Duct noise control by spectral matching

Abstract

Noise pollution significantly impacts life quality, necessitating noise control measures to mitigate its effects. This study aims to control noise by spectral matching and to design noise control devices for noise sources with different spectral characteristics. Firstly, for broadband noise in a flow duct, the silencer intruding the flow passage are effective, and the larger the silencer, the better the noise reduction. However, larger silencers cause higher airflow pressure loss, requiring a trade-off between noise reduction and pressure loss. To address this, we apply a validated empirical pressure loss formula and an acoustic eigenvalue solution discretized via the Chebyshev collocation method, enabling consideration of acoustic performance within specified pressure loss constraints. Furthermore, by strategically placing perforated panels, the attenuation curve is matched with the noise spectrum. For a typical muffler with a 2-meter length and an incident spectrum range of 80 Hz to 8 kHz, the optimization method can reduce noise by 4-5 dB under equal pressure loss conditions, consistent with experimental results from a full-scale sample. Rotational machinery in ducts generates additional tonal noise superimposed on the original broadband noise, which can be effectively mitigated by employing multiple resonators to generate multi-peak absorption and transmission spectra. However, this will take up ample space and is not tunable. Instead, we applied a resonator utilizing the electromechanical coupling effect, the Shunted Electromechanical Diaphragm (SEMD). Using a single resonator, the frequencies and number of peaks can be arbitrarily reconfigured by adjusting circuit parameters connected with SEMD. Impedance tube experiments demonstrate that fine-tuning the circuit parameters makes it possible to successfully construct and reconfigure comb-like spectra with three absorption coefficients exceeding 0.9. Furthermore, we found that mechanical resonator arrays can also achieve comb-like acoustic filters, forming a peak-dip structure for broadband sound absorption. While the mechanism behind peak formation is well elucidated, the genesis of dips remains shrouded. We aimed to explain the cause of the dips observed in the absorption spectrum. The sharp drops are due to the velocity dipole response of coupled resonators, resulting in the generation of evanescent waves and the effect of excessive acoustic resistance. This issue can be mitigated by introducing additional local acoustic resistance. The mechanisms behind the sharp declines in broadband absorption spectra and the improvement measures were further elucidated through analytical, numerical, and experimental studies, which are concordant. Different from the above approaches of spatially arranging resonators for specific passive noise control, temporally serial resonances are explored to reconfigure broadband sound absorption spectra, significantly saving space. A Time-Varying Shunted Electromechanical diaphragm (TV-SEMD) is introduced to achieve adaptive and flexible absorption spectra, using digitally controlled Metal–oxide–semiconductor Field-effect Transistor (MOSFET) switches to activate different circuit branches. Numerical simulation and experimental validation confirm that the TV-SEMD can adapt its sound absorption characteristics to match varying noise spectra, providing a flexible and practical approach to noise control. Overall, this study derives a spectrum-matching-based noise control for different characteristic spectra. Theoretical methods, numerical simulations, and experimental results all show high consistency.