Detection of crack growth in rail steel using acoustic emission

Increased traffic speeds and axle loads on modern railways enhance rail track degradation. To eliminate track failure due to rail defects, a condition monitoring system requires methods for the early detection of defects which grow in service. Acoustic emission (AE) monitoring is the only non-destructive technique which might be applied online to study the defect growth under traffic loading. However, a high level of traffic noise and a limited signal from crack growth, especially at low crack growth rates, significantly complicate the AE signal analysis. In the present work, the AE monitoring of rail steel fatigue was carried out in a 'noisy' laboratory environment using different methods of signal analysis. Signal parameters of AE for machine noise, sample deformation and crack growth were identified. The crack growth related AE signature was found to be dependent on fracture mode.