Cyclic Stress-Strain Curve Determination: Testing Methods, Standards, and Applications

Cyclic stress-strain curve determination is a core laboratory technique for characterizing how metallic materials harden or soften under repeated elastic-plastic loading, providing the cyclic strength coefficient and cyclic strain hardening exponent essential for low-cycle fatigue design. Following recognized standards such as ASTM E606 and ISO 12106, this article explains the testing principle, specimen requirements, test procedures, data reduction methods, and typical industrial applications of cyclic stress-strain curve measurement.

Technical Overview

The cyclic stress-strain curve describes the stabilized relationship between stress amplitude and strain amplitude when a material is subjected to repeated plastic straining. Unlike a monotonic tensile curve, which is obtained from a single loading to fracture, the cyclic curve reflects the steady-state deformation resistance that a material develops after the transient stage of cyclic hardening or cyclic softening has saturated. For components subjected to low-cycle fatigue (LCF), where local plastic strains at notches, welds, and thermal transient locations control service life, the cyclic curve is far more representative of actual material behavior than the monotonic curve.

Under fully reversed strain-controlled loading, each cycle traces a closed hysteresis loop on the stress-strain plane. The total strain amplitude consists of an elastic part (stress amplitude divided by Young's modulus) and a plastic part (measured from the width of the loop at zero stress). After an initial transient response, typically tens to hundreds of cycles, the loop stabilizes. The locus of stabilized loop tips measured at different strain amplitudes defines the cyclic stress-strain curve, which is universally expressed by the Ramberg-Osgood relationship:

Δε/2 = Δσ/2E + (Δσ/2K′)^(1/n′)

where Δε/2 is the strain amplitude, Δσ/2 is the stress amplitude, E is the elastic modulus, K′ is the cyclic strength coefficient, and n′ is the cyclic strain hardening exponent. Comparing the cyclic curve with the monotonic curve classifies material behavior: annealed low-strength steels, aluminum alloys, and many austenitic stainless steels tend to cyclically harden, while cold-worked materials and high-strength quenched-and-tempered steels tend to cyclically soften. This distinction is critical because a softening material loaded near its monotonic yield strength may undergo progressive plastic ratcheting in service.

The principal standards governing the test are ASTM E606/E606M (Standard Test Method for Strain-Controlled Fatigue Testing), ISO 12106 (Metallic materials — Fatigue testing — Axial-strain-controlled method), and the Chinese national standard GB/T 26077. Supporting standards include ASTM E467 for dynamic force verification, ASTM E1012 for load-train alignment, and ASTM E83 for extensometer classification. Testing should be performed in accredited third-party laboratories or in-house facilities with calibrated equipment and documented quality systems.

Testing Samples

Cylindrical Specimens

Cylindrical specimens are the most common geometry and are machined from bars, forgings, castings, or sections cut from actual components. Two configurations are used: uniform-gauge specimens, which allow strain measurement over a parallel length and are preferred when buckling is not a concern, and hourglass specimens, which concentrate deformation at the minimum cross section and are preferred at large strain amplitudes or elevated temperature. Gauge diameters typically range from 6 mm to 12.5 mm, with generous fillet radii to avoid stress concentration at the shoulders.

Flat Sheet and Plate Specimens

For sheet and plate products that cannot yield round bars, flat specimens with rectangular cross sections are used. Because thin sections are prone to compressive buckling, anti-buckling guides are often required, and gauge lengths are kept short. Edge preparation and surface condition are more critical for flat specimens since fatigue cracks frequently initiate at machined edges.

Specimen Preparation and Surface Condition

All specimens must be finished by low-stress machining followed by progressive grinding and final axial polishing to remove circumferential machining marks. A surface roughness of Ra 0.2 μm or better is typically specified for baseline material characterization. Specimens must be identified by material pedigree, including heat number, heat treatment condition, product form, and sampling orientation, because cyclic behavior is strongly influenced by metallurgical condition and processing history.

Number of Specimens

• Multiple specimen method: typically 10 to 20 specimens per material condition to construct a reliable cyclic curve, with more required when the full strain-life curve is also needed.
• Incremental step method: a single specimen is sufficient for a first approximation, plus optional verification specimens.
• Replicate tests at selected amplitudes are recommended to quantify data scatter.

Testing Items

A complete cyclic stress-strain curve determination campaign typically delivers the following results:

• Stabilized hysteresis loops at each strain amplitude.
• Stress amplitude versus elastic, plastic, and total strain amplitude data.
• Cyclic strength coefficient K′ and cyclic strain hardening exponent n′ from Ramberg-Osgood fitting.
• Cyclic yield strength determined at 0.2% offset from the cyclic curve.
• Cyclic hardening or softening ratio, comparing cyclic and monotonic proof stress.
• Hysteresis loop width and plastic strain energy per cycle.
• Companion strain-life parameters where applicable: fatigue strength coefficient σf′, fatigue strength exponent b, fatigue ductility coefficient εf′, and fatigue ductility exponent c.
• Transition fatigue life 2Nt, at which elastic and plastic strain amplitudes are equal.
• Cycles to failure Nf at each tested amplitude, with the failure criterion clearly defined.

