Design of Launch Pad for Mitigating Acoustic Loads on Launch Vehicle at Liftoff 우주발사체 발사 시 음향하중 저감을 위한 발사대 설계

Home , Epsilon (rocket), Launch pad

한국음향학회지 제39권 제4호 pp. 331～341 (2020) The Journal of the Acoustical Society of Korea Vol.39, No.4 (2020) pISSN : 1225-4428 https://doi.org/10.7776/ASK.2020.39.4.331 eISSN : 2287-3775

Design of launch pad for mitigating acoustic loads on launch vehicle at liftoff 우주발사체 발사 시 음향하중 저감을 위한 발사대 설계

Seiji Tsutsumi1†

1Japan Aerospace Exploration Agency (Received June 17, 2020; accepted July 18, 2020)

ABSTRACT: At liftoff, launch vehicles are subject to harmful acoustic loads due to the intense acoustic waves generated by propulsion systems. Because these waves can cause electronic and mechanical components of launch vehicles and payloads to fail, predicting and mitigating acoustic loads is an important design issue. This article presents the latest information about the generation of acoustic waves and the acoustic design methods applicable to the launch pad. The development of the Japanese Epsilon solid launcher is given as an example of the new methodology for launch pad design. Computational fluid dynamics together with 1/42 scale model testing were performed for this development. Effectiveness of the launch pad design to reduce acoustic loads was confirmed by the post-flight analysis. Keywords: Launch vehicle, Aeroacoustics, Computational fluid dynamics, Computational aeroacoustics

PACS numbers: 43.50.Nm, 43.28.Js

초 록: 우주발사체는 발사 시 추진장치에서 발생하는 고강도 소음에 의한 음향하중의 영향을 받는다. 로켓소음은 발사체와 페이로드 내 전자 및 기계 부품의 손상 및 오작동을 유발할 수 있기 때문에 음향하중의 예측 및 저감은 설계에 있어 중요한 고려사항이다. 본 논문에서는 로켓 소음의 생성 및 발사대의 음향설계 기법에 대한 최신 연구동향을 논하 였다. 특히, 새로운 발사대 설계 방법론의 예로서 일본 Epsilon 로켓 발사대의 개발과정을 기술하였다. 전산유체역학 모사 및 1/42 축소모형 실험을 통하여 설계된 발사대의 음향하중 저감 효과를 Epsilon 로켓의 실제 비행 데이터 분석을 통하여 검증하였다. 핵심용어: 우주발사체, 공력음향학, 전산유체역학, 전산공력음향학

I. Introduction dynamic pressure flight regimes. The aerodynamic vibration is generated by oscillating shock waves, a Launch vehicles are exposed to severe vibration and turbulent boundary layer, and separated flow behind shock during flight. Fig. 1 shows an example of vibration protuberances. This study focuses on the acoustic loads measured with an accelerometer mounted on the upper that appear at liftoff. stage of a launch vehicle. There are two major peaks, The retired M-V solid launcher generated 3,700 kN of which appeared at 0 s ~ 5 s and 45 s ~ 80 s. The former is thrust at liftoff, and overall acoustic power radiated from caused by acoustic waves generated from exhaust jets of the exhaust jet was estimated to be 197 dB. The Overall the propulsion systems. The latter is due to aerodynamic Sound Pressure Level (OASPL) measured by a micro- vibration that occurs under transonic or maximum phone located outside the payload fairing reached 157.6

†Corresponding author: Seiji Tsutsumi (tsutsumi.seiji@jaxa.jp) Aerospace Research and Development Directorate, Research Unit III, Japan Aerospace Exploration Agency (JAXA), 3-1-1 Yoshinodai, Chuuou, Sagamihara, Kanagawa, 252-5210, Japan (Tel: 81-50-3362-3151, Fax: 81-50-3362-7189) Copyrightⓒ2020 The Acoustical Society of Korea. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

