Ugo d'Elia, Giuseppe Pelosi, Stefano Selleri and Ruggero Taddei
A design procedure for multi-layer absorbers based on carbon nanotubes (CNT) frequency selective surfaces (FSS) sheets is here developed. The paper aims to discuss there issues…
Abstract
Purpose
A design procedure for multi-layer absorbers based on carbon nanotubes (CNT) frequency selective surfaces (FSS) sheets is here developed. The paper aims to discuss there issues.
Design/methodology/approach
Single layer FSS are first analyzed via finite element (FE). Then equivalent sheets admittances are extracted in a transmission line model. Neural networks (NNs) interpolation over this data and subsequent multi-objective genetic algorithm (GA) based optimizations are then performed to design multiple layers absorbing structures. Designs are finally validated via full wave FEM simulations.
Findings
In this paper, some absorbing structures made of three or four FSS sheets with total thicknesses around 6 mm are synthesized.
Research limitations/implications
NNs' accuracy used in the equivalent model can be refined with further training.
Practical implications
Low profile absorbing materials are of relevant industrial interest both for radar cloaking and anechoic chambers.
Originality/value
The transmission line model combined with NNs and GA optimization is capable of speeding up the design procedure with respect to a conventional full-wave FEM approach.
Details
Keywords
Giacomo Guarnieri, Giuseppe Pelosi, Lorenzo Rossi and Stefano Selleri
The paper's aim is to devise a fast method for microwave waveguide filter permittivity tolerance analysis.
Abstract
Purpose
The paper's aim is to devise a fast method for microwave waveguide filter permittivity tolerance analysis.
Design/methodology/approach
A 2D finite elements (FEs) formulation is combined via a Schur complement‐based domain decomposition (DD) technique to reduce the tolerance affected part of the analysis to a smaller domain.
Findings
The paper shows how to combine FEs and DD in an efficient way for material parameters tolerance analyses in microwave waveguide filters, showing speedup results.
Research limitations/implications
The formulation here presented is 2D but can be easily extended to 3D.
Originality/value
The application of DD to solve numerically large problem is well‐known, the idea and organization of the algorithm to allow iteration on parameter values on a single sub‐domain is here proposed.
Details
Keywords
The aim of this paper is to show the effectiveness of the finite element method (FEM) to study the properties of different kinds of photonic crystal fibers (PCFs), presenting…
Abstract
Purpose
The aim of this paper is to show the effectiveness of the finite element method (FEM) to study the properties of different kinds of photonic crystal fibers (PCFs), presenting results which highlight the FEM flexibility, exploited according to the particular PCF feature under investigation.
Design/methodology/approach
The FEM has been applied to a new emerging class of optical fibers, the so‐called PCFs, also known as microstructured or holey fibers.
Findings
It has been shown how to design and customize the PCF cross‐section to achieve desired values of dispersion, confinement loss, nonlinear or amplification properties. Reported examples prove the FEM ability to deal with complex geometries, arbitrary refractive index steps and distribution, and to be integrated with other approaches for a better and accurate analysis of the considered fiber.
Research limitations/implications
Limitation in the FEM use can be given by the required computation effort in terms of memory occupancy and time, even if computational power of modern workstations can attenuate this aspect.
Practical implications
The FEM can be a very powerful tool to investigate and design actual structures to be used in several fields, as telecom, sensing, fiber lasers, spectroscopy.
Originality/value
The novelty of the paper is given by the exploitation of the FEM feature to design a new emerging class of optical fibers, considering all numerical aspects given by the unusual characteristics of the domain and problem under investigation.