Valutazione della robustezza di un sistema di misura ferroviario mediante tecniche di “progettazione al calcolatore degli esperimenti" // Evaluation of the robustness of a railway measurement system by means of “design of computer experiments” techniques

In railway operations, data provided by wayside measuring systems could be useful for many reasons as ensuring safety and facilitating maintenance of the tracks. For example, measuring devices of wheel-rail contact forces allow the monitoring of any kind of rolling stock under specific operating conditions, like curves at low speed or twisted tracks, which are always present in stations, depots and marshalling yards. For that reason, in this paper we present the evaluation of the robustness of the SMCT wayside measurement system. The system is based on a particular strain gauges configuration on the rail foot, able to isolate the effects of the lateral force from those of the vertical one and record the lateral force using a proportional coefficient between that force and strains. The system has been validated with several field tests directly on a railway line in service: about 4000 transit and different types of trains, up to 110 km/h. The aim of the work is to build a fitting regression model able to predict the value of the ratio between the lateral force and strains in order to test the robustness of the measurement method and its operating range, modifying parameters that experimentally could not have been changed.