Multidisciplinary Optimization of Aircraft Aerodynamics for Distributed Propulsion Configurations

Abstract

The combination of different aerodynamic configurations and propulsion systems, namely, aero-propulsion, affects flight performance differently. These effects are closely related to multidisciplinary collaborative aspects (aerodynamic configuration, propulsion, energy, control systems, etc.) and determine the overall energy consumption of an aircraft. The potential benefits of distributed propulsion (DP) involve propulsive efficiency, energy-saving, and emissions reduction. In particular, wake filling is maximized when the trailing edge of a blended wing body (BWB) is fully covered by propulsion systems that employ boundary layer ingestion (BLI). Nonetheless, the thrust–drag imbalance that frequently arises at the trailing edge, excessive energy consumption, and flow distortions during propulsion remain unsolved challenges. These after-effects imply the complexity of DP systems in multidisciplinary optimization (MDO). To coordinate the different functions of the aero-propulsive configuration, the application of MDO is essential for intellectualized modulate layout, thrust manipulation, and energy efficiency. This paper presents the research challenges of ultra-high-dimensional optimization objectives and design variables in the current literature in aerodynamic configuration integrated DP. The benefits and defects of various coupled conditions and feasible proposals have been listed. Contemporary advanced energy systems, propulsion control, and influential technologies that are energy-saving are discussed. Based on the proposed technical benchmarks and the algorithm of MDO, the propulsive configuration that might affect energy efficiency is summarized. Moreover, suggestions are drawn for forthcoming exploitation and studies.

1. Introduction

The current methods to cut energy use in new planes often rely on innovative propulsion systems. This includes blended wing body (BWB) designs, distributed propulsion (DP), and boundary layer ingestion (BLI) [1]. A NASA report shows that at least $70\%$ of carbon emissions from civil aircraft can be reduced by replacing fuel with electricity [2]. However, carbon emissions (CO₂₎ and energy consumption $\left(J\right)$ are two different physical quantities. Using different energy sources (fuel, electricity, or hybrid) does not reduce total energy use. One potential method for reducing energy consumption involves recharging the jet flow on the boundary layer into the free stream to backfill the overall wake of an aircraft. This approach aims to improve thrust efficiency and contribute to the goal of green aviation [3]. To backfill the overall wake of an aircraft, numerous small engines must be placed evenly along the trailing edge as part of the distributed propulsion (DP) system to achieve wake recovery effects during propulsion [4,5]. This technology is considered a revolutionary configuration for future aircraft propulsion systems. In 1929, the aviation industry began to test DP aircraft designs. Only recently, its refined design and research work have started as a new potential type of aircraft configuration [6].

In recent years, the National Aeronautics and Space Administration (NASA) has achieved numerous successful discoveries in aero-propulsive technology, such as propulsive fuselage, turboelectric DP, non-planar systems, BWB-DP, etc. [4,5,7,8,9]. Pioneering institutions such as the University of Cambridge and the Massachusetts Institute of Technology (MIT) have successively explored some advanced DP technology and its corresponding aircraft configurations, including the silent aircraft program and the energy sources of future aircraft [4,10,11]. The concepts of these configurations are mainly to reduce the noise generated by large turbine engines and to reduce harmful emissions generated by combustion processes. Further research in this sector demonstrates a feasible integrated scheme of distributed electric propulsion [12]. It means that the all-electric aircraft will no longer be a distant dream [9]. However, the specific total amount of energy storage, the hazards of conductive materials emission, and the technical challenges of the integrated layout have not yet been fully specified.

