Tensile testing speed, more precisely expressed as strain rate, has a direct and measurable influence on the plasticity indicators reported for metallic and polymeric materials, including elongation at fracture, uniform elongation, and reduction of area. This article analyzes the deformation mechanisms that link loading rate to plastic behavior, explains how ISO 6892-1, ASTM E8/E8M and related standards control test rates, and provides practical guidance for selecting and reporting rates so that results remain reproducible and comparable between laboratories.
Technical Overview
The rate at which a tensile test is performed is not a trivial procedural detail: it is one of the most influential test parameters governing the measured mechanical response of a material. In uniaxial tensile testing, the relevant quantity is the strain rate, defined as the change in engineering strain per unit time and typically expressed in units of s⁻¹. Conventional quasi-static testing covers the range of approximately 10⁻⁵ to 10⁻² s⁻¹, intermediate rates extend to roughly 10 s⁻¹, and dynamic testing regimes exceed 10² s⁻¹. Because plastic deformation is a time-dependent process at the microstructural scale, changing the strain rate can shift the measured yield strength, tensile strength, elongation, and reduction of area of the same material by margins large enough to affect acceptance decisions.
For the majority of metals tested at room temperature in the quasi-static range, increasing the strain rate raises the flow stress. Dislocation motion relies partly on thermally activated processes such as overcoming short-range obstacles; when there is less time available for thermal activation, a higher applied stress is required to sustain the same deformation rate. This produces the well-known positive strain rate sensitivity observed in face-centered cubic (FCC) alloys such as aluminum and copper, and in most body-centered cubic (BCC) steels at room temperature. The magnitude of the effect is quantified by the strain rate sensitivity exponent m, defined as m = ∂(ln σ)/∂(ln ε̇), which typically falls between 0.01 and 0.05 for structural metals at room temperature but can exceed 0.3 in superplastic alloys.
Fundamental Mechanisms of Rate-Dependent Plasticity
Three coupled mechanisms explain why plasticity indicators vary with testing speed. First, thermally activated dislocation glide causes the critical resolved shear stress to increase with strain rate, elevating both yield and flow stresses. Second, strain rate hardening delays the onset of necking instability according to the Considère-Hart criterion, meaning that a material with high positive m-value can exhibit greater uniform elongation at higher rates. Third, at very high rates the deformation becomes adiabatic, so that plastic work converted into heat raises the local temperature, softens the material, and in extreme cases promotes localized adiabatic shear bands that sharply reduce ductility. The observed ductility at any given rate reflects the competition among these strengthening and softening mechanisms.
How Testing Speed Affects Key Plasticity Indicators
- Elongation after fracture (A): For many ductile steels and aluminum alloys in the quasi-static range, elongation changes only moderately with rate, often a few percent relative, while strength changes are more pronounced. In rate-sensitive materials, faster testing can either increase or decrease A depending on whether strain rate hardening or adiabatic localization dominates.
- Uniform elongation (Ag) and total extensometer strain at maximum force (Agt): These indicators generally respond more strongly to rate than total elongation, because they are governed by the necking instability, which is directly stabilized by positive strain rate sensitivity.
- Reduction of area (Z): Z is usually less rate-sensitive than elongation in the quasi-static regime, but it can drop significantly once adiabatic shear localization or dynamic strain aging conditions are reached.
- Yield phenomena: Distinct yield points and yield-point elongation (Lüders band propagation) in low-carbon steels can be suppressed or amplified depending on the imposed rate, affecting measured lower yield strength (ReL) and elongation values.
Special Cases: Dynamic Strain Aging and Adiabatic Effects
Certain material-rate-temperature combinations produce anomalous behavior. Dynamic strain aging (DSA), driven by the interaction between diffusing interstitial atoms and mobile dislocations, causes negative strain rate sensitivity within specific temperature windows, for example in low-carbon steels around the blue-brittleness region and in some aluminum-magnesium alloys at room temperature where Portevin-Le Chatelier serrations appear. Under DSA conditions, faster testing can paradoxically lower the flow stress and degrade plasticity. At strain rates above roughly 10 s⁻¹, adiabatic heating becomes significant, and above 10³ s⁻¹, wave propagation and inertia effects require specialized high-rate techniques. Understanding these regimes is essential when laboratory data are used to predict behavior in service conditions such as crash loading, impact, or high-speed forming.
