Computational hybrid framework for wear of polyurethane coated GFRP composites in offshore and onshore environments

Purpose This study aims to examine the wear performance of polyurethane (PU)-coated GFRP composites under onshore and offshore conditions. With the world’s growing reliance on wind energy, leading-edge erosion of turbine blades particularly offshore has become a pressing challenge, directly impacting durability and efficiency. This work evaluates PU coatings protective effectiveness and their potential to extend turbine blade service life. Design/methodology/approach A hybrid methodology was used to evaluate the erosion wear by using design of experiments (Taguchi L25) and artificial neural network (ANN) modelling. To identify the wear mechanisms, SEM microstructural methods were used and the influence of three important parameters including impact velocity, impact angle and the duration of exposure were also investigated systematically. Findings PU coatings provided good protection in both environments as the average S/N ratios were 43.9 and 40.1 under onshore and offshore, respectively. DOE-ANN had a ranking of parameter influence in Impact Velocity > Impact Angle > Exposure Duration and offshore conditions resulted in erosion which was more severe in salty conditions. ANN prediction accuracy was high, achieving R² values of 0.9931 (onshore) and 0.9851 (offshore). Hybrid crow search algorithm and grey wolf optimizer identified optimal parameters that reduced mass loss to 0.00145 g (onshore) and 0.0044 g (offshore). SEM observations revealed plowing, matrix removal, fiber exposure, and deeper grooves. Originality/value The paper integrates experimental design, ANN-based prediction, parameter validation, and hybrid metaheuristic optimization to address wear behavior. This comprehensive approach provides practical insights for reducing wear and enhancing the service life of wind turbine blades, particularly under offshore operating conditions. Peer review The peer review history for this article is available at: Link to the website of publons