Numerical investigation of the powered take-off configuration of civil transport using an improved thrust-drag bookkeeping method

This paper offers a detailed examination of the powered nacelle’s effect on take-off configuration of a civil transport. Using computational fluid dynamics (CFD) techniques, we analyzed two specific geometries from the Propulsion Aerodynamics Workshop and High-lift Prediction Workshop, comparing our results with existing experimental data. We incorporated configurations of both the Through-flow Nacelle (TFN) and Power Nacelle (PN) for a comprehensive comparison. Our approach significantly refined the prevailing thrust-drag accounting method. Employing surface flow visualizations and spatial particle trajectories, we highlighted the powered nacelle’s influence on the aircraft flow field. Simulations show that the engine exhaust’s flow acceleration boosts lift and reduces flow separation at high angles of attack. However, power intervention notably increases the overall drag. Specifically, at a 14° angle of attack, power effects raise the drag coefficient of the take-off configuration by 275 counts, resulting in a 0.76 decrease in the lift-to-drag ratio.


Introduction
The tragic crashes of the Boeing 737 MAX in 2018 and 2019 brought sharp focus to specific design considerations.The aircraft's redesigned nacelle inadvertently introduced lift at high angles of attack, thereby increasing the risk of stalling.Boeing's remedy, the Maneuvering Characteristics Augmentation System (MCAS), came under scrutiny following the incidents, underscoring the paramount importance of meticulous aerodynamic design and comprehensive engine-airframe research [1].A myriad of studies have concentrated on the evolution towards high-bypass ratio engines and their intimate incorporation with the airframe.Landmark European collaborative endeavors such as DUPRIN I & II [2], ENIFAIR [3], AIRDATA, and the Clean Sky 2's [4] ongoing IVANHOE [5] project have been at the forefront of Propulsion/Airframe Integration (PAI) research since the early 1990s.They traverse a spectrum of subjects, with a particular emphasis on the precision of aerodynamic force predictions.
The American Institute of Aeronautics and Astronautics (AIAA) has been an inspiration, with its Drag Prediction Workshops (DPW) [6] propelling advancements in Computational Fluid Dynamics (CFD) predictions.These workshops have established benchmark geometries, enabling a side-by-side comparison of CFD methodologies with empirical results.Notably, AIAA's High Lift Prediction Workshop, convened four times, zeroes in on the fidelity of CFD predictions concerning aircraft wings under high-lift scenarios.This includes the dissection of various civil transport models, such as DLR-F11, the model from the Japan Aerospace Exploration Agency, and the widely recognized NASA Common Research Model (CRM).The latest incarnation of CRM, symbolic of a twin-engine, long-haul aircraft, has gained traction in research circles exploring the integration of ultra-high bypass ratio engines [7][8][9][10][11].
Upon reviewing the aforementioned studies, we've identified three core challenges in numerical research on engine integration: 1.The precise prediction of aerodynamic forces.
2. The realistic simulation of engine intake and exhaust dynamics.
3. Post simulation, the extraction of exact aerodynamic forces and engine thrust.This intricate procedure is encapsulated in the term -the Thrust-Drag Bookkeeping (TDB) method.While there have been numerous attempts to research engine integration, the vast majority of experiments and numerical studies have opted for the through-flow nacelle (TFN) as their model.To simulate the effects of engine exhaust and intake on the flow field more realistically, it's essential to incorporate a powered nacelle in the CFD calculations that can emulate these effects.This not only complicates the aerodynamic computations but also becomes particularly challenging when dealing with complex high-lift configurations.Given the inherent geometric complexity of these configurations, coupled with the need to simulate the real effects of the engine on the flow field, both grid generation and numerical convergence become daunting tasks.This might explain the relatively few analytical articles published in this domain.
Currently, there are two main areas of concern in the studies: 1.Many attempts to simplify complexities led researchers to pare down geometric models, which inadvertently resulted in the neglect of vital nacelle strakes.2. Present methods of thrust-drag bookkeeping require further refinement, especially in high-lift configurations.This paper will address the aforementioned issues, starting with refining the existing TDB method, followed by the correct integration of the nacelle strake in take-off configurations, and culminating with a numerical analysis of powered take-off configurations.

CFD method
In this study, we utilized COMAC's proprietary S-Flow code for all computations.S-Flow, a finitevolume solver, is built upon the Reynolds-averaged Navier-Stokes (RANS) equations.It employs Roe's scheme for spatial discretization and the LU-SGS method for temporal advancement.The software offers a range of turbulence models, including the Spalart-Allmaras (SA) and k-ω shear stress transport (SST).To handle large-scale problems efficiently, S-Flow incorporates a massively parallel computing framework using the message passing interface (MPI) and leverages multigrid techniques to expedite numerical convergence.This enables S-Flow to deliver highly accurate simulations for complex aerodynamic studies, making it a valuable tool in aeronautical research and development.

