Wind tunnel investigation of different engine layouts of a blended-wing-body transport

Zheng CUI, Guojun LAI, Qifeng WANG, Yu LIANG, Yihu CAO

a School of Aeronautic Science and Engineering, Beihang University, Beijing 100083, China

b COMAC Beijing Aircraft Technology Research Institute, Beijing 102211, China

KEYWORDS PAI;Aerodynamics;BWB;CFD;Civil aviation;Engine layout;Transport aircraft;Wind tunnels

Abstract A 2% scale, cruising version of a 450-seat class Blended-Wing-Body (BWB) transport was tested in the China Aerodynamic Research and Development Center’s FL-26 2.4-by-2.4-meter subsonic wind tunnel.The focus of the wind tunnel test was to investigate the aerodynamic performance of the latest BWB transport design, which would also aid in choosing a final engine arrangement in the three most potential engine integration layouts.The wind tunnel model can be tested with and without the nacelle and has three sets of different nacelle/tail integration positions.Computational Fluid Dynamics (CFD) simulations were performed in engine-aircraft integration design to find appropriate nacelle installing parameters of each layout.The comparison of CFD with experimental results shows good agreement.Wind tunnel measurements indicate that the tail-mounted engine layout produces the minimum drag penalty, while the fuselage-mounted engine layout increases drag the most.Experimental pressure measurement illustrates the effect of nacelle integration on the wing-body surface pressure distribution.This experimental and numerical research provides a reference for future BWB Propulsion-Airframe Integration (PAI) design.

1.Introduction

The United States and Europe have carried out research projects such as Environmentally Responsible Aviation (ERA)1–4 and Clean Sky5intending to maintain the leading edge of the aviation industry and seize the commanding heights of the future aviation market competition, and dozens of unconventional aircraft layout transport configurations such as Blended-Wing-Body (BWB) transport have been proposed and investigated.6–10The BWB layout refers to the shape of a full-lifting surface aircraft in which the wings and the fuselage are highly fused.11–15The BWB layout can reduce the wetted area of the whole aircraft under a given load requiring,and thus reduce the frictional drag.16–20Compared with the traditional conventional layout, the BWB cruise efficiency is improved by 15%-20%,21–24and it has the advantages of reducing noise, emissions, and structural weight, and other potential advantages.These advantages are represented by Boeing’s BWB-450 concept25and X-48 series demonstrator.26,27The academia and industry’s research on BWB is gradually deepening.The US X-48B and X-48C sub-scale BWB flying demonstrator verifies,by flight test,the feasibility,flight control, noise, fuel consumption, and other aspects of BWB layout,and the rich design experience and flight test data have been obtained.In NASA’s ERA project, the ERA-0009H1 is the result of refinement conducted under the ERA-0009A design that was developed under the ERA Advanced Vehicle Concepts (AVC) contract with Boeing.28Later, the Boeing-MIT team proposed BWB layout design H3 for NASA’s N+3 plan.29European ACFA2020 research project also selected BWB layout for next-generation civil aircraft.30Airbus built a demonstrator named MAVERIC(Model Aircraft for Validation and Experimentation of Robust Innovative Controls)for a possible full-scale BWB airliner and claimed that this design can reduce up to 20% of fuel.31The Russian Central Aerohydrodynamic Institute(TsAGI) launched the 900-seat BWB layout airliner program FW-900 back in 1996, and continued to design hybrid wingbody aircraft including Hybrid Wing-Body(TsAGI-IWB),lifting body layout(TsAGI-LF)and Flying Wing layout(TsAGIFW).32A series of research works and wind tunnel experiments have also been performed in Chinese academia, such as BWB-150, BWB-300 and BWB-330 proposed by Northwestern Polytechnical University (NPU).33,34

