Gearbox failure analysis is the practice of diagnosing why a gearbox failed by reading the damage patterns on the gear teeth. Pitting, spalling, scuffing, and wear each leave a distinct signature in a specific place on the tooth, and that signature points back to the root cause, overload, poor lubrication, misalignment, or contamination, so you can fix the cause instead of just replacing the gear.
A failed gear is a written record. The teeth do not lie: the location, shape, and texture of the damage tell you what killed the gearbox, and that is far more useful than the fact that it stopped turning. Swap the gear without reading the teeth and you install a new part into the same conditions that destroyed the old one. This guide walks the main tooth failure modes, where each shows up, and how to run the analysis so the next gearbox lasts.
Why read the gear teeth instead of just replacing the gear?
Because the failure mode tells you the root cause, and the root cause is what recurs. A gearbox that pits its teeth has a different problem than one that scuffs them, the first is a contact-stress or lubrication issue, the second is a lubricant-film breakdown under load and speed. If you cannot tell them apart, you cannot fix either, and the replacement gear fails the same way.
Reading the teeth is a core reliability skill, the gear equivalent of the diagnosis in our bearing failure modes work. It turns a breakdown into a root cause analysis and, done consistently, into rising mean time between failures as the underlying causes get eliminated. The standards world has built a shared vocabulary for exactly this so that a user and a gear manufacturer describe the same damage the same way.
What are the main gear tooth failure modes?
Gear tooth damage sorts into a handful of modes grouped by mechanism: surface fatigue (pitting, micropitting, spalling), lubrication-related (scuffing, wear), and bending fatigue (tooth breakage). Each looks different and comes from a different cause.
Pitting is surface-fatigue damage: repeated contact stress fatigues the surface until small craters break out, typically in a band at or just below the pitchline where contact stress peaks. A little initial pitting can be normal on new gears; progressive pitting that grows is a contact-stress or lubrication problem. Micropitting is its fine cousin, a frosted, matte grey band from very small-scale surface fatigue, common where the oil film is too thin. Spalling is advanced surface fatigue where larger flakes break away, often from a subsurface origin under heavy load.
Scuffing (the term the standards prefer over the old word "scoring") is not fatigue at all, it is adhesive damage from the lubricant film collapsing under high load and sliding speed, letting metal contact and tear, leaving radial score marks near the tooth tip and root where sliding is highest. Abrasive wear gradually thins the tooth profile, usually from contamination in the oil acting as grinding grit. Tooth breakage is bending fatigue: a crack starts at the root fillet where bending stress concentrates and grows until the tooth snaps off, the most catastrophic mode.
Where on the tooth does each mode appear?
Location is half the diagnosis, because the position of the damage maps directly to the type of stress that caused it. Contact-fatigue damage lives on the working flank; bending-fatigue damage lives at the root.
The practical shortcut: damage on the working flank (pitting, micropitting, spalling, scuffing, wear) points to contact stress, lubrication, or contamination. A crack at the root fillet points to bending fatigue from overload, shock loading, or a stress raiser. Uneven damage across the face width points to misalignment the load is concentrated on one end of the teeth instead of spread evenly.
How do you run a gearbox failure analysis?
Work the analysis in order and preserve evidence as you go. The temptation is to pull the broken gear and order a replacement; resist it until you have read the whole story.
- Preserve the evidence before you disturb it. Photograph the teeth, the oil, and the debris before cleaning anything. Note which teeth are damaged and how the damage is distributed around the gear and across the face. Once you wire-brush a tooth, the story is gone.
- Identify the failure mode from the tooth surface. Match the damage to a mode, pitting, micropitting, spalling, scuffing, wear, or breakage, using appearance and location. This is the single most informative step.
- Read the load distribution across the face. Even damage across the full face width suggests a load or lubrication issue; damage biased to one end suggests misalignment or shaft deflection. This alone catches a huge share of gearbox failures.
- Examine the lubricant and debris. Check oil level, condition, and contamination, and inspect the magnetic plug and filter. Oil analysis and wear-particle data tell you whether the lube film ever had a chance. Fine grit means abrasive wear; large flakes mean spalling.
- Inspect the bearings, seals, and shafts. A gearbox failure is often a bearing or seal failure that let contamination in or allowed misalignment. Check them against your bearing failure modes reference, and check splines and couplings for fretting.
- Trace back to the root cause. Connect the mode and its location to a cause: overload, thin oil film, wrong viscosity, contamination, misalignment, or an assembly error. Use five whys to get past the surface answer.
- Correct the cause, then verify. Fix the actual cause, align the shafts, change the lubricant, add filtration, correct the load, not just the gear, and confirm the fix by monitoring the replacement. A repair you cannot verify is a guess.
What do the other clues tell you?
The teeth are the headline, but the supporting evidence confirms the diagnosis. Oil analysis reveals the invisible: wear-metal trends, water or coolant ingress, and viscosity loss that explain a lubrication-related mode before you ever open the case. The bearings and seals tell you how contamination got in or how misalignment developed. Vibration data, tracked under a predictive maintenance program, often shows the gear-mesh or bearing signature climbing for weeks before the failure, the early warning that a condition-based maintenance plan is built to catch.
How do you prevent gearbox failures?
Most gearbox failures trace back to four preventable causes: lubrication, alignment, contamination, and overload. Get the lubrication right, correct viscosity, correct level, clean oil, changed on interval, and you eliminate the largest share of failures, because most surface modes are film-thickness problems in disguise. Align the shafts and couplings to spread load evenly across the face width. Keep contamination out with good seals, breathers, and filtration. And keep the gearbox within its rated load, watching for the shock loads that crack root fillets. Tie all four into the preventive maintenance schedule so they happen on a plan, not after a failure.
What do the standards say?
Gear failure has a codified language, which matters because a vague term sends the analysis the wrong way.
- ISO 10825-1:2022 provides the international nomenclature and appearance descriptions for gear wear and failure modes, pitting, micropitting, scuffing, wear, spalling, and tooth breakage, specifically to improve communication between equipment users and gear manufacturers during failure analysis (ISO 10825-1:2022).
- The standard deliberately discourages the term "scoring," because it has been applied to both adhesive and abrasive wear; the correct term for lubricant-film-collapse damage is scuffing. Using the right word keeps the diagnosis pointed at the right cause.
- The rating side, how gears are sized against contact (pitting) and bending (breakage) fatigue, is covered in the ANSI/AGMA 2001 framework, the reference gear engineers use to connect a failure mode back to a design margin (AGMA standards).
The through-line: name the mode with the standard vocabulary, place it on the tooth, and the physics tells you the cause. A gearbox that keeps failing is a gearbox whose root cause was never read off the teeth.
Where do the records live?
Gearbox diagnosis is a longitudinal problem. The most valuable signal is not this one failure but the pattern, the same gearbox pitting at every rebuild, the oil analysis trending toward contamination for three quarters, the vibration climbing before every failure. That pattern only exists if the teardown photos, the failure-mode call, the oil-analysis results, and the vibration history all accumulate against the asset. When they scatter across a technician's phone, a lab PDF, and a paper rebuild sheet, the story that would prevent the next failure is never assembled.
Harmony's role is to keep that record in one place: capture the failure-mode findings and photos at the asset during the teardown, hold the oil and vibration history tied to the equipment, and surface the recurring pattern instead of losing it across systems. It layers onto the tools a plant already runs. No rip-and-replace. The CLS case study shows the move from paper records to real-time capture, and the platform overview shows how the pieces connect. The teeth always record the cause; the plants that stop repeat failures are the ones that write down what they read and keep the history where the next analyst can find it.