How to Determine Tightening Torque and Choose a Tightening Method for a Bolted Joint

Torque Is a Proxy for Clamp Load, Not the Goal

A torque value is not the objective; the clamp load it produces is. When a bolt is turned down its thread helix, rotary motion becomes linear motion, the bolt stretches and the clamped parts compress. But most of the input torque is spent overcoming friction rather than stretching the bolt ⁸. The breakdown is roughly 10% useful preload, 50% underhead friction and 40% thread friction ⁸. That is why the same torque can produce very different clamp loads depending on the friction condition of the joint. Start from the required preload instead. One vendor-tier source reports that most fasteners are torqued to reach 65–80% of proof load, and works an M6 steel bolt at 75% of proof load ¹⁶. For that bolt — proof load about 700 MPa over a tensile stress area of about 20 mm², so proof load is 14,000 N and the 75% target is 10,500 N — the source reports the target reached at roughly 7–8 N·m when lubricated ¹⁶. Treat that as a source-specific observation for a lubricated M6, not a value to copy onto a dry joint: the same bolt dry would need a different torque for the same clamp load. Before you compute anything, ask what preload the joint actually needs and how wide the usable window is. That window, not the torque figure, is what decides the method in the next step.

Computing a First-Pass Torque from Clamp Load and Nut Factor

The torque-preload relationship can be written as Torque = Load × Nominal Stud Size × Nut Factor / 12, where the nut factor K rolls thread geometry, pitch, nut-face friction and thread friction into one number ¹⁵. K is experimentally determined and varies with lubrication, so the same load gives different torques under different friction conditions ¹⁵. The worked instance in the source: for 50,000 lb of load on a 1-1/8 in stud using an anti-seize with a nut factor of 0.17, torque = 50,000 × 1.125 × 0.17 / 12 = 797 ft-lb ¹⁵. Anti-seize nut factors usually fall between 0.15 and 0.20 ¹⁵. That K is the weak link. It is not a constant you can assume — it must be measured for your fastener, coating and lubricant combination, or taken from the applicable standard. If you cannot obtain a measured K for your joint, the computed torque is only a starting estimate.

Matching the Tightening Method to the Preload Window

The usable preload window divided by each method's scatter determines which methods can hit the target. If a ±30% window fits, torque control is sufficient ¹⁹ (supplier-reported figure); if only ±10–15% fits, torque-angle is required ²³. The scatter figures below are representative ranges for competently executed work, not guarantees ¹⁹.

Adjusting Torque for Lubrication and Coating

Lubrication lowers friction, so it lowers the torque needed to reach the same preload ⁸. Applying a dry-steel torque spec to a lubricated fastener therefore over-tensions the joint — the bolt reaches a higher clamp load than the spec intended ²². If you are specifying a coated fastener, the torque value must account for the coating ²². Zinc flake coatings with a topcoat hold a friction coefficient typically between 0.09 and 0.16 depending on the topcoat specification, which gives a more predictable torque-tension relationship than untreated steel ¹¹. That consistency is the point: a controlled friction coefficient makes the torque-preload conversion more repeatable, so the torque you compute is closer to the preload you get. What the sources do not give you is a K value or torque reduction factor for each lubricant type. You will need to obtain that from the lubricant supplier or the applicable standard, and verify it on a sample joint.

Matching the Tool to the Method

The tool can dominate the scatter the method was chosen to achieve. Tool control accuracies range from ±0.25–5% for a worm-gear torque wrench to ±10–30% for an impact wrench, with computer-controlled tensioning at ±0.5–2% ^{12 19} (supplier-reported figure). A hydraulic tensioner is more accurate (±1–5%) than most torque wrenches (±3–15% bar, ±10–30% impact) ^{12 19} (supplier-reported figure). If you specify torque-angle control for a ±10–15% window but tighten with an impact wrench, the tool's own scatter can exceed the method's. Match the tool's accuracy class to the method before committing to production. One supplier notes that torque tightening is the oldest, simplest and most cost-effective method for controlled tightening ¹⁷. And another supplier reports that digital torque wrenches with real-time torque curves can detect abnormal tightening conditions ²⁵. Those are supplier positions, not independent verification of tool performance on your joint.

Planning the Verification Check

For safety-critical high-strength assemblies, the joint loses preload after tightening, so a final torque check immediately after installation can read a value that will not hold. Allow a relaxation period before the final check.

• Allow a 24–48 h relaxation period after initial tightening before the final torque check on safety-critical high-strength assemblies ²⁴.
• During that period, a fastener with early-stage hydrogen embrittlement will lose clamp load or fracture, so the check catches it ²⁴.
• Use calibrated torque wrenches or torque-angle equipment so fasteners are not overtightened beyond the specified preload ²⁴.
• Re-torque verification after the relaxation period confirms joint integrity ²⁴.

