Asynchronous Fast Adaptive Composite-Grid Methods for Elliptic Problems: Theoretical Foundations

Accurate numerical modeling of complex physical, chemical, and biological systems requires numerical simulation capability over a large range of length scales, with the ability to capture rapidly varying phenomena localized in space and/or time. Adaptive mesh refinement (AMR) is a numerical process for dynamically introducing local fine resolution on computational grids during the solution process, in response to unresolved error in a computation. Fast adaptive compositegrid (FAC) methods are a class of algorithms that exploit the multilevel structure of AMR grids to solve elliptic problems efficiently. This paper develops a theoretical foundation for AFACx, an asynchronous FAC method. A new multilevel condition number estimate establishes that the convergence rate of the AFACx algorithm does not degrade as the number of refinement levels in the AMR hierarchy increases.