UCLA

The study of magnetic phenomena at the nanoscale requires modeling frameworks that bridge the gap between computationally expensive quantum mechanical treatments and spatially averaged continuum micromagnetics. This dissertation presents the development of a fully coupled, atomistic magnetoelastic energy method implemented within a hybrid Molecular Dynamics–Atomic Spin Dynamics (MD–ASD) framework. By treating individual atoms as classical particles and magnetic moments as classical spin vectors subject to quantum-informed energetics, the model enables the direct coupling of atomic displacements and spin evolution. The framework is implemented as a custom pair potential in the LAMMPS software package, incorporating exchange, magnetocrystalline anisotropy, dipole–dipole interactions, and a novel atomistic formulation of cubic magnetoelastic energy. The model is validated against established micromagnetic solvers for BCC iron, demonstrating its capability to capture both linear and nonlinear magnetoelastic responses. Specifically, simulations of displacement-induced magnetic reorientation reveal significant deviations from linear elastic behavior. The results show that the nonlinear spin-rotation pathway drives a distinct mechanical response in transverse displacements and system stiffness, with the latter exhibiting a local minimum and maximums as the spins settle into new energy wells. These findings underscore the importance of atomistic resolution in describing strain-mediated magnetization dynamics in nanoscale multiferroic systems and provide a predictive tool for the design of next-generation spintronic and memory devices.