Validation of Surrogate Model-Based Life Prediction for a Composite Rotorcraft Hub Component

Abstract

A composite rotorcraft hub component is subjected to fatigue due to thrust, centrifugal and bending loads during flights, leading to delamination between individual plies at the thick-to-taper transition resin pocket. In order to evaluate the delamination fatigue life of the rotorcraft hub, a finite element analysis (FEA) model is constructed for a quarter symmetry of the hub geometry. The virtual crack closure technique (VCCT) is used to assess the strain energy release rate associated with the fatigue delamination of the component under given geometry, material property, and loads. The fatigue life distribution is then obtained by propagating the uncertainty of material and loads through the simulation model. Validation of the life prediction model is required before the model is applied to the rotorcraft hub design. In this paper, a model validation methodology is developed for the fatigue life prediction model of the composite rotorcraft hub component. Since directly using the FEA model for fatigue life prediction is computationally expensive, a Kriging surrogate modelbased method is employed to reduce the computational burden. To validate the surrogate model-based fatigue life prediction model, in this work, Bayesian hypothesis testing is extended to include both statistical uncertainty and surrogate model uncertainty during model validation. The results show that the involvement of statistical uncertainty and surrogate uncertainty in model validation may reduce our confidence in rejecting or accepting the fatigue life prediction model of the composite component.