The existence of complex microscopic features in the bolted structure leads to inevitable uncertainty in the numerical model.Aiming at the quantification of parameter uncertainty in bolt design,this paper proposes a stochastic model updating method for assembled structures with bolted joints based on a high-precision modal synthesis method.First,considering the influence of higher-order residual modes,a high-precision modal synthesis method suitable for the elastic interface of bolted joints is derived.Then,under the framework of Bayesian inference,using the natural frequency and mode shape probability distribution of the assembled structure,the posterior probability density function of the bolt parameters is established,and the parameter uncertainty is quantified with delayed rejection adaptive metropolis(DRAM)sampling method.The numerical examples indicate that for structures with large differences in natural frequencies among the substructures,compared with the general modal synthesis methods,the method proposed in this paper can ensure good model updating performance in the case of using a small number of substructure modes.
关键词
组合结构/随机模型更新/贝叶斯方法/模态综合法
Key words
assembled structure/stochastic model updating/Bayesian inference/component mode synthesis
维普
万方数据