Diagnosis of internal cracks in concrete using vibro-acoustic modulation and machine learning

Abstract

This paper investigates the utility of physics-informed machine learning models for vibro-acoustic modulation (VAM)–based damage localization in concrete structures. Vibro-acoustic modulation is a nonlinear dynamics-based non-destructive testing method, which was initially developed to perform damage detection and later extended to accomplish damage localization. The VAM-based damage (hidden crack) diagnosis is performed by analyzing the damage index pattern on the surface of the component to arrive at the size and location of the hidden damage. Past investigations have employed heuristically selected damage index thresholds as well as computationally expensive Bayesian estimation methods for VAM-based damage localization in two (surface) dimensions. Compared to these studies, the proposed methodology automates the threshold selection (algorithmic instead of heuristic), increases the speed of the probabilistic damage diagnosis process, and enables the estimation of damage depth. We generate training data (damage index) for the machine learning models using the pertinent nonlinear dynamics (finite element) models using different combinations of test parameters. The (supervised) machine learning models are thus informed by computational physics models. These include two types of artificial neural network (ANN) models: classification models that identify whether a sensor location is damaged or not and regression models that enable Bayesian estimation to obtain the posterior probability distribution of damage location and size. The accuracy of machine learning-based diagnosis is evaluated using both numerical and laboratory experiments. The proposed physics-informed machine learning models for VAM-based damage diagnosis are able to achieve an accuracy of about 60–64% in the validation experiments, indicating the potential of these methods for internal crack detection. The results show that for complex (nonlinear dynamics-driven) diagnostic methods, damage index patterns learned from physics models could be successfully used for damage detection as well as localization.

A novel time domain-based multiple breathing crack diagnosis technique using Quadratic-Teager Kaiser Energy (Q-TKE)

“…The advantages of bitone over single tone in nonlinear feature extraction have been demonstrated very clearly in the literature. 5–7…”

“…It can be observed from the above Equations ( 6)-( 8) that the amplitudes of each harmonic component of the time signal largely increase (i.e., A m is multiplied with the frequency w n ) in TKE due to the energy modulation effect. The energy modulation effect refers to the energy component resulted out of modulation of higher order harmonic (nth superharmonic) and first order harmonic in Equation (7). In Q-TKE, the amplitudes are further increased.…”

“…Identification of fatigue-breathing cracks at an early stage offers critical information about the health of the structures, aiding in preventing catastrophic failure and saving human lives. Several vibration-based methods have been proposed by researchers [1][2][3][4][5][6][7] for breathing crack identification (addressing all three steps: detection, localization, and characterization) using single tone, bitone, and multi-tone harmonic vibration data. Due to the periodic opening and closing of fatigue cracks under single-tone harmonic excitation, the structure vibrates at the excitation frequency and higherorder harmonics.…”

“…While the dynamic behavior of beams with a single breathing crack has received much attention, [3][4][5][6][7][8][9][10][11][12] very limited investigations have been reported on multiple breathing cracks. [13][14][15][16][17][18][19] As the number of breathing cracks present in the beams and their spatial locations is usually not known a priori, detection and localization of multiple breathing cracks are much more challenging.…”

“…There exists a lot of research evidence on extraction of low amplitude nonlinear features such as superharmonics and mixed harmonics or intermodulations from the vibration response using frequency domain approaches. [1][2][3][4][5][6][7][8] Only very few studies are completely time-domain approaches [22][23][24][25] and they mostly aim at addressing single breathing crack identification. The extraction and enhancement of the nonlinear features in the time domain especially toward multiple breathing crack localization is highly complex and challenging.…”