| Traditional empirical potentials used in molecular dynamics (MD) simulations replace an explicit treatment of the electronic structure with an appropriate interatomic potential energy expression. This enables MD simulations to model atomistic processes, such as dislocation dynamics and plastic deformation, which typically require size and time domains exceeding what is currently feasible with computationally-demanding first principles techniques. However, discarding the electronic degrees of freedom prevents MD simulations from properly resolving certain phenomena which are dominated by electronic interactions. One example is thermal transport in metals, which is often underestimated by orders of magnitude in MD simulations. A recently-developed multi-scale simulation approach, allowing ad-hoc feedback from continuum heat flow solutions to thermostat atoms in an MD simulation, is used to model Joule-heating in nano-scale metallic contacts under electromagnetic stress. The simulations are carried out under conditions representative of contact surfaces in Radio Frequency Electromechanical Switches (RF MEMS) and rail/armature components of Electromagnetic Launchers (EMLs) and are used to speculate on the mechanisms for experimentally-observed material transfer.;Another phenomenon that is typically neglected in MD simulations is charge transfer between atoms of dissimilar electronegativity. A common approach to incorporating a dynamic treatment of charge in a classical potential simulation is to solve atomic charges using an equalization of electronegativity in the charge equilibration (QEq) method. The current work studies the effectiveness of the QEq to mimic the charge distribution properties of f-center defects in a sodium chloride crystal. The results indicate that the QEq is able to replicate some of the electrostatic energy features of an f-center, which include an extremely localized potential well in the vicinity of the defect. |
