Applications of Response Surface-Based Methods to Noise Analysis in the Conceptual Design of Revolutionary Aircraft

Due to the growing problem of noise in today's air transportation system, there have arisen needs to incorporate noise considerations in the conceptual design of revolutionary aircraft. Through the use of response surfaces, complex noise models may be converted into polynomial equations for rapid and simplified evaluation. This conversion allows many of the commonly used response surface-based trade space exploration methods to be applied to noise analysis. This methodology is demonstrated using a noise model of a notional 300 passenger Blended-Wing-Body (BWB) transport. Response surfaces are created relating source noise levels of the BWB vehicle to its corresponding FAR-36 certification noise levels and the resulting trade space is explored. Methods demonstrated include: single point analysis, parametric study, an optimization technique for inverse analysis, sensitivity studies, and probabilistic analysis. Extended applications of response surface-based methods in noise analysis are also discussed. rd octave frequency and spherical directivity. The SPL distribution of each source may be obtained by prediction from analytical tools, or from test data. Turbofan engines are usually broken down into five sources: inlet radiated fan noise, exhaust radiated fan noise, combustor or core noise, turbine noise, and jet noise. Distinct sources of a transport airframe include: wing trailing edge, flaps, slats, and landing gear. When all the sources are characterized, the second step in the analysis is to fly them along the vehicle's flight path and propagate the emitted noise to ground observers via ray tracing methods. The propagation analysis can take into account the effects of atmospheric attenuation, ground absorption, and reflection. The noise at the observer may then be converted into a certification or community noise metric that can take into account the human response to frequency, discrete tones, and duration of the noise event. Due to the logarithmic nature of noise measurement, the total vehicle noise will be dominated by the strongest sources and less influential ones, even by a few decibels, will make very little of a contribution. For example, a noise source with an SPL 10 dB below another will increase the combined noise by less than 1/2 dB. Total vehicle noise prediction and subsequent assessment therefore demands that the relative proportions of the noise levels of all the discrete sources be accurately predicted. It is also important that noise reduction technologies and design concepts be evaluated in the context of the entire system to assess their community noise impact. Failure to do so may lead to misleading conclusions about the influence of a particular source or the effectiveness of a technology or design concept on the vehicle. Such accurate system predictions often do not exist in the conceptual design phase due to uncertainty in source level predictions, and lack of sufficient technology and vehicle definition. For these reasons, it is useful to develop and explore a trade space in which different source level proportionalities may be