331 332 Seiji Tsutsumi

Fig. 1. Typical time history of vibration measured on launch vehicle.[1] dB, and OASPL measured inside the payload fairing was 144 dB.[2] Since electronic and mechanical components in the satellite and the launch vehicle are exposed to an acoustic environment exceeding 140 dB, predicting and reducing liftoff acoustics is one of the important design considerations in developing launch vehicles. Prediction and reduction of liftoff acoustics require knowledge in two domains, aeroacoustics and vibro- Movable Launch acoustics. Aeroacoustics deals with the generation of Platform (ML) acoustic waves from exhaust jets, their interaction with the Exhaust holes launch pad, and the propagation to the payload fairing. Flame duct On the other hand, vibroacoustics is concerned with Deflector the transmission of acoustic waves through the fairing Fig. 2. Schematic of typical launch pad. structure and the effects on the satellite inside. This article focuses on aeroacoustics and introduces the latest studies safely deflects the hot exhaust downstream. on the mechanism of acoustic generation and design The exit Mach number of the exhaust jet from a rocket methodologies of the launch pad. Then, the example of engine is at least 3, and the exit static temperature can be launch pad design for the Epsilon solid launcher, over 1000 K. The Mach wave is radiated from the free jet developed by the Japan Aerospace Exploration Agency region of the exhaust jet due to its large-scale turbulent (JAXA), is introduced with post-flight analysis. structure.[3] Empirical models proposed by Eldred et al. and Varnier are representative of the acoustic prediction II. Mechanisms of method.[4,5] According to the Varnier’s model, the major acoustic generation source of the Mach wave is generally around 10D downstream from the nozzle exit. Note that D is the nozzle Fig. 2 is a schematic of a typical launch pad. The launch exit diameter. The major acoustic-emission direction of vehicle is mounted on a Movable Launch pad (ML) and the Mach wave is roughly 50 deg from the direction of transported from a vehicle assembly building to the final jet flow. As shown in Fig. 2, when the launch vehicle rises launch pad location. The ML has exhaust holes to direct to 10D above the launch pad, the region of the jet that the exhaust jets into a flame duct. The flame duct is usually radiates the Mach waves appears from the exhaust hole of constructed underground. A deflector in the flame duct the ML then begins to contribute to the acoustic load on the

한국음향학회지 제39권 제4호 (2020) Design of launch pad for mitigating acoustic loads on launch vehicle at liftoff 333

-0.1 0.1 -0.03 ∆͊⁄ ͊ 0.03

0.1 0.4 0.8 2.0

Acoustic wave due to jet impingement

Mach wave radiation from wall jet

Fig. 4. (Color available online) Acoustic wave generated due to jet impingement on deflector.[8] (a) Instantaneous result the nozzle exit from the ground were 2D and 20D, ⁄ -0.02 ∆͊ ͕͊ 0.02 respectively. The Mach waves are radiated from the

Reflected Mach wave or large-scale turbulent structure in the free jet shear layer acoustic wave due to jet interaction with exhaust hole and propagate obliquely downstream. 1/1 octave band filter is applied to the time series data, and results at a center frequency of 4000 Hz (Strouhal number St = 0.026) is shown in Fig. 3(b). The acoustic waves returning to the Mach wave radiation from free jet launch vehicle are clear. The acoustic waves originated in the Mach waves reflect off the flat plate. The free jet grows up as it proceeds downstream and interacts with the edge of the exhaust hole. This interaction is another source of (b) 1/1 octave bandpass filtered result at St = 0.026 acoustics propagating to the vehicle.[7] It is not clear how Fig. 3. (Color available online) Acoustic wave generated much each contributes to the acoustic environment of the above ML.[6] launch vehicle. launch vehicle. The maximum acoustic level is usually Immediately after liftoff, the high-speed exhaust jet observed when the launch vehicle is elevated to approxi- enters the flame duct through the exhaust hole then mately 20D. To investigate the acoustic field appearing impinges on the deflector. This causes the different above the ML, numerical analysis using Computational acoustic waves from the Mach wave radiation from the Fluid Dynamics (CFD) was conducted to understand the free jet. CFD is applied to analyze the hydrodynamic and hydrodynamic and acoustic field generated by impinge- acoustic field generated when a supersonic cold jet with ment of a supersonic jet exhausted from a 1/42-scale solid the Mach number of 1.8 impinges on the deflector.[8] Fig. 4 motor on an infinite flat plate with an exhaust hole.[6] shows the result. Same with Fig. 3, the hydrodynamic and

Instantaneous CFD result of the hydrodynamic field acoustic field is visualized by , and ∆, visualized by density  normalized by ambient density , respectively. The deflector studied here was initially and acoustic field by pressure fluctuation ∆ nor- inclined at 45 deg then smoothly connected to the ground. malized by ambient pressure  are shown in Fig. 3(a). In The distance from the nozzle exit to the deflector was set to this result, the diameter of the exhaust hole and height of 5D. Two types of the acoustic waves can be seen in Fig. 4.

