RAILROAD GEOMETRIC DEFECTS: A METHODOLOGY TO DEVELOP DEGRADATION MODELS FROM EXPERIMENTAL DATA

In recent years, almost all railway management companies are equipping themselves with more complex high performance devices in order to monitor track condition. Periodic control of railway characteristics is strongly advised to prevent and reduce transportation risk. Wear defects, as well as the erroneous design of railway structures, are among the main causes of derailment. Various approaches have been considered to analyze, manage and reduce risk. Local transportation companies should have available and reliable predictive degradation models, based on empirical and mechanistic considerations obtained from experimentally collected rail defect data. These would be useful to predict the evolution of track conditions to be used in decision support systems for Railway Track Maintenance. In this paper an analysis approach following a large experimental measurement campaign carried out on a suburban narrow-gauge railway network operating in the Naples metropolitan area is proposed. The data have been obtained by the Plasser device (POS/TG System), directly mounted on a vehicle Fe ETR 220’s wheelset during its regular operation. Track cant, level, gauge and twist have been collected almost every month in a year since 2007 to 2013 with a 25 cm sampling interval on track sections up to 40 Km long. Preliminary results seem to indicate that by coupling track geometric data analyzed so far with alignment and traffic, a more effective evaluation of track defect can be obtained.