1. Quick Summary
Fatigue is failure under repeated loading. A part can carry a load comfortably if it is applied once, and still break after that same load is applied and removed enough times. Nothing about the material has degraded globally; the damage is local and cumulative.
It proceeds in stages. A crack initiates at a stress concentration, grows a tiny amount with each load cycle, and eventually reaches a size where the remaining cross-section cannot support the load and fails in a single cycle.
The final break looks sudden, which is why the term is confusing. The component did not weaken gradually in any way an operator could feel — the crack reached a critical size and the rest happened fast.
Because the early stages are invisible without inspection, managing fatigue means designing against crack initiation and scheduling inspection against crack growth, rather than waiting for symptoms.
2. What It Means
Stress concentration is the practical driver. A sharp corner, a hole, a scratch, a weld defect or a corrosion pit multiplies local stress well above the nominal average. Cracks start where geometry or surface condition concentrates load.
Crack growth per cycle is often extremely small — fractions of a micrometre — which is why fatigue life can run to millions of cycles and why the damage is undetectable until late.
3. Why It Happens
The mean stress matters as well as the alternating part. A component under steady tension with a small fluctuation behaves worse than one cycled symmetrically about zero, because the tensile mean holds the crack faces open and lets it advance.
Environment accelerates it dramatically. Corrosion fatigue combines chemical attack with cyclic loading, and the two reinforce each other: corrosion creates pits that act as crack starters, and cracking exposes fresh surface to further attack.
Manufacturing and service damage dominate real-world cases. Machining marks, inclusions in the metal, improper welds, fretting between contacting parts, and maintenance dents are far more often the origin than the bulk material’s inherent strength.
Inspection is probabilistic. Non-destructive testing methods have detection thresholds, so a crack can exist below the size the inspection can reliably find. Inspection intervals are set so that a missed crack cannot grow to critical size before the next check.
4. Real Examples
Aircraft structures: designed on a damage-tolerance basis, assuming cracks exist and setting inspection intervals from measured growth rates.
Rotating machinery: shafts and turbine blades see enormous cycle counts, so small stress fluctuations accumulate quickly.
Bridges and cranes: welded details are common fatigue origins, and traffic loading provides the cycles.
Medical implants: expected to survive tens of millions of cycles in a corrosive environment, which is a demanding combination.
5. How It Affects Us
For engineers, the useful levers are geometry and surface condition — generous radii, smooth finishes, careful weld detailing — which often buy more life than a stronger alloy does.
For operators, load history matters. Usage beyond the design assumption shortens life faster than intuition suggests, because cycle count is what counts.
For maintenance programmes, the key question is whether inspection intervals are derived from crack growth rates or merely inherited from convention.
For anyone assessing a failure, the fracture surface usually tells the story: fatigue leaves characteristic marks that distinguish it from a single overload break.
6. Key Takeaways
- Fatigue is cumulative damage from repeated loading, not gradual weakening of the whole part.
- Cracks start at stress concentrations — geometric, surface or corrosion-related — and grow invisibly.
- Steady tensile load plus fluctuation is worse than symmetric cycling because it holds cracks open.
- Corrosion and cyclic load reinforce each other, which is why the combination is especially damaging.
- Management relies on design against initiation and inspection intervals set from crack growth rates.