Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
499993 | Computer Methods in Applied Mechanics and Engineering | 2008 | 12 Pages |
Abstract
This paper presents multi-element stochastic reduced basis methods (ME-SRBMs) for solving linear stochastic partial differential equations. In ME-SRBMs, the domain of definition of the random inputs is decomposed into smaller subdomains or random elements. Stochastic reduced basis methods (SRBMs) are employed in each random element to evaluate the response statistics. These elemental statistics are assimilated to compute the overall statistics. The effectiveness of the method is demonstrated by solving the stochastic steady-state heat transfer equation on two geometries involving different types of boundary conditions. Numerical studies are conducted to investigate the h-convergence rates of global and local preconditioning strategies.
Keywords
Related Topics
Physical Sciences and Engineering
Computer Science
Computer Science Applications
Authors
P. Surya Mohan, Prasanth B. Nair, Andy J. Keane,