PSI - Issue 54

Francisco Afonso et al. / Procedia Structural Integrity 54 (2024) 545–552 Francisco Afonso / Structural Integrity Procedia 00 (2019) 000 – 000

546

2

1. Introduction Besides contributing to condition monitoring, the timely detection of failures in the railway system can be invaluable for human and material safety, by anticipating potential hazards. As indicated by Davari et al. (2021), preventing failures in high operational risk industries such as railways is crucial to improve efficiency and effectiveness in several levels. In such an environment, the implementation of an efficient maintenance approach that helps minimizing operational costs and increase productivity is essential (Davari et al., 2021). Podofillini et al. (2006) note in their study that neglecting the presence of defects such as cracks and track misalignments poses a significant challenge to the railway system operation, worsening over time if not corrected, and may even lead to accidents such as derailment due to complete rail breakage. Davari et al. (2021) suggest that the prevention of such occurrences can be accomplished through the introduction of cyber-physical systems in Industry 4.0. These systems typically include the integration of systems and tools through intercommunication systems and intelligent data processing. This paper focuses its developments towards two major areas of railway operation, the railway tracks and the train’s wheels. According to Castillo-Mingorance et al. (2020) in their review, the monitoring of railway tracks is important since track operability directly impacts maintenance time, resources and costs. Monitoring these components is essential when tracking the infrastructure’s condition and can be achieved, as previously mentioned, by implementing sm art sensors. Additionally, regarding the train wheels, Yang and Létourneau (2005), state that wheel failures, in the global train industry, account for half of all train derailments by accelerating rail deterioration, in some cases, causing the rail to break prematurely. The development and implementation of systems devised for the railway is expected to enable the monitoring of the railway component’s physical condition over time, allowing intervention when necessary to avoid complications. The systems discussed in this article were developed as part of a further encompassing project, Ferrovia 4.0 , which aimed to conceive and develop new tools that can be implemented in the Portuguese railway system, improving its overall quality in the transport service, asset management, safety and passenger comfort. In the scope of this project, with the intent of detecting surface defects in important components, two different systems were developed, the train wheel and railway track defect detection systems. While the railway track surface defect detection system is to be implemented in a fully assembled train, the train wheel surface defect detection system was developed to be implemented in a maintenance context. Both systems must also comply with a few requirements, such as, operations that may compromise the physical integrity of the train, such as welding or drilling, cannot be implemented on the train structure; the technician must be alerted of the detected surface defects as the system executes its analysis and the railway track system must also communicate with a maintenance platform, indicating when a surface defect has been detected.

Nomenclature RCF

Rolling contact fatigue Track recording vehicles

TRV FOV MD AVG MED MM AMM MQTT

Field of view

Maximum difference.

Average. Median.

Metric calculated by subtracting the value of the median from the value of the average.

Absolute value of MM.

Message Queuing Telemetry Transport communication protocol.

2. Defect detecting sensors As explained in the review conducted by Falamarzi et al. (2019), when it comes to railway tracks, there are many different types of defects, such as internal head defects, web and foot defects, surface defects, defective wheels and