Among the numerous aero-propulsive configurations, the most revolutionary is undoubtedly the BWB-DP configuration scheme. The development of DP has initially been concentrated on large-width BWB. The purpose is to maximize the carrying capacity of an aircraft while reducing wake losses caused by regional thrust. The synergy of BWB-DP can improve the lift-to-drag ratio by approximately $20\%$. It has a large capacity volume, small surface friction drag, and low manufacturing cost [7,8]. A study found that a 16-propeller aircraft could reduce its take-off distance by at least $50\%$ compared to a 2-propeller one [13]. However, the total amount of energy consumption is unspecified. Moreover, there are still some technical difficulties with the integration of BWB with DP. For instance, gaps between circular propulsors cause wake loss and energy waste. Wake losses can be reduced with an expanded nacelle propulsion system, but more propulsors have to be disposed to cover the trailing edge of the BWB. This aero-propulsive configuration would exhaust more energies, which would eventually not be worth the candle. Therefore, some researchers shifted the focus of DP technology to the low laterally surface tube wing body (TWB) aircraft. Additionally, the performance of each configuration needs to be known, such as propulsion and energy efficiency, maneuverability, control capabilities, structural strength, and aero-elasticity [14,15]. The complex coupling between these variables involves multidisciplinary aspects. It is hard to analyze the performance without considering the effects of different disciplines.

Based on the energy-saving scheme, the existing DP works of literature are sorted and analyzed. The differences and advantages of the thrust features of DP systems in conventional (TWB) configurations and the futuristic BWB are listed. The current research trends, applications, developments, and problems in the aviation field of propulsion are presented. In light of the high-capacity requirements of future civil aircraft, this paper eventually focused on several technical challenges of aero-propulsive design in DP and BWB configuration. The application of multidisciplinary design optimization (MDO) and several unresolved dilemmas of the BWB-DP have also been listed. The development trend of the BWB-DP in energy-saving and the layout issues derived from this aspect are explored.

2. Development and Tendency of Aero-Propulsion

Conventional aero-propulsive designs consist of TWB and few concentrated engines. Novel aero-propulsive designs generally profit from the integration between aerodynamics and propulsion systems.

In the context of novel propulsion systems, distributed propulsion (DP) refers to maximizing the distribution of thrust across an aircraft to minimize wake losses caused by its configuration. This approach aims to achieve the optimal balance between aerodynamic thrust and drag [10,16]. In other words, the magnitude of kinetic energy and stream velocity must be consistent as much as possible after flitting over the trailing edge. The reason for arranging the propulsion in this way has been proven in marine propulsion, where inefficiencies in propulsion are found when wake losses occur at the trailing edge of a moving craft. This is due to an uneven thrust distribution on the trailing edge of the object [17]. When multiple propulsors are distributed, the differences in locations, Mach numbers, and exhaust port shapes might all be reasons for an uneven thrust distribution. When the thrust distribution is uniform, DP could provide more lift with less surface to increase the lift coefficient for a given flight speed and wing planform area [7,18,19,20]. The flight noise can also be highly reduced due to the propulsor size and the arrangement of boundary layer ingestion (BLI) [21,22,23].

For aerodynamics, we must evenly distribute the aircraft’s weight. This reduces energy loss between thrust and lift. In 1986, Liebeck re-explored BWB to reduce an aircraft’s weight [24]. This was a trend in developing future civil aircraft. The overly concentrated weight on a tube has been identified as one of the causes of uneven DP, leading to energy losses. A huge portion of drag on conventional aircraft comes from the middle aft. The conventional side thrust cannot cover it. In 2005, Schrauf G. illustrated that friction drag constitutes more than half of the total TWB transport aircraft drag. The percentages of friction drag in the fuselage (including the horizontal tail and fin), wings, and nacelles are 29%, 18%, and 3%, respectively. The idea of reducing drag on wing area by laminarity is possible, but it is impossible to distribute it in a fuselage with a very large Reynolds number [25]. As opposed to TWB, BWB could reduce takeoff weight by $15\%$, fuel consumption per seat per mile by $27\%$, empty weight by $12\%$, and increase the lift-to-drag ratio by $20\%$ [26]. It is believed that DP could cover the drag from uniform weight.

