Aerodynamic Investigation of Propulsion Integration for a Low Noise Hybrid Wing-Body with Podded UHBR Turbofan Engines

The presented work examines the aerodynamic challenges of podded ultra high-bypass engines installed on a short/mid range low noise hybrid wing body configuration. After assessing the initial engine integration design, sensitivity studies regarding the impact of the freestream Mach number, engine position, and the engine incidence angle on the interference drag are discussed. Then, shape modifications of the nacelle and the center-body based on 2D RANS optimizations are presented. The results indicate that in order to reduce the initially massive interference drag, the overall aircraft design either has to allow for a modified engine position or for reshaping of the center-body’s upper surface. With the latter, an overall drag reduction of –44 % was achieved.


Introduction
With noise reduction being one of the key requirements according to Flightpath 2050, the DLR is investigating a novel low noise aircraft concept for short-and medium range.The main objective thereby lies in the minimization of noise immissions, i.e. the noise perceived on the ground while also achieving an acceptable aircraft performance.The hybrid wing body (HWB) concept with over the body mounted engines therefore appears to be a promising approach due to several reasons including engine noise shielding effects and potentially low requirements regarding high-lift capabilities.The installation of the engines for this type of configuration can be either realized with podded engines or buried engines.While buried engines are thought to be the more efficient option in particular when exploiting the benefits of boundary layer ingestion, it may pose a challenge to the engine design.Moreover, it clearly removes available space from the compartment below the engine, in this case the center-body.The podded installation is thought to be the more conservative approach, as it presumably poses less challenges to the airframe and engine design while promising less aero-propulsive efficiency.Nevertheless, engines mounted above lifting surfaces and more precisely over the body of a HWB provide integration challenges as well.Hooker et al. showed that while over the wing integration can yield benefits in aerodynamic efficiency when an ultra high bypass ratio (UHBR) engine is mounted at or beyond the wing trailing edge it substantially increases overall drag when it is integrated ahead of the trailing edge [1].Flamm et al. [2] and Xin et al. [3] respectively found flow separation and strong shock waves associated to engines mounted above the body of HWB designs.
The presented work investigates the aerodynamic challenges of podded UHBR engines installed on a small/medium range low noise HWB configuration.First, sensitivity studies regarding the impact of the freestream Mach number, engine position, and the engine incidence angle on the interference drag are discussed.Then, shape modifications of the nacelle and the center-body based on 2D RANS optimizations are presented.

Numerical Methods
The numerical simulations based on the Reynolds-averaged Navier-Stokes (RANS) equations have been carried out with the DLR TAU code [4].The code relies on an unstructured finite volume approach for solving the RANS equations.For the present investigation, a central scheme and a second order Roe upwind scheme were used for the spatial discretization of the inviscid mean flow fluxes and the turbulent convective fluxes, respectively.The turbulence effects were modeled with the Spalart-Allmaras formulation (SA) [5] with vortical and rotational flow correction based on the Spalart-Shur correction [6].As shown in several AIAA drag prediction workshops, the TAU code can predict total drag coefficients by as close as ∆C D /C D ≈ 3 % (compared to wind tunnel data).Moreover, differences in installation drag of 1 ≤ ∆C D × 10 4 ≤ 2 due to configurational changes such as engine position/size can be analysed correctly [7].The meshes were created based on best practice procedures having around 40 million nodes and first cell heights of y + << 1.

Reference area
174.The geometries investigated in the present study are derived from an overall aircraft design carried out within the SIAM project [8].The aircraft is designed for 170 passengers and a mission range of 2600 nm at a cruise flight Mach number of M cr = 0.78.As the design is driven by low noise considerations, a low noise and thus slotless high-lift system [9] was selected resulting in a comparably large wing size.Flight mechanic stability without artificial support for all flight conditions was an additional prerequisite during the design process.As a result, the aircraft concept features a t-tail.The stability requirement as well as the aim to utilize engine noise shielding effects led to the UHBR engines being located rather towards the front with the streamwise engine position being at 63 % of the center-body length.Moreover, the engines' bypass ratio is comparably large with 14.6 ≤ BP R ≤ 16 making engine integration even more challenging.Figure 1 depicts the basic aircraft geometry of the hybrid wing body.Table 1 summarizes the basic aircraft parameters.

