Multi-physics Design Optimization of Multifunctional Composites

The design of multifunctional composites is an arduous task due to the presence of conflicting physical property demands from integrated functions. The realization of successful multifunctional architecture requires the development of new design methodologies that balance opposing needs and properties. The current design methodologies for multifunctional composites are immature since the majority of them only concentrate on the design of microstructures. In fact, designing only the microstructure of multifunctional composite is typically not sufficient for creating a successful multifunctional architecture. This dissertation targets the development of a multi-levels, multi-physics, multi-objectives design framework for multifunctional composites to advance the design methodologies of multifunctional materials. This design framework is equipped with some general features (applicable for the design of any multifunctional composite systems) and some problem-specific features (need to be modified based on the problem at hand). General features include multi-objectives schemes, uncertainty quantification algorithms, surrogate-based models, advanced numerical solvers, and large-scale high-performance computing algorithms. The problem-specific features contain efficient reduced-order multi-physics models (ROMs) and innovative optimization schemes. To demonstrate the framework, this thesis concentrates on a particular class of multifunctional composite systems, namely, the structural battery composite-microvascular composite (SBC-MVC) system as a case study. SBC-MVC system includes two various types of multifunctional composites: (i) structural battery composite (SBC), which can work as a Li-ion battery, and (ii) microvascular composites (MVC) which can provide thermal regulation for SBC. The key distinctions made in the developed design framework compared to the other related methods available in the literature for MVC and SBC are as follows: (i) looking at the design problem more comprehensively by considering it at multiple levels and (ii) giving more flexibility to the optimizer to provide better and more efficient design solutions by equipping the design framework with the aforementioned features. These tasks have been accomplished by (i) extending the design spaces in comparison with the available design approaches in the literature for MVC and SBC, (ii) facilitating the design process (especially from the computational point of view) with aid of advanced computational libraries and ROM-based models, and (iii) integrating uncertainty quantification algorithms allowing for having reliable designs. Structural battery composites and their extension to the SBC-MVC system are currently in their early stages of development and to date, there are very limited studies on their design. Thus, the results of this research can be insightful and may accelerate the development process of this new generation of energy-harvesting materials. The overall results of this research reveal that the developed computational design framework in this study advances the current design methodologies for SBC, MVC, and SBC-MVC systems.