Efficient Predictive Model for High-Frequency Fatigue Life of High-Speed Railway Fastening Clips Using Particle Swarm Optimization Algorithm

Abstract

Fastening clips, though seemingly minor components, are crucial for maintaining the stability and smoothness of railway tracks. However, clip fatigue fractures have been frequently encountered in high-speed railways due to high-frequency excitations. This study develops an efficient predictive model for clip fatigue life assessment using a particle swarm optimization algorithm (PSOA). Firstly, the central axis of an ω-type fastening clip with spatially variable curvature is mathematically described via ten independent parameters. By employing the modal superposition method (MSM) and PSOA, we have derived vibration equations of a prestressed clip. This approach simplifies the nonlinear contact behavior, reduces degrees of freedom, and achieves rapid numerical convergence compared to traditional 3D finite element (FE) models, enabling efficient prediction of fatigue life due to vehicle-track dynamic interactions. The model's reliability has been validated by existing numerical results and experimental studies. Subsequently, high vibration, high stress, and low fatigue life regions, as well as the specific location most prone to fracture, are identified based on the spatial distribution of clip dynamic responses. The potential causes of stress concentration at the critical location are explored by examining the clip's geometric characteristics, including curvature, torsion, and their derivatives. Finally, control thresholds for the amplitude of short-wavelength irregularities and vertical vibration displacement of clips are proposed to meet the fatigue life requirement. The current work provides guidance for the maintenance and management of high-speed railway fastening clips and inspires the optimization of the clip configuration.

Keywords: High-speed railway fastening clips, Particle swarm optimization algorithm, Efficient predictive model, High-frequency fatigue life, Vibration control thresholds