Open Access Open Access  Restricted Access Subscription or Fee Access

Damage Localization in Plate-like Structures Using Guided Ultrasonic Waves Edge Reflections

A. EBRAHIMKHANLOU, B. DUBUC, S. SALAMONE

Abstract


This paper proposes a novel idea for ultrasonic imaging of plate structures with a small number of permanently attached ultrasonic sensors. Guided ultrasonic waves excited and captured by a single actuator-sensor pair suffice as input for the proposed imaging technique. Taking advantage of several reflections from boundaries and geometric features of plates, the number of inspection lines is artificially increased. Only the first antisymmetric mode (A0) is used for imaging, because it has higher excitability than the first symmetric mode (S0) and simpler reflection behavior. Mode conversion from reflections is avoided by exciting below the first cut-off frequency of higher order propagating antisymmetric modes. The path and distance traveled by each echo is calculated using a ray tracking model, and a drop in energy of each echo compared to its baseline is associated with the presence of damage. Dispersion curves are used to calculate wave velocity and expected arrival times of the reflected echoes. Based on the expected time of arrival, a gate is defined for each echo, and an energy comparison is conducted for the high energy gated portion of the echo. A probabilistic method for image reconstruction is also proposed to locate damage. The proposed imaging method combines information from several inspection paths using a Bayesian probabilistic framework. To validate the approach, experiments have been carried out on an aluminum plate, instrumented with only two permanently attached low profile circular piezoelectric sensors. A tone burst packet is used an input excitation, and multiple echoed packets have been recorded at a receiving sensor. A large C-clamp is used to locally scatter the waves and simulate damage. Damage is simulated in nine locations, and the proposed method achieves notable success in localizing the applied damage with only two sensors.

doi: 10.12783/SHM2015/313


Full Text:

PDF