Figure 1 shows that regardless of which aircraft configuration is integrated with the DP, the problem of an imbalanced thrust–drag distribution still exists on the trailing edge. It is worth noting that most of the propulsors in active-duty TWB are placed on the wings, where the aft has typically been ignored. This configuration is probably the solution adopted after measuring the gains and losses of thrust and load in the presence/absence of aft wakes [24,27,28,29]. Assuming that the thrust is distributed evenly on the wings of the BWB and TWB, attributable to geometric issues, the thrust of the TWB has to be divided into three sections (two semi-span wings and a fuselage), which leads to a more complicated propulsor installation. Conversely, BWB can easily achieve a smooth integrated trailing edge on the Y-axis. In the absence of BLI, both aircraft configurations present a side with a potential severe wake dissipation at the Z-axis. Even if a circumferential propulsor is implanted on the trailing edge of TWB, the wake dissipation may occur at the exact center point of the circumferential propulsor, which is detrimental to flight performance. In short, the main purpose of DP is not to distribute the propulsors in fancy locations but to distribute the thrust of the aircraft’s trailing edge in the most coordinated and balanced possible way.

3. Performance and Coupling Effects of Aero-Propulsion

At present, it is relatively difficult to take profit from thrust distribution on TWB [30]. The problem is that when the thrust forces are equally distributed along its wing and aft, the speed and mass of the jet streams present big discrepancies due to their position at different parts of the aircraft. If a comparatively large thrust is centralized on the aft, a large amount of wake dissipation appears on both sides of the wing. Otherwise, when the larger thrust is concentrated on both sides of the wing, it is difficult for the small thrust at the aft to compensate for the wake dissipation caused by the larger thrust [27]. Since the performance of TWB aircraft, such as lift-to-drag ratio and fuel-saving, is already enhanced close to the limit, this problem can only be solved by modifying the aero-propulsive configuration.

In the case of the BWB, the distance difference between the wing and trailing edge is comparatively small, and it is easier to deal with the propulsion distribution [26]. Thus, it is not difficult to predict that integration of the DP system and BWB has great potential to become a general civil aircraft in the future [7,8,31].

3.1. Configuration Design of Distributed Propulsion

The DP system has a variety of configuration methods like thrust split, multi-fan common core, distributed exhaust system, etc. (refer to Figure 2) [5,10].

Table 1. Distributed propulsion in multi-fan common core and individual core systems.

Selected Aircraft	Engine Location	Number of Engines (Fans)	Mach Number	Wing Area (m2)	MTOM (kg)	Payload (kg)	Max. Seat	L/D (Cruise)	Range (km)	Cumulative Noise (EPNdB)	Mission Fuel (kg)
SAX-12	BLI (IPS)	4 (4)	0.80	-	154,290	-	215	21.90	14,816	-	-
SAX-40	BLI (MPS)	3 (9)	0.80	836	150,847	23,405	215	20.10	9260	185.0	33,253
N2A-EXTE	PA (IPS)	2 (2)	0.81	925	213,914	46,720	-	21.42	11,112	251.0	65,453
N2B	BLI (MPS)	3 (9)	0.80	846	216,545	46,720	-	21.55	11,112	274.4	66,587
N3-X	BLI (MPS)	2 (15∼21)	0.84	-	-	107,048	300	-	13,890	-	-

NOTE: PA = pylons-above-wing, BLI = boundary layer ingestion, IPS = individual core propulsion system, MPS = multi-fan common core propulsion system. Compiled from sources: [3,35,36,37,38].

shows the difference in mission fuel consumption and range of the well-known conceptual aircraft in regional/distributed propulsion, BLI/pylon-above-wing (PA), and single/multi-fan common core system (MPS). It can be seen that the aircraft with DP, BLI, and MPS (N2B) has not achieved a reduction in mission fuel or cumulative noise. These data show that flight performance improves slightly by reducing the wing area and Mach number. Therefore, this configuration is likely feasible. To achieve the benefits of wake filling, overlaying the entire trailing edge of the BWB with an overdraft amount of circular propulsors is unreasonable and energy-consuming [11,32,33,34]. The currently developed DP system has reached the limit in the effectiveness of BWB.

