Loosening Is Not About Insufficient Tightening

The most common first response to a loose bolt is to tighten it harder. This almost never solves the problem. If a bolt loosens on its own under vibration, applying more torque only delays the failure - and risks yielding the bolt. The reason is that bolt loosening is rarely caused by the preload being too low. It is caused by the thread pair rotating in the unscrewing direction, which happens when transverse slip occurs between the clamped parts. Once the bolt starts rotating backwards, it unscrews quickly regardless of how high the initial preload was.

This was demonstrated definitively by the Junker vibration test, which shows that a standard bolted joint under transverse vibration can lose all preload in a few seconds, even if it was initially torqued to 90% of yield. The test also shows that once preload drops below a critical threshold, the bolt rotates freely and loosens completely. No amount of initial torque prevents this; the only cure is to stop the rotation.

Three Mechanisms of Loosening

Rotational self-loosening is the most damaging mode. Transverse vibration causes the clamped plates to slide back and forth. This slip overcomes the friction under the bolt head and in the threads, allowing the nut to rotate backward (unscrew) one thread pitch at a time. Once started, it accelerates until the joint is completely loose. This is the loosening you see on engine mounts, chassis bolts, and rail fasteners.

Slackening or embedding happens when microscopic roughness peaks on the clamped surfaces crush under preload, causing the bolt to shorten slightly and lose a small amount of preload. This occurs in the first few hundred hours of service and is usually manageable by re-torquing after run-in. It is not the catastrophic loosening that causes failures.

Preload decay or relaxation occurs when the clamped material creeps (in plastic joints), relaxes under high temperature (in high-temperature joints), or stress-relieves over time. The bolt loses preload gradually without rotating. This is common in gasketed joints, plastic housings, and high-temperature flanges.

Why Vibration Is the Catalyst

Not all vibration causes loosening. Longitudinal vibration (along the bolt axis) tends to compress the clamped parts but does not generate the transverse slip needed to overcome thread friction. Transverse vibration (perpendicular to the bolt axis) is the killer: it causes the joint plates to slide sideways, which directly overcomes the friction under the bolt head and allows rotation. This is why Junker testing uses transverse excitation, and why bolts in engines, transmissions, and rail tracks - which see strong transverse vibration - are the ones that loosen.

Mainstream Anti-Loosening Solutions Compared

Note that the standard spring lock washer is largely ineffective against transverse vibration - the washer flattens under preload and provides negligible locking force. Engineering literature and Junker test data consistently show that spring washers do not prevent rotational loosening in vibrating joints. If you are still using spring washers as your primary anti-loosening method, that is likely the root cause of your loosening problems.

The Combined Solution for a Permanent Fix

A robust anti-loosening design usually combines two or more methods. For most automotive and machinery applications, an all-metal prevailing-torque nut per GB/T 3098.9 plus controlled torque is sufficient. For high-vibration joints such as engine mounts, add pre-applied thread adhesive. For high-temperature joints where adhesive and nylon fail, use wedge-locking washers or all-metal prevailing-torque nuts. For safety-critical joints, use a mechanical positive lock such as wire locking or castellated nuts with cotter pins.

The key principle is: stop the rotation, not the preload. Methods that only increase friction or add spring tension do not stop rotational loosening under transverse vibration. Methods that physically prevent the nut from rotating - prevailing torque, adhesive curing, wedge geometry, or mechanical wiring - are the ones that work. Diagnose which loosening mechanism your joint sees, then choose a method that directly addresses it.

Diagnosing your specific loosening problem. Before choosing an anti-loosening method, observe how the bolt comes loose. If the nut has rotated backwards (you can see thread exposed past the nut face), it is rotational self-loosening - use a prevailing-torque nut, wedge washer, or adhesive. If the bolt remains tight visually but the joint leaks or rattles, it is preload decay from embedding or relaxation - re-torque after run-in or use Belleville washers to compensate. If the bolt shears, it is overloading, not loosening. Matching the fix to the failure mode prevents wasting money on the wrong anti-loosening hardware.

The Junker test and what it tells you. The Junker transverse vibration test (per ISO 16130) is the gold standard for anti-loosening performance. A bolted joint is mounted on a test rig that applies transverse vibration at a defined displacement, and the residual preload is measured continuously. A standard bolt with no locking device loses all preload in seconds. A spring lock washer delays the loss but does not prevent it. A prevailing-torque nut retains most of the preload. A wedge lock washer retains nearly all of it. When selecting an anti-loosening method for a vibrating joint, ask the supplier for Junker test data rather than relying on marketing claims. The test results tell you which method actually works for your load level and vibration frequency.

Re-torquing as a maintenance practice. For joints that embed or relax over time, planned re-torquing is a legitimate maintenance strategy. Many heavy equipment and automotive OEMs specify re-torque after the first 50-100 hours of operation, then periodically. Belleville washers can reduce the frequency of re-torquing by compensating for embedding. For gasketed joints, re-torque after the first heat cycle is standard practice. Document the re-torque schedule in the maintenance manual so that field service does it systematically rather than waiting for a failure. This is cheaper than warranty claims and safer than in-service loosening.

Diagnosing your specific loosening problem. Before choosing an anti-loosening method, observe how the bolt comes loose. If the nut has rotated backwards (you can see thread exposed past the nut face), it is rotational self-loosening - use a prevailing-torque nut, wedge washer, or adhesive. If the bolt remains tight visually but the joint leaks or rattles, it is preload decay from embedding or relaxation - re-torque after run-in or use Belleville washers to compensate. If the bolt shears, it is overloading, not loosening. Matching the fix to the failure mode prevents wasting money on the wrong anti-loosening hardware.

The Junker test and what it tells you. The Junker transverse vibration test (per ISO 16130) is the gold standard for anti-loosening performance. A bolted joint is mounted on a test rig that applies transverse vibration at a defined displacement, and the residual preload is measured continuously. A standard bolt with no locking device loses all preload in seconds. A spring lock washer delays the loss but does not prevent it. A prevailing-torque nut retains most of the preload. A wedge lock washer retains nearly all of it. When selecting an anti-loosening method for a vibrating joint, ask the supplier for Junker test data rather than relying on marketing claims. The test results tell you which method actually works for your load level and vibration frequency.

Re-torquing as a maintenance practice. For joints that embed or relax over time, planned re-torquing is a legitimate maintenance strategy. Many heavy equipment and automotive OEMs specify re-torque after the first 50-100 hours of operation, then periodically. Belleville washers can reduce the frequency of re-torquing by compensating for embedding. For gasketed joints, re-torque after the first heat cycle is standard practice. Document the re-torque schedule in the maintenance manual so that field service does it systematically rather than waiting for a failure. This is cheaper than warranty claims and safer than in-service loosening.