers are the practical option where they are not.
Is the turn-of-nut angle here the same as the structural one?
No, and confusing them is dangerous. The angle shown here is the nut rotation corresponding to the bolt elongation alone, from a snug start. The structural turn-of-nut method specifies much larger angles, a third of a turn and up, because it also compresses the joint itself and deliberately takes the bolt past yield into the plastic region. They are different procedures answering different questions and their numbers should never be compared.
Why is a big Grade 2 bolt weaker than a smaller one?
Because SAE Grade 2 proof strength drops from 55,000 to 33,000 psi above three-quarters of an inch, so torque can go down as the bolt gets bigger across that boundary. Grade 5 does the same thing above one inch, falling from 85,000 to 74,000. Metric class 4.6 has the equivalent problem. It is a genuine trap: the fastener looks more capable and is specified for less.
Is a stronger bolt always the safer choice?
No. Grade 8 and class 12.9 are harder and stronger but markedly less ductile than Grade 5 and 8.8, so they tolerate shock loading, bending and misalignment less well. In a joint that flexes, the harder fastener is the one that snaps, and it snaps without the visible stretching that would have warned you. Match the grade to the joint rather than reaching for the strongest thing in the drawer.
Why does stainless need special treatment?
Because it galls. Under pressure the thread surfaces cold-weld to each other and the fastener seizes partway down, at which point it will neither tighten nor come out and usually has to be cut off. Anti-seize and slow assembly are close to mandatory, which also means the dry torque figures never applied to stainless in the first place. Stainless is also mechanically unremarkable: 18-8 is weaker than Grade 5 despite costing more.
What is tensile stress area and why not just use the diameter?
It is the effective load-carrying area of a threaded fastener, sitting between the major and minor diameters rather than at either one. Using the major diameter overstates a bolt capacity by around 25 per cent, which is not a rounding error. It is also why a fine thread takes more preload than a coarse thread of the same nominal size: cutting a shallower thread removes less material, so more of the shank is left to carry load.
🔒 This tool runs entirely in your browser. Nothing you enter is uploaded, logged, or stored.