You are here

Multiscale Methods and Data-driven Models

Xingjie Li, University of North Carolina-Charlotte
 
Yue Carol Yu, Lehigh University
 
Multiphysics and multiscale problems are challenging algorithmically and taxing computationally. Hence, high order and high-fidelity discretization methods combined with accurate and stable coupling strategies are needed, and analysis tools as well as acceleration (stabilization) techniques are of critical importance.  Recently, data-driven models and stochastic methods have been explored and developed to overcome these challenges. These models have also indicated many successes and promises in various applications.
 
In this minisymposium, we will review a number of new ideas for developing efficient and reliable coupling algorithms, data-driven methods and stochastic simulation tools in computational mechanics and related applications. The topics of interest include and are not limited to:  fluid–structure interaction, atomistic-to-continuum coupling problems, nonlocal-to-nonlocal/local coupling problems, quantum mechanics, micromechanics of materials, concurrent coupling approaches, multiphysics/multiscale coupling methods.