Material defects as the basis of fatigue design

Abstract Microstructural inhomogeneities can exist in metals, such as precipitates and inclusions. These can be introduced purposely to strengthen the metal matrix (e.g.: CuAl2 in α-phase aluminium) or by faulty manufacture (e.g.: large inclusions). A major part of the history of metal fatigue has been to eliminate, or at least reduce, detrimental precipitates, inclusions and manufacture defects such as scratches, surface roughness and shrinkages in cast irons since these can initiate fatigue cracks. The deleterious nature of foreign bodies or other forms of inhomogeneities, e.g.: pores or de-bonded zones within a microstructure are related to their size, position, shape, orientation and physical properties. Small sizes of precipitates and inclusions are to be preferred to large ones; such defects being less detrimental inside a grain rather than at grain boundaries where they can simultaneously affect two or three near-neighbour grains. The orientation of a defect is dangerous should it be inclined to the future direction of Stage I and/or Stage II crack growth planes. An important duty of a metallurgist is to reduce the effectiveness of these different forms of micro-defects produced during manufacture whilst a mechanical engineer is required to derive a suitable form of fracture mechanics in order to account for their behaviour in a quantitative fatigue fracture analysis. In this paper, some important types of microstructural defects will be illustrated and discussed in terms of their size, position, shape, orientation and properties, together with their effect on the fatigue resistance of a material. This will be coupled with a fracture mechanics (FMs) approach that quantifies their behaviour in terms of their relationship to fatigue crack propagation. In this approach, the boundary condition between failure and non-failure is explored using FM as a threshold condition for a small crack coupled with the Vickers hardness HV of the material that represents the condition (ΔKth) for the onset of micro-plasticity (yielding) required for the growth of a crack from the defect and the non-propagation behaviour of the crack. Statistical scatters of microstructures, defects and inclusions are the major factors of statistical scatters of fatigue strength and fatigue life. Directions for optimizing microstructure to improve fatigue strength are explained from the viewpoint of equality control based on the statistics of extremes of defects and inclusions. A new efficient and reliable inclusion rating method for high strength steels based on the statistics of extremes using the phenomenon of hydrogen embrittlement is proposed.