PSI - Issue 66

Vivek Vishwakarma et al. / Procedia Structural Integrity 66 (2024) 381–387 Vishwakarma and Ray/ Structural Integrity Procedia 00 (2025) 000–000

386

6

fatigue testing were categorized using the same AF and RA criteria established after monotonic testing. This allowed for consistent identification of mode-I cracking events under both loading regimes. The primary goal was to estimate the maximum crack length within each temporal bin, acknowledging the challenges posed by the continuous opening and closing of the crack under fatigue. The results demonstrated the feasibility of applying the spatial binning approach to fatigue loading, estimating crack growth progression under cyclic loading conditions. Beams are tested under two different load ratios, 0.6 and 0.7 of constant amplitude. The frequency of the cyclic load is kept at 2 Hz. The crack profile obtained through the proposed methods is shown in Figure 6

R = 0.6

R = 0.7 Figure 6 Crack path and 3D crack profile in case of fatigue loading

5. Conclusions This study introduces a novel application of Acoustic Emission (AE) technology for real-time crack monitoring in lightly reinforced concrete beams, utilizing both spatial and temporal binning techniques alongside Gaussian Mixture Model (GMM) clustering. The GMM approach effectively classifies AE signals based on their AF and RA values, enabling the selection of highly probable Mode I (tensile) crack events based on their posterior probabilities. This probabilistic filtering ensures that only the most relevant AE events are considered for further analysis, enhancing the accuracy of crack detection. The spatial binning method organizes AE data into location-based bins and offers a novel way to simulate crack paths accurately throughout the material's depth. This approach, validated against Digital Image Correlation measurements, provides a more comprehensive view of crack propagation than surface-based methods alone. The temporal binning strategy introduces a reliable means of tracking crack length evolution over time, using a monotonic constraint to filter noise and secondary failure modes. Together, these methods improve the capability of AE technology to monitor and predict crack behaviour in reinforced concrete structures. Future work will extend these methods to more complex cracking patterns, further expanding their applicability in long-term durability assessments.