A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
The tensile test is the most fundamental mechanical characterization method, but extracting meaningful design parameters from the raw load-displacement data requires understanding two different stress-strain formulations. Engineering stress and strain use the original specimen dimensions (sigma_e = F/A_0, epsilon_e = deltaL/L_0) and are straightforward to calculate, but they give misleading results after necking begins because the cross-section is no longer uniform. True stress and true strain account for the instantaneous dimensions (sigma_t = F/A_inst, epsilon_t = ln(L/L_0)), giving a continuously rising curve that reflects the material's actual work-hardening behavior. Before necking, the two are related by sigma_t = sigma_e(1 + epsilon_e) and epsilon_t = ln(1 + epsilon_e). Key properties extracted include: yield strength (by the 0.2% offset method for materials without a distinct yield point, or upper/lower yield points for low-carbon steels exhibiting Luders band behavior), ultimate tensile strength (the engineering stress peak corresponding to necking onset), percent elongation and percent reduction in area (ductility measures), and the elastic modulus (slope of the initial linear region). The area under the engineering stress-strain curve up to fracture represents the toughness — the energy per unit volume the material can absorb before failure.
Work through a complete tensile test dataset: convert raw load-displacement to both engineering and true stress-strain, identify the 0.2% offset yield strength, locate the UTS and necking point, and calculate elongation and reduction in area. Compare the engineering and true curves on the same plot to see where they diverge. Test or examine data for materials with distinct yield point behavior (mild steel) versus those requiring the offset method (aluminum alloy).
The tensile test stretches a standardized specimen until it breaks, recording force and displacement throughout. From the stress-strain behavior you studied as a prerequisite, you already know what the curve looks like: an initial linear elastic region, a yield point where permanent plastic deformation begins, a region of work hardening where the material gets stronger as it deforms, and eventually fracture. The tensile test is simply the experimental apparatus that generates this data — but extracting reliable, meaningful material properties from raw load-displacement output requires understanding why there are two different stress-strain formulations, and which one to use for which purpose.
Engineering stress (σ_e = F/A₀) uses the original cross-sectional area throughout the test. This is convenient — you measure A₀ once before testing — but it becomes misleading after significant plastic deformation. As the specimen stretches, the cross-section narrows (conserving volume), so the actual stress on the material is higher than σ_e reports. Before necking, the error is modest and both formulations track each other closely. After necking begins, engineering stress falls (because load F drops), suggesting the material is getting weaker. It is not — it is still work hardening, but the area in the neck is shrinking faster than the flow stress is rising. True stress (σ_t = F/A_inst) corrects for this by using the actual instantaneous area, giving a curve that rises monotonically until fracture and accurately reflects the material's work-hardening behavior. For design at small strains, engineering stress is practical and sufficient. For modeling large deformations (forming, crash simulations), true stress and true strain are essential.
Yield strength is the stress at which permanent plastic deformation begins. For metals with a clear yield drop (mild steel shows an upper yield point followed by a drop to a lower plateau where Lüders bands propagate across the specimen), reading yield strength is straightforward. For most metals — aluminum alloys, stainless steels, most high-strength steels — the transition from elastic to plastic behavior is gradual, with no distinct inflection point. The 0.2% offset method solves this: draw a line parallel to the initial elastic slope, but starting from 0.002 strain (0.2%). Where this line intersects the stress-strain curve is defined as the yield strength. The 0.002 strain offset is a convention; it represents an acceptable small amount of permanent strain for engineering purposes, not a physical threshold.
Ultimate tensile strength (UTS) is the engineering stress at the peak of the engineering stress-strain curve. It corresponds to the onset of necking — the point where plastic instability begins and deformation localizes into a narrow region. After the UTS, load decreases as the neck thins rapidly, even though the true flow stress in the neck is still rising. The UTS is widely used in design as a strength reference, but it is less fundamental than yield strength: the UTS is where the specimen becomes geometrically unstable, not where the material reaches any intrinsic limit.
Ductility captures how much plastic deformation occurs before fracture. Percent elongation (total extension divided by gauge length, expressed as percent) and percent reduction in area ((A₀ − A_f)/A₀ × 100%) both measure ductility but are sensitive to different aspects of the test geometry. A brittle material fractures with little plastic deformation, showing a steep rise and sudden drop. A ductile material shows extensive necking and elongation before fracture. Toughness — the area under the engineering stress-strain curve to fracture — integrates both strength and ductility and represents the energy per unit volume the material can absorb before failure. High toughness requires both strength and ductility, which are often in tension with each other: processing that increases yield strength typically reduces ductility.