Asymptotic Sampling for reliability analysis of adhesive bonded stepped lap composite joints

Reliability analysis coupled with finite element analysis (FEA) of composite structures is computationally very demanding and requires a large number of simulations to achieve an accurate prediction of the probability of failure with a small standard error. In this paper Asymptotic Sampling, which is a promising and time efficient tool to calculate the probability of failure, is underutilised, and a probabilistic model for the reliability analysis of adhesive bonded stepped lap composite joints, representative for the main laminate in a wind turbine blade subjected to static flapwise bending load, is presented. Three dimensional (3D) FEA is used for the structural analysis together with a design equation that is associated with a deterministic code-based design equation where reliability is secured by partial safety factors. The Tsai–Wu and the maximum principal stress failure criteria are used to predict failure in the composite and adhesive layers, respectively, and the results are compared with the target reliability level implicitly used in the wind turbine standard IEC 61400-1. The accuracy and efficiency of Asymptotic Sampling is investigated by comparing the results with predictions obtained using the Monte Carlo simulation technique. Finally, the partial safety factors are calibrated, and it is shown that the methodology can be further applied to general calibration of partial safety factors to be used in deterministic design.