The last article talked about stress, strain and what they mean. This article continues the discussion by talking about how strength tests are performed, some of the considerations and how the data can be interpreted.

To ensure consistency between tests, the American Society for Testing Materials (ASTM) International publishes standards that mimic most real-life loading situations. There are ASTM tests for determining the force required to pull a grommet out of a tarp, or how much shear loading an adhesive can withstand before failure, as well as many other such standards.

Tensile and compression tests

Tensile tests stretch material and compression tests try to squash material. Metals are equally strong in tension and compression, generally speaking, so tensile tests are perhaps more common.

The tensile test machine has two separate jaws, one of which will move away from the other at a controlled rate during the test. As it moves away, it is pulling on the material, stretching it until it fails. The controlled rate is called the “strain rate,” and materials will behave differently under different strain rates. Strain rates must be specified when comparing materials.

A tensile test machine that is holding a sample with clamps. The top jaw moves, stretching the material. Source: Kerina yin/Public domain

Compression tests are often performed with a hydraulic system that forces a ram down on a cylindrical sample. As the load increases, the cylinder will “barrel” outward, eventually failing. Compression tests are more common for ceramic materials, such as concrete, cement, asphalt, stone and other building materials that are often loaded in compression during everyday use.

Regardless of the test, the control system detects when the material has failed. Once the material fails, the controller stops applying the load, keeping the jaw from moving quickly and dangerously.

To gather true stress and strain, an extensometer is placed on the sample to measure how much it is strained, especially after the yield stress has been reached and the material begins to plastically deform. Some systems will also be equipped with a laser or machine vision system to determine the dimensions of the cross-sectional area during necking.

Interpreting the curve

The strain is applied by the tensile test machine, and the stress is measured. Initially, this will be a linear relationship, as shown in the graph, where the slope is known as the Young’s Modulus. This can be used to determine the stiffness of the material.

An engineering stress-strain curve. Source: Breakdown/CC BY-SA 3.0

Consider a stretchy foam. The slope would be very shallow and the Young’s modulus small, as the elongation would be large, but the material would easily stretch. For a stiffer metal, the slope will be much higher.

For brittle materials, the yield strength, ultimate tensile strength (UTS) and the fracture strength may be very close together, or even immeasurably close. For ductile materials, the sample will yield, then strain-harden, which means it will get a little stronger with additional load, before reaching the UTS. At the UTS, the sample will begin to neck, reducing the cross-sectional area. The true stress will increase, but because the sample’s dimensions have rapidly changed, engineering stress will drop. This is a function of ease in measurement. The reality is that the stress continues to grow after UTS.

The area under the curve is called the toughness of the material. Strong, ductile materials will have a lot of area under this curve, but weak or brittle materials will not. Tough materials can be plastically deformed before failing catastrophically, making them a good choice for “leak-before-break” applications.

Fracture surfaces

The tensile test can also qualitatively reveal more information about the material. Brittle failures tend to be shiny and granular, with little or no necking. Ductile materials tend to fail in a “cup and cone” appearance, with the sample necking extensively. Once the ductile material fails, one side will look like a cone and the other like a cup.

This cast iron tensile test sample broke in brittle fashion. Notice the lack of necking and the grainy surface texture. This particular sample was screwed into the test apparatus rather than clamped, as is often the case with strong materials. Source: Sigmund/CC BY-SA 3.0

For composites, fracture surfaces can reveal which component failed. Consider a fiber-reinforced material. If the fracture surface has a bunch of fibers sticking out (called fiber pullout), the interface between the fiber and the matrix is what has failed. This often means the fibers are making the material stronger. However, if the material looks like it fails evenly across the matrix and fibers together, this can be an indication that the fiber is not making the material stronger at that strain rate.

Factors that affect results

The ASTM standards specify sample geometries, surface treatments and other conditions for standardized testing. However, other factors may cause results to vary between tests, such as sample temperature and critical flaw size.

Temperature

Temperature is a big consideration during strength testing. For accurate strength measurements, testing should occur at the service temperatures experienced by the final product. This is particularly a concern for polymers and composites, as many of them are especially sensitive to small changes in temperature, whereas metals and ceramics do not typically lose strength until much higher temperatures. However, some metals undergo a ductile-to-brittle transition at lower temperatures.

To accommodate testing under different temperatures, portable furnaces can be placed over the sample under test. The ones designed for use with the tensile test machine are best, as they are designed to handle the violent breaking motion of the sample. An improvised furnace may be damaged during the test. For testing at lower temperatures, samples are placed in a cooler, freezer or liquid nitrogen and removed and mounted before testing begins.

Critical flaw size

Besides temperature, critical flaw size can cause variation in testing data. If there is a small crack, pore or impurity in the material, it will become a stress concentrator, reducing the amount of force that can be applied. In the strictest sense, consider a large, flat, internal crack, invisible to the naked eye. It reduces the cross-sectional area, but would not be easily detected or accounted for during manual calculations. Ceramics are particularly sensitive to this failure mode, as they are not as strong in tension, and a small pore or impurity during manufacturing will lead to failure.

To combat the effects of critical flaw size, samples are inspected before test to look for nicks, machine marks, scratches, pores, discolorations or anything else that might indicate a problem. Furthermore, samples that fail in the machine’s grips indicate a critical-flaw related failure, not actual tensile data. Furthermore, tests are conducted in high enough numbers (at least n=30) to perform statistical analysis on the data. Weibull distribution is a popular way to view tensile testing in ceramics for this reason.

Final thoughts

There are hundreds of procedures for testing materials, all of which are defined by the ASTM standards. This allows for accurate comparison between different materials so that engineers can make the best design decisions. By choosing the proper standard, performing the test, and correctly interpreting the data, the best materials can be chosen.