Bolted-joint decision support

Torque vs Tension: Comparing Bolt Preload Methods

A method-level comparison of how bolt preload is produced and how conformity is evidenced, written for the point in a project before a specific tightening or measurement product has been selected.

Topic
Bolted joints
Reading time
8 minutes
Published
Revised
Editorial author
WindInspectTech · Source policy
Technical reviewer
Not attributed — no named technical reviewer is claimed
Source basis
RCSC Specification for Structural Joints Using High-Strength Bolts (2020), ISO 16047:2005, VDI 2230 Part 1 and the NASA Fastener Design Manual.

The distinction

The joint needs tension. Torque is one way to get there.

Preload — the axial force clamping the joint — is what the design calculation requires. Torque is an input used to produce it, not a synonym for it. Most of the torque applied to a fastener is consumed overcoming friction at the thread flanks and under the turned element; only the remainder converts to axial force. That is why two bolts tightened to the same torque value can end up at different tensions.

Input, length change and force are different quantities. A calibrated relationship is needed to connect an observed change to preload.
Torque: rotational input

Torque · rotational input

The tightening moment acts through the joint. Friction and the assembly procedure affect the resulting preload.

Elongation: change in length

Elongation · length change

A loaded bolt changes length. Ultrasonic time of flight is an acoustic observation and must be distinguished from a direct mechanical length reading.

Preload: axial forceCalibration

Preload · calibrated force estimate

Interpret the measurement with the applicable calibration, material, geometry, reference state, temperature and uncertainty.

Conceptual diagrams, not to scale. There is no universal torque → elongation → preload conversion. A recorded value is not continuous in-service monitoring or proof of defect-free material.

NASA’s official Fastener Design Manual identifies friction at threads and bearing surfaces, coatings and lubricants among the variables that complicate torque-to-preload selection. ISO 16047:2005 exists because that relationship has to be established by test rather than assumed: it specifies the conditions for torque/clamp-force testing of threaded fasteners.

VDI 2230 Part 1, Systematic calculation of highly stressed bolted joints — Joints with one cylindrical bolt, handles the remaining spread explicitly. It carries the achieved preload from a minimum assembly value up through a tightening factor that reflects the imprecision of the assembly procedure and friction. The design does not assume the tightening method is exact; it budgets for how inexact it is.

What this changes in a specification

A requirement written as a torque value specifies an input. A requirement written as a preload value with an acceptable range specifies an outcome, and then has to say how that outcome will be evidenced. The two are not interchangeable, and only the second can be checked on the installed fastener.

Method comparison

Each method controls a different variable.

The named structural methods come from Section 8.2 of the RCSC Specification for Structural Joints Using High-Strength Bolts (June 11, 2020). Hydraulic tensioning and ultrasonic measurement are listed alongside them because they appear in wind, pressure and machinery work, not because RCSC names them. The comparison is conditional: it describes what each approach controls, not a ranking that holds for every joint.

What each preload approach controls and what remains inferred
ApproachDirectly controlledPreload is…Principal sensitivity
Turn-of-nut (RCSC 8.2.1)Relative rotation between the turned and unturned element, from a snug-tight startInferred from bolt elongation over a specified rotationThe snug-tight starting condition, grip length and joint compaction
Calibrated wrench (RCSC 8.2.2)Applied torque, with the 2020 edition restricting the method to rotation of the nutInferred through a torque–clamp-force relationship established by daily testingFriction, lubrication condition, fastener lot and tool calibration drift
Twist-off tension control bolt (RCSC 8.2.3)Shearing of a spline end by the installation wrenchInferred from the assembly’s designed shear pointAssembly condition and lot, since the shear point is built into the matched assembly
Direct tension indicator (RCSC 8.2.4)Compression of protrusions on an indicating washerIndicated by a mechanical response at a designed loadWasher orientation, seating and correct gap interpretation
Combined method (RCSC 8.2.5)An initial torque followed by a designated rotationInferred from both inputs, added in the 2020 edition with its own inspection provisionsWhether the initial torque achieves the required initial tension, verified in pre-installation testing
Hydraulic tensioningHydraulic pressure acting over a known tool areaApplied as tension, then transferred to the fastener as the nut is seatedLoad transfer at seating, tool fit, pressure calibration and sequence relaxation
Ultrasonic axial-force measurementTime of flight through the bolt, against a referenceCalculated for the individual fastener from calibrated bolt inputsBaseline capture, material, temperature, geometry, probe and coupling

For the calibration controls, workflow and procurement questions specific to the last row, see the ultrasonic bolt tension measurement guide. This page stops at method selection.

Evidence

Most methods evidence the process, not the force.

This is the distinction that decides whether a preload measurement is needed, and it is stated in the specification rather than inferred from it. Under RCSC Section 9.2.1, inspection of a turn-of-nut installation requires the Inspector to observe the pre-installation verification testing, verify by routine observation that the snug-tight condition was reached, and verify that the crew rotated the turned element by the specified amount. The specification then states plainly that no further evidence of conformity is required.

That is a defensible and long-established basis for structural steel. It is also a statement about what the record contains: an observed process, applied to a lot, not an axial force value for each installed fastener.

Where the force does get measured

RCSC Section 7 requires pre-installation verification before bolting assemblies of verified lots go into the work — at the site of installation, on a sample of not fewer than three complete assemblies for each combination of diameter, length, grade and lot, and daily for the calibrated wrench method. The commentary is explicit about why: experience on many projects showed that non-conforming assemblies would have been identified had they been tested as an assembly in a bolt tension measurement device.

So the standard does put fasteners into a tension-measuring device — as a qualification of the lot and the method, before installation. It does not, through these provisions, produce a measured preload for a specific bolt already in the structure.

