Detecting early indications of head checks in rail systems is paramount for effective service and maintenance, presenting substantial cost savings through the anticipation of crack growth behaviour. Current methodologies, such as wayside monitoring employing equipment like laser vibrometers or specialized test trains, have been in operation for several years. However, these approaches often prove cost-prohibitive, overly complex, or overly sensitive to environmental variations. This creates a critical void in the availability of an affordable and resilient sensor network specifically designed for installation in crucial rail locations.
Piezoelectric sensors emerge as promising candidates due to their cost-effectiveness and versatility. While existing literature introduces some piezo-based systems, they frequently grapple with the intricacies of signal complexity or impractical installation locations on the rail profile, such as the sides of the head.
The current project aims at addressing these issues by implementing a pitch-catch network of permanently installed piezo sensors, presenting an affordable solution for rail network operators. Positioned on the web of the rail, these sensors avoid collision with passing wheels, enhancing practicality. A dedicated hardware/software setup is designed to excite, collect, and analyze signals, overcoming the limitations of passive systems.
To validate the setup's performance, a pearlite rail sample underwent artificial loading designed to provoke head checks at a special wheel-rail test rig. Surface acoustic waves generated and collected by the piezo sensors every 10,000 cycles were analyzed. The results demonstrate the promising potential of the proposed setup for field applications, providing a reliable and cost-effective solution for early head check detection in rail systems.

Condition monitoring; rail head checks; surface acoustic waves