3. Principal Physical Test Methods
3.1 Introduction
Rubber has a wide range of unique properties and it is necessary to utilize them in a vari- ety of products and applications, like tires, hoses, conveyor belts, bridge bearings, shoe soles, cables, mounts, gaskets, seals, rocket insulation, etc.
Rubber differs considerably from other engineering materials—for example, it is a highly deformable material, exhibiting virtually complete recovery, and it is nearly incompress- ible with a bulk modulus some thousand times greater than its shear modulus.
Why is it necessary to carry out many complicated tests on rubber-like materials? The simplest answer is that rubber cannot be described by the mathematical laws applicable to ideal materials such as Hookean solids and Newtonian fluids, and is a material that pos- sesses complex properties that are not related to each other.
Rubber differs from metals in that it does not obey Hooke’s law except at very small extensions (up to 50%), and there is no elastic limit that can be specified for “safe loads.”
The stress/strain curve for vulcanized rubber is “S” shaped, and this shape can be modi- fied by the temperature of the test, speed of testing, dimensions of sample, and its previous history. There is a large hysteresis effect when rubber is deformed. This is largely due to the Gough-Joule effect that is evidenced by the fact that rubber heats upon stretching and cools on retraction. The evolution of heat utilizes some of the energy input, and the retrac- tion stress/strain curve does not follow the same course as the extension curve.
One of the chief differences between rubber and other engineering materials is in the large elastic strains to which the former can be subjected without rupture. The engineer wishing to make use of this capacity of strain encounters two difficulties:
1. The normal methods of engineering calculations fail.
2. The material properties, on which normal engineering design is founded, cannot be measured easily or directly under conditions of considerable strain.
The reasons for failure of the rules of normal engineering design are that most of these rules assume that change of shape under load is small and that deformation at the loading or gripping points is also small. Neither assumption is true for rubber. The result is that the average engineer cannot forecast the mechanical behavior of a sample of rubber, even if he has been supplied with the best available data regarding its fundamental physical properties of stiffness and viscosity.
The basis for collaboration between a conventional engineer and a rubber technologist lies in the understanding of each other’s problems. For instance, the rubber technologist expects that the engineer should be familiar to some degree with the nature of rubber-like
materials, the general effects of compounding ingredients, and the difficulties of fabrica- tion and mold design. The engineer expects the rubber technologist to have some knowl- edge of basic applied mechanics in order for the engineering factors of the problem to be understood. The rubber technologist has often been found to prefer to experiment with rubbers of different compositions, when it becomes clear that the original mechanical design requires modification.
The general engineer’s first difficulty arises because with metals he has become used to a small strain proportional to stress. Young’s modulus or the modulus of rigidity (stress/
strain) is readily determined and available for use in design calculations, within the safe working range of metal. For rubber in either extension or compression, the stress/strain relationship is variable. By modulus, the rubber technologist means the stress at a given elongation (e.g., 100%, 300%). Flexure to an engineer means bending, but to some rubber technologists it means any form of straining. Aging to an engineer means stress relieving before final machining; to the rubber technologist aging means deterioration with age.
Resilience to an engineer means the energy stored per unit volume; in rubber it means the rebound property or coefficient of restitution. It is important that these and other similar differences are appreciated.
With such terms as hardness, fatigue, and creep, we are at least on fairly common ground. Hardness with a metal is assessed by measurement of the permanent indentation produced by a hardened ball or diamond at a given load. Rubber hardness is resistance to elastic indentation. Creep, which is change of strain at constant stress, normally occurs with rubber but is not noticeable with metals except at elevated temperatures.
Fatigue is probably the most important factor in the dynamic application of rubber or a metal. With a metal, stress reversals do no serious harm; in rubber design they should be avoided by pre-loading. Poisson’s ratio applies to both metals and rubber. It is important that Poisson’s ratio of 0.5 makes rubber virtually incompressible.
The conventional engineer should know that rubber has the following advantages:
strength; good energy absorption; the ability to undergo large deformations and recover (160,000 times more elastic than steel in shear); good electrical resistance properties; resis- tance to fatigue, abrasion, and corrosion; and moldability.
The disadvantages are that it can be attacked by oils and greases; depending on the structure, it is susceptible to aging which is accelerated by exposure to heat or light; it can be attacked by the ozone; and it is a poor conductor of heat. The first three of the above can be more or less avoided by using a suitable type of synthetic rubber.
From the above considerations it is clear that the elastic behavior of rubbers differs funda- mentally from that of metals. Deformation of a metal involves changes in the inter-atomic distances, and very large forces are needed to change these distances, hence very high elas- tic modulus characteristics of a metal. The forces involved are so great that before the defor- mation reaches a few percent, other actions come into place involving slippage between adjacent crystals. In other words, a metal shows a yield point above which the deformation increases much more rapidly than the stress, so that the stress/strain curve turns away from the stress axis (Figure 3.1). Moreover, the deformation above this point is irreversible.
With rubber, on the other hand, the stress/strain curve (Figure 3.2) bends the other way, as explained above, and there is no yield point. The rubber recovers to almost its original form from any point on the stress/strain curve. Moreover, the deformation of rubber does not involve any straining of the inter-atomic bonds, and hence the force required is lower than that of a metal.
It is essential that selected tests in some way relate to the functions a rubber has to perform. In other words, we should be able to assume that the results are related to
performance at least to the extent that if the properties tested remain constant (or above predetermined values), its performance will be satisfactory. This is the crux of the whole question of laboratory tests, because the relation between basic properties and perfor- mance is only imperfectly understood, so that the selection of the most suitable tests calls for careful judgement.
Laboratory tests can be divided into three main groups:
1. Simple destruction tests and those involving non-recurrent cycles of loading (e.g., tensile strength, compression and shear modulus, bond strength, tear resistance, plasticity, hardness, resilience, permanent set and creep, low-temperature flexibil- ity, and swelling)
2. Service tests (e.g., fatigue, heat buildup, abrasion, flexing, and aging)
3. Development tests (e.g., dynamic stiffness and dynamic resilience, serviceability tests)
Strain Run Run
Rise
Yield strength
Ultimate strength
Fracture Necking
Strain hardening
Stress
Young’s Modulus = Rise = Slope
FIGURE 3.1
Typical stress/strain curve of metal.
Extension Unloading Loading
Force
FIGURE 3.2
Typical stress/strain curve of rubber.