Multiaxial Fatigue Testing of Materials: Methods, Standards, and Applications

Multiaxial fatigue testing evaluates the durability of metallic and non-metallic materials under combined cyclic loading conditions such as tension-torsion, biaxial tension, and axial loading with internal pressure. By reproducing the complex, non-proportional stress states found in real service components, these tests provide the reliable fatigue life data required for the safe design of shafts, pressure equipment, aerospace structures, and powertrain components.

Technical Overview

Multiaxial fatigue testing is a specialized mechanical testing discipline in which a material specimen is subjected to two or more independently controlled cyclic load components—typically axial force, torque, internal pressure, or biaxial tension—applied simultaneously. In actual service, engineering components such as crankshafts, gas turbine discs, pressure vessels, drive shafts, and aircraft landing gear rarely experience pure uniaxial loading. Instead, they operate under combined stress states in which the principal stress directions may remain fixed (proportional or in-phase loading) or rotate continuously throughout each loading cycle (non-proportional or out-of-phase loading).

Non-proportional loading is particularly damaging because the rotation of principal stress and strain directions activates additional slip systems in the material, producing a phenomenon known as non-proportional additional hardening. This cyclic hardening accelerates crack initiation and can reduce fatigue life by a factor of several compared with proportional loading at the same equivalent strain amplitude. Consequently, uniaxial fatigue data alone cannot be reliably extrapolated to predict component life under complex loading, and dedicated multiaxial fatigue tests are essential for accurate life assessment.

Test results are typically analyzed using equivalent stress or strain approaches (von Mises or Tresca criteria) or, more accurately, critical plane approaches such as the Fatemi-Socie, Smith-Watson-Topper (SWT), Brown-Miller, and Findley parameters, which relate crack initiation and early growth to the stress and strain history on the plane where cracks physically form. Accredited third-party laboratories commonly conduct these tests in accordance with recognized standards organizations including ASTM International, ISO, and the Chinese national GB series.

Testing Samples

Thin-Walled Tubular Specimens

Thin-walled tubes are the most widely used geometry for combined axial-torsional fatigue testing. The thin wall ensures a nearly uniform stress distribution through the thickness, allowing independent control of normal stress and shear stress. This specimen type is specified in ASTM E2207 for strain-controlled axial-torsional fatigue testing and enables precise characterization of proportional and non-proportional material behavior.

Cruciform (Cross-Shaped) Specimens

Flat cruciform specimens are loaded along two perpendicular axes in biaxial test frames. They are used to evaluate in-plane biaxial tension-compression behavior, including equibiaxial loading states that cannot be achieved with tubular specimens, and are valuable for sheet metals, welded joints, and thin-walled structures.

Solid Cylindrical Specimens

Solid round bars are employed primarily for pure torsional fatigue or tension-torsion tests where wall-thickness limitations of tubes are unacceptable. Stress gradients through the section must be accounted for during data reduction.

Notched Specimens and Full-Scale Components

Notched tubular or cruciform specimens introduce controlled stress concentrations for evaluating multiaxial notch effects and crack initiation behavior. Structural-scale tests on actual components—such as shafts, flanged joints, and pressure fittings—validate design assumptions under realistic multiaxial load spectra.

Testing Items

Typical multiaxial fatigue testing programs conducted by accredited third-party laboratories include:

• Proportional (in-phase) axial-torsional fatigue testing under stress or strain control
• Non-proportional (out-of-phase) fatigue testing with phase angles from 45° to 180°
• Biaxial tension-compression fatigue testing on cruciform specimens
• Combined axial load, torsion, and internal pressure fatigue testing of tubular specimens
• Torsional fatigue testing under torque control in accordance with ISO 1352 and GB/T 12443
• Axial strain-controlled fatigue testing in accordance with ASTM E606, ISO 12106, and GB/T 26077
• Force-controlled axial fatigue testing in accordance with ASTM E466 and GB/T 3075
• Multiaxial fatigue life (S-N and ε-N) curve determination
• Non-proportional additional hardening characterization and cyclic stress-strain response measurement
• Crack initiation life, crack orientation, and fracture surface analysis
• Mean stress and mean strain effects under combined loading
• Variable-amplitude and spectrum multiaxial loading tests
• Thermomechanical and elevated-temperature multiaxial fatigue testing

Testing Methods

Multiaxial fatigue tests are performed on servo-hydraulic axial-torsional test systems equipped with two independent actuators, or on planar biaxial frames with four actuators for cruciform specimens. Closed-loop digital controllers coordinate the load channels so that specified amplitudes, phase angles, and load path shapes are maintained throughout the test. Key methodological elements include:

• Specimen alignment and verification to minimize bending and parasitic stresses, in line with ASTM E1012 practice for alignment verification
• Instrumentation with biaxial extensometers capable of independently measuring axial strain and engineering shear strain on tubular specimens
• Control mode selection: force control, strain control, or mixed axial-strain/torque control depending on the expected service condition and specimen behavior
• Load path programming, including straight-line proportional paths and circular, elliptical, square, L-shaped, and box-shaped non-proportional strain paths
• Phase angle control between axial and torsional channels, with 90° out-of-phase sinusoidal loading being the most common non-proportional condition
• Failure definition, typically based on a specified drop in peak load or stiffness (for example, 10 percent), the detected initiation of a surface crack of defined length, or total separation
• Data reduction using equivalent stress and strain amplitudes as well as critical plane damage parameters (Fatemi-Socie, SWT, Brown-Miller, Findley) to correlate lives across different loading paths
• Post-test fractography by optical and scanning electron microscopy to identify crack initiation sites, crack planes, and propagation modes

Ambient-temperature tests are most common, but specialized systems with induction heating or environmental chambers allow multiaxial fatigue evaluation at elevated temperatures or in corrosive media. Test frequencies typically range from 0.1 Hz to 20 Hz depending on control mode, specimen response, and the requirement to avoid self-heating.

Standards and Reference Documents

Testing is generally performed with reference to internationally and nationally recognized standards, including:

• ASTM E2207 – Standard Practice for Strain-Controlled Axial-Torsional Fatigue Testing with Thin-Walled Tubular Specimens
• ASTM E466 – Standard Practice for Force-Controlled Constant Amplitude Axial Fatigue Testing of Metallic Materials
• ASTM E606 – Standard Test Method for Strain-Controlled Fatigue Testing
• ASTM E468 – Standard Practice for Presentation of Constant Amplitude Fatigue Test Results for Metallic Materials
• ASTM E1012 – Standard Practice for Verification of Testing Frame and Specimen Alignment Under Tensile and Compressive Axial Force Application
• ISO 12106 – Metallic Materials, Fatigue Testing, Axial-Strain-Controlled Fatigue Method
• ISO 1352 – Steel, Torque-Controlled Fatigue Testing
• GB/T 3075 – Metallic Materials, Fatigue Testing, Axial Force Control Method
• GB/T 26077 – Metallic Materials, Fatigue Testing, Axial-Strain-Controlled Method
• GB/T 12443 – Metallic Materials, Torque-Controlled Fatigue Testing

Applications

Multiaxial fatigue testing supports design, qualification, and failure analysis across many industries:

• Aerospace: turbine discs and blades, landing gear, wing attachments, and engine shafts subjected to combined centrifugal, thermal, and vibratory loads
• Automotive: crankshafts, connecting rods, drive shafts, constant-velocity joints, suspension arms, and wheel hubs under combined bending and torsion
• Power generation: pressure vessels, boiler tubes, piping systems, steam and gas turbine rotors, and nuclear components experiencing thermal-mechanical cycling
• Oil and gas: drill strings, wellhead equipment, risers, and threaded connections under tension, torsion, and internal pressure
• Rail transportation: axles, bogie frames, and rail fastening systems under combined bending, torsion, and contact loads
• Biomedical engineering: orthopedic implants such as hip stems and bone plates loaded in combined bending and torsion
• Additive manufacturing and welding research: qualification of processed materials and welded joints under realistic multiaxial stress states

FAQ

What is the difference between uniaxial and multiaxial fatigue testing?

Uniaxial fatigue testing applies cyclic loading along a single axis, producing one principal stress direction. Multiaxial fatigue testing applies two or more independent cyclic load components simultaneously—such as axial force plus torque or biaxial tension—creating combined normal and shear stress states in which principal directions may rotate during each cycle. This reproduces real service conditions far more accurately.

Why does non-proportional loading reduce fatigue life?

When axial and torsional loads are out of phase, the principal stress and strain axes rotate continuously within each cycle. This forces the material to activate multiple slip systems, causing additional cyclic hardening, faster accumulation of dislocation damage, and earlier crack initiation compared with in-phase loading at the same equivalent amplitude.

Which specimen geometry is most commonly used?

The thin-walled tubular specimen is the most common choice for axial-torsional testing because it provides a nearly uniform biaxial stress state through the wall thickness and permits independent control of normal and shear components. Cruciform specimens are preferred for planar biaxial tension studies of sheet materials.

What failure criteria are used to terminate a test?

Tests are commonly terminated when the peak tensile load drops by a specified percentage (often 10 percent relative to stabilized values), when a surface crack of a defined length is detected, or at complete specimen separation. The selected criterion must be reported together with the data, as it directly affects reported fatigue life.

How are multiaxial fatigue results analyzed?

Simple analyses use equivalent stress or strain amplitudes based on von Mises or Tresca criteria, suitable for proportional loading. For non-proportional conditions, critical plane approaches such as Fatemi-Socie for shear-dominated failure and Smith-Watson-Topper for tension-dominated failure provide much better life correlations because they track damage on the physical cracking plane.

What industries require multiaxial fatigue data?

Aerospace, automotive, power generation, oil and gas, rail transportation, and medical device industries all rely on multiaxial fatigue data for component design validation, durability sign-off, remaining life assessment, and root-cause failure investigation.