The Journal of the Acoustical Society of Korea Vol.39, No.4 (2020) 334 Seiji Tsutsumi

One propagates in all directions from the impingement Empirical method point, while the other is generated from the wall jet and CFD1 Simplified model test propagates obliquely downstream. From the directivity (limited frequency) angle, the latter is probably the Mach wave generated by Flight data the large-scale vortex structure of the shear layer, similar Preliminary design to that of the free jet. While the former is thought to be CFD2 generated by the interaction between the vortex structure Subscale model test (high fidelity) of the jet shear layer and a plate shock wave appeared at the jet impingement region, the interaction of the vortex Detailed design structure with tail shock waves generated inside the wall Fig. 5. Procedure of acoustic design. jet also contributes to the former acoustic wave.[9-12] However, the cause of the waves due to jet impingement preliminary and detailed designs. Conventionally, the has not been clarified, and further research is necessary. At preliminary stage of design uses the empirical method, a the launch pad shown in Fig. 2, the acoustic waves simplified model test, and past flight data if available. The generated due to the jet impingement propagate in all detailed design mainly involves a subscale model test. directions then are shielded by the ML located above the The empirical method described in NASA SP-8072[4] is deflector. Therefore, the acoustic waves due to the jet used in the preliminary design. This method was based on impingement do not propagate directly to the launch results obtained from static-firing tests and flights. It can vehicle. Together with the Mach waves radiated from the predict the SPL at the observation point easily, but this wall jet, the acoustic waves due to the jet impingement method is applicable mainly to the Mach wave radiation propagate to the launch vehicle through the exit of the from the free jet. The other source of acoustics described in flame duct. Section II were modeled by changing the acoustic As seen above, the acoustic generation mechanism efficiency; their generation mechanisms were not con- changes with the elevation of the launch vehicle. Panda et sidered. (The acoustic efficiency is defined as the ratio of al. obtained a similar conclusion from visualizing the the sound power to the mechanical power of the exhaust acoustic sources of an actual flight by the beam-forming jet.) Many studies have been made to improve the technique.[13] In designing the launch pad, it is necessary to prediction accuracy in the framework of the empirical consider the effect of acoustic sources that change with the method.[5,14-16] elevation of the launch vehicle. In preliminary design, a simplified model test can also be used instead of the empirical method. The MARTEL III. Acoustic design methods test stand of the Centre National d'Etudes Spatiales (CNES) is a representative type. Acoustic testing using over- In the development of launch vehicles, design re- expanded hot jets having an exit Mach number of about 3 quirement for the Sound Pressure Level (SPL) inside the can be done in a semi-anechoic test stand. Gaseous payload fairing is determined. Then the SPL outside the hydrogen and air are employed. Detailed specification of payload fairing is determined, considering the trans- the jets is given in the references.[17,18] Although an mission loss due to the payload fairing. The launch pad is acoustic environment similar to the full-scale launcher is designed to satisfy the SPL outside the payload fairing. not possible using the MARTEL test stand, acoustic The acoustic design procedure for the launch pad is testing can easily be done repeatedly. Parametric studies shown in Fig. 5. The launch pad design is divided into for the launch pad configuration including water injection

한국음향학회지 제39권 제4호 (2020) Design of launch pad for mitigating acoustic loads on launch vehicle at liftoff 335

system for Ariane 5, VEGA, and Ariane 6 have been done Table 1. CFD methods used in development of Epsilon so far.[17,19,20] solid launcher. The launch pad used in the past has been modified and CFD1 CFD2 reused without need for developing a new one in some Method Implicit LES Zonal RANS/LES 6th order Space 2nd order cases. In such case, the preliminary design (very (2nd order for discontinuity) occasionally, the detailed design, too) is performed based Time 2nd order implicit / explicit on past flight data. Viscosity 2nd order central difference Subscale model testing is generally done in detailed RANS: Spalart-Allmaras Turbulence - design. In the development of launch vehicles such as the LES: Space Shuttle, Ares I, Space Launch System (SLS), Ariane Standard Smagorinsky 5, and VEGA, subscale model test having 1/15 to 1/20 [7,19,21-24] scale have been performed. model tests, to design a water injection system.[7,17,20,22,26-28] The development of the Space Shuttle used a 6.4 The mechanism of acoustic attenuation by water injection %-scale model test. Assuming that the same nozzle exit has not been clarified. In the subscale model test, the effect velocity is obtained between the subscale and full-scale of the scale size on the attenuation level should be engines, the acoustic power W and frequency f are carefully examined.[29] expressed as follows using scale factor S. With the improvement of CFD technology and the computing performance increase of High Performance      (1) Computers (HPCs), improved CFD-based design has become available. As mentioned above, the empirical