The cross-flow fans from FanWing Ltd (London, UK) can generate airflow. They evenly cover the entire trailing edge. Their blustering angle is ${90}^{\circ }$ different from the ordinary fan, consisting of a long strip rotor and blades encased in a customizable rectangular nacelle with an inlet and outlet [39]. This configuration completes the wake ingestion that the circular fan could not reach. It has a relatively complete wake filling, which provides a relevant improvement of the DP. Another promising DP configuration is the homeopathic jet flap. It takes advantage of the exhaust flow. It has a high-velocity thrust from a flap tip, especially when its installation is spanwise along the trailing edge. This could maximize lift and improve take-off and landing. On the other hand, the distributed exhaust system (DES) generally exhausts heat flow, some of which might be exhausted through the central axis of a fan duct connected to an engine. This means that the exhaust ducts allocated between fans are likely to cause an imbalance in the speed and thrust of jet streams between the central and side of the multi-fan common core system. In addition, the blended effect between the heat and cold flows from the exhaust and fan ducts on propulsion is still unknown [40].

According to the above inferences, a group of propulsors should be sequenced. This is based on the thrust needed at different positions of the aircraft’s trailing edge. If propulsors with the same thrust are rashly sequenced side by side, it would be hard to achieve the goal of reducing energy consumption [14,41]. Therefore, distributed thrust in a more balanced way to decrease drag is still a challenge.

3.2. Layout Features of Distributed Propulsion

The main target of the DP system is to provide a uniform distribution of thrust force along the trailing edge and to avoid empty zones as much as possible. When thrust is launching, the so-called thrust vacancy does not refer to the vacant part of the engine but refers to the wake dissipation that occurs after the trailing edge. When the geometry between the trailing edge and the wing sweep angle changes, the wake changes. Figure 3 shows a qualitative prediction of the wake corresponding to a launching regional propulsion system [42,43,44,45]. For a conventional aircraft, its thrust does not consider the state of power balance. Therefore, regardless of velocity or mass, its thrust condition tends to result in significant wake dissipation. To fix this, NASA developed STARC-ABL. It reduces excess propulsion a bit [43]. It is noteworthy that the thrust on its aft-mounted propulsors is lower than the propulsors on the wings to avoid greater velocity residuals due to the distance difference between propulsors; therefore, after an aft-mounted propulsor is installed, the velocity residual and wake dissipation can be decreased favorably. However, it is undeniable that the velocity residual will still be more or less continuously caused by the large span from wing to aft. This situation is also one of the reasons that the aft-mounted propulsor is difficult to use as a collaborative engine.

3.3. Aero-Propulsive Coupling Effects

Aero-propulsion could gain from coupling aerodynamics with the thrust stream in certain conditions [46]. Specific types of equipment with low defect rates that could significantly impact aero-propulsion are discussed (refer to

Table 2. Benefits and defects of Aero-propulsive configuration in specific conditions.

