Dry vs lubricated torque: the K factor and how much to reduce
For the same clamp load, an oiled steel bolt needs about 25% less torque than a plain dry one: K drops from 0.20 to 0.15. Copper anti-seize is 20% less and moly paste 45% less. Apply a dry torque to lubricated threads and the bolt is overloaded by the same ratio.
Contents8 sections
Key takeaways
- K, the nut factor, is not a friction coefficient. It combines thread geometry, thread friction and underhead friction, and is found by test 1.
- Fastenal's lab K ranges: black or plain 0.20 to 0.30, zinc 0.17 to 0.22, lubricated 0.12 to 0.16, cadmium 0.11 to 0.15 1.
- Same clamp load means T_lubed = T_dry × K_lubed ÷ K_dry. From K 0.20: oil −25%, anti-seize −20%, moly −45%.
- NASA warns that the common K = 0.2 'should not be used blindly' and suggests 0.15 as a more realistic typical value for steel on steel 2.
- Vehicle wheel specs are dry-thread specs: GM and Ford manuals say no oil or grease, clean threads 3 4.
Quick facts
| Plain dry | K 0.20 1 |
|---|---|
| Zinc-plated dry | K 0.18 1 |
| Oil or grease | K 0.15 1 |
| Copper anti-seize | K 0.16 5 |
| Moly paste | K 0.11 6 |
Why lubrication changes the torque
A torque wrench measures how hard the fastener resists turning, not how hard it clamps. Most of that resistance is friction. Fastenal estimates that about 90% of the input energy goes into friction under the head and in the threads, and only about 10% into stretching the bolt 1. Bolt Science puts the friction share at usually 85 to 95% of the applied torque 7.
Lubricant lowers the friction. The same torque then turns the nut further and stretches the bolt more. That is good if you lower the torque to match, and bad if you don’t.
The short-form equation captures it: T = K × D × F 1. K is the nut factor. Fastenal is careful to say it is “not to be confused with the frictional coefficient”: it combines a geometric factor, a thread-friction factor and an underhead-friction factor, and for real accuracy it is measured on a torque-tension load cell 1. NASA RP-1228 gives the link to friction for equal thread and underhead coefficients 2:
| Friction coefficient | K |
|---|---|
| 0.05 | 0.074 |
| 0.10 | 0.133 |
| 0.15 | 0.189 |
| 0.20 | 0.250 |
Friction coefficients themselves range from 0.04 to 1.10 depending on materials and lubricant, according to the same manual 2.
Published K values
Nobody publishes one K for “dry” or one for “oiled”. These are the values with a named source:
| Condition | K | Source |
|---|---|---|
| Non-plated, black finish | 0.20 to 0.30 | Fastenal lab estimate 1 |
| Plain and dry (Fastenal’s torque tables) | 0.20 | 1 |
| Zinc-plated | 0.17 to 0.22 | Fastenal lab estimate 1 |
| Zinc-plated and dry (Fastenal’s tables) | 0.18 | 1 |
| Lubricated | 0.12 to 0.16 | Fastenal lab estimate 1 |
| Lubricated (Fastenal’s tables) | 0.15 | 1 |
| Cadmium-plated | 0.11 to 0.15 | Fastenal lab estimate 1 |
| Steel on steel, “typical” | 0.15 | NASA RP-1228 2 |
| Copper anti-seize (LOCTITE C5-A), 1/2 in. Grade 8 bolt | 0.16 | Henkel TDS 5 |
| Moly paste (LOCTITE LB 8012), 1/2 in. Grade 8 bolt | 0.11 | Henkel TDS 6 |
| Same hardware, solvent cleaned, no lubricant | 0.27 | Henkel TDS 5 |
Note the last row. Henkel’s solvent-cleaned, unlubricated baseline is 0.27, well above the 0.20 most tables use for “dry”. On phosphate-coated Grade 5 hardware Henkel measured 0.19 for the same clean condition 8. “Dry” is not one condition.
This site uses one value per condition (the five in the quick facts) so its charts and calculators stay consistent with each other.
