Estimation of loading conditions of failed crane‐hook: an image‐based approach with knowledge and simulation

The main topic of this study is an estimation of loading conditions of crane‐hooks. We examine the relation between the applied load condition and hook's deformation by FEM analysis. The relation is recorded into the load‐deformation database. The feature points are detected on failed crane‐hook image in order to compare the image with the recorded deformation information in the load‐deformation database. The analysis record corresponding to the failed crane‐hook image is then obtained by means of a difference‐minimisation approach. On the basis of the Bayesian theory, we estimate the probability distribution of the load condition, that is, the applied point and the direction of the critical load that causes the failure. The following tendency of the critical load is observed based on the obtained results in this study: the load applied point is not at the most downward position and the load direction is not the normal direction to the tangential line of the hook curve.