Research Outline

Ducted fans are the only vertical takeoff and landing vehicles capable of efficient hover and cruise at equal power. This work reports on design, wind tunnel tests, and transition simulation of a high-efficiency ducted fan with cyclic pitch actuation designed to use equal power in hover and cruise. The rotor shroud is a ring wing designed for efficient cruise. Conventional duct leading edges are also tested in hover, and with a crosswind. The results of this research may be used to increase the cruise efficiency and pitch control authority of unmanned aerial vehicles, while reducing design constraints imposed by the transition corridor.

In this work, the following hypotheses are tested:

-Airfoil section data may be used with vortex lattice codes to accurately predict the aerodynamic coefficients of a ring wing.

-Use of a ring wing as the rotor shroud of a ducted fan results in high lift to drag ratios and useful lift coefficients at low angle of attack in a near-axial flow flight condition.

-Use of cyclic pitch for moment control is adequate for stabilization in hover and cruise.

Abstract

The ducted fan is unique among VTOL vehicles in that it can cruise and hover efficiently at the hover power setting. A graph of ducted fan power required in cruise as a function of lift coefficient and L:D ratio is shown in Fig. 2.

Experimental wind tunnel data for a new ring wing (shown in Fig. 2) is presented. At a chord Reynolds number of 250,000, the biplane-cambered aspect ratio 3.1 ring wing has a lift to drag ratio greater than 12, developed at 4 degrees angle of attack. High lift to drag ratio is due to the biplane-cambered airfoil stack, shown in Fig. 3. Ring wing lift, drag and moment data are taken from 0 to 90 degrees of the angle of attack. The ring's minimum internal diameter is 600 mm.

The ring wing in use as a duct is shown in Fig. 4. Untapered, untwisted NACA 0015 helicopter blades are used for the rotor. Force and moment data were taken for the ducted fan in hover, with and without a 10 m/s headwind, and in near axial-flow conditions at 15 m/s.

Thrust and power coefficients are presented for the no-headwind hover case, with three different duct lips, a circular lip, a trumpet lip, and the lip of the ring wing tested by itself. Front views of all duct lips are shown in Fig. 5. The interchangeable duct lips are shown in cross section in Fig. 6.

Ducted fan maximum test Reynolds number is 190,000. Rotor collective pitch settings from 0 to 25 degrees, and helicopter-type cyclic pitch settings from 0 to 7.5 degrees were tested. Ducted fan maximum lift to drag ratio with unfaired struts and centerbody was 4. This value could easily be improved to 8 or higher by fairing the struts and centerbody. Cylic pitch response was found to be linear below blade stall. Large moments were produced in hover and cruise, sufficient to stabilize the vehicle and provide maneuvering capability.

Large control moments available from cyclic pitch in hover and near-axial flow conditions may enable dynamic transitions from hover to cruise. The simple, accurate models developed for power-off shroud aerodynamic performance enable UAV design for longer loiter and more efficient cruise flight.

Wind Tunnel Test Overview

Configurations tested in the wind tunnel are shown in Fig. 7. The test range covers crosswind and axial flow for the ring wing and ring wing with rotor and the different duct lips. For the ring wing, intermediate angles of attack are also checked.

Results from Wind Tunnel Testing

A ducted fan with axial asymmetry has been tested for hover performance with a variety of duct lips at low Reynolds numbers. The experimental results provide the first published force and moment data for cyclic pitch-actuated ducts with axial asymmetry.

Aspect ratio 3.1 ring wing polars were accurately predicted using the AVL vortex lattice code on a wing with circumferentially varying camber. Wind tunnel lift and drag coefficients for the tested ring wing are shown in Fig. 8. Lift to drag ratios greater than 12 occur at a CL range of 1-1.2, near 4 degreee angle of attack. Calculated polar and span efficiency factor is compared with the experimental ring wing lift polar in Fig. 9. It can be seen that section methods provide a good trust region for lift coefficient estimation. The CD0 offset for the AVL drag figure is a circumferential average of section drag coefficients at zero angle of attack, multiplied by pi.

Solidity-weighted thrust coefficient versus power coefficient for no-crosswind hover are presented in Fig. 12, with comparison data from NACA TN 626. It can be seen that the trumpet and circular rotor shrouds increase efficiency beyond the induced power theoretical limit, for the trumpet and circular duct lips. In addition, at constant solidity, useful CT range is increased by the use of a duct.

In crosswinds, rotor offloading for the trumpet and circular duct lips is reduced, and pitching moments are increased relative to the airfoil duct lip.

