Fractional calculus in dynamic analysis of truck frame: a special focus on finite element modelling

This paper investigates an innovative method to analyse and improve the dynamic performance of truck frames. Firstly, a 3D model of the frame was constructed by computer-aided design (CAD), and then meshing and material property sets were carried out by finite element analysis software to simulate the behaviour of the frame under variable working conditions. Fractional calculus is introduced to capture the nonlinear dynamic characteristics of the frame more accurately, and the Grunwald-Letnikov method is used to solve the relevant equations. Deep learning techniques extract key features from FEA data to identify stress concentration areas. Through dimensional optimisation experiments, this study also demonstrates how to adjust the frame plate thickness using gradient-based optimisation algorithms to achieve lightweight and performance improvement. Notably, the maximum stress value under bending conditions increases by 29.6 MPa, while under torsional conditions, it decreases by 65.7 MPa. This study highlights the synergistic potential of artificial intelligence, fractional calculus, finite element modelling, and deep learning techniques in advancing the dynamic analysis of truck frames, setting the stage for more resilient and efficient truck designs in the future.