What Is Fatigue? Causes, Failure, and Testing
Representative Infinita Engineering Visual explaining the four-step workflow for What Is Fatigue? Causes, Failure, and Testing.What Is Material Fatigue?
Material fatigue is the progressive structural damage that occurs when a material is subjected to repeated cyclic loading below its static ultimate strength. A component that can withstand a single application of a given load may fail after thousands or millions of applications of the same load – or even a much smaller load. Fatigue is responsible for the majority of mechanical failures in service and is the governing failure mode for rotating machinery, vehicle structures, aircraft components, pressure vessels under cyclic pressure, and any part subject to repeated stress.
Fatigue failure is insidious because it occurs without visible warning – a fatigue crack can propagate through a cross-section while the component continues to function normally, until the remaining cross-section is too small to carry the load and fracture occurs suddenly. The fracture surface shows distinctive beach marks and ratchet marks that identify the origin and propagation of the fatigue crack, distinguishing fatigue failure from overload fracture in failure analysis.
The Fatigue Process
Crack Initiation
Fatigue cracks initiate at stress concentrations – notches, holes, surface scratches, inclusions, weld toes, thread roots, or any geometric feature that locally amplifies stress. In smooth specimens without stress concentrations, cracks initiate at persistent slip bands on the surface where cyclic plastic deformation accumulates. The number of cycles to crack initiation depends on the stress amplitude, the stress concentration factor, the surface finish, and the material’s resistance to cyclic plastic deformation. High surface finish and compressive residual stresses (from shot peening or surface rolling) extend crack initiation life significantly.
Also Read – Mechanical Engineering Testing Standards: ASTM, ISO & ASME Reference List
Crack Propagation
Once initiated, a fatigue crack propagates incrementally with each load cycle. The crack growth rate (da/dN) is governed by the stress intensity factor range (delta K), described by the Paris law: da/dN = C*(delta K)^m, where C and m are material constants. This stage produces the beach marks visible on fatigue fracture surfaces – they are the crack front at successive positions during propagation. Propagation life is the domain of fracture mechanics analysis. Variables that affect propagation rate include mean stress (R-ratio), environment (corrosion fatigue), and temperature.
Final Fracture
When the fatigue crack has grown to the point where the remaining cross-section cannot support the applied load, final fracture occurs. In ductile materials, the final fracture zone shows dimpled rupture morphology by SEM. The relative areas of the fatigue propagation zone and the final fracture zone on the fracture surface reveal the stress level – a small final fracture zone relative to the total cross-section indicates the crack grew under low cyclic stress; a large final fracture zone indicates high stress. This analysis is a standard part of failure investigation.
Fatigue Testing Methods
Stress-Life (S-N) Testing
Stress-life (S-N) testing, also called the Wohler test, subjects specimens to fully reversed or constant-amplitude cyclic stress at various stress levels and records the number of cycles to failure (N) at each stress level (S). The data is plotted as an S-N curve. For steels, there is a stress level (the fatigue limit or endurance limit) below which failure does not occur within 10^7 cycles. For aluminium and most non-ferrous metals, there is no true fatigue limit – the S-N curve continues to decrease. ASTM E466 covers S-N testing of metallic materials under constant-amplitude axial loading.
Strain-Life (e-N) Testing
Strain-life testing controls strain amplitude rather than stress, which is more relevant for low-cycle fatigue where plastic strain is significant. The specimen is cycled at constant strain amplitude, and the number of cycles to crack initiation is recorded. The Coffin-Manson relationship describes the plastic strain amplitude versus cycles to failure. Strain-life data is used for automotive component fatigue analysis where local plastic strain at notches drives fatigue life. ASTM E606 covers strain-controlled fatigue testing.
Fatigue Crack Growth Rate Testing
Fracture mechanics-based fatigue testing measures crack growth rate (da/dN) as a function of stress intensity factor range (delta K) using compact tension (CT) or middle-tension (MT) specimens with a machined notch and fatigue pre-crack. The test produces a Paris law curve from threshold (delta K_th below which cracks do not grow) through stable propagation to fracture. ASTM E647 covers fatigue crack growth rate measurement. The data is used to predict remaining life of structures with known crack sizes and to set inspection intervals.
