Technical Library

Tightening torque

Torque is a proxy for clamp load, and a poor one. Roughly 85 to 90% of the torque you apply is spent overcoming friction under the head and in the threads. Which is why the finish on the part changes the answer more than the grade does.

Indicative torque, dry / as-plated

Computed as T = K · d · F, where F = 0.70 · R_p0.2 · A_s - that is, tightened to 70% of proof - with a nut factor K = 0.20, typical of a plain or zinc-plated fastener assembled dry. Values in newton metres.

Size	8.8	10.9	12.9	A2-70 / A4-70	A2-80 / A4-80
M3	1.4	2.0	2.3	1.0	1.3
M4	3.1	4.6	5.4	2.2	3.0
M5	6.4	9.3	11	4.5	6.0
M6	11	16	19	7.6	10
M8	26	39	45	18	25
M10	52	76	89	37	49
M12	91	133	156	64	85
M14	144	212	248	101	135
M16	225	331	387	158	211
M20	439	645	755	309	412
M24	759	1115	1305	534	712

The same table, lubricated

With a lubricant, a wax, or an integrated lubricant in a zinc flake coating, K falls to roughly 0.14. For the same clamp load you need about 30% less torque. Apply the dry figures to a lubricated part and you will over-tighten it, potentially past yield.

Size	8.8	10.9	12.9	A2-70 / A4-70	A2-80 / A4-80
M3	0.9	1.4	1.6	0.7	0.9
M4	2.2	3.2	3.8	1.5	2.1
M5	4.5	6.5	7.7	3.1	4.2
M6	7.6	11	13	5.3	7.1
M8	18	27	32	13	17
M10	36	53	63	26	34
M12	63	93	109	45	59
M14	101	148	174	71	95
M16	158	231	271	111	148
M20	307	451	528	216	288
M24	531	780	913	374	498

Read this before using either table. These are calculated reference figures on stated assumptions, not a torque specification. Real values depend on the coating and its lubricity, the joint material and whether it yields under the bearing face, washer use, thread engagement length, reuse, temperature and tool accuracy. A joint that matters should be validated by test, not by table. Where a customer drawing specifies torque, the drawing governs.

Practical notes

Coating changes torque more than grade does. Moving from plain zinc to zinc flake with integrated lubricant can shift required torque by 30% at identical clamp load.
Soft joints limit you before the bolt does. In aluminium, magnesium or plastic the bearing face fails first, so torque is set by the joint material, not by the fastener class.
Thread engagement. As a rule of thumb, engagement of about one diameter in steel and closer to two in aluminium, so the thread strips after the bolt yields rather than before.
Reuse. Friction changes after the first tightening. Prevailing-torque locking nuts and patch-coated screws have a stated number of reuses; beyond it, replace.
Stainless galls. Austenitic stainless can cold-weld during tightening. See the stainless page for how to avoid it.

How to use these tables. The values here are standard reference figures published to help you specify and compare parts. They are guidance, not a substitute for your own engineering standard, and several of them (torque above all) depend on assumptions about friction, lubrication and joint material that will differ in your application. Where a number matters, confirm it against the governing specification and against the part drawing. We will supply the plant's actual test data with any quotation.

Request a Quote Back to the Technical Library