Inversion analysis of constitutive relations of blade spar laminates with wrinkle defects

The spar is the primary load-bearing component of wind turbine blades, where laminate wrinkles are common defects. During loading, such defects often serve as initiation points for damage, significantly reducing the spar’s load-carrying capacity. However, accurately testing the mechanical properties of laminated composites is challenging, which adversely affects the reliability of blade safety assessments. To address this, a method combining finite element analysis and experimental simulation is proposed to optimize the inversion analysis of the true stress–strain constitutive relationship for the spar laminate. Four sets of uniaxial tension tests were conducted on specimens with varying defect sizes to obtain their corresponding load-displacement curves. A hybrid optimization method integrating Bayesian regularized backpropagation neural networks and particle swarm optimization was developed to inversely determine the constitutive parameters of composite laminates. The relative errors of the identified parameters were below 5%, validating the accuracy of the inverse model. Finite element simulations based on the inversely obtained parameters demonstrated excellent agreement with experimental load-displacement curves (errors <8%). It was found that the unloading displacement at initial damage increases with the amplitude-to-length ratio, indicating that wrinkle defects exhibit strain-sensitive monitoring nodes in structural health monitoring. The proposed method proves effective for mechanical property characterization of wind turbine spars, enhancing safety assessment accuracy and operational reliability.