Misalignment damage

1. What misalignment does at the flank

Every degree of misalignment forces each hub tooth to slide back and forth along its mating sleeve tooth once per revolution, under load, with only a grease film between the flanks.

The published capacity is a survival limit, not an operating target. A coupling running near it is inside specification and wearing many times faster than one running near zero. Misalignment damage is what that accelerated sliding leaves behind.

2. The signature set

• Patch driven to one tooth end — articulation has exceeded what the crown radius was designed to accommodate.
• Worm tracking — a fine wavy line traced along the flank by the contact point migrating. Frequently mistaken for a machining mark.
• Wear concentrated at the tooth ends rather than distributed across the working face.
• Pointed tooth tips in advanced cases, as material is lost from one side only.

2.1 Asymmetry is the decisive one

None of those is decisive on its own. The finding that is: the two meshes disagree.

A double-engagement coupling has a mesh at each end. Both carry the same torque, both see the same grease, both have run the same hours. A difference between them cannot be explained by duty, lubricant, service factor or age — it can only be geometric. Where one hub is markedly worse, the misalignment is concentrated at that end, and the cause is local to that machine rather than general to the train.

That comparison bounds the investigation, and it costs nothing but the discipline of examining both ends before forming a view. It is also the step most often skipped: the failed hub gets photographed, and the good one gets cleaned and put back.

3. Angular and offset leave different marks

Less reliable than the asymmetry test, but useful. Angular misalignment articulates the hub about a point, so the patch migrates in a way that varies around the circumference. Parallel offset is produced by two opposed angular deflections, so it tends to move both patches, in opposite directions.

Heavy one-sided wear at one mesh with light central wear at the other is angular at that end. Moderate opposed patterns at both meshes is offset — and the offset capacity at that size is the figure the installation failed to hold.

4. Ruling out the imposters

5. Where the misalignment came from

Once established, the useful question — and the answer is almost never the coupling.

• Soft foot distorts a casing as soon as its bolt is tightened, so any alignment measured afterwards was a measurement of a distorted machine.
• Pipe strain applies the load of pipework pulled into position, and keeps applying it as the line heats and cools.
• Foundations settle. Bearings move.
• Thermal growth lifts a running centreline above where it sat at commissioning.

Which is why a train aligned to zero cold is misaligned hot, in the condition it spends nearly all its life. An alignment report from the day of commissioning is evidence about that morning and about nothing since — the cold offset exists precisely for this.

6. Confirming it before stripping

1. Photograph both hubs before anything is cleaned.
2. Note where each patch sits, and whether the two agree.
3. Take alignment readings as found, before any bolt is slackened — slackening a bolt on a machine with soft foot destroys the evidence.
4. Check the feet and the pipework for the conditions in §5.
5. Measure backlash and compare with the previous inspection.

Only then take it apart. And realign the train as part of the repair rather than as a separate job — a new coupling fitted to the same geometry wears the same way.