Mechanical Testing

Mechanical testing is a general term which refers to a broad range of activity involved with
the determination of mechanical properties and behavior of material, structures and machines.
Typically, this involves applying some type of stimulus and measuring a response. The stimulus
may be either static or dynamic in nature and may act in a localized or a distributed manner at
single or multiple locations. In a similar way, the response measurement may be made at one or
more points or in a distributed manner.
Expanding on this definition, mechanical testing is the process of applying forces, pressures,
displacements, heat, or similar stimuli to a mechanical system and then measuring its response.
The objective, or course, is to characterize the behavior of the system either in order to obtain
assurance of the performance or to verify the one hand, mechanical testing may be carried out
primarily to verify the overall performance of a mechanical system, i.e., the load carrying
capability of an aircraft wing, or the thrust production of a rocket motor. On the other hand,
mechanical testing may be used to verify the appropriateness or accuracy of a particular theory.
In this case, the testing is usually directed at a simplified component or model that may actually
bear no outward resemblance to a practical design. A third case could be defined to include
those tests carried out primarily to characterize the mechanical properties of a new material. As
with the second case, the testing usually involves working with rather abstract specimens which
bear no resemblance to a practical product.
In all of these situations, the test methodology may be either deterministic or probabilistic in
nature. In the first case, the approach is to carry out thorough and extensive tests with a single or
at most a very few specimens which are assumed to be representative of a larger set. This is
often preferred when the specimen is costly or the test setup is complex. The other approach is
to use a large number of nominally identical specimens which are assumed to constitute a
suitable subset or sample of the larger set of all possible specimens. Statistical methods are then
used to analyze the data and to draw conclusions about the properties of the whole set.
In the Structures Lab courses, the major emphasis will be directed towards examining the
various techniques, instruments and interpretative methods for carrying out mechanical testing.
At the same time, this knowledge will be used to perform various mechanical tests for the
purpose of verifying the theories developed in the companion courses. Relatively little emphasis
will be placed on examining methods for overall performance testing.

Test Machines
Test machine or testing machines is the name used to refer to any one of a variety of devices
for applying a controlled load to a specimen. The loads may range from a few grams to millions
of pounds using machines currently available. The loads may either be static in nature (changed
slowly from one level to another), or may vary rapidly at rates up to tens or hundreds of cycles
per second (Hertz).
In general there are two principal components to a test machine: a loading mechanism, and a
weighing or load-measuring system. A number of different approaches have been used in the
design on each of these components, and consequently, it is logical to describe test machine
designs by examining each component separately. The section will be concluded by describing
the characteristics of several of the test machines available in the Aerospace Engineering
laboratories.

AE3145 Test Machines Page 1

Loading Mechanisms
Load generation in most test machines is done either hydraulically or mechanically.
Electromagnetic and thermal techniques have been employed in a few specialized cases.

Hydraulic:
Hydraulic loading mechanisms use a piston in a cylinder actuated by pressurized oil or
hydraulic fluid. As shown in Fig. 1, the cylinder is bored in a large massive structure which
forms the stationary base of the test machine. For simplicity and ruggedness, the piston is design
for unidirectional operation, that is, for applying load in only one direction. It is connected to a
moving table on which the test specimen can be placed. A fixed crosshead is supported above on
columns attached to the base so that a compressive load can be applied to a specimen located
between the table and crosshead. The test machine “daylight space” is the maximum vertical
opening between the table and crosshead. Typically, the maximum piston stroke is about 2”, and
in order to accommodate a variety of specimen heights, the fixed crosshead can be adjusted
vertically. This is usually done by using large threaded rods for the supporting columns and
fitting motorized nuts in the crosshead. Tension loads can be applied if a driven crosshead
attached to the table and piston is located above the fixed crosshead. Specimens are then
attached between the fixed and driven crossheads. This type of design is frequently referred to
as a Universal Test Machine because of its ability to generate either tension or compression
loads. Hydraulic design are usually the least expensive in the larger machine sizes and are
capable of generating the highest loads available.

Figure 1. Universal Hydraulic Testing Machine

The load control is obtained by regulating the hydraulic pressure in the cylinder beneath the
piston. This is done by means of two needle type flow control valves. One valve is connected
between pump and the cylinder and is used to control the buildup of pressure. The other is
connected between the cylinder and the reservoir and is used to control the bleedoff of pressure.
The two valve functions are sometimes incorporated in a single control valve, or in other designs
the control unit is simply fitted with separate “loading” and “unloading” valves. Due to the
double acting nature of the controls and the small leakage always present around the piston, it is

AE3145 Test Machines Page 2

nearly impossible to maintain a constant load or crosshead position. However, the load can
usually be applied smoothly especially at high levels.
The newest type of hydraulic test machine is the electrohydraulic design. The load is
generated in the same manner as above but the control is accomplished by means of an
electrically activated 4-way valve. A typical configuration is shown in Fig. 2. Due to the
reversible 4-way control valve, it is now possible to use a double acting hydraulic piston or
“actuator” which is capable of forcefully advancing or retracting. This is accomplished because
the control valve can continuously vary the flow from the pump to either side of the piston while
connecting the opposite side to the tank. As a result, it is not necessary to provide separate
tension and compression spaces. Both tensile and compressive loads can be applied to the
specimen without having to relocate or reattach it. In fact, the load can obviously be reversed
from tension to compression or vice versa during the same test.

Figure 2. Typical Electrohydrualic Test machine

Control of an electrohydraulic machine is considerably more complex but the resulting

capabilities are far greater. Since the valve is electrically operated, it can readily be controlled
by electronic circuitry or even a computer. The typical control systems used a concept called
feedback control that is commonly employed in a large number of process and industrial control
applications as well as for many purely electronic circuit designs. In this approach, an electrical
signal generated by the weighing system is fed back to the control system where it is compare
with a command signal representing the desired load. If the difference or error signal is zero, the
valve position remains unchanged, but if it is nonzero, the valve will be electrically driven in a
direction to reduce the error. With proper stable adjustment, the load will closely follow the
valve specified by the command signal which itself may be generated by a manually operated
knob, a function generator, a computer, or a previously recorded signal (on magnetic tape, for
example). In typical test machine designs, the valve can be driven at frequencies up to several
tens of Hertz so that dynamically varying loads can be generated. This is invaluable for fatigue
and fracture testing. On a more general note, the feedback signal need not be proportional to
load but can be derived from actuator displacement, velocity, or to specimen strain, for example.
These variables then become the controlled or independent variable in the test. As a final note,
this same technology was originally developed to provide boost forces for aircraft control

AE3145 Test Machines Page 3

surfaces in the early 1950’s and is used today in “fly-by-wire” designs in which the only pilot
connection to a control surface is by means of the electrical command signal.

Mechanical:
Mechanical loading mechanisms employ a screw system to generate a load in much the same
way as a screw-jack operates. In the most common configuration, as shown in Fig. 3, the load is
produced by the movement of a driven crosshead towards or away from a stationary crosshead or
the machine table. The driven crosshead is activated by threaded supporting columns which are
rotated by a motor and gear drive mounted in the machine base. The figure shows a “universal”
configuration with separate spaces for compression and tension. Most modern machines use a
single opening and rive the crosshead in either direction. This is simpler but requires special
nuts in the crosshead which do not exhibit any mechanical hysterisis or slack when the load is
reversed. Mechanical or screw-jack type machines are generally limited to less than about
500,000 lbs. capacity and are more costly than hydraulic machines in the larger capacity ranges.
Mechanical designs are often cheaper in the small sizes of less than about 2000 lbs. capacity.
Mechanical machines are somewhat harder to control at very high loading levels but they can
hold a constant crosshead position easily and thus can be used to maintain loads for long periods
of time.

Figure 3. Typical Mechanical (Screw-jack) Test Machine

Control of mechanical loading mechanisms can be accomplished in the simplest case by

using a reversible variable speed motor with mechanically or electrically selectable gear ratios.
A more complex approach involves using a form of synchro control called and “amplidyne”.
The shaft position of a synchro motor is slaved to a much smaller synchro transmitter which
provides the control input. As the shaft of the synchro transmitter is rotated manually, the slaved
synchro motor follows exactly. This motor is connected through gearing to drive the screw
columns. This makes for simple and convenient manual control. Automatic control can be
provided by rotating the transmitter with a small geared clock motor or else the synchro signals
can be electrically created in a computer controlled synchro synthesizer.

AE3145 Test Machines Page 4

Other:
Many other techniques have been used to produce controlled forces but few are commonly
used in test machine designs. Two techniques deserve mention, however. Electromagnetic
actuators similar in concept to audio speakers have been employed to generate forces for test
purposes. The single advantage in this approach is the dynamic performance obtained, but this
has in recent designs been nearly equalled by electrohydraulic machines. The disadvantage is
that maximum forces are typically in the range of a few hundred pounds. Another approach
involves using the thermal expansion of a confined material to produce a force. A suitable
material is slowly heated or cooled to produce an expansion or contraction. If it along with the
specimen is confined between two rigid crossheads, the result will be a compressive or tensile
load in both the specimen and the driving material. This approach resembles the mechanical
screw type machines, but it is capable of much slower loading rates with little of the hysterisis
often encountered in mechanical machines because of the screw/nut interface.

