Tightening Torque Calculator
Compute the tightening torque and preload for a metric bolted joint — a quick K-factor estimate, or the detailed VDI 2230 method that separates thread friction from bearing-face friction.
Pick a thread size, strength class and surface/lubrication condition, and the torque updates live. The Quick method is the fastest way to get a working number using a single empirical nut factor K. The Precise (VDI 2230) method breaks the same physics into its two real components — friction at the thread flank and friction under the bolt head or nut — and shows a live breakdown of where the applied torque actually goes, which is the part most torque charts never show you.
Quick method — nut factor K
MA = K · d · Fv — one empirical factor covers thread and bearing friction together.
K = 0.20
D_Km defaults to the standard VDI 2230 approximation, D_Km ≈ 1.36 · d. Edit it if you know the actual bearing-face and clearance-hole diameters for your fastener — see the standard components tool for dimensions. µG/µK reset to the table value whenever you change the surface condition above; edit them afterwards to override.
Guidance only. For safety-critical joints (pressure equipment, structural connections, anything with a real consequence of failure), verify against the fastener manufacturer's torque specification or a full engineering analysis — friction coefficients from a table are a starting estimate for your specific parts, not a guarantee. See the FAQ below.
How the numbers are calculated
Both methods start from the same target: a preload Fv sized to a fraction β (70–90%) of the bolt's proof stress Rp0.2, applied over its tensile stress area As (computed from the pitch and minor diameter, ISO 898-1) — Fv = β · Rp0.2 · As. They differ in how they turn that target preload into a torque number.
Quick method (nut factor K)
K bundles thread-flank friction, bearing-face friction and the thread's own mechanical advantage into a single empirical number, typically 0.10–0.20 depending on surface finish and lubrication — see the surface picker above. It's fast, but it hides where the torque actually goes and — because K itself varies from joint to joint even under nominally identical conditions — the resulting preload commonly scatters by around ±25% at a fixed torque, which is why the result above is shown as a band, not a single number.
Precise method (VDI 2230)
MG = Fv · ( P⁄(2π) + µG · d₂⁄(2·cos 30°) )
MK = Fv · µK · D_Km⁄2
MG is the torque needed to advance the nut along the thread's helix — the first term (P⁄2π) is the only part of the whole calculation doing real mechanical work against the preload; the second term is friction at the 30° thread flank. MK is friction under the bolt head or nut face, over the mean bearing diameter D_Km. Keeping the two friction terms separate lets you see, live, in the breakdown bar above, how much of MA is useful work versus friction — for a typical dry M10 8.8 bolt it's roughly 10–15% useful, ~35% thread friction, ~50% bearing friction, which matches the "80–90% of tightening torque is lost to friction" figure often quoted in fastening engineering literature.
Torque scatter and tightening method accuracy
The ±25% band above is about friction scatter — how much the same nut factor K actually varies between nominally identical joints — not about how precisely a tool can hit a torque setpoint. Those are two different sources of error, both real, that stack on top of each other: a well-calibrated torque wrench is commonly accurate to within a few percent of its own reading, but the resulting preload still scatters by roughly ±25% because of friction variation. Angle-controlled tightening (torque to a snug point, then a further fixed rotation) and hydraulic/mechanical tensioning largely sidestep the friction uncertainty, which is why they're preferred over plain torque control for fatigue-critical or highly-loaded joints.
Glossary
- K (nut factor)
- Empirical constant bundling thread and bearing-face friction plus thread lead into a single number for the Quick method, MA = K·d·Fv. Depends mainly on surface finish and lubrication.
- Fv (preload / clamp force)
- The axial tension the bolt is stretched to during tightening — the actual clamping force holding the joint together. Both methods target the same Fv = β · Rp0.2 · As.
- Rp0.2 (proof stress)
- The 0.2% offset yield strength for the bolt's ISO 898-1 property class (e.g. 660 MPa for class 8.8) — the stress limit the preload is sized against.
- As (tensile stress area)
- The effective cross-section area used for stress calculations on a threaded fastener, As = (π/4)·((d₂+d₃)/2)² — larger than the minor-diameter area, smaller than the nominal one, because it accounts for the helical thread geometry.
