Failure of metal structures is caused by cracks appearing in the material and growing until the strength of the structure is compromised. Such cracks are formed and grow under stresses far lower than the stress at which unstable fracture occurs. It is useful for designers and engineers to be able to predict how long a material will last under certain operating conditions, so that safe structures may be manufactured. Such structures include suspension bridges, ships, bicycles and aircraft.
The way in which cracks grow in metal has been researched extensively, for different types of metal, and for different operating conditions. The great majority of this work has concentrated on long cracks, that is cracks of the order of 1mm or longer. Short cracks do not adhere well to the models which work for the long cracks. In addition, there tend to be many short cracks within a specimen which grow into one another. This interaction between short cracks is termed coalescence, and results in the formation of a dominant long crack.
Reliability modelling is concerned with making probability statements about how long an object will survive. This dissertation is concerned with modelling reliability by developing a model for short crack propagation and coalescence in steel.
The models have been developed in collaboration with, and the data analysed comes from material experiments carried out at the Department of Mechanical and Manufacturing Engineering, Trinity College Dublin.