How much to reduce
The reduction comes straight from the ratio of nut factors. Keep the clamp load, change K:
T_lubricated = T_dry × K_lubricated ÷ K_dry
| Condition | K | Multiply dry torque by | Reduction |
|---|---|---|---|
| Zinc-plated dry | 0.18 | 0.90 | 10% |
| Copper anti-seize | 0.16 | 0.80 | 20% |
| Oil or grease | 0.15 | 0.75 | 25% |
| Moly paste | 0.11 | 0.55 | 45% |
Scroll the table sideways to see every column.
Two real sizes from the site’s charts, at 75% of proof load:
- 1/2-13 Grade 8: 107 ft-lb (144 N·m) dry, 80 ft-lb (108 N·m) oiled, for 12,780 lbf.
- M10 x 1.5 class 8.8: 37 ft-lb (50 N·m) dry, 28 ft-lb (38 N·m) oiled, for 25.2 kN.
NASA gives one reduction outright: silver-plated nuts coated in wax, where “the normal dry torque values of the torque tables should be reduced by 50 percent” 2.
The lubricated torque calculator does the arithmetic for any pair of conditions.
What happens if you get it wrong
Dry torque on lubricated threads. The clamp load rises by K_dry ÷ K_lubed. Oil: a third more. Moly: 82% more, which takes a bolt aimed at 75% of proof to about 136%. Fastenal works this through for a 1/2-13 Grade 5 bolt: at its normal 905 in-lb, an excessively lubricated K of 0.10 gives 18,100 lb, more than the bolt’s tensile strength 1.
Lubricated torque on dry threads. The clamp load falls by K_lubed ÷ K_dry. Using an oiled figure on a dry bolt leaves it a quarter short. Fastenal points out that under-clamping causes field failures too 1.
Which condition is your spec written for?
- Manufacturer specs say. GM’s owner’s manual: “Never use oil or grease on bolts or nuts because the nuts might come loose” 3. Ford: “Torque specifications are for nut and bolt threads free of dirt and rust” 4. Those are dry-thread specs, and the instruction is to keep them dry, not to recalculate.
- Engine procedures often specify oil. Some head bolt procedures call for oil on the threads and some forbid it. See torque sequence.
- Generic charts should state K. Every chart on this site prints its K under the table. A chart that does not is ambiguous by about 25%.
Things that change K besides the lubricant
Fastenal lists the variables: hardness of all parts, materials, class of fit, plating type and thickness, surface finish, cut or rolled threads, washers present or not, the tightening tool, tightening speed, whether the nut or bolt is turned, how many times the fastener has been used, and the type, amount, condition, application, contamination and temperature of any lubricant 1. NASA-STD-5020B adds that lubricating only under the head behaves differently from lubricating under the head and the nut 9.
The practical rule: decide the thread condition, apply it the same way to every fastener in the joint, and use the K for that condition. For compound-specific guidance, read anti-seize and torque and thread locker and torque.
Sources
- 1
- 2
- 3
- 4
- 5
- 6Henkel Corporation (LOCTITE), LOCTITE LB 8012 Moly Paste TDS, April 2015Checked 21 Oct 2026
- 7Bolt Science Ltd, Bolt Science, Methods of Tightening Threaded FastenersChecked 21 Oct 2026
- 8
- 9
Spotted a value that doesn’t match its source? Report an error
Related
- Lubricated torqueConvert a dry torque spec to oiled, zinc, anti-seize or moly threads by the ratio of nut factors. 100 ft-lb dry becomes 80 ft-lb with copper anti-seize.
- Anti-seize torqueHow much to cut torque with anti-seize, from LOCTITE data sheets: copper and metal-free K 0.16, marine and silver 0.18, moly 0.11. And why NGK says no on plugs.
- Thread locker torqueLOCTITE 242 changes the nut factor by no more than ±10%, so torque stays at the dry value. Henkel publishes no figure for 243, 262 or 271. What the data sheets do say.
- Torque vs tensionTorque only estimates bolt tension: about 90% goes into friction, so preload scatters ±25% or more. The figures from NASA-STD-5020B, NASA RP-1228, Fastenal and RCSC.