Coated Bolts: Ask for Torque–Tension Test Conditions
Share
A torque–tension report is not a property of the coating name alone. The measured clamp force also depends on the bolt and nut, lubricant or topcoat, thread fit, bearing surface, washer, fixture and test procedure. Two reports can both say “zinc-flake coated” while describing materially different friction interfaces.
This guide helps a buyer compare laboratory evidence. It does not set an installation torque, select a corrosion coating or approve a bolted joint. Keep the component’s general purpose with the existing track bolt and nut overview. The engineering authority for the actual joint must decide what test conditions and validation are suitable.
Define the joint and fastener condition
Identify every test specimen before reading the plotted curve or reported coefficient. Record manufacturer, part number, production lot, fastener type, nominal thread size and pitch, tolerance class, property class, material and geometry. Include thread length, under-head form and bearing diameter because these can affect the fixture and contact interface.
Describe the coating as a system rather than a colour. Record the base coating, conversion layer or sealant where applicable, topcoat, integral lubricant and any additional lubricant applied before the test. Capture product references, specified coating thickness, application process, batch and the time or conditioning between coating and testing. “As received” is useful only when the receiving state and lot are traceable.
Record the mating internal thread with the same care. A coated bolt tested with a plain steel test nut does not reproduce a production joint that uses a coated nut or a tapped component with another finish. Identify nut property class, thread tolerance, material, coating, lubricant and whether it was new or reused for each run.
The bearing interface needs its own fields. State whether the rotating surface was the bolt head or nut, whether a washer was used, and the washer or bearing plate’s material, hardness, finish, roughness and geometry when required by the method. A change in effective bearing diameter changes the frictional torque distribution even if the thread pair is unchanged.
Also record temperature, humidity or other conditioning specified by the test, plus cleaning and handling. Solvent cleaning can remove a lubricant; fingerprints, corrosion or storage can alter a surface. If the procedure calls for tests in delivery condition, document how specimens were protected between sampling and the laboratory.
The public scope for ISO 16047:2005 says it specifies conditions for torque/clamp-force testing and describes its basic range of applicable threaded fasteners and exclusions. ISO also lists a 2012 amendment. A report should name the edition and amendment basis actually used rather than relying on the phrase “tested to ISO.”
Read what the test fixture measures
Draw the torque path through the fixture. Applied torque is used in stretching the bolt and overcoming friction in the threads and at the rotating bearing surface. A torque transducer and load cell can record total tightening torque and clamp force. Additional instrumentation or a defined calculation may be required to separate thread torque from bearing torque.
List the channels reported by the laboratory: total torque, clamp force, angle or rotation, thread torque, bearing torque, time and any other signal. For each, record sensor range, calibration status, sampling or filtering information and measurement uncertainty when supplied. A graph without axis units, sample identity or channel definition is not a complete result.
Identify whether the test controls torque, clamp force, angle or a sequence of conditions. Record tightening speed, target, stop rule and any preload or seating step. Dynamic friction behaviour can vary with speed and surface state. A result obtained after repeated tightening may not represent a new fastener’s first installation.
Fixture geometry must correspond to the calculation. Record grip length, engaged thread length, test nut or tapped insert, bearing plate, hole diameter, alignment and effective bearing dimensions used by the method. If a washer rotates unexpectedly or the non-rotating member moves, the assumed interface can change during the run.
Do not infer outputs that were not measured or validly derived. Total torque and clamp force can support a torque–clamp-force relationship or a reported torque coefficient under stated definitions. They do not automatically reveal separate thread and bearing coefficients. Ask the laboratory for the equation, input dimensions and method clause behind any derived value.
Bolt Science’s test-service overview distinguishes torque–tension testing used to determine a relationship and nut factor from tests that separate thread and head friction torques to determine thread and bearing coefficients. This is a useful explanation of different outputs, but it does not define the acceptance criteria for an order.
Check the standard’s applicability to the specimen. ISO 16047’s public scope lists basic thread-size and fastener categories and excludes certain fasteners, including types with additional self-locking features. When the production joint lies outside the named scope or uses a modified procedure, the report should describe the deviation and the engineering basis for the method.
Compare repetitions and reporting conventions
Read the individual results before the average. Record the number of specimens, number of tightening cycles per specimen and how specimens were sampled from the coating lot. A large number of cycles on one bolt is not the same population as one cycle on many production fasteners.
