A study of the assumptions used in statistical energy analysis

Thibault Lafont (1) , Nicolas Totaro (2) , Alain Le Bot (1)

1 LTDS - Laboratoire de Tribologie et Dynamique des Systèmes
2 LVA - Laboratoire Vibrations Acoustique

This paper is a discussion on the coupling strength, the frequency bandwidth and indirect coupling loss factor in statistical energy analysis. The example of coupled oscillators is taken. First, it is shown that the coupling power proportionality is always valid for two coupled oscillators for any strength of coupling as long as the resonant modes are within the frequency band. Then, the case of three coupled oscillators indicated that the apparition of indirect coupling loss factors depends on the coupling strength. Finally, it is shown that modes outside the frequency bandwidth of excitation can lead to invalid the reciprocity relation. Some test cases are presented to illustrate these conclusions.

Mots clés: Statistical energy analysis; coupling power proportionality; indirect coupling; coupling loss factors

Domaines: Acoustique [physics.class-ph]

Thibault Lafont, Nicolas Totaro, Alain Le Bot. A study of the assumptions used in statistical energy analysis. Inter.noise 2014, Nov 2014, Melbourne, Australia. pp.662. ⟨hal-01086611⟩