Engine Integration Design Approach
In order to reduce the installation drag of the SIAM hybrid wing body (HWB), the engine integration was modified based on the following approach: • In a first step, a sensitivity study was carried out to assess the impact of the vertical and streamwise engine position, the engine incidence angle, and the freestream Mach number.• In parallel, the nacelle was designed for cruise flight conditions with 2-D and 3-D RANS computations.Low speed off-design conditions were thereby considered as well.• Then, the 2-D section at the engine's center line was parameterized and optimized via 2-D-RANS computations.Since the representative 2-D geometry includes the highly swept center-body and the nacelle, considerable effort was dedicated to find reasonable 2-D boundary conditions.Subsequent to that, two approaches have been followed for the geometric parametrization.In the first approach, only the upper rear part of the center-body section and the nacelle were modified ("local modification").The cabin dimension has not been considered in the first approach.10 design parameters were used to parameterize the section.In the second approach, the entire center-body section was modified while the cabin dimensions were considered as geometric constraint.For this approach, 24 design parameters were used.• Based on the results of the 2-D-RANS optimizations, the parameterized 3-D CAD model was adapted.The model was then modified in spanwise direction with the help of 3-D-RANS computations.Figure 2 shows the parameterized CAD model for the "local modification" approach.
• In parallel to the spanwise modifications, the engine pylon was designed and integrated.The initial computations indi cated a substantial variation in drag and thus thrust requirement.In or der to limit the effect of the thrust on the aerodynamic behavior, the design studies were carried out with constant thrust settings yielding a gross thrust of around 32 kN .More over, the 3-D computations were performed at constant lift condi tions of C L = 0.231 (mid cruise).

Baseline
From an aerodynamic perspective, the initial streamwise engine posi tion poses significant integration challenges.Figure 3 shows the flow field in terms of Mach number and the surface pressure distribution of the center-body at the engine's center line sec tion.The black line thereby represents the Cp distribution without engine and the red line the one with engine.The figure reveals a deceleration of the flow ahead of the nacelle.At the position of the nacelle lip the flow is deflected and accelerated.On the lower side of the nacelle the flow is additionally accelerated due to the channel effect between the nacelle and the center-body.Further downstream, the flow within the channel experiences a shock as the ambient pressure at the nacelle's trailing edge is too high for the shape of this Laval nozzle  type channel, causing a sudden increase in the surface pressure.Towards the trailing edge, the flow then decelerates further while the pressure recovery appears less effective compared to the case without engine, which eventually results in a lower stagnation pressure at the trailing edge.Figure 4 compares the streamwise drag distribution without engine and with engine.The solid lines thereby represent the drag contribution from the wing-body and the dash-dotted line the one from the nacelle.The plot shows that with engine, the deceleration ahead of the nacelle first causes a local reduction in drag.The channel effect however leads to a significant local drag increase on the wing-body which continues downstream of the nacelle trailing edge due to the lower pressure on the upper surface.With the trailing edge stagnation pressure being lower compared to the one of the case without engine, the local drag remains larger for the rest of the distribution.As a result, the overall drag in cruise flight increases by ∆C D = 0.0111 (83 %) compared to the combination of the isolated wing-body and isolated nacelle.Several sensitivity studies were carried out with a preliminary nacelle shape.

Sensitivity Study
Figure 5 shows the change in the drag coefficient C D depending on the change in the streamwise engine position with respect to the baseline position.The black line thereby represents computations without tail whereas the blue lines show the behavior when the tail geometry is considered in the computations.Without tail, C D steadily decreases with increasing engine position.The lowest C D value is therefore achieved when the engine is located beyond the center-body's trailing edge.Compared to the initial position, the drag coefficient is reduced by ∆C D = −0.0094(−44 %).With tail, the maximum drag reduction is limited to −∆C D = 0.0078 (compared to the baseline position).In this case, the engine position with the lowest drag is shifted towards the front to ∆X/c loc = 0.31.Moving the engine further downstream leads to a notable drag increase of the tail (dotted line), which in turn increases the overall drag.However, moving the engine position in rearward direction will also move the center of gravity towards the rear.As a result, the static stability margin is reduced.Moreover, the potential of engine noise shielding will be severely reduced [10].
The effect of altering the vertical engine position is shown in figure 6.Again, a steady decrease in C D can be observed with increasing vertical engine position.However, increasing the vertical position will lead to larger thrust-induced pitching moments.Moreover, the structural weight, viscous drag of the pylon, and noise shielding towards the ground will be adversely affected [10].Reasonable vertical engine positions will therefore only lead to moderate drag reductions.
Figure 7 depicts the effect of the engine's incidence angle on the drag coefficient for three vertical heights.While it can be expected that the streamwise engine position has a significant influence on the optimal incidence angle due to the shape of the center-body's upper surface, the effect of the vertical position appears to be negligible.The curves of all investigated engine heights show the same trend.With rising incidence angle, the drag coefficient is reduced until a minimum is reached.Beyond this point, the drag coefficient increases again with rising incidence angle.The optimal incidence angle is at γ = 8 • for all investigated engine heights (for the baseline streamwise engine position), yielding a drag reduction of −∆C D ≤ 0.0026 compared to γ = 0 • .

