ABSTRACT: For decades, grid adaptive methods have been recognized as an enabling technology for computational fluid dynamics (CFD).  However, with only a few exceptions, most CFD simulations today are performed on a priori grids (i.e. grids based upon heuristic meshing requirements from past experience on similar problems) as opposed to solution-based adaptive grids.  In this talk, we will discuss key ingredients for solution-based adaptive methods and review the current state of these methods.  Then we will focus on one particular approach from our research utilizing output error estimation, finite element methods, and metric-based adaptive meshing.  Finally, we consider the potential of this approach for time-dependent flows and highlight the particular issues associated with chaotic behavior of turbulent flows.

BIO: David Darmofal is the Jerome C. Hunsaker Professor of Aeronautics and Astronautics at MIT and a member of the MIT Aerospace Computational Design Laboratory (ACDL) and the MIT Center for Computational Science & Engineering (CCSE).  His principal areas of interest are computational methods for partial differential equations, especially fluid dynamics; and engineering education innovation.  He has written approximately 80 technical publications in peer-reviewed journals and conferences.  He teaches graduate and undergraduate courses in fluid dynamics and computational methods.