Probabilistic design of mechanical components

During the design stage, realistic life time predictions for components help to reduce test cycles and allow for an optimization with respect to both production and repair costs. However, frequently used calculation procedures rely on the introduction of empirically or semi-empirically derived safety factors and often lead to a very conservative design. In this paper, a probabilistic fracture mechanics approach is proposed, that allows for accurate predictions of the life time distribution. In order to reduce the computational effort, the computational procedure is divided into a deterministic and a stochastic part. While for the deterministic part, the complexity of the model can be rather high, the stochastic part requires simplifying assumptions and efficient simulation schemes. These assumptions and their consequences are in the focus of this paper. The quality of the proposed approach is assessed by comparison of results obtained for industrial applications with experiments.