Malgorzata Peszynska - Department of Mathematics - Oregon State University - Research



Links:
Publications
CV
Research
Teaching


Research Links:

My research interests are in mathematical modeling and numerical analysis and simulation of phenomena of flow and transport in porous media. Such phenomena are typically described by nonlinear coupled PDEs whose data have highly heterogeneous character, frequently at multiple scales. A typical example is the coupled multiphase multicomponent flow and transport in fractured porous media. I am interested in analysis of solutions to such models as well as in various ways to characterize and describe the heterogeneity and uncertainty of the data. The upscaled models which arise, for example, by way of homogenization, have typically a nonlocal character in time (memory terms) and/or in space. These present significant challenges when analyzed and solved numerically using finite element or finite difference methods.

I have always been interested in computing. This includes both: rigorous numerical analysis and, in paticular, derivation of a-priori and a-posteriori error estimates, as well as high-performance computing issues including domain decomposition and parallel solution techniques. I became interested in Parallel Computing even before it became a reality (80's); back then temporal logic applied to concurrent processes stated that "Sometime" is sometimes "not never". (L. Lamport). (Temporal logic is one of the tools used in proving accuracy and non-blocking of parallel (concurrent) processes.) These days computing has reached "terascale" levels; I have been involved in various terascale reservoir simulation and characterization projects which involve data-intense parallel and grid computing, optimization and visualization techniques.

Continuing my early interest in domain decomposition (Schwartz) methods, I have worked on multiphysics and multinumerics couplings of models of multiphase multicomponent flow, transport, and coupled elastic deformation in porous media. Recently I have proposed adaptivity of such couplings based on a-posteriori error estimators for mortar mixed finite element methods.


My mathematical genealogy and the tree

Grants and projects:


Recent/upcoming events and conferences:

2006: