Why Preload Is the Working Parameter of a Fastening System
The clamping force of a rail fastening system, usually expressed as the toe load per clip, is what actually holds the rail in gauge, resists rail creep and keeps the fastening system stiff. It is produced by preload: the tension created in a bolt, screw spike or clip leg when it is tightened. The design target is set by the fastening system standard, for example the toe load and vertical stiffness requirements of EN 13481-1 for systems on concrete sleepers, and the clamping force itself is measured on a test rig. In service, the maintenance engineer cannot see preload directly, so it is managed through torque, deflection or gap checks defined by the system drawing. Understanding how preload is generated, where it is lost and how it relaxes over time is therefore the basis of both correct installation and correct maintenance scheduling.
The Transmission Path and Where Preload Is Lost
Preload travels from the tightening tool to the rail through a chain of interfaces, and each interface consumes a share of the input force through friction and local deformation. The table below summarises the typical loss ranges reported in fastening practice for a bolted elastic fastening system; the values are engineering estimates from track component testing and should be confirmed for the specific system and lubrication state.
| Interface in the load path | Typical preload loss | Main influence |
|---|---|---|
| Bolt thread engagement | 15-20 percent | Thread friction coefficient, lubrication, thread fit |
| Nut face on pressure plate or washer | 10-15 percent | Face flatness, surface roughness, contact area |
| Pressure plate to clip contact | 8-12 percent | Clip arc geometry, contact stress distribution |
| Clip to rail foot | 5-10 percent dynamic loss | Wheel-rail vibration, dynamic load history |
After the four interfaces are summed, the effective clamping force reaching the rail is typically 50 to 60 percent of the value implied by the nominal tightening torque. This is not a defect; it is the normal conversion efficiency of a threaded joint, and the tightening specification already accounts for it. What matters in practice is consistency: the same torque must produce the same clamping force every time, which is only true if the friction state is controlled.
Torque, Friction and Clamping Force
The relationship between tightening torque T, clamping force F and thread pitch diameter d is the classic formula T = K x F x d, where K is the torque coefficient that bundles thread friction, face friction and thread helix effects. K typically lies between 0.15 and 0.25 depending on surface treatment and lubrication: a clean dry steel thread is near the top of the range, a lubricated or zinc-flake coated thread is near the bottom. The coefficient is not a constant; it changes with plating, lubrication, reuse of the bolt and environmental conditions, which is why ISO 16047 prescribes a standard test for measuring the torque-clamping force relationship of a fastener assembly, and why critical joints are validated on a rig before the torque specification is frozen. The tightening method also sets the scatter: torque control holds torque but not clamping force, with a scatter of about plus or minus 25 percent on the resulting preload; angle-controlled tightening reduces scatter to about 15 percent; yield-controlled tightening, used on the highest-integrity joints, holds preload to about 8 percent scatter. For track fastenings, torque or deflection control is standard because it is fast and tool-verifiable, and the friction state is managed by specifying the coating and lubrication of the components.
Controlling Friction at Each Interface
Because friction decides how much torque becomes clamping force, the practical controls are applied at the interfaces. On the thread, a defined lubricant or coating stabilises the thread friction coefficient, typically in the range 0.10 to 0.15 for a coated and lubricated thread, and prevents the galling that appears on unlubricated plated threads. On the nut face and the clip contact, flatness and roughness are controlled by machining; a flatness deviation of about 0.05 mm or less and a seating area of 95 percent or more avoid local high spots that consume preload unevenly. On the clip-to-rail contact, the geometry of the clip toe and the rail foot is fixed by the system drawing, and the interface carries dynamic load, so wear and fretting here appear first in service. Elastic washers and conical spring washers are used where the joint must absorb settlement or thermal movement, because they convert part of the elastic deflection into retained preload instead of losing it. The general rule is that each interface is specified, tested and controlled exactly like a material property, because the scatter of K, not the nominal value, determines whether the installed clamping force meets the standard.
Stress Matching Across Bolts, Clips and Pads
The preload must be compatible with the strength of every component in the load path. Bolts and screw spikes are selected by property class, typically 8.8 or 10.9 per ISO 898-1 for track bolts, so that the preload corresponding to the specified torque stays below the proof stress with a design margin; the common design practice is to keep working stress at 60 to 70 percent of yield for static joints. Spring clips are made from hardened spring steel, for example 60Si2MnA to GB/T 1222, and their working stress is limited by the fatigue requirement rather than by static strength, so the permissible stress is typically 50 to 60 percent of the elastic limit and the clip deflection is set by the toe-load target. Pads and insulating components are the soft link in the chain: their compressive stress is limited by the pad material, and the pad area is sized so the preload plus traffic load stays below the pad stress limit, otherwise the pad creeps and preload is lost. The design task is to match the stiffness gradient through the system, stiffer at the bolt and clip, softer at the pad, so that the preload is shared without overloading any single component.
Heavy-Haul and High-Speed Differences
Heavy-haul lines specify a higher initial preload than conventional lines for three reasons. The wheel loads are higher, so the dynamic share of the clamping force demand is larger. The friction coefficients of the interfaces tend to rise with the higher interface pressures, which increases the transmission loss, typically 10 to 15 percentage points more than on a conventional line. And the stronger vibration accelerates relaxation, so the initial preload must include a relaxation reserve that keeps the effective clamping force above the minimum for the life of the system. On high-speed lines the same logic runs in the opposite direction: the vertical forces are moderate, but the system must hold a precise track geometry at speed, so the priority is stiffness consistency and uniform preload from clip to clip rather than maximum preload. In both cases the maintenance consequence is the same: measure clamping force or torque on a sampling basis, watch the attenuation trend and re-tighten before the effective preload falls below the minimum required by the system standard.
Frequently Asked Questions
Q1: Why does only about half of the torque become clamping force?
Because the applied torque first overcomes thread friction, then nut face friction, then the friction and deformation of the clip interfaces. These losses are the normal conversion inefficiency of the joint, typically leaving 50 to 60 percent of the nominal force as effective clamping force.
Q2: What does the torque coefficient K mean?
K in the formula T = K x F x d bundles all friction effects of the joint. A K of 0.2 means 20 percent of the torque is consumed by friction. Lower K gives more clamping force for the same torque and less scatter.
Q3: Why do the same torque values give different clamping forces?
Because friction state varies with plating, lubrication, thread condition and reuse. This is why ISO 16047 tests exist and why critical joints are validated on a rig before the torque value is written into the installation spec.
Q4: When should angle-controlled tightening be used?
When scatter must be minimised, for example on safety-critical joints where the clamping force band is narrow. Angle control removes the influence of friction, but it requires an elastic joint where the turn-to-preload relationship is stable.
Q5: How does pad creep affect preload?
When the pad takes permanent set, the clip working point moves and the toe load drops. This is the dominant preload loss mechanism on soft pad systems and the main reason the first re-tightening is scheduled a few months after line opening.
Q6: Why do heavy-haul fastenings need more initial preload?
Because the interface losses are higher, the dynamic demand is larger and vibration-driven relaxation is faster. The extra initial preload is a reserve that keeps the effective clamping force above the minimum through the service life.