1. Physics-Based Modeling of Failure of Novel Light-Weight Materials: Applications to Computational Design of Materials
- Author
-
Amirian, Benhour
- Subjects
- Cermet, Phase-field approach, Finite element method, Physics-based model, FEniCS, Monolithic scheme, Brittle materials
- Abstract
Abstract: Advanced physics-based computational models have been developed in this thesis to study dynamic failure in novel light-weight materials used in impact applications, focusing on applying different computational techniques to study three material systems: 1. novel self-propagating high-temperature synthesized (γ + α2)-TiAl/Ti3Al- Al2O3 ceramic-metal composites, 2. single crystal magnesium, and 3. nano-grained boron carbide. For the (γ + α2)-TiAl/Ti3Al- Al2O3 cermet, a progressive study has been carried out by developing a three-dimensional microstructure-based finite element model. The models have been developed to investigate the rate-dependent mechanical response features (e.g., compressive strength, flow stress hardening, and energy absorbing efficiency) and predominant failure mechanisms (e.g., void deformation and growth, particle cracking, and interface decohesion), and have been validated with high fidelity experimental data. Validating the numerical model has enabled predicting the material response outside of the experimentally-accessible conditions. To date, few studies have been made to bridge ceramic-metal material models under different loading conditions with experimental inputs, such as porosity, particle volume fraction, elastic modulus of the constitutions, and void clusters. To address this, a modified variational formulation of the Gurson model has been used to allow for the damage caused by the voids. The strain hardening components, as one of the most unknown material properties, have been calibrated by matching the modeling results with experimental data. Following validation, the effects and implications of different parameters such as particle volume fraction, porosity, unit cell size ratio, and the variability of the inclusions have been presented and discussed. After that, the model was extended to explore the response of the α2(Ti3Al)+γ(TiAl)-submicron grained alumina cermet by considering high particle volume fraction and under high strain-rate loading applications. In this part of the study, the energy absorbing efficiency, as one of the important factors in using the ceramic-metal composites in high-rate applications, has been studied. The results have indicated that the particle shapes and void volume fraction play an important role on energy absorption capabilities of the (γ + α2)-TiAl/Ti3Al-Al2O3 cermets' damage tolerant design. These new understandings have informed the fabrication and refinement of new generations of the cermet variant for commercial use in protection products by industry partners. In the second thrust of my research involving magnesium and boron carbide, I have developed an advanced physics-based time-dependent phase-field model to predict deformation and failure mechanisms (e.g., fracture and twinning) in these anisotropic materials. Computationally, a monolithic scheme has been used as a powerful technique to increase the accuracy of the solver and high-performance computer clusters have been employed to solve large-scale problems associated with the research. The coupled differential equations have been implemented in an open-source high-level Python interface, FEniCS. First, the model is extensively validated for an intrinsically brittle material, magnesium. Comparing with molecular dynamics simulation, the twin interface velocity is explored in order to obtain the kinetic coefficient. Finding this velocity-related parameter is important as one of the main factors for determining the driving force of twinning for magnesium. In addition, the spatial distribution of the shear stress field in the parent and twinned phases is investigated. The result provides insights into the effect of twin's thickness on further twin nucleation and growth. Next, the critical strain and initial twin embryo size required for propagation and growth of a single twin embryo in magnesium are predicted by the current phase-field approach. After that, the effect of twin-twin and twin-defect interactions is explored because this may increase the likelihood for crack and failure leading to reduction of material lifetime. Then, the phase-field approach is extended to predict various deformation mechanisms in nano-grained B4C (e.g., fracture and twinning). In addition, the crack propagation under compressive loading is treated by using a new decomposition for the strain energy density, which represents a valuable contribution to the literature. Overall, this thesis consists mainly of three parts adapted from the two published and the two submitted journal articles: 1- A physics-based model to capture the mechanical response of (γ+α2)-TiAl/Ti3Al- Al2O3 cermet under quasi-static and dynamic loading, 2- An advanced phase-field approach to evolution and interaction of twins to unravel time-evolved twinning behavior in magnesium at nanoscale, and 3- A comprehensive calibrated and validated phase-field model for studying the deformation mechanisms of nanocrystalline Mg and B4C in order to provide guidance for material refinement via tailoring their mechanical properties and microstructure.
- Published
- 2021