Numerical simulations provide crucial theoretical support for optimizing design and operational management. However, their practical application is limited by infeasibility and high computational costs. Machine learning (ML) methods, with their notable efficiency and accuracy, have emerged as powerful tools to address these challenges. A novel framework is proposed to predict the impact responses of foam-filled lattice composite cylinders (FLCCs), integrating precise numerical simulation analyses, metamodels, fast Fourier transform (FFT), and inverse fast Fourier transform (IFFT) methods. Initially, robust numerical models were developed to evaluate the crashworthiness of three different types of FLCCs subjected to impact loading, accompanied by energy transformation analysis. Subsequently, in combination with FFT and IFFT techniques, various metamodels were employed to predict the force–time and displacement–time histories of the FLCCs. Each FLCC type included more than 1000 frequency points, and all constructed metamodels achieved R-square (R2) values greater than 0.95. These results indicate that the proposed framework can effectively predict impact duration and response characteristics in the frequency domain. Furthermore, sensitivity analysis revealed that higher peak impact forces (PIFs) are associated with greater resistance to impact deformation. An increase in glass fiber reinforced polymer (GFRP) thickness led to a marked enhancement in the resistance to impact deformation.
扫码关注我们
求助内容:
应助结果提醒方式:
