Abstract
It has been noted in Part I of this series (referred to hereafter as I), that if a nicked specimen of a natural rubber vulcanizate is slowly stretched, tearing occurs at the tip for quite small applied forces. In the initial stages, this tearing continues only as long as the deformation of the specimen is being increased, and virtually ceases if the deformation is held constant. This tearing is essentially time independent, and is termed “static” cut growth. If, however, the deformation is continued until the cut has grown by a few hundredths of a millimeter the growth becomes time dependent and catastrophic tearing takes place, the cut suddenly increasing in length by perhaps a millimeter or so. If a nicked specimen is alternately stretched and relaxed to the unstrained state, the cut gradually grows even though the applied force is less than that required to produce catastrophic tearing. This phenomenon is termed “dynamic” cut growth. This behavior can be compared to that of gum GR-S vulcanizates described in Part III, where static cut growth of the above type does not occur, a dead load on a test piece producing a more or less steady rate of cut growth. In the present paper, measurements on natural rubber gum vulcanizates only are described, and the numerical results expressed in terms of the theory developed in previous papers (Parts I, II and III). It has been shown in I and II that the tear behavior of differently shaped test pieces cut from thin sheets of thickness t may be correlated by means of the concept of the energy for tearing. This is defined as the value of T[=(1/t)(∂W/∂c)l] at the instant of tear, and is denoted by Tc. In the definition of T, is the total elastic energy stored in the test piece, c the length of the cut, and the subscript l indicates that the differentiation is to be carried out at constant displacement of those parts of the boundary that are not force-free. It was also shown that a convenient and direct method of obtaining Tc is by the use of the “simple extension” tear test piece described in I and shown in Figure 1, and this has been used for most of the experiments. Under most conditions, T for this test piece is nearly independent of the cut length, width of the test piece, and modulus of the rubber; T is very nearly equal to 2F/twhere F is the force applied to the arms. In the cases where the use of the above approximate relation between T and F introduces an appreciable error, the exact theory given in I was used.
A Study of Contact Interface for an Anti-Rotation Rivet Using Ultrasonic Detection
