Physics-informed process optimization of compression-molded titanium/carbon fiber metal laminates for enhanced flexural and interlaminar performance
Abstract
The mechanical performance of titanium fiber–metal laminates is strongly influenced by consolidation parameters that govern resin infiltration, void formation, and interfacial bonding. This study presents a physics-informed framework for the process optimization of compression-molded Ti6Al4V/carbon fiber metal laminates with a 3/2 symmetric architecture. A D-optimal experimental design was employed to investigate the effects of molding temperature, pressure, dwell time, and thermal ramp rates on effective flexural rigidity, interlaminar shear strength, and void fraction. Laminate bending mechanics incorporating porosity-dependent stiffness degradation were integrated within a residual learning scheme to ensure physically consistent predictions. The developed PIML surrogate showed good cross-validated predictive performance, with R 2 values of 0.960, 0.959, and 0.887 for effective flexural rigidity, interlaminar shear strength, and void fraction, respectively. Multi-objective optimization identified a processing window that simultaneously maintained high flexural rigidity and interlaminar shear strength while minimizing porosity, with experimental deviations below 3% for the mechanical responses and below 4.2% for the void fraction. Fractographic analysis revealed matrix shear yielding, interfacial decohesion, and fiber bridging, confirming progressive damage mechanisms. The proposed framework establishes a robust processing–structure–performance linkage for titanium-based hybrid laminates and supports manufacturing optimization of reinforced composite systems.