A FE/BE coupling for the 3D time-dependent eddy current problem. I: A priori error estimates (Q985736)

The paper gives a discontinuous Galerkin method in time for coupling conforming FE and BE methods together with an optimal a priori estimates for the problem: \[ \begin{aligned} \mu\partial_{t}\mathbf{H}+\text{curl} \mathbf{E}=0 \qquad &\text{in}\quad (0,T)\times\mathbb{R}^{3},\\ \text{curl} \mathbf{H}=\mathbf{J}+\sigma\mathbf{E}\qquad &\text{in}\quad (0,T)\times\mathbb{R}^{3},\\ \int_\Gamma(\epsilon \mathbf{E})\cdot\mathbf{n}=0 \qquad &\text{on}\quad \Gamma,\\ \mathbf{H}(0,x)=\mathbf{H}_{0}(x),\quad \mathbf{E}(0,x)=\mathbf{E}_{0}(x)\quad&\text{in}\quad \mathbb{R}^{3},\\ [\mathbf{E}\times \mathbf{n}]_{\Gamma}=[\mathbf{H}\times\mathbf{n}]_{\Gamma}=0\qquad &\text{on}\quad \Gamma,\\ \mathbf{H}(t,x)=\mathbf{E}(t,x)=\mathcal{O}(|x|^{-1})\qquad &|x|\rightarrow\infty. \end{aligned} \] Here \(\mathbf{J}(x,t)\in\mathbf{L}^{2}(\mathbb{R}^{3})\) is the current density. The authors use a quasi-uniform and shape regular triangulation with mesh size \(h\) of the domain \(\Omega\). Besides the standard FB/BE Galerkin method they perform a discontinuous Galerkin method in time. Besides a priori error estimates the paper contains also numerical experiments.