Testing Methods

Equipment and Setup

Testing is performed on servo-hydraulic or electrodynamic axial fatigue frames equipped with a calibrated load cell and a digital closed-loop controller. Strain is measured with an axial extensometer of at least Class B-1 accuracy per ASTM E83 (or equivalent per ISO 9513), attached directly to the gauge section. Load-train alignment is verified per ASTM E1012, with bending strain typically limited to 5% or less of the axial strain at the gauge section. Specimen temperature is monitored, and strain rates, typically in the range of 10⁻³ to 10⁻² s⁻¹, are selected to limit self-heating. A triangular command waveform is used to maintain constant strain rate throughout the cycle.

Multiple Specimen (Companion) Method

In the reference method, several identical specimens are each cycled at a different constant total strain amplitude, most commonly fully reversed with a strain ratio of Rε = −1. Amplitudes typically span about 0.2% to 2.0% total strain, adjusted according to material ductility. Hysteresis loops are recorded at logarithmic cycle intervals, and the loop at half the failure life (N50) is taken as representative of stabilized behavior. The tips of these half-life loops are plotted to construct the cyclic stress-strain curve, and the Ramberg-Osgood parameters are obtained by log-log regression of stress amplitude against plastic strain amplitude.

Incremental Step Test Method

In this economical single-specimen approach, the strain amplitude is increased block by block to a maximum value and then decreased symmetrically, forming one complete block. Ten to twenty blocks are applied until the response stabilizes. The tensile halves of the final stabilized block are averaged to construct the cyclic curve. The method is well suited to comparative studies and material screening, although results for strongly history-dependent materials may deviate slightly from companion-specimen data.

Procedure and Data Reduction

• Mount the specimen, attach the extensometer, and verify zero load and strain.
• Select strain amplitudes, strain ratio, waveform, and strain rate, and document the test plan.
• Run each test under total strain control, recording peak and valley stresses and periodic hysteresis loops.
• Define failure by complete separation, a specified load drop (commonly 10%, 25%, or 50%), or a crack initiation criterion agreed with the client.
• Extract σa = Δσ/2 and plastic strain amplitude εpa from each stabilized loop.
• Fit the cyclic stress-strain relationship Δσ/2 = K′(Δεp/2)^n′ and report E, temperature, strain rate, failure definition, and statistical scatter.
• Optionally verify Masing or non-Masing behavior by checking whether stabilized half-loops match when translated to a common origin.

Applications

Cyclic stress-strain data are a fundamental input to fatigue design and structural integrity assessment across many industries. Representative applications include:

• Low-cycle fatigue design of gas turbine discs, blades, and combustor hardware subjected to startup-shutdown thermal-mechanical loading.
• Pressure vessels and piping in nuclear and thermal power plants, where thermal transients and stratification produce cyclic plastic strains.
• Automotive powertrain, exhaust, and suspension components experiencing repeated thermal and mechanical cycling.
• Notch and stress-concentration analysis using the local strain approach combined with Neuber's rule.
• Calibration of cyclic plasticity models, such as Chaboche-type kinematic hardening, for finite element simulation of notched and welded structures.
• Qualification of welded joints, castings, forgings, and additively manufactured alloys, where cyclic behavior may differ substantially from wrought baseline data.
• Heat-treatment optimization, supplier qualification, and generation of material design allowables for engineering databases.

FAQ

What is the difference between a monotonic and a cyclic stress-strain curve?

A monotonic curve is measured during a single continuous load application to fracture, while a cyclic curve describes the stabilized stress-strain response after repeated loading. Materials that harden cyclically show a cyclic curve above the monotonic curve, and softening materials show the opposite. Using monotonic properties in low-cycle fatigue analysis can therefore be unconservative for softening materials.

Which standards govern cyclic stress-strain curve determination?

The primary standards are ASTM E606/E606M and ISO 12106, together with the Chinese national standard GB/T 26077. Related standards include ASTM E467, ASTM E1012, and ASTM E83 for machine, alignment, and extensometer verification.

Why is the half-life hysteresis loop used?

At half the failure life, transient hardening or softening is complete and macro-crack propagation has not yet significantly reduced the load-carrying cross section, so the loop best represents stabilized bulk material behavior.

How many specimens are needed?

The incremental step method requires only one specimen, making it attractive for screening. The multiple specimen method typically requires 10 to 20 specimens to define the cyclic curve with acceptable confidence, particularly when strain-life parameters are also required.

Which materials cyclically harden and which soften?

Generally, soft annealed metals such as low-carbon steel, copper, and aluminum alloys harden cyclically, while cold-worked or high-strength quenched-and-tempered steels soften. Some materials, including certain stainless steels and duplex grades, display mixed behavior depending on strain amplitude and temperature.

Can the test be performed at elevated temperature?

Yes. Elevated-temperature testing uses induction or furnace heating with high-temperature extensometers and thermal soak procedures. Strain rates are often reduced to minimize temperature gradients and self-heating, in accordance with the requirements of ISO 12106 and ASTM E606.

How are the results used in fatigue life prediction?

The cyclic stress-strain curve is combined with the strain-life Coffin-Manson relationship, Δε/2 = (σf′/E)(2N)^b + εf′(2N)^c, within the local strain approach. This allows notched component lives to be estimated from the cyclic stress-strain response at the critical location, forming the basis of many modern durability assessment workflows.