Testing Samples
Test piece geometry and preparation directly interact with rate effects, because the true strain rate experienced by the gauge section depends on the gauge length and the cross-section in addition to the actuator speed. Accredited third-party laboratories machine and condition specimens strictly according to the geometry requirements of the governing standard so that rate-dependent results remain traceable and comparable.
Metallic Test Pieces
Metallic specimens are prepared in accordance with ISO 6892-1, ASTM E8/E8M, EN ISO 6892-1, or GB/T 228.1, depending on the specification invoked. Common configurations include machined round bars with threaded or shoulder heads, rectangular flat coupons for sheet and plate, and proportional test pieces where the gauge length is calculated as 5.65 times the square root of the original cross-sectional area. The parallel length, fillet radii, and surface finish are controlled because stress concentrators amplify rate-sensitive local deformation and can artificially depress ductility measurements.
Polymeric Specimens
Polymers exhibit pronounced viscoelasticity, making them far more rate-sensitive than metals. Specimens are typically dumbbell-shaped type 1A or 1B test pieces per ISO 527-1 and ISO 527-2, or Type I specimens per ASTM D638. Because polymer ductility can change dramatically between slow and fast crosshead speeds, the specified testing speed for each material class must be strictly followed.
Product Forms and Full-Section Samples
- Wire and bar products tested at full section in accordance with ASTM A370 or the relevant product standard, where the strain rate is estimated from crosshead speed and the parallel length.
- Fasteners loaded axially per ISO 898-1 or ASTM F606, with rate limits specified to avoid non-conservative yield measurements.
- Tubular sections, castings, forgings, and additively manufactured coupons, where orientation and build direction must be recorded alongside the applied rate.
Testing Items
A rate-controlled tensile test campaign delivers a complete set of strength and plasticity parameters, each with its own rate dependence. The core measured and derived items include:
- Upper and lower yield strength (ReH, ReL) or proof strength (Rp0.2, Rt0.5) as the rate-sensitive strength indicators
- Tensile strength (Rm) and force at fracture
- Percentage elongation after fracture (A) with a stated gauge length, such as A50 mm or A5.65√S0
- Percentage uniform elongation (Ag) and percentage total extension at maximum force (Agt)
- Percentage reduction of area (Z) for test pieces that fail with a cup-and-cone or ductile morphology
- Strain rate sensitivity exponent (m) determined from tests at two or more rates
- Strain hardening exponent (n) derived from the true stress-strain curve at each rate
- Young's modulus, determined at a specifically low and tightly controlled rate
Testing Methods
Modern standards approach rate control with precision precisely because of the effects described above. Selecting and documenting the control mode is as important as selecting the rate itself.
Rate Control Modes
ISO 6892-1 defines three control philosophies. Closed-loop strain rate control (Method A) uses feedback from the extensometer to maintain a constant strain rate within the gauge section and provides the best reproducibility. Open-loop control based on estimated strain rate over the parallel length (Method B) uses a calculated crosshead speed and depends on the testing machine stiffness. Stress rate control is applied only within the elastic region before yield, typically at 2 to 20 N/mm² per second depending on the material strength class. ASTM E8/E8M similarly permits control of crosshead separation rate or strain rate, with nominal strain rate values of 0.015 s⁻¹, 0.003 s⁻¹, and lower for different test stages.
Standardized Rate Schedules
ISO 6892-1 structures the test sequence into stages, each with its own permissible rate: an initial low rate for modulus determination if required, a controlled rate up to yield, and a higher permissible rate after yield because post-yield rate sensitivity is lower. Following the staged schedule ensures that the yield point, the most rate-sensitive parameter, is always measured under equivalent conditions between laboratories. Interlaboratory studies have repeatedly shown that failing to coordinate rates is a leading cause of scattered yield strength and elongation results for identical materials.