The aft engine location of the BWB allows for plenty of installation options.Candidate installation concepts include podded with pylon, upper or lower surface inlet with S-duct,Boundary Layer Ingestion (BLI), or distributed propulsion.The aft engine location on the BWB offers the opportunity to ingest the boundary layer generated on the center body forward of the inlets.Boeing Company under the Ultra Efficient Engine Technology/Propulsion Airframe Integration Project has performed the study to assess the improvements in the BWB’s mission range that occur when podded engines are replaced with the embedded BLI system.35–37Numerical and experimental activities conducted by French Aerospace Lab(ONERA) within the RAPRO2 project allow an in-depth investigation and quantification of the mechanisms of BLI.Also, the experimental results are used to establish and validate a numerical simulation approach for future BLI propulsion system design.38Researchers39–41have been examining distributed propulsion systems that break the traditional twin turbofan engine configuration into arrays of many smaller engines.These arrangements create the opportunity for drag reduction and noise improvement.But the smaller turbofans are generally less efficient than large ones due to the practical limits of the core size and turbine inlet temperature.Distributed propulsion configurations will therefore require innovative electric turbo compound propulsion solutions to minimize these penalties.Simply mounting the engine on pylon is the most practical option, but increased wetted area and weight plus nose-down thrust moment are detractors of this installation.Lockheed Martin (LM) has teamed with both the Air Force Research Laboratory (AFRL) and the National Aeronautics and Space Administration (NASA) to develop a new strategic freighter called the Hybrid Wing Body (HWB) and validate its performance.42A Boeing 757 sized HWB configuration with over-wing engine nacelles was proposed and investigated.The results documented in this report indicate that an HWB-based commercial freighter with over-wing engine layout would use approximately about 7% less fuel than an advanced tube-and-wing configuration using the same technology levels.

To sum up, many BWB civil transport configurations have been proposed and investigated,covering a variety of research on aerodynamics, structure, stability, and flight control.The previous research has considerably improved the engineering feasibility of the BWB transport concept.However, unlike conventional civil aircraft that generally integrates engines by wing-mounted arrangement, the academia and industry still do not have a general agreement on the question: which is the best way to integrate engines onto the BWB aircraft.And there is a devastating shortage of related wind tunnel data to support the engine integration design.So, in this research,we have investigated the three most popular engine arrangements of BWB civil transport configurations, which are wing-mounted layout, fuselage-mounted layout, and tailmounted layout.We hope that this work will give a good reference for future BWB aircraft design and push the BWB forward as a relatively clear pathway for future aircraft development.

2.Experimental setup

2.1.Test description

Based on the needs of the future civil aircraft market, Commercial Aircraft Corporation of China(COMAC)Beijing Aircraft Technology Research Institute (BATRI) has launched pre-researches on a new generation of passenger aircraft and conducted comprehensive trade-offs of various layouts for future subsonic transport.The prototype of this research is a 450-seat BWB configuration with a cruise Mach number of 0.85 and a cruise altitude of 11000 m(Fig.1).The wind tunnel research is carried out in the FL-26 wind tunnel of the China Aerodynamics Development and Research Center.The FL-26 wind tunnel is the present largest ejector-type semirecirculation transient transonic wind tunnel in China.The size of its test section is 2.4 m×2.4 m,and the maximum Reynolds number per unit chord length test can reach 70 million/m with the capability to test Mach numbers from 0.4 to 1.4.Two primary goals exist for this research.The first is to study the subsonic and transonic aerodynamic characteristics of the BATRI BWB configuration,and the second is to investigate the effects of different engine integration layouts on the aircraft’s overall aerodynamic performance.

The design of the wind tunnel test model followed the idea of building blocks and modular design.The wind tunnel test starts with the basic wing-body assembly configuration, and then gradually adds other components for comparison,including the very basic wing-body-tail assembly (baseline) configuration, wing-mounted engine configuration, fuselagemounted engine configuration, and tail-mounted engine configuration.

Fig.1 BATRI 450-seat class BWB transport.

Conventional static tests are conducted at a dynamic pressure of 11.4×103Pa,corresponding to a Reynolds number of 8.99 × 106based on mean aerodynamic chord.Data are acquired at angle of attack from —2° to 8°.For all tests, sixcomponent force and moment data were acquired using an internal strain-gauge balance.To minimize Reynolds number effects, transition tripping strip is used to fix transition at 7% chord on the upper and lower surface of the wing, Vtail, and nacelle.

