Based on the acoustic-electric analogy,an acoustic structure can be designed by an electric filter.However,its filtering performances usually deviate from those of the electric filter.In order to improve the accuracy of predicting filtering performances of the acoustic structure,the strategy for modifying the sound inductance and sound capacitance of the acoustic structure is proposed.First,the transfer function model of the acoustic filter is established based on the acoustic-electric analogy.Second,the acoustic filter is designed by the first-order Butterworth band-stop electric filter with the given center frequency and stop-band width.The finite element method is used to simulate transmission losses of the acoustic filter,and relative errors,between the acoustic filter and the electric filter,of the center frequency and stop-band width are analyzed.And then,an iterative optimization algorithm is constructed by adopting least square method to fit finite element simulation data and to establish a new transfer function model of the acoustic filter based on the data fitting.Subsequently,the strategy for modifying the sound inductance and sound capacitance of the acoustic filter is proposed by analyzing differences between the two transfer function models.Finally,the first-and third-order Butterworth band-stop acoustic filters are designed to verify the modification strategy.The research shows that,for the acoustic structure designed based on the acoustic-electric analogy,its center frequency and stop-band width deviate from the given design requirements,and relative errors increase with the increase in the given stop-band width;for the modified first-order Butterworth band-stop acoustic filter,average relative errors of its center frequency and stop-band width are both less than 2%;for the modified third-order Butterworth band-stop acoustic filter,relative errors of its center frequency and stopband width reduce from 4.6%to 0.6%,and from 7.5%to 2.5%,separately.
Acoustic-electric analogyAcoustic structureFiltering performanceTransfer function
万方数据
维普
