Stress Fatigue Testing and Determination: Methods, Standards, and Applications

Stress fatigue determination is a fundamental mechanical testing discipline that quantifies how materials behave under repeated cyclic loading, producing the S-N curves, fatigue limits, and fatigue strength data essential for safe and reliable structural design. This technical guide explains the testing principle, applicable ISO, ASTM, EN, JIS, and GB standards, specimen requirements, test methods, and typical industrial applications of stress fatigue evaluation performed in accredited third-party laboratories.

Technical Overview

Fatigue is the progressive, localized, and permanent structural damage that develops when a material is subjected to repeated or fluctuating stresses at levels well below its ultimate tensile strength. Engineering failure surveys consistently attribute the majority of in-service mechanical failures to fatigue rather than static overload, which makes fatigue data a critical input for design, qualification, and failure analysis. Stress fatigue determination, commonly known as the stress-life (S-N) approach, measures a material's resistance to cyclic loading by applying controlled, constant-amplitude stress cycles to standardized specimens until failure occurs or a runout condition is reached.

The primary deliverable of stress fatigue determination is the S-N (Wöhler) curve, which plots stress amplitude or maximum stress against the number of cycles to failure on semi-logarithmic coordinates. For many ferrous alloys, the curve approaches a horizontal asymptote known as the fatigue limit or endurance limit, below which the material can theoretically sustain an effectively infinite number of cycles. Non-ferrous alloys such as aluminum generally show no true fatigue limit, so a fatigue strength at a specified number of cycles (for example, 10^7 cycles) is reported instead. In the finite-life region, results are frequently fitted with the Basquin power-law relationship to support damage-tolerant and safe-life design calculations.

Stress fatigue determination primarily addresses the high-cycle fatigue regime, roughly 10^4 to 10^7 cycles or beyond, in which nominal stresses remain nominally elastic. Tests are conducted on servo-hydraulic, electrodynamic, or resonant fatigue testing systems whose static force calibration follows ASTM E4 and whose dynamic force verification follows ASTM E467. Bending errors introduced by load-frame misalignment are controlled and quantified according to ASTM E1012, because even small secondary bending stresses can significantly shift fatigue results.

Testing Samples

Machined Metallic Specimens

Cylindrical hourglass specimens or flat rectangular specimens are machined from bars, plates, forgings, or castings in accordance with ASTM E466, ISO 1099, and GB/T 3075. Specimen preparation is critical: gauge surfaces are longitudinally polished to a roughness typically at or below Ra 0.2 µm, sharp circumferential tool marks are avoided, and machining is performed in stages with light final cuts to prevent introducing residual stresses that would distort the fatigue response.

Welded Joint Specimens

Cross-section specimens containing the weld, heat-affected zone, and base metal are tested with the weld reinforcement retained or removed depending on the applicable design code. Welded details are fatigue-critical because geometric stress concentration, residual welding stresses, and possible internal imperfections substantially reduce fatigue strength relative to the parent material.

Fasteners and Threaded Components

Bolts, screws, and studs are tested in the as-finished condition under axial cyclic loading following ISO 3800 and related fastener fatigue procedures. Test stress is defined on the tensile stress area of the thread, and results are used to verify fatigue performance of high-strength fastener grades and to evaluate locking devices and plating effects.

Cast, Forged, Powder Metallurgy, and Additive Manufactured Materials

These product forms are evaluated to characterize the influence of microstructure, porosity, surface condition, and build orientation on fatigue performance. Additive manufactured alloys in particular often require dedicated S-N characterization in both as-built and machined or post-processed surface conditions.

Full-Scale Components and Springs

When coupon data are insufficient, actual components such as springs, shafts, brackets, and medical implants are tested under simulated service load spectra to validate durability and demonstrate compliance with product standards such as ISO 14801 for dental implants.

Testing Items

Typical testing items within a stress fatigue determination program include:

• S-N (stress-life) curve generation at one or more stress ratios
• Fatigue limit or endurance limit determination for ferrous alloys
• Fatigue strength at a specified number of cycles (10^5, 10^6, 10^7, or higher)
• High-cycle fatigue life verification at defined stress amplitudes
• Mean stress and stress ratio (R) effects, including Haigh diagram construction with Goodman or Gerber corrections
• Fatigue notch factor (Kf) and notch sensitivity evaluation
• Effect of surface finish, coatings, and treatments such as shot peening or nitriding
• Low-cycle strain-controlled fatigue properties where required (cyclic stress-strain and strain-life data)
• Fatigue crack growth rate (da/dN) and threshold stress intensity factor range (ΔKth)
• Fractographic examination of failed specimens to identify initiation sites and failure mechanisms

Testing Methods

Axial Force-Controlled Fatigue Testing

Axial loading is the most widely applied method and is governed by ASTM E466, ISO 1099, and GB/T 3075. Specimens are cycled under constant-amplitude sinusoidal force control, most commonly at a fully reversed stress ratio of R = -1 or a tension-tension ratio of R = 0.1. Testing frequencies typically range from a few hertz up to approximately 100 Hz depending on the system and on controlling specimen heating. Testing is terminated at specimen separation, at defined stiffness drop criteria, or at the runout cycle count, commonly 10^7 cycles.

