A simple tool for the optimization of 1D phononic and photonic bandgap filters

Prasanna Salasiya, Bojan B. Guzina · 2024

Read A simple tool for the optimization of 1D phononic and photonic bandgap filters by Prasanna Salasiya, Bojan B. Guzina (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-02037.txt.

Introduction

Prasanna Salasiya and Bojan Guzina build a design tool for one-dimensional wave filters in a 2024 paper. Such filters use layered periodic media to block or pass sound and light waves. Testing each design with full simulation is slow. Their tool cuts the work by a wide margin.

The layer is stitched from segments cut out of known periodic media. In each segment, waves take the form of Bloch waves drawn from the mother medium. A transfer matrix step then solves for reflected and transmitted waves. Designers can swap segment order, length, and offset at low cost, since the wave shapes stay fixed. The shortcut matched full simulations and sped design searches greatly.

Catalog record

CREATORPrasanna Salasiya, Bojan B. Guzina
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE48 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-02037

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A simple tool for the optimization of 1D phononic and photonic bandgap filters Prasanna Salasiya, Bojan B. Guzina Abstract We develop an effective computational tool for simulating the scattering of one-dimensional (1D) waves by a composite layer architected in an otherwise homogeneous medium. The layer is designed as the union of segments cut from various mother periodic media, which allows us to describe the wavefield in each segment in terms of the “left” and “right” (propagating or evanescent) Bloch waves. For a given periodic medium and frequency of oscillations, the latter are computed by solving the quadratic eigenvalue problem (QEP) which seeks the (real- or complex-valued) wavenumber – and affiliated eigenstate – of a Bloch wave. In this way the scattering problem is reduced to a low-dimensional algebraic problem, solved via the transfer matrix approach, that seeks the amplitudes of the featured Bloch waves (two per segment), amplitude of the reflected wave, and that of the transmitted wave. Such an approach inherently caters for an optimal filter (e.g. rainbow trap) design as it enables rapid exploration of the design space with respect to segment (i) permutations (with or without repetition), (ii) cut lengths, and (iii) cut offsets relative to the mother periodic media. Specifically, under (i)–(iii) the Bloch eigenstates remain invariant, so that only the transfer matrices need to be recomputed. The reduced order model is found to be in excellent agreement with numerical simulations. Example simulations demonstrate 40x computational speedup when optimizing a 1D filter for minimum transmission via a genetic algorithm (GA) approach that entails O ⁡ ( 10 6 ) O(10^{6}) trial configurations. Relative to the classical rainbow trap design where the unit cells of the mother periodic media are arranged in a “linear” fashion according to their dispersive characteristics, the GA-optimized (rearranged) configuration yields a 40 % 40\% reduction in filter transmissibility over the target frequency range, for the same filter thickness. 1 Introduction Wave motion in periodic media has garnered significant research interest since early in the last century due to its relevance to various macroscopic phenomena such as wave dispersion, negative index of refraction, and band gaps [ 4 , 18 , 34 ] . In short, band gap refers to a frequency range where the incident waves are exponentially attenuated (more specifically reflected backward) due to destructive interference of the scattered wavefield; periodic media that are capable of realizing such exotic behavior in the sub-wavelength regime are commonly referred to as metamaterials [ 24 , 26 ] . Many applications of metamaterials such as negative refraction, superlens, cloaking, energy harvesting, vibration isolation and seismic protection, rely on the location and width of the band gaps [ 15 , 3 , 11 ] . Notwithstanding the fact that the wave manipulation devices designed for such purposes are necessarily bounded (i.e. compactly supported) and possibly aperiodic, as in the case of rainbow traps [ 31 , 32 ] , their performance is often interpreted through the lens of the frequency-wavenumber dispersion relationship computed for an infinite periodic medium. This dichotomy inherently necessitates numerical or physical simulations of the full scattering problem, even for simple 1D systems, to achieve optimal design of metamaterial-based wave manipulation devices. As noted in [ 14 ] , having at least three to four unit cells is necessary to reflect – at least to some degree – the dispersive characteristics of an infinite periodic medium; however many metamaterial systems, such as rainbow traps, entail single unit cells extracted from various periodic media, rendering traditional methods of interpretation inadequate. The scarcity of analytical tools for bounded, aperiodic metamaterial systems poses significant challenges for the optimal design of wave-manipulation devices. In the context of 1D problems, the existing literature predominantly focuses on either bounded cutouts of a single periodic medium, or very specific filter designs. For instance, the authors in [ 9 , 6 ] deploy a transfer matrix method combined with the Bloch analysis to examine the reflectivity of a finite segment of a 1D periodic medium. Similarly, [ 2 ] focuses on the analysis of linearly-graded internal resonators in finite 1D chains, while [ 13 ] deals with metamaterial beams with a priori chosen nonlinear gradation of local resonators via the spectral element method catering for flexural waves. However, these techniques are ill-suited for batch simulations targeting large sets of trial system configurations, highlighting the need for a universal reduced-order model that can be used as a computational backbone for the optimal design of more complex 1D metamaterial systems. By building on the recent works in [ 29 , 27 ] , this paper aims to address the gap by developing a semi-analytical tool to effectively compute the effects of 1D scattering by an aperiodic layer in an otherwise homogeneous infinite medium (see Fig. 1 (c)), relevant to e.g. rainbow trapping or energy harvesting, using a combination of the Bloch wave analysis [ 10 , 5 , 7 ] and transfer matrix approach [ 22 , 30 , 8 ] . Specifically, our objective is to enable rapid exploration of the design space of a broad class of 1D (passive, metamaterial-based) wave manipulation devices. To this end, we assume that the architected layer is constructed as a 1D sequence of bonded strips of dissimilar mother periodic media, see Fig. 1 . In our approach the wave motion in each mother periodic medium is precomputed, in terms of the “left” and “right” (propagating or evanescent) Bloch waves, by solving a quadratic eigenvalue problem (QEP) [ 33 ] which seeks the wavenumber – and affiliated eigenfunction – of a Bloch wave for a given frequency of wave motion. While the linear eigenvalue problem (LEP), seeking the frequency of a Bloch wave for a given wavenumber, is typically used to compute the dispersion relationship [ 19 ] , the QEP offers a distinct advantage for it enables computation of the wave motion inside a band gap, that is essential for applications such as rainbow trapping and energy harvesting. In the context of the 1D wave equation, it was shown explicitly in [ 29 ] that the QEP yields exactly two (real- or complex-valued) wavenumbers for a given periodic medium. This allows us to express the solution of the wave equation with periodic coefficients, á la d’Alembert, via superposition of the left- and right-going Bloch waves for an arbitrary segment of a periodic medium – irrespective of its length and boundary conditions. With such result in place, 1D scattering by an architected layer is then computed via the transfer matrix approach used as a tool to “stitch” the segments of different mother periodic media. The key advantage of such an approach lies in the fact that the QEP (solved independently for each mother periodic medium) – the most computationally taxing step – can be precomputed and used to rapidly solve the scattering problem for numerous combinations of the segment lengths, offsets, and permutations (with or without repetition). For each such trial architecture, the only item that needs recomputing are the transfer matrices which are then used to solve for the amplitudes of the germane Bloch waves (two per segment), amplitude of the reflected wave, and that of the transmitted wave. Thanks to its efficiency, this

Read the complete text