AuthorM. S. Shephard and E. S. Seol
TitleToward a Multi-Model Hierarchy to Support Multiscale Simulations
AbstractThere is a long history on the development of mathematical representations capable of providing behavioral predictions of physical parameters on the atomic, molecular, microscopic, and macroscopic scales. Over the past half century, simulation programs have been developed to support the computerized solution of these mathematical representations which, in some cases, are discretized with billions of degrees of freedom and solved on massively parallel computers with thousands of processors. Historically, scientists and engineers have applied these models (simulation programs) to solve problems on a single physical scale. However, in recent years it has become clear that to continue to make advances in the areas of nanotechnology and biotechnology and to develop new products and treatments based on those advances, scientists and engineers must be able to solve sets of coupled models active over multiple interacting scales. For example, the development of new materials will require the design of structure and function across a hierarchy of scales starting at the molecular scales to define nanoscale building blocks that are used to define mesoscale features that are combined into micron-scale weaves that used in the manufacturing of the complete part (Figure 1). Such capabilities are clearly central to the development of nanoelectronics devices and future drug delivery systems (Figure 2), as well as many of other future products. For example, consider new automotive skins made of nano-reinforced materials in which the material interfaces are strong at strain rates consistent with normal usage, leading to high stiffness so the little dents are avoided, while at high strain rates the interfaces demonstrate substantial local damage, thus providing high energy absorption under impact loading to keep the individuals in the passenger compartment safe.
PDF File Download