Source author record

I. G. Graham

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

math.NA Numerical Analysis

Catalog footprint

Research overview

3works

2topics

4close collaborators

Actions

Is this your work?

Open graph Browse papers

preprint2020arXiv

Domain Decomposition with local impedance conditions for the Helmholtz equation with absorption

We consider one-level additive Schwarz preconditioners for a family of Helmholtz problems with absorption and increasing wavenumber $k$ . These problems are discretized using the Galerkin method with nodal conforming finite elements of any (fixed) order on meshes with diameter $h = h(k)$ , chosen to maintain accuracy as $k$ increases. The action of the preconditioner requires solution of independent (parallel) subproblems (with impedance boundary conditions) on overlapping subdomains of diameter $H$ and overlap $δ\leq H$ . The solutions of these subproblems are linked together using prolongation/restriction operators defined using a partition of unity. In numerical experiments (with $δ\sim H$ ) for a model interior impedance problem, we observe robust (i.e. $k-$ independent) GMRES convergence as $k$ increases. This provides a highly-parallel, $k-$ robust one-level domain decomposition method. We provide supporting theory by studying the preconditioner applied to a range of absorptive problems, $k^2\mapsto k^2+ \mathrm{i} \varepsilon$ , with absorption parameter $\varepsilon$ . Working in the Helmholtz ``energy'' inner product, and using the underlying theory of Helmholtz boundary-value problems, we prove a $k-$ independent upper bound on the norm of the preconditioned matrix, valid for all $\vert \varepsilon\vert \lesssim k^2$ . We also prove a strictly-positive lower bound on the distance of the field of values of the preconditioned matrix from the origin which holds when $\varepsilon/k$ is constant or growing arbitrarily slowly with $k$ . These results imply robustness of the preconditioner for the corresponding absorptive problem as k increases and give theoretical support for the observed robustness of the preconditioner for the pure Helmholtz problem.

preprint2016arXiv

Recent Results on Domain Decomposition Preconditioning for the High-frequency Helmholtz Equation using Absorption

In this paper we present an overview of recent progress on the development and analysis of domain decomposition preconditioners for discretised Helmholtz problems, where the preconditioner is constructed from the corresponding problem with added absorption. Our preconditioners incorporate local subproblems that can have various boundary conditions, and include the possibility of a global coarse mesh. While the rigorous analysis describes preconditioners for the Helmholtz problem with added absorption, this theory also informs the development of efficient multilevel solvers for the "pure" Helmholtz problem without absorption. For this case, 2D experiments for problems containing up to about $50$ wavelengths are presented. The experiments show iteration counts of order about $\mathcal{O}(n^{0.2})$ and times (on a serial machine) of order about $\mathcal{O}(n^α)$ , { with $α\in [1.3,1.4]$ } for solving systems of dimension $n$ . This holds both in the pollution-free case corresponding to meshes with grid size $\mathcal{O}(k^{-3/2})$ (as the wavenumber $k$ increases), and also for discretisations with a fixed number of grid points per wavelength, commonly used in applications. Parallelisation of the algorithms is also briefly discussed.

preprint2012arXiv

Filon-Clenshaw-Curtis rules for highly-oscillatory integrals with algebraic singularities and stationary points

In this paper we propose and analyse composite Filon-Clenshaw-Curtis quadrature rules for integrals of the form $I_{k}^{[a,b]}(f,g) := \int_a^b f(x) \exp(\mathrm{i}kg(x)) \rd x $, where $k \geq 0$ , $f$ may have integrable singularities and $g$ may have stationary points. Our composite rule is defined on a mesh with $M$ subintervals and requires $MN+1$ evaluations of $f$ . It satisfies an error estimate of the form $C_N k^{-r} M^{-N-1 + r}$ , where $r$ is determined by the strength of any singularity in $f$ and the order of any stationary points in $g$ and $C_N$ is a constant which is independent of $k$ and $M$ , but depends on $N$ . The regularity requirements on $f$ and $g$ are explicit in the error estimates. For fixed $k$ , the rate of convergence of the rule as $M \rightarrow \infty$ is the same as would be obtained if $f$ was smooth. Moreover, the quadrature error decays at least as fast as $k \rightarrow \infty$ as does the original integral $I_{k}^{[a,b]}(f,g)$ . For the case of nonlinear oscillators $g$ , the algorithm requires the evaluation of $g^{-1}$ at non-stationary points. Numerical results demonstrate the sharpness of the theory. An application to the implementation of boundary integral methods for the high-frequency Helmholtz equation is given.

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.