To obtain engineering properties of materials, designers usually refer more to information obtained from tensile testing than to that obtained from practically all other types of testing combined. The more important properties are tensile ultimate and yield stresses, modulus of elasticity, the corresponding compressive and shearing properties and Poisson's ration.
Fatigue Strength of Engineering Materials.
Machine parts frequently are loaded in a dynamic fashion in which the stresses vary with time, usually periodically. These stresses may vary periodically from a given value of tensile stresses to the same magnitude of compressive stress, or from a particular value of shear in one direction to the same value in the opposite direction. Either of these two cases would be examples of fully reversed stress variation. This periodic variation may be harmonic or non-harmonic. Even though a stress variation may be periodic, it may not be fully reversed. Stress variation from one magnitude of tension to a different magnitude of compression are examples of non-fully reversed stress variations. Each of these possible variations in stresses produces what is known as fatigue loading on the machine member.
Experimental data indicate that repeated loading and unloading of a material will cause fatigue failure at a stress much lower than the ultimate strength of the same material as determined by static testing. The number of stress repetitions necessary to produce failure depends upon both the maximum stress imposed and the nature of the stress variation. Experimental evidence indicates that there is a limiting stress below which failure essentially will not occur even though the stress may be repeated an "infinite" number of times. This limiting value of stress often is referred to as endurance limit of the material.
Endurance-limit data are very difficult to determine by experimental or other means, particularly for an infinite number of fatigue cycle. For this reason, experimentally determined endurance limits generally are expressed by an average amplitude of fully reversed harmonically varying stress as determined from several test specimens in which failure occurs in a specified number of cycles. If the method of loading is nor specified, fully reversed harmonically varying flexure in fatigue is assumed. The usually employed endurance-limit cycle is 5 million for wrought ferrous materials, 10 million for cast ferrous materials, 1 million or less for magnesium alloys, up to 500 million for aluminum alloys and 2 million for wood.