A Novel Optimization Strategy for Curing Thick Composite Laminates Based on Multi‐Physics Coupled Physics‐Informed Neural Operator

ABSTRACT This work developed a novel, nondata‐driven multiobjective optimization strategy for the curing process of thick composite laminates, employing a multiphysics coupled physics‐informed neural operator (PINO) framework to predict temperature, degree of cure (DoC), and resin flow. Three coupled Fourier Neural Operators (FNO), namely temperature FNO ( T ‐FNO), DoC FNO ( α ‐FNO), and resin flow FNO ( P _r ‐FNO), were developed within a function‐to‐function framework to accurately predict the curing process. Subsequently, the Curing Fourier Neural Operator framework ( C ‐FNO) was integrated with the nondominated sorting genetic algorithm‐III (NSGA‐III) to optimize the cure time ( t _cycle ), maximum temperature gradient (Δ T _max ) and DoC gradient (Δ α _max ) simultaneously. The framework was evaluated using several aerospace‐grade prepreg systems, and satisfactory prediction accuracy was obtained under different curing kinetics and material conditions. Experimental studies on thick laminates further verified the predicted temperature evolution and the effectiveness of the optimized curing cycles. In comparison with conventional and data‐driven approaches, the proposed method led to more uniform curing and improved mechanical performance while maintaining stable prediction accuracy. These results indicate that the proposed strategy is applicable to the curing optimization of thick composite structures.