Fatigue and fracture of engineering materials and structures in material selection
Why fatigue and fracture matter in material selection
Fatigue and fracture of engineering materials and structures is the practical study of how components crack, accumulate damage, and ultimately fail under service conditions. A material with high static strength can still be a poor choice if it contains defects, sees repeated loading, operates in a corrosive environment, or has low fracture toughness at the intended service temperature. For designers, buyers, and materials engineers, the question is not only “Which material is strongest?” It is “Which material, geometry, manufacturing route, and inspection plan can tolerate expected flaws and cyclic loading?”
That is why fatigue strength, crack growth rate, fracture toughness, surface condition, residual stress, and environmental exposure need to be reviewed together. For more material behavior topics, see our Properties section.
What the term covers
Fatigue and fracture are related, but they are not the same. Fatigue describes progressive damage caused by repeated or fluctuating loads. Fracture describes crack extension and separation when a material can no longer carry the applied stress. In service, the two often appear as one sequence: a small discontinuity forms or is already present, cyclic loading grows it, and final fracture occurs when the remaining section cannot resist the stress intensity at the crack tip.
The phrase also overlaps with the title of the peer-reviewed journal Fatigue & Fracture of Engineering Materials & Structures, which focuses on structural integrity, the mechanics of fatigue and fracture, and the reliability of materials and components. In an engineering content context, however, the phrase is broader. It covers metals, polymers, composites, ceramics, welded joints, additively manufactured parts, infrastructure materials, pressure equipment, transportation systems, and other structures where cracks can affect safety or service life.
Authoritative references commonly used in this field include ASTM fatigue and fracture standards, ISO fatigue data guidance, and ASM Handbook Volume 19, Fatigue and Fracture. These sources point to the same core principle: fatigue and fracture performance cannot be reduced to one material number without considering loading, geometry, defect population, environment, and statistical scatter.
How fatigue damage develops
Fatigue failure usually starts locally rather than uniformly across a part. Stress raisers such as notches, holes, weld toes, inclusions, machining marks, corrosion pits, pores, and sharp transitions can concentrate stress and create preferred sites for crack initiation. Even when the nominal stress is below the static yield strength, microscopic cyclic plasticity can occur at these local features.
Crack initiation
Crack initiation is the period in which microscopic damage accumulates until a physically detectable crack exists. In high-cycle fatigue, this stage can account for much of the total life of a smooth, well-manufactured specimen. In welded joints, castings, powder metallurgy parts, and some additively manufactured materials, pre-existing imperfections may shorten the initiation period because a flaw already behaves like a small crack.
Crack propagation
Once a crack is established, the engineering focus often shifts from initiation life to crack growth rate. Fatigue crack growth is commonly evaluated through the relationship between crack extension per cycle and the stress intensity factor range. ASTM E647 is a widely recognized method for measuring fatigue crack growth rates. The resulting data help engineers compare materials, evaluate damage-tolerant designs, and plan inspection intervals when combined with stress analysis, defect characterization, and fracture toughness data.
Final fracture
Final fracture occurs when a crack reaches a size at which the material and geometry can no longer resist rapid extension. A ductile material may show stable tearing before failure, while a brittle material, or a material operating under low-temperature or high-constraint conditions, may fail with little warning. This distinction matters in structures where leak-before-break behavior, visible deformation, or scheduled inspection is part of the safety concept.
Key properties and tests engineers compare
Fatigue and fracture assessment uses several data types. Each answers a different engineering question, and none should be treated as a universal measure of structural safety.
