1

Flow Separation Control and Lift-to-Drag Ratio Improvement Using Repetitive Laser Pulses

IAPS 2019Double Tree by Hilton, Naha Shuri Castle

1) Department of Aerospace Engineering, Tohoku University2) Department of Aerospace Engineering, Nagoya University 3) School of Engineering, University of Glasgow

Masayuki Takahashi1, Manabu Myokan2, Syouya Kubota2, Francesca Gnani3, Henny Bottini2, Akira Iwakawa2, Naofumi Ohnishi1, Akihiro Sasoh2

March 9, 2019

Flow Separation During Supersonic Flight

2

• Oblique shock wave at the trailing edge forms an inverse pressure gradient on the wing surface because positive pressure of the oblique shock wave propagates toward the upstream in the inside of the subsonic boundary layer

• Flow separation is induced even if AoA is small because of the oblique shock wave, which decreases a lift-to-drag ratio (L/D) of the airfoil

oblique shock wave expansion wave oblique shock wave

diamond wing

separation region@P

@x> 0

uniform flow

Conventional Techniques Based on Perturbation Generation

3

• Dielectric-barrier-discharge plasma actuator (PA) was proposed for flow control, but a flow control force was insufficient for supersonic flow

• Interaction duration between flowfield and perturbation generated by PA or laser was too short when speed of the main stream is fast

• It is necessary to propose new concept to control high-speed flow

A. Iwakawa et al. (2016)

Aerospace 2016, 3, 13 5 of 13

suppressed, and the foot of the shock wave is clearer than in Figure 4b. This modulation is the effect ofthe repetitive energy deposition already mentioned in [16].!"#$%&'(")!"#$!"*!"#$" %"&'"#$"

%&'()*&+,

-./0123456()*&+,

!

"

748('9&(6:65;<(=6>&)?@?&: 7+8

7+8

7%8

7%8#"(AA B(AA

B(AA

23456()*&+,

7=8('9(C"(,DE(6:65;<(=6>&)?@?&:

728

768728

%F%%36

7%8768#"(AA B(AA

B(AA"

G?;F56( HI" ()*+,-." '.&/" 0123,1324" &'" 154"6&7+8-."7&94." &:1-+849" :)" '2-7+8;" 154" <,5.+4248" +7-;4=" ">4?" /+15&31" 4842;)" 94*&0+1+&8@" >%?" ,.&04A3*" -2&389" 154" '.-24" 04,1+&8" /+15&31" 4842;)" 94*&0+1+&8@" ">+?" ,.&04A3*" -2&389" 154" 01-;8-1+&8" *&+81"/+15&31" 4842;)" 94*&0+1+&8@" >=?" +4" B" $C" DEF@" >6?" ,.&04A3*"-2&389"154"'.-24"04,1+&8"/+15"+4"B"$C"DEF@">2?",.&04A3*"-2&389"154"01-;8-1+&8"*&+81"/+15"+4"B"$C"DEFG"

*,-,)."'/)01'#")012(32'34$5)

(54"0,547-1+,"+..3012-1+&8"-89"154"08-*05&10"&'"154"<,5.+4248"+7-;4"&'"154"<5-..&/"7&94."-24"05&/8"+8"H+;324"%-!:!"240*4,1+I4.)G"H2&7"H+;324"%:!"4I48"15&3;5"154"'.-24"05&,D"01-890"+8"'2&81"&'"154" '.-24" 04,1+&8" -89" 042+&30" :&389-2)A.-)42" 04*-2-1+&8" 9&40" 8&1" &,,32!" 154" '.-24" 05&,D" &0,+..-140"&/+8;" 1&" 154" '.3,13-1+&8" +893,49" :)" 154" :&389-2)A.-)42" +8142-,1+&8" -1" 154" 05&,D" '&&1G" (5+0"&0,+..-1+&8"'24J348,)"+0"24.-1+I4.)".&/"-89"154"&0,+..-1+&8"-7*.+1394"+0"07-..G"

H+;324"K"05&/0"154"*&/42"0*4,1237"9480+1)"9+012+:31+&8"-.&8;"#L1"B"CGMN">#"B"OGM"77?"/+15&31"-89"/+15"4842;)"94*&0+1+&8G"(54"5&2+F&81-."-P+0"6",&2240*&890"1&"154"5&2+F&81-."-P+0"&'"H+;324"%:!"-89"154"0*4,1237"'24J348,)"&'"154"I421+,-."-P+0",&2240*&890"1&"154"5&2+F&81-."-P+0"&'"H+;324"$:G"(54",&.&2",&81&320",&2240*&89"1&"154"0*4,1237"9480+1)G"H2&7"H+;324"K-!"154"<5-..&/"7&94."5-0"-"012&8;"0*4,1237" +8" 154" #Q$" DEF" 2-8;4" &/+8;" 1&" 154" '.-24" 05&,D" '.3,13-1+&8G" (54" 012&8;401" 0*4,1237" +0"