   (2) model proposed by NASA SP-8072 is based on the Mach wave radiation from the free jet, and the limitation in the The subscripts full and sub represent the full scale and range of application is clear. As shown on the right side of subscale, respectively. Using above equations, the flight Fig. 5, CFD can be used both in the preliminary and data of the Space Shuttle was compared with the subscale detailed designs. The CFD methods used for the test results. An accuracy of roughly 5 dB in 1/3 octave preliminary and detailed designs must be different. They band Sound Pressure Level (SPL) was achieved.[21] In the are named as CFD1 and CFD2, respectively, here. Table 1 development of the Ares I, a 1/20-scale model test was shows an example of CFD methods used in the conducted. Very good prediction accuracy of approxi- development of the Epsilon.[30,31] Parametric study for the mately 2 dB in OASPL was achieved.[22] launch pad configuration is done in the preliminary design, In addition to predicting the full-scale acoustic level, the so the need for CFD1 is throughput rather than fidelity. acoustic loads also needed to be reduced to satisfy design Therefore, implicit Large-Eddy Simulation (LES) con- requirement. In the development of the Ariane 5 and sidering only the numerical viscosity originating in grid VEGA, experiment was conducted at MARTEL test stand resolution was employed. Second-order accuracy was to determine whether the flame duct should be covered or achieved for the convective and viscous terms and for time uncovered to reduce the acoustic level.[17,20] In many integration. The cut-off frequency of CFD1 was set to 150 launch pads, a large amount of water, several times the Hz, which is not enough to analyze the acoustic total flow rate of the engine exhaust jets, is injected for environment of the launch vehicle. However, using 3072 acoustic attenuation.[25] Many studies have been cores of the former JAXA’s HPC, JSS (Fujitsu FX1, 31 conducted, using simplified model tests and subscale TFlops) operated from 2009 to 2014, results could be

The Journal of the Acoustical Society of Korea Vol.39, No.4 (2020) 336 Seiji Tsutsumi obtained within 3 days per case. CFD2 used in the detailed design requires fidelity rather than throughput. The zonal hybrid LES / Reynolds-averaged Navier-Stokes (RANS) Hydraulic cylinder method was employed. The turbulent boundary layer on the deflector was computed by RANS, and off-wall region Load cell Farfield mics. were computed by LES. Convective terms were evaluated Microphones th Solid motor by a high-resolution method with 6 order accuracy, and Mic.W Mic.E the cut-off frequency reached 800 Hz using 2048 cores of N Mic.S 50D H E W the JSS (20 TFlops) for 2 to 5 months. Validation of CFD2 S Launch pad was performed using the results of the subscale model test. 50D Prediction accuracy of approximately 4 dB in OASPL was N E [30,32] confirmed. Note that JAXA’s HPC is expected to be W S renewed this fiscal year, and will achieve higher com- Fig. 6. Schematic of test stand used in detailed design puting performance by 144 times than the former JSS. for Epsilon solid launcher. CFD analyses will be performed in a shorter time. CFD has site was dug into the scale of tens of meters to create the also been used in the development of the H3 launch vehicle flame duct. This result violated the other design re- under development in Japan, and in the development of quirement to reuse the existing launch pad for the M-V as VEGA.[6,33,34] much as possible. The application range of the empirical method was known to be limited, so CFD methods listed in IV. Launch pad design for Table 1 were used heavily in both the preliminary and epsilon solid launcher detailed designs. The detailed design also used subscale model testing. Considering the similarity of acoustic Design requirements imposed at the beginning of the spectra, a 1/42-scale solid motor was employed in that test. Epsilon development were that: (1) the acoustic level The nozzle exit velocity was designed to be the same as the inside and outside the payload fairing would be 140 dB and full-scale motor, and the Reynolds number based on the 147 dB in OASPL, respectively, and (2) the existing nozzle exit conditions was 2.2 × 105. The propellant of the launch pad for the M-V in Uchinoura Space Center (USC) 1/42-scale motor had almost the same properties as the would be utilized as much as possible. Since the thrust at full-scale motor, so that the acoustic power Wfull and liftoff was 3,700 kN for the M-V and 2,000 kN for the frequency ffull were estimated by Eqs. (1) and (2). The Epsilon, the acoustic power generated at the exhaust jet schematic of a test stand is shown in Fig. 6. The 1/42-scale was estimated to be 3 dB lower for the Epsilon. The solid motor was hung from the top of the test stand via a OASPL at the outside of the payload fairing was 157.6 dB load cell for measuring thrust and a hydraulic cylinder for for the M-V. Even considering the difference in the engine, changing the elevation. The subscale launch pad was a reduction of 8 dB was required for the launch pad design. located below the solid motor. Three microphones were mounted around the payload fairing. Additional micro- 4.1 Methods phones were located 60D from the jet impingement point At the beginning, the empirical method described in to obtain validation data for CFD and to investigate NASA SP-8072 was employed. The required acoustic acoustic feature of the launch pad. Sufficient repeatability level could not be satisfied unless the cliff of the launch was confirmed in advance. The measurement error of the

