Data-driven analysis and prediction of railway curve squeal

Investigating the factors influencing railway curve squeal, developing predictive models, and optimising effective mitigation strategies to reduce noise pollution and improve railway system performance.

Goal

The research objective is to investigate the factors influencing railway curve squeal, develop predictive models, and optimize effective mitigation strategies to reduce noise pollution and improve railway system performance.

Project status and results

Railway curve squeal is a prominent tonal noise resulting from instabilities in the wheel-rail interface as railway vehicles negotiate curved track sections. This squealing produces high noise levels, with dominant tonal components in the frequency range where human hearing is most sensitive, leading to significant noise pollution in areas near curved railway tracks.
Currently, the most effective mitigation strategy involves applying friction modifiers on the rail head to optimize friction conditions, reducing wheel-rail contact instabilities while maintaining sufficient traction and braking performance. However, automatic rail lubrication systems are prone to defects, require frequent maintenance, and the friction modifiers are generally not environmentally friendly. Consequently, there is a critical need to explore alternative approaches to mitigate curve squeal and improve existing friction modification techniques.

A deeper understanding of the mechanisms driving curve squeal is critical for developing improved solutions. This PhD project will focus on identifying the root causes of curve squeal, investigating the impact of operational, examine the influence of operational, environmental, and infrastructural factors on its occurrence, and evaluating potential mitigation strategies. The study will utilize dynamic simulations, field measurements, experimental tests, and statistical analyses. Moreover, the research involves developing a practical track side monitoring system to assess the statistical tendencies of curve squeal.

In 2024, a wayside squeal measurement system was deployed on a curved track segment with a known history of squeal noise. Algorithms squeal noise detection and tonal analysis from rail vibration and acoustic signals were developed and presented at the Inter-Noise 2024 conference [1] and at the Vibration Day 2024 organised by the Scandinavian Vibration Society. Current efforts are focused on data acquisition and statistical analysis of squeal propensity, with an emphasis on environmental and operational parameters. Furthermore, field measurements are being conducted to assess the efficacy of various friction modification strategies in mitigating curve squeal.