Ref.	Section	Specific Condition	Benefits	Defects
[48]  [49]	Fuselage/ body	TWB	★ Flight performance is easier to discriminate by     accumulated long-term experiences.	★ The fuselage has to be elongated to embed     a circle aft propulsion system, resulting     in higher costs with low benefits.
			★ No maneuverability and integration problems;     the division of operation for each unit is clear.	★ The non-integrated “body" is difficult to     deploy DP systems.
[50]		BWB	★ The load distribution is more uniform and     significant in load advantages. ★ Easier to implant DP in an integrated platform.	★ Maneuverability problems have to be addressed     for aircraft without rudder and flaps. ★ The interaction between wing, control     surfaces, and DP adds complexity to the     design.
[51]  [52]	Wing	Jet wing	★ Reduces lift-induced drag. ★ The pitching moment of blades could change     the direction and magnitude of the thrust vector.     It could replace flaps and slats to improve flight     performance in stability, controllability, and safety.	★ In the case of replacing a rudder with wing     jets, when the energy transmission process     cannot be started on time, there is a     dilemma that the control wheel cannot be     driven.
		Jet flap	★ It provides an active circulation control of the     airfoil through the suction-side blowing. ★ Ejects high-velocity air from a downward slot     so flaps can obtain a high lift.	★ It has a thinner exhaust port with lower     exhaust flow volume, the effect is not as     significant as other methods.
[53]  [54]		Split wing	★ Ingesting low-momentum air from the boundary     layer can significantly reduce the required     kinetic energy, while maintaining the required thrust     velocity and thus increasing propulsion efficiency. ★ The truss structure of the split-wing provides     adequate load-bearing capacity. It could     significantly reduce the wing weight and     bending moment. ★ It could garner a high lift coefficient via powered-     lift.	★ The inherent complexity of inboard     structures in DP systems must be     considered.
[55]  [56]		Under wing	★ Can improve the aerodynamic efficiency of the     L/D ratio.	★ The under-wing duct is susceptible to foreign     object damage, and its noise is relatively large.
[57]			★ Provide additional bending relief to the main     wing structure.
[58]  [59]		BLI	★ Reduced the ram drag at the engine inlet and     improved propulsive efficiency (energy-saving). ★ Produce wake filling to reduce the required     thrust and energy, further reducing emissions. ★ Effectively balances the CG and offsets part of     the load. ★ A noise-shielded configuration.	★ Anergy comes with energy. It is almost     impossible to achieve complete wake filling. ★ The straight duct does not achieve an actual     function of BLI, while flow distortion     occurs on the embedded S-duct.
[60]		Pylon	★ Provides an appropriate external surface to     allow the intake and exhaust integration,     access for inspection, removal, and     maintenance requirements.	★ Wetted area and load are increased,     and the nose is downward at the     thrust moment. ★ A drastic reduction in CG range     happens via displaced thrust axis. ★ Have a high static instability.
[61]  [62]	Trailing edge	Parallel	★ A parallel trailing edge with nearly 0° of wing     sweep makes it easier to arrange the DP to attain     momentum balance and reduce wake losses.	★ Low L/D ratio.
[63]		Angular	★ The trailing edge with an angular back sweep     Wing has a higher L/D ratio.	★ The angular trailing edge makes it difficult to shape     DP in alignment. It becomes more complex     to allocate the thrust speed to meet the     momentum balance.
		Circumfere- ntial	★ Removing a significant share of fuselage drag     could substantially increase the L/D ratio. ★ Provides BLI to serve the wake filling (energy-     saving.)	★ To account for the axisymmetric contraction     of the aft towards the fan inlet, fuselage     length has to increase by 2.0 m. ★ The horizontal tail needs to be increased to     balance the pitching moment of an aircraft;     it leads the fuselage to a bending moment,     and the weight is increased greatly.
[48]  [49]	Propulsor	Regional	★ The Mach number is generally ≈ 0.85, and the     number of engines is less than 4. The thrust     force is concentrated and easier to estimate.	★ Wake losses occur in places without thrust,     resulting in high energy consumption.
[50]  [64]		Distributed	★ A system with more engines has redundancy for     increased safety and reliability. ★ Noise can be significantly reduced due to the     low bypass ratio of the engine. ★ Smaller and lightly changeable propulsors     could improve affordability and reduce the     life-cycle cost by up to 50%.	★ Have to ensure that the wake filling is filled     precisely and not excessively. ★ Mach number is between 0.75 and 0.8, and     more than 4 engines. The more dispersed     the engines, the more the propulsion effect is     significantly reduced.

). Generally, the unconstrained geometric BWB can equip DP arbitrarily. Its thrust can be adjusted according to the position requirement of the discretely distributed multi-vector propulsors to achieve synergy effects and thrust balance on the overall trailing edge. This thereby reduces the crisis caused by an engine stall and also improves the safety and reliability of the aircraft. The advantage of this synergy can improve the aerodynamic characteristics, propulsion efficiency, and aero-elasticity of BWB [14,15]. As for TWB, a well-known example is the encircled aft propulsor of the STARC-ABL, which is designed to re-energize airflow from the boundary layer and release thrust along the wake generated by the fuselage [47]. This setup favors regional propulsion over DP, since its thrust is eventually concentrated on the “tube”. Moreover, although the thrust of ECO-150 is distributed well on both wing sides, its crucial aft is not equipped with a centralized thrust. If it has an aft-mounted propulsor, the residual velocity might be severe on both wing sides. This propulsion configuration has higher wake loss than conventional aircraft.

