A fictitious domain approach to the numerical solution of PDEs in stochastic domains (Q2454706)

Numerical methods are studied for the numerical treatment of boundary value problems of elliptic partial differential equations (PDEs) \[ - \Delta u = f \] in bounded random domains \(D=D(\omega) \in {\mathbb R}^2\) with homogeneous boundary conditions \(u| _{\partial D} = 0\) on sufficiently smooth boundary \(\partial D\), where \(f \in L^2\). The authors are interested in an appropriate concept of weak solutions for these problems exploiting saddle-point formulations and Lagrange multipliers. A combination of a fictitious domain approach and a polynomial chaos expansion is suggested. The PDE is solved in a larger, fixed domain (called the fictitious domain), with the original boundary condition enforced via a Lagrange multiplier acting on a random manifold inside the new domain. A (generalized) Wiener expansion is invoked to convert such a randomized problem into a deterministic one, depending on an extra set of real random variables. Discretization is accomplished by mixed finite elements in the physical variables and a Galerkin projection method which coincides with a randomized collocation scheme. Some stability and convergence analysis of the presented randomized numerical methods are provided. Eventually, some numerical experiments illustrate their randomized approach.