A hexagonal all-metal locking nut can feel resistant during installation yet still provide insufficient locking resistance after removal and reuse. ISO 2320 evaluates this behavior through a controlled prevailing-torque test: the nut is installed and removed under specified conditions, and defined torque measurements are compared with the applicable limits. For inspection decisions, the essential distinction is that prevailing torque measures resistance from the locking feature—not the torque used to clamp a joint.

When an incoming batch feels different from an approved batch, or a maintenance procedure allows repeated assembly, hand feel alone cannot establish conformity. Excessive installation resistance can create assembly problems, while insufficient removal resistance can indicate inadequate prevailing-torque performance. The standard addresses both concerns, but a passing result does not prove that a joint will resist loosening under every vibration, temperature, or loading condition.

What the test measures—and what it does not

All-metal prevailing-torque nuts generate resistance through controlled interference in their metallic locking feature. Depending on the design, this may involve locally deformed threads or another metallic interference arrangement. Unlike an ordinary free-running nut, the locking nut therefore requires torque to rotate even before its bearing face contacts the joint.

Prevailing torque is measured while the locking feature is engaged but the nut is not producing clamping force against the test assembly. Once the nut seats and tightening develops bolt tension, the measured tightening torque includes other contributions, particularly thread friction and bearing-face friction. These values are not interchangeable.

A report that lists only “locking torque” without identifying direction, cycle, measurement interval, and loading condition is therefore ambiguous. Before accepting the result, establish which quantity the laboratory actually measured.

Establish the applicable requirements before testing

ISO 2320 covers functional properties of prevailing-torque steel nuts within its stated scope. A hexagonal shape or an all-metal construction does not, by itself, establish that every requirement applies to a particular product. Confirm the thread designation, pitch, property class, material, product standard, and specified surface condition against the applicable edition.

The acceptance limits are not one universal torque range. They depend on the relevant designation and table. A limit taken from another diameter, pitch, property class, or nut category can produce a false acceptance or rejection. The laboratory instruction should identify the exact standard edition and applicable requirements rather than referring loosely to “ISO locking-nut testing.”

Separate the roles of the documents. The product specification establishes what has been ordered; ISO 2320 provides applicable functional requirements and test methods; the inspection or purchasing agreement establishes matters such as lot sampling and required reporting. A successful test on selected specimens is not automatically a certification of every nut in a shipment.

Construction type also matters when comparing alternatives. A product identified as DIN985 Nylon Insert Hex Locking Nuts uses a non-metallic insert rather than an all-metal locking feature. It should not be substituted into an all-metal inspection specification merely because both products are hexagonal lock nuts. The relevant product requirements and intended service conditions need separate confirmation.

Prepare the test assembly so the result is meaningful

A prevailing-torque result describes the interaction between the nut and its mating thread under the test conditions. It is not independent of the bolt, coating, lubrication, or equipment. An uncontrolled setup can make a conforming nut appear defective—or conceal a problem.

• Use the specified mating test thread. Confirm the required dimensions, tolerance, mechanical properties, and surface characteristics. A convenient production bolt is not necessarily a compliant test bolt.
• Preserve the required surface condition. Added oil, solvent cleaning, coating damage, or an unapproved lubricant can materially change torque. Record any conditioning required by the applicable procedure.
• Control rotation and alignment. Use equipment capable of the specified rotational conditions without introducing side loading or substantial torque overshoot.
• Verify measurement capability. The torque instrument must have suitable range, accuracy, and calibration status. Where the procedure requires a defined axial test load, the load-measurement system also needs appropriate verification.
• Provide adequate engagement and travel. The locking feature must engage fully, while the designated prevailing-torque interval must remain clear of bearing-face contact.

Inspect the test threads before starting. Burrs, damaged lead threads, contamination, or wear can introduce resistance unrelated to the intended locking feature. Follow the standard’s requirements for test-bolt use and replacement rather than repeatedly using one bolt without control.

For an incoming inspection dispute, retaining unused samples is useful. Once a nut has been cycled, its locking interface has changed; that specimen can no longer represent an untouched first-installation condition.

Following the installation and removal sequence

The prevailing-torque assessment uses a prescribed sequence extending through five installation/removal cycles. The characteristic acceptance checks include a maximum first-installation prevailing-on torque and minimum prevailing-off torque values at the first and fifth removals. The detailed loading sequence, travel, speed, and measuring intervals must come from the applicable edition, not from a simplified shop-floor summary.

