A dislocation barrier model for fatigue limit - as determined by crack non-initiation and crack non-propagation

摘要

Fatigue damage is caused by cyclic slip, and cyclic slip isdriven by dislocation glide force. In order to cause fatigue damage, thecyclic glide force has to overcome the resistance of the primary and secondarydislocation barriers. Based on this cyclic damage process, the followingdiverse fatigue phenomena are synthesized into an integral and self-consistent analysis: Fatigue damage occurs in persistent slip bands (Hempel,1956; Smith, 1957; Forsyth, 1957, 1961, 1963), and a nucleated fatiguemicro crack is a shear crack (Forsyth, 1961). Pre-cracking fatigue damage isconfined to the surface layer of a stressed metal (Wood et al., 1963).Fatigue limit is inversely related to grain size as shown in brass (Sinclairand Craig, 1952), in mild steels (Klesnil, Rolzmann, Luka's and P. Rys 1965;Yoshikawa and Sugeno, 1965; and Taira, Tanaka, and Roshina, 1979), andin ferritic-martensitic steel, (Kunio, Shimizu, and Yamada, 1969). Forrestand Tate (1964) found fatigue cracks infinegrained brass at an alternating stresseven below the fatigue limit. The cracks were within the boundaries of singlegrains. But they found no cracks in coarse-grained brass below the fatiguelimit. The analysis synthesizes all of these experimental observations.The analysis is based on a realistic physical model. With a betterunderstanding of the model and an improved calculation of the glide force,quantitative evaluations of the resistance of the dislocation barrierswould eventually be possible. The needsfor additional research are pointed out. A number of means of improvingfatigue strength, based on the analysis, are suggested or explained.