Impact of damping uncertainty on SEAmodel response variance

摘要

Statistical Energy Analysis (SEA) is commonly used to predict high-frequencyrnvibroacoustic levels. This statistical approach provides the mean response over an ensemblernof random subsystems that share the same gross system properties such as density, size,rnand damping. Recently, techniques have been developed to predict the ensemble variancernas well as the mean response. However these techniques do not account for uncertainties inrnthe system properties. In the present paper uncertainty in the damping loss factor isrnpropagated through SEA to obtain more realistic prediction bounds that account for bothrnensemble and damping variance. The analysis is performed on a floor-equipped cylindricalrntest article that resembles an aircraft fuselage. Realistic bounds on the damping loss factorrnare determined from measurements acquired on the sidewall of the test article. Thernanalysis demonstrates that uncertainties in damping have the potential to significantlyrnimpact the mean and variance of the predicted response.