Failure Analysis

Tooth wear

A worn coupling records the conditions it actually ran at, not the ones it was commissioned to. Three observations — where the contact patch sat, whether the wear is symmetric across the two meshes, and what the debris in the grease looks like — separate a lubrication problem from a misalignment problem before anything is measured.

Author
Priyansh Thummar, Editor
Dates
Published · Last updated
Reading time
4 minutes

1. Wear is the design condition

A gear coupling wears. That is not a fault condition — it is the mechanism the design accepts in exchange for its torque density. Every degree of misalignment makes each tooth slide along its mate and back once per revolution, under load, with only a grease film between the flanks. On a machine running continuously at 1500 rev/min that is over two million sliding cycles a day.

The engineering question is never whether the teeth wear, but whether they are wearing at the rate the design intended and in the pattern it predicts.

That is why a worn coupling is worth reading before it is replaced. It records the conditions the machine actually ran at, written on the flanks over the whole service life — usually more reliable than the alignment report in the file, which records the conditions on one morning, cold, before the machine had run.

2. Reading the contact patch

Hub teeth are crowned — barrelled along their length — so a correctly aligned coupling makes contact near the middle of the flank and keeps it there as the coupling articulates. The patch is polished, symmetric about the tooth centre, and comfortably inside both ends.

Anything else is telling you something.

Contact patch position on a coupling tooth flank under four conditionsFour tooth flanks drawn side by side, each a tall rectangle representing one tooth seen on its working face. On the aligned flank the contact patch is a band centred along the tooth length. Under excess angular misalignment the band is pushed toward one end of the tooth. Under overload the band spreads across almost the whole tooth length. Under reversing duty the band is centred but appears on both the leading and trailing flanks.Alignedwithin capacityAngularoutside capacityOverloadoutside capacityReversingoutside capacity
  • AlignedCentred, symmetric, well inside both ends
  • AngularDriven to one end by articulation at the mesh
  • OverloadCrown flattened; patch spreads the full length
  • ReversingCentred but on both flanks — backlash opens up
Figure 1. The tooth flank seen on its working face, with the contact patch drawn where each condition puts it. Schematic — patch proportions are illustrative. What matters in the field is the position relative to the tooth ends and whether the two meshes agree.

3. The patterns and what each means

Table 1 — Wear patterns, the condition each indicates, and how to confirm it
What you seeMost likely causeConfirming evidenceWhat it rules out
Polished band, centred, both meshes alikeNormal service wearDepth consistent with running hours; backlash change smallA specific fault — investigate duty and interval instead
Band pushed to one end of the toothAngular misalignment beyond the mesh capacityCompare meshes; check for soft foot, pipe strain, foundation movementGrease specification as the primary cause
Band spread over nearly the full tooth lengthCrown flattened — sustained overload or very long serviceTorque history, service factor, tooth thickness lossA recent change — this takes time to develop
Wear on both leading and trailing flanksTorque reversal in serviceBacklash growing faster than flank loss explainsSteady unidirectional duty
Pitting on the flankLubricant film too thin for the contact stressGrease condition, water content, interval overrunContact fatigue in the gear-tooth sense
Fine wavy line traced along the flankWorm tracking — contact migrating under articulationAlmost always accompanies a misalignment findingA manufacturing or machining mark
Pointed or knife-edged tooth tipsWear has passed the point of safe engagementAny of the above, left in service too longNothing — this coupling is at end of life
Patterns combine. A misaligned coupling on an overdue grease interval shows both, and addressing only one of them puts the same coupling back into the same condition.

3.1 Worm tracking

Worm tracking is the pattern most often misread. It appears as a fine wavy or serpentine line traced along the flank, and it is not a machining mark. It is the track left by a contact point migrating along the tooth as the coupling articulates through a misalignment large enough that the sliding is no longer confined to the crown.

Where it appears, the cause is misalignment rather than lubrication — although the two arrive together often enough that both need addressing.

4. The symmetry test

This is the observation that localises the fault, and it costs nothing but the discipline of looking at both ends before forming a view. A double-engagement coupling has a mesh at each end and both see the same torque.

