摘要

To evaluate the effects of structural parameters on sound absorption performance of the aluminum foams with semiopen cells, a generalized regression neural network (GRNN) model has been utilized. In this paper, parameters of bulk density, static flow resistivity were analyzed. Results indicate that the effects of bulk density on sound absorption properties of the semiopen-celled foams are limited. On the contrary, the effects of static flow resistivity on absorption coefficients are complicated. When frequency is 500 Hz, there has no significant relationship between static flow resistivity and sound absorption coefficients. When the frequency is higher than 1000 Hz, with flow resistivity increasing, absorption coefficients of the aluminum foams increased significantly from 0.2 to exceed 0.8. However, when the static flow resistivity reaches 140 kPa s/m(2), still increasing static flow resistivity could not improve the acoustic performance. Analysis revealed that the GRNN model provides a robust method for predicting acoustic absorption properties.