Weighing Mechanisms
Weighing or load measuring mechanisms are employed in test machines to indicate the
magnitude of the force generated. In contrast to the case for loading mechanisms, weighing
systems are considerably more varied in concept although the current designs almost universally
utilize electrical resistance strain gage load cells. The different approaches are described in the
following sections.

Mechanical:
Mechanical weighing systems employ an arrangement of beams, levers, and balances to
indicate load in much the same way that balance beam platform weighing scales operate. The
schematic in Fig. 3 shows a simple balance beam and poise system used with a screw type
loading mechanism. This requires manual positioning of the pose and is very cumbersome. The
approach is more commonly designed with an automated means for positioning the poise. One
method is to attach the poise to a threaded rod or lead screw running the length of the balance
beam and driven by a small electric motor. The position of the beam is sensed electrically and if
not in balance, current is supplied to the motor to move the poise in a direction to return to a
balanced state. This principle of measurement in which the system is always returned to (or
maintained in) the same balanced state is often referred to as a “null balance” measurement and
can be implemented in either mechanical or electrical terms. The balance beam approach often
yields high accuracies and by change of poise weight can easily be extended in range. It is,
however, quite slow. The concept is often used in wind tunnels to measure model forces (the
systems are referred to as “tunnel balances”).
The balance beam can be replaced with a directly calibrated indicating system. In this
approach the scaled down force proportional to load that is transmitted by the beam is applied to
a load sensing element and the resulting deflection is used to indicate the load. A common
method is to use a pendulum which swings up in response to a moment generated by the beam
force. The pendulum position is indicated on a large load dial. Elastic members such as
cantilevered beams and torque tubes can also be used. In these cases the load is indicated by the
resulting elastic deformation produced by the scaled forces.

AE3145 Test Machines Page 5

Hydraulic:
Hydraulic weighing systems are of two main types. The first type is usable only with
hydraulic loading systems and simply involves measuring the fluid pressure in the cylinder
beneath a single acting piston. The pressure is sensed by a conventional pressure gage whose
dial is calibrated in units of load. (The most common gages, called bourdon tube types, employ a
hood shaped thin tube which when pressurized tends to uncurl and by means of gears and
linkages to rotate a dial pointer.) This direct approach is by far the simplest, but due to friction
forces between the piston and cylinder, the maximum in accuracy may not be obtained. This
concept is illustrated in Fig. 1. Note that the method cannot be used with bi-directional cylinders
or actuators in which an opposing pressure may be acting on the other side of the piston.

Figure 4. Hydraulic Test Machine with Load Weighing Capsule

The other type of hydraulic weighing system is completely separate from the loading
mechanism. A hydraulic capsule (a sealed bellows, for example) is placed somewhere in the
load path and the internal pressure is sensed by a pressure gage or transducer. As shown in Fig.
4, the capsule can be located in a position to measure the full load reacted by the stationary
crosshead. The use of a capsule avoids the problem of piston friction noted above and a similar
simple bourdon tube pressure gage can also be used to indicate the load.
For either type, several variations in the load indicating system have been developed. One
approach is actually a combination of hydraulic and mechanical techniques. The hydraulic
pressure is used to activate a small piston which in turn deflects a balance beam. The beam can
be automatically balanced as noted earlier or it can be coupled to a pendulum. This approach is
often used because various indicator ranges can be incorporated simply by varying the balance
beam fulcrum location. Another approach involves using a bourdon tube in a null balance
configuration. A pneumatic actuator (bellows) is connected to the tip of the bourdon tube and is
used to hold the tip at its undeflected or null position. Thus, as the load increases, more force
must be exerted by the bellows to maintain the bourdon tube tip at the null position. The load
indication is derived from the deflection of an elastic member in the bellows load path. This

AE3145 Test Machines Page 6

design is called a “Tate-Emory” weighing system and has the advantage of allowing or change in
fluid volume takes place in the bourdon tube.
Strain Gage Load Cells: By far the most popular approach used in current test machine
weighing systems is to employ strain gage type force transducers (or load cells) to measure the
loads. These units will be described in much greater detail in subsequent sections when strain
measuring techniques are discussed. Basically, however, the units consist of an elastic load
sensing element whose resulting deformation under load is sensed by electrical resistance strain
gages. The resistance change is detected by simple electronic circuitry and amplified to either
operate a large dial indicator or a digital display. The signal is also available for recording or
processing by a computer. These load cells are generally bi-directional so they are ideally suited
to electrohydraulic test machines or other machines with a single test space.

Test Machine Characteristics

Hydraulic and mechanical test machines have other distinguishing characteristics that may be
more or less desirable for specific applications. Some of these have already been noted. Two,
however, deserve special treatment.

