What is tribology, and what does it have to do with machinery failures?
Tribology is the science and engineering of surfaces in relative motion: friction between them, wear of them, and the lubrication that separates them. Most mechanical failures that reach a failure investigation, seized bearings, spalled gear teeth, scored journals, shafts whose fatigue cracks started at a wear mark or a fretted fit, are tribological failures, which makes the discipline central to failure analysis even where nobody uses the word.
The reason the word matters is that tribology supplies the reading skills. A worn surface is a record of the loads, alignment, lubrication, and contamination the machine actually experienced, as opposed to what it was designed for or what the maintenance log says it experienced. An investigator who can identify the wear mechanism can usually say something disciplined about the conditions that produced it, and that is where a cause determination starts.
Where did tribology get its name?
The name tribology was coined in Britain in 1966, when the government commissioned Peter Jost, an engineer with a background in lubrication, to examine the machinery failures then afflicting industry, steelworks prominently among them. The Jost report argued that friction and wear were costing British industry an estimated 500 million pounds a year, proposed a unified discipline for the problem, and named it from the Greek tribo, to rub.
The practice is far older than the name. Notes on friction appear in Leonardo da Vinci's notebooks, and Egyptian tomb art shows workers pouring liquid ahead of sledges hauling stone. What the Jost report created was an organized field, and sixty years on it remains nearly invisible next to its economic weight. A Financial Times science column by Anjana Ahuja made that point this week, quoting the University of Southampton's Robert Wood, who calls tribology "one of the enabling technologies that silently underpins industrial productivity."
What are the main wear mechanisms in bearings and gears?
The wear mechanisms that matter most in bearings and gears are adhesive wear, abrasive wear, rolling-contact fatigue, and fretting, and each leaves a signature that can be distinguished under magnification. Adhesive wear is metal transferring between surfaces that touched because the lubricant film failed to keep them apart. Abrasive wear is cutting by hard third-body particles, which usually means contamination. Rolling-contact fatigue initiates cracks at or below the surface under repeated contact stress until material spalls out of a raceway or a tooth flank. Fretting is micro-motion damage at fits and joints that were never meant to move at all, and it is a common birthplace of fatigue cracks.
Two further families earn their place in modern investigations. Corrosion converts the contact surface itself, often where water entered the lubricant or a machine sat idle, and electrical discharge damage, the fluting left when current crosses a bearing, appears regularly in equipment driven through variable frequency drives. Identifying the mechanism is the first step of the analysis; locating what caused that mechanism to operate is the second, and the two steps should not be collapsed into one.
Why do lubrication failures look sudden when they are not?
Lubrication failures look sudden because the final event is fast, a bearing that seizes or a gear set that welds and fractures within minutes, while the path to that event is almost always slow. The recurring paths are starvation, where a blocked passage, a failed seal, or a missed interval leaves the contact running dry; contamination, where water, dirt, or wear metal turns the lubricant into an abrasive carrier; degradation, where oxidation, additive depletion, or overheated grease leaves the film unable to carry its load; and specification error, where the lubricant in the machine is simply not the lubricant the duty requires.
Each of those paths writes a record months long: rising particle counts, wear metals accumulating in oil samples, viscosity drifting, a vibration signature changing character. By the time the machine stops, the question is rarely whether lubrication failed, which the surfaces usually show plainly, but which path it took, when the record first showed it, and what the maintenance program did with that information.
Why do wind turbine gearboxes fail before their design life?
Wind turbine gearboxes have a documented record of bearing failures well short of the twenty-year design life of the turbines that carry them, and the candidate explanations under active research are tribological. Wind loading is relentlessly transient: gusts, direction changes, braking, and grid events all pass through the drivetrain, and a lubricant film sized for steady conditions can thin and shear under transient ones. Bearing steels in this service also show cracking and microstructural alteration whose root cause the research literature still debates, which is a reminder that a field sixty years old retains open questions of real commercial consequence.
The economics explain the research attention. A gearbox exchange on an offshore turbine is a vessel and crane operation scheduled around weather, so the difference between detecting a failing bearing early and discovering it at failure is measured in months of downtime. That is why this corner of tribology has moved fastest toward continuous monitoring and prediction.
Can wear be predicted before a machine fails?
Increasingly, yes: wear is one of the few failure processes that can be read continuously while the machine runs. Oil analysis tracks wear metals, particle counts, viscosity, and water in trend form. Debris analysis reads the size and shape of captured particles, which change as a contact deteriorates. Vibration monitoring picks up the characteristic frequencies of a defective bearing long before it fails, and thermography sees the heat of a contact working too hard. The declared ambition of the field is to predict and prevent wear rather than measure it after the fact.
That ambition lines up with the central finding of the Institute's research on the 100 worst engineering failures of the modern era: in 67 of the 100, the knowledge needed to prevent the failure existed before the event. Tribological failures belong to that pattern more than most failure types, because the evidence of decline is continuous and measurable while the machine is still running. The full analysis is published in Known in Advance.
What should be preserved after a bearing, gear, or shaft failure?
Preserve the lubricant first: a sample drawn from the failed machine, a sample of fresh stock from the same supply, and the filters, because the oil and what it carries are evidence as surely as the fracture surface is. Keep the failed components uncleaned and unassembled, keep mating parts and fasteners together and labeled, and resist the urge to wipe a wear surface clean to see it better, since the deposit being wiped away may be the finding.
Alongside the hardware, preserve the paper and the data: lubrication and maintenance records, oil analysis history if the machine was on a program, alarms and operating data around the event, and photographs of the machine as found. Restarting a damaged machine to see whether it still runs converts a readable record into an unreadable one, and it is the single most common way this class of evidence is lost in the first week.