Optimization of Acoustic Source Localization in Large Plates

K. GRABOWSKI, H. NAKATANI, M. GAWRONSKI, P. PACKO, W. J. STASZEWSKI, T. UHL, T. KUNDU

Abstract


Acoustic Emission (AE) is a phenomenon of great importance for structural health monitoring (SHM). One of important applications of AE in SHM is the localization of a source of acoustic energy in the structure. For certain types of structures and for large propagation distances where dispersion phenomenon is prominent, it is hard to estimate proper location of the damage. This paper discusses some aspects of AE source localization techniques. It is shown how an acoustic source in large plates can be localized using only six sensors and without apriori knowledge of any material pa- rameter, thus omitting the need of considering dispersion phenomena. The method is based on a predefined three-sensor setup, required to determine the Lamb wave propagation direction by detecting the angle at which the Lamb wave hits the array. As introduced in earlier works, the propagation direction of the Lamb wave is detected using arrival time differences between sensors, measured by correlating only the first dip and peak of the recorded signals. With another propagation direction detected by another sensor cluster the acoustic source can be localized as an intersection point of the two directions for planar structures. In this paper special orientation of the sensor cluster is suggested in order to apply the technique to a narrow plate. This modification enables to evaluate a wider area and avoid possible errors in detected propagation directions from some acoustic source locations. The method is compared with another source localization technique based on combined time-domain dispersion removal approach and localization. The dispersion removal technique reconstructs the source signal compensating distortions resulting from dispersion effects and transforms the time domain response directly into the distance one. Both methods are verified numerically and experimentally through artificial sources.

doi: 10.12783/SHM2015/194


Full Text:

 Subscribers Only