Total recorded moment due to cyclic pitch (changing the angle of attack of the rotor blades as they rotate) is shown in Figs. 10 and 11. Based on hover and crosswind hover wind tunnel tests of cyclic pitch actuation, cyclic pitch actuation is adequate for moment control, even with crosswinds of 0.14 times the blade tip speed.

Conclusions from Longitudinal Simulation of Ducted Fan Flight

Cruise condition lift to drag for the ducted fan, shown in Fig. 13, reaches a maximum of about 4 at the highest allowable angle of attack. This is due to the high bluff body drag of the tested configuration. Lift to drag ratios greater than 8 should be easily achievable by the use of faired struts and centerbody, instead of the square struts and centerbody shown in Fig. 4.

Open-loop rotor collective pitch doublet response is unstable, as shown in Fig. 14. The trimmed ducted fan (xcg=.25c) position has an unstable phugoid mode. The collective pitch control offers direct control of airspeed. As seen in Fig. 15, the cyclic pitch control is an effective control for the duct attitude, and also offers immediate response in flight path angle. The cyclic doublet response shows undesirable coupling to the flight path velocity.

Discussion

This work has presented selected results from aerodynamic analysis and wind tunnel tests of a ducted fan. The duct was designed for efficient lift in a near axial flow condition. The design is limited to UAVs due to the need to vary fuselage horizon angle from 90 to 0 degree from hover to cruise.

The use of airfoil section data and vortex lattice codes for shroud design was validated with experimental data for a new ring wing. The ring wing has high efficiency in axial flow at low chord Reynolds numbers.

In hover, it is found that duct lips with small curvature give useful rotor offloading at helicopter thrust coefficients. Comparison of induced power to thrust show that the induced efficiency gain is equivalent to increased rotor area, but without the increased profile drag penalty. Airfoil-shaped duct lips do not offer offloading, but have lower crosswind pitching moment.

Experimental results show that large control moments can be obtained by cyclic pitch application in hover and cruise. Large control moments are useful for rapid transitions from hover to cruise and back. Such transitions might be used by UAVs, eliminating trim requirements in the separated flow fuselage angle of attack range.

It is shown in simulation that the control-input responses are coupled. Cyclic and collective pitch are effective controls in simulations using experimental wind tunnel data for the shrouds tested. Finally, the desired lift to drag ratio, while not achieved in tests, could be gained by strut and centerbody streamlining.

Summary

-A new, cruise-efficient shroud is shown to match predicted high efficiency in axial flow conditions. Increased loiter capability and range are predicted for ducted fans using shrouds of the type presented here.

-High moment generation capability was reported using cyclic pitching of the blades. If control responses are properly decoupled, control surfaces will no longer need to be sized for sustained flight in the transition region between cruise and hover flight, allowing further increases in aerodynamic and structural efficiency.

Figure 1: Perspective View of Ring Wing

Figure 2: Ring Wing with Heli Fuselage and Rotor Blades

Figure 3: New Ring Wing Airfoil Stack

Figure 4: Power fraction of hover for cruising ducted fan as a function of lift coefficient and lift to drag ratio. Assumptions: 30% rotor offloading in hover and aspect ratio 3.

Figure 5: Front view of airfoil (l), and circular and trumpet lips (r), mounted on ducted fan as tested in wind tunnel

Figure 6: Cross-section detail of interchangeable lip. Top to bottom: airfoil, circular, trumpet.

Figure 7: Tested configurations. To show the angle of attack convention, α=15° is shown by a velocity arrow. For tests with a rotor, collective pitch settings are given.

Figure 8: Ring Wing Lift and Drag coefficients from α = -10 ° (lhs of plot) to α = 90 ° (at plot center).

Figure 9: Athena Vortex Lattice polar agrees with experiment; span efficiency factor $e$ reaches 1.86 (based on projected area)

Figure 10: Ducted fan hover pitching moment as a function of applied cyclic pitch.

Generated pitching moments are presented for hover and hover with crosswind. 'tr' means trumpet lip, 'c' means circle lip, and 'E' means airfoil lip. Lip name is followed by the collective pitch setting.

Figure 11: Solidity-weighted thrust coefficient versus solidity-weighted power coefficient. e68 is the ring wing with rotor and centerbody.

Figure 12: Crosswind hover pitching moment as a function of applied cyclic pitch. Cyclic pitch travel extends from -9 to +9 degrees, so the expected crosswind pitching moment can be countered.

Figure 13: Lift and drag coefficients in the cruise condition based on simulation using wind tunnel data. In the upper plot, lift, drag and aerodynamic power in 100s of watts are plotted against freestream velocity. In the lower plot, thrust and power coefficient are plotted against freestream velocity.

