Programmable time-serial resonances for broadband spectrum matching

Abstract

Spatially parallel and cascading acoustic liners are commonly used to treat duct noise. Their performance depends on the spectral match with source characteristics. Such liners rely on subtle geometrical and mechanical designs which are not difficult to adapt to different noise sources in a passive device. Here, we propose a temporal analogy, which is a shunted electromechanical diaphragm (SEMD) having different resonant frequencies in different time segments, namely, a time-serial resonator. Its sound absorption spectrum can be easily adapted by a program and yet the device functions passively. The acoustic impedance of the SEMD is determined by the shunt circuit, hence its resonant frequency and absorption peak. A multiple set of branch circuits are used to shunt the diaphragm. A MOSFET, which is a voltage-controlled ultrafast electronic switch, is introduced to cascade and switch each circuit branch. By activating circuit branches in a pre-defined time sequence, we obtain a series of resonance and the absorption which are effective for specific frequency bands at different time segments. When viewed over a long period, the averaged effective sound absorption spectrum is broadened and can be easily shaped by the working duty cycle of each circuit branch. Note that the switching voltage changes the circuit states without supplying power to the SEMD. Both numerical results and experimental demonstrations are presented. This study would open the new era of temporal design of acoustic liners.