Figure 4. Typical flow structure of the Nominal model obtained by framing the Schlieren image:(a) without energy deposition; (b) close-up around the flare section without energy deposition;(c) close-up around the stagnation point without energy deposition; (d) f e = 30 kHz; (e) close-uparound the flare section with f e = 30 kHz; (f) close-up around the stagnation point with f e = 30 kHz.

3.1. Weak Flare Fluctuation

The schematic illustration and the snapshots of the Schlieren image of the Shallow model areshown in Figure 5a,b, respectively. From Figure 5b, even though the flare shock stands in front of theflare section and serious boundary-layer separation does not occur, the flare shock oscillates owing tothe fluctuation induced by the boundary-layer interaction at the shock foot. This oscillation frequencyis relatively low and the oscillation amplitude is small.

Figure 6 shows the power spectrum density distribution along r/l = 0.42 (r = 8.4 mm) withoutand with energy deposition. The horizontal axis z corresponds to the horizontal axis of Figure 5b, and

focal point

second cone

diamond wing

plasma actuator or laser pulse

4

https://response.jp/imgs/zoom1/921821.jpg

beam source

New Flow Control Concept Proposed by Us

• Expansion wave is induced from the trailing edge when the blast wave propagates to the upper surface, which interacted with the separation

• A source of the repetitive pulses is equipped on the aircraft• After a computational fluid dynamics (CFD) simulation, we show the

experimental results using the repetitive pulses in this presentation

expansion occurs if the blast wave jet is supersonic

blast wave

wing

interaction with separation

Objective of Our Study

5

• Phase 1: Checking feasibility of our flow control concept using CFD

• Phase 2: Proving our concept by conducting a wind-tunnel experiment

• Phase 3: Conducting actual flight test by equipping the repetitive pulses

Establishing new flow control concept based on propagation of the blast wave

Governing Equation and Numerical Methods

6

• Governing equation　　2D compressible Navier-Stokes equation

• Spatial discritization cell-centered finite volume method

• Numerical flux AUSM-DV with 2nd-order MUSCL method

• Viscous flux 2nd-order central difference method

• Time integration 1st-order Euler explicit method

• Turbulence model N/A to examine pure shock wave dynamics

• Chemical reaction N/A to examine pure shock wave dynamics

繰り返しパルス照射による剥離流れ制御の数値的研究⃝高橋聖幸,大西直文 (東北大工)

Numerical Study for Separation-Flow Control by Irradiating Repetitive Pulses

Masayuki Takahashi and Naofumi Ohnishi

Department of Aerospace Engineering, Tohoku University, Sendai 980-8579, JAPAN

Abstract

Contactless flow-control device was proposed based on an energy-deposition technique using the repetitivelaser pulses, which was numerically verified by integrating the two-dimensional Navier-Stokes equations.Separation flow was induced under an uniform flow having Mach number of 1.7 because of an adversepressure gradient caused by an oblique shock wave on the upper surface of a diamond wing. Thicknessof the separation region became thinner when the repetitive pulses were irradiated onto the under surfaceof the diamond wing because interactions between an expansion wave and the sepration region relaxed theadverse pressure gradient on the upper surface. A better lift-to drag ratio was obtained when the blast waveinteracted with the boundary layer because a Cp profile was improved on the upper surface. Our simulationresults indicated an effectiveness of the proposed contractless flow-control device based on interactionsbetween the boundary layer and the blast wave induced by the repetitive pulses.

Keywords: Repetitive pulses, Separation flow, Contactless flow control, Blast wave-boudary layerinteractions

