Simulation and optimization of hot press molding for polypropylene-based continuous carbon fiber broadening fabric prepregs

Abstract

Undoubtedly, continuous carbon fiber-reinforced thermoplastics (CCFRTP), with their excellent specific stiffness, specific strength, recyclability, and high design precision, can effectively meet the automotive industry’s demand for lightweight and energy-saving solutions. For automotive B-pillar reinforcement plates, side-impact energy absorption is critical to passenger safety, and CCFRTP’s superior performance makes it an ideal choice for enhancing structural integrity. However, the hot press molding process for B-pillar manufacturing is relatively complex, involving fiber-matrix interaction, fiber orientation variation, and resin flow. Additionally, unreasonable molding parameters may even cause fiber breakage and tearing. In this study, a self-developed non-orthotropic anisotropic constitutive model for CCFRTP was employed to simulate the complex multi-physics field coupling behavior during hot press molding. Meanwhile, artificial neural networks (ANN) were utilized to optimize key process parameters, aiming to reduce fiber shear angle and stack thickness. This integrated approach effectively addresses the process control challenges arising from such coupled effects in CCFRTP hot press molding. After optimization, the maximum fiber shear angle was reduced by 6.513%, stack thickness deviation by 5.746%, and fiber wrinkling was significantly mitigated. Notably, the wrinkle area of B-pillars fabricated with the optimized parameters was nearly consistent with simulation results, which effectively verifies the optimization method’s effectiveness and reliability. This study demonstrates that the method combining simulation and ANN optimization plays an irreplaceable role in the hot press molding process of CCFRTP.