2.2.Model features

According to the width of the test section of FL-26,the scale of the model is set to 1:50.The model has a wingspan of 1.6 m and the average aerodynamic chord length is 509 mm.The support form of the model is abdominal support.The installation of the test model is shown in Fig.2.

The maximum thickness position of the BWB test model is about 135.7 mm, and its distance from the nose is about 312 mm.An internal strain-gauge balance is installed inside the central lift body and connected to the abdominal support.The test model has a total of 181 pressure measuring holes with an inner diameter of 0.4–0.6 mm, which is respectively arranged on eight spanwise positions distributed along the wingspan direction.Their relative positions to the symmetry plane of the aircraft are 0%, 17%, 23.5%, 30%, 36.5%,48%, 67.5% and 93.5%, respectively.

The prototype BWB transport is designed as a twin-engine with V-tail(see Fig.2).The V-tails are installed on the back of the center lift body.Compared with an integrated V-tail at the wingtip, this tail layout can achieve a higher tail capacity and better aerodynamic and structural efficiency of the V-tail itself.That can further benefit the directional trim under the condition of single-engine failure and also increase the directional stability of the aircraft when flying at a high angle of attack.The three engine integration layouts being investigated herein are the fuselage-mounted, wing-mounted, and tail-mounted layouts.

The fuselage-mounted layout puts nacelles on the back of the fuselage between the V-tails, which is the most common engine arrangement in the present BWB transport design (see Fig.3 and Fig.4).

The second is the wing-mounted layout.This layout is to put the nacelle over the wing through a pylon that is connected to the wing’s trailing edge and extends backward to avoid the mutual aerodynamic interference between the nacelle and the wing as much as possible (see Fig.5 and Fig.6).

Fig.2 Installation of wind tunnel model.

Fig.3 Fuselage-mounted layout geometry.

Fig.4 Fuselage-mounted layout model.

Fig.5 Wing-mounted layout geometry.

Fig.6 Wing-mounted layout model.

Fig.7 Tail-mounted layout geometry.

The last one is the tail-mounted layout.This layout integrates the engine nacelle at the root of the V-tail.Then the nacelle-tail assembly is installed as a whole part (see Fig.7 and Fig.8).

For the above three engine layouts, the fuselage-mounted layout has the acoustic benefits of the shielding effect of the engines,42and neither of the other two layouts will be equally effective since only part of the engine was shielded; the wingmounted layout has an unloading effect on the wing, which can produce a structural benefit; the tail-mounted configuration has structural reinforcement on the root of V-tail; the fuselage-mounted one also has the advantage of increasing less yaw moment at the time of one engine failure.All the above highly complex factors have to be considered comprehensively in the design progress.However, this wind tunnel test only focuses on the aerodynamic characteristics and the trade-offs between aerodynamics,and other aspects are beyond the scope of this paper.

2.3.CFD results overview

CFD simulations were performed to aid in Propulsion-Airframe Integration (PAI) design, and an in-house software S-Flow package based on the Reynolds average Navier-Stokes (RANS) equation was used to simulate the Three-Dimensional (3D), compressible, steady flow over the BWB prototype.Roe’s scheme is used for space discretization.The Lower-Upper Symmetric Gauss-Seidel (LU-SGS) method is used for time advancement.And k-ω Shear Stress Transport(SST) equation is used for the turbulence model.

For wing-body assembly design, we use a Free-Form Deformation (FFD) approach to parametrize the geometry.The FFD volume parametrizes the geometry changes rather than the geometry itself,resulting in a more efficient and compact set of geometry design variables, thus making it easier to handle complex geometric manipulations.In the aspect of the optimization algorithm, we use a gradient-based optimizer combined with adjoint gradient evaluations to solve the problem efficiently.Afterward,the CFD-assisted design progress of the nacelle installation position is carried out for each engine layout.The investigated variables include the nacelle height and chordwise position.For each engine layout,at least 30 sets of installation positions are calculated by CFD at the corresponding spanwise position (see Fig.9 and Fig.10).Finally,the full configuration of the integrated nacelle, pylon, and Vtail are evaluated.The overall design principle of nacelle integration is to avoid the boundary layer on the upper wing surface being sucked into the nacelle, and at the same time to ensure that the drag increase caused by the nacelle/pylon integration is minimized.The selected nacelle installation positions of different engine layouts are shown in Fig.11, and the CFD drag prediction results of each integrated configuration under the cruising condition are shown in Table 1.

