Development of Microstructure-Informed Computational Models for the Design of Cold-Sprayed Additively Manufactured Metal-Ceramic Composite Materials

Abstract

This thesis develops microstructure-based finite element (FE) models of cold-sprayed additively manufactured (CSAM) metal-ceramic composites – Al-Al2O3 in this research – to fill a gap in our understanding of the microscale failure progression in the material in correlation with its macroscale response across stress states and strain rates. Representative volume elements (RVEs) of the material are generated by Digimat based on the features of Al2O3 ceramic particles, including weight fraction, size distribution, and clustering in addition to porosity in the metal matrix that are achieved by the scanning electron microscopy (SEM) characterization. Stress state- and strain rate-dependent constitutive models are implemented via VUMAT subroutines in Abaqus/Explicit FE solver to capture the growth of failure mechanisms, namely debonding of interfaces, particle cracking, and matrix failure. The micromechanical models are validated by the experimental data both quantitatively (i.e., stress-strain curves) and qualitatively (i.e., the manifestation of failure mechanisms in fractography analysis). The validated micromechanical model is leveraged to quantify the growth history of failure mechanisms in association with the stress-strain response of the material under different stress states and strain rates, which is not readily within reach by the available experimental mechanics approaches. Such quantification framework is next employed to systematically track the interaction and propagation of failure mechanisms as a function of microstructural characteristics (i.e., particle weight fraction, particle size, and particle clustering), stress state (i.e., uniaxial compression, uniaxial tension, and pure shear), and strain rate (i.e., quasi-static and dynamic loading covering 10−3 to 10−2 s−1 and 170 to 3200 s−1, respectively). This thesis is the first of its kind that provides a foundational understanding of the relationships between the microstructure, growth history of microscale failure mechanisms, and stress-strain response – as the general descriptor of the macro scale behavior – of CSAM metal-ceramic composites through establishing experimentally validated micromechanical FE models. As a comprehensive dataset generator covering the space of design across microstructures, stress states, and strain rates, the current computational framework will lay the foundation for the development of computationally efficient machine learning-based surrogate models of the microstructure-property-performance relationships of the material in addition to accelerating concurrent multiscale simulations for the design and optimization of CSAM metal-ceramic composites as structural materials and coatings with an application-specific fine-tuned mechanical performance.