Quantum-mechanical-calculations-based multiscale modeling

In the community of modeling and simulation, computational mechanics and materials science has become an indispensable ally of experimental research for the advancement of scientific knowledge and engineering practice. A comprehensive, efficient and accurate description for materials properties and behaviors in cutting-edge technologies and applications demands an understanding over multiple length and time scales. From the quantum and nanoscale to the mesoscale and macroscale, activity on one level influences behavior on the others. It requires that descriptions at all levels be consistent with each other, which can be particularly demanding for advanced and complex materials. In this sense, the core of this work is bridging the gap between electronic/atomic level quantum mechanical calculations and multiscale simulations.;This research starts from quantum mechanical calculations and arrives at a newly developed concurrent atomic/continuum multiscale field theory. First principles calculations provide the database for interatomic potential developments. Further, the interatomic potential is utilized in all scale-levels to determine the material properties and involved in governing equations of motion ranging from microscopic to macroscopic. This strategy enables large scale atomistic computations and concurrent atomic/continuum simulations for real functional materials.;It is well known that first principles calculations are the most accurate theoretical method to describe essential properties of functional materials, such as ferroelectrics. In this work, quantum mechanical calculations (QMC) were performed within density functional theory. Through QMC, all of the fundamental information about lattice energies, lattice structures, and phonon dispersion relations are obtained. Sequentially, QMC-based interatomic potential developments are accomplished through multi-objective optimizations with genetic algorithms. Energies and phonon dispersion relations are considered as the objectives in the optimization.;Employing the developed interatomic potential from QMC, this work is scaled up to atomistic modeling and finite element analysis. The smart material in piezoelectric applications, perovskite oxide BiScO3 (BSO), and the prototype material for a large group of ionic oxides MgO are studied. The atomistic modeling of BSO, including molecular dynamics (MD) simulations, structural optimizations and elastic property calculations are performed. As the numerical implementation of the multiscale field theory, the response of a finite nano-size MgO continuum to an electromagnetic wave is simulated by a concurrent atomic/continuum model.