Search

As the cracking mode (tensile or shear) of a crack is related to the underlying physical mechanisms, crack mode classification is a very useful method to identify the damage state of a structure for proper maintenance to enhance structural safety and durability. Acoustic Emission (AE) is a passive structural health monitoring technique based on the stress wave generated due to cracking in a structure. A framework has been designed in this study for automated probabilistic classification of the cracks in cementitious components based on the AE signals. With this approach, unlabeled hand designed waveform parameters, i.e. RA values (RA) and Average frequency (AF) are clustered using density dictated unsupervised clustering algorithm. Intersecting clusters in the data were then separated through a hyperplane created using Support Vector Machine (SVM) algorithm. Based on physical insight obtained from labeled data, unlabeled data was classified into events corresponding to different cracking modes. The framework was applied to the analysis of AE data from Steel Fiber Reinforced Concrete (SFRC) beam under bending and Strain Hardening Cementitious Composite (SHCC) samples under direct tension. The cracking modes obtained from the proposed machine learning approach are found to be in good agreement with expectations based on composite theory. With good qualitative prediction, the proposed approach shows promise for the prediction of damage state in structures based on unlabeled data obtained in the field.

Characterization of surface cracks is one of the key steps towards condititional assessment and understanding the durability of strain hardening cementitious composites (SHCCs). Under laboratory conditions, surface crack statistics can be obtained from images of specimen surfaces through manual inspection. This is subjective, time consuming and laborious. In order to automate this process a framework encompassing a deep learning model for rapid identification and computation of crack parameters of thin SHCC cracks in presented in this work. A tailored deep convolutiontional neural network (TDCNN) was trained to detect thin cracks and then crack parameters were computed using image processing technique. The proposed technique does not require optimal lighting conditions, proper surface treatment, and prior (manual) selection of the correct region for proper inference.. The results from the controlled study suggest that the inference ability of TDCNN is reasonably good, resilient against epistemic uncertainty and tunable for completely independent but adverse observations. From the crack pattern computed using TDCCN, crack parameters-average crack width (ACW) and crack density (CD) can be calculated to facilitate conditional and durability assessment in a practical environment.

Accurate time of arrival (TOA) detection is essential for successful wave and vibration-based structural health monitoring (SHM) system including Acoustic Emission (AE). AE is the foremost among passive wave-based monitoring systems. During high burst rate results, AE waves could overlap. Overlapping of AE waves leads to overlap in both time and frequency. As a result, the TOA information is lost in the coda of the previous wave and cannot be (reliably) detected by conventional techniques. To this end, ICD is introduced. ICD involves 3 cascaded systems a) (Overlapping) Identification b) (Overlap) Cleaning c) (Arrival) Detection. A novel 3D Fingerprint is meticulously designed to autogenously and injectively identify the overlapping waves. Positive identification activates the cleaning system which eradicates the influence of the Intersecting Wave using newly proposed adaptive spectral subtraction (ASpS). Then, TOA for the Intersected Wave was identified. The approaches were verified in controlled as well as source localization tests. The results from controlled study validate robustness in low IRR values for a wide range of waveform parameters. Finally, source localization results from laboratory testing confidently display the scientific applicability of the proposed system. This system could enhance detectability, reliability, and accuracy of a conventional system.

Strain Hardening cementitious composites (SHCCs) have been known to produce excellent crack width control and very good tensile strain capacity. As a result, strategically introducing SHCCs would lead to significant carbon saving through the creation of structures that are durable and could potentially undergo autogenous self-healing. However, for wider acceptability of any new material requires long term empirical data. This may be circumvented by continually ensuring a structure made with this new material is serviceable i.e. material has not failed. We present an approach to predict failure time without any prior information on operating conditions and material information for SHCCs. This method is based on the modeling of ‘finite time singularity’ based on mathematical laws. The method produces fairly accurate results when applied to estimate failure for SHCCs with a wide range of failure times and to both linear and non-linear loading regimes. The practical applicability of this method was thus demonstrated.

In strain-hardening cementitious composites, the bridging effect of fibers can control crack openings to very small values so the cracks can be self-healed in the presence of water. To rely on autogeneous self-healing to enhance the durability of structures, a method to assess the state of self-healing is required. A new strategy is presented in this paper for in-situ measurement of the healing effect without pre-knowledge of the environmental condition (such as moisture). Exploiting the fact that the electrical conductivity in the crack is lower than that of the undamaged section, the effective macroscopic cracking effect is first distinguished by frequency dependent impedance measurements both along (parallel to) and across the cracks. Dielectric spectroscopy along the cracks is found to correlate well with gravimetric measurements and can therefore, be used for inverse approximation of moisture content. The effect of moisture change on the equivalent resistance in the direction across the cracks can then be derived approximately from measurements under various moisture contents. The approximated resistance vs moisture curve is found to reflect the evolution of autogenous self-healing, and a curve indicating almost complete healing agrees well with the recovery in member stiffness measured from loading test.

For strain-hardening cementitious composites (SHCCs) the knowledge of operating regime is important for maintenance planning and monitoring of the element. This paper presents a methodology of regime discrimination for SHCC from acoustic emission (AE) signals. From an AE signal, various damage indicators (DIs) can be derived. A new DI called Power Spectral Entropy (PSE) is developed in this paper. New benchmarks are developed to quantify possible effect of external factors on the measurement accuracy. Theoretical results indicate that the PSE is Signal to Noise Ratio (SNR) invariant, insensitive to the choice of subjective parameters and can be performed in real time. PSE was then obtained from the test result of SHCC with elastic behavior followed by strain hardening and softening. The test results indicate that the PSE varies with strain in a very similar way to the applied load. An approach to distinguish between different operating regimes of a SHCC component based on PSE is then proposed and validated. The practical applicability of PSE is hence demonstrated.