Share this article
Continue reading with these related posts
Fatigue failures do not announce themselves. A shaft carries its rated load for months, then fractures at a load it has handled thousands of times before. The failure looks sudden. The process was not — it began as a microscopic crack at a stress concentration, growing incrementally with each cycle until the remaining section could no longer carry the load.
Corrosion is not a single mechanism — it is a family of electrochemical attack processes, each with distinct visual signatures, distinct causes, and distinct prevention strategies. A pit that looks like surface roughness can initiate a fatigue crack at a third of the expected stress. A weld that looks structurally sound can be corroding from the inside along the heat-affected zone.
The fracture surface of a failed part is a record of how the failure progressed. Beach marks, chevron patterns, shear lips, and ratchet marks each tell a specific story about load type, crack origin, and propagation rate. Learning to read these features is the most cost-effective skill in failure analysis.
Wear is the progressive removal of material from a surface through mechanical contact. Unlike fracture, which is sudden, or corrosion, which is chemical, wear is gradual and mechanical — but it is no less serious for being slow. A bearing that loses 0.05 mm of material per thousand hours of operation will eventually reach a clearance that allows vibration, misalignment, or seizure. A gear tooth worn past its contact geometry will transmit load incorrectly and fail in bending or surface fatigue long before its nominal design life.
The engineering challenge is that wear is not a single mechanism. Abrasive wear, adhesive wear, and surface fatigue all remove material, but they do so through completely different physical processes, leave completely different surface evidence, and respond to completely different prevention strategies. Applying the right prevention to the wrong mechanism produces no improvement. Understanding which mechanism is operating — from the surface evidence — is the first step toward effective design.
Wear occurs at the interface between two surfaces in relative motion. At the microscopic level, real surfaces are not smooth — they consist of asperities, small peaks and valleys, whose heights are measured in microns. When two surfaces contact each other under load, the true contact area is a small fraction of the apparent contact area, and the local contact stress at asperity peaks can be many times the nominal contact pressure.
The wear mechanism that results depends on four factors: the surface hardness and material properties of both bodies, the contact stress, the relative motion (sliding, rolling, oscillating), and the presence or absence of lubricant and abrasive particles. Each combination of these factors favours a different wear mechanism, and in service, multiple mechanisms often operate simultaneously.
The Archard Wear Equation
The simplest model of wear is the Archard equation: wear volume = K × (normal force × sliding distance) / hardness. The wear coefficient K captures the mechanism and material pairing. Abrasive wear has a K roughly ten to one thousand times higher than well-lubricated adhesive wear. This is why hardness alone does not prevent wear — if the mechanism is wrong, hardness improvement produces marginal benefit. Reducing K by changing the mechanism produces orders-of-magnitude improvement.
Abrasive wear occurs when hard particles or hard surface asperities cut, plough, or scratch the softer surface they contact. It is the most common wear mechanism in equipment exposed to soil, sand, mineral slurry, or contaminated lubricant.
Two sub-types are important to distinguish. Two-body abrasion occurs when a hard surface asperity on one component scratches the opposing surface directly — like a file cutting metal. Three-body abrasion occurs when hard particles become trapped between two surfaces and are dragged across them, cutting both simultaneously. Three-body abrasion from abrasive particles in lubricant is responsible for a significant proportion of premature bearing and gear wear in equipment that operates in dusty or contaminated environments.
Recognition: parallel scratch marks on the worn surface, aligned with the direction of relative motion. Under a 10× loupe, the scratches have a consistent directionality, with ridges of displaced material along the scratch flanks. The surface has a roughened, matte texture compared to an unworn surface. In severe abrasion, material may be removed in chips visible to the naked eye.
The hardness ratio between the abrasive particle and the worn surface is the controlling parameter. If the abrasive is harder than the surface by a factor of 1.2 or more, significant abrasive wear occurs. If the surface is harder than the abrasive, wear rate drops sharply. This is why hardening the worn surface — through case hardening, nitriding, or hard facing — is the primary design response to abrasive wear.
Lubricant Contamination Converts Adhesive to Abrasive Wear
A machine designed with adequate lubrication for adhesive wear resistance can suffer catastrophic abrasive wear if the lubricant becomes contaminated with hard particles — metal fines, dirt, or wear debris. Contamination from a single ingress event can degrade a bearing from a projected life of 20,000 hours to a few hundred. Contamination control — sealing, filtration, and scheduled oil sampling — is as important as lubricant selection for wear prevention in contaminated environments.