Shape Modifications
In order to adapt the geometric shape for engine integration 2-D RANS optimizations of the middle section as shown in figure 2 were performed.As most geometric parameters such as the vertical and streamwise engine position have implications on aspects other than aerodynamic performance, several optimizations with varying geometric constraints were performed.Table 2 summarizes the changes in geometric constraints during the 2-D optimizations compared to the baseline values for selected cases.Besides the dimensionless vertical and streamwise engine position (∆Z E /D N and ∆X E /c ref ), the dimensionless minimum vertical height of the center-body's upper surface at the position of the nacelle (∆Z CB /D N ) is an important parameter that has been varied.∆Z CB /D N potentially affects the cabin size and structural weight.As mentioned before, the optimization of the entire section (ES) considers the cabin dimension as geometric constraint.
Figure 9 shows the shapes of the optimized   The comparison reveals that the local flow velocities between the the center-body and the nacelle are notably reduced due to the optimization.The area of supersonic flow is now limited to a small area at the nacelle lip and does not extent to the center-body surface anymore.Moreover, the flow separation that exits in the initial case has vanished.Comparison of surface pressure coefficient distributions of base line and optimized (4-LM-SM) geometric shape Subsequent to the 2-D optimizations, the resulting 2-D shapes were transferred to the 3-D CAD model and 3-D RANS computations were performed.The shape was thereby further adapted by manually modifying the design parameters in the inboard and outboard sections as shown in figure 2. Additionally, the center-body planform was slightly altered to eliminate flow separation at the center section and the engine pylon was integrated.Figure 11 visualizes the 3-D flow field in vicinity of the engine for two selected optimization results after spanwise modifications (SM) and pylon integration.In both cases, the streamlines colored by the Mach number indicate subsonic flow in the lower part of the channel between the center-body and the nacelle.Similar to the 2-D-computation of case 4-LM, the flow field only indicates supersonic flow at the nacelle lip.While for 4-LM-SM, the flow is also fully subsonic ahead of the nacelle, case 6-ES-SM reveals supersonic flow at the location of highest curvature where the contraction of the center-body begins.Moreover, the streamlines as well as the surface pressure coefficient distribution indicate a shock at this location.
Figure 12 compares the surface pressure distribution of the 4-LM-SM case with the one of the baseline.While the comparison confirms a substantial reduction of the pressure increase ahead of the nacelle and a removal of the low pressure zone and the accompanied shock below the nacelle for the 4-LM-SM case, it also reveals inhomogeneities in the C P -distribution at the wing root that may still affect the performance adversely.
Figure 13 summarizes the optimization results of the different cases in terms of change in overall drag coefficient compared to the baseline.The drag difference thereby includes the change in airframe drag as well as the change in net thrust (at constant gross thrust).Besides the optimization results, the total drag of the isolated wing-body plus the isolated engine (without pylon) (ID=0) as well as the case with changes in the engine vertical position and incidence angle but without any shape modification (ID=1) are shown.Without any shape modification, the drag reduction is rather low (∆C D = −0.0015).Modifying the shape reduces C D notably.Interestingly, cases 2-LM, 3-LM, and 5-LM thereby yield similar drag reductions ranging between −0.0074 ≤ ∆C D ≤ −0.0072.In contrast, lessening the constraint on the minimum vertical height of the center-body at the position of the nacelle leads to a stronger drag reduction of −∆C D = 0.0087 as shown by case 4-LM.The drag reduction is further increased by spanwise modifications (4-LM-SM) as indicated by the green symbol with the magenta outline (−∆C D = 0.0109).This value is close to the drag difference of the isolated case (−∆C D = 0.0111).Integrating the engine pylon increases the overall drag coefficient by only ∆C D = 0.0001.The case with the optimization of the entire 2-D section (6-ES-SM), where the cabin dimension remains unaffected, yields a lower drag reduction (−∆C D = 0.0100) compared to 4-LM-SM.On the one hand, the shock ahead of the nacelle leads to additional drag.On the other hand, the spanwise modifications have not been carried out in such detail as it was done for 4-LM-SM most likely resulting in a further drag increase compared to 4-LM-SM.

