Tensile strength is measured by pulling a specially shaped specimen in a testing machine until it breaks, then dividing the maximum force recorded by the specimen’s original cross-sectional area. The result tells you the most stress a material can withstand before failure, reported in megapascals (MPa) or pounds per square inch (psi). The process sounds simple, but getting an accurate number depends on specimen preparation, machine setup, strain measurement, and testing speed.
The Core Formula
The calculation itself is straightforward: divide the maximum load the specimen sustained by its original cross-sectional area. If a round metal bar with a cross-section of 100 square millimeters broke under a load of 50,000 newtons, its ultimate tensile strength is 500 MPa. You always use the original area, not the narrowed area at the moment of fracture, because the standard measurement convention is based on “engineering stress” rather than “true stress.” This distinction matters: the specimen thins considerably before it snaps, but the original dimensions are what get plugged into the formula.
Equipment You Need
The standard tool is a universal testing machine (UTM). It has three essential components. The crosshead moves up or down, powered by an electric motor or hydraulic system, to pull the specimen apart. A load cell attached to the crosshead provides a real-time force reading throughout the test, calibrated to an accuracy of plus or minus 1% of the indicated value. And grips hold the specimen in place during the pull.
Grip selection depends on the material and the forces involved. Wedge grips, pneumatic grips, and vice grips are all common. The wrong grip is one of the most frequent sources of error in tensile testing: if the specimen slips even slightly, the machine records extra travel that inflates your elongation measurement and distorts the stress-strain curve. For brittle materials or thin films, specialized fixtures prevent crushing the sample before the test even starts.
Preparing the Specimen
You can’t just clamp a random chunk of material and pull. Specimens need a precise, standardized shape so results are comparable across labs. For metals, the governing standard is ASTM E8. The most common specimen is a flat “dogbone” shape, wide at the ends where the grips hold it and narrow in the middle where failure is supposed to occur. A standard flat dogbone has a width of 12.5 mm and a gauge length (the measured section in the middle) of 50 mm. Round specimens are also common, typically with diameters between 6.35 and 12.7 mm and gauge lengths of 25.4 to 50.8 mm.
Plastics follow a different standard, ASTM D638, which uses dumbbell-shaped specimens and accommodates materials up to 14 mm thick. Anything thicker gets machined down. For thin films under 1.0 mm, a separate standard (D882) applies because the gripping and handling requirements are completely different. The key point is that tensile properties in plastics are highly dependent on how the specimen was prepared, the testing speed, the temperature, and even the material’s history, so following the standard precisely is more critical than it might seem.
Running the Test Step by Step
A tensile test follows a consistent sequence regardless of the material being tested:
- Condition the specimen. Some standards require holding the sample at a specific temperature and humidity for a set period before testing, especially for plastics and polymers.
- Calibrate the machine. Zero the load cell and verify that the crosshead displacement reads zero with no load applied.
- Mount the specimen. Center it carefully in the grips. Misalignment introduces bending forces that act on only part of the cross-section, skewing results and potentially damaging equipment.
- Set the test parameters. This means choosing the crosshead speed (or strain rate) specified by the relevant standard for your material.
- Start the test. The crosshead moves, the specimen stretches, and the machine continuously records force and displacement.
- Continue to fracture. The test runs until the specimen breaks. The machine logs the complete force-displacement history.
- Inspect and report. Examine the fracture surface, calculate stress and strain values, and document the results.
How Strain Gets Measured
Force measurement is the easy part: the load cell handles it directly. Strain measurement is where things get tricky. The simplest approach is to track how far the crosshead travels, but this is surprisingly inaccurate. Not all crosshead movement translates into specimen deformation. Some motion goes into compressing the grips, flexing the load cell, and deflecting the test frame itself. Subtracting out this “system compliance” is difficult, and even corrected crosshead displacement values carry high uncertainty.
For accurate strain data, you need an extensometer: a device that clips directly onto the specimen’s gauge length and measures only the stretch happening in that section. Contact extensometers use knife edges that rest against the specimen surface. Non-contact options, like video extensometers and laser systems, track strain without touching the sample at all, which is useful for delicate materials or tests at extreme temperatures. If your goal is just to find ultimate tensile strength (the peak force divided by area), crosshead displacement can work. But if you need a reliable stress-strain curve, yield strength, or elongation values, an extensometer is essential.
Reading the Stress-Strain Curve
The machine’s output is a stress-strain curve: stress (force divided by area) on the vertical axis, strain (change in length divided by original length) on the horizontal axis. Several key points on this curve each tell you something different about the material.
The curve starts as a straight line. In this region, the material is elastic, meaning it snaps back to its original shape if you release the load. The slope of this straight section is the elastic modulus, a measure of stiffness. The point where the curve stops being straight is the proportional limit. Very close to it (often identical in practice) is the elastic limit, the last point at which the material fully recovers its shape.
Just beyond that is the yield strength, the stress at which permanent deformation begins. Because pinpointing the exact onset of plastic deformation is difficult, yield strength is conventionally defined using a 0.2% offset method: you draw a line parallel to the initial straight portion of the curve, starting at 0.2% strain on the horizontal axis, and the point where that line intersects the curve is the offset yield stress.
As strain continues to increase, the curve rises more gradually until it reaches its highest point. That peak is the ultimate tensile strength, the single number most people mean when they say “tensile strength.” Beyond this peak, the material begins to neck (thin locally), the engineering stress drops, and eventually the specimen fractures.
Why Testing Speed Matters
The rate at which you pull the specimen changes the results, sometimes dramatically. This effect is especially pronounced in plastics and fiber-reinforced composites. In carbon fiber-reinforced polymer laminates, for example, tensile strength is sensitive to strain rate while stiffness often is not. In one study of composite specimens oriented at 90 degrees to the fiber direction, tensile strength increased by 114% when the strain rate jumped from a slow quasi-static pace to a moderate dynamic rate. Even at 45-degree orientations, the increase was 36% over the same range.
Metals are generally less sensitive to strain rate at room temperature, but the effect still exists. This is why every tensile testing standard specifies a crosshead speed or strain rate: it ensures that a value measured in one lab can be meaningfully compared to a value measured in another. Deviating from the specified speed, even unintentionally, can make a material appear stronger or weaker than it actually is under standard conditions.
Common Sources of Error
Three problems account for most inaccurate tensile test results. Grip slippage is the first: if the specimen slides in the jaws during the test, the machine records more travel than the specimen actually experienced, producing false elongation values. Using grips rated for the force range of your test and checking that the specimen hasn’t shifted after the test are basic safeguards.
Axial misalignment is the second. If the specimen isn’t perfectly centered, it experiences a combination of tension and bending rather than pure tension. This concentrates stress on one side of the cross-section, leading to premature failure and an artificially low tensile strength reading.
The third is uncompensated system deflection. Every component between the crosshead and the specimen, including the load cell, grips, and frame, flexes slightly under load. If you’re relying on crosshead displacement for strain data and don’t account for this machine compliance, your strain values will be inflated. An extensometer mounted directly on the gauge length bypasses this problem entirely.