Torque coefficient testing evaluates the relationship between the applied tightening torque and the resulting axial preload in high-strength large hexagon head bolt assemblies used in steel structural connections. Performed in accordance with standards such as GB/T 1231, GB/T 16823.3, and ISO 16047, this test verifies that torque-controlled tightening on site will reliably achieve the design clamp force. This article provides a comprehensive technical reference covering testing principles, sample requirements, test procedures, calculation methods, and acceptance criteria for bolt assemblies intended for friction-type steel connections.
Technical Overview
High-strength large hexagon head bolts, together with their matching nuts and washers, form complete connection sets widely used in friction-type steel structure connections. These assemblies are typically supplied in property classes 8.8S and 10.9S, with nominal thread sizes ranging from M12 to M30, as specified in GB/T 1231. The load-carrying capacity of a friction-type connection depends directly on the axial preload, or clamp force, generated during tightening.
Because preload cannot be measured directly during routine installation, it is controlled indirectly through the applied tightening torque. The torque coefficient K quantifies the relationship between the applied torque T, the achieved axial preload F, and the nominal thread diameter d, expressed as T = K × F × d. The coefficient K is not a material constant; it is influenced by thread geometry and tolerance, surface coating, lubrication, thread friction, and the condition of the bearing surfaces. For this reason, each production lot of bolt assemblies must undergo torque coefficient testing before delivery and installation.
Testing is generally carried out by accredited third-party laboratories, manufacturer laboratories, or construction quality inspection institutions using calibrated torque-preload testing equipment. The governing methods are defined in GB/T 1231 and GB/T 16823.3 for the Chinese market, with ISO 16047 providing internationally recognized procedures for torque and clamp force testing of threaded fasteners. The test results establish the tightening torque values used on site and support acceptance of bolt assemblies under construction quality standards such as GB 50205.
Testing Samples
Bolt Connection Sets
The test sample is a complete connection set consisting of one large hexagon head bolt, one high-strength nut, and two high-strength washers. All components must originate from the same production lot, with the same property class, nominal size, thread pitch, surface treatment, and lubrication condition as the assemblies intended for actual installation. Mixing components from different lots invalidates the test because the torque coefficient is a property of the assembly as delivered, not of the bolt alone.
Sample Quantity and Conditioning
In accordance with GB/T 1231, a minimum of eight complete connection sets shall be sampled from each production lot for torque coefficient testing. Samples shall be free of visible corrosion, thread damage, and contamination, and shall not be additionally lubricated unless the lubricant is part of the as-delivered condition. Prior to testing, samples shall be conditioned in the laboratory at a temperature of 10°C to 35°C until temperature equilibrium is reached. Tested assemblies shall never be returned to stock or reused in structures.
Testing Items
The torque coefficient test program covers the following measurement and evaluation items, which together characterize both the central tendency and the consistency of the torque-preload relationship of the lot.
- Torque coefficient K of each individual bolt connection set
- Average torque coefficient of the tested sample group
- Standard deviation of the torque coefficient across the tested sets
- Axial preload achieved at the specified tightening torque
- Verification of thread engagement, coating condition, and assembly integrity factors that influence K
Testing Methods
Test Apparatus
Testing is performed on a dedicated torque coefficient testing machine or a calibrated torque wrench system integrated with an axial force measurement device. The nut-turning torque and the resulting bolt preload are recorded simultaneously during tightening. Torque measurement error should generally not exceed ±2% of the measured torque, and axial preload measurement error should not exceed ±1%, with all sensors calibrated and traceable to national metrology standards. ISO 16047 additionally permits the use of instrumented bolts with strain gauges for preload measurement where friction coefficient characterization is required.
Test Procedure
- Assemble each connection set with one washer installed under the bolt head and one washer under the nut, with the marked or chamfered faces oriented as specified in GB/T 1231.
- Mount the assembly in the testing machine so that the axial force sensor measures the bolt preload along the bolt axis without eccentric loading.
- Apply the tightening torque by rotating the nut at a slow, steady, and continuous rate to avoid dynamic effects on the measurements.
- Record the applied torque and the axial preload simultaneously, taking readings when the preload reaches the value specified for the nominal size and property class in the applicable standard.
- Calculate the torque coefficient K for each set using the recorded torque, preload, and nominal diameter.
- Repeat the procedure for all eight connection sets from the lot.
- Calculate the arithmetic mean and the standard deviation of the eight individual torque coefficient values.
Calculation and Acceptance Criteria
The torque coefficient of each individual assembly is calculated as K = T / (F × d), where T is the applied torque in newton-millimeters, F is the measured axial preload in newtons, and d is the nominal thread diameter in millimeters. In accordance with GB/T 1231, the average torque coefficient of the tested sets shall fall within the range of 0.110 to 0.150, and the standard deviation of the torque coefficient shall not exceed 0.010. The test report shall state the individual K values, the average value, the standard deviation, the test temperature, and the equipment calibration status. Assemblies that fail the criteria are rejected for torque-controlled installation and may only be used with an alternative verified tightening method where permitted.
Test Environment and Precautions
The laboratory temperature shall be maintained between 10°C and 35°C, since temperature affects lubricant behavior and friction conditions. Tightening shall be performed in a single continuous operation, and no re-tightening of an already tightened assembly is permitted, as preload relaxation and surface wear would distort the measured coefficient. Operators shall avoid contaminating threads and bearing surfaces with hand oils, dust, or cleaning agents during sample handling.
Applications
Torque coefficient testing is mandatory wherever high-strength large hexagon head bolt assemblies are installed using the torque control method in load-bearing steel structures. Typical fields of application include the following.
- Steel frameworks of high-rise and super high-rise buildings
- Steel bridge girders, box girder segments, and orthotropic bridge decks
- Wind turbine towers, transmission towers, and telecommunication masts
- Industrial plants, power plants, and petrochemical facilities
- Stadiums, exhibition halls, and other large-span spatial steel structures
- Railway and rail transit steel structures
FAQ
What is the torque coefficient and why is it important?
The torque coefficient K is an empirical factor linking the applied tightening torque to the axial preload generated in the bolt, defined by the relationship T = K × F × d. It is important because field installation relies on torque control: the design preload can only be achieved reliably if K is known, stable, and within specification for the delivered lot of assemblies.
What are the acceptance criteria for the torque coefficient?
Under GB/T 1231, the average torque coefficient of the tested connection sets shall be between 0.110 and 0.150, and the standard deviation shall not exceed 0.010. Both conditions must be satisfied simultaneously; a lot with a compliant average but excessive scatter is still rejected.
Why is the standard deviation of K so tightly controlled?
A small standard deviation indicates that the torque-preload relationship is consistent across the lot. If the scatter is large, identical tightening torques would produce significantly different preloads in different bolts, leading to non-uniform clamp forces and unpredictable slip resistance in the connection.
How often should torque coefficient testing be performed?
Testing is required for every production lot at the factory. On construction sites, retesting is required before installation if bolt assemblies have exceeded their specified storage period, typically more than six months, or if there is any doubt about their condition, as required by GB 50205.
What factors can cause the torque coefficient to fall outside specification?
Common causes include corrosion or moisture damage to the lubricating coating, thread damage or contamination, missing or degraded surface lubrication, mixing of nuts, bolts, and washers from different lots, and exposure to excessive heat or solvents during storage and transport. Proper sealed packaging, dry storage, and lot integrity control are essential to preserve the certified torque coefficient.