Figure 14: Collective doublet response

Figure 15: Cyclic doublet response

修士(工学)コルマン マイルス リチャード 提出の論文はPerformance, Stability and Control of a VTOL Ducted Fan「垂直離着陸ダクテッドファンの安定性と性能に関する研究」と題し、英文で書かれ、8章からなっている。

ダクテッドファンは、円筒状のダクトでファンを覆うことで、ファンの推進効率を上げるとともに、騒音低減にも効果があり、航空機およびホバークラフト等における推進装置として使用されている。また、近年、垂直離陸型UAV(Unmanned Aerial Vehicle)の推進装置としても実用化されている。ダクテッドファンUAVは、ヘリコプターよりもホバー時の高い推進効率と、低騒音特性、ファンが内蔵されるための安全性を備えているが、水平飛行時の効率はヘリコプターよりも劣るため、十分な航続距離の確保が困難であるという課題がある。筆者は、円筒状のダクトを円環翼(Ring Wing)として利用することで、水平飛行時の効率をさらに向上させ、ファンをヘリコプターのようなピッチコントロール可能なローターブレードにすることで制御能力を向上させることを提案し、その空力特性の解析および計測を実施し、制御能力の評価を試みている。

第1章は序論で、研究の背景を整理するとともに、研究の動機、目的を述べ、最後に本論文の構成をまとめている。

第2章は先行研究の解説で、円環翼に関する過去の風洞計測事例、その空力特性の予測手法を整理し、ダクテッドファンUAVに関する過去の研究事例を紹介している。

第3章ではダクテッドファンの構成を述べるとともに、空力計算法を示し、風洞試験の概要をまとめている。検討するダクテッドファンは円環翼の内部にローターブレードを配置したもので、円環翼は断面を軸対象とするのではなく、水平飛行時の揚力発生を考慮し、水平飛行時の上部断面と下部断面にキャンバー翼型(Eppler 68)を採用し、側部断面には対称翼型(NACA0006)を用い、途中断面には両翼型を線形補間した翼型を用いている。円環翼の空力特性の推定には、翼断面の2次元特性と3次元渦格子法を組み合わせた手法を用いている。最後に風洞実験のセットアップに関して整理している。

第4章では、円環翼の空力特性に関する解析、計測結果をまとめている。レイノルズ数247,000において、アスペクト比3.1の円環翼の最大揚力係数は1.5、スパン効率は1.86と高い値を示すことが確認され、第3章で示された計算法においても良好な精度で予測が可能であることが示されている。

第5章では、ローターブレードをダクトに取り付けたダクテッドファン形態において、姿勢角を0度から90度まで変化させ、トランペットタイプ、対称翼タイプおよび第4章で試験した円環翼のダクトに関する風洞実験を実施し、その特性を分析している。ホバー姿勢における推進効率に関しては、前縁の曲率を大きくしたトランペットタイプが高い値を示し、提案する円環翼に前縁フラップを装着することで同等の効果を期待できることを示唆している。ローターブレードのコレクティブピッチ、サイクリックピッチの効果も計測し、特にサイクリックピッチに関しては、横風ホバー時、水平飛行時ともに十分なモーメントを生成でき、ピッチ角に対してその大きさは比例的であることを確認している。

第6章は、プロペラを固定した主翼の取り付け角(チルト角)を変更するチルトウィング機に関して、垂直上昇から水平飛行への遷移飛行の飛行シミュレーションを実施している。このシミュレーションによって、ホバー状態からチルト角を一定速度で変化させる場合、高度を維持して水平飛行に遷移させるためにはプロペラ出力を著しく増加させる必要があり、逆にプロペラ出力を一定にして遷移させると大きな高度低下を伴うことを明らかにした。

第7章は、提案するダクテッドファンに関する飛行シミュレーションであり、トリム条件を明らかにするとともに、ローターブレードのサイクリックピッチは姿勢変化を行うために十分な制御能力を持つことが示され、第6章で検討したチルトウィング形態よりも優れた遷移性能を発揮できる可能性を示している。

第8章では、本研究の成果をまとめると同時に、さらなる研究課題について述べている。

以上、要するに、本論文は、円環翼にヘリコプタータイプのローターブレードを取り付けたダクテッドファン型垂直離着陸機(VTOL)をUAVの形態として提案し、その空力特性、制御能力を解析、計測し、その成立性を飛行シミュレーションによって明らかにした。これらの成果は、航空工学上貢献するところが大きい。

よって本論文は博士(工学)の学位請求論文として合格と認められる。