The Tension and Compression Test

The strength of a material depends on its ability to sustain a load

without undue deformation or failure. This property is inherent in the
material itself and must be determined by experiment. One of the most
important tests to perform in this regard is the tension or compression
test. Although several important mechanical properties of a material
can be determined from this test, it is used primarily to determine the
relationship between the average normal stress and average normal
strain in many engineering materials such as metals, ceramics, polymers,
and composites.

To perform a tension or compression test a specimen of the material is

made into a “standard” shape and size. It has a constant circular cross
section with enlarged ends, so that failure will not occur at the grips.
Before testing, two small punch marks are placed along the specimen’s
uniform length. Measurements are taken of both the specimen’s initial
cross-sectional area, and the gauge-length distance between the
punch marks. For example, when a metal specimen is used in a tension
test it generally has an initial diameter of (13 mm) and a
gauge length of (50 mm), Fig. 3–1. In order to apply an axial
load with no bending of the specimen, the ends are usually seated into
ball-and-socket joints.A testing machine like the one shown in Fig. 3–2 is
then used to stretch the specimen at a very slow, constant rate until it
fails. The machine is designed to read the load required to maintain this
uniform stretching.
At frequent intervals during the test, data is recorded of the applied
load P, as read on the dial of the machine or taken from a digital readout.
Also, the elongation between the punch marks on the
specimen may be measured using either a caliper or a mechanical or
optical device called an extensometer.This value of (delta) is then used
to calculate the average normal strain in the specimen. Sometimes,
however, this measurement is not taken, since it is also possible to read
the strain directly by using an electrical-resistance strain gauge, which
looks like the one shown in Fig. 3–3. The operation of this gauge is
based on the change in electrical resistance of a very thin wire or piece
of metal foil under strain. Essentially the gauge is cemented to the
specimen along its length. If the cement is very strong in comparison to
the gauge, then the gauge is in effect an integral part of the specimen,
so that when the specimen is strained in the direction of the gauge, the
wire and specimen will experience the same strain. By measuring the
electrical resistance of the wire, the gauge may be calibrated to read
values of normal strain directly.
The Stress–Strain Diagram
It is not feasible to prepare a test specimen to match the size, and
of each structural member. Rather, the test results must be reported so
they apply to a member of any size. To achieve this, the load and
corresponding deformation data are used to calculate various values of
the stress and corresponding strain in the specimen. A plot of the results
produces a curve called the stress–strain diagram.There are two ways in
which it is normally described.

Conventional Stress–Strain Diagram. We can determine the

nominal or engineering stress by dividing the applied load P by the
specimen’s original cross-sectional area This calculation assumes that
the stress is constant over the cross section and throughout the gauge
length.We have

Likewise, the nominal or engineering strain is found directly from the

strain gauge reading, or by dividing the change in the specimen’s gauge
length, by the specimen’s original gauge length Here the strain is
assumed to be constant throughout the region between the gauge points.
Thus,

If the corresponding values of and are plotted so that the vertical

axis is the stress and the horizontal axis is the strain, the resulting curve is
called a conventional stress–strain diagram. Realize, however, that two
stress–strain diagrams for a particular material will be quite similar, but
will never be exactly the same.This is because the results actually depend
on variables such as the material’s composition, microscopic
imperfections, the way it is manufactured, the rate of loading, and the
temperature during the time of the test.
We will now discuss the characteristics of the conventional stress–strain
curve as it pertains to steel, a commonly used material for fabricating
both structural members and mechanical elements. Using the method
described above, the characteristic stress–strain diagram for a steel
specimen is shown in Fig. 3–4. From this curve we can identify four
different ways in which the material behaves, depending on the amount
of strain induced in the material.

Elastic Behavior. Elastic behavior of the material occurs when

the strains in the specimen are within the light orange region shown in
Fig. 3–4. Here the curve is actually a straight line throughout most of
this region, so that the stress is proportional to the strain.The material
in this region is said to be linear elastic. The upper stress limit to this
linear relationship is called the proportional limit, If the stress
slightly exceeds the proportional limit, the curve tends to bend and
flatten out as shown.This continues until the stress reaches the elastic
limit. Upon reaching this point, if the load is removed the specimen
will still return back to its original shape. Normally for steel, however,
the elastic limit is seldom determined, since it is very close to the
proportional limit and therefore rather difficult to detect.
Yielding. A slight increase in stress above the elastic limit will result
in a breakdown of the material and cause it to deform permanently.
This behavior is called yielding, and it is indicated by the rectangular
dark orange region of the curve.The stress that causes yielding is called
the yield stress or yield point, and the deformation that occurs
is called plastic deformation. Although not shown in Fig. 3–4, for lowcarbon
steels or those that are hot rolled, the yield point is often
distinguished by two values. The upper yield point occurs first,
followed by a sudden decrease in load-carrying capacity to a lower
yield point. Notice that once the yield point is reached, then as shown
in Fig. 3–4, the specimen will continue to elongate (strain) without any
increase in load.When the material is in this state, it is often referred to
as being perfectly plastic.

True Stress–Strain Diagram. Instead of always using the original

cross-sectional area and specimen length to calculate the (engineering)
stress and strain, we could have used the actual cross-sectional area and
specimen length at the instant the load is measured. The values of stress
and strain found from these measurements are called true stress and true
strain, and a plot of their values is called the true stress–strain diagram.
When this diagram is plotted it has a form shown by the light-blue
curve in Fig. 3–4. Note that the conventional and true diagrams
are practically coincident when the strain is small. The differences
between the diagrams begin to appear in the strain-hardening range,
where the magnitude of strain becomes more significant. In particular,
there is a large divergence within the necking region. Here it can be
seen from the conventional diagram that the specimen actually
supports a decreasing load, since is constant when calculating
engineering stress, However, from the true diagram,
the actual area A within the necking region is always decreasing until
fracture, and so the material actually sustains increasing stress,
since s = P>A.

Although the true and conventional stress–strain diagrams are

different, most engineering design is done so that the material supports
a stress within the elastic range. This is because the deformation of the
material is generally not severe and the material will restore itself
when the load is removed. The true strain up to the elastic limit will
remain small enough so that the error in using the engineering values
of and is very small (about 0.1%) compared with their true values.
This is one of the primary reasons for using conventional stress–strain
diagrams.