Acoustic Emission Monitoring of Progressive Damage of Reinforced Concrete T-Beams under Four-Point Bending

Abstract

Acoustic Emission (AE) is revealed to be highly adapted to monitor materials and structures in materials research and for site monitoring. AE-features can be either analyzed by means of physical considerations (geophysics/seismology) or through their time/frequency waveform characteristics. However, the multitude of definitions related to the different parameters as well as the processing methods makes it necessary to develop a comparative analysis in the case of a heterogeneous material such as civil engineering concrete. This paper aimed to study the micro-cracking behavior of steel fiber-reinforced reinforced concrete T-beams subjected to mechanical tests. For this purpose, four-points bending tests, carried out at different displacement velocities, were performed in the presence of an acoustic emission sensors network. Besides, a comparison between the sensitivity to damage of three definitions corresponding to the b-value parameter was performed and completed by the evolution of the RA-value and average frequency (AF) as a function of loading time. This work also discussed the use of the support-vector machine (SVM) approach to define different damage zones in the load-displacement curve. This work shows the limits of this approach and proposes the use of an unsupervised learning approach to cluster AE data according to physical and time/frequency parameters. The paper ends with a conclusion on the advantages and limitations of the different methods and parameters used in connection with the micro/macro tensile and shear mechanisms involved in concrete cracking for the purpose of in situ monitoring of concrete structures.

Keywords: RA-value; acoustic emission (AE); average frequency; b-value; four-point bending; reinforced concrete beam; support-vectors machine; unsupervised machine learning.

Figure 1

Schematic representation of (a) an AE signal and some important AE parameters, (b) tensile crack, and (c) shear crack.

Figure 2

SVM, support vectors, hyperplane (schematic).

Figure 3

Flow chart showing steps involved in the adopted unsupervised learning scheme.

Figure 4

Dimensions of the reinforced concrete T-beam (in mm): (a) longitudinal and (b) cross sectional dimensions.

Figure 5

Reinforced concrete T-beam subjected to four-point bending and acoustic monitoring.

Figure 6

Experimental set-up: UTM, sample beam, and AE setup.

Figure 7

(a) Evolution of load as a function of displacement, (b) evolution of load as a function time, (c) initial cracks, and (d) rupture of the beam.

Figure 8

Schematic presentation of cracks and the corresponding dominating AE signals: (a) cracks during initial stages of loading and AE signal, and (b) cracks during final stages of loading and AE signal.

Figure 9

(a) Variation of load and AE amplitude with time (this observation is found to be consistent for each of the samples) and (b) variation in cumulative AE hits with time and corresponding linear fit.

Figure 10

Evolution of the three types of b-values during the performed mechanical loading at (a) 1 mm/s, (b) 2 mm/s, and (c) 4 mm/s.

Figure 11

Variation of load, average frequency (AF), and RA value with respect to time: (a) sample 1, (b) sample 2, and (c) sample 3.

Figure 11

Variation of load, average frequency (AF), and RA value with respect to time: (a) sample 1, (b) sample 2, and (c) sample 3.

Figure 12

Labeling of the data set into three separate classes (i.e., Zone 1, Zone 2, and Zone 3).

Figure 13

Classification using SVM (for sample 1 data): (a) using linear kernel and (b) using Gaussian kernel.

Figure 14

Two labels of SVM classification using Gaussian kernel in the case of sample 1.

Figure 15

Different steps of the adopted unsupervised learning method (shown for AE data of sample 1): (a) Standard deviation of features, (b) feature selection using Laplacian scores, (c) screen plot obtained from PCA analysis, and (d) selection of the optimal number of clusters for k-means using cluster validity indices, i.e., Davies–Bouldin (DB) index and Silhouette Coefficient (SC).

Figure 16

Clustering using k-means and significance of the obtained clusters (shown for AE data of sample 2): (a) obtained clusters, (b) variation in cumulative hits of the clusters; a typical wavelet transformed signal of (c) cluster 1, (d) cluster 2, and (e) cluster 3.