Advanced Acoustic Emission-based SHM for Concrete Structures: Real-Time, High-Precision Imaging of Crack Geometry and Damage Source-type Using Moment Tensor Inversion

Abstract

Listen

Translate article

Acoustic Emission (AE) monitoring is a powerful technique for continuously inspecting the health of structures, thereby aiding in the prevention of potential failures. AE consists of elastic waves that are produced and emitted during fracture processes and are captured by sensors. Utilizing quantitative geophysics-based methods, these recorded signals can be processed to monitor and delineate the spatiotemporal progression of fractures in opaque materials, such as concrete, in real-time. Moreover, the classification of damage source types, namely shear versus tensile, by processing AE waveform features, is crucial for appropriate structural treatment. Owing to the complex nature of the recorded elastic signals and the heterogeneity of the media, data are typically processed manually, covering both AE localization and source type classification. This, coupled with the high processing costs of extensive datasets, often exceeding terabytes, has limited the method's practical applications. In this paper, we introduce a novel automated and high-precision AE monitoring algorithm and software (Momeni et al., 2021a, b) suitable for SHM. It is designed to work with various standard data formats and is capable of handling both trigger-based and continuous data, and medium heterogeneity. We present results from implementing this software in the AE monitoring of two 4.88-meter-long concrete beams in a laboratory setting, comparing it with manually processed AE data (Mhamdi, 2015). Our approach enabled the identification of over three times more AE sources than manual processing, achieving higher precision. By processing waveform features in both time and frequency domains, we successfully classified the damage sources into three categories: tensile, shear, and mixed-mode, at different stages of the experiment. With adequate processing units, the software can operate in parallel, facilitating real-time SHM with exceptional precision in imaging both crack geometry and source types. This provides invaluable information to decision-makers regarding the nature of the data captured.