Detecting corrugation defects in harbour railway networks using axle-box acceleration data

Sea- and inner ports are intermodal traffic nodes that play an important role in transportation, especially in the transportation of goods. The appearance of track defects in a harbour railway network has a negative impact on safety, cost and comfort (for example due to noise emission). The analysis of data obtained by embedded acceleration sensors, which are installed at the axle box of an equipped in-service vehicle, allows for continuous condition monitoring of the track infrastructure. The German Aerospace Center (DLR) develops prototypical modular multi-sensor systems that are used in different operational environments, including on a shunter locomotive operating in an industrial harbour railway network in Braunschweig, Germany. Within the HavenZuG research project, extensive rail longitudinal profile and track geometry measurements have been performed using established inspection methods to obtain the true underlying condition of the railway network. In the present paper, methods for gaining relevant information from the axle-box acceleration (ABA) data are presented and validated with the given reference data. The focus is on detecting defects that are visible in the rail longitudinal profile, mainly rail corrugation. It can be shown that ABA data gathered during everyday shunting operation can be used for detecting corrugation and for inferring rail longitudinal profile parameters.