Title:
From deterministic to stochastic, from mesoscale to macroscopic: 
multiscale modeling of grain growth

Abstract: 
Many problems in science and engineering call for multiscale modeling 
and simulations. In some applications the main issues are related to 
efficient computing, in others -- the challenge is gaining further 
insight into the problem. Multiscale methodology serves for both, which 
will be illustrated on the  example of modeling of grain growth in 
materials science. The hierachy of models will be discussed: starting 
with a very detailed deterministic mesoscale model of curvuture driven 
growth. The coarsest model is a macroscopic stochastic model that 
describes evolution of distribution functions.