A crack only fractions of a millimetre wide can materially reduce remaining service life when it occurs at a welded connection, reinforcing-bar detail, bearing component, or pavement layer under repeated loading. Fatigue is progressive damage caused by stress cycles, often at stresses below a material’s short-term static strength. The key questions are not only the size of the load, but how often it occurs, where stress concentrates, and whether environmental exposure speeds crack initiation or growth.
In transport infrastructure, fatigue is often associated with bridge decks and steel details, yet it also affects reinforced and prestressed concrete, asphalt layers, rail components, fasteners, foundations, and tunnel fittings. A structure may have adequate capacity under a one-time proof load while still accumulating damage under traffic, vibration, wind, thermal movement, wave action, or machinery-induced cycles.
What fatigue damage means in structural materials
Fatigue generally develops through three overlapping stages: crack initiation, stable crack propagation, and final fracture or loss of function. Cracks usually begin at local discontinuities rather than in an ideal smooth section. Weld toes, bolt holes, abrupt changes in geometry, corrosion pits, surface defects, poorly consolidated concrete, and material interfaces all increase local stress and can become likely initiation points.
Once a crack has formed, each stress cycle can extend it by a very small amount. Growth may remain slow through much of a component’s life, then accelerate as the uncracked section reduces and stress intensity at the crack tip rises. The absence of visible deformation therefore does not rule out fatigue risk.
Fatigue performance is commonly expressed through an S–N relationship, which relates stress range to cycles to failure under defined test conditions. Stress range is the difference between maximum and minimum stress in a cycle. Infrastructure is rarely exposed to constant-amplitude loading, so design and assessment need to consider variable traffic spectra, occasional heavy loads, load sequence, dynamic amplification, and realistic boundary conditions.
Where fatigue concentrates in civil infrastructure
Steel bridges and ancillary steelwork
Steel is sensitive to detail quality because welded and bolted assemblies contain geometric discontinuities. Fatigue-prone locations commonly include welded stiffener ends, diaphragm connections, cover-plate terminations, transverse welds, gusset plates, attachment welds, orthotropic deck details, and bearing restraints. Nominal member stress can appear modest while local stress at a weld toe is considerably higher.
Traffic produces repeated bending and distortion cycles. In some details, fatigue is driven less by primary girder bending than by local out-of-plane deformation of plates or secondary members. Inspection planning should therefore include a detail-level review of load paths rather than relying only on global structural calculations.
Reinforced and prestressed concrete
Concrete fatigue involves repeated compression, tension, and shear, while reinforcement fatigue depends on cyclic stress range in bars or strands. Repeated wheel loading can gradually widen flexural cracks in decks and slabs. Where cracks admit chlorides or moisture, corrosion may reduce reinforcement area and create pitting, sharply increasing local stress concentration. This combined deterioration can be more significant than mechanical fatigue alone.
Prestressed members require close attention at tendons, anchorages, deviators, ducts, and areas of local cracking that may alter stress distribution. Assessment should account for prestress loss, corrosion condition, actual load history, and the possibility that damaged protective systems have changed the exposure environment.
Pavements and rail systems
In flexible pavements, repeated axle loading creates tensile strain near the bottom of asphalt layers and can initiate bottom-up cracking. Surface-initiated cracking may occur where tyre contact stress, binder aging, temperature effects, and weak interfaces govern behaviour. In rigid pavements, repeated slab bending can contribute to cracking, joint deterioration, and faulting, particularly where support is non-uniform.
Rail fatigue can affect rails, welds, switches, clips, fasteners, and supporting steelwork. Wheel–rail contact produces high local stresses over a large number of cycles. Defects may be linked to rolling contact fatigue, residual stress, poor weld quality, inadequate support, or loading amplified by geometry irregularities. Ground settlement can worsen these effects; settlement monitoring for urban transport infrastructure is therefore relevant where track or approach slabs are sensitive to differential movement.
Factors that accelerate fatigue
- Stress concentration: Sharp corners, attachment terminations, holes, weld profiles, notches, and corrosion pits increase local cyclic stress.
- High stress range: A large difference between peak and minimum stress generally causes damage to accumulate faster than the same number of lower-range cycles.
- Cycle count: Frequent loading can be significant even when individual vehicles or events are not exceptional.
- Dynamic effects: Impact, vibration, resonance, uneven road surfaces, rail irregularities, and loose components can amplify local response.
- Corrosion and moisture: Corrosion fatigue can trigger cracking earlier and increase crack growth through pitting, section loss, and deteriorated protective systems.
- Temperature and restraint: Thermal cycles add stress where expansion, contraction, or shrinkage is restrained.
- Construction and repair quality: Weld defects, poor surface preparation, unsuitable welding sequence, and incompatible repairs can introduce residual stress or create new fatigue-critical details.
