Sensitivity analysis of a sound absorption model with correlated inputs. (28th April 2017)

• Record Type: Journal Article
• Authors: Chai, W.; Christen, J.-L.; Zine, A.-M.; Ichchou, M.
• Is Part Of: Journal of sound and vibration. Volume 394(2017)
• Page Start: 75
• Page End: 89
• Publication Date: 2017-04-28
• Subjects: Global sensitivity analysis -- Correlated inputs -- FAST -- JCA model -- Micro-macro model; Sound -- Periodicals; Vibration -- Periodicals; Son -- Périodiques; Vibration -- Périodiques; Sound; Vibration; Periodicals; Electronic journals; 620.205
• DOI: 10.1016/j.jsv.2017.01.022
• Languages: English
• ISSNs: 0022-460X

Abstract

Abstract: Sound absorption in porous media is a complex phenomenon, which is usually addressed with homogenized models, depending on macroscopic parameters. Since these parameters emerge from the structure at microscopic scale, they may be correlated. This paper deals with sensitivity analysis methods of a sound absorption model with correlated inputs. Specifically, the Johnson-Champoux-Allard model (JCA) is chosen as the objective model with correlation effects generated by a secondary micro-macro semi-empirical model. To deal with this case, a relatively new sensitivity analysis method Fourier Amplitude Sensitivity Test with Correlation design (FASTC), based on Iman's transform, is taken into application. This method requires a priori information such as variables' marginal distribution functions and their correlation matrix. The results are compared to the Correlation Ratio Method (CRM) for reference and validation. The distribution of the macroscopic variables arising from the microstructure, as well as their correlation matrix are studied. Finally the results of tests shows that the correlation has a very important impact on the results of sensitivity analysis. Assessment of correlation strength among input variables on the sensitivity analysis is also achieved.