You can get some erroneous test results if you don't pay careful attention to test specification requirements and laboratory testing details.

Dual-screw electromechanical universal testing machine

A tensile test is performed to determine the strength of a metal under predictable loading conditions. Tensile test results show how materials should perform in actual service, as designed. In other words, the tensile test is a good predictor of how the part will react under real life loading. With most metals, thermal processing can alter those properties greatly so proper testing can assure that the part was processed correctly. Hardness testing using Rockwell and Brinell methods is also useful, and often is used in conjunction with a tensile test to assure quality. ASTM standards provide procedures of testing metals (see sidebar).

Fig 1 Representative stress-strain curve

The points of interest for a tensile test are illustrated on an engineering stress-strain curve (Fig. 1), including tensile strength (TS), or ultimate tensile strength (UTS) as the maximum load value; offset yield strength (OYS), Young's Modulus (E) and total elongation, the point where the specimen fractures.

## Basic definitions

The stress-strain curve is a graphical description of the amount of deflection under load for a given material. Engineering stress (S) is given by equation 1 where P is the load at any given time and Ao is the cross-sectional area of a specimen before testing. Engineering strain (e) is given by equation 2 where l(f) is gage length after the test and l(0) is the original gage length of the specimen.

Young's modulus (E), also called Modulus of Elasticity, is defined Delta S/Delta e, and provides a measure of a metal's stiffness, below the proportional limit. Young's modulus is the slope of the curve in the linear stress region; that is, E = (S2-S1)/(e2-e1) as defined by the ASTM method. It commonly is used in designing structures to evaluate the amount of allowable deflection under service loads.

Offset yield strength is defined by the ASTM method required for the material being tested. In a typical metals tensile test, OYS is taken at the point where strain = 0.2% strain offset from initial proportional portion of the test curve. Failure to measure this offset strain accurately can result in erroneous test results.

## Real-life challenges in reporting results

In the laboratory, there are many things that can trip you up, and if you are signing the report, always question the assumptions and methods used to obtain the results. Following is a list of factors to consider when determining if your results are correct. Start with right methods

Check the ASTM test method or other test specification and be sure the correct test speeds, loading profile and calculations are being used. Many times, it has been discovered that a technician was not following the test specification. A frequent response as to why is "This is the way we have been doing it for years." This can be avoided by studying and understanding the specification and being familiar with its important points to discuss with operators and correct their procedures if they are in error. Know your testing machine

ASTM requires that load and strain measuring devices are calibrated annually. More frequent calibrations should be performed if a device has been damaged or is subjected to excessive use. Calibration records should be kept to determine when it's time to replace or refurbish a device. Each device has its own certified range and, therefore, results should not be reported outside certified ranges.

Fig 2 Hydraulic universal testing machine

## Basic testing machine design

Testing machines can be either hydraulic driven (Fig. 2) or screw driven. This article discusses only static or quasi-static machines. In simplest terms, the test load is applied through a frame and a drive system that sets the testing speed. Hydraulic system speeds can be set using an operator-adjustable needle valve that controls the flow of fluid, or by using a servovalve that is under the control of closed-loop electronics.

Electromechanical screw-driven machines can have one to four screws depending on capacity, and these are driven by a variable speed motor through gearing or belt-drives, under closed-loop control electronics. Characteristics of static testing machines are given in Table 1.

In general, hydraulically driven frames are cost-effective for higher load tests (above 60,000 lbf, or 267 kN) and screw-driven machines are used for lower loads and longer crosshead displacements. Both can use electronic load cells for reading the applied force. However, some hydraulic machines use a transducer that measures the hydraulic piston pressure, which then is converted into applied test force.