In the wing part, BLI propulsors mainly absorb and reduce the energy loss of the wake. They use less fuel to produce the same thrust at the same flight conditions. This improves fuel efficiency and cuts energy costs. Since the BLI engine is now closer to the central axis after being embedded in the wing, it could reduce the extra pitching moment from the pylon propulsor. This would cut the control surface size, skin friction drag, and balance the thrust–drag requirements.

Additionally, BLI with an embedded duct could encapsulate a high bypass ratio engine in a body, which reduces and shields engine noise from the wings. The shape design parameters that could directly affect the mentioned performance of BLI include the width to height ratio, the shape of the embedded nacelle, the bump air inlet, and swept lip technology [65].

Regarding the propulsion system, the thrust distribution of the DP system with smaller engines can contribute to increase the maximum take-off weight for a large thrust-to-weight ratio. A systematic research study from BWB Boeing company shows that aft-mounted propulsion could save 3–5% of fuel consumption compared with the pylon-mounted propulsion [5,7,8,47]. Due to the small size engine, a profusion of high-strength composite materials can be used to reduce engine weight and save manufacturing costs. However, it also faces some technical challenges, such as low Reynolds number flow effects, strict engine manufacturing tolerance, and lower combustion efficiency [66]. Other factors that might affect the structure or mechanical design of small engines include high rotational speed, which may exceed 2 million r.p.m. A rotational speed this high not only demands non-lubricated air bearings but also introduces numerous new challenges in the fields of aerodynamics, materials, and manufacturing.

The forenamed various aero-propulsive configurations (listed in Table 2) can be combined to improve further propulsion efficiency and flight performance. NASA N3-X is a model that precisely integrates the BWB, BLI, jet wing, and parallel trailing edge. Its merits are extracted and evolved from three former models of N2A, SAX-40, and cruise-efficient short takeoff and landing (CESTOL). It has the characteristics of high controllability, high wake filing, low noise emission, and high fuel efficiency. From the predictions of NASA, its reduction of NOx in the take-off and landing phases could be up to $85\%$ of the Tier 6-CAEP/6 standard. Furthermore, the cumulative noise can be reduced by a minimum of $12EPNdB$ compared with their N+3 environmental goal [44,67]. The dilemma is that this model is an all-electric-driven aircraft, and the current development situation of all-electric-driven aircraft is still immature. Additionally, the storage space and electric current conversion between AC and DC do not yet have the necessary performance to convert the output instantaneously. On the other hand, DP consists of multiple propulsors, which means the conventional energy loops that should be integrated with DP might also face the supply–demand problem on multiple propulsors instantaneously. From the comparison and detailed description in Table 2, it can be found that the DP and BLI systems can also be integrated with TWB. A parametrization study on the power system of propulsive fuselage indicates that the potential of energy-saving in an identical fan inlet area is up to $10\%$ compared to a conventional podded power system. The fuel combustion of an economic and design mission of the turbo-electric concept can be reduced by $7\%$ and $12\%$, respectively [63]. However, it is not easy to implement the characteristics of small engines into the TWB since it is a concentrated tube body. The only possibility is to place the DP at a position with low thrust–drag areas (wing), but this may result in excess thrust.

4. Multidisciplinary Analysis and Optimization in Aero-Propulsive Design

Nowadays, most aero-propulsion systems are designed using the multidisciplinary analysis and optimization (MDAO) procedure. After the global analysis, the integration and optimization of the aerodynamics layout is conducted to figure out an efficient coupling configuration to achieve efficient wake filling and to reduce the mutual interference between the components [8,22,68]. Despite all this, arranging the DP system on the TWB wings would increase unnecessary energy consumption, while wake loss cannot be covered without adding a DP system on the wing portion. This situation puts the TWB at a standstill in balancing thrust and drag. Meanwhile, the BWB aligns well with the integration needs of DP. This section describes the case analysis of BWB-DP, which has the most potential for future civil aircraft. The content includes the schemes and solutions of excellent integrated bodies, layout features, MDO tools and calculation methods, etc.

4.1. Interaction of Aero-Propulsion

To ascertain the interaction and aero-propulsive performance of various configurations, 59 BWB civil aircraft with DPs were selected for analysