Fatigue Fundamentals

Technical provenance

Applicable standards / specifications

• BS 7608 (2014+A1:2015) — Guide to fatigue design and assessment of steel products — Use only where the material/detail and project basis fall within scope.

References

• Schijve, J. — Fatigue of Structures and Materials — Background reference for fatigue mechanisms, S–N behaviour, mean-stress effects and variable-amplitude loading.
• BS 7608 — Guide to fatigue design and assessment of steel products — Applicable principally to fatigue assessment of steel products and welded details.

What Is It?

Fatigue is progressive structural damage caused by repeated loading, often at stress levels well below the static strength of the material. A single application of the load may produce no visible damage, but after thousands or millions of repetitions, a crack can initiate, grow and ultimately cause fracture. Fatigue is the dominant failure mechanism in many engineering structures — aircraft, vehicles, machinery, bridges and consumer products alike.

Why It Matters

A structure that is demonstrably adequate for its maximum static load can still fail in service if that load, or smaller loads, are applied repeatedly. Fatigue failures often occur without warning — there is no permanent set, no visible distortion and no progressive loss of stiffness until a crack has grown to a significant fraction of the load-bearing section. Understanding fatigue is essential for any structure that experiences cyclic loading during its service life.

Static pass does not mean fatigue pass. A structure can withstand its maximum load once and still fail after that load has been repeated many thousands or millions of times.

Physical Mechanism

Fatigue develops in three broad stages. First, microscopic damage accumulates at a highly stressed location — typically a surface or a stress concentration. Persistent cyclic slip creates intrusions and extrusions at the grain level. Second, a crack initiates and begins to grow, initially along slip planes and then perpendicular to the principal tensile stress. Third, the remaining cross-section can no longer sustain the applied load and final fracture occurs — often a sudden, brittle-appearing overload of the remaining ligament.

Local cyclic stress → microscopic damage → crack initiation → crack growth → remaining section overload → fracture

Where Fatigue Starts

Fatigue cracks typically initiate at locations where local stress is elevated or where the surface condition is imperfect. Understanding where to look is as important as understanding the mechanism.

• Holes — fastener holes, access holes, drainage holes
• Notches — fillets, grooves, threads, keyways
• Surface defects — machining marks, scratches, nicks
• Weld toes — geometric and metallurgical discontinuities
• Fretting interfaces — clamped joints with micro-slip
• Manufacturing defects — inclusions, porosity, laps
• Corrosion pits — surface damage that concentrates stress

Key Parameters

Fatigue is governed by the stress cycle — the variation of stress between maximum and minimum values. Several quantities are used to characterise a cycle, and each has physical significance.

Stress range:       Δσ  =  σ_max  −  σ_min
Stress amplitude:   σ_a  =  (σ_max − σ_min) / 2
Mean stress:        σ_m  =  (σ_max + σ_min) / 2
Stress ratio:       R    =  σ_min / σ_max

where:
σ_max  =  maximum stress in the cycle
σ_min  =  minimum stress in the cycle

High-Cycle vs Low-Cycle Fatigue

Fatigue is broadly divided into two regimes. In high-cycle fatigue (HCF), stresses are nominally elastic and lives are long — hundreds of thousands to millions of cycles. The S-N (stress-life) method is typically used. In low-cycle fatigue (LCF), local plastic strain occurs on each cycle and lives are short — hundreds to tens of thousands of cycles. The strain-life (ε-N) method is more appropriate. The boundary between the two regimes is not a fixed number — it depends on the material, geometry and stress level.

Static Strength vs Fatigue

Static strength and fatigue life answer fundamentally different engineering questions. Static strength asks: can the structure carry the maximum load once? Fatigue asks: can the structure carry repeated loads for the required number of cycles? A positive static margin does not imply adequate fatigue life. A positive fatigue margin does not imply adequate static strength. Both must be assessed separately.

Common Mistakes

• Treating a positive static margin as evidence of adequate fatigue life
• Using the maximum stress from a single load case as a fatigue input without considering the full load history
• Assuming that because the stress is below yield, fatigue is not a concern
• Applying a single generic "fatigue factor" to a static stress without understanding the cyclic loading
• Confusing high-cycle and low-cycle fatigue regimes and applying the wrong method

Uncertainty

Fatigue is inherently scatter-prone. Identical specimens tested under identical conditions can fail at lives differing by factors of two to ten. This scatter arises from material variability, surface condition differences, microstructural variation and the stochastic nature of crack initiation. Fatigue analysis should acknowledge this scatter — typically by using design-allowable curves (e.g. A-basis or B-basis S-N data) rather than mean curves, and by considering the consequence of premature failure.

Key Takeaways

• Fatigue is progressive damage from repeated loading, often at stresses well below static strength
• Cracks initiate at stress concentrations and surface defects, then grow until the remaining section fractures
• Static strength and fatigue life answer different questions — both must be assessed
• High-cycle fatigue uses stress-life; low-cycle fatigue uses strain-life — the regimes are not interchangeable
• Fatigue is scatter-prone — design-allowable curves and consideration of variability are essential