  • Both hubs alike — the cause is general: duty, running hours, service factor, grease specification.
  • One hub markedly worse — the misalignment is concentrated at that end, and the cause is local to that machine: soft foot, pipe strain on a pump casing, a settled foundation, a bearing that has moved.

That single comparison converts an open-ended investigation into a bounded one. If the answer is "one end", the per-mesh capacity at that size is the figure the installation failed to hold.

5. What the grease is telling you

The third source of evidence, and the one routinely thrown away before anyone looks.

  • Fine grey or black paste — normal wear debris. Expected in moderate quantity.
  • Bright metallic flakes — scuffing or spalling rather than steady abrasion. Something more aggressive is happening.
  • Gritty, dark, hardened residue separated from the base oil — grease that has been overheated or has exceeded its service life. A lubrication failure regardless of what the flanks look like.
  • Rust-coloured fine powder at the bore or keyway — fretting, not tooth wear. A different investigation entirely.

6. An inspection sequence

  1. Photograph both hubs before cleaning anything.
  2. Note where the patch sits on each, and whether the two agree.
  3. Sample the grease before it is wiped away.
  4. Measure backlash and compare with the previous inspection — the change is more informative than the value.
  5. Only then check tooth thickness.
  6. Before the coupling goes back, answer two questions: is the wear consistent with the hours run, and if not, what changed?

Re-greasing a coupling that wore because it was misaligned resets nothing. The same misalignment is still there, and the wear resumes from where it stopped — realignment is part of the repair, not a separate job.

Frequently asked

How much gear coupling tooth wear is acceptable before replacement?
There is no single figure, because what matters is the rate and the pattern rather than the depth. A coupling that has lost a measurable amount of tooth thickness evenly over ten years is behaving normally; the same loss over ten months is a fault that will recur on the replacement. The practical triggers are backlash that has grown noticeably since the last inspection, wear that has reached or passed the crown so contact is no longer central, and any evidence that the tooth tip is becoming pointed.
What does one-sided tooth wear on a coupling indicate?
That the coupling was running at more angular misalignment than the mesh could accommodate, so the contact patch was pushed toward one end of the tooth instead of staying near the crown. Confirm it by comparing the two meshes: if one hub shows the pattern and the other does not, the misalignment is at that end and the cause is local — soft foot, pipe strain, a settled foundation — rather than a general alignment error.
Is pitting on coupling teeth the same as gear pitting?
The mechanism is related but the diagnosis differs. Coupling teeth do not roll against each other the way gear teeth do; they slide a short distance back and forth once per revolution. Pitting on a coupling flank therefore usually points to a lubricant film that was too thin for the contact stress — wrong grease, an overdue interval, or water contamination — rather than to a Hertzian contact fatigue limit having been reached in normal service.
Can a worn gear coupling be re-greased and put back in service?
Only after the pattern has been read and the cause addressed. Re-greasing a coupling that wore because it was misaligned resets nothing: the same misalignment is still there and the wear continues from where it stopped. Re-greasing is the right response when the pattern is symmetric, central and consistent with the running hours, and the grease itself is degraded or contaminated.

References

  1. AGMA 1010-F14, Appearance of Gear Teeth — Terminology of Wear and Failure, for the wear nomenclature used throughout this article.
  2. ISO 14691:2008, Petroleum, petrochemical and natural gas industries — Flexible couplings for mechanical power transmission — General-purpose applications.
  3. Editorial note: this article is written from wear nomenclature and mechanism, and Figure 1 is a schematic. It does not yet carry the photographed failures with service history and root cause that the failure-analysis cluster is built around. Photographs of returned units — with hours run, duty, grease type and interval attached — are to be added from Super Mech Industries service records and confirmed at technical review. Until then, treat the patterns here as a framework for what to look for, not as a substitute for examining your own coupling.
  • Maintenance

    Lubrication guide

    Grease selection, fill quantity and the intervals that prevent most of this.

  • Selection & Sizing

    Service factor guide

    Where uniform wear on a correctly aligned coupling usually originates.