Adhesive wear occurs when the surfaces of two materials come into direct contact, form adhesive junctions at asperity contact points, and then rupture those junctions as the surfaces continue to move. Material is transferred from the softer surface to the harder surface, or torn from both surfaces simultaneously. The transferred material forms lumps or flakes that can become abrasive particles, accelerating wear further.
At low severity, adhesive wear produces a polished, burnished surface appearance. At moderate severity, it produces smearing — patches of transferred material on the surface, visible as bright metallic deposits on the opposing geometry. At high severity, it produces galling: gross material transfer with deep tears, surface welding, and in the worst case, seizure where the two surfaces fuse and relative motion stops entirely.
Recognition: transferred material on the wear surface (metallic deposits that are brighter or different in colour than the base material), surface tearing rather than scratching, and a directionality consistent with the sliding direction. Galling shows larger-scale tears and lumps, often with a rough, torn texture rather than the smooth scratched texture of abrasive wear.
The material pairing governs susceptibility. Same-material pairings — steel on steel, aluminium on aluminium — have high adhesion tendency because the materials share similar atomic structure and bond readily under contact stress. Dissimilar-material pairings with different crystal structures (steel on bronze, steel on PTFE composite) have much lower adhesion tendency. This is why plain bearings use different materials for the shaft and the bearing insert, and why coating one surface with a low-adhesion material (PTFE, DLC, molybdenum disulphide) reduces adhesive wear dramatically.
Recognise Galling Early — It Escalates Quickly
Galling begins as localised adhesive wear at a few asperity contacts and progresses rapidly if not addressed. The transferred material increases the local contact stress at adjacent locations, spreading the damage. A component showing early galling — small bright metallic deposits, slight surface roughening — will progress to gross material transfer and potential seizure within hours to days under continued operation at the same conditions. Early detection during inspection is the intervention point; operating through early galling signs leads to component destruction.
Surface fatigue occurs in rolling contact — in gears, rolling element bearings, cam followers, and rail-wheel contact. Unlike sliding wear, which produces continuous material removal, surface fatigue produces discrete damage events: pits and spalls that form when subsurface cracks propagate to the surface and release material fragments.
The mechanism is subsurface crack initiation driven by the cyclic Hertzian contact stress. At each rolling contact event, a small volume of material just below the surface experiences a complex stress cycle, including shear stresses that are maximum at a depth of approximately 0.47 × contact half-width. Over many cycles, fatigue cracks initiate at this depth (at inclusions, carbides, or other microstructural stress concentrations) and grow until they reach the surface, releasing a spall fragment.
Recognition: pits and spalls on the rolling surface. Early pitting produces small, shallow depressions with smooth floors and sharp boundaries. Advanced pitting (spalling) produces larger, deeper craters with rough, jagged walls, and in severe cases, complete removal of the hardened case layer on a case-hardened component. In bearings, spalling produces a characteristic rumbling or vibration change detectable by vibration monitoring.
The size and distribution of pits communicates information about the failure stage. Fine, distributed pitting on a gear tooth contact band indicates early surface fatigue from inadequate lubricant film thickness. Large, deep spalls localised to a specific zone indicate subsurface defects, case hardening deficiencies, or localised overload from misalignment.
Oil Film Thickness Controls Surface Fatigue Life
Surface fatigue life in rolling contact is not purely a function of load — it is strongly controlled by the lubricant film parameter lambda (λ), the ratio of oil film thickness to composite surface roughness. When lambda is below 1, asperity contact is frequent and surface fatigue life is reduced by orders of magnitude from the theoretical calculated life. Specifying the correct lubricant viscosity for the operating speed and temperature to achieve lambda above 2 is as important as the contact stress calculation in rolling contact design.
Fretting wear is a distinct wear mechanism that occurs at the interface between two surfaces under load when small-amplitude oscillatory sliding motion exists between them. "Small-amplitude" means motion in the range of 1 to 300 micrometres — far smaller than the relative motion in conventional sliding wear. Fretting occurs at press-fit interfaces, clamped joints, bolted connections, and any location where vibration-induced micro-motion exists between clamped surfaces.
The mechanism combines adhesive wear, abrasive wear, and oxidation. Adhesive wear at asperity contacts produces fine metallic particles. These particles oxidise rapidly, becoming hard, abrasive oxide debris. The abrasive debris is trapped at the interface under the clamping load and continues to abrade both surfaces, producing a characteristic reddish-brown iron oxide powder at steel interfaces and black oxide powder at aluminium interfaces.
