|
The properties and physics of complex flow, such as liquid crystal and polymer solutions, are studied from different time scales and length scales. Navier-Stokes-like Equations describe macroscopic fluid mechanics. Smoluchowski equation is a phenomenological equation describing Brownian motion et. al. Although great success has been made due to these equations, phenomenological input such as boundary conditions and constitutive laws are very complex or remain unclear. Singularity or large gradients in velocity and density is hard to handle. In worst case the continuum description is no longer valid. Fortunately, molecular simulation methods have the great strengths to avoid these problems and prove to be very useful tools for investigating material properties and its physics. We concentrate on classical molecular dynamics (MD) method and problems in liquid crystal and polymer. Molecular dynamics is a standard computational technique used in condensed matter physics, materials science, chemistry, and other fields. The idea is a simple one: calculate the forces acting on the particles in a molecular system and analyze their motion. When enough information on the motion of the individual atoms has been gathered, it is possible to condense it all using the methods of statistical mechanic to deduce the bulk properties of the material. These properties include the structure (e.g. crystal structure, predicted x-ray and neutron diffraction patterns), thermodynamics (e.g. enthalpy, temperature, pressure) and transport properties (e.g. thermal conductivity, viscosity, diffusion). In addition molecular dynamics can be used to investigate the detailed atomistic mechanisms underlying these properties and compare them with theory. It is a valuable bridge between experiment and theory. MD provides a wealth of detailed information on the structure and dynamics of molecular. However, it suffers certain limitations: the force field and computational demanding. The possibility to observe certain properties is directly related to the quality of the force field and, whether or not it has been parameterized for the system simulated. But it must be kept simple for reasons of computational efficiency. Efficient simulation methods that make use of parallel computers should be developed so we are able to simulation extremely large system composed of many molecules and take advantage of today's supercomputers. If you are interested in our work or if you have any question, do NOT hesitate to email us. You can write to Wu Congmin (cmwu@math.pku.edu.cn) or Wu Yibiao (patriot@pku.edu.cn). |
|
