Influence of crack tip and crack offset distance on crack interaction and growth direction in multiple cracks

Purpose – This paper aims to deal with the study of interaction between multiple cracks in an aluminum alloy under static loading. Self-similar as well as non-self-similar crack growth has been observed which depends on the relative crack positions defined by crack offset distance and crack tip distance. On the basis of experimental observations, the conditions for crack coalescence, crack shielding, crack interaction, crack initiation, etc. are discussed with respect to crack position parameters. Considering crack tip distance, crack offset distance, crack size and crack inclination with loading axis as input parameter and crack initiation direction as output parameter, an artificial neural network (ANN) model is developed. The model results were then compared with the experimental results. It was observed that the model predicts the crack initiation direction under monotonic loading within a scatter band of ±0.5°. Design/methodology/approach – The study is based on the experimental observations. Growth studies are made from the growth initiation from two cracks in a rectangular aluminium plate under static loading. The present study is focused on the influence of crack position defined by crack offset distance and crack tip distance on growth direction. In addition to this, ANN has been used to predict crack growth direction in multiple crack geometry under static loading. The predicted results have been compared with the experimental data. Findings – The influence of the interaction between multiple cracks on crack extension angle greatly depends on the relative position of cracks defined by crack tip distance S, crack offset distance H and crack inclinations with respect to loading direction. The intensity of the crack interaction can be described according to degree of crack extension angle and relative crack position factors. It is also observed that the progress of the outer and inner crack tip direction is different which mainly depends on the relative crack position. Research limitations/implications – It is limited to static loading only. Under fatigue loading findings may differ. Practical implications – It is important to investigate the growth behaviour under multiple cracks and also to know the effect of crack statistics on the growth behaviour to estimate the component life. The study also focused on the development of a high quality predictive method. Originality/value – The results show trends that vary with crack geometry condition and the ANN and empirical solution provides a possible solution to assess crack initiation angle under multiple crack geometry.