Abstract
The aim of this work is to derive sharp quantitative estimates of the qualitative convergence results developed by Li et al. (2015) for regularized full- and partial-cloaks through transformation optics approach. Let Γ0 be a compact set in ℝ3 and Γδ be a δ-neighborhood of Γ0 for δ∈ℝ+. Γδ represents the virtual domain used for the blow-up construction. By incorporating suitably designed lossy layers, it is shown that if the generating set Γ0 is a generic curve, then one would have an approximate full-cloak within δ2 to the perfect full-cloak, whereas if Γ0 is the closure of an open subset on a flat surface, then one would have an approximate partial-cloak within δ to its perfect counterpart. The estimates derived are independent of the contents being cloaked; that is, the cloaking devices are capable of nearly cloaking an arbitrary content. Furthermore, as a significant by-product, our argument allows the relaxation of the convexity requirement on Γ0 by Li et al. (2015), which is critical for the Mosco convergence argument therein.
| Original language | English |
|---|---|
| Pages (from-to) | 821-851 |
| Number of pages | 31 |
| Journal | Communications in Partial Differential Equations |
| Volume | 42 |
| Issue number | 6 |
| Early online date | 8 May 2017 |
| DOIs | |
| Publication status | Published - 3 Jun 2017 |
User-Defined Keywords
- Asymptotic estimates
- invisibility cloaking
- partial and full cloaks
- regularization
- scattering
- transformation optics