Wear is the progressive loss of material from a surface in relative motion, and in machinery it happens through four mechanisms: adhesive, abrasive, surface fatigue, and corrosive wear. Each one removes metal a different way and leaves a different mark, so naming the mechanism is the first step in stopping it.
The four-mechanism list traces back to a 1957 survey by Burwell and has been the working vocabulary of tribology, the study of friction, wear, and lubrication, ever since. This guide walks through each mechanism, the fingerprint it leaves on a failed part, why one machine often shows two at once, and the three levers you actually pull to slow wear down. It is a foundation piece for the broader work of equipment reliability.
What are the four wear mechanisms?
The four mechanisms are adhesive, abrasive, surface fatigue, and corrosive wear. They differ in how material leaves the surface, not in how bad they are.
- Adhesive wear happens when two surfaces slide and the high spots, asperities, cold-weld together under load, then tear apart as motion continues. Material transfers from one surface to the other and eventually breaks free as debris. Its severe form is scuffing or galling, the smeared, torn look you see on a seized shaft or a gear tooth that lost its oil film.
- Abrasive wear is cutting and ploughing. A harder surface or a loose hard particle gouges a softer one, leaving parallel scratches in the sliding direction. Two-body abrasion is a hard counterface cutting directly; three-body abrasion is grit trapped between the surfaces rolling and cutting as it goes. Contaminated oil turns almost any bearing into a three-body abrasion problem.
- Surface fatigue wear is failure from repeated contact stress, not sliding. Under rolling contact, bearings, gears, cam followers, subsurface cracks start, grow, and eventually throw off a flake of metal. The fingerprint is pitting and spalling: small craters where the surface has fractured away rather than been rubbed off.
- Corrosive wear is a chemical-plus-mechanical tag team. The surface reacts with its environment, moisture, acids, oxidized oil, to form a brittle film, then sliding wipes that film off and exposes fresh metal to react again. Neither corrosion nor rubbing alone would remove as much as the two do together. Fretting corrosion, at tight clamped joints under vibration, is a common special case.
How do you tell which wear mechanism you have?
Read the surface and the debris. Each mechanism leaves a distinct fingerprint, and the pattern usually points to one dominant cause even when several are present.
| Mechanism | What the surface looks like | Where it shows up | Debris signature |
|---|---|---|---|
| Adhesive | Smeared, torn metal; scuffing, galling, cold-weld drag marks | Sliding contacts starved of oil: gears, plain bearings, pistons | Large, irregular flakes; transferred material |
| Abrasive | Parallel scratches and grooves along the sliding direction | Dirty environments, contaminated oil, unfiltered systems | Cutting chips; silica and hard-particle grit in oil |
| Surface fatigue | Pits, craters, spalling; cracks under a still-smooth surface | Rolling contacts: rolling-element bearings, gear flanks, cams | Sudden flakes and spheres; spikes in ferrous debris count |
| Corrosive | Etched, discolored, rusty, or frosted patches; pitting | Wet, chemical, or high-temperature service; stagnant joints | Fine oxide particles; rust-colored fines in oil |
Two tools turn this from art into evidence. Oil analysis reads the metal the machine is shedding, the mix of elements and the shape of the particles distinguishes cutting (abrasive) from flaking (fatigue) from oxide fines (corrosive). And machine monitoring catches the trend: a bearing entering fatigue spalling shows a rising vibration signature at its bearing defect frequencies long before it comes apart. For the failure-mode language behind all of this, see bearing failure modes.
Why does one machine show two mechanisms at once?
Because wear cascades. One mechanism opens the door for the next, so a part that started with a single problem often reads like several by the time it fails.
A classic chain: contaminated oil starts abrasive three-body wear on a journal bearing. The grit opens up the clearance. A wider clearance thins the oil film until, at start-up, asperities touch and adhesive scuffing begins. The scuffed surface runs hotter, the oil oxidizes and turns acidic, and now corrosive attack joins in. Three mechanisms, one root cause, dirt in the oil. This is why chasing the label on the final failure is a trap; you want the mechanism that started the cascade, which is a job for root cause analysis. It is also why lubrication failure modes sit underneath so many wear problems: the oil film is the one defense common to three of the four mechanisms.
Can you predict wear with an equation?
Roughly, yes, the Archard equation is the workhorse. It says the volume of material worn away rises with the load pressing the surfaces together and the distance they slide, and falls as the material gets harder:
Wear volume ≈ K × (load × sliding distance) ÷ hardness. K is a dimensionless wear coefficient that captures how aggressive the specific contact is.
Treat Archard as a way to reason, not a number to trust to three decimals. K spans orders of magnitude between a well-lubricated bearing and a dry sliding contact, and it is the film that moves it. Full hydrodynamic lubrication, surfaces floating on a wedge of oil, never touching, drops the wear coefficient toward negligible. Let that film collapse and K jumps by a factor of thousands. That single fact is why lubrication management returns more wear reduction per dollar than harder metal on most machines.
How do you slow wear down? A five-step loop
You cannot eliminate wear, but you can push a machine from severe to mild and buy years of life. There are really only three levers, and the useful thing about them is that one lever often defends against several mechanisms at once, which is why the order you pull them in matters as much as the levers themselves.
Work the levers in order, the early steps are cheap and cover the most ground.
- Name the dominant mechanism first. Pull the part, read the surface, and confirm with oil analysis or vibration. Every step after this depends on getting the mechanism right; a countermeasure aimed at the wrong one wastes money.
- Protect the oil film. Right viscosity for the load and speed, right quantity, clean and dry. This one move attacks adhesive, abrasive, and corrosive wear at once, because all three are held off by keeping surfaces apart and contaminants out.
- Control contamination. Filtration, sealing, breathers, and clean handling. Abrasive three-body wear is almost entirely a contamination problem, and dirt is the trigger for most wear cascades. Set target cleanliness and measure it.
- Match materials and surfaces to the load. Harden, coat, or nitride surfaces under high contact stress; improve the surface finish so asperities are lower to begin with. This is where you spend real money, so do it only after the film and cleanliness are handled.
- Monitor the survivors and close the loop. Trend vibration and wear-metal data on the assets that matter, feed confirmed failures back into a failure reporting system and move repeat offenders onto condition-based or predictive maintenance. Wear you can see coming is wear you can plan around.
What the numbers say
- Friction and wear are not a rounding error. A widely cited 2017 study in the journal Friction estimated that about 23% of the world's total energy use traces to tribological contacts, roughly 20% to overcome friction and 3% to remanufacture worn parts and replace equipment lost to wear (Holmberg & Erdemir, Influence of tribology on global energy consumption, costs and emissions Friction 5(3), 2017). Wear is a national-scale cost that shows up one bearing at a time.
- Catching wear before it becomes a breakdown pays. The U.S. Department of Energy's FEMP O&M guidance, maintained by PNNL, reports that condition-based programs save 8–12% over preventive-only maintenance, with the opportunity versus reactive operation reaching 30–40% (PNNL, O&M Best Practices: Maintenance Approaches). Reading wear early is how those savings materialize.
Wear is slow, quiet, and completely legible if you know the four fingerprints. Name the mechanism, protect the film, control the dirt, and monitor what is left, and most wear moves from the failure column to the plan column. To see how one plant got trustworthy floor and equipment data to work problems like these, read the CLS case study or explore how Harmony connects machine signals into one real-time layer on the features overview.