Fig.8 Tail-mounted layout model.

Fig.9 Nacelle installation position.

Fig.10 CFD drag prediction of nacelle installation.

3.Results

3.1.CFD and experimental comparison

Fig.11 Top view and back view of nacelle installation position.

Table 1 CFD drag prediction of different engine layouts.

For conventional tube-and-wing configuration,the model support system can be attached to the tubular fuselage.Since the support is far away from the wing, the aerodynamic interference of the support on the wing can be well avoided.The BWB configuration, however, uses an integrated lift surface design where the fuselage and wing are highly fused.The aerodynamic interference of the model support on the wing-body assembly is particularly obvious.Therefore, the Reynolds number correction and support interference correction of the wind tunnel test data of the BWB configuration are quite difficult, and the relevant test data correction methods are still immature.Considering the above reasons, it is very difficult to directly compare the CFD data of the prototype with the wind tunnel test data.To verify the accuracy of the CFD numerical method, we have conducted a CFD analysis of the wing-body assembly model that has the model support integrated.The Reynolds number of the numerical simulation is consistent with the actual experimental conditions.Fig.12 shows the structured grid used in the above CFD simulation,the total cell number is 11.7 × 106, and the simulated Mach number and Reynolds number are 0.85 and 8.99 × 106,respectively.

Fig.13 shows the comparison of CFD and experimental data for lift and drag coefficients, where CLrepresents the lift coefficient,CDrepresents the drag coefficient,AOA represents the angle of attack, respectively.It can be seen that the CFD prediction of the lift coefficient is in good agreement with the test results,and the CFD prediction of the drag coefficient is slightly higher than the test results, but the deviation is within an acceptable range.Fig.14 and Fig.15 show the comparison between the result of CFD simulated and experimental surface pressure measurement, where Cprepresents the pressure coefficient, X/C represents the local chord-wise position divided by the local chord length.The corresponding spanwise positions are 23.5% and 67.5%, respectively.The CFD and experimental comparisons above show that the S-Flow results are in excellent agreement with experimental results both for the force coefficients and surface pressure measurement.This agreement further proved that the grid generation strategy and numerical method can correctly predict overall aerodynamic characteristics and well support the design of BWB configuration.

Fig.12 Structured grid of wing-body assembly model with support.

Fig.13 Comparison of lift coefficient and drag coefficient between CFD and experiment.

Fig.14 Pressure coefficient comparison at spanwise position 23.5%.

Fig.15 Pressure coefficient comparison at spanwise position 67.5%.

3.2.Force measurement results

Fig.16 Lift coefficients comparison.

Fig.17 Polar curves comparison.

Fig.18 Pitching moment coefficients comparison.

The wind tunnel test results of lift coefficient, polar curve,pitch moment coefficient,and lift-to-drag ratio K versus Angle of Attack(AOA)of the baseline configuration and three different nacelle-integrated configurations are shown in Figs.16-19.It should be noted that the geometric irregularity of the BWB fuselage makes it quite challenging to arrange a regular-size internal strain-gauge balance in the test model, so a ponysize internal strain-gauge balance is used in this experiment.However, during the experiment, we found that the stiffness of the pony-size balance was not enough to suppress test model oscillation at angles of attack over 6°.For safety reasons, in the later tests of the tail-mounted configuration and the fuselage-mounted configuration,the test angle of attack is limited to 5.5°.This adjustment resulted in inconsistencies in the angle of attack range of the data.However, we do not think it will affect the validity of the experimental data, because we did not focus on the high angle of attack characteristics in this experiment.

