Frontiers of Engineering: Reports on Leading-Edge Engineering from the 2004 NAE Symposium on Frontiers of Engineering (2005)
Chapter: MULTISCALE MODELINGIntroduction--Grant S. Heffelfinger and Dimitrios Maroudas
Introduction
GRANT S. HEFFELFINGER
Sandia National Laboratories
Albuquerque, New Mexico
DIMITRIOS MAROUDAS
University of Massachusetts
Amherst, Massachussetts
Multiscale modeling and simulation are used in scientific and engineering research in biology and the health sciences, materials and processing sciences, and earth and atmospheric sciences. This emerging integrated, computational approach can lead to better understanding, analysis, and quantitative predictions of the behavior of realistic, complex systems. Multiscale modeling and simulation links atomic-scale phenomena with macroscopic responses over time scales relevant to humans (from minutes to centuries) by establishing rigorous links between widely different theoretical formalisms and computational methods that capture a very broad range of space and time scales—from electronic, molecular, and mesoscopic or microstructural scales to continuum or macroscopic scales.
The core capabilities of multiscale modeling and simulation include computational quantum mechanics, statistical mechanics, and continuum mechanics combined with applied and computational mathematics, such as numerical analysis, nonlinear analysis of dynamic systems, optimization, and control theory. Some of these methods, which have been developed over many decades, are quite mature. However, the rigorous coupling of these methods to produce predictive models of phenomena that span multiple length and time scales remains a significant challenge. Examples range from linking molecular phenomena, such as DNA transcription and gene expression, with cellular, tissue, and organ response to connecting atomic-scale and grain-scale dynamics with the macroscopic response of engineering materials, such as metals, semiconductors, and polymers.
The ultimate goal of multiscale modeling and simulation is to produce a global framework for system-level analyses of processes and phenomena relevant to human scales that are governed by phenomena occurring at much finer length and time scales.
Achieving this goal will require advances in coupling various time and length scales, as well as in mathematics, computer science, and computational science. This session focuses on the state of the art in several aspects of multiscale modeling and simulation, including the coupling of modeling and simulation methods across time and length scales for specific applications in the science and processing of engineered materials (e.g., semiconductors, metals, and polymers) and in biology and health science. Presentations address recent advances in the theoretical, mathematical, and computer science underpinnings of multiscale modeling, as well as computational science challenges, such as tera- and peta-scale computing, advanced visualization, and enabling technologies.
