Research and Challenges of Multi-Scale Mechanics and Modeling
K P. Chong, Nicos Martys
One of the great challenges in the material sciences is to predict long term performance of materials and structures. Recent efforts have focused on linking nano/micro scalebehavior to meso-scale and macroscale behavior through computational modeling.In addition, accelerated tests are needed to simulate various environmental factors. In many cases, parallel computing is needed to account for the large number of variables and to relate short term test results to long term life-cycle behavior.It is impossible to do a detailed nano-molecular dynamics simulation to characterize macroscopic behavior. Large-scale atomistic simulations can account only for 500,000 atoms for a few nanoseconds. On the other hand it is possible to characterize certain macroscopic properties of materials such as the viscosity of fluids with smaller simulations. Clearly, the approach taken depends on understanding the underlying physics controlling the macroscopic behavior, e.g. for crack propagation one carries out detailed atomistic simulations near the tip of the crack and use such information to determine boundary conditions for continuum equation. Information from the continuum simulation can then be fed back into the microscopic equations. Other challenges and research concerning multi-scale mechanics are presented.
Micromechanics Europe 2002 Workshop, proceedings 13th