Fig.19 Lift-to-drag ratio comparison.

From the comparison of lift coefficients (see Fig.16), we can see that the integration of the engine nacelle will lead to a reduction in the lift of the basic configuration.Among them,the lift loss caused by the fuselage-mounted and tail-mounted engine layouts is about the same,and the lift coefficient reduction caused by the wing-mounted nacelle is the most obvious.It indicates that the wing-mounted engine layout deteriorates wing performance.So, the wing/engine integrative aerodynamic optimization progress should be performed if this engine arrangement is adapted.From the polar curve(Fig.17), we can see an overall drag increase after integrating the engine nacelle.Compared with the baseline configuration at the lift coefficient of 0.1, the drag increment for the fuselage-mounted, wing-mounted, and tail-mounted nacelle configuration is 55.3, 8.7, and 7.3 counts, respectively.So,the fuselage-mounted engine layout has brought a significantly higher drag cost than the other two engine-integrated configurations.The CFD simulation results also indicate that the fuselage-mounted engine arrangement leads to higher wave drag and interference drag because of the very close distances between nacelles and nacelle to the V-tail.From the comparison of the surface pressure coefficient (see Fig.20 and Fig.21), it can be observed that shock waves appear on the nacelle of the wing-mounted configuration.We believe that this phenomenon is caused by the mutual aerodynamic interference of the nacelle and tail and increases overall drag.

From the comparison of the moment coefficients (see Fig.18), the pitching moment coefficient maintains a linear change in the range of 0°–5° angles of attack, and all the moment coefficient curves have an upturn at a 5° angle of attack.The three nacelle integration methods have different effects on the pitching moment, but the wind tunnel test data is still inadequate in evaluating which engine layout has a better moment character because the engine thrust influence has to be taken into account.The lift-to-drag ratio varies with the angle of attack as shown in Fig.19.Here we must point out that the test results of the lift-to-drag ratio of each configuration are significantly lower than the CFD results at the designed Reynolds number.This is because the Reynolds number of this test is significantly lower than the flight Reynolds number of the prototype.The flight angle of attack corresponding to the max lift-to-drag ratio is also about 2° higher because of the Reynolds number effects.(The test model has its best lift-to-drag performance at an AOA of 5°,but the prototype aircraft was designed for cruising at an AOA of 3°).At an angle of attack of 5°,the tail-mounted nacelle configuration has the best performance on lift-to-drag ratio,compared to the baseline configuration, it has a lift-to-drag deduction of 0.55,and the wing-mounted nacelle configuration and fuselagemounted configuration have a lift-to-drag deduction of 0.7 and 3.35, respectively.

Fig.20 Pressure coefficient of tail-mounted configuration.

Fig.21 Pressure coefficient of wing-mounted configuration.

3.3.Pressure measurement results

Fig.22 indicates the spanwise position(η)of the pressure measurement wing section.Figs.23-30 show the pressure measurement results of each wing section under Mach number 0.85 at an angle of attack of 5°.Generally speaking, the influence of the nacelle integration on the pressure distribution mainly concentrates on the inner wing upper surface and center lift body upper surface.From spanwise positions 0%-48%(see Figs.23-28), we can find that the nacelle integration will increase the pressure local near the trailing edge.That is because the nacelle has a blocking effect on the incoming flow, and the speed decrease of incoming flow leads to a pressure increase.The pressure distribution of the tail-mounted engine layout is the closest to the baseline pressure distribution,indicating that this nacelle integration approach has the mildest effect on baseline aerodynamics.On the contrary, the fuselage-mounted nacelle configuration has the most obvious effect on the upper surface of the center lift body (see Figs.23-25), but at spanwise positions of 30%-48% (see Figs.26-28), the effects of the wingmounted nacelle configuration exceed the influence of other configurations.It is also worth noting that at the spanwise position 67.5% wing section (see Fig.29) there is the opposite trend.The integration of the engine leads to a reduction in the pressure on the outer wing’s upper surface, which means that the wing-mounted and fuselage-mounted nacelle has increased the lift on the outer wing, and this is quite different compared to its effects on the inner parts.The reason for this inverted change is the airflow on the upper surface of the inner wing and the center lift body is partly blocked by the nacelle,resulting in more air flow being directed to the outer wing part.From this, we can know that the two nacelle integration layouts, the fuselage-mounted and wing-mounted layouts will lead to a lift decrease at the inner wing and lift increase at the outer wing, so the spanwise load distribution moves outward.The pressure distribution of the tail-mounted engine layout is the closest to the baseline configuration, resulting in the least impact on the spanwise load distribution.

