The Foldy-Lax Approximation for the Full Electromagnetic Scattering by Small Conductive Bodies of Arbitrary Shapes

TL;DR

This paper derives a validated Foldy-Lax approximation for electromagnetic scattering by small conductive bodies of arbitrary shapes, providing explicit error estimates and conditions for validity in the harmonic regime.

Contribution

It introduces a general condition for the approximation's validity, explicit error bounds, and a mathematical framework for analyzing electromagnetic scattering by small inhomogeneities.

Findings

Valid approximation under specific size and distance conditions

Explicit error estimates in terms of parameters and frequency

Mathematical proof of system invertibility and density estimates

Abstract

We deal with the electromagnetic waves propagation in the harmonic regime. We derive the Foldy-Lax approximation of the scattered fields generated by a cluster of small conductive inhomogeneities arbitrarily distributed in a bounded domain $\Omega$ of $\mathbb{R}^3$. This approximation is valid under a sufficient but general condition on the number of such inhomogeneities $m$, their maximum radii $\epsilon$ and the minimum distances between them $\delta$, the form $$(\ln m)^{\frac{1}{3}}\frac{\epsilon}{\delta} \leq C,$$ where $C$ is a constant depending only on the Lipschitz characters of the scaled inhomogeneities. In addition, we provide explicit error estimates of this approximation in terms of aforementioned parameters, $m, \epsilon, \delta$ but also the used frequencies $k$ under the Rayleigh regime. Both the far-fields and the near-fields (stated at a distance $\delta$ to the cluster) are estimated. In particular, for a moderate number of small inhomogeneities $m$, the derived expansions are valid in the mesoscale regime where $\delta \sim \epsilon$. At the mathematical analysis level and based on integral equation methods, we prove a priori estimates of the densities in the $L^{2,Div}_t$ spaces instead of the usual $L^2$ spaces (which are not enough). A key point in such a proof is a derivation of a particular Helmholtz type decomposition of the densities. Those estimates allow to obtain the needed qualitative as well as quantitative estimates while refining the approximation. Finally, to prove the invertibility of the Foldy-Lax linear algebraic system, we reduce the coercivity inequality to the one related to the scalar Helmholtz model. As this linear algebraic system comes from the boundary conditions, such a reduction is not straightforward.