Envelope Spectrum Analysis With Algorithm Simulations To Detect Railway Wheel Out-of-roundness Defects

Resumo

The objective of this work is to identify defects in railway vehicle wheels using an envelope spectrum
analysis technique combined with spectral kurtosis analysis. The analysis is performed on data collected by a
simulated wayside monitoring system situated on the track. A dynamic 3D interaction model is implemented using
the in-house software VSI-Vehicle-Structure Interaction Analysis to simulate the dynamic interactions between a
passing train and the track. In this work, an Alfa Pendular passenger vehicle and a section of the Portuguese
Northern Line serve as the basis for the numerical models of the train and track, respectively. In the simulations,
three types of wheel flat profiles and three periodic polygonal wheel profiles are analysed, along with results
obtained from dynamic analyses conducted on wheels that do not exhibit any type of damage. Additionally, the
simulation considers track irregularity profiles generated based on the Federal Railroad Administration (FRA)
standards. The dynamic responses of several strain gauges and accelerometer sensors located on the rail between
sleepers are then evaluated through numerical calculations. To further validate the methodology, the influence of
different train speeds, track unevenness profiles, and wheel fault severity are considered. For validation purposes,
the right wheel of the first wheelset is modelled with a defect, but the detection methodology is effective for
damaged wheels modelled in any position. The application of the methodology demonstrates that envelope
spectrum analysis can successfully differentiate between healthy and defective wheels.