Capture every individual torque–clamp-force curve or result where available. Then record the mean, range, standard deviation or other stated scatter measure. If the laboratory excludes an outlier, require the pre-defined rule or technical explanation and preserve both the original and revised analysis. Removing the worst run because it is inconvenient can hide a real coating or assembly problem.
Keep the reported quantity’s name and definition. Nut factor, torque coefficient, total coefficient of friction, thread coefficient and bearing coefficient are not interchangeable labels. Even when two outputs are dimensionless, they can use different equations, diameters and assumptions. A procurement comparison should never align them merely because the numbers look similar.
Record the point or range over which each coefficient was calculated. A single value may describe a target clamp force, a fitted part of a curve or an average across runs. Check whether seating behaviour is excluded and whether the relationship remains approximately linear through the range of interest.
Bolt Science discusses non-linear torque–preload behaviour and why a simple constant-factor model can fail under some circumstances. A curved or changing result is not repaired by reporting one convenient factor. Preserve the curve and ask which mechanism, such as changing friction or interface behaviour, was investigated.
Compare reports only after normalising the descriptions, not the results. Put fastener lot, coating, lubricant, nut, bearing plate, washer, speed, cycle number, target range and calculation convention side by side. If one field is unknown, the comparison remains conditional. A later report with lower scatter is not automatically better if it used a different surface or population.
Review failures and abnormal observations. Galling, coating pick-up, stick-slip, thread damage, bearing embedment, yielding or fixture slippage can invalidate a run or reveal a relevant application risk. The laboratory should identify the event and its disposition rather than replacing the result silently.
Keep laboratory evidence within its application
Define the configuration to which the report can reasonably apply. Link the tested fasteners to production lots and verify that coating source, topcoat, lubricant, nut and bearing interface match the controlled purchase and assembly specifications. A supplier certificate naming only the coating family does not establish that match.
Compare the laboratory fixture with the production joint. Consider mating thread material and finish, engaged length, washer, bearing diameter, joint stiffness, surface condition and whether tightening occurs under similar speed and tool behaviour. Differences do not always make the report useless, but they must be assessed by the responsible engineer.
Do not transfer a laboratory torque directly to field work. Installation also depends on target clamp force, joint design, allowable bolt and thread stresses, tool accuracy, tightening sequence, access, reuse policy and service conditions. The test report can be an input to that engineering decision; it is not the completed instruction.
Separate corrosion performance from friction performance. A coating may meet one requirement and need different evidence for the other. Environmental exposure, coating damage during assembly and compatibility with the joint materials require their own review. This article does not recommend a coating system.
Set revalidation triggers. A new coating supplier, bath, topcoat, lubricant, application process, curing condition, fastener source, nut, washer or bearing finish may change the torque–clamp-force relationship. Define which changes require document review, confirmation testing or full joint validation.
Preserve the report version, raw results, calculation sheet, specimen traceability, deviations and authorised decision. If the purchasing specification sets an acceptance range, record the exact clause and verify that the laboratory used the same definition and conditions. Do not invent a generic “good friction range.”
| Case | Fastener and interfaces | Fixture and method | Results | Application decision |
|---|---|---|---|---|
| Normal | Bolt, lot, coating, lubricant, nut and bearing surface are fully identified and match the requested configuration | Method, amendment, geometry, speed, sensors and cycle sequence are documented | Individual curves, sample count, scatter and coefficient definitions are available | Ready for authorised joint-specific review; no installation torque is issued by this record |
| Missing | Coating trade name is shown, but topcoat, lubricant, nut or washer finish is absent | Report cites ISO 16047 without edition, fixture dimensions or cycle sequence | Only an average coefficient appears; individual runs and calculation definition are missing | Keep applicability open and request the condition and raw-result package |
| Conflict | Report sample lot or nut finish differs from the delivered production configuration | One report uses total torque while another separates thread and bearing torque | Curves are non-linear or an excluded result has no documented rule | Do not merge the coefficients; reconcile the method and retest the required configuration if authorised |
The review is complete when a reader can reconstruct what was tightened, against which surfaces, by which procedure, how many times and how each reported number was defined. Keep unknown conditions visible. A result becomes useful for a production joint only after the responsible authority confirms that the tested interfaces and the real assembly are sufficiently aligned.