Industry Specifications
- Stress-Life Testing: ASTM E466 (constant amplitude axial), ASTM E468 (presentation of S-N data), ASTM E739 (statistical analysis)
- Strain-Life Testing: ASTM E606 (strain-controlled fatigue), ASTM E1012 (verification of test alignment)
- Crack Growth Rate: ASTM E647 (fatigue crack growth), ASTM E399 (fracture toughness, related)
- Rotating Bending: ASTM E468, ISO 1143 (rotating bar bending fatigue)
- Automotive: SAE J1099 (technical report on fatigue properties), SAE J2816
- Aerospace: MIL-HDBK-5 (fatigue data for aerospace alloys), FAR 25.571 (damage tolerance)
Also Read – Metallic Materials Testing, Explained
Conclusion
Fatigue is the most common cause of mechanical failure in service, yet it is preventable through material selection, design for low stress concentration, surface treatment, and fatigue testing during product development. S-N testing establishes the stress level for infinite life. Strain-life testing characterises low-cycle fatigue at notches. Crack growth testing provides propagation rate data for damage-tolerant design. Fracture surface analysis identifies the failure mechanism after the fact. Each method addresses a different stage or aspect of the fatigue process, and together they provide the data needed to design against fatigue failure.
What is the fatigue limit and which materials have one? The fatigue limit (endurance limit) is the stress amplitude below which a material can endure an infinite number of cycles without fatigue failure. Steels and some titanium alloys exhibit a true fatigue limit - the S-N curve becomes horizontal at approximately 10^6 to 10^7 cycles. Aluminum alloys, copper alloys, and most non-ferrous metals do not have a true fatigue limit - the S-N curve continues to decrease slowly even at very high cycle counts. For these materials, a fatigue strength at a defined life (10^7 or 10^8 cycles) is reported instead of a true fatigue limit.
How does surface condition affect fatigue life? Surface condition has a dominant effect on fatigue life because fatigue cracks initiate at the surface. A polished surface has the highest fatigue strength; machined surfaces are lower; forged or as-cast surfaces are lower still; surfaces with corrosion pits or grinding burns are significantly lower. Surface treatments that introduce compressive residual stress - shot peening, laser peening, cold rolling, nitriding - extend fatigue life by impeding crack initiation and early growth. A single machining scratch oriented transverse to the cyclic stress can reduce fatigue life by a factor of 2 to 5.
What is the difference between high-cycle and low-cycle fatigue? High-cycle fatigue (HCF) involves stress levels below the yield strength where deformation is nominally elastic, and failure occurs after 10^4 to 10^8 or more cycles. Rotating machinery components, springs, and fasteners typically operate in this regime. Low-cycle fatigue (LCF) involves stress levels at or above yield where significant cyclic plastic strain accumulates, and failure occurs in fewer than 10^4 cycles. Pressure vessels under repeated pressurization, engine components, and structures subject to large infrequent overloads operate in the LCF regime. HCF is analyzed by stress-life methods; LCF by strain-life methods.
What does the R-ratio mean in fatigue testing? The R-ratio is the ratio of minimum stress to maximum stress in a fatigue cycle: R = S_min / S_max. Fully reversed loading (equal tension and compression) has R = -1. Tension-tension loading with no compressive component has R = 0 to 1. Compressive mean stress (R < -1) extends fatigue life; tensile mean stress (R > 0, positive mean stress) reduces it. This is described by the Goodman, Gerber, or Soderberg mean stress correction diagrams. Most published fatigue data is for R = -1 or R = 0.1; fatigue data must be corrected for actual R-ratio in service if it differs.
How is fatigue failure identified on a fracture surface? Fatigue fracture surfaces have characteristic features visible at the macro scale: beach marks (also called clamshell marks) are arc-shaped lines centered on the crack origin, formed by changes in loading or brief crack arrest periods. Ratchet marks are ridges between multiple initiation sites that merge into a single crack front. The final fracture zone shows the rough, fibrous or crystalline appearance of overload fracture. At the SEM scale, fatigue striations - one per load cycle - are visible on the fatigue propagation zone. The combination of beach marks, ratchet marks, initiation site, propagation zone, and final fracture zone uniquely identifies fatigue as the failure mechanism.
Other Categories