Rotating Bending Fatigue Testing

Rotating bar bending fatigue, performed according to ISO 1143, JIS Z 2274, and GB/T 4337, applies a constant bending moment to a rotating specimen so that every surface fiber experiences fully reversed stresses each revolution. This method is economical for screening fatigue limits of bar and wire products and remains widely used for steel quality control.

Torsional and Combined Loading Fatigue Testing

Components loaded in shear, such as shafts and torsion bars, are evaluated under torque-controlled cycling in accordance with ISO 1352. Where service loading is multiaxial, combined axial-torsional programs can be developed to reproduce realistic stress states.

Strain-Controlled Low-Cycle Fatigue Testing

When cyclic strains become appreciably plastic, strain-controlled testing per ASTM E606, ISO 12106, and GB/T 15248 is used to obtain cyclic stress-strain curves and strain-life (Coffin-Manson) parameters that complement the stress-based data and support fatigue analysis in the low-cycle regime.

Fatigue Crack Growth Testing

Crack propagation behavior is measured on pre-cracked compact or bend specimens following ASTM E647, ISO 12108, and GB/T 6398. The resulting da/dN versus ΔK curves and threshold values support damage-tolerance assessments and inspection interval planning.

Statistical Planning and Data Analysis

Because fatigue data are inherently scattered, test programs are planned and analyzed statistically according to ISO 12107, ASTM E739, and GB/T 24176. A complete S-N curve typically requires 8 to 15 specimens distributed across stress levels, while the staircase method applied to 15 or more specimens provides a statistically defensible estimate of the fatigue limit with defined confidence bounds.

Applications

Stress fatigue determination underpins durability assurance across virtually every industry in which structures and machines experience repeated loading. Representative applications include:

• Aerospace: airframe structures, landing gear, engine disks, and fasteners, where fatigue substantiation is mandatory for certification
• Automotive: crankshafts, connecting rods, suspension springs, wheels, and powertrain components subjected to road-load spectra
• Rail transportation: axles and wheelset components evaluated against EN 13261 requirements
• Energy and pressure equipment: pressure vessels, piping, turbine blades, and wind turbine drivetrain components
• Medical devices: orthopedic implants, spinal constructs, and dental implants qualified under standards such as ISO 14801
• Additive manufacturing: build-orientation-dependent fatigue qualification of printed structural alloys
• Materials development: ranking of alloys, heat treatments, coatings, and surface treatments during research and procurement
• Failure analysis: comparing the fatigue strength of a failed component against expected performance to establish root cause

FAQ

What is stress fatigue determination?

It is the laboratory process of applying controlled, constant-amplitude cyclic stresses to material specimens or components and recording the number of cycles sustained until failure or runout, in order to establish S-N curves, fatigue limits, and fatigue strength values used in design and qualification.

What is the difference between stress-based and strain-based fatigue testing?

Stress-based (S-N) testing controls applied stress and suits the high-cycle regime where strains are nominally elastic. Strain-based testing controls total strain amplitude and is used for low-cycle fatigue, where plastic deformation each cycle governs behavior and Coffin-Manson strain-life parameters are required.

How many specimens are needed to build an S-N curve?

A preliminary S-N curve typically requires 8 to 15 specimens distributed over several stress levels. Statistically rigorous programs, including fatigue limit determination by the staircase method, may require 20 to 30 or more specimens as guided by ISO 12107.

What is a runout?

A runout is a specimen that survives a predetermined cycle count, commonly 10^7 cycles for steels, without failure. Runout data are treated separately in the S-N analysis and indicate stresses at or below the fatigue limit.

Which stress ratios are used?

Fully reversed loading (R = -1) is the classical reference condition, while R = 0.1 is widely used in axial testing because it avoids compressive buckling. Other ratios are selected to reproduce specific service mean stresses, and results are often summarized on a Haigh diagram.

Can stress fatigue testing be performed under special environments?

Yes. Environmental chambers allow testing at elevated or sub-ambient temperatures, in corrosive media, or in controlled humidity. Elevated-temperature and creep-fatigue interaction testing can additionally follow ASTM E2714 when applicable.

How long does fatigue testing take?

Duration depends on frequency and target cycles: a single 10^7-cycle test at 50 Hz takes roughly two to three days, while resonant or ultrasonic systems reaching 10^9 cycles for very-high-cycle-fatigue studies can shorten programs substantially. Full S-N programs are typically scheduled over several weeks.