Fastener preload is the tension deliberately created in a bolt or stud to clamp a joint. Tightening methods do not measure preload equally. NASA-STD-5020B orders four common methods by increasing labor, cost, and accuracy: torque control, turn angle, fastener length change, and strain gauges. Its ordering is qualitative; it does not define one universal accuracy percentage for each method.

Bolt preload method accuracy comparison

What each method controls

Torque control is economical and familiar, but applied torque is affected by friction at the threads and bearing surfaces. Lubrication location can change the torque-to-preload relationship, so the same torque does not guarantee the same bolt tension. A calibrated wrench checks the applied torque; it does not directly measure clamp load.

Turn-of-nut and turn-angle methods control rotation after the assembly reaches a defined seated condition. Bolt-stretch methods estimate tension from the fastener’s elongation. Strain gauges measure strain on the bolt body and can give a more direct indication, but require suitable access, instrumentation, and calibration. NASA’s qualitative ranking places them in increasing accuracy order along with increasing labor and cost.

The word “accuracy” here refers to controlling the intended preload, not the accuracy of the wrench or sensor by itself. A torque tool can be calibrated and still produce variable tension if friction differs. Likewise, a stretch or strain measurement is only useful when the gauge length, material properties, temperature, installation sequence, and reading method are appropriate for the actual fastener.

Why method accuracy depends on the joint

NASA-STD-5020B requires the nominal preload relationship to be substantiated by testing the fastening-system hardware and the parameter controlled during installation. It also calls out configuration-specific behavior for torque, turn-of-nut, turn-angle, and bolt stretch. A test should reproduce the relevant fastener, thread form, washers, materials, lubricant placement, and assembly process. The standard discusses spaceflight hardware, so its requirements should not be applied as a substitute for another industry’s governing code.

For an illustrative joint requiring a specified initial tension, first choose a method compatible with access, production rate, inspection, and the consequence of preload variation. Establish the seated condition, test representative assemblies, measure the resulting tension, and use the observed range in the design. If direct preload measurement is not practical, account for the method’s uncertainty and verify the installed joint using the governing procedure. A single torque reading is not proof of a known preload.

The standard’s test guidance calls for a representative sample and requires nominal preload to be supported by test data. In addition to the initial result, engineering checks may need to include relaxation, creep, temperature effects, and the minimum and maximum preload that the joint can experience. Those changes matter because a joint can lose clamp load after assembly or see higher bolt tension as temperature changes. Select a method and acceptance limit based on the joint’s function and consequences, not merely installation convenience.

Trade-offs and limits

Torque control is usually easiest to apply at scale but is most sensitive to friction. Rotation-based methods need a repeatable snug or seated reference. Length measurement requires a measurable bolt and an accurate baseline. Strain gauges add sensor preparation and wiring but directly monitor elastic strain during installation. Hydraulic tensioners are another specialized approach for accessible studs; the NASA technical report describes accounting for relaxation when the tensioner is removed. No method eliminates the need to consider embedment, creep, thermal effects, or load sharing.

Direct hydraulic tensioning can pull an exposed stud to a target elongation while the nut is seated; once the tensioner is released, some relaxation must be accounted for. This method is useful only when the hardware and access permit it, and it is not the same operation as tightening a bolt through its head. Application-specific tooling and procedure qualification remain essential.

Do not use the ordinal ranking as an accuracy guarantee or a universal specification. Test the actual fastener and joint, and use the governing engineering standard or qualified fastening procedure for safety-critical work.

Sources

The method list, relative order, and preload test discussion come from NASA-STD-5020B. The hydraulic direct-tensioning description is from the NASA technical report Direct Tensioning.