Reproducibility of railway noise measurements—influence of weather and test site conditions

Requirements on monitoring stations to guarantee good reproducibility of measured railway noise data from one test site to another are investigated by field measurements and numerical simulations. Based on data from a monitoring system installed at Deutsch-Wagram in Austria, the influence of weather conditions (temperature, rain and snow) on pass-by noise levels and rail vibration is assessed. The influence of temperature on track decay rates (TDR) has been determined from pass-by analysis (PBA). It is concluded that the track superstructure design used at Deutsch-Wagram (60 kg/m rails, rail pads with dynamic stiffness in the order of 700–1000 kN/mm, and monobloc concrete sleepers on ballast) is appropriate for noise monitoring since the variation of measured noise levels for the studied range of weather conditions has a standard deviation of less than 2 dB. For such a track design, it is not essential to apply a correction procedure to account for seasonal variations (except in the case of snow). Further, a numerical parameter study of the influence of TDR and ground conditions (horizontal ground surface level relative to top of rail, and ground impedance based on a single value effective flow resistivity parameter) on predicted noise level is performed using TWINS. It is found that for ground surfaces with low sound reflection, the influence of ground surface level (when varying from 0.2 to 2 m) on predicted pass-by noise level is in the order of 1 dB