Active Sound Power Attenuation with a Ring of Harmonic Acoustic Pneumatic Sources fOr Destructive Interference (RHAPSODI) and near field in-duct microphones

Philippe Micheau 1 , Julien Drant 2 and Alain Berry 3 CRASH-UdeS Mechanical Engineering Dpt Université de Sherbrooke 2500 Blvd de l’université, Sherbrooke, QC, J1K 2R1, Canada ABSTRACT

Much research has been conducted to investigate active noise control of turbofans with loudspeak- ers as noise cancellation sources. However, the required power consumption, fragility, and weight and volume penalty make them unsuitable for engine nacelle applications. An alternative source tech- nology, developed at Sherbrooke, is a harmonic acoustic pneumatic source (HAPS). They are mounted on a ring to perform multimodal active noise control in duct by destructive interference of primary noise (RHAPSODI). A dedicated MIMO harmonic control strategy based on complex enve- lopes is required to control the RHAPSODI with in-duct microphones located at a very close distance from it or with external in-duct microphones located 1 meter away from the duct mouth. During a training phase, active radiated power minimization is performed with the external microphones for different modal components of the primary harmonic noise. The optimal HAPS control and the opti- mal signals from the in-duct microphones are then used to tune the specific MIMO controller with a field compensation matrix close to the in-duct microphone signals. A typical result shows that RHAP- SODI using 5 in-duct microphones and 5 HAPS can provide high attenuation of radiated sound power (22 dB-SPL, 1 kHz, primary sound power 130 dB).

1. INTRODUCTION

The need to reduce aircraft noise was highlighted by the Advisory Council for Aviation Research and Innovation in Europe [1]. During takeoff, the harmonic noise of the turbofan is the main acoustic nuisance for people in the vicinity of airports. Many researches have been conducted to perform active noise control of turbofans with loudspeakers as sources of anti-noise[2,3,4]. Although, the high noise levels involved pose a challenge for the technology of loudspeakers in harsh environments such as aircraft engine nacelles [4]. The required energy consumption, fragility as well as the weight penalty makes them unfit for engine nacelle applications.

1 Philippe.Micheau@USherbrooke.ca

2 Julien.Drant@USherbrooke.ca

3 Alain.Berry @USherbrooke.ca

a 2022

An alternative technology, proposed by Micheau et al. [5], is to use a Harmonic Anti-Noise Pneu- matic source (HAPS) as anti-noise source. It is based on the concept of the pneumatic loudspeaker [6,7,8,9]. It is known that flow modulations of a high-pressure airstream can provide the most effec- tive sound sources to test satellites in large reverberant test chambers [10]. But few research has been conducted on the use of pneumatic sources as secondary actuators in active control systems. A sub- sonic pneumatic source has been investigated by Blondel [11] for active noise control; but, its fre- quency was below 100 Hz. In order to overcome this upper frequency limit, the option is to use a rotary flow chopper to open and close an orifice that modulates the air flow [12]. Figure 1 presents the developed RHAPSODI of 6 HAPS mounted on a circular duct. The 6 perfectly controlled pulsed jets are ejected in the duct from 6 holes in the duct.

a 2022

Figure 1: Ring of Harmonic Acoustic Pneumatic Sources fOr Destructive Interference (RHAP- SODI) with 6 HAPS. The rotation of the flow chopper with multiple holes generates a periodic sound with its first har- monic controllable in amplitude and phase [13]. With such innovative device, it is possible to achieve SISO active sound control of a harmonic disturbance with a dedicated complex envelop controller [13,14]. Because the sound attenuation in far field is based on destructive interference, this project has been called RHAPSODI for Ring of HAPS fOr Destructive Interference.

2. RHAPSODI

2.1. Harmonic Acoustic Pneumatic Source (HAPS)

This section presents the principle of the HAPS and the configuration of the RAPSODI as a ring of HAPS around a duct.

Figure 2 shows a description of the main components of HAP. The two mechanical components of a harmonic acoustic pneumatic source (HAPS) are a valve for flow regulation and a flow chopper using a rotating perforated cage [5]. The chopper and the valve are equivalent to a throat orifice between a plenum of high pressure and the HAPS exhaust at the atmospheric pressure. Since the flow is supposed to become sonic through the equivalent orifice, it is possible to explain the mass flow rate exhausted from the device as a function of the motor angle and valve opening, [25]. The time variation of the orifice due to the rotation of the chopper generates a periodic volume velocity. The phase of the pulsed flow is then directly related to the rotation of the flow chopper. The Phase-Locked Loop (PLL) commands the chopper speed in order to synchronize the measured instantaneous angle of the chopper with the instantaneous reference angle constructed with the phase command input.

Finally, the periodic mass flow rate generates a periodic anti-noise (pressure fluctuation in the far field). For the first harmonic, the magnitude of the generated anti-noise is commanded by the servo- valve, since the phase of the anti-noise is commanded by the motor. Hence, each HAPS is controlled by two signals (angle command and amplitude command) in order to physically perform a complex- amplitude modulation of the generated anti-noise.

a 2022

Figure 2: Noise generation with an HAPS.