Prescribed Variable:
Hydraulic and mechanical test machines differ in their loading characteristics in one major
way. Mechanical designs utilize a prescribed displacement of a crosshead to produce a load on a
confined specimen. In this case the displacement is the independent variable and the resulting
force depends upon the stiffness of the specimen:
F = ksx
where, ks is the specimen stiffness and x is the crosshead displacement. The forces are higher for
stiffer specimens and vice versa. Hydraulic designs on the other hand utilize hydraulic pressure
to produce a prescribed force on the confined specimen. In this case, force is the independent
variable and the resulting deflection depends on the specimen stiffness.
x = F/ks
The deflections are thus lower for stiffer specimens and vice versa.
Electrohydraulic test machines are somewhat unique is this respect because their
characteristics can be changed depending on the choice of control variable. If the load as
measured by a load cell is used as the feedback signal, then load is the independent variable and
the machine behaves like a conventional hydraulic machine. On the other hand, if actuator
stroke is used for the feedback signal, then displacement is the independent variable and the
machine behaves like a screw type design.
These characteristics are particularly important depending on the test requirements.
Hydraulic machines are preferable in applications where a relatively constant load must be
maintained even though small incremental or continuing deformation occurs. This is desirable,
for example, in certain creep tests where it is necessary to measure the specimen elongation with
time while under a constant load at an elevated temperature. Mechanical screw type machines
are preferable in applications where load instabilities exits, that is, in situations where an
increment in load produces deformations that can only be maintained if the load is reduced. This
occurs in column buckling tests as well as in tension tests of material samples to determine their
mechanical properties. The latter case is illustrated in Fig. 5 for a typical ductile material. If the
tensile load is prescribed (hydraulic machine) then the vertical axis marks the independent

AE3145 Test Machines Page 7

variable. The situation is stable until the loading reaches A. Beyond this point, the specimen can
support no further increase in load so that what happens is that path A-B is never recorded.
Instead, the specimen rapidly elongates and failure occurs almost instantly. Another way of
visualizing this is to note that for loads above Po, there are two equilibrium deflections for a
given load. It is thus possible at any time for the specimen to jump over to a larger displacement
state provided the incremental force is momentarily available (due for instance to a perturbation
from outside forces). The key here is the multivalued nature of the behavior for prescribed
loading. If a mechanical test machine is used, the displacement marked on the horizontal axis is
the independent variable. In this case there is no multivaluedness and instability is not a
problem. Instead, as the crosshead is advanced, the load on the specimen rises to A and then
falls off to the failure point at B.

Figure 5. Typical Load-Deflection Curve for a Ductile Material

Machine Stiffness:
A more subtle factor in test machine operation is the mechanical stiffness of the machine
through the load path. Two problems can arise if the machine is not very stiff compared to the
specimen. First, for a mechanical screw type machine, this finite stiffness acts like a spring in
series with the specimen whose stiffness can also be represented by another spring as shown in
Fig. 6. If the crosshead motion is x and the specimen load is F, then it follows that
F = k s es = k m em = kx
And the relative deflections of the specimen and machine are thus given by:

Specimen: es = F / k s

Machine: em = F / k m
It also follows that:
x = F / k = F / [k s k m / (k s + k m )]
Thus the specimen deflection (shortening in this case) is given as:

AE3145 Test Machines Page 8

 es = F / k s = xk / k s = xk m / (k s + k m )
or by taking time derivatives, the deflection rate is:
es = xk m / (k s + k m )
This result show that for a finite machine stiffness, the actual specimen deflection will be less
that the crosshead motion which is prescribed. For example, it the specimen and machine are of
equal stiffness, the actual deflection rate will be one half the set crosshead rate. Generally, it is
desirable to use a machine whose stiffness is at least an order of magnitude or better than the
specimen stiffness.

Figure 6. Stiffness Analysis of a Typical Test Machine

The second factor concerns the buildup of internal or strain energy within the frame of the
test machine. With reference to the above results, the strain energy in the test machine is the
strain energy in the equivalent spring:
U = 1 k m (em ) = 1 F 2 / k m
2
2 2
The stiffer the machine, the less strain energy is stored for a given load. The presence and
availability of this energy can be a cause for concern in testing where instabilities are present.
Given the right trigger, such as an outside impulse or shock, this energy is available to quickly
upset the equilibrium of the specimen and possible result in its sudden failure.
Machines that are very stiff are called “hard” test machines, while those with a low stiffness
are referred to as “soft”. With a very hard screw type machine it is often possible to watch the
ultimate tensile failure of a ductile specimen as a gradual tearing apart of the material. The same
process observed in a soft machine is usually a rapid and sometimes explosive fracture that is
impossible to see with the unaided eye.

AE3145 Test Machines Page 9