Shape optimization on constrained single-layer sound absorber by using GA method and mathematical gradient methods

Abstract

The problem of space constraint in sound absorber design occasionally occurs in practical design work for the requisite maintenance and access. To ultimate the performance of sound absorption under space constraint, the shape optimization to maximize the absorber's performance is thus arising, accordingly. In this paper, the numerical optimal studies on single-layer sound absorber are presented. Two categories of numerical approaches, including the genetic algorithm ( GA) method and traditional gradient methods, are applied into the muffler design work. The acoustic impedance of sound absorber in evaluating the sound absorption coefficient is in conjunction with these numerical techniques. Here, GA technique and traditional gradient methods are programmed by MATLAB and FORTRAN individually. A numerical case on the single-layer sound absorber in dealing with the pure tone noise is introduced. Before optimization, one example is tested and compared with the experimental data for an accuracy check of the mathematical model. It indicates that the result is in good agreements. Consequently, the novel scheme, genetic algorithm, is capable of solving the shape optimization of a single-layer sound absorber under boundary constraint.