The 1940 failure of the Tacoma Narrows Bridge is widely cited because repeated dynamic action and aerodynamic instability overwhelmed assumptions based on static strength. the documented history of the original Tacoma Narrows Bridge illustrates why dynamic behaviour must be treated as a structural design and operational concern, not a secondary effect.
Assessing fatigue: from load history to remaining life
A credible fatigue assessment begins with the actual structural detail and the service actions acting on it. Engineers establish relevant load paths, support conditions, geometry, material properties, inspection findings, and environmental exposure, then estimate or measure stress ranges at fatigue-sensitive locations.
For variable-amplitude loading, cumulative-damage methods are often used. One common screening method divides the observed or predicted number of cycles at each stress range by the allowable cycles for that range, then sums the resulting damage fractions. This can be useful, but its result depends on appropriate fatigue data, accurate stress estimates, suitable treatment of load sequence, and conservative interpretation where uncertainty is high.
| Assessment input | Why it matters | Typical evidence |
|---|---|---|
| Traffic or operational spectrum | Defines cycle frequency and load range | Weigh-in-motion data, axle counts, train records, operating logs |
| Detail geometry | Controls stress concentration and crack location | Drawings, laser survey, close visual inspection |
| Material and connection condition | Identifies defects, deterioration, and section loss | Material records, corrosion mapping, NDT results |
| Actual structural response | Checks whether analytical assumptions match service behavior | Strain gauges, accelerometers, displacement monitoring |
| Crack dimensions and orientation | Supports crack-growth and fracture assessment | Ultrasonic testing, magnetic particle testing, dye penetrant testing |
Where a crack has been identified, a fracture-mechanics assessment can be more informative than an S–N approach alone. It examines whether the detected crack could grow to a critical size under anticipated stress cycles. The work requires reliable information on crack shape, material toughness, residual stress, and actual loading. It does not replace inspection: the result is only as reliable as the defect data used in the model.
Inspection and monitoring strategies
Visual inspection remains essential because it can reveal coating failure, rust staining, deformation, water paths, loose fasteners, spalling, and crack patterns that guide targeted testing. Many fatigue cracks, however, begin in inaccessible areas or remain too small to confirm visually. Inspection methods should be chosen for the material and the expected orientation of the defect.
- Magnetic particle testing is useful for surface and near-surface discontinuities in ferromagnetic steel.
- Dye penetrant testing can reveal surface-breaking cracks on suitably prepared, non-porous surfaces.
- Ultrasonic testing can identify and size internal or subsurface flaws, although access, geometry, and operator capability affect reliability.
- Eddy-current methods are useful for near-surface flaws in conductive materials and in selected rail applications.
- Strain and vibration monitoring can identify changes in demand, local behaviour, or structural response, but cannot directly determine crack size unless supported by a validated interpretation model.
Monitoring works best when it addresses a specific engineering question: whether a known crack is growing, whether changed traffic has increased stress range, or whether a repaired detail is performing as intended. Continuous data collection without thresholds, baseline measurements, assigned maintenance responsibility, and review procedures can produce large volumes of data without improving safety decisions.
Repair decisions and fatigue-resistant detailing
Repair must address the mechanism rather than conceal the visible symptom. Grinding out a crack and repainting the surface may achieve little if the crack arose from continuing distortion, an unresolved stress concentration, poor drainage, or an overloaded connection. A welded reinforcement can also introduce new fatigue-sensitive weld terminations if its geometry and installation sequence are not properly engineered.
Interventions may include removing or arresting a crack where appropriate, modifying geometry to reduce stress concentration, improving load distribution, restoring section loss, replacing damaged components, and controlling water or corrosive contaminants. Temporary restrictions may be justified while the cause and growth rate are evaluated, but their effectiveness should be checked against the actual critical load effect.
Failure investigations are most useful when they connect observed symptoms to underlying mechanisms rather than attributing an event to one superficial cause. The blog’s guide to finding root causes in transport infrastructure failure case studies explains the value of combining records, inspection evidence, loading history, and competing hypotheses before corrective work is selected.
Managing uncertainty in service-life decisions
Fatigue assessments involve uncertainty in future traffic, material variability, residual stress, hidden defects, and the representativeness of monitoring data. Conservative assumptions may be justified for inaccessible critical details, but they should be explicit and traceable. A practical asset-management record identifies each fatigue-sensitive detail, its inspection method, condition state, expected loading, consequence of failure, and trigger for reassessment.
For a detected crack, a useful field record includes its precise location referenced to a durable datum, length and orientation, surface condition, inspection method, images at a consistent scale, date, any load restrictions, and the next inspection interval. Repeating the same measurement method at the same location often provides a clearer indication of growth than a single more sophisticated reading with no comparable baseline.