First installation: measure resistance before seating

Start the nut correctly on the test bolt and advance it until the locking feature is engaged as required. Measure the first prevailing-on torque over the prescribed interval, while the bearing face remains unloaded. Compare the required measured value with the applicable maximum.

This upper limit matters because stronger resistance is not automatically better. Excessive interference can complicate assembly, increase tool demand, and affect the relationship between applied tightening torque and achieved preload. A nut that exceeds the installation limit does not become acceptable merely because its removal resistance is high.

Apply the specified loading stage, then measure removal

Continue the first-cycle procedure through the prescribed seating and axial-loading stage. This exposes the nut to the required test condition before removal evaluation. Do not replace the specified load with an arbitrary tightening-torque setting: friction differences can produce different axial loads at the same applied torque.

During removal, release the clamping load and measure prevailing-off torque in the interval defined by the standard. The initial torque needed to break a tightened joint loose is not the required reading. Capturing that peak can make a weak locking feature appear satisfactory because the instrument is measuring loaded-joint effects instead.

Complete the remaining cycles without changing the method

Continue through the prescribed repeated installation and removal sequence, including the specified treatment of the intermediate cycles. Maintain the required test conditions and determine the fifth-removal value as directed. Compare it with its own minimum requirement; do not assume that the first-removal minimum also applies to the fifth.

The repeated-cycle assessment checks retained prevailing-torque performance after controlled use. It does not establish unlimited reusability. Nor does it mean that maintenance personnel may reinstall a nut five times in any application: service reuse remains subject to the equipment specification, joint requirements, and condition of the fastener.

Turning readings into an acceptance decision

Each required limit must be satisfied independently. A low installation torque cannot compensate for insufficient removal torque, and a strong first-removal result cannot compensate for failure at the fifth removal. Averaging these different measurements into one “locking performance” value removes the distinction the test is designed to evaluate.

A usable inspection record should make the decision traceable without requiring someone to reconstruct the test from memory. Include:

• Nut identification, lot reference, thread designation, property class, and specified finish.
• The standard edition, applicable acceptance limits, and specimen identification.
• Test-bolt and fixture information sufficient to verify conformity with the method.
• Equipment identification, relevant settings, and required axial-load information.
• First prevailing-on torque and the required first- and fifth-removal results, clearly labeled.
• Deviations, abnormal behavior, and the acceptance decision under the agreed sampling plan.

Results close to a limit deserve particular care. Apply the agreed conformity decision rule, including treatment of measurement uncertainty where relevant. Do not resolve a borderline result by rounding favorably or repeating the test until a passing value appears. Any retest needs an authorized basis and a clear distinction between fresh specimens and previously cycled ones.

When results disagree with assembly experience

High first-installation torque calls for checking test-thread dimensions, coating condition, lubrication, alignment, and the nut’s locking-feature geometry. Confirm that the measurement did not extend into seating. A sudden spike accompanied by metal pickup may indicate thread damage or galling rather than useful prevailing resistance.

A satisfactory first removal followed by low fifth-removal torque suggests loss of interference during cycling, but the test setup still needs examination. Verify the cycle sequence, measuring intervals, and condition of the mating thread before assigning the cause to the nut manufacturing process. Keep the individual specimen results rather than reporting only a batch average.

Large variation between specimens may reflect product variation, inconsistent finish, contaminated threads, or measurement instability. Torque-versus-angle records, where available, can help distinguish sustained locking resistance from isolated spikes. They supplement the required measurements; they do not replace the standard’s acceptance method.

A discrepancy between laboratory results and production-tool readings is not necessarily contradictory. Production tools may report total tightening torque, use different speeds, or operate with different mating bolts and lubrication. Compare equivalent quantities before concluding that either measurement is wrong.

Where ISO 2320 evidence stops

A passing test supports a defined claim: the tested specimens met the applicable prevailing-torque requirements under the prescribed conditions. It does not independently demonstrate resistance to transverse vibration, preservation of joint preload during thermal cycling, corrosion durability, or acceptable behavior at elevated temperature.

Those distinctions matter in safety-related joints. Prevailing resistance may remain after a joint has lost much of its preload; conversely, a compliant nut cannot correct inadequate bolt tension, joint settlement, or unsuitable joint geometry. Locking-function evidence and joint-integrity evidence answer different questions.

When approving a nut for service, keep the ISO 2320 report alongside the application’s tightening specification and any required joint-level validation. If the actual concern is vibration loosening, temperature exposure, or repeated maintenance reuse, specify evidence addressing that condition directly. Use the standard to make a controlled, traceable acceptance decision—not to extend a laboratory torque result into an unsupported service guarantee.