A simplified model for the prediction of airplane interior noise by the statistical energy analysis (SEA) method

Abstract

Prediction of interior noise essentially represents a definition of attenuation of noise through a sidewall structure. It is a combination of response of the fuselage primary structure to the excitation and subsequent vibration and reradiation. In the cabin noise study, the cabin interior space is an indirectly excited subsystem, the surrounding exterior space is the directly excited subsystem and the fuselage shell structure forms the barrier. The important paths of vibrational energy exchange are mechanical and acoustical. The mechanical connections between the fuselage skin, stringer, frame, vibration isolator, and trim panels provide for the transmission of energy from the exterior space to the interior space by a mechanical path. The flow of vibrational energy via the sidewall cavity is by an acoustical path. The frames and stringers as well as the skin are the elastic structural elements that are capable of being excited under resonant, coincidence and nonresonant conditions. For a typical fuselage structure the frame coincidence frequency is about 80 Hz and the stringer coincidence is about 300 Hz. A typical skin coincidence frequency is 12 000 Hz. This indicates that in an ultimately desired basic SEA model, an appropriate representation for the primary fuselage structure, particularly for adequate low to mid frequency region, is to preserve the identity of all the three basic elements skin, frame and stringer. This makes the model very complex (Fig. S-1). Initially as a first step to SEA capability an alternate simplified model of intermediate complexity is considered. In this simplified model (Fig. S-2) the frame and stringer stiffened shell is replaced by an orthotropic shell of an equivalent stiffness. In addition it is assumed that mechanical path is secondary in importance to the acoustic path and can be eliminated as a first approximation. The method for the energy flow calculations and the evaluation of the parameters for this model has been programmed on the CDC6600 computer. The SEA cabin noise computer program has been used to calculate 737 fuselage skin vibration and interior noise levels. Figure S-3 shows calculated and measured 737 skin vibration levels for takeoff. The predicted values of the present analysis, with orthotropic representation for the fuselage shell are seen to compare quite well with the measured ones. Figure S-4 compares calculated and measured 737 interior noise levels for a ground engine run. In the low- to midfrequency range the SEA predicted interior noise levels is about 2 dB higher than the noise levels measured on a 737 during a ground engine run test (Fig. S-4). The limitations of the simplified model are discussed and the further planned analytical modifications are outlined.