Use of Power Spectral Density for Effective Turnout Geometry Maintenance

Abstract

The purpose of this study was to assess the geometrical quality of railway turnouts through application of power spectral density (PSD). PSD in the form of a continuous curve can show turnout irregularities by means of wavelength and amplitude. As a case study, the data related to longitudinal level parameter measurements for four turnouts in the Swedish Railways from 2007 to 2011 were acquired. The PSD curves for these measurements were calculated by means of ProVal software. The PSD curves were compared by PSD standard in a program written in MATLAB and turnout irregularities were determined. The results show that PSD is able to demonstrate the condition of turnouts individually and can also identify the wavelength faults present in different sections of a turnout. More important, PSD can be applied to prioritize maintenance needs of turnouts on the basis of their criticality. Use of PSD can help identify tamping needs more precisely. As a result, the geometrical quality can be restored before the defined maintenance thresholds are exceeded. This maintenance action also would reduce the total downtime, obligatory speed reductions, and maintenance costs.