Powered engine simulation
In the referenced studies [12][13][14][15][16], simulations of the powered engine nacelle predominantly overlooked the internal flow of the engine while computing aircraft outflow with engine powered condition.The primary focus of these studies was to determine the alignment of engine inlet and outlet flows with realworld operational conditions.In our approach, the fan is distilled down to a model represented by inlet and outlet boundary disks, while the turbo is modelled solely as an inlet boundary disk.Specifications for mass flow rate and total pressure at the engine boundaries are based on data provided by the engine manufacturer.

The improvement of the existing thrust-drag bookkeeping method
The thrust-drag bookkeeping method mentioned in paper [17] originates from the control volume theory of powered fans.In this method, the key issue lies in the accurate prediction of the velocity coefficient of the powered nacelle.Current methods calibrate the velocity coefficient of the powered nacelle based on CFD calculations of an isolated powered nacelle under given inlet and outlet boundary conditions.However, this approach neglects the flow disturbances caused by the wing, engine pylon, and high-lift devices on the engine intake and exhaust when the nacelle is integrated with the aircraft.This might influence the calibration of the flow coefficient.
Furthermore, during the process of extracting force coefficients, the basic thrust-drag bookkeeping method assumes that the engine thrust direction is consistent with the drag direction and does not consider the influence of the angle of attack.In cruise conditions, the aircraft's angle of attack is small, so this method won't cause significant errors.However, at low speeds, the aircraft has a larger angle of attack, and the engine thrust will produce components in both the drag and lift directions.Therefore, when performing low-speed thrust-drag bookkeeping, there is a need to improve the current method to obtain more accurate results in low-speed conditions.The improved thrust-drag decomposition method is as follows: 1. Calibration of the engine velocity coefficient: Calculate the engine thrust Fc and velocity coefficient Cv under static air freestream boundary conditions.Using the whole aircraft configuration instead of just the nacelle for engine flow coefficient calibration provides a better assessment of the flow disturbances due to full aircraft integration.
The relationship between engine thrust and axial integrated force Fx can be obtained from the equation: Fx can be derived from CFD calculations.β and χ are the installation angle and the internal cant angle of the nacelle, respectively.
The calculation method for the velocity coefficient C V of the dual-flow nozzle is given by equation ( 2), where ̇  and ̇  represent the mass flow rates of the fan and the engine core, respectively, and    and    are the ideal expansion speeds at the outlets.
Due to convergence issues in actual calculations when using completely static far-field boundary conditions, a very low inflow speed is generally set in the computation to ensure numerical convergence.The velocity coefficient correction term D W in the market takes this inflow speed into account and can be obtained from equation (3).
The ideal expansion speeds    and    can be derived from equation (4).γ and R are the ratios of the specific heat and gas constant, respectively.Pt and Tt are the total pressure and total temperature of the nozzle outlet, respectively.2. Whole aircraft CFD calculation with powered nacelle: Compute the powered nacelle's whole aircraft CFD under low-speed conditions.Calculate the combined force L' and D' and the total pitch moment M' of all aircraft wall boundaries and engine intake and exhaust boundaries in the wind axis system.
3. Whole aircraft thrust-drag bookkeeping: The engine's total static thrust can be obtained from equation (5), where F N represents the total static thrust, treated numerically as a scalar.Where the far-field inflow speed is  0 , the engine inlet mass flow rate is ̇1.The flight angle of attack is α.The arm length of the thrust in the pitching direction is   .
( ) The total aerodynamic force of the aircraft can then be calculated using equation (6).
The aerodynamic force coefficient can be represented as in equation (7), where S REF is the wing reference area, and if a half-model is used for CFD calculation, S REF is the wing reference area of the half-model, and mac is the average aerodynamic chord of the wing.With this, the engine thrust-drag bookkeeping is complete.

Propulsion Aerodynamics Workshop (PAW) case verification
To validate the jet simulation method, CFD studies are conducted, and the results from different CFD solvers are compared with available experimental data from the first Propulsion Aerodynamics Workshop [18,19].The nozzle geometry is shown in Figure 2. Figure 3 shows the discharge coefficients of the CFD, experimental and analytical results plotted against the nozzle pressure ratio (NPR).In simulating the nozzle, S-Flow demonstrated strong consistency with experimental results, validating that the engine nozzle boundary modeling method used in this study is suitable for further analysis.

