Efficient wideband numerical simulations for nanostructures employing a Drude-Critical Points (DCP) dispersive model

Qiang Ren, Jogender Nagar, Lei Kang, Yusheng Bian, Ping Werner, Douglas H. Werner

Research output: Contribution to journalArticlepeer-review

3 Scopus citations


A highly efficient numerical approach for simulating the wideband optical response of nano-architectures comprised of Drude-Critical Points (DCP) media (e.g., gold and silver) is proposed and validated through comparing with commercial computational software. The kernel of this algorithm is the subdomain level discontinuous Galerkin time domain (DGTD) method, which can be viewed as a hybrid of the spectral-element time-domain method (SETD) and the finite-element time-domain (FETD) method. An hp-refinement technique is applied to decrease the Degrees-of-Freedom (DoFs) and computational requirements. The collocated E-J scheme facilitates solving the auxiliary equations by converting the inversions of matrices to simpler vector manipulations. A new hybrid time stepping approach, which couples the Runge-Kutta and Newmark methods, is proposed to solve the temporal auxiliary differential equations (ADEs) with a high degree of efficiency. The advantages of this new approach, in terms of computational resource overhead and accuracy, are validated through comparison with well-known commercial software for three diverse cases, which cover both near-field and far-field properties with plane wave and lumped port sources. The presented work provides the missing link between DCP dispersive models and FETD and/or SETD based algorithms. It is a competitive candidate for numerically studying the wideband plasmonic properties of DCP media.

Original languageEnglish (US)
Article number2126
JournalScientific reports
Issue number1
StatePublished - Dec 1 2017

All Science Journal Classification (ASJC) codes

  • General


Dive into the research topics of 'Efficient wideband numerical simulations for nanostructures employing a Drude-Critical Points (DCP) dispersive model'. Together they form a unique fingerprint.

Cite this