Some recent developments at "GALCIT" concerning a theory of metal fatigue: recents developements a "GALCIT"

Abstract

The proposed Theory of Metal Fatigue is based on the proposition that the fatigue problem is better understood, if attention is focused on the propagation of a "dominant crack" presumed responsible for the ultimate catastrophic failure of the specimens. The length of the crack at the instant of the catastrophic failure is assumed to be governed by the Griffith condition. The advantage of this formulation appears to be that one is in a position to keep track of the fatigue damage in terms of a growing crack. Consequently, by referring the influence of the different variables on the growth of this dominant crack, it has been found possible to treat the various aspects of fatigue in a fairly systematic manner. The theory distinguishes the order of application of load in two step cumulative damage, and in a fairly simple manner treats the problem of program type loading and random loading and intermixing of stress, temperature and frequency. While analytically the computations for completely random loading became too involved, physically the problem can be fairly well understood in terms of the growth of the dominant crack. One of the fundamental assumptions made is that concerning the mean value of an "internal stress" that is presumed to oppose the motion of dislocations. In the present theory, this stress is formally identified with the "endurance limit" in fatigue. By invoking recovery mechanisms to discuss the decrease of this internal stress as a function of temperature and time, it has been found possible to develop a first order theory of elevated temperature fatigue which gives fair agreement with the test data. The inherent statistical nature of the fatigue problem is attributed in the present theory, to the statistical nature of the internal stresses and the resulting statistical nature of the submicroscopic crack initiation ahead of the dominant crack. Finally an attempt is made to evaluate some of the known shortcomings of the proposed theory.