Boeing - Scale Models give Full-size Noise Data

Boeing's name is synonymous with the production of state-of-the-art aircraft that are sold to airlines throughout the world. There is constantly increasing focus on the environmental effects of aircraft noise - from authorities, airlines and the general public.

Boeing's relationship with Brüel & Kjær began over thirty years ago and, since the mid 1990s, Boeing has used arrays of Brüel & Kjær microphones in hard-wall pressurised wind tunnels to determine the airframe noise of aircraft designs. The acoustic measurements are currently made at frequencies up to 88 kHz.

Noise Engineering Laboratory
Boeing's Noise Engineering Laboratory in Seattle, Washington, was established coincident with the introduction of the Boeing 707 jet transport in the late 1950s. It measures all aspects of both civil and military airplane noise.

James R. Underbrink is an Associate Technical Fellow in Boeing's Noise Engineering Laboratory Dynamic Data Systems and Methods Group. He says, "On new aircraft, environmental issues are very important and airlines want to avoid fees for exceeding noise limits imposed at airports. Airframe noise is a critical factor during approach and landing. Over the years, engine noise has been consistently reduced. Now, the noise generated by the airframe is about the same as from the engine at some overhead positions and therefore there is great focus on this".

Acoustic Measurements in Pressurised Wind Tunnels
Jim explains, "Back in the 1970s, airframe noise research was generally carried out using the whole-airplane approach in specialised acoustic test facilities with anechoic test sections. Aeroacoustic testing in pressurised wind tunnels was not feasible but recent advances in acoustic measurement technology using phased arrays of microphones placed flush in the wall of the test section have enabled us to acquire noise data using scale models in hard-wall pressurised wind tunnels. We can accurately emulate the full-size aircraft, using full-scale wind speeds. Airframe noise research is generally focused on the components that produce noise such as flaps, landing gear, and high lift leading edge devices. The noise data can be included at the design stage of a new aircraft".

Jim continues, "It is necessary to work at very high acoustic frequencies as the frequency of the noise created using a scale model is inversely proportional to the size of the model. For example, if noise is measured at 50 kHz using a one-tenth-scale model, this equates to 5 kHz on the full size object. Typical airframe noise is a significant contributor to certification noise levels within the range of 500 Hz to 6 kHz."

Microphone Development
Microphones are generally designed to have a flat frequency response. Variations in atmospheric pressure are known to affect the microphone's response. With the advent of the use of pressurised wind tunnels for acoustic testing, Boeing asked Brüel & Kjær to investigate the response of its 1/4-inch Microphone Type 4136 at several pressures. It was found that, at a pressure of six atmospheres, the microphone has a gain of 5 dB in sensitivity at 15 kHz. At 50 kHz, the microphone has a drop in sensitivity of some 25 dB.

As a result of these investigations, Brüel & Kjær began development of a new 1/4-inch microphone. The goal - to have an improved, flatter frequency response under high air pressures at high acoustic frequencies. The result was the 1/4-inch Microphone Type 4938-W-001.

The maximum loss in sensitivity is only 15dB and correction factors for the microphone's response under various pressures are applied to all test data. The sensitivity of the new microphone starts rolling-off at about 60 kHz but it is usable beyond 80 kHz.

Testing
Jim says, "Many microphones are used in each array because the noise sources we're trying to measure are typically much lower in level than the boundary layer noise at the microphone diaphragms. More microphones also improve the array's dynamic range, and its ability to distinguish a source from the array noise floor. I have designed and used logarithmic spiral arrays with up to 251 microphones".

Logarithmic spirals of microphones perform well over a sufficiently wide frequency range to give quality test data. The multi-arm spiral has been found to be superior to the single-arm configuration.

Source location measurements at 80 kHz require that microphone positions relative to one another be known with an accuracy of up to 0.003 inches (0.076mm). Coordinate-measuring machines are used to survey manufactured array panels to achieve this level of accuracy. The entire array is calibrated in-place using a calibration source that is omnidirectional over the array aperture.

Jim says, "In the future, we would like to consider using prepolarized microphones. This could greatly decrease the cost per channel of an acoustic phased array measurement capability, allowing us to increase to several hundred or even a thousand channels. We would also like to increase the maximum frequency at which we can measure and so enable us to test our scale models at the high end of the frequency range corresponding to full-scale frequencies of interest. This is currently limited to 88.3 kHz by the complete data acquisition system".

Benefits
Jim concludes, "The work we do in the Noise Engineering Laboratory improves the decision-making process. It enables us to reduce the development time on new designs and derivatives of existing aircraft types. We can develop more efficient lower-cost designs and choose the configuration that gives the lowest noise level with the required aerodynamic performance. We are continually improving our knowledge and techniques, and the data from the model tests is used to design quieter airplanes".

Brüel & Kjær thanks James R. Underbrink of Boeing's Noise Engineering Laboratory for his help with the writing and editing of this article.

Pictures reproduced with the kind permission of the Boeing Company.