Open Access Open Access  Restricted Access Subscription or Fee Access

Accurate Source Localization Using Highly Narrowband and Densely Populated MEMS Acoustic Emission Sensors

M. KABIR, H. SABOONCHI, D. OZEVIN

Abstract


The real time source detection and localization are the fundamental advantages of Acoustic Emission (AE) method. The multi-dimensional localization requires an array of sensors distributed and positioned strategically on the structure. The sensor positions are controlled by the attenuation characteristics of the structure in order to have minimum required wave arrivals to sensors (e.g., minimum three sensors detecting an event for 2D source localization) for an AE event. While minimum three sensors are sufficient for 2D source localization, redundant data collection increases the accuracy of source location, and cleans the data set from non-relevant signals. Additionally, accurate prediction of wave velocity results in better localization result. Current AE sensors are piezoelectric type, and while they are designed to exhibit resonant behavior, they have a certain bandwidth that opens the response to a range of frequencies. In this study, highly narrowband Micro-Electro-Mechanical Systems (MEMS) sensors are presented with more accurate source localization ability due to detecting single frequency, and potential to densely populate on structures due to lightweight, and lowcost properties. The MEMS AE sensors are tuned to a range of frequencies between 60 kHz – 200 kHz for different structural applications (e.g., 60 kHz for concrete and 150 kHz for steel). The accurate prediction of source localization is tested on aluminum plate, and wave velocity of particular frequency is obtained using the dispersion curves. The source localization ability of the MEMS AE sensors is compared with conventional piezoelectric sensors. The efficiency of MEMS AE sensors with densely positioned scheme is discussed to monitor large-scale structures.

doi: 10.12783/SHM2015/372


Full Text:

PDF