An Approach for Active Acoustic Metamaterial Design Implementation and Analysis

Abstract

Acoustic metamaterials are artificially structured materials exhibiting extraordinary acoustical properties unavailable in natural media. Currently, most of the acoustic metamaterials developed are passive. Although passive metamaterials are easy to fabricate with 3D printers, there are fundamental limitations associated with them. Firstly, the properties of all passive metamaterials cannot be changed after fabrication. Secondly, passive metamaterials typically employ resonant structures, which reduce the working bandwidth of the metamaterial. Moreover, the intense fields created by resonances usually mean that the energy dissipated in the material is considerably high. These limitations hinder the applicability of metamaterials because most applications require broadband acoustic signals. On the contrary, the active acoustic metamaterials have the potential to solve these problems by using embedded tunable electronics. However, the acoustic properties and functionality of most active metamaterial designs reported in the past are limited by the requirement that these metamaterials must be stable. To enforce stability, past designs have narrow bandwidth. This thesis demonstrates a novel design of active acoustic metamaterial based on Willis material concepts that eliminate the above limitations. Willis materials are complex media in which the unusual coupling between stress and particle velocity on one hand and linear momentum and strain, on the other hand, enable unique properties. In this work, we concentrate on creating stable yet broadband active acoustic metamaterials. We present the methodology and theory for designing Willis materials based on an active acoustic metamaterial architecture, in which Willis coupling parameters are allowed to vary independently. The proposed active acoustic metamaterial presented here is designed with three major improvements: broadband response; stable linear response; and modular design. After the design phase, we demonstrate the circuit schematic of the unit-cell for implementing such active Willis materials. Then we show the applications of the designed active metamaterial in the following order. Firstly, we show the acoustic non-reciprocity devices achieved from active Willis materials, and we also demonstrate the noise reduction ability of the designed non-reciprocity devices. Secondly, the application of designed metamaterial is extended to achieve acoustic anomalous reflectors. These applications are validated through experiments and numerical simulations. In addition, an experimental procedure to extract the effective material parameters from near-field measurements and to express material parameters in terms of acoustic polarizabilities are demonstrated to better understand the physics of the design.

The material parameter retrieval method proposed earlier is analytical and only works the best for 1D scenarios, because the results from 2D measurements could be contaminated by diffraction and become inaccurate. For this reason, we extend the research on extracting the effective material parameters of small samples by applying machine learning methods. Since the scattered fields measured can be essentially presented as images, the machine learning method employs convolutional neural networks, which are good at associating image patterns to salient macroscopic material properties that determined these patterns. We demonstrate that convolutional neural networks (CNNs) can interpret the diffraction patterns and learn the mapping between the scattered fields and all the effective material parameters including mass density and stiffness tensors from a small set of numerical simulations. We show that the CNN-based methods provide insight into the dynamic behavior of matter including quantitative measures of the scattered fields sensitivity to each material property and how the sensitivity changes from material to material.