The active minimisation of harmonic enclosed sound fields, with particular application to propeller induced cabin noise
This work considers the use of secondary acoustic sources for the reduction of noise levels in harmonically excited enclosed sound fields. This possibility is studied initially through the use of an analytical model of a single frequency `two-dimensional' rectangular, enclosed sound field of low modal density. Quadratic optimisation theory is used to predict the effectiveness of these active techniques for reducing sound levels. It is shown that if one chooses to minimise the total acoustic potential energy of the sound field then substantial global reductions in sound pressure may be achieved. The conditions required to achieve these reductions are discussed, and in particular the effect of secondary source location is demonstrated. The unsuitability of this as a practical control system cost function is discussed, and as an alternative it is suggested that one can minimise the sum of the squared pressures over a number of discrete sensors. Both theoretical and experimental results from using this cost function are presented. These show good agreement. The importance of both secondary source and error sensor locations is discussed. The reduction of low frequency propeller induced cabin noise is a possible application of active noise control. Simplified analytical models of the structural and internal acoustic response of a 48 seat, twin turboprop passenger aircraft are presented. These consist of the structural response of an isotropic, thin cylindrical shell of finite length and the acoustic response of the enclosed cyclindrical sound field. It is shown that, provided the external acoustic pressure forcing of the shell is modelled representatively of measured propeller pressure fields, then these simple models provide results which demonstrate good agreement with measured data, at least for frequencies encompassing the first two propeller blade passenger harmonics (88 Hz and 176 Hz). These models are used to predict the effectiveness of active noise control when it is applied to reduce the average sound pressure level over a typical seated head height plane in the passenger cabin. Results for the first two blade passage harmonics are presented for systems consisting of up to 24 secondary sources and 48 error sensors. In general the predictions suggest that active noise control shows promise as a potential method of reducing low frequency propeller induced cabin noise.