1. 緒言高迎角時に生ずる翼面上の剥離は，失速を招く為に防がねばならぬ現象である．従来はボルティックスジェネレータを使用する事で剥離抑制が試みられてきたが，ボルティックスジェネレータを使用時は翼表面積が増え，摩擦抗力が増大する事が問題となる．また，低迎角で飛行している際は剥離を生じ難い為にボルティックスジェネレータは不必要であるが，飛行中の状態に応じて着脱する事が出来ないという難点がある．そこで，飛行機翼の剥離流制御の為の新たなデバイスとして誘電体バリア放電を用いたプラズマアクチュエータが提案されている(1)．プラズマアクチュエータは，被覆電極と露出電極の 2つを用いてイオン風を誘起し，流れ場に運動量を与える事で剥離流を制御する．或は，パルス状の電圧を印加する事でブラスト波を誘起しつつ擾乱を生成し，乱流化を促す事で再付着を促進する(2)．従来のボルティックスジェネレータとは異なり，翼の摩擦抗力を増やす事無く翼性能を改善出来，さらに剥離を生じ易い高迎角時のみの作動が可能であり，注目されている．しかし過去の数値計算や実験によると，プラズマアクチュエータは流れ場に与える事の出来る体積力が小さく，10の 6乗を超える高レイノルズ数流れでは有効ではない．印加電圧を増加すればより大きな体積力を獲得可能だと予期出来るが，電極や翼へのダメージが避けられず，印加可能電圧が制限される．高密度プラズマが電極に触れる事で機器の損耗が生ずる為，プラズマが翼や電極と非接触である事が望ましい．本研究では，レーザーパルス照射により駆動されるブラスト波に注目して，剥離流を制御する新たな非接

触気流制御デバイスを提案する．本手法では，翼面上に設置した電極を用いずに翼面と非接触にプラズマを生成出来る為，荷電粒子が翼壁面に衝突する際に生ずるダメージを軽減出来る．テスト問題として超音速気流中にダイアモンド翼を設置して剥離流を発生させ，翼周辺に繰り返しパルス照射によるエネルギー投入を行う事で，提案した非接触流れ場制御法の有効性を検証する．

2. 流れ場の数値計算法と計算条件2次元 Navier-Stokes方程式を数値的に解く事で，剥離流と繰り返しパルス照射時のブラスト波-剥離流相互作用を再現する．

∂Q∂t+∂(E − E)∂x

+∂(F − F)∂y

= S， (1)

ここで，Q は保存量ベクトル，E と F は x 及び y 方向成分の非粘性流束ベクトル，E と F は x 及び y 方向成分の粘性流束ベクトル，Sはパルス照射のソース項である．支配方程式の離散化は，セル中心有限体積法によって行う．数値流束は AUSM-DV スキームを用いて評価し，空間精度は minmod リミッターを用いた MUSCL 法により 2 次精度化した．高次精度化の際に使用する値としては特性量が選択される．粘性項は 2 次精度中心差分を用いて計算し，粘性係数はSutherlandの式を用いて評価する．また，本研究では乱流モデルは使用しない．時間積分は 1 次精度 Euler陽解法を用いる．本研究ではダイアモンド翼を計算対象とし，Fig. 1に示す 2次元格子を用いる．境界条件設定の複雑さを避けるために，マルチブロック法を使用して流れ場を

Simulation Conditions

7

• Flow Mach number 1.7• Ambient gas 20 km param. of air• Angle of attack 2 degree• Reynolds number • Airfoil diamond wing (t/c=0.2)• Minimal grid size

Flow condition (D. Maruyama et al. (2007))

Pulse laser condition (A. Iwakawa et al. (2016))diamond wing

beam’s focus point

• Pulse energy 50 mJ/m• Pulse frequency 80 kHz• Focal point lower surface

1/25 · (5Lref)/pRe

1.26⇥ 105

8

• Separation region on the upper surface became smaller when the repetitive pulses were irradiated to the lower surface

Separation Flow Control by Repetitive Pulses

0.15

0.151

0.152

0.153

0.154

0.155

-100 0 100 200 300 400 500 600y

[m]

flow velocity in x-direction [m/s]

u w/o pulsesu- w/ 50 mJ/m, 80 kHz

w/o pulses

w/ 50 mJ/m, 80 kHzy

y

time-averaged value

time-averaged value

beam’s focus point

Mechanism of Separation Control Using Repetitive Pulses

9

• Prantdle-Mayer expansion occurred at the trailing edge when the blast wave with supersonic speed propagated from the lower to upper surface

• The inverse pressure gradient on the upper surface was relaxed when the expansion wave at the trailing edge interacted with the separation region

expansion occurs if the blast wave jet is supersonic

blast wave

wing

interaction with separation

strong expansionbeam’s focus point

instantaneous value

Summary

10

• CFD simulation was conducted to check a feasibility of flow control method using the repetitive laser pulses

• An adverse pressure gradient was relaxed on the upper surface because an expansion wave induced from the trailing edge when the blast wave propagated from the lower to upper surfaces

• L/D performance was improved by combining the pressure decrease on the upper surface and pressure increase on the lower surface

• The lift increment was experimentally captured when the repetitive pulses were irradiated on the lower surface

• The lift becomes higher with an increase in the laser power