Ballast particle dynamic response as a function of subgrade support using statistical methods

Abstract

Railroad track substructure condition has historically been difficult to monitor since the substructure is not visible without displacing ballast, making it challenging for engineers to observe degradation and to optimally schedule maintenance. Many methods to accurately classify track substructure condition require invasive measurement techniques that disrupt normal railroad operations. Nondestructive testing methods for the track substructure have been developed to avoid disrupting ballast sections but often require expensive equipment and delay track operations. Many railroads opt to conduct maintenance on a conservative timeline instead of routinely investigating the substructure condition such that the track remains in good operating condition. Research using SmartRocks (battery-powered, wireless sensors that physically resemble ballast particles in shape, size, and texture) has shown that SmartRocks can be used as a tool for maintenance engineers to better understand railroad track substructure behavior. Ballast inter-particle interactions under dynamic wheel loading were quantified for two track sections with different support conditions using SmartRock technology. Correlation statistics and K-means clustering were used to identify time series’ characteristics that detail SmartRock vibration and movement along and around the track coordinate axes. The analysis showed a clear dampening effect on SmartRock vibration from the use of under-sleeper tie pads. Furthermore, large vertical vibrations of the SmartRocks associated with passing trains led to large rotational displacements. The measured SmartRock behavior indicates a strong correlation between support conditions and the magnitude of ballast inter-particle movement, suggesting that SmartRocks could be used as a tool to quantify ballast behavior without using invasive measurement devices and disrupting railroad operations.