The question to answer before selecting a method

Does the requirement end at demonstrating that a qualified method was correctly applied, or does something later — a re-check after service, a dispute, a joint that cannot be re-accessed cheaply, or an asset owner’s retained record — need a force value for an identified fastener? The first is satisfied by process evidence. The second is not.

Application

Wind-turbine joints add re-checking to the problem.

RCSC governs structural steel building and bridge connections; a wind turbine’s tower flange, blade root and yaw or pitch bearing bolts are designed and maintained under their own engineering procedures. The methods above still describe the available physics, but the maintenance question changes the balance: these joints are re-entered over a service life, often at height, and an owner may need to know whether preload has changed rather than whether the original installation was performed correctly.

  1. 01State the requirement as an outcome

    Write the target preload and its acceptable range, then separately state how conformity will be evidenced at installation and at any later interval.

  2. 02Check what the access allows

    Tool clearance, the ability to reach both ends of the fastener, working position and the time window all rule methods in or out before accuracy does.

  3. 03Decide whether re-measurement is needed

    If a later comparison against the installed state is required, the method has to leave something behind that a later measurement can be compared to.

  4. 04Qualify the chosen method for the actual fastener

    Diameter, length, grade, coating, lubrication and lot all belong in the qualification, whichever method is selected.

Where the answer is that a force value is needed, it can be bought as a measurement engagement instead of an instrument — see ultrasonic bolt load verification as a service. Where the question is the arithmetic rather than the method, the bolt torque chart and preload calculator shows what the nut factor does to the same target.

Preload verification and defect detection are different questions and should not be merged into one result — the verification versus monitoring section sets out where that boundary sits, and the PAUT guide covers inspection for indications rather than force.

Residual or breakaway torque is not a preload reading.

A request for residual torque or breakaway torque needs a defined procedure: direction, the rotation being observed, elapsed time and the condition of the installed fastener. A torque result should not be relabelled as measured axial force.

RCSC Section 10 and its commentary explain why after-the-fact torque arbitration is sensitive to lubrication, bearing conditions, time and exposure. The specified process uses representative assemblies; it does not endorse a universal check torque. Rotation during a check can also require re-pretensioning and reinspection under that procedure.

RCSC addresses structural steel joints. For a wind-turbine joint, use the asset’s approved maintenance and acceptance procedure. If the question is retained axial force, define the reference state and calibrated measurement method with the bolt measurement service review. No generic residual-torque acceptance number is provided here.

Evidence record

Sources and limitations

Method names, section numbers and inspection provisions were read from the published specifications cited below. No vendor material was used as a source for this comparison.

  1. RCSC, Specification for Structural Joints Using High-Strength Bolts, June 11, 2020 — Sections 7.1, 8.2 and 9.2 for the named pretensioning methods, pre-installation verification and inspection provisions.
  2. ISO 16047:2005 — official scope for torque/clamp-force testing conditions on threaded fasteners.
  3. VDI 2230 Part 1 — systematic calculation of highly stressed bolted joints, including the tightening factor applied to assembly preload.
  4. NASA Fastener Design Manual — official engineering reference describing preload and torque-control variables.

Limitation: RCSC governs structural steel joints and does not set requirements for wind-turbine bolting, which is covered by the relevant design and maintenance procedures for the asset. No accuracy ranking between methods is stated here, because achievable scatter depends on the fastener, its condition, the tooling and the qualification actually performed. Nothing on this page substitutes for the governing specification or the Engineer of Record.

See the source and editorial policy for how supplied product manuals, external method sources and claim limitations are kept separate.

Decision support

Frequently asked questions

What is the difference between torque and tension in bolting?

Tension, or preload, is the axial force the joint actually needs. Torque is one input used to produce it. Because a large share of applied torque is consumed by friction at the threads and under the turned element, torque is related to preload through a model whose inputs — friction, coating, lubricant, surface condition and tool calibration — vary between assemblies.

Can you calculate bolt tension from applied torque?

A torque–clamp-force relationship can be established for a defined fastener condition under a standardised test. ISO 16047:2005 specifies the conditions for that testing. The result applies to the tested combination and condition, so it is a qualified relationship rather than a general conversion, and VDI 2230 Part 1 handles the remaining variability through a tightening factor that depends on the assembly procedure.

What are the recognised bolt pretensioning methods?

For structural joints, the RCSC Specification (2020) names five in Section 8.2: the turn-of-nut method, the calibrated wrench method, the twist-off tension control bolt method, the direct tension indicator method, and a combined method added in the 2020 edition. Hydraulic tensioning and ultrasonic axial-force measurement are used in other sectors under their own engineering procedures.

Does an inspected bolt have a measured preload?

Not necessarily. Under RCSC Section 9.2.1, inspection of a turn-of-nut installation is satisfied by observing the pre-installation verification test, the snug-tight condition and the specified rotation; the specification states that no further evidence of conformity is required. That evidences the process, not the axial force in each installed fastener. Measuring the force in a specific bolt is a separate activity.

Engineering review

Say what the joint has to evidence, not which tool to quote.

Tell us the joint, the fastener grade and size, the governing specification, the access available and whether preload has to be re-checked in service. We’ll identify which methods can evidence that requirement.

Technical enquiry

sales@WindInspectTech.comWhatsApp sales
At least 20 characters. Do not include passwords, credentials, export-controlled files or confidential design data.
Company and destination (optional)
Useful technical inputs (optional)

Component, operating condition, required output and preferred data interface.

Inquiry route
Use email insteadBy submitting, you confirm that you have read the Privacy Notice. Product values remain configuration-specific and require application review.