Reliability analysis method of the hull structure based on generalized polynomial chaos

The complexity of hull-structure calculation and the diversity of random distributions of uncertainty factors necessitate structural reliability analysis methods that are cost-effective, efficient and adaptable to multiple distributions. This paper proposes a novel reliability analysis method of the hull structure based on Generalized Polynomial Chaos (GPC). GPC surrogate models for performance functions are developed for two typical cases with different distributions within the hull structure. The variations in mean, standard deviation, reliability index and failure probability with respect to GPC expansion order and sample size are analysed and compared with traditional methods, including Monte Carlo (MC) and the First-Order Reliability Method (FORM). The results indicate that the GPC-based method is effective for handling random variables of various distributions in hull structures. Moreover, the proposed method demonstrates superior convergence and accuracy compared to MC and FORM. By circumventing the need for extensive sample calculations for real structural models, the GPC-based method enhances computational efficiency. The feasibility and efficiency of the GPC-based structural reliability analysis method are validated, offering a promising new approach for assessing the reliability of hull structures.