2.1. Control of the RHAPSODI

Due to the specificity of the RHAPSODI, the real-time implementation of the MIMO harmonic controller is achieved with the complex envelops of inputs and outputs. This approach was success- fully implemented for the active control of narrowband controllers working with harmonic signals [21, 22, 23, 24, 5]. The MIMO harmonic controller with complex envelops is equivalent to a narrow- band MIMO feedback control [26].

In the targeted application of turbofans, the error microphones should be located very close to the HAPS [15]. With such configuration, it can be expected that the active control should be relatively robust against model uncertainties [16]. But, a significant portion of the error signals is due to non-

Noise emission co Pulsed jet = P Resolver motor Chopper Servo-valve Compressed Air Pulsed flow |

propagating components of the HAPS: minimizing error microphone signals does not ensure an op- timal attenuation of the radiated sound. Hence, the classical minimization of the error microphone [17] is not the optimal solution. Such problem can require remote sensing approach [18] or control of acoustic power [19]. To solve the problem a local compensation matrix M (or equivalent to a virtual impedance approach [20]) is included in the MIMO active harmonic control using error mi- crophones located close to the RHAPSODI. However, the theoretical analytical form of the compen- sation matrix M as a function of transfer matrices is not useful in practical applications. Hence, a dedicated learning method was implemented to tune the compensation matrix with M a set of optimal commands defined as the minimization of the far-field radiated harmonic sound measured with mi- crophones located out of the duct [27].

3. EXPERIMENTATIONS

a 2022

3.1. Experimental set-up

The cylindrical duct, which allows the propagation of radial and circumferential acoustic modes, is often used in the literature to approximate an acoustic response of the nacelle. This simplified configuration is used in this paper to develop an experimental set-up presented on Figure 3. The vein is composed of 3 sections : a section with the primary noise using a ring of loudspeakers, a section hosting the RHAPSODI, and a last section with in-duct microphones. The left termination of the duct is closed. Therefore, the sound is emitted in a semi-anechoic chamber only from the right termination of the duct. The radiated sound power is measured in the far field with out-duct external microphones.

The presented tests were performed with 5 HAPS. The in-duct microphones were located at 100 mm from HAPS. The 10 error microphones were located on the hemispheric support at 1 m from the duct end. The disturbance frequency was 1 kHz and the supply pressure was 69kPa.

Figure 3: Experimental set-up in the Anechoic room: Righ: (1) duct with the RHAPSODI, (2) drivers for the RHAPSODI, (3) real-time controller, (4) amplifiers for primary sources ; Left : (1-2) absorbing acoustic materials, (3) hemispheric support of microphones in far field, (4) hard wall, (5) end of the duct.

3.2. Typical experimental result

The frequency spectrum of the radiated sound power with and without active control is presented in Figure 4. The radiated acoustic power without active control is very large (acoustic power Lp=130dB at 1 kHz). However, due to this large sound level, undesirable higher harmonics are gen- erated (2 kHz, 3 kHz) by the compression chambers. With active control, an important attenuation of the targeted first harmonic is reached (attenuation of 24 dB). Other experiments (with 6 or 5 HAPS, distance between the RHAPS and the error microphones at 150 and 200 mm, different configurations) confirmed that the developed system is efficient to achieve the desired power attenuation of 20 dB of the first harmonic [27,28].

a 2022

Figure 4: Frequency spectrum of the radiated power for a) Control off b) Control on 4. CONCLUSIONS

This work has established the feasibility of active multimodal noise control in a duct using a RHAPSODI and in-duct microphones located at a close distance from it. The results shown that the presented strategy can provide the attenuation of 20 dB in the far field. Hence, the active control method with a RHAPSODI is a good candidate to develop an active harmonic noise control system of turbofans.

5. ACKNOWLEDGEMENTS

We gratefully acknowledge Marc Versaevel form Safran Nacelles (Le Havre, France) for support- ing this work.

Acoustic power level Acoustic power level (4B re 1pW) oo Spectral density of acoustic radiated power without active control 500 1000 1500 2000 2500 3000 Frequency(Hz) Spectral density of acoustic radiated power with active control 3500 500 1000 1500 2000 2500 3000 Frequency(Hz) 3500