Intermediate and High-Rate Testing
Servo-hydraulic frames equipped with high-response control loops and drop towers extend coverage into the 1 to 500 s⁻¹ range relevant to crash and impact analysis, while split Hopkinson pressure bar systems address rates above 10³ s⁻¹. In these regimes, force measurement, wave propagation effects, signal synchronization, and adiabatic assumptions must be validated, and results should be reported together with the strain rate history rather than a single nominal value.
Instrumentation, Validation, and Reporting
- Calibrated class 0.5 or better extensometers, or verified class 1 devices with the device class recorded, attached within the gauge length for direct strain measurement
- Verification of testing machine force and displacement calibration traceable to national metrology institutes
- Specimen temperature conditioning, since temperature shifts the strain rate sensitivity of most alloys
- Full reporting of the control method, nominal strain rate for each test stage, extensometer class, specimen geometry, and the numerical basis of any plasticity comparison
Applications
Rate-plasticity analysis supports engineering decisions wherever the speed of loading in service differs from the speed of a conventional laboratory test. Typical applications include:
- Automotive crashworthiness programs, where high-rate tensile data feed constitutive models used in explicit finite element simulation of structural deformation
- Sheet metal forming and stamping process design, where press speeds are orders of magnitude faster than standard laboratory rates and forming limit behavior must be rated correctly
- Pipeline, pressure vessel, and structural steel qualification, verifying that specified minimum elongation and yield values are met under the rate defined by the governing standard
- Fastener and wire drawing industries, in which ductility at production strain rates governs drawability and failure risk
- Additive manufacturing qualification, comparing as-built and heat-treated materials across rates to detect process-induced rate sensitivity anomalies
- Generation of calibrated input parameters for rate-dependent material models such as Johnson-Cook and Cowper-Symonds formulations
- Supplier and batch comparisons, ensuring that plasticity data used for acceptance are generated at identical rates by accredited third-party laboratories
FAQ
Does a faster tensile test always reduce the measured elongation?
No. The direction of the change depends on the material. In polymers and materials prone to adiabatic shear localization, higher rates generally reduce ductility. However, in metals with strong positive strain rate sensitivity, faster deformation raises flow stress and delays necking, which can increase uniform elongation. In the dynamic strain aging regime, the relationship can even invert. This is why rate-controlled comparisons, rather than single-rate assumptions, are required for sound conclusions.
What strain rates do ISO 6892-1 and ASTM E8/E8M actually specify?
ISO 6892-1 defines permissible strain rate ranges for each stage, with the lowest range typically around 0.00007 s⁻¹ and commonly used ranges extending through 0.0025 s⁻¹ for post-yield deformation, along with stress-rate limits in the elastic region. ASTM E8/E8M prescribes nominal strain rates such as 0.015 s⁻¹ and 0.003 s⁻¹ depending on the quantity being measured. Both standards allow faster rates after yield because post-yield rate sensitivity is smaller.
Is crosshead speed the same as strain rate?
No. Crosshead speed is a machine displacement command, while strain rate is the deformation rate experienced by the specimen gauge section. The two coincide only approximately, and the relationship depends on gauge length, machine stiffness, and specimen geometry. Closed-loop strain rate control using an extensometer eliminates this ambiguity and is the preferred method for reproducible testing.
How is the strain rate sensitivity exponent m determined?
The exponent m is obtained by performing tensile tests at a minimum of two, and preferably three or more, strain rates separated by at least one decade. The flow stress at a fixed true plastic strain is plotted against strain rate on logarithmic axes, and the slope of the fitted line yields m. Alternatively, rate-jump tests, in which the rate is stepped during a single test, provide m directly from the instantaneous stress response and are less sensitive to specimen-to-specimen variation.
Can results from tests performed at different speeds be compared directly?
Direct comparison is not recommended. Differences in rate can shift yield strength by several percent per decade of strain rate and alter elongation indicators measurably, which can mask or mimic real material differences. When comparing data from different sources, always verify that the control mode and rate ranges were equivalent, and where they were not, consider retesting at a common standardized rate before drawing engineering conclusions.