Types of rail roughness and the selection of vibration isolation measures

Abstract

A procedure is outlined for quantifying the significance of various types of rail roughness for the purposes of predicting and controlling low-frequency ground vibration near railway lines. The effectiveness of resilient elements inserted beneath the rail is investigated. It is found that roughness mechanisms can be classified according to the effectiveness of added resilience. Vibration from a track with "Class A" roughness will be well controlled by resilient rail support; a track with "Class B" roughness cannot be controlled; and vibration from a track with "Class C" roughness will be increased. Examples of "Class A" roughness are those due to variations in trackbed roughness and in rail-support stiffness and these can be controlled through the insertion of additional under-rail resilience. The reduction in roughness level can easily result in a 10dB reduction in perceived roughness down to 5Hz for an urban metro system. This reduction is additional to any reduction due to the mass-on-spring effect of the unsprung mass on its resilient track. This not only explains the good insertion performance often measured of under-rail countermeasures, but also why there is often far less amplification at the track-system resonance than might otherwise be expected. In some cases the effect of adding resilience has less effect than in others. For example there is no beneficial effect found for the case of roughness due to a rail that is naturally bent (a condition built-in after leaving the rail-straightener in the rolling mill). © 2008 Springer-Verlag Berlin Heidelberg.