Keywords: track surveying, track monitoring, axlebox acceleration, lateral dynamics, wheel-rail interaction.

This study focuses on the detection of short wavelength lateral track irregularities. For this, lateral axlebox accelerations from trains during commercial service have been gathered. This data has been further characterised in the time-frequency domain by means of STFT. This yields some useful information in terms of excited frequencies and their respective reached amplitudes. Specific locations along a railway track with noticeable loss of horizontal alignment are compared with some track stretches with absence of defects. Lateral axlebox accelerations are firstly analysed in the time domain in order to identify the location of the main peaks. Peaks are classified upon their absolute magnitude in three levels. i.e. Alert Limit (AL), Intervention Limit (IL) and Immediate Action Limit (IAL). In terms of amplitudes, it is found that regular track stretches (i.e. with no defects) show amplitudes below -20 dB for all the analysed frequency range. In opposition, spots with track defects show amplitudes higher than -10 dB for some specific frequencies. These frequencies mainly are featured by a main isolated peak at 25 Hz, a secondary peak at 150 Hz and a set of peaks in the band of 200-350 Hz. Minor differences among the analysed stretches have been found for frequencies above 400 Hz. This empirical featuring of lateral vehicle-track dynamics may be used for the classification an identification of a large variety of short length horizontal track defects, such as misaligned fishplate joints, bended or worn rails. This can also be used for validating a smart detection system based on image processing techniques.

download the full-text of this paper (PDF, 763 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description