On the application of UPV in health monitoring of Indian Rail section using AE Technique

Abstract

The health monitoring of rail sections from the perspective of safety is a significant field of research in the context of the Indian Railway, which is thought to be Asia's second-largest rail network. The main focus of health monitoring is to assess damages and determine their location in order to initiate the appropriate measures. One of the well-known Non-Destructive Techniques (NDTs), the Acoustic Emission (AE) technique can be used for this purpose. However, the literature suggests that the appropriate localization of damage, which is in fact AE source, is challenging due to the complex geometry of the rail section. Continuous research going on in this direction and the pencil lead break (PLB) method often serves the purpose of simulated AE source in the experimental investigation of these research works. There are a few limitations present in the PLB method due to its manual operation. In the practical scenario, PLB requires expertise to operate perfectly. Moreover, the complexity of the application of PLB is increased in case of uneven surfaces. In this context, PLB can be replaced by an Ultrasonic Pulse Velocity (UPV) instrument. UPV instrument produces an ultrasonic signal which is very similar to PLB. In this study, an effort is paid to detect and localize the simulated AE sources produced by a UPV instrument in the Indian rail section under laboratory conditions using AE technique. The localisation of such simulated AE sources is done using an Artificial Neural Network (ANN) based methodology. The ANN model is developed using the basic parameters of the AE signal. Further, a set of PLB sources are also applied to the rail section and localization of these sources is also done. Localisation results of both simulated AE sources, PLBs and UPVgenerated ones, are compared. It is found that the localization using ANN of the simulated AE source is done with a minimal error of 1.3%. The technique is promising and there is a good scope for future research and application in health monitoring of different types of structures.