Recognition: reddish-brown or black powdery debris at the interface, surface pitting and grooving confined to the contact zone, and a reduction in clamp load or interference fit interference as material is removed. The debris colour and texture immediately identifies fretting — no other wear mechanism produces the same fine, coloured oxide powder in quantity at a clamped joint.
Prevention focuses on eliminating the micro-motion. Options include increasing the interface normal force (higher clamping torque, greater interference fit), changing the surface condition (shot peening, friction-increasing coatings, serrated surfaces), or providing a fretting-resistant surface treatment (molybdenum disulphide coating, electroless nickel). In applications where micro-motion cannot be eliminated, using a sacrificial fretting-resistant insert at the interface limits damage to the replaceable component.
A conveyor system used in a mineral processing plant was experiencing accelerated wear of the head pulley lagging within six months of service, compared to an expected service life of eighteen months. The lagging was 20 mm thick natural rubber bonded to the steel pulley shell.
Visual examination of the worn lagging revealed parallel scratch marks aligned with the belt travel direction, consistent ridging and grooving across the full face of the lagging, and a roughened surface texture with material removed in strips. No pitting, no transferred material, and no discolouration from heat were observed.
The scratch pattern and surface morphology identified the mechanism as three-body abrasive wear, not adhesive wear or surface fatigue. The abrasive agent was the material being conveyed — fine mineral particles entraining between the belt and the lagging face.
Operating condition review revealed that the lagging had been specified for a clean conveyor carrying dry limestone. The application had changed twelve months prior to use the conveyor for wet iron ore fines with a high percentage of particles under 1 mm — a significantly more abrasive service than the original design basis.
The diagnosis was material selection mismatch: the lagging specification was not updated when the conveyed material changed. Natural rubber has a Shore A hardness of approximately 60–70, suitable for impact loading but relatively soft against hard abrasives. The iron ore fines, predominantly haematite with Mohs hardness 5.5–6.5, were harder than the lagging by a wide margin.
Corrective action: replace the lagging with ceramic-reinforced polyurethane (Shore D 60), which has a hardness approximately three times that of natural rubber and substantially higher abrasion resistance against hard mineral particles. Revised life estimate: twenty-four months. First replacement cycle confirmed: twenty-two months of service before first lagging change.
For abrasive wear: increase surface hardness above the hardness of the abrasive particles. Case hardening, nitriding, hard chromium plating, and tungsten carbide thermal spray are common approaches. Hardness alone is not sufficient if the application involves gouging rather than scratching — ductility and toughness matter in high-impact abrasive service. Contamination control (sealing, filtration) prevents three-body abrasion in lubricated systems.
For adhesive wear: use dissimilar material pairings at sliding contacts. Specify adequate lubrication to maintain a separating film between surfaces. Apply anti-galling coatings (molybdenum disulphide, DLC, PTFE composite) where direct metal contact is unavoidable. Reduce surface roughness at sliding interfaces — a smoother surface has fewer asperity contacts and lower adhesion tendency.
For surface fatigue: control the lambda ratio through lubricant viscosity selection. Specify the surface hardness and case depth appropriate for the contact stress. Control subsurface cleanliness through material specification — bearing quality steel with controlled inclusion content has higher surface fatigue life than standard structural steel. Minimise misalignment, which concentrates contact stress at one end of the contact zone.
For fretting: increase interface normal force, reduce relative motion, or separate the surfaces with a fretting-resistant coating or insert. Peening the interface surfaces introduces compressive residual stress that delays fretting crack initiation.
Match Prevention to Mechanism
The most common error in wear prevention is applying the wrong solution: specifying harder material for a surface fatigue problem (hardness does not address oil film thickness), or improving lubrication for a three-body abrasive wear problem (lubricant does not remove abrasive particles already entrained). Diagnosing the wear mechanism first, from the surface evidence, takes ten minutes. Applying the right solution to the right mechanism typically produces a tenfold or greater improvement in wear life.
Wear mechanisms remove material gradually over many operating cycles. The next post examines a different failure mode at the opposite end of the time scale: overload and impact failure, where a single load event or a small number of events exceeds the material's strength. Reading the fracture surface features of overload failures provides immediate information about the load level, material condition, and whether the failure was expected or premature.
