Tightening Torque Calculator
Torque from thread geometry and property class, with the friction condition shown as a control instead of buried in a constant. Every figure is computed from this database's own pitch and minor diameters, not from a pasted table.
66.8 N·m (49.3 lb·ft)
Target preload 33.4 kN · stress area 58.0 mm² · K = 0.2
Where that number comes from
check the arithmetic before you set a wrench| Stress area As | π/4 × ((9.030 + 8.160) / 2)² = 58.02 mm² |
| Yield strength Rp0.2 | 640 MPa |
| Target preload (90% of yield) | 0.9 × 640 × 58.02 = 33,420 N = 33.4 kN |
| Torque coefficient K | 0.2 |
| Tightening torque | 0.2 × 10.000 × 33,420 ÷ 1000 = 66.8 N·m |
Only about 10% of the torque you apply becomes clamping force — roughly half is lost to thread friction and 40% to the bearing face.
Friction is the biggest variable here
same M10 × 1.5 at Property Class 8.8, four lubricant conditions| Condition | Note | K | Torque | vs dry |
|---|---|---|---|---|
| Dry / as received | Plain steel, no lubricant | 0.2 | 66.8 N·m | 100% |
| Lightly oiled | Oil on threads and bearing face | 0.16 | 53.5 N·m | 80% |
| Zinc plated | Electroplated, unlubricated | 0.22 | 73.5 N·m | 110% |
| MoS₂ paste | Molybdenum disulphide lubricant | 0.12 | 40.1 N·m | 60% |
Setting the same torque on a lubricated joint as on a dry one overloads the fastener: the lubricant lets more of that torque turn into clamp force. This is why a torque figure is meaningless without stating the lubrication.
Torque to 90% of yield — dry, unlubricated
metric coarse threads, K = 0.20, in N·m| Thread | 4.8 | 8.8 | 10.9 | 12.9 | A2-70 Stainless |
|---|---|---|---|---|---|
| M1.6 × 0.35 | 0.1 | 0.2 | 0.3 | 0.4 | 0.2 |
| M2 × 0.4 | 0.3 | 0.5 | 0.7 | 0.8 | 0.3 |
| M2.5 × 0.45 | 0.5 | 1.0 | 1.4 | 1.7 | 0.7 |
| M3 × 0.5 | 0.9 | 1.7 | 2.6 | 3.0 | 1.2 |
| M3.5 × 0.6 | 1.4 | 2.7 | 4.0 | 4.7 | 1.9 |
| M4 × 0.7 | 2.2 | 4.0 | 5.9 | 7.0 | 2.8 |
| M5 × 0.8 | 4.3 | 8.2 | 12.0 | 14.0 | 5.7 |
| M6 × 1 | 7.4 | 13.9 | 20.4 | 23.9 | 9.8 |
Computed from the thread geometry and the Rp0.2 of each class, at 90% of yield with a friction coefficient of 0.20. Torque-preload scatter is typically ±25% with the K-factor method. For a critical joint, calibrate against the actual fastener and lubricant.
Using these numbers
Torque is not a material property
Two identical bolts in the same joint need different torque if one is plated and the other is not. Always state the lubrication and coating alongside the figure, or the number travels badly.
Torque controls preload only loosely
About 90% of what you apply goes into friction, and friction varies with surface finish, coating and how the parts were handled. Expect ±25% scatter even when you do everything right.
When the joint is critical
Move past torque control: use a torque-and-angle sequence, a load-indicating washer, or measure bolt elongation directly. For anything structural, follow the governing standard rather than a calculator.
Metric vs inch
The formula is unit-agnostic once everything is in millimetres and newtons — this page converts the inch threads before calculating. The result is always N·m, with lb·ft alongside.
Frequently asked questionstorque, preload and the friction between them
What is the tightening torque for an M10 8.8 bolt?
About 67 N·m dry and unlubricated, based on 90% of yield (Rp0.2 = 640 MPa) and a friction coefficient of 0.20. Lubricate the threads and the same bolt needs closer to 53 N·m — the reduction comes from the lubricant, not from any change in the bolt.
Does lubrication really change the torque that much?
Yes, and it is the single largest variable on this page. Approximately 90% of the torque you apply is absorbed by friction — roughly half in the threads and 40% under the head. Change the friction and you change how much of that input becomes clamping force, so the same torque can over-tension a lubricated joint.
Why is the preload calculated from yield strength rather than tensile strength?
Because preload is limited by when the bolt starts to yield, not by when it breaks. Property class 8.8 has Rp0.2 of 640 MPa and Rm of 800 MPa; using the tensile figure would overstate the safe preload by 25% and produce a torque that stretches the bolt permanently on the first tightening.
Is torque the same thing as clamping force?
No. Torque is what you apply with the wrench; clamping force is what the joint actually feels. The conversion passes through friction, which is why two joints tightened to the same torque can differ in clamp load by tens of percent. Torque is a means of controlling preload, not a measurement of it.
Can I use these figures for stainless steel fasteners?
Use the stainless property classes (A2-70, A4-80) shown in the selector, and always lubricate. Stainless galls — cold-welds — against stainless under load, and an unlubricated stainless joint can seize well before it reaches the calculated torque. Anti-seize also brings the friction coefficient down, so the required torque drops with it.
How accurate is a torque calculation?
The K-factor method carries roughly ±25% scatter in the preload it actually produces, even with the arithmetic done perfectly, because friction varies with surface finish, coating, and how clean the parts were. For a critical joint, calibrate against the actual fasteners and lubricant, or move to a method that measures preload directly.