- β (utilization factor)
- The fraction of Rp0.2 the target preload is allowed to reach, typically 70–90%. Higher β means more clamp force but less margin before yielding.
- µG, µK (friction coefficients)
- µG is friction at the thread flank; µK is friction under the bolt head or nut bearing face. Used separately by the Precise/VDI 2230 method — the Quick method lumps both into K.
- D_Km (mean bearing diameter)
- The mean diameter of the annular contact area under the bolt head or nut, roughly the average of the bearing-face outer diameter and the clearance-hole diameter. Approximated here as 1.36·d when not entered directly.
- MG, MK, MA
- Thread-friction torque, bearing-friction torque, and their sum — the total tightening torque to apply.
Frequently asked questions
What torque should I use for a bolt?
It depends on the bolt's diameter and pitch, its strength class, the friction at the thread and under the head, and how much of the material's yield strength you're willing to use. There is no single universal number for "an M10 bolt" — the same M10 8.8 bolt needs a different torque dry than it does lightly oiled. Pick the thread size, strength class and surface condition that match your actual joint below.
What is the difference between the Quick and Precise methods?
The Quick method lumps thread friction, bearing friction and the thread's mechanical advantage into a single empirical "nut factor" K, in the spirit of the simplified torque-tension relationship commonly used alongside ISO 898-1 strength data (MA = K · d · Fv). The Precise method, from VDI 2230, keeps thread friction (MG) and bearing-face friction (MK) as two separate terms with their own friction coefficients (µG, µK) and geometry, which is more accurate and lets you see where the torque actually goes — see the breakdown bar in Precise mode.
Why does the same bolt need different torque depending on surface finish?
Torque doesn't directly control preload — friction does, and friction depends heavily on surface finish and lubrication. For the same target preload, a bolt with a low-friction coating (MoS2, oiled) needs noticeably less torque than a dry, uncoated one, because less of the applied torque is lost overcoming friction at the thread and under the head. Mixing up the surface condition is one of the most common real-world causes of over- or under-tightened joints.
What does the ±25% scatter band mean?
It's the typical spread in actual preload you get from torque-controlled tightening at a fixed K factor, even with a perfectly calibrated torque wrench — because K itself varies joint to joint (surface roughness, exact lubricant film, thread condition all vary slightly even within the same batch). It is not the same thing as torque-wrench accuracy: a wrench can be accurate to ±3% and the resulting preload can still scatter by ±25%, because the uncertainty is in the friction, not the tool. Angle-controlled and hydraulic tensioning methods largely avoid this source of scatter, which is why they're preferred for critical joints.
Is this calculator accurate enough for safety-critical joints?
Use it for estimating, checking orders of magnitude, and learning how the numbers relate to each other. For a safety-critical joint (pressure equipment, structural connections, anything with a real consequence of failure), verify against the fastener manufacturer's torque specification, a full VDI 2230 analysis including fatigue and combined tension-torsion yield checks, or testing on the actual joint — friction coefficients from a table are a starting estimate, not a guarantee for your specific parts.
Where do I find D_Km (bearing diameter) for my actual bolt?
D_Km is the mean diameter of the annular contact area under the bolt head or nut, roughly the average of the bearing-face outer diameter and the clearance-hole diameter. This calculator defaults to the common VDI 2230 approximation D_Km ≈ 1.36 · d, which is close enough for most standard hex and socket-head bolts. If you know your fastener's actual head and hole dimensions — the standard-components tool below has technical drawings and dimension tables for hex bolts, socket-head bolts and countersunk bolts — enter D_Km directly for a more accurate result.
References & further reading
- NASA Reference Publication 1228 — Fastener Design Manual — public-domain NASA manual covering torque formulas, preload and fastener strength derivations.
- VDI 2230 Blatt 1 — Systematic calculation of highly stressed bolted joints — the reference German engineering guideline for detailed bolted-joint design.
- Preload (engineering) — Wikipedia — background on why bolts are tightened to a target preload rather than just "snug".
Carry this calculator in your pocket
MechaHandbook has this torque calculator and more — thread charts, tolerances, standard components and unit conversions — all offline, no internet required.
Download the app