Does a higher K value mean more or less friction?

A higher K value corresponds to more friction in the joint. Because more torque is consumed overcoming friction before any of it converts into clamp force, a higher-K joint requires more applied torque to reach the same target preload than a lower-K joint would.

Can K ever be greater than 1 or negative?

K is a positive dimensionless number, and for common fastener conditions it typically falls in a range roughly between 0.10 and 0.30. Values well outside this range would indicate either unusual friction conditions or an error in how K was measured or applied.

Is K the same thing as a coefficient of friction?

Not exactly — K is closely related to friction but is a combined empirical factor that also accounts for thread geometry and the specific way torque converts to axial force through the thread helix, not a single, direct coefficient-of-friction measurement on its own.

Engineering·6 min read·August 14, 2026

🔧 The Nut Factor K: Where Bolt Torque Actually Goes

Why only about 10% of applied bolt torque converts into clamp force, where the rest is lost to friction, and why the nut factor K has to be measured rather than calculated.

By EngineersUniverse Editorial Team

Engineering Editorial & Technical Review Team

Published August 14, 2026

A Surprising Breakdown of Applied Torque

When a torque wrench applies torque to tighten a bolt, it is natural to assume most of that effort goes into stretching the bolt and clamping the joint together. In reality, only a small fraction of applied torque — commonly cited around 10 percent — actually converts into useful bolt stretch and clamp force (preload). The overwhelming majority is consumed by friction before it ever contributes to clamping the joint.

The Three-Way Split of Applied Torque

Applied torque divides roughly three ways: friction under the bolt head or nut bearing face typically consumes the largest share, commonly cited around 40 to 50 percent; friction in the engaged threads consumes another large share, commonly cited around 35 to 45 percent; and only the remaining roughly 10 percent actually converts into the bolt stretch that produces real clamp force. This breakdown explains why torque, despite being the easiest quantity to measure at the wrench, is really a proxy for clamp force rather than a direct measurement of it.

Why the Nut Factor K Has to Absorb All of This

The nut factor K in the standard torque-tension equation T = K times D times F is a single empirical number that lumps together every one of these friction contributions — head/nut bearing friction, thread friction, and the geometric efficiency of the thread helix angle — into one coefficient. This is a deliberate simplification: rather than modeling each friction source separately (which would require detailed knowledge of surface roughness, lubrication, contact pressure, and other hard-to-predict variables), the torque-tension equation absorbs all of that complexity into K, making the formula usable with a single measured or referenced value.

Why K Cannot Be Calculated From First Principles

Because K represents the combined effect of friction phenomena that depend on surface finish, lubrication, contact pressure, material combination, and even installation speed, it cannot be reliably derived from a theoretical formula the way, say, section modulus can be calculated from pure geometry. K has to be determined experimentally — either through controlled laboratory testing of the specific fastener, coating, and lubrication combination, or estimated from published reference values for broadly similar conditions (dry steel, lubricated steel, galvanized steel, and so on).

Why This Makes Torque-Based Preload Control Inherently Approximate

Since K is an empirical average representing typical friction behavior for a given condition, but actual friction varies somewhat from one specific bolt, nut, and washer combination to the next — even within the same batch, coating lot, and lubrication method — using a reference K value in the torque-tension formula produces a good estimate, not an exact prediction, of actual achieved preload. This inherent scatter, covered in more depth in the companion article on torque scatter, is a direct and unavoidable consequence of K being a friction proxy rather than a precisely measured, joint-specific quantity.

Why Understanding This Changes How You Read a Torque Spec

Recognizing that K absorbs friction, not geometry, explains why a torque specification is only meaningful when paired with the surface condition it was derived for — a torque value calculated or specified for a dry, as-received bolt does not apply to the same bolt once it has been lubricated, coated, or reused, because the underlying friction behavior K represents has genuinely changed even though the bolt geometry has not.

Topics covered

nut factor K bolt torquebolt torque friction losstorque tension relationshipwhere does bolt torque go