Fatigue and Corrosion Fatigue with Special Reference to Service Breakages

Abstract

There is need for closer co-operation between those who encounter service breakages and those engaged in research on the strength of materials, because various causes are conspiring to enhance liability to fatigue failure in modern prime movers and machines. In the author's investigations, special attention has been paid to the crack spreading process with the twofold object (1) of interpreting the features found on fatigue fractures, and (2) as a method of attacking the problem of crack formation from a new angle. The fatigue testing machines and material used are briefly described. Crack spread in plain specimens has been found to follow definite laws of a very simple character. There is a fixed relationship between the markings on a fracture produced in rotary bending and the direction of rotation, which the author calls “crack slip.” Matters become more complex when “stress raisers” are present, as also when cracks converge from several separate origins. The stresses in revolving beams of square section, and of round section with equal parallel flats, are discussed at some length, and theory and experiment are shown to be in complete agreement. Reference is made to the fatigue fracture of the main haulage shaft at a colliery and to a danger which must be reckoned with in the blading of steam turbines. Attention is drawn to the weakening effect of the local concentrations of stress produced by holes, shoulders, keyways, etc., and two examples are included of an attempt to locate and assess the diminution of fatigue resistance at the weakest spot of a projected design by a single fatigue test of a suitable model. The history of corrosion fatigue is briefly reviewed, with comments as to why its discovery is of such recent origin. An extensive list of practical instances where corrosion fatigue is the known or suspected cause of failure is given. Examples of the pitting and cracking processes peculiar to corrosion fatigue are illustrated and discussed, the corrosive agent being a small steam jet in air, and attention is drawn to the differences which are found according as the limits of cyclic stress are near, or much below, the fatigue limit in air.