Prediction of contact stress in bolted joints using the Polynomial Chaos-Kriging model

•PCK surrogate model exhibited superior performance in contact stress prediction.•It can be used as an alternative for FEA to reduce the computational cost.•The powerful solution effectively combines the design and using of structures. Bolted joints are one of the most widely used forms of connection. As the weak link, its contact stress directly influences the mechanical properties of a structure. In this work, Polynomial Chaos-Kriging (PCK) surrogate model is adopted to predict the contact stress in bolted joints. Firstly, a Finite Element (FE) model of the bolted joint is established using a parametric modeling method. On this basis, the stress datasets for the contact area of the plate are obtained. Subsequently, the generated stress datasets are imported into the PCK surrogate model for training and validation. Maximum axial stress and influence coefficient are used to conduct the characterization analysis of the contact stress. Results show that the PCK surrogate model is suitable to predict these two parameters. In the case that compressive stress does not exist in all contact areas, quantitative analysis of the contact stress can be further carried out by using this model. The main influencing parameters of contact stress are extracted according to the correlation analysis results. The trained PCK surrogate model can be used as an alternative to finite element analysis for the prediction of contact stress. Design of bolted joints can also be achieved by the parametric control using this model.