한국음향학회지 제39권 제4호 (2020) Design of launch pad for mitigating acoustic loads on launch vehicle at liftoff 337 microphones was estimated to be 2 dB. Details of the shown here, but three strategies for acoustic attenuation subscale test can be found in the reference.[30] are explained below.

4.2 Strategies for acoustic-load mitigation 4.2.1 Position and shape of launch mount As described in Section II, the dominant acoustic waves As mentioned in Section II, when the vehicle is just after liftoff are generated by the jet impingement on elevated, the Mach wave radiated from the free jet is the deflector and the Mach wave radiation from the wall reflected from the ground or launch complex, and then jet. As the launch vehicle is elevated, the dominant sources returns back to the launch vehicle, which imposes an of acoustics change to the reflection of the Mach wave acoustic load. To increase the propagation distance of the from the free jet and the interaction between the free jet Mach wave and reduce its effect on the launch vehicle, the and the ML. The acoustic sources that change according to launch mount was determined to be located as high as the vehicle’s elevation should be taken into account in the possible from the ground. The Epsilon was planned to be design of launch pads. The baseline configurations for the assembled in the existing vehicle assembly building. The preliminary and detailed designs are shown in Fig. 7. Fig. height of the launch mount was constrained by the height 7(a) is based on the launch pad for the Minotaur 1 at the of the building. Wallops Flight Facility of NASA. On this type of launch Additional consideration for designing the launch pad, the launch vehicle is mounted at a certain height mount was to minimize the interaction of the jet with the above the ground. As described in Section 4.2.1, the effect launch mount, because the interaction generates the of the Mach wave radiation from the free jet can be acoustic waves. In the Epsilon, the projected area of the reduced by raising the launch mount. Based on the CFD launch mount was minimized, and the launch mount was results and structural design, the shape of the launch pad in tapered to reduce the interaction. the detailed design was determined by considering the installation on the existing launch pad. Finally, the 4.2.2 Deflector shape configuration shown in Fig 7(b) was obtained as the Just after liftoff, the dominant acoustic waves are baseline for detailed design. Due to space limitations, not generated by the impingement of the exhaust jet on the all the results of the preliminary and detailed designs are deflector, as discussed in Section II. The acoustic waves due to this are related to the strength of the plate shock wave. As the mounting angle of the deflector to the ground increases, the strength of the plate shock wave weakens, so the acoustic waves due to jet impingement can be reduced.[35,36] The tail shock waves are also a source of acoustics. By increasing the curvature of the deflector and Ve r t i c a l flame gradually changing the flow direction, the tail shock waves duct are weakened, and eventually the acoustic level can be reduced.[8,35] In the preliminary design, the effect on the acoustic level was evaluated by CFD for three explanatory variables: (1) the height of the launch mount, (2) the initial inclination angle of the deflector, and (3) the curvature of (a) Preliminary design (b) Detailed design the deflector. About twenty deflector shapes were Fig. 7. Baseline configuration of launch pad for Epsilon evaluated. solid launcher.

The Journal of the Acoustical Society of Korea Vol.39, No.4 (2020) 338 Seiji Tsutsumi

-0.15 0.15 Launch mount and vertical flame duct 0.0 Mach number 4.0

Deflector

Ground flame duct

Fig. 9. (Color available online) Launch pad for Epsilon solid launcher.

