Corrosion monitoring in steel bars using Laser ultrasonic guided waves and advanced signal processing

Ultrasonic scanning can present a detailed map of the invisible internal degradation of objects. However, their utility is limited in the case of built facilities as their dimensions are too large for the currently used piezoelectric transducers to penetrate. This paper presents an investigation with high power Laser ultrasonic guided waves combined with an advanced signal processing for discerning corrosion in structural steel bars. The pros and cons of Laser ultrasonic monitoring vis-à-vis traditional piezoelectric systems have been highlighted. Dispersion curves for the steel bars have been deduced to understand its modal characteristics. The variations in time-frequency spectra due to the spread in corrosion have been correlated with the predominant propagation modes in the bar. The time-frequency spectral pattern obtained from the proposed signal processing techniques are calibrated with the mass loss due to corrosion. A K-means cluster-based approach has been used to predict the state of corrosion.