DLR-F11 case verification
We opted to validate our solver using the DLR-F11's config4 as our benchmark model [20,21].The DLR-F11 config4 configuration includes a full-span slotted wing, full-span slat, fuselage, slat root fairing bulge, five slat tracks, and seven slotted wing brackets.Its high Reynolds number test conditions are Ma=0.175and Re=1.51×10 7 .Using the commercial software ICEM-CFD, we generated an intricate structured grid for this complex configuration, totaling 50 million cells.The height of the first wallnormal layer was chosen to be 0.00037 mm, with a y + value of approximately 2/3.Its geometric configuration and the associated surface mesh are displayed in Figure 4.In conclusion, considering the complex geometry of high-lift configuration, the solver accurately reflects flow field characteristics with a high degree of precision.Therefore, it forms a reliable foundation for subsequent aerodynamic studies of take-off configuration in engine powered conditions.

Configurations
This study centers on a conventional twin-engine, double-aisle jetliner with a bypass ratio (BPR) of 10:1 and moderately swept wings.The cruising Reynolds number is approximately  .For the high-lift configuration, a droop nose device (DND) [22,23,24] is affixed to the wing's inboard leading edge, while a single-slot slat is positioned on the outboard.The wing's trailing edge incorporates two adaptive drooped hinge flaps (ADHFs) [25][26][27][28].Further outboard, two sets of ailerons are placed adjacent to the flaps.Figure 6 depicts the aircraft in its takeoff configuration.In this study, we've incorporated an inboard nacelle strake with meticulously chosen shape and installation parameters.CFD simulations indicate that the inclusion of the nacelle strake enhances the maximum lift coefficient by 0.33 and delays the stalling angle by 4°. Figure 7 displays the total pressure coefficient contours and particle trajectories both before and after the integration of the nacelle strake, corresponding to an angle of attack (AoA) of 14°.

The power effect on force coefficients
Utilizing the TDB method outlined in the preceding chapter combined with CFD outcomes, we can deduce the force coefficients for the take-off configuration with powered nacelle (TK-PN).A comparison of the lift force coefficients between take-off configuration with through flow nacelle (TK-TFN) and TK-PN is illustrated in Figure 8 (a).The power effect extends the stall angle by 3°.As a result, there's an enhancement in the maximum lift coefficient by 0.3.The power effect also notably amplifies the drag at elevated AoAs.Once the AoA surpasses 8°, an increase exceeding 100 counts in total drag is detected.Specifically, at a 14° angle of attack, power effects raise the drag coefficient of the take-off configuration by 275 counts, resulting in a 0.76 decrease in the lift-to-drag ratio, as showcased in Figure 8

The power effect on the wing
Figure 9 showcases the pressure coefficient distributions and surface streamlines on the inboard wing for both TK-TFN and TK-PN configurations at an AoA of 19°.
Figure 9. Distribution of pressure coefficients and inboard wing surface streamlines.The surface streamlines of the TK-TFN configuration demonstrate a flow separation on the inboard wing when subjected to a 19° AoA.This separation begins at the wing's leading edge and extends towards its trailing edge.On the other hand, the TK-PN setup exhibits no evident flow separation at the same angle.While past studies suggested that the variance in flow separation was attributed to the upper surface flow separation of the nacelle, our findings present a different narrative.In our research, there's no observed flow separation on the nacelle, and the upper surface flow dynamics for both TFN and PN are remarkably consistent.This leads us to infer that the nacelle's upper surface isn't the primary cause of the differing flow separations on the wing.It's plausible that earlier studies identified nacelle flow separation because they missed incorporating the nacelle strake in their numerical models.
Figure 10 presents the pressure coefficient distribution at a 36.39%spanwise location (the wing's kink position) with a 19° AoA.Depicted streamlines trace particles moving through the wing leading edge's low-pressure region.In the TK-PN setup, the flow pattern highlights that these particles traverse the engine exhaust's high-pressure region.From the above examination, it's discerned that the proximity of the close-coupled nacelle's nozzle to the wing's leading edge allows the engine's power effect to infuse additional flow energy.This explains the power effect's capability to deter flow separation and amplify lift.The influence of this flow acceleration is more pronounced at elevated AoAs, suggesting that the power effect can lessen flow separation in the post-stall region, thereby elevating the maximum lift for takeoff configurations.