Thrust-Drag Bookkeeping
After correcting the thrust to achieve a balanced, level flight under cruise conditions and performing thrust drag bookkeeping (nacelle/engine component drag inside of the nacelle's leading edge stagnation line was accounted to the net thrust), the 4-LM-SM case (without tail) yields a lift-to-drag ratio of L/D = 20.7.

Conclusions
Numerical simulations of a hybrid wing body configuration with podded UHBR engines have been performed in order to assess and improve the aerodynamic performance with regard to engine integration.The initial engine integration leads to a strong shock within the channel between the center-body and the nacelle affecting the flow downstream and eventually causing a substantial drag increase.The magnitude of the drag increase is thereby adversely affected by the streamwise engine position that was chosen due to flight stability and engine noise considerations.Sensitivity studies indicate that the drag can be substantially reduced by alteration of the vertical engine position and the engine incidence angle.Most notably, however, the drag is reduced by moving the engine towards the trailing edge.With the engine located beyond the center-body's trailing edge, the drag is reduced by −44 % compared to the initial design if the tail is not considered.With tail, the drag reduction is limited to −37 %.Maintaining the initial streamwise engine position, significant drag reduction can still be achieved by modifying the shape of the nacelle and the center-body's upper surface.A rather simple approach by means of 2-D-RANS optimizations of the center-body's section at the engine's center in conjunction with subsequent manual spanwise modifications leads to notable results.By only modifying the area in vicinity of the engine and allowing a slight reduction of the cabin size, the drag can be reduced by −44 %, yielding an overall drag penalty of 1.7 % compared to the sum of the isolated wing body's and engine's drag coefficients.It was found that selecting representative flow conditions for the 2-D computations is thereby essential for achieving substantial drag reductions.
Optimizing the entire center-body's section while maintaining the cabin dimension leads to a drag reduction of −41 %.While there may be additional potential for improvement, it is generally assumed that the drag reduction in this case is adversely affected by the thicker front part of the center-body eventually causing a shock ahead of the nacelle.

Figure 3 .
Figure 3. Flow field and surface pressure distribution of center-body at engine's center line section

Figure 5 .
Figure 5. Change in total drag (vs.baseline) depending on streamwise engine position

Figure 6 .
Figure 6.Change in total drag (vs.baseline) depending on vertical engine position

Figure 7 .
Figure 7. Change in total drag (vs.baseline) depending on engine incidence angle

Figure 8 .
Figure 8. Change in total drag (vs.baseline) depending on Mach number to the initial shape (black line).The green line thereby represents the shape of case 4-LM and the blue line the one of case 6-ES.The comparison illustrates how the optimizer handles the additional geometric constraint due to the cabin dimension (magenta line) in the 6-ES case by increasing the center-body's thickness ahead of the nacelle.Moreover, the limitation of ∆Z CB /D N becomes obsolete as the minimum vertical height of the center-body at the position of the nacelle is defined by the cabin.

Figure 9 .
Figure 9. Shape of optimized 2-D-section at engine center position.Black: initial shape, green: 4-LM, blue: 6-ES, magenta: cabin Figure 10 compares the flowfield in terms of the Mach number in the region of the engine for the initial shape and the optimized shape of case 4-LM.The comparison reveals that the local flow velocities between the the center-body and the nacelle are notably reduced due to the optimization.The area of supersonic flow is now limited to a small area at the nacelle lip and does not extent to the center-body surface anymore.Moreover, the flow separation that exits in the initial case has vanished.

Figure 10 .Figure 11 .
Figure 10.Flow field in vicinity of nacelle in 2-D section at engine center

Figure 13 .
Figure 13.Change in C D of 3D model due to shape modifications

Table 2 .
Change in geometric constraints during 2-D optimizations compared to the baseline values