This Is AuburnElectronic Theses and Dissertations

Design and Analysis of Acoustic Metamaterials in Order to Isolate Microelectromechanical Systems Devices from High Frequency Acoustic Noise




Griffin, Brittany

Type of Degree

PhD Dissertation


Mechanical Engineering


The effects of high frequency acoustic noise can be rather damaging to Microelectromechanical Systems (MEMS) devices. The use of MEMS devices is widespread; therefore, some operational environments for MEMS devices can be harsh with regards to high frequency acoustic noise. This research has developed a compact, configurable, omnidirectional, and passive packaging that isolates MEMS devices from damaging acoustic environments by utilizing acoustic metamaterials (AMM). An AMM is a material that affects sound waves and has properties not found in nature. Commonly, the properties are a function of the material’s geometric structure instead of material composition. In order to ensure acoustic mitigation across a broad spectrum of sensors and applications, sensor packaging was the focus of this research. The packaging developed during this research acoustically isolates the sensor using a combination of perforated nozzles in conjunction with interconnected cavities. The design is modular therefore the packaging can be tailored to fit any size required by the application. The design and experimental testing process was an iterative approach that was broken into three phases. The first phase focused on a series of nozzles with various features and of various sizes in a single input single output (SISO) configuration. Multiple designs were manufactured and experimentally tested. The testing concluded that a four-stage system of nozzles with radial and axial perforations and a constricted opening at the output plane of the system was the optimal configuration. This configuration is referred to as the Base Feature (BF) AMM. The nozzles in the early SISO configurations were two inches in length, but testing proved the system was operational when the nozzles were one-quarter inches in length meaning the overall thickness of the BF AMM configuration is one inch. From the control configuration to the BF AMM configuration, the average transmission loss (TLavg) across the acoustic spectrum went from 4.8 dB to 33.7 dB. Additionally, the BF performed well in the 2 kHz to 20 kHz range, with the TLavg over 35 dB. These experimental results for the BF AMM formed the basis for the second phase of testing. The next phase of testing focused on multiple input multiple output (MIMO) testing. A series of interconnected cavities surrounding the BF AMM were added to the design. The series of nozzles and accompanying interconnected cavities, called a cell, were repeated in parallel to form a panel. The panels consisted of 37 cells and were 7.5 inches in diameter. Multiple configurations of interconnected cavities of varying sizes were tested. The optimal configuration consisted of 12 interconnected cavities surrounding the BF AMM. This configuration is referred to as the Small Cavity AMM configuration. From the control panel to the Small Cavity AMM panel, the TLavg increased from 16.4 dB to 33.8 dB across the acoustic spectrum. The Small Cavity AMM performed exceptionally in the 2 kHz to 20 kHz range, with the TLavg equaling approximately 40 dB. These results formed the basis for the final phase of testing. The final stage of testing focused on an AMM Sphere. The Small Cavity AMM cell was modified to have curvature for use in the AMM Sphere. The AMM Sphere testing was performed at various distances as well as between multiple sound sources. The AMM Sphere test results were comparable in all the experimental test configurations with a TLavg at approximately 28 dB. The AMM Sphere performed exceptionally well in the 2 kHz to 20 kHz. The system performance when subjected to multiple sound sources was not affected. Acoustic simulations were performed with MSC ACTRAN. The theory and methodology of the acoustic simulations was thoroughly considered to ensure that any assumptions and the overall model development were accurate. Comparison of the acoustic simulations with the experimental results showed correlation and allowed for validation of the acoustic models. The experimental results were encapsulated within 1 standard deviation of the acoustic simulation mean results. The acoustic models demonstrated the capability to predict performance in operational environments. Several supplemental features are to be noted. The system is air permeable, therefore, adverse conditions such as stored heat, or the blockage of airflow are not introduced to the system. All of the configurations were tested in the Converging Nozzle and Diverging Nozzle configurations with similar results. This is indicative of the structure having the ability to keep sound in as well as out. This research has contributed to the field of AMM in several ways. The system’s capability to mitigate noise in the upper end of the acoustic spectrum is a significant contribution. Currently, AMM literature above 10 kHz is very rare. Within the current literature, AMM research above 10 kHz tends to be configured for a specific system and operates in a very narrow bandwidth or is theoretical in nature. The system’s ability to successfully operate from 2 kHz to 20 kHz is innovative. Additionally, the system operates omnidirectionally and with multiple sound sources. Current literature focuses on systems that operate in laboratory settings where the location of the single sound source is predetermined. Finally, the system can be configured to work with any MEMS device as well as within any application size.