prefaceIn recent years, with the increasing attention of human beings to environmental noise, developed countries around the world have formulated noise standards for civil aircraft takeoff and landing. Due to its excellent noise performance, the Boeing 777 launched in the mid-1990s became the preferred choice for many long-distance routes. Although the Boeing 777, which entered the international aviation market in 1995, had achieved all design goals, it was soon discovered that it often emitted a whistling like sound during takeoff and landing. The frequency of the whistling sound was quickly measured to be around 2000 hertz, but Boeing engineers were unable to determine where the whistling came from? In addition to the huge engine during flight, the vibration of various other components on the aircraft and the friction between the fuselage and air can produce noise. It is difficult to decompose such a complex noise source from a high-speed flying object one by one. Boeing engineers helplessly referred to this whistling as the '2000 hertz mystery tone'.

Figure 1.There is a row of small holes at the front edge of the Boeing 777 main wing. In cold environments, the airflow passing through the small holes will be heated and then circulated inside the wing to prevent moisture from freezing on the wing

Years later, Boeing researchers deployed a spiral microphone array with a diameter of up to 150 feet on the runway of the airport to record the noise emitted by the Boeing 777 flying over. The results of repeated experiments indicate that the 2000 Hz howl comes from the leading edges of both wings. The front edge of the Boeing 777 main wing has a row of small holes (see Figure 1). In cold environments, the airflow passing through the holes will be heated and then circulated inside the wing to prevent moisture from freezing on the wing. Based on the results of microphone array detection and analysis of the wing structure, Boeing researchers suspect that the "culprit" behind the mysterious whistling is the two rows of small holes. When the incoming airflow passes through the neatly shaped and arranged small holes, it resonates with the wings at 2000 hertz, just like people playing a flute. To confirm this speculation, acoustic expert Rob Stoker convinced relevant departments of Boeing to seal all the small holes at the leading edge of one wing with tape, and then compare the noise intensity of the two wing leading edges.

Figure 2. After sealing the small hole at the leading edge of one wing, the 2000 Hz whistling almost completely disappeared

Figure 2 shows the test results after sealing the small hole on one side. As predicted by Stoker, the 2000 hertz whistling disappeared completely after the small hole at the leading edge of one wing was sealed. Boeing engineers have finally solved their long-standing mystery! Based on the results of microphone array detection, Boeing designers have redesigned the shape and arrangement of the anti freezing holes at the leading edge of the wing. The improved Boeing 777-300ER series completely eliminates the "mysterious two thousand week" roar.

Figure 3. Mr. Stoker, an excited Boeing acoustics expert after success

The story of the successful application of acoustic cameras on Boeing aircraft quickly spread to Airbus in Europe and other aircraft manufacturing companies around the world. Nowadays, microphone arrays are not only used to study noise sources on airplanes and cars, but also in noise research in industries such as submarines, construction, and home appliances.

The following article will briefly introduce the principle, application examples, and common problems in the application of acoustic cameras. In order to enable non professional readers to have a comprehensive understanding of this emerging technology, this article will try to avoid using professional vocabulary and mathematical formulas as much as possible, so some assumptions may appear too ideal and some explanations may appear too simple.

Detection principleFigure 4 is a schematic diagram of Boeing's testing. The black dots inside the circle represent the array of acoustic cameras. The red line represents the transmission path of sound waves from the sound source s (t) to the microphone. Due to the unequal distance between the sound source and each microphone, the sound waves received by each microphone have different time delays t_iIn the frequency domain, it is called phase difference, which can be roughly described mathematically as x_i(t) = s(t-t_i)+N_i(t) Here x_i(t) And N_i(t) Representing the signal received by the i-th microphone and the interference with a mean of zero. Because the structure of the microphone array and the speed of sound propagation are known, we can use our learned knowledge of trigonometry and geometry to solve for a set of corresponding time delays {t} for each sound source at each position in space_i}. If the signal x received by each microphone_i(t) Compensate separately for t_i(That is, x_i(t+t_i））Align the sound waves from the sound source and then add all M compensated signals x_i(t+t_i）Add them together, and finally, the interference tends to zero (because the interference N_i(t) The mean of the enhanced sound wave Ms (t) is zero.

Figure 4. Boeing Testing Diagram

As mentioned above, sound sources from different directions correspond to a unique set of time delays, whereas each set of time delays points to a unique sound source. So, using the one-to-one correspondence between sound wave delay and sound source position, we can calculate the spatial sound intensity distribution map point by point by compensating for the delay of each received signal and then adding them up. In such applications, microphone arrays can be regarded as "acoustic cameras". However, the lens of a regular camera focuses on light waves, while the microphone array of an acoustic camera focuses on sound waves.

Figure 5 The spatial distribution of sound intensity calculated from rectangular (left image) and spiral (right image) microphone arrays. The spiral microphone array accurately detects three sound sources, while the rectangular microphone array mixes multiple false sound sources that do not exist in the real world while detecting three real sound sources.