The launch pad for the Epsilon built at USC is shown in Fig. 9. The launch mount and the vertical flame duct were attached to a boom that carries the launch vehicle from the vehicle assembly building to the launch pad. The ground Fig. 8. (Color available online) CFD result of baseline shape for detailed design. Elevation of launch vehicle flame duct was constructed on the ground. A triangular is 14D. The Mach number plot is shown only for the roof was attached to the ground flame duct to reduce the supersonic region. reflection of the Mach waves radiated from the free jet. 4.2.3 Flame duct The deflector inside the ground flame duct was designed The vertical flame duct of the baseline shapes shown in based on the CFD results obtained in the preliminary Fig. 7 had four side walls, but the side wall facing the design. The launch pad of the Epsilon does not have a large deflector outlet was fully open. Fig. 8 shows the CFD underground flame duct or water injection system, result for Fig. 7(b). The Mach waves radiated from the jet resulting in a compact launch pad compared to others in flowing into the vertical flame duct propagate outward the world. from open side, reflect off the ground, and head back to the launch vehicle. Based on this, a front cover was installed 4.3 Post-Flight Analysis on the open side to shield against the Mach waves. As a In the Epsilon’s first flight, microphones were installed result, reduction of OASPL by 3 dB at the payload fairing both inside and outside the payload fairing to validate the was obtained in the subscale model test when the elevation launch pad design. Results of two microphones attached of the launch vehicle was 14D. On the other hand, further inside the payload fairing are shown in Fig. 10(a). reduction was required at low elevations to satisfy the Although there are differences in the two measurements design requirement. Therefore, a ground flame duct was due to the different locations, those results satisfied the designed to shield the acoustic waves due to the jet design requirements at all frequencies. The launch pad impinging on the deflector and the Mach waves radiated designed for the Epsilon was compact without an from the wall jet. As a result, reduction by 3 dB in OASPL underground flame duct or water injection system, but the at the payload fairing was also obtained in the subscale flight results demonstrated a quieter acoustic environment model test, when the elevation of the launch vehicle was than other small- and medium-sized launch vehicles. Fig. 8D. 10(b) compares the measurements by the microphones

한국음향학회지 제39권 제4호 (2020) Design of launch pad for mitigating acoustic loads on launch vehicle at liftoff 339

mic.1 (132.0dB) requirement (140dB) mechanism of acoustic waves generated by launch mic.2 (132.6dB) 140 vehicles at liftoff, as well as introducing acoustic design 135 methods for the launch pad. Then, as an example, 130 125 development of the launch pad for the Japanese Epsilon 120 solid launcher is outlined. The launch pad for the Epsilon 115 110 was designed by using CFD together with subscale model

1/1 octave band SPL, dB 105 100 testing. Through the post-flight analysis, it was found that 10 100 1000 reduction of the acoustic level by 10 dB compared to the frequency, Hz (a) Inside payload fairing previous M-V launcher was achieved, in addition to

mic.1 (143.6dB) M-V (157.6dB) satisfying the design requirement. mic.2 (144.4dB) requirement (147.5dB) 155 The launch vehicle is an expensive transportation 150 system that cannot fail. Cost and reliability are often the 145 140 main development goals. Reducing the acoustic loads on 135 the satellite leads to the reduction of satellite development 130 cost. Therefore, reducing the acoustic loads at liftoff is also

1/1 octave band SPL, dB 125 120 an important design issue that increases the value of the 10 100 1000 10000 frequency, Hz launch vehicle. (b) Outside payload fairing It is also a fact that mechanisms of acoustic waves Fig. 10. Microphone measurements of Epsilon’s first generated at liftoff of a launch vehicle are not sufficiently flight. OASPL of each data is shown in legends. understood yet. Study of acoustic generation is still an attached at different locations outside the payload fairing interesting topic to academic researchers. In addition, there with the design requirements and the measurement of the is a possibility for engineers to achieve further reduction of M-V. The results of the Epsilon satisfied the design the acoustic loads by modifying the shape of the launch requirement in all frequencies. In addition, the results were pad. I hope this article will help to promote cooperation significantly lower than that of the M-V. OASPL between Korea and Japan in the research and development measured outside the payload fairing of the Epsilon was at of liftoff acoustics. most 144.4 dB, which means that the acoustic environment was reduced by about 13 dB compared with the M-V. Acknowledgements Although the difference of the acoustic level due to the motor size was 3 dB as mentioned earlier, the acoustic The author wishes to thank Dr. Kyoichi Ui (JAXA) and level could be reduced by 10 dB through the improvement Dr. Tatsuya Ishii (JAXA) for their help to the research and of the launch pad. The flight data usually have large development of the launch pad for the Epsilon solid variations, but the results of four Epsilon’s flights so far launcher. did not show remarkable variations. Therefore, the effectiveness of the launch pad designed using CFD was References confirmed. 1. S. J. Archer, “Structural vibration prediction,” NASA V. Summary Tech. Rep., SP-8050, 1970. 2. J. Onoda and K. Minesugi, “Estimation of mechanical environment of M-V satellite launcher,” Proc. JSASS/ This article has presented the latest information of the JSME Structures Conference, 229-232 (1997).