6. REFERENCES

1. M. Darecki, C. Edelstenne, T. Enders, E. Fernandez, P. Hartman, J.-P. Herteman, M. Kerkloh, I. King, P. Ky, M. Mathieu, G. Orsi,, G. Schotman, C. Smith, and J.-D. W¨orner, “Flightpath 2050 – report on europe’s vision for aviation,” 2011. 2. F. G. Pla, Z. Hu, and D. L. Sutliff, “Active control of fan noise: Feasibility study - active fan noise cancellation in the nasa lewis active noise control fan facility,” 3 (1996). 3. N. Yamasaki and H. Tajima, “Active control of fan noise,” Journal of Thermal Science , 17 , 163– 169, (2008). 4. E. J. Kors, “Openair, optimisation for low environmental noise impact aircraft internoise 2010,”, INTER-NOISE 2010 , 5 , 4163–4167 (2014). 5. P. Micheau, B. Burel, R. Robert, and P. Masson, “Control of amplitude, frequency and phase of an harmonic anti-noise pneumatic source,” International Congress on Sound and Vibration , 1–8, (2017). 6. W. T. Fiala, J. K. Hilliard, J. A. Renkus, and J. J. V. Houten, “Electropneumatic acoustic gener- ator,” The Journal of the Acoustical Society of America, vol. 37, pp. 1177–1178, 6 1965.Poulsen, T. Influence of session length on judged annoyance. Journal of Sound and Vibration, 145(2) , 217–224 (1991). 7. R. J. Hooker and R. H. Rumble, “Noise characteristics of a pulsed jet,” Noise Control Engineer- ing, vol. 17, Nov.-Dec. 1981, p. 113-119. Research sponsored by the Australian Mineral Indus- tries Research Association , 17 , 113–119, (1981). 8. C. Chapman and A. Glendinning, “A theorical analysis of a compressed air loudspeaker,” Journal of Sound and Vibration , 138 , 493–499, (1990). 9. M. Prek, “Flow-induced noise generation in a simple fluid valve model,” Noise Control Engi- neering Journal , 52 , (2004). 10. A. Grewal, R. Ramakrishnan, W. O. Hughes, B. Woyski, G. Elfstrom, C. Mech, and Y. Chen, “High intensity noise generation for extremely large reverberant room test applications,” Confer- ence Proceedings of the Society for Experimental Mechanics Series , 6 , 103–118, (2011). 11. L. A. Blondel and S. J. Elliott, “Electropneumatic transducers as secondary actuators for active noise control, part ii: Experimental analysis of the subsonic source,” Journal of Sound and Vibra- tion , 219 , 429–449 (1999). 12. R. C. Jones, “A fifty horsepower siren,” The Journal of the Acoustical Society of America , 18 , 371–387, (1946). 13. J. Drant, P. Micheau, and A. Berry, “Active noise control in duct with a harmonic acoustic pneu- matic source,” Applied Acoustics , 176 , (2021). 14. M. Michau, A. Berry, P. Herzog, and P. Micheau, “Optimal virtual mechanical impedances for the vibroacoustic active control of a thin plate,” Journal of the Acoustical Society of America , 137(1) , 199–207, (2015). 15. J.-P. Rami, H. Hurlin, J.-B. Goulard, M. Versaevel, P. Micheau, J. Drant, A. Allard, A. Berry, and R. Robert, “Active device for attenuating acoustic emissions for a turbojet comprising con- trolled turbines w02019/171000a1,” Patent, 2019. 16. A. P. Berkhoff, “Control strategies for active noise barriers using nearfield error sensing,” The Journal of the Acoustical Society of America , 118(9) , 1469–1479 (2005). 17. S. Elliott, Signal Processing for Active Control. Academic Pres s, (2001). 18. D. Moreau, B. Cazzolato, A. Zander, and C. Petersen, “A review of virtual sensing algorithms for active noise control,” Algorithms , 1 ,69–99, (2008).

a 2022

19. S. D. Sommerfeldt and K. L. Gee, “Multi-channel active control system and method for the re- duction of tonal noise from an axial fan, patent US20030219132, (2007). 20. M. Michau, A. Berry, P. Micheau, and P. Herzog, “Optimal virtual mechanical impedances for the vibroacoustic active control of a thin plate,” The Journal of the Acoustical Society of America , 137 , 199–207, (2014). 21. P. Micheau and P. Coirault, “Adaptive controller using filter banks to reject multi-sinusoidal dis- turbance,” Automatica , 36 , 1659–1664, (2000). 22. M. Baudry, P. Micheau, and A. Berry, “Decentralized harmonic active vibration control of a flex- ible plate using piezoelectric actuator-sensor pairs,” Citation: The Journal of the Acoustical Soci- ety of America , 119 , 262, (2006). 23. P. Micheau, R. Louviot, and A. Berry, “Decentralized resonant controller for vibroacoustic active control,” 2007 Mediterranean Conference on Control and Automation , August 2007. 24. P. Micheau, R. Boulandet, M. Michau, and A. Berry, “Active structural acoustic control of a composite panel with sensori-actuators and complex envelope controllers,” 22nd International Congress on Sound and Vibration , ICSV 2015, pp. 12–16, August 2015. 25. A. Allard, N. Atalla, and P. Micheau, “The equivalent fluctuating area for modelling harmonic acoustic pneumatic sources (HAPS),” Applied Acoustics , 172 , (2021). 26. P. Micheau and S. Renault, “Active control of the complex envelope associated with a low damped mode,” Mechanical Systems and Signal Processing, 20 , 646–661, (2006). 27. J. Drant, “Développement et validation expérimentale du concept de l'atténuation active de la puissance acoustique rayonnée par un turbofan avec un RHAPS et des microphones en champ proche,” PhD thesis of Université de Sherbrooke , In French (2020). 28. J. Drant, P. Micheau, and A. Berry, “Active noise control of higher modes in a duct using near field compensation and a ring of harmonic acoustic pneumatic sources,” Applied Acoustics , 188 , (2022).

a 2022