A bolted joint leaks, or the bolt breaks in service, even though the wrench clicked at the target torque. The torque was set, but the actual bolt tension was wrong. Torque is not preload. Tightening a bolt converts torque into clamp force, but friction in the threads and under the nut eats most of it. Understanding preload is the difference between a joint that holds and one that fails.
What preload does
A bolted joint works by clamping the parts together. The bolt stretches slightly and acts like a stiff spring, pressing the flanges together. This clamp force, the preload, keeps the joint sealed and prevents the parts from sliding or separating. If preload is too low, the joint leaks or the bolt sees dynamic load and fatigues. If too high, the bolt yields or the clamped parts crush. The goal is a controlled clamp near, but below, the bolt’s yield.
Torque is a proxy for preload
Only a fraction of the applied torque becomes bolt tension. Most torque overcomes friction under the nut and in the threads. Roughly 50 percent goes to nut-face friction, 40 percent to thread friction, and only 10 percent to stretching the bolt. This is why the same torque can give very different preload depending on lubrication, surface condition, and bolt finish. A dry bolt and an oiled bolt at the same torque can differ by a factor of two in clamp force.
Use the right torque and lubrication
Torque values assume a defined friction condition, usually a specified lubricant. If you lubricate a bolt rated for dry torque, you over-tighten and may yield it. If you run a lubricated bolt dry, you under-tighten. Always use the lubrication the torque table assumes, and record it. For critical joints, measure preload directly with a load cell or ultrasonic elongation rather than relying on torque alone.
Yield and proof load
High-strength bolts are tightened to a percentage of their proof or yield load, often 70 percent, to leave margin. Tightening past yield stretches the bolt permanently and ruins the clamp. Torque wrenches that click are good, but they do not detect a seized thread or a nut that bottoms out. For critical joints, use torque plus angle, or measure elongation. Stop before yield.
Why joints loosen
Vibration, thermal cycling, and embedment can loosen a bolted joint. If preload is insufficient, parts slide and the bolt sees shear or fatigue. Self-loosening happens when the nut rotates back under transverse vibration. Lock washers, nylon insert nuts, or thread locker help, but the best prevention is adequate initial preload. A properly preloaded bolt stays clamped; a low-preload joint needs every locking trick in the book.
Torque pattern and multiple bolts
Tighten multiple bolts in a crisscross pattern, in passes, so clamp force builds evenly. Tightening one bolt fully before the next distorts the flange and leaves uneven load. Repeat the pass at increasing torque. For flanges, this is essential for a uniform gasket load. Skip the pattern and one side stays loose even though every bolt clicked.
Thermal expansion and relaxation
Differential thermal expansion between bolt and clamped parts changes preload as temperature changes. Embedment, where surface roughness flattens, also relaxes preload. Retorque after initial run-in where the design allows. For hot joints, choose materials and lubricants that retain preload, and account for expansion differences.
A worked torque check
Take an M12 property 8.8 bolt with a target preload of about 70 percent of proof load. The torque table, assuming lubricated threads, gives a torque around 85 Nm. If the mechanic runs the bolt dry because the table says lubricated, the actual preload may be half the target, and the joint can slip or leak. If he uses anti-seize where a plain oil was assumed, preload can rise past yield. The torque value is meaningless without the friction condition. Record the lubricant on the job, and the torque becomes repeatable.
Torque plus angle control
Critical joints use torque plus turn: snug the bolt to a low seating torque, then turn it a set angle. This is more repeatable than torque alone because it controls bolt stretch beyond the scatter of friction. It is common in automotive and structural joints. The angle places the bolt into the preload range without measuring elongation directly. Set the snug torque and angle from the joint specification, and mark the bolt so turn is visible.
Direct preload measurement
Where the cost of failure is high, measure preload directly. Ultrasonic bolts measure stretch with sound velocity; load cells or washers measure clamp force. These remove the friction guesswork. They cost more, but for pressure vessels, critical machinery, or structural joints, the data is worth it. Don’t measure every bolt on a farm implement, but do measure the ones that hurt people if they fail.
Locking hardware limits
Nylon nuts, lock washers, and thread locker help only when preload is already adequate. A lock nut on a loose joint still vibrates loose. Serrated washers can damage surfaces and create false torque. Use locking hardware as insurance, not as a substitute for preload. If a joint repeatedly loosens, first check the preload, bolt stiffness, and joint separation before blaming the locknut.
Bolt stiffness and joint member stiffness
The joint behaves as two springs in series: the bolt in tension and the clamped parts in compression. If the clamped members are soft or gaskets compress, the bolt sees more of the external load and can fatigue. A stiff joint, with thick members and a short bolt, keeps the load on the clamp. Long, slender bolts are more flexible and less forgiving. Design the joint stiffness, not just the bolt size.
Inspection after tightening
After assembly, mark each bolt and nut with a witness line. This lets inspection see whether a nut has rotated. Re-check torque after the first run-in or heat cycle. For flanges, follow the retorque schedule. Record who tightened, to what torque, and with what lubricant. These records matter when a joint fails in service.
Common mistakes
Using the wrong lubrication, tightening one bolt at a time, relying on lock washers instead of preload, ignoring yield, and skipping retorque are the recurring errors. Torque is a means to preload, not the end. Control the clamp force, and the bolted joint stays sealed and tight.
Choosing bolt grade
Bolt strength grades, such as 8.8 or 10.9, set the yield and proof load. Higher grades allow more preload for the same bolt size, but they are also more notch-sensitive and less forgiving of over-tightening. A 10.9 bolt at the same torque as an 8.8 may be near yield if the torque was set for the lower grade. Never mix grades in the same joint, and don’t upgrade a bolt without rechecking the torque and the clamped member strength. Stronger bolts do not solve a weak flange; the clamped parts must also hold.
When to retorque
Retorque after initial embedment, typically after the joint has seen its first load or heat cycle, and on gasketed joints after warm-up. Don’t retorque a live pressurized hot joint without isolation. Mark the bolts and record the retorque. A joint that has been properly retorque once stays put; one that has not relaxes and leaks.
Field inspection of loose joints
When a bolted joint repeatedly loosens, don’t just retorque it again. Look for joint separation under load, soft gaskets, creeping clamped parts, or a bolt that is too long and flexible. Measure whether the parts move under load. If the joint separates, preload is inadequate and the bolt will fatigue. Increase preload within yield, shorten the bolt grip, or stiffen the members. Retorquing a separating joint only delays the break.
Finally, document the torque, lubricant, and tightening pattern on the assembly drawing or work order. A bolted joint that is tightened differently each time it is opened is a future failure. Consistent documentation turns preload from guesswork into a repeatable process, which is the real point of controlling bolt tension.
Finally, document the torque, lubricant, and tightening pattern on the assembly drawing or work order. A bolted joint that is tightened differently each time it is opened is a future failure. Consistent documentation turns preload from guesswork into a repeatable process, which is the real point of controlling bolt tension in production and maintenance.
Document the torque, lubricant, and tightening pattern on the drawing. A joint tightened differently each time is a future failure. Consistent records make preload repeatable.
Document the torque, lubricant, and tightening pattern on the drawing. A joint tightened differently each time is a future failure. Records make preload repeatable.
Bottom line
Bolt preload is the clamp force that holds a joint, and torque is only an approximate proxy. Use the correct lubrication, tighten to a safe fraction of yield, follow a crisscross pass pattern, and rely on adequate preload rather than locking hardware to resist loosening. For critical joints, measure elongation. Torque that clicks at the wrench does not guarantee the right clamp. Treat the joint as a spring and control the stretch, not just the wrench reading.
