確率的均質化法による多孔質材料音響特性の不確実性定量化

抄録

To improve the acoustic characteristics of porous materials, we calculated the sound absorption coefficient using the homogenization method. The homogenization method can evaluate the sound absorption coefficient directly from the microstructure of porous materials. On the other hand, since this method uses representative volume elements, it cannot calculate the microscopic variation of porous materials. In this study, we proposed methods to calculate the probability distribution of the sound absorption coefficient considering the microscopic variation by using the homogenization method and probabilistic methods. We applied four stochastic methods, Monte Carlo simulation, Perturbation method, Gaussian quadrature method, andB ayesian quadrature method, with the homogenization method, and compared the probability distributions of the sound absorption coefficient. This comparison clarified the characteristics of each stochastic method. The results showed that the perturbation method had the lowest computational cost, but the accuracy was not sufficient when the variability of random variables was large. The Gaussian quadrature method and theB ayesian quadrature method were able to predict the sound absorption coefficient distribution with high accuracy even when the variation of random variables was large. Still, the computational cost tended to be larger than that of the perturbation method. By choosing these methods appropriately, we can evaluate how the variation of the microscopic structure of porous materials affects the sound absorption coefficient.