Fig.22 Spanwise positions of pressure measurement.

Fig.23 Spanwise position: 0%.

Fig.24 Spanwise position: 17%.

Fig.25 Spanwise position: 23.5%.

Fig.26 Spanwise position: 30%.

Fig.27 Spanwise position: 36.5%.

Fig.28 Spanwise position: 48%.

Fig.29 Spanwise position: 67.5%.

Fig.30 Spanwise position: 93.5%.

Fig.31 shows the spanwise Cpdistributions near the 60%chord of the wing-body assembly at an AOA of 5°.The background image shows the location of the spanwise pressure taps as well as the engine nacelle to help interpret the pressure distributions.As expected,the suction pressures on the main wing vary with the change of engine layouts.The effect of the engine nacelle is noticeable, especially in the upstream spanwise Cpdistributions.The dip in the Cpdistributions varying with the nacelle/pylon integration position corresponds to the aerodynamic interference of nacelles and pylons on the wing-body assembly.The spanwise Cpdistributions of wing-mounted and fuselage-mounted layout appear to be dramatically suction pressure drop, implying a potential lift loss induced by the engine integration.

Fig.31 Variation of spanwise Cp distribution along 60%chord.

Fig.32 Spanwise lift distribution of CFD results.

The spanwise lift distribution of the CFD results is illustrated in Fig.32,and the CFD calculation conditions are consistent with the wind tunnel test conditions.Comparing Fig.32 with Fig.31, we can see that the changing trend of the spanwise lift is consistent with the changing trend of the pressure coefficient.The nacelle integration does have a significant effect on the spanwise lift distribution.The wing root bending moment will also change according to the different integration methods of the nacelle.

4.Conclusions

A 2% scale version of a 450-seat class BWB transport wind tunnel test model was successfully tested in the FL-26 wind tunnel of the China Aerodynamics Development and Research Center.Wind tunnel measurements included force and pressure data as well as pressure measurement by 181 pressure holes.

This study investigates the three mainstream engine integration layouts around the cruising Mach number 0.85.Force measurements indicate performance degradation after nacelles are integrated with baseline configuration.The lift coefficient reduction caused by the wing-mounted nacelle is the most obvious.An overall drag increase after integrating the engine nacelle is observed.Compared with the baseline configuration at the lift coefficient of 0.1,the drag increment for the fuselagemounted, wing-mounted, and tail-mounted nacelle configuration is 55.3, 8.7, and 7.3 counts, respectively.The tailmounted nacelle configuration has the best performance on lift-to-drag ratio, compared to the baseline configuration, it has a lift-to-drag deduction of 0.55, and the wing-mounted nacelle configuration and fuselage-mounted configuration have a lift-to-drag deduction of 0.7 and 3.35, respectively.

Pressure measurement results indicate that the influence of the nacelle integration on the pressure distribution mainly concentrates on the inner wing upper surface and center lift body upper surface.The pressure distribution of the tail-mounted engine layout is the closest to the baseline pressure distribution, indicating that this nacelle integration approach has the mildest effect on baseline aerodynamics.However,other undesired effects brought about by tail-mounted layout require further investigation, including the control performance loss,structural strength, and vibration of the V-tail caused by the engine installation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This study was co-supported by the Beijing Key Laboratory of Simulation Technology for Civil Aircraft Design, China (No.11WX08), and Innovation Foundation of COMAC Beijing Aircraft Technology Research Institute, China (No.Y16QT01).