This article investigates the application of vibro-acoustic modulation testing for diagnosing damage in concrete structures. The vibro-acoustic modulation technique employs two excitation frequencies on a structure. The interaction of these excitations in the measured response indicates damage through the presence of sidebands in the frequency spectra. Past studies using this technique have mostly focused on metals and composites (thin plates or laminates). Our research focuses on concrete, which is a highly heterogeneous material susceptible to a variety of chemical, physical, and mechanical damage processes. In particular, this article investigates diagnosing cracking in concrete from an expansive gel produced by an alkali–silica reaction in the presence of moisture. Past studies have been limited to damage detection using vibro-acoustic modulation testing, whereas this article extends the technique to damage localization. A cement slab with pockets of reactive aggregate is used to investigate the diagnosis technique. The effects of different testing parameters, such as locations, magnitudes, and frequencies of the two excitations, are analyzed and incorporated in the damage localization methodology. A Bayesian probabilistic methodology is developed to fuse the information from multiple test configurations in order to construct damage probability maps for the test specimen. The results of vibro-acoustic modulation–based damage localization are validated by petrographic study of cores taken from the slab.

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.

No abstract

A primary challenge for the current fleet of light water reactors in the United States is age-related degradation of their passive assets that include concrete, cables, piping, and the reactor pressure vessel. Various nondestructive techniques exist for locally assessing degradation in passive structures in nuclear power plants. This paper presents results from the analysis of acoustic data acquired using the vibro-acoustic modulation (VAM) technique on a medium-sized concrete slab (2 × 2 × 0.5 ft), which has undergone degradation due to alkali-silica reaction (ASR). VAM is a nonlinear vibration technique in which the structure of interest is excited using a combination of specific frequencies and the response is recorded. VAM assumes that an undamaged structure can be represented by a linear system while the representation of a damaged structure must include nonlinearity. Two piezo-stack actuators are used to excite the slab. One stack is dedicated to excite the slab at a low frequency, referred to as pump frequency, and another stack is dedicated to excite the slab at a high frequency, referred to as probe frequency. The concrete slab was cured in an environmental chamber to accelerate ASR-related degradation. In this paper, the measured acoustic signals are analyzed in both the time and frequency domains to understand state of health of the concrete specimen.

No abstract