The Journal of the Acoustical Society of Korea Vol.39, No.4 (2020) 340 Seiji Tsutsumi

3. C. K. W. Tam, “Mach wave radiation from high-speed application to the ariane 5 launch vehicle,” AIAA jets,” AIAA J. 47, 2440-2448 (2009). Paper 200-2026 (2000). 4. K. M. Eldred, “Acoustic loads generated by the pro- 18. N. Cacqueray, C. Bogey, and C. Bailly, “Investigation pulsion system,” NASA Tech. Rep., SP-8072, 1971. of a high-mach-number overexpanded jet using large- 5. J. Varnier, “Experimental study and simulation of eddy simulation,” AIAA J. 49, 2171-2182 (2011). rocket engine freejet noise,” AIAA J. 39, 1851-1859 19. J. Varnier, J. F. Piet, D. Gély, G. Elias, and S. (2001). Radulovic, “Modeling of the acoustic environment on 6. S. Tsutsumi, R. Takaki, H. Ikaida, and K. Terashima, the ariane 5 fairing using small scale test data,” AIAA “Numerical aeroacoustics analysis of a scaled solid jet Paper 1996-1721 (1996). impinging on flat plate with exhaust hole,” 30th 20. D. Gély, G. Elias, F. Mascanzoni, and H. Foulon, International Symposium on Space Technology and “Acoustic environment of the vega launch vehicle at Science, 2015-o-2-05 (2015). lift-off,” Proc. 4th European Congress on Acoustics, 7. J. Panda and R. Mosher, “Microphone phased array to 821-826 (2005). identify liftoff noise sources in model-scale tests,” J. 21. N. Dougherty and S. Guest, “A correlation of scale Spacecraft Rockets, 50, 1002-1012 (2013). model and flight aeroacoustic data for the space 8. S. Tsutsumi, R. Takaki, Y. Nakanishi, K. Okamoto, shuttle vehicle,” AIAA Paper 1984-2351 (1984). and S. Teramoto, “Acoustic generation mechanism of 22. D. D. Counter and J. Houston, “Ares I scale model a supersonic jet impinging on deflectors,” AIAA Paper acoustic test lift-off acoustics,” J. Acoust. Soc. Am. 2014-0882 (2014). 130, 2542 (2011). 9. T. Nonomura, Y. Goto, and K. Fujii, “Aeroacoustic 23. D. Casalino, S. Santini, M. Genito, and V. Ferrara, waves generated from a supersonic jet impinging on “Rocket noise sources localization through a tailored an inclined flat plate,” Int. J. Aeroacoustics, 10, 401- beam-forming technique” AIAA J. 50, 2146-2158 425 (2011). (2012). 10. M. Akamine, Y. Nakanishi, K. Okamoto, S. Teramoto, 24. J. Houston, D. Counter, and C. Giacomoni, “SLS scale T. Okunuki, and S. Tsutsumi, “Acoustic phenomena model acoustic test liftoff results and comparisons,” from correctly expanded supersonic jet impinging on 29th Aerospace Testing Seminar (2015). inclined plate,” AIAA J. 53, 2061-2067 (2015). 25. H. Himelblau, J. E. Manning, A. G. Piersol, and S. 11. C. Brehm, J. A. Housman, and C. C. Kiris, “Noise Rubin, “Dynamic environmental criteria,” NASA Tech. generation mechanisms for a supersonic jet impinging Rep., HDBK-7005, 2001. on an inclined plate,” J. Fluid Mech. 797, 802-850 26. J. K. Ignatius, S. Sathiyavageeswaran, and S. R. (2016). Chakravarthy, “Hot-flow simulation of aeroacoustics 12. M. Kurokawa, S. Teramoto, and K. Okamoto, “Acoustic and suppression by water injection during rocket lift- wave generation from two-dimensional supersonic off,” AIAA Journal, 53, 235-245 (2015). inviscid jet impinging on inclined plate,” AIAA J., 27. S. Sankaran, J. K. Ignatius, R. Ramkumar, T. N. V. (2020). Satyanarayana, S. R. Chakravarthy, and N. R. 13. J. Panda, R. N. Mosher, and B. J. Porter, “Noise source Panchapakesan, “Suppression of high mach number identification during rocket engine test firings and a rocket jet noise by water injection,” J. Spacecraft rocket launch,” J. Spacecraft Rockets, 51, 1761-1772 Rockets, 46, 1164-1170 (2009). (2014). 28. W. Sarae, K. Terashima, S. Tsutsumi, M. Takegoshi, 14. J. Varnier and W. Raguenet, “Experimental charac- H. Kobayashi, and A. Mori, “Summary of acoustic terization of the sound power radiated by impinging design for H3 launch complex,” J. Acoust. Soc. Am. supersonic jets,” AIAA J. 40, 825-831 (2002). 146, 3001 (2019). 15. V. V. Koudriavtsev, and A. V. Safronov, “Noise ge- 29. T. Shimizu, M. Hirai, S. Tsutsumi, R. Takaki, and M. neration at supersonic jet interaction with inclined Arita, “Study on attenuation of sound by droplets,” J. deflector,” Int. Congress and Exhibition on Noise Japan Society for Aeronautical and Space Sciences, Control Engineering, 139-142 (2001). 57, 71-76 (2009). 16. J. Haynes and R. J. Kenny, “Modifications to the 30. S. Tsutsumi, T. Ishii, K. Ui, S. Tokudome, and K. NASA SP-8072 distributed source method II for ares I Wada, “Study on acoustic prediction and reduction of lift-off environment predictions,” AIAA Paper 2009- epsilon launch vehicle at liftoff,” J. Spacecraft Rockets, 3160 (2009). 52, 350-361 (2015). 17. D. Gély, G. Elias, C. Bresson, H. Foulon, S. Radulovic, 31. S. Tsutsumi, T. Ishii, K. Ui, S. Tokudome, and K. and P. Roux, “Reduction of supersonic jet noise. Wada, “Assessing prediction and reduction technique