Conclusions
Utilizing the RANS solver and a multiblock structured grid, we conducted a numerical investigation into the takeoff configuration equipped with a powered nacelle.The nacelle strake is meticulously integrated into the takeoff setting.A modified thrust-drag bookkeeping approach is introduced, optimized for low-speed conditions combined with high angles of attack, and this revamped method is incorporated into our analysis.The major conclusions drawn from our research include: 1. Based on experimental data from the PAW and DLR-FR11 cases, the numerical method adopted in this paper has been well-validated.It demonstrates a strong consistency with experimental outcomes, suggesting the reliability of the numerical method.2. Numerical analysis of the takeoff configuration indicates that in terms of the paramount lift characteristics, engine power effects heighten the lift coefficient and enhance the maximum lift coefficient.3.In terms of drag, the influence of engine power augments the overall drag of the aircraft.Taking an angle of attack of 14° as an instance, the overall drag has risen by 275 counts.4. The power effect serves as a deterrent to flow separation and wing stalling.During instances when the aircraft maneuvers at low velocities accompanied by steep angles of attack, the flow speed-up resulting from the engine exhaust interacts with the wing's leading edge.This accelerated flow is verifiable through the streamlines' trajectory and velocity shifts, proving its role in mitigating flow separation and postponing wing stall.5.The impact of power also refines the linearity of the longitudinal pitching moment at significant angles of attack.This enhancement is highly advantageous for the aircraft's maneuverability and stability characteristics.In essence, the thrust-drag bookkeeping method proposed, combined with the findings for the takeoff configuration, offers valuable insights and can be referenced during the design phase of takeoff configurations that factor in engine integration.In the next stage of our research, we need to closely examine how the powered nacelle affects each high-lift devices component.Additionally, when the time is right, we plan to test our numerical findings with TPS(Turbo Propulsion Simulator) wind tunnel experiments to make sure they're accurate.

Figure 3 .
Figure 3.Comparison of the numerical results to the experimental and analytical data.In simulating the nozzle, S-Flow demonstrated strong consistency with experimental results, validating that the engine nozzle boundary modeling method used in this study is suitable for further analysis.

Figure 4 .
Figure 4.The mesh of DLR-F11's config4.Figures 5 showcase the comparison between calculated and experimental lift and drag characteristics of the DLR-F11 config4.The graphs indicate that the calculations closely match experimental results, particularly in the pre-stall angle range.

Figure 5 .
Figure 5. lift and drag characteristics of the DLR-F11 config4.In conclusion, considering the complex geometry of high-lift configuration, the solver accurately reflects flow field characteristics with a high degree of precision.Therefore, it forms a reliable foundation for subsequent aerodynamic studies of take-off configuration in engine powered conditions.

Figure 6 .
Figure 6.Investigated take-off configuration.In this study, we've incorporated an inboard nacelle strake with meticulously chosen shape and installation parameters.CFD simulations indicate that the inclusion of the nacelle strake enhances the maximum lift coefficient by 0.33 and delays the stalling angle by 4°.Figure7displays the total pressure coefficient contours and particle trajectories both before and after the integration of the nacelle strake, corresponding to an angle of attack (AoA) of 14°.

Figure 7 .
Figure 7.The total pressure coefficient and particle traces of the nacelle.
(b).This trend aligns with wind tunnel test findings.With regard to the longitudinal pitching moment coefficient (visualized in Figure8 (c)), there's a noticeable rise for TK-PN, accompanied by enhanced linearity in the vicinity of the stall region.This modification augments the longitudinal static stability during takeoff.

Figure 10 .
Figure 10.Particle tracking and pressure distribution at the kink position.Figure 11 further visualizes the streamlines shown in Figure 10.For the TK-PN configuration, the Mach number on the wing's leading-edge up to 0.3.Relative to the TK-TFN setup, particles experience a significant acceleration after navigating the high-pressure engine jet area, signaling an added momentum from the jet.This rapid acceleration is absent in the TK-TFN configuration, emphasizing that the powered nacelle likely drives the flow acceleration.

Figure 11
Figure 10.Particle tracking and pressure distribution at the kink position.Figure 11 further visualizes the streamlines shown in Figure 10.For the TK-PN configuration, the Mach number on the wing's leading-edge up to 0.3.Relative to the TK-TFN setup, particles experience a significant acceleration after navigating the high-pressure engine jet area, signaling an added momentum from the jet.This rapid acceleration is absent in the TK-TFN configuration, emphasizing that the powered nacelle likely drives the flow acceleration.

Figure 11 .
Figure 11.Particle trajectories for TK-PN and TK-TFN.From the above examination, it's discerned that the proximity of the close-coupled nacelle's nozzle to the wing's leading edge allows the engine's power effect to infuse additional flow energy.This explains the power effect's capability to deter flow separation and amplify lift.The influence of this flow acceleration is more pronounced at elevated AoAs, suggesting that the power effect can lessen flow separation in the post-stall region, thereby elevating the maximum lift for takeoff configurations.

and
Operations Conference.[26] James U and Nhan N 2013 A Mission Adaptive Variable Camber Flap Control System to Optimize High Lift and Cruise Lift to Drag Ratios of Future N+3 Transport Aircraft 51st AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition.[27] Strüber H 2014 The aerodynamic design of the A350 XWB-900 high lift system 29th international congress of the aeronautical sciences.[28] Qiang J I Zhang Y Haixin C et al. 2022 Aerodynamic optimization of a high-lift system with adaptive dropped hinge flap Chinese Journal of Aeronautics 35(11) pp191-208.