The resolution of acoustic camera images is closely related to the number of microphones and the shape of the array. Generally speaking, the more microphones there are, the higher the resolution. The relationship between resolution and microphone array shape is quite complex, except for the simple analytical relationship between cross shaped and rectangular arrays and resolution, the relationship between other shape arrays and resolution is not clear at a glance. The cross shaped and rectangular array structures are simple and easy to install; The structure of a spiral is complex, but mathematically it can be proven that the performance of a spiral is optimal. Figure 5 shows the spatial distribution of sound intensity calculated using rectangular (left image) and spiral (right image) microphone arrays, respectively. Although the number of microphones is the same, the result of a spiral microphone array is significantly better than that of a rectangular one. The spiral microphone array accurately detects three sound sources, while the rectangular microphone array mixes multiple false sound sources that do not exist in the real world while detecting three real sound sources.

Figure 6. Spiral microphone array

Before the advent of acoustic cameras, people used Near Field Acoustic Holography to test the spatial distribution of noise intensity. Compared to acoustic cameras, acoustic holography typically requires an array of microphones with an area at least as large as the surface of the object being measured. In addition, acoustic holography requires that the distance between the microphone and the object being measured must be sufficiently small (usually within 10 centimeters). However, in Boeing's applications, the tested aircraft is typically located about 150 meters above the microphone array. So acoustic holography cannot meet Boeing's needs. The advantage of acoustic holography is that its low-frequency resolution is fixed and does not vary with frequency; And acoustic cameras, the lower the frequency, the worse the resolution.

Due to the focusing function of microphone arrays, in addition to acoustic cameras, they are also widely used as spatial filters to enhance sound waves from specified directions. Nowadays, many video or conference call devices (including pure software products such as Microsoft Vista and XP) have microphone array functionality. In those applications, microphone arrays are used to enhance sound waves from the speaker's direction and suppress interference from all other directions. The sound intensity distribution image obtained by acoustic cameras is two-dimensional, while the output of spatial filters based on microphone arrays is usually one-dimensional sound signals.

Figure 7. The noise generated by the friction between the rear wheels and the ground of the Porsche 911 will be much stronger than the noise generated by other parts of the car.

Application CasesSo far, acoustic cameras have mainly been used to identify the location of noise sources. In addition to the various types of noise emitted by aircraft during flight introduced at the beginning of this article, acoustic cameras have also been successfully used to study the noise of moving cars, electric locomotives, and maglev trains. Figure 7 shows the test results of the Porsche 911 while driving, with a speed exceeding forty miles per hour in this experiment. People usually think that the noise of a car during high-speed driving mainly comes from the engine and exhaust pipe, but the images produced by acoustic cameras tell us that the noise generated by the friction between the wheels and the ground is much greater than the noise emitted by other parts of the car. The Porsche 911 is rear wheel drive, with the engine located in the trunk and the car's center of gravity tilted towards the rear. Therefore, the noise generated by the friction between the rear wheels and the ground will be much stronger than the noise generated by other parts of the car.

Figure 8. TECLAB Spiral Acoustic Camera for Noise Localization and Diagnosis of Great Wall Motors Engine Room. It was found that the 500Hz noise originated from the engine body, the 1000Hz noise originated from the pulley, and the 2500Hz noise originated from the pipeline system

Figure 9. TECLAB spiral acoustic camera was used to analyze the noise source of wind turbines, and multiple abnormal noises were found with frequencies between 187Hz-206Hz

Figure 10. After conducting noise analysis on the Harmony high-speed train using the TECLAB spiral acoustic camera, joint modifications can be made to key areas such as wheels, rails, and carriages based on the analysis results to improve the design structure and achieve the goal of reducing overall noise.

Figure 11. After conducting noise analysis on the TECLAB spiral acoustic camera for detecting fans, air conditioners, and range hoods, the optimization design was used to achieve the goal of reducing noise and improving the quietness effect of household appliances, thereby enhancing user experience and comfort.

Figure 12. TECLAB spiral acoustic camera measured in an environment with strong noise interference in an industrial workshop and found strong vibration in a 7MPa high-pressure gas storage tank. The vibration source was accurately identified, reflecting the safety hazards of the gas storage tank.

expectationFrom Boeing's successful use of acoustic cameras for early noise research on the Boeing 777 to today, the application of acoustic cameras has rapidly expanded from expensive large passenger aircraft to small automotive air conditioning systems. Acoustic cameras have evolved from one-dimensional arrays and cross arrays to the latest and most optimal spiral arrays, with continuous upgrades and improvements in detection accuracy and effectiveness. Keeping our lives away from noise pollution and continuously improving production safety. In the future, acoustic cameras will exert their powerful and unique functions in more fields, improving our quality of life.