한국음향학회지 제39권 제4호 (2020) Design of launch pad for mitigating acoustic loads on launch vehicle at liftoff 341

of lift-off acoustics using epsilon flight data,” AIAA Paper 2015-1007 (2015). 32. S. Tsutsumi and K. Terashima, “Validation and veri- fication of a numerical prediction method for lift-off acoustics of launch vehicles,” Trans. Japan Society for Aeronautical and Space Sciences, Aerospace Technology Japan, 15, 7-12 (2017). 33. S. Tsutsumi, W. Sarae, and K. Terashima, “Effect of rocket engine clustering on acoustic level of H3 launch vehicle at lift-off,” Proc. 32th International Symposium on Space Technology and Science, 2019-g-12 (2019). 34. D. Palmieri, D. Nicolini, P. M. Roviera, C. Kiris, M. Barad, B. Vu, and D. Chesnutt, “Design and validation of VEGA launch pad modifications to reduce payload acoustic environment at lift-off,” Proc. 68th Interna- tional Astronautical Congress, IAC-17-D2.2.2 (2017). 35. S. Tsutsumi, S. Kato, K. Fukuda, R. Takaki, and K. Ui, “Effect of deflector shape on acoustic field of launch vehicle at lift-off,” AIAA Paper 2009-0328 (2009). 36. T. Nonomura, H. Honda, Y. Nagata, M. Yamamoto, S. Morizawa, S. Obayashi, and K. Fujii, “Plate-angle effects on acoustic waves from supersonic jets impinging on inclined plates,” AIAA J. 54, 816-827 (2016).

Profile

▸Seiji Tsutsumi He graduat ed wi t h a DEng f rom Depart ment of Aeronautics and Astronautics, University of Tokyo in 2006. He has been working for JAXA si nce t hen, where he is currentl y an associate senior researcher. His research interest includes liftoff acoustics, high speed aerodynamics, CFD technique using HPC, and application of machine learning to analyze fluid phenomena.

The Journal of the Acoustical Society of Korea Vol.39, No.4 (2020)