Advanced fibrous composite materials exhibit outstanding thermomechanical performance under extreme environments, which make them ideal for structural components that are used in a wide range of aerospace, nuclear, and defense applications. The integrity and residual useful life of these components, however, are strongly influenced by their inherent material flaws and defects resulting from the complex fabrication processes. These defects exist across multiple length scales and govern several scale-dependent inelastic deformation mechanisms of each of the constituents as well as their composite damage anisotropy. Tailoring structural components for optimal performance requires addressing the knowledge gap regarding the microstructural material morphology that governs the structural scale damage and failure response. Therefore, there is a need for a high-fidelity multiscale modeling framework and scale-specific in-situ experimental characterization that can capture complex inelastic mechanisms, including damage initiation and propagation across multiple length scales. This dissertation presents a novel multiscale computational framework that accounts for experimental information pertinent to microstructure morphology and architectural variabilities to investigate the response of ceramic matrix composites (CMCs) with manufacturing-induced defects. First, a three-dimensional orthotropic viscoplasticity creep formulation is developed to capture the complex temperature- and time-dependent constituent load transfer mechanisms in different CMC material systems. The framework also accounts for a reformulated fracture mechanics-informed matrix damage model and the Curtin progressive fiber damage model to capture the complex scale-dependent damage and failure mechanisms through crack kinetics and porosity growth. Next, in-situ experiments using digital image correlation (DIC) are performed to capture the damage and failure mechanisms in CMCs and to validate the high-fidelity modeling results. The dissertation also presents an exhaustive experimental investigation into the effects of temperature and manufacturing-induced defects on toughened epoxy adhesives and hybrid composite-metallic bonded joints. Nondestructive evaluation techniques are utilized to characterize the inherent defects morphology of the bulk adhesives and bonded interface. This is followed by quasi-static tensile tests conducted at extreme hot and cold temperature conditions. The damage mechanisms and failure modes are investigated using in-situ DIC and a high-resolution camera. The information from the morphology characterization studies is used to reconstruct high-fidelity geometries of the test specimens for finite element analysis.

This thesis encompasses research performed in the focus area of structural health monitoring. More specifically, this research focuses on high velocity impact testing of carbon fiber reinforced structures, especially plates, and evaluating the damage post-impact. To this end, various non-destructive evaluation techniques such as ultrasonic C-scan testing and flash thermography were utilized for post-impact analysis. MATLAB algorithms were written and refined for the localization and quantification of damage in plates using data from sensors such as piezoelectric and fiber Bragg gratings sensors. Throughout the thesis, the general plate theory and laminate plate theory, the operations and optimization of the gas gun, and the theory used for the damage localization algorithms will be discussed. Additional quantifiable results are to come in future semesters of experimentation, but this thesis outlines the framework upon which all the research will continue to advance.

This thesis encompasses research performed in the focus area of structural health monitoring. More specifically, this research focuses on high velocity impact testing of carbon fiber reinforced structures, especially plates, and evaluating the damage post-impact. To this end, various non-destructive evaluation techniques such as ultrasonic C-scan testing and flash thermography were utilized for post-impact analysis. MATLAB algorithms were written and refined for the localization and quantification of damage in plates using data from sensors such as piezoelectric and fiber Bragg gratings sensors. Throughout the thesis, the general plate theory and laminate plate theory, the operations and optimization of the gas gun, and the theory used for the damage localization algorithms will be discussed. Additional quantifiable results are to come in future semesters of experimentation, but this thesis outlines the framework upon which all the research will continue to advance.

Advanced aerospace materials, including fiber reinforced polymer and ceramic matrix composites, are increasingly being used in critical and demanding applications, challenging the current damage prediction, detection, and quantification methodologies. Multiscale computational models offer key advantages over traditional analysis techniques and can provide the necessary capabilities for the development of a comprehensive virtual structural health monitoring (SHM) framework. Virtual SHM has the potential to drastically improve the design and analysis of aerospace components through coupling the complementary capabilities of models able to predict the initiation and propagation of damage under a wide range of loading and environmental scenarios, simulate interrogation methods for damage detection and quantification, and assess the health of a structure. A major component of the virtual SHM framework involves having micromechanics-based multiscale composite models that can provide the elastic, inelastic, and damage behavior of composite material systems under mechanical and thermal loading conditions and in the presence of microstructural complexity and variability. Quantification of the role geometric and architectural variability in the composite microstructure plays in the local and global composite behavior is essential to the development of appropriate scale-dependent unit cells and boundary conditions for the multiscale model. Once the composite behavior is predicted and variability effects assessed, wave-based SHM simulation models serve to provide knowledge on the probability of detection and characterization accuracy of damage present in the composite. The research presented in this dissertation provides the foundation for a comprehensive SHM framework for advanced aerospace materials. The developed models enhance the prediction of damage formation as a result of ceramic matrix composite processing, improve the understanding of the effects of architectural and geometric variability in polymer matrix composites, and provide an accurate and computational efficient modeling scheme for simulating guided wave excitation, propagation, interaction with damage, and sensing in a range of materials. The methodologies presented in this research represent substantial progress toward the development of an accurate and generalized virtual SHM framework.