Lubrication failure

1. Four ways the film is lost

A gear coupling can be full of what looks like grease and be, in the way that matters, unlubricated. That defeats the obvious inspection — opening a sleeve, finding grease in it, and concluding lubrication was not the problem.

1.1 Centrifugal separation

This is the mechanism most specific to couplings, and the one least familiar to people who come to couplings from bearings. Grease in a rotating sleeve experiences an outward acceleration proportional to radius and to the square of speed — hundreds of times gravity on a medium coupling at ordinary industrial speed, applied continuously for the whole service life.

An ordinary bearing grease separates under that loading within weeks. Coupling greases are formulated to resist it, which is the property being paid for and the reason ANSI/AGMA 9001 classifies them separately as CG-1, CG-2 and CG-3. The lubrication guide covers the selection side of this.

2. The seal is usually where it started

Whichever mechanism is nominally responsible, start the investigation at the seal. A gear coupling depends completely on grease staying where it was packed and dirt staying out, and the radial shaft seal doing that job is a simple, cheap, ageing component.

When it hardens, loses lip contact and begins to pass, the coupling does not fail immediately and does not announce anything. Lip material is a temperature decision, and a nitrile seal specified on a machine that runs hot will harden well inside its expected life.

3. The delay that makes it dangerous

Nothing changes in vibration or temperature at the moment the seal goes. The mesh runs on for weeks or months on a film that is thinning and getting dirtier, and by the time the symptom reaches an operator — grease thrown around the guard, a temperature rise at the sleeve — the flank damage has already happened.

A great many coupling failures recorded as lubrication failures are, at root, seal failures that nobody inspected. That is not a distinction without a difference: replacing the grease without replacing the seal schedules the next one.

4. Reading the grease that comes out

The best evidence available, routinely wiped away before anyone looks at it. Sample it before the coupling is cleaned — the sample is the diagnosis.

• Clean, homogeneous, fine grey debris — a coupling behaving normally.
• Free oil at the wall, stiff residue at the teeth — separation, caught in the act.
• Dark, hardened, gritty, parted from its oil — thermal degradation or an interval overrun.
• Bright metallic flakes — something more aggressive than steady abrasion.
• Milky or emulsified — water.

5. Telling it from a mechanical failure

Mostly a matter of symmetry. Lubrication is a condition of the whole coupling: both meshes see the same grease, so both hubs degrade at a similar rate and in a similar position on the flank.

Misalignment is geometric and usually concentrated at one end, so one hub is markedly worse and its patch sits toward a tooth end rather than centrally. Overload flattens the crown and spreads the patch across nearly the whole flank on both hubs, which can read as a lubrication problem until the torque history is checked.

The two arrive together often enough to state plainly: misalignment raises sliding velocity at the flank, which raises the demand on the film, which brings forward the point at which an adequate grease and an adequate interval stop being adequate. A coupling that is both misaligned and overdue shows both signatures — and addressing only the grease returns it to service in the condition that caused the failure.

6. What to change

Follows from which mechanism applied, per Table 1. Two points that are easy to get wrong:

• For separation, check for an AGMA CG designation rather than an NLGI grade. NLGI describes stiffness and says nothing about behaviour under centrifugal loading.
• For starvation, the fill quantity is published per size, not judged at the machine with a grease gun. Overfilling pressurises the sleeve and pushes grease past the seals, which presents as a leak rather than as a filling error.

In every case, replace the seals while the coupling is open. They are the cheapest component in the assembly and the most likely origin of the next occurrence.