Tensile Testing

Every stress-strain graph in this topic came from somewhere real: a specimen, gripped at both ends, pulled slowly and steadily apart while a machine records exactly how hard it's pulling and how far the material has stretched. That procedure is tensile testing, and it's worth understanding as an experiment in its own right, not just as the source of a graph shape.

Apparatus and method

A tensile test is the standard experimental procedure used to obtain the stress-strain data covered in the previous lesson. A specimen of the material under test — often manufactured into a standardised dumbbell (dog-bone) shape, with a narrower, uniform central section of precisely known length (the gauge length) and wider ends for gripping — is clamped securely at both ends into a tensile testing machine.

The machine then steadily stretches the specimen at a controlled, constant rate, applying a continuously increasing tensile force along its axis. Throughout the test, the machine records two quantities simultaneously:

• Applied force, measured directly by a load cell built into the machine's grips.
• Extension, measured over the specimen's gauge length specifically — often using a device called an extensometer (or a strain gauge) clipped or bonded directly onto the gauge section, rather than simply tracking the movement of the machine's grips, which would also pick up any slack or slipping at the clamps and give a less accurate reading of the specimen's actual extension.

From force-extension data to a stress-strain graph

The machine directly measures force and extension, but as established earlier in this topic, it's stress and strain — not raw force and extension — that describe the underlying material behaviour independent of the specific specimen tested. Converting the raw data is simply a matter of applying the definitions from the first two lessons of this topic:

where $A_0$ is the specimen's original cross-sectional area (measured before testing begins) and $L_0$ is its original gauge length. Every force reading from the test is divided by $A_0$ to give a stress value, and every extension reading is divided by $L_0$ to give a strain value — producing the stress-strain graph examined in the previous lesson, point by point, directly from the raw test data.

Why the specimen shape and size matter

Using a standardised specimen shape and size is not a minor procedural detail — it's what makes tensile testing useful for comparing materials and results at all. If different labs tested wildly different specimen shapes, or used different gauge lengths, their raw force-extension curves would not be directly comparable, even for identical materials, simply because of geometry. Standardising the specimen removes this variable, and — combined with the normalisation to stress and strain — ensures that the resulting graph genuinely reflects the material's properties, not the arbitrary choices made in preparing that particular test piece. This is precisely the same principle that motivated defining stress and strain in the first place, applied here at the level of experimental procedure.

Tensile tests are also run all the way to complete fracture, not stopped once the material starts to yield or even once it passes its ultimate tensile stress. Running the test to failure is essential to capture the material's full behaviour — its UTS, its necking behaviour, and its breaking stress — all of which are needed to fully characterise the material and are used directly in later lessons on ductile/brittle behaviour and fracture.

Mark as complete