Confirming the Governing Standard Before Release

The sources establish that standards govern torque tables and calculation rules, but they do not reproduce them. Before you release a joint for production, confirm which standard or specification applies and obtain the actual table or calculation rule. The torque-preload relationship and the 65–80% of proof load target are general guidance ¹⁶; the specific torque value for your bolt size, grade and lubrication condition comes from the governing document, not from this article. If your joint is safety-critical or high-strength, the standard may also set requirements for relaxation time, verification method and documentation. Confirm those at the same time you confirm the torque value.

Where the Sources Disagree

The sources in this set do not disagree on the torque-preload relationship or the method scatter figures. The one dispute that touches this decision is about baking duration for zinc electroplated high-strength fasteners, which affects whether a coating process step is sufficient — not the torque value itself.

Method Accuracy in Practice: Scatter Bands, Structural Audit Methods, and Tool Calibration

The method table sets expectations for preload scatter, but on a real joint the achieved accuracy also depends on how the method is audited — daily calibration against measured tension, match-marked rotation, or a physical indicator. This section covers the reference accuracy hierarchy, the RCSC structural pretensioning methods and their verification discipline, and how torque wrench calibration drift changes the delivered torque over time.

Reference engineering tables place the common tightening methods on a consistent accuracy ladder: hand-feel ±35%, torque wrench ±25%, turn-of-nut ±15%, load-indicating washer ±10%, bolt elongation ±3–5%, and strain gauges ±1% — a hierarchy adapted from the NASA Fastener Design Manual, MIL-HDBK-60, and Machinery's Handbook (MechaniCalc). Independent method lists show the same ordering, with the elongation method at ±3–5% and strain gauges at ±1% (Alekvs). The width of the ladder reflects how indirectly each method controls preload: friction scatter in the torque-preload relationship can reach ±30% (MechCodex), which is why a torque wrench is described as 'only ±25% accurate' in the NASA chart (FiveFlute). Where preload is critical, sources recommend a method with more process control than torque alone (Flybear Fastener).

For structural steel, the RCSC specification defines four pretensioning methods — turn-of-nut, calibrated wrench, twist-off tension-control (TC) bolts, and direct-tension-indicator (DTI) washers — all starting from snug-tight, and split into control methods (rotation, calibrated torque) versus indicating methods (spline shears, washer bumps crush) (AnvilField). Turn-of-nut controls bolt stretch directly: from snug-tight, required rotation is 1/3 turn up to 4 diameters of bolt length, 1/2 turn from 4 to 8 diameters, and 2/3 turn from 8 to 12 diameters (SteelCalculator). The calibrated-wrench method is audited through tension, not a chart value: three representative bolts from the working lot are calibrated daily in a Skidmore-Wilhelm calibrator, and the wrench must be recalibrated when the lot changes, the air pressure changes, or after 4 hours of continuous use (SteelCalculator). Running a published torque value without that calibration is not the calibrated-wrench method at all (AnvilField).

TC bolts move the accuracy check into the fastener itself: a shear wrench counter-rotates the nut against the splined tail, and the spline shearing off at a calibrated groove gives visual confirmation of pretension (LRK Industries). The calibration lives in the assembly's proprietary lubrication, which is why field re-lubrication of TC bolts is not permissible — only the manufacturer may re-lubricate the assembly (Haydon Bolts). Separate from the method, the tool itself drifts: a click-type wrench that met ±3% when new may degrade to ±5% or worse with heavy use, detectable only through calibration against a known standard (Accredited Labs). Common acceptance guidance is ±4% error for general use and ±2% for critical applications (ToolsGearLab); no fixed recalibration interval is universal, since frequency of use and torque level both drive the schedule (Ten Supplies).

No candidate source gives a numeric scatter band for TC bolt installation (only qualitative 'Excellent'/'Very good' ratings), nor fixed recalibration intervals in time units for hand torque wrenches; the 4-hour/lott-change recalibration rule is documented only for the RCSC calibrated-wrench method. Tool calibration costs ($30–80 per wrench, SurgePV) reflect one commercial source and were not generalized.

FAQ

Related pages

• Nuts product page
• Calculating Bolt Preload and Clamp Force: Torque, Friction and Tightening Method
• Installing a Bolt to a Specified Torque: Procedure, Tooling and Verification
• Bolt Head Markings: What the Lines, Numbers and Letters Claim — and How to Verify the Grade Before Accepting a Lot
• Proof Load vs Yield and Tensile Strength in Bolt Selection: What the Evidence Settles