A tightening torque is turned into clamp load through the short form T = K·D·F, where D is the bolt's diameter, F the clamp load and K, the nut factor, lumps together the friction in the threads and under the head. Fastenal's own chart says "Torque is only an indirect indication of tension." We held one M12 class 8.8 bolt at 88 N·m, the dry torque Fastenal's metric chart prints for it (75% of proof load at K 0.20), and solved for clamp load at each K published by NASA, Fastenal, Bossard's VDI 2230 tables and the Fastener Facts table. The answers ran from 24.4 to 63.3 kN, more than twice. "Lubricated" alone is published as 0.15 by Fastenal and 0.18 in Fastener Facts. The high end is 129% of the bolt's 48.9 kN proof load, so it is arithmetic rather than a load the joint would hold; a lower K simply means the same torque stretches the bolt further.
Dry steel at a friction of about 0.15 corresponds to the textbook nut factor
NASA's Fastener Design Manual, Reference Publication 1228 by Richard Barrett (1990, NTRS 19900009424), says "the commonly assumed value for K is 0.2, but this value should not be used blindly", and offers 0.15 as a more realistic typical value for steel on steel. Its Table VI, with thread and head friction equal, puts a friction coefficient of 0.15 at K 0.189.
The Fastener Facts table, published in 1985 by Bowman Distribution-Barnes Group, is the four-row list repeated in Shigley's textbook; we read it in SOLIDWORKS Help and count its copies as one source. That page says average friction values are "approximately 0.15" and that "if the bolt condition is unknown, K = 0.20 is recommended". Fastenal also uses 0.20 for plain, dry conditions.
Zinc plating eases thread friction in some tables and matches plain steel in others
Fastenal's inch chart gives K 0.17 "for zinc plated and dry conditions", below its plain dry 0.20. Fastener Facts lists zinc-plated at 0.20, the same as plain dry, and reserves 0.30 for non-plated black finish. Bossard's worked example for an electro-zinc-plated M12 class 8.8 screw gives friction of 0.14–0.24, straddling the 0.15 that corresponds to K 0.2. Across the three, zinc does not sit consistently below plain steel.
Until 19 October 2026 our bolt torque calculator offered three nut factors, two of them unsourced: 0.10 labelled "lubricated" and 0.30 labelled "zinc-plated". An M12 class 8.8 zinc bolt at 75% of proof came out at 132.0 N·m, which at Fastenal's 0.17 clamps 132% of proof load. Any preload percentage passed (200% gave 234.7 N·m). Its prose suggested less preload for vibration-sensitive joints, where NASA's manual says higher load increases vibration resistance, and called angle tightening independent of friction.
The revision committed that day lists each published K with its source, accepts a K of one's own between 0 and 1, and refuses a preload at or below zero or above 100% of proof. Fastenal's zinc 0.17 now gives 74.8 N·m for that bolt. With M12, class 8.8, "Plain and dry" (Fastenal) and 75%, it returns 88.0 N·m and shows this torque clamping 48.9 kN at K 0.15 and 24.4 kN at 0.30.
Oil and wax share the lubricated label across several nut factors
Fastenal's metric chart estimates K at 0.13 for Ecoguard, uses 0.15 for "lubricated" and 0.20 for "dry", and notes that waxed zinc-plated Top Lock nuts can vary between 0.12 and 0.16. Fastener Facts gives lubricated as 0.18.
At 88 N·m on the M12, by our arithmetic, Fastenal's lubricated 0.15 gives 48.9 kN and the 0.18 gives 40.7 kN. The lowest K we used, 0.116, is ours: Bossard publishes 1.39 N·m per kN for an M12 at friction 0.08, which we divided by the diameter.
Wheel-nut threads go without oil or grease in Toyota's Japanese manuals
Toyota's Japanese owner's manual for the Yaris Cross hybrid (vehicles from August 2026) gives a tightening torque of 103 N·m (1050 kgf·cm) and says ねじ部にオイルやグリースを塗らない, do not apply oil or grease to the threads. It warns that otherwise the nut is tightened more than necessary and the bolt could break. Toyota's Japanese Mirai manual gives 140 N·m (1428 kgf·cm) with the same warning.
Tesla's 2017–2023 Model 3 service manual reads: "Fully tighten the lug nuts in a criss-cross star pattern. 175 Nm (129.1 lbs-ft)." That page names no thread condition. None of the three pages states a stud diameter, pitch or class, and each figure applies to the model and manual named. Engine torque, the other everyday N·m, is the subject of our horsepower and torque guide.
Steel rubbing under the bolt head absorbs about half the torque
The long form splits torque into a pitch term, which stretches the bolt, and two friction terms. Bossard's preload tables are based on VDI 2230 (2015 edition) and assume medium clearance holes to ISO 273. We computed it for an M12 coarse thread (pitch 1.75 mm), assuming an 18 mm head across flats, a 13.5 mm hole and equal thread and head friction.
| Friction | Long-form K (ours) | Bossard, as K (ours) | Stretching | Under the head |
|---|---|---|---|---|
| 0.08 | 0.118 | 0.116 | 20% | 45% |
| 0.10 | 0.142 | 0.138 | 16% | 46% |
| 0.12 | 0.165 | 0.162 | 14% | 48% |
| 0.14 | 0.189 | 0.185 | 12% | 49% |
Our long form lands within about 3% of Bossard's table, an agreement that rests on the head and hole sizes we assumed. On it, only 12–20% of the torque stretches the bolt, and the threads take 36–39%. A split repeated in references we checked, about 15% to the bolt, 39% to the threads and 46% under the head for coarse thread, matches this arithmetic.
Class 8.8 steel changes its proof stress partway up the size range
Fastenal's ISO 898-1 chart gives an M12 class 8.8 bolt a minimum proof load of 48,894 N. Dividing its proof loads by the stress areas gives 580 MPa up to M16 and 600 MPa at M18, on our arithmetic. Fastenal's inch chart calculates clamp load as 75% of proof load. By proof stress derived from that chart, class 8.8 sits near the US SAE Grade 5 (about 586 MPa up to an inch).
Threads tightened by torque alone vary by about a quarter in clamp load
In tables reproduced in NASA's manual, the Industrial Fasteners Institute's Fastener Standards (1970) gives preload accuracy of ±25% with a torque wrench and ±15% with turn of the nut. Machine Design (19 November 1981) gives ±15 to ±30% when torque is controlled and the same band when turn is controlled.
For commercial, modern torque wrenches Bossard applies a tightening factor of 1.6 to 2.0, the ratio of highest to lowest preload: ±23% to ±33% around a midpoint, on our arithmetic. In its zinc-plated M12 example a factor of 1.8 takes a 41.9 kN maximum down to a 23.3 kN minimum. Because clamp load varies with 1/K, the spread is lopsided: K from 0.15 to 0.25 around 0.20 moves clamp load from 33% above to 20% below.
View author profile → · Editorial policy
