Why Thread Size Sets the Preload Ceiling
Preload is the tension the bolt carries after tightening, and it is what actually clamps the fishplate or the clip shoulder. The maximum usable preload is limited by the stress area of the threaded section, which grows roughly with the square of the diameter. An M16 bolt has a stress area of about 157 mm2; an M36 bolt has a stress area of about 817 mm2, roughly 5.2 times larger, so at the same utilisation of its proof strength the M36 can hold about five times the preload of the M16. In round numbers for rail fastening service, an M16 bolt in grade 8.8 is tightened to a preload in the region of 60 to 80 kN, while an M36 in grade 10.9 can be tightened toward 450 kN. The jump between adjacent sizes is substantial: stepping from M16 to M20 to M24 to M30 increases the available preload by 50 to 70 percent at each step, which is why heavy-haul fishplate joints move to M24 and above as soon as the loading demands it.
Torque Is Only a Means to Preload
In track work the installer cannot measure bolt tension directly, so torque is used as the practical proxy. The relationship between torque and preload is governed by the thread pitch, the thread friction and the head friction, summarised in the classic torque equation used by VDI 2230: a rule of thumb for steel bolts with normal lubrication is that only about 10 to 15 percent of the applied torque becomes preload, and the rest is consumed by friction. Because friction varies with coating, lubrication and surface condition, the same torque can produce different preloads on different bolts; this is why torque reference values are always quoted for a defined friction state and why the effect of the surface treatment must be accounted for.
For hot-dip galvanized bolts the thread friction coefficient is noticeably different from black-oiled bolts, so the tightening torque must be re-derived for the actual coating and lubrication state rather than copied from a general table. In practice, engineering organisations set the installation torque by calibration on the actual bolt, nut, washer and coating combination, and the values in the quick reference table below assume clean, lightly oiled threads.
Practical Torque Reference Values
The following values are typical installation torques for rail fishplate and fastening bolts at the commonly used strength grades, assuming clean threads and normal lubrication; they are starting points for a calibration, not substitutes for the project torque specification.
| Thread | Grade | Typical preload range (kN) | Typical tightening torque (N-m) |
|---|---|---|---|
| M16 | 8.8 | 60-80 | 250-300 |
| M20 | 8.8 | 95-125 | 400-450 |
| M24 | 10.9 | 160-200 | 800-900 |
| M30 | 10.9 | 250-310 | 1500-1600 |
The jump from M24 to M30 roughly doubles the torque and the preload, which is why the tightening method must change with the specification.
Why Hydraulic Wrenches Are Required for M24 and Above
A manual click-type torque wrench commonly used on track has a practical upper range around 400 N-m, which is below the design torque of an M24 grade 10.9 bolt at 800 to 900 N-m and far below the 1500 N-m region of an M30. Attempting to reach those values by hand is neither accurate nor safe. Hydraulic torque wrenches cover the 500 to 5000 N-m range, apply torque smoothly without shock, and hold the target value with a repeatability typically within about 5 percent. They also remove the human factor: with a manual wrench, each installer applies a different force, and the scatter in preload across a fishplate joint is large enough to leave some bolts near yield and others barely tight. Hydraulic tightening, ideally followed by a torque audit, keeps the preload scatter small and the joint behaviour predictable.
What Goes Wrong When Thread and Torque Do Not Match
Over-torquing a small bolt is the classic installation error. The bolt yields in the threaded section, the permanent elongation means the preload cannot be maintained, and under service vibration the joint loses its clamping force; the bolt then works under bending and fatigue and eventually fractures, usually at the first engaged thread. Under-torquing a large bolt leaves the joint loose from the start: the fishplate gap opens and closes under every wheel, the bolt sees alternating shear and bending, the rail ends batter, and the joint geometry deteriorates. A joint that is loose in this way consumes maintenance hours out of proportion to its length and is a known source of rail-end defects. The engineering fix is always the same: select the thread size so the required preload is comfortably within the bolt capacity, and set the torque so the preload is achieved without exceeding about 80 to 90 percent of proof stress.
Long Threads and Preload Stability on Heavy-Haul Lines
On heavy-haul joints the length of the engaged thread also matters. A longer thread engagement increases the number of loaded threads, spreads the load more evenly, reduces the stress concentration at the first thread, and lowers the shear stress per unit of engaged length. Bolts with longer threaded sections also tolerate more elastic strain, which helps the joint hold preload through temperature cycles and repeated impacts. This is the engineering reason why heavy-haul fishplate bolts are specified with longer thread lengths and why substitution with a short-thread automotive bolt is never acceptable in track service.
FAQ
Why is preload more important than torque itself?
Because preload is the actual clamping force that keeps the joint tight and the components from moving; torque is only a practical way to produce preload, and the same torque can give different preloads under different friction conditions.
What torque is required for an M24 grade 10.9 rail bolt?
Typical installation torque is in the range of 800 to 900 N-m with normal lubrication, which produces a preload of roughly 160 to 200 kN; the value should be calibrated for the actual coating and lubricant.
Can a manual torque wrench tighten M30 bolts correctly?
No. The design torque of an M30 grade 10.9 bolt is around 1500 N-m and above, far beyond the practical range of hand-held click wrenches; a hydraulic torque wrench is required.
Why do galvanized bolts need different torque values?
Because the zinc layer changes the thread friction coefficient, and therefore changes the fraction of torque that converts into preload; the tightening torque must be re-derived for the galvanized surface condition, usually with the threads lubricated.
What happens if a bolt is tightened past yield?
The bolt takes a permanent stretch, the preload relaxes below the design value, and under vibration the joint loosens; the bolt then carries fatigue cycles in bending and typically fails at the thread root.
How is the correct preload verified on site?
With a torque audit using a calibrated wrench, by measuring the additional angle of turn, or with a direct tension indicator such as a load-indicating washer; the acceptance criterion is a preload within the specified band, not simply the wrench reading.