To bridge the mechanical performance gap between polyethylene terephthalate (PET) foam cores and balsa wood in wind turbine blades, this study proposes a hierarchical groove-perforation design for structural optimization. A finite element model integrating PET foam and epoxy resin was developed and validated against experimental shear modulus data (α < 0.5%). Machine learning combined with a multi-island genetic algorithm (MIGA) optimized groove parameters (spacing: 7.5-30 mm, width: 0.9-2 mm, depth: 0-23.5 mm, perforation angle: 45-90°) under constant resin infusion. The optimal configuration (width: 1 mm, spacing: 15 mm, angle: 65°) increased the shear modulus by 9.2% (from 125 MPa to 137.1 MPa) and enhanced compressive/tensile modulus by 10.7% compared to conventional designs, without increasing core mass. Stress distribution analysis demonstrated that secondary grooves improved resin infiltration uniformity and interfacial stress transfer, reducing localized strain concentration. Further integration of machine learning with MIGA for parameter optimization enabled the shear modulus to reach 150 MPa while minimizing weight gain, achieving a balance between structural performance and material efficiency. This hierarchical optimization strategy offers a cost-effective and lightweight alternative to balsa, promoting broader application of PET foam cores in wind energy and other high-performance composite structures.
Keywords: genetic algorithm; groove and perforating; optimal design; pet foam core; shear performance.