Early detection of damage in concrete infrastructure is critical to prolonging structural service life through timely maintenance, ensuring safety and preventing failure. Current management practices rely on regular visual inspections, which can be subjective and limited to accessible locations. In the field of structural health monitoring (SHM), there remain key limitations: (1) Indirect damage sensing, which requires complex physics-based models and algorithms to correlate structural response measurements to damage state; (2) Point-based sensing, which cannot accurately identify spatially distributed damage such as cracking and corrosion. Researchers have begun to explore distributed sensing methods such as ultrasonic guided waves and acoustic emission. These methods require expensive instrumentation, intricate network of sensors, and suffer from data contamination due to background noise and secondary sources. Most importantly, they are significantly more difficult to apply to concrete structures than steel structures. Distributed sensing that can provide the spatial resolution necessary to localize and quantify the severity of concrete infrastructure deterioration and damage is direly needed.

In lieu of reliance on point-based sensors, this dissertation develops a distributed, direct damage and strain sensing approach based on novel multifunctional strain-hardening and self-sensing cementitious materials (MSCs). In this work, multifunctional cementitious materials are encoded with damage tolerance and spatial self-sensing capacity. The sequential formation of steady-state microcracks, rather than localized cracking, enables a prolonged and intrinsically controlled damage process, while allowing detection of microcracking damage level in the material long before failure occurs. The beauty of multifunctional concrete materials is two-fold: First, it serves as a major structural material component for infrastructure systems with greatly improved resistance to deterioration under service loading conditions, and to fracture failure under extreme events. Second, it offers capacity for distributed and direct sensing of cracking and straining wherever the material is located, while eliminating the need for sensor installation and maintenance.

The dissertation research is driven by two central hypotheses: (I) Cementitious materials exhibit an AC frequency-dependent electrical response, which would depend on its heterogeneous microstructure. Mechanical straining and damage process would lead to a change in the microstructure, thus affecting the frequency-dependent electrical response. A strong, high signal-to-noise coupling between cementitious material electrical response and mechanical behavior at different length scales would enable a self-sensing functionality during elastic and post-cracking stages; (II) Using electrical stimulation and advanced tomography methods, spatial mapping offering a visual depiction of concrete damage and deterioration can be gained. Spatial sensing of damage location and level inside a structural element can be achieved through electrical probing only from the boundaries.

This dissertation generates new understandings on how the composition and microstructure affect the frequency-dependent electrical response of cementitious materials. 4-point AC impedance spectroscopy is integrated with equivalent circuit modeling analysis to reveal the electrical microstructures and properties of cementitious materials, and their interfaces with electrodes. The work also elucidates whether and how cracking and healing (as a reversed damage process) lead to changes in material electrical response. An idealized model circuit is formulated to predict the frequency-dependent electrical behavior of cementitious composite materials at various cracking and healing levels. The model is tested and validated through a series of experimental measurements. Analyzing the changes of model parameters due to material composition, mechanical strain, and damage processes reveal the mechanisms that contribute to the overall electromechanical response of cementitious materials.

This dissertation develops novel multifunctional cementitious composite materials that integrate self-sensing functionality with a pseudo-strain-hardening behavior accompanied by multiple steady-state microcracking process. The macroscopic electro-mechanical properties, which are strongly coupled through high signal-to-noise ratios, are achieved by tailoring the material electrical microstructure as well as micromechanical parameters. Through experimental investigation, the dissertation reveals the material strain sensing behavior at elastic, pseudo-strain-hardening, and tension softening stages under various loading scenarios and environmental conditions. In addition, a modeling framework is established to link length scales from single fiber/matrix interfacial electromechanical behavior, to a single crack bridged by numerous fibers with statistically random embedment lengths and orientation, and to multiple cracking process and final localized failure. The analytical model couples micromechanics theory with equivalent circuit model, to bring mechanistic insights into the material electromechanical response as the basis for self-sensing.

Furthermore, this dissertation realizes spatial sensing and visual depiction of damage through combining with advanced electrical impedance tomography (EIT) methods. It establishes two EIT methods that are suitable for solid heterogeneous concrete elements: time-difference EIT and frequency-difference EIT. It evaluates different variables, e.g., material type, real or imaginary parts of impedance, probing frequency, and finite element algorithm, on the effectiveness of EIT methods and image reconstruction. Impedance reconstruction is an ill-posed inverse problem. Finite element models that describe the forward problem are implemented, and the inverse solution is solved using regularized least square analysis (e.g., Tikhonov regularization). This work makes it possible to spatially visualize material impedance and damage upon voltage measurements collected from probe locations at element boundaries. It demonstrates the effectiveness of the EIT methods using 2D (plate) and 3D (beam) specimens with different levels of cracking damage. Finally, the EIT methods are further extended for corrosion sensing in reinforced concrete specimens.