Programmed fatigue of a low carbon steel

Abstract

Fatigue crack propagation has been investigated in single edge notched mild steel specimens under constant and programmed loading conditions. Results have been analysed by the fracture mechanics approach and the Palmgren-Miner summation. Fracture surfaces were examined with a Cambridge Stereoscan scanning electron microscope. In tests where both load rises and load drops were included, the cumulative cycle ratio was always slightly greater than unity, irrespective of block size, load level, rate of load change or sequence. When the programme consisted only of load rises, the cumulative cycle ratio decreased below unity. These results have been explained in terms of residual stresses developed at the crack tip during load changes. The fracture mechanics analysis was found to provide a satisfactory method of data correlation, provided the effects of material and experimental variables were fully appreciated. However, no single value of slope of the log∆K-log da/dN relationship will apply to all materials and conditions. In the absence of residual stresses due to manufacturing processes, computer integration of the constant amplitude fracture mechanics data will provide a satisfactory method of life prediction. Fractographic observations showed that striation spacings were often larger than the macroscopic crack growth rate. This effect was discussed in terms of discontinuous crack front movement. Striation spacing measurements were not therefore useful as a technique of correlating macroscopic crack growth rate with the microscopic observations. However, two distinct modes of fracture were observed to occur. The first (designated stage Ila) was structure sensitive, whilst the second (designated stage IIb) was structure insensitive. The transition from stage Ila to stage llb occurred when the radius of the plastic zone ahead of the crack tip reached a value of about four times the mean grain diameter. Under programmed loading conditions, stage IIa was found to occur after a high to low load change even though the prevailing stress intensity was at the level sufficient to produce stage IIb in a single stress level test. It is suggested that this effect is caused by substantial residual compressive stresses acting in opposition to the applied stress intensity and estimates of the minimal levels of these residual stresses have been made. The mechanisms of residual stress formation and decay have been explained in terms of sizes of plastic zones at the crack tip.