UCLA

Multi-scale modeling has been growing rapidly from the descriptive to predictive functions and employing modeling and simulations has become a crucial mission of science that helps to deliver an uninterrupted growth of the modern society. Here we integrated modeling works on different level of scale regarding four different material behaviors, with each illustrated in a chapter. They are ordered as atomic scale modeling, oxidation kinetics modeling for tungsten, reaction rates for 1D motion particles(verified with Kinetic Monte Carlo) and modeling work on the non-monotonic strain response of nanoporous multiferroic composites with Finite Element Method.We used Density Functional Theory(DFT) to obtain different kinds of intrinsic material parameters based on the calculations of the electronic configuration of a system, part of which is then used in the oxidation model for pure tungsten. In the oxidation numerical model, we predict oxide scale growth on tungsten surfaces under exposure to oxygen at high temperatures. The model captures the formation of four thermodynamically-compatible oxide sublayers, WO2, WO2.72, WO2.9, and WO3, on top of the metal substrate. Oxide layer growth is simulated by tracking the oxide/oxide and oxide/metal interfaces using a sharp-interface Stefan model coupled to diffusion kinetics. We simulate oxide growth at temperatures of 600◦C and above, where we find deviation from classical parabolic growth in several cases. A comparison of the model predictions with an extensive experimental data set, shows reasonable agreement at most temperatures. Our development of reaction rate for 1d motion particle was initiated due to the asymmetry in diffusion dimensionality between self-interstitial atom (SIA) clusters and vacancies is a fundamental feature of irradiation damage in crystals. While SIA clusters perform one-dimensional motion along mostly rectilinear trajectories, a complete set of kinetic coefficients, including coagulation reaction rates and sink strengths, does not exist for 1D-moving objects. We derive analytical expressions for these coefficients from continuum diffusion theory particularized to 1D motion and carry out kinetic Monte Carlo simulations of numerical replicas of the geometry of diffusing particles and sinks to validate the proposed solutions. Our simulations, which are conducted entirely independently from the analytical derivations, reveal excellent agreement with the proposed expressions, adding confidence to their validity. We compare the 1D and 3D cases and discuss their relevance for kinetic codes for damage accumulation calculations. In the work that uses Finite Element Method (FEM) to study the non-monotonic strain response of nanoporous multiferroic composites under electric field control, we simulate and analyze the mechanical response of a class of multiferroic materials consisting of a templated porous nanostructure made out of cobalt ferrite (CFO) partially filled by atomic layer deposition (ALD) with a ferroelectric phase of lead zirconate titanate (PZT). Our numerical results show that the non-monotonic mechanical response’s causes e.g. The increase of the strain due to a reduced system stiffness; And a larger mass fraction of PZT due to decreased porosity. A nonlinear piezoelectric response for PZT leads to an improved agreement with the experimental data, consistent with ex situ poling of the nanostructure prior to magnetic measurements.