Railway track degradation: The contribution of a spatially variant support stiffness - Local variation

This study addresses the contribution of spatial variance in the railway track support stiffness to the expected long-term track degradation. Hereto, a novel frequency-domain model is developed with a double periodicity ‘layer’, capable of dealing with both sleeper periodicity and arbitrary non-uniformity in track properties. The model application focuses on a locally reduced support stiffness (hanging sleeper) along the track. The resulting susceptibility to degradation is assessed by quantifying the mechanical energy dissipated over the influence length under a moving train axle. Different descriptions of this energy amount are benchmarked with respect to their predictive value. In the presence of a degraded sleeper support, hanging sleepers are found to develop faster with increasing train speed; the speed effect may be estimated as roughly linear. Moreover, degradation increases progressively with an increasing local relative stiffness reduction. Coincidence of the train speed corresponding to the sleeper passing frequency with the first resonance peak of the system leads to severely increased degradation; increased damping however attenuates dissipation peaks at resonant speeds and shifts their position upwards. The effect of a degraded support is most significant on soft subgrades. The effect of multiple degraded sleeper supports increases up to three sleepers, for any train speed. With respect to the system parameters, particularly the railpad stiffness has significant effect; especially for high-speed tracks a high pad stiffness is very unfavorable. Other effective control parameters in the case of degraded sleeper supports are the sleeper spacing and the rail cross-sectional properties; for example replacing a 54E1 with a 60E1 rail profile may reduce energy dissipation with roughly 30% on high-speed track. An increasing unsprung vehicle mass is unfavorable for track degradation, again with the effect increasing with the train speed. The developed methodology is shown to have significant potential with respect to railway track design in terms of multi-parametric optimization for concrete cases with a given input in terms of soil properties and operational regime.