| Property or data type | What it helps answer | Typical reference framework | Important limitation |
|---|---|---|---|
| S-N fatigue data | How many cycles a specimen may survive at a given stress amplitude | Constant-amplitude fatigue testing such as ASTM E466, with presentation and statistical treatment guided by fatigue data practices | Specimen results may not represent full-size parts, variable amplitude loading, or severe environments |
| Fatigue crack growth rate | How quickly an existing crack may grow under cyclic loading | ASTM E647 for fatigue crack growth rate measurement | Requires representative stress ratio, environment, temperature, thickness, and crack geometry assumptions |
| Plane-strain fracture toughness | How resistant a metallic material is to fracture under linear-elastic, high-constraint conditions | ASTM E399 for KIC determination | Valid results require specimen size and constraint conditions; thin or highly ductile materials may need other approaches |
| Elastic-plastic fracture toughness | How a material resists crack extension when plasticity is significant | ASTM E1820 for K, J, and CTOD-based toughness evaluation | Specimen dimensions, crack extension measurement, and tearing behavior influence interpretation |
| Statistical fatigue analysis | How scatter, confidence, and test planning affect fatigue data use | ISO 12107 and related fatigue data planning approaches | Requires enough valid tests to support statistical claims |
The practical point is straightforward: fatigue life estimates depend on both the test method and the assumptions used to transfer specimen data to a component. A polished laboratory specimen, a welded bracket, and a corroded field component may all be made from the same alloy, yet show very different fatigue behavior.
Why strength alone is not enough
Material selection often starts with yield strength, tensile strength, hardness, density, cost, corrosion resistance, and availability. Those properties matter, but they do not fully describe crack-sensitive performance. Increasing strength can sometimes reduce fracture toughness or increase defect sensitivity, depending on alloy system, heat treatment, microstructure, and processing route. A stronger grade is therefore not automatically a safer fatigue choice.
Geometry can dominate the result. A sharp corner, undersized fillet radius, drilled hole, keyway, thread root, or weld toe may raise local stress enough to control fatigue life. This is why fatigue design commonly focuses on reducing stress concentration, improving surface finish, controlling weld profile, and avoiding abrupt load paths. Surface treatments such as polishing, shot peening, carburizing, nitriding, and coating can improve or harm performance depending on residual stress, surface integrity, and environmental compatibility.
Environment is another major variable. Corrosion fatigue, hydrogen embrittlement, stress corrosion cracking, fretting fatigue, oxidation, moisture ingress, and thermal cycling can all change crack initiation or growth behavior. For polymers and composites, temperature, moisture, ultraviolet exposure, matrix cracking, fiber-matrix interface behavior, and loading direction may be as important as nominal strength. For ceramics, flaw size distribution and fracture toughness dominate because brittle materials tolerate little plastic redistribution at crack tips.
Material groups respond differently
Metals remain the most documented class for fatigue and fracture design because standardized test methods and extensive historical data exist for steels, aluminum alloys, titanium alloys, nickel alloys, and other structural metals. Microstructure, inclusions, grain size, heat treatment, welding, forming, and residual stress can all change fatigue crack initiation and propagation.
Composites require a different approach. Instead of one crack growing through a uniform material, damage may include matrix cracking, fiber breakage, delamination, fiber pull-out, and impact-related defects. The direction of loading relative to fiber orientation strongly affects performance. A composite laminate that performs well in one loading direction may be vulnerable under out-of-plane impact, bearing loads, or moisture-assisted degradation.
Polymers and elastomers introduce time, temperature, and rate effects. Creep, stress relaxation, aging, and environmental exposure can interact with fatigue. For polymers used in moving parts, seals, medical devices, or electrical housings, fatigue design should account for service temperature range, chemical contact, and long-duration loading. See also: Application.
Ceramics and glassy materials are typically governed by flaw populations and low fracture toughness compared with ductile metals. They can offer hardness, wear resistance, high-temperature capability, or chemical stability, but design must control tensile stress and flaw size. Proof testing, surface quality, and compressive residual stress may be central to reliability.
Additively manufactured materials deserve special attention because build orientation, porosity, lack-of-fusion defects, surface roughness, heat treatment, and post-processing can significantly affect fatigue behavior. NIST and other research organizations have studied fatigue and fracture of additively manufactured metallic materials because conventional wrought-material assumptions may not transfer directly to printed components.
A practical workflow for comparing materials and structures
A useful fatigue and fracture review should move from general material properties to application-specific evidence. The following workflow helps avoid over-reliance on a single data sheet value.
- Define the loading spectrum. Identify whether the component sees constant amplitude, variable amplitude, impact, vibration, thermal cycling, pressure cycling, start-stop operation, or random service loads.
- Identify critical locations. Use drawings, stress analysis, inspection history, or finite element analysis to locate notches, holes, welds, transitions, interfaces, and contact zones.
- Separate initiation-controlled and crack-growth-controlled design. Smooth parts often require S-N or strain-life analysis, while damage-tolerant structures require crack growth and fracture toughness assessment.
- Choose representative test data. Match material form, heat treatment, surface condition, thickness, orientation, environment, and temperature as closely as possible.
- Check the standard behind the data. Data generated under ASTM E466, E647, E399, E1820, ISO 12107, or comparable frameworks are easier to audit than unsupported numbers.
- Account for scatter and reliability. Fatigue data naturally scatter. Design allowables, safety factors, inspection intervals, and statistical confidence should reflect the consequence of failure.
- Validate with component-level evidence when needed. Full-scale or subcomponent tests may be necessary where geometry, assembly, residual stress, or multiaxial loading cannot be represented by simple specimens.
This workflow is especially important for sectors such as aerospace, rail, automotive, wind energy, bridges, pressure equipment, medical implants, heavy machinery, and electronics hardware. In these applications, small cracks can become business-critical, safety-critical, or maintenance-critical long before static strength is exceeded.
Common mistakes in interpreting fatigue and fracture data
One common mistake is treating a fatigue limit or endurance strength as a fixed universal property. Some materials show an apparent endurance limit under certain laboratory conditions, while others do not. Even where an endurance limit is reported, surface finish, size, stress ratio, mean stress, corrosion, temperature, and variable-amplitude loading can change the practical result.
A second mistake is comparing fracture toughness values without checking specimen validity and test conditions. ASTM E399 focuses on plane-strain fracture toughness under defined conditions. ASTM E1820 is used where elastic-plastic behavior, J-integral, CTOD, or resistance-curve evaluation is appropriate. Comparing values across methods without understanding the underlying assumptions can lead to misleading material rankings.
A third mistake is ignoring defects created by manufacturing. Weld discontinuities, casting porosity, machining damage, grinding burns, inclusions, additive manufacturing pores, and residual tensile stresses may reduce fatigue performance even when the base material specification looks adequate. In many real structures, the manufacturing route controls fatigue reliability as much as the alloy name.
A fourth mistake is assuming that a simulation result is more accurate than the input data. Finite element analysis can locate stress concentrations and support crack-driving-force calculations, but it cannot correct unrepresentative fatigue properties, unrealistic boundary conditions, or unknown defect populations. Simulation and testing work best when used as complementary evidence.
Frequently asked questions
Is fatigue failure possible below the yield strength?
Yes. Fatigue can occur when repeated local stress and strain cycles create microscopic damage even though the nominal stress is below the material’s static yield strength. Stress concentrations, surface defects, corrosion pits, and residual tensile stress can accelerate this process.
What is the difference between fatigue strength and fracture toughness?
Fatigue strength describes resistance to failure under repeated loading, often expressed through stress-life data. Fracture toughness describes resistance to crack extension under specified crack-tip conditions. A material can have good fatigue strength in smooth specimens but still be vulnerable if it has low toughness or contains sharp defects.
Why do fatigue test results scatter so much?
Fatigue is sensitive to microstructure, surface condition, specimen preparation, environment, alignment, loading history, and small defects. Because cracks often start at local features, two specimens with the same nominal material and stress level may fail after different numbers of cycles.
Can one standard test prove a structure is safe?
No. Standard tests provide controlled material data, but structural safety also depends on geometry, manufacturing quality, loading spectrum, inspection capability, environment, and consequence of failure. Critical applications usually require a combination of material tests, analysis, quality control, and inspection planning.
How should engineers use published fatigue and fracture data?
Published data are best used as a starting point for screening and comparison. Before using them in design, engineers should check whether the data match the actual material form, heat treatment, thickness, surface finish, temperature, environment, stress ratio, and loading condition of the intended structure.
