Abstract
Railway noise from conventional meter-gauge lines in Japan mainly consists of rolling noise and traction-motor fan noise. Rolling noise is generated by vertical vibration of the wheel and rail, which is induced by relative displacement between the two due to the roughness on their surfaces. Through field tests, it was found that the excitation of rolling noise is determined by both the wheel/rail roughness and the vibratory behaviour of rolling stock and tracks. A theoretical model for rolling noise (such as TWINS) was then applied to Japanese railways, with the predictions showing a close correlation to the measured values. In terms of noise spectra, the rail was found to contribute more to rolling noise than the wheel in much of the frequency range. An attempt to estimate the effect of wheel and track parameters on rolling noise was also made using the TWINS model. The stiffness of the rail pad was found to affect the balance between the rail and sleeper components of noise, and additional damping for the rail was deemed effective in reducing the rail component of noise.