## How test machines affect results

The shape and magnitude of the stress-strain diagram can be affected by the test speed. Some materials show an appreciable increase in strength using faster test speeds. Make sure the load rate is in accordance with the specific test method. Worn machine components can result in misalignment, which creates bending stresses that lower tensile stress readings. Check the test machine's alignment and play to ensure concentricity of the crosshead over the full travel. With the advent of microprocessor-based test systems, applied loads can inadvertently be "zeroed out," resulting in lower stress readings. This can be prevented by clamping the specimen in the upper grip, "zeroing" the load, and closing the lower grip.

## Potential pitfalls in measuring strain

When testing most metals, strains usually are too small to be measured using testing machine crosshead or piston displacement. Measuring small strains typical of a high-strength metals test (<0.0001 in.) is the task of an extensometer and strain-measuring electronics.

Select an extensometer having a gage length, travel range and magnification ratio appropriate for the metal being tested and anticipated elongation. Standard sizes are one-, two-, and eight-inch gage lengths. Extensometer travel typically is I5, 10, 20, 50 and 100%. Stiffer metals require less travel, and an extensometer having too much travel can have insufficient resolution to read correctly. Many test methods require extensometers to have certain accuracy characteristics (see ASTM E83).

The extensometer must be mounted securely onto the specimen at the start of a test. Extensometers are held on the specimen using springs, clips and elastic bands, and they should not be old or weakened from excessive use. Slippage is a common cause of errors in tensile testing.

Mechanical stops define initial gage length. Care must be taken when mounting an extensometer to establish the initial gage length, otherwise reading errors can result.

It pays to examine the "knife edges" of an extensometer to ensure they are properly installed and not worn from excessive use. Again, slippage can result in reading errors.

Match the knife edges to the specimen. When extensometers are used on different size or shape specimens, knife edges having the proper dimension and configuration must be installed to accommodate flat or round specimens.

Periodically check extensometers for worn knife edges or clips and springs. Also be sure there is no physical damage such as bent arms or frayed wires.

Fig 3 Vee wedge grips for a Baldwin testing machine

## Gripping metal specimens

Wedge action grips are the most common style used in metals testing (Fig. 3). As the axial load increases, the wedge serves to increase the squeezing pressure applied to the specimen. Wedge grips are manually, pneumatically and hydraulically actuated. Pneumatic or hydraulic actuated grips are recommended for high-volume testing. Worn or dirty grip faces can result in specimen slippage, which often renders the stress-strain diagram useless. The grip faces should be inspected periodically. Worn inserts should be replaced and dirty inserts cleaned using a wire brush. Proper alignment of the grips and the specimen when clamped in the grips is important. Offsets in alignment will create bending stresses and lower tensile stress readings. It may even cause the specimen to fracture outside the gage length. Some test machines require backlash nuts to hold the grips in place. The backlash nuts should be tightened while a specimen loaded to machine capacity is installed in the machine.

## Test specimen considerations

Most ASTM or similar test methods require a shaped specimen, which will concentrate the stress within the gage length. If the specimen is improperly machined, fracture could occur outside the gage length, resulting in strain errors. Improper reading of specimen dimensions will create stress measurement errors. Worn micrometers or calipers should be replaced and care should be taken when recording specimen dimensions. Some computer-based test systems will read the micrometer or caliper directly, thus eliminating data entry errors. Although most tests are conducted at ambient temperature, some specifications require materials to be tested at a specific temperature. Metals properties are affected by temperature which, if ignored, can result in testing errors.

## Can you rely on software to do it all?

Today's software algorithms are good at calculating the correct results for a given stress vs. strain curve. It is usually the data input into the algorithm (ie. the shape and magnitude of the stress vs. strain curve) that produces the bad results. That is why the test technician's job is so important. The technician must be responsible for performing the test according to specification. He/she should be aware of the potential errors introduced by the machine, extensometer, grips and specimen irregularities; and should alert the lab supervisor when problems arise. In short, the technician should be trained in correctly generating the stress vs. strain curve of Figure 1 for a given test method. One indirect way of checking your results is to send specimens to at least two other testing laboratories so that you can compare results.