The method of fundamental solutions for Brinkman flows. Part II. Interior domains

• Published in: Journal of engineering mathematics Vol. 127; no. 1
• Main Authors: Karageorghis, Andreas, Lesnic, Daniel, Marin, Liviu
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.04.2021; Springer Nature B.V
• Subjects: Applications of Mathematics; Barriers; Computational fluid dynamics; Computational Mathematics and Numerical Analysis; Domains; Fluid flow; Incompressible flow; Inverse problems; Mathematical and Computational Engineering; Mathematical Modeling and Industrial Mathematics; Mathematics; Mathematics and Statistics; Optimization; Porous media; Porous media flow; Regularization; Theoretical and Applied Mechanics; Thermal expansion; Viscous flow; Primary 65N35; Nonlinear minimization; Brinkman flow; Method of fundamental solutions; 65N38; Secondary 65N21; Inverse problem
• ISSN: 0022-0833; 1573-2703
• DOI: 10.1007/s10665-020-10083-2

In part I, we considered the application of the method of fundamental solutions (MFS) for solving numerically the Brinkman fluid flow in the unbounded porous medium outside obstacles of known or unknown shapes. In this companion paper we consider the corresponding interior problem for the Brinkman flow in a bounded porous medium which contains an unknown rigid inclusion D ⊂ Ω . The inclusion D is to be identified by a pair of Cauchy data represented by the fluid velocity and traction on the boundary ∂ Ω . The fluid velocity and pressure of the incompressible viscous flow in the porous medium Ω \ D ¯ are approximated by linear combinations of fundamentals solutions for the Brinkman system with sources ( singularities ) placed outside the closure of the solution domain, i.e. in D ∪ ( R 2 \ Ω ¯ ) , assuming, for simplicity, that we analyse planar domains. By further assuming that the unknown obstacle D is star-shaped (with respect to the origin), the inverse problem recasts as the minimization of the nonlinear Tikhonov’s regularization functional with respect to the MFS expansion coefficients and the discretized polar radii defining D . This minimization subject to simple bounds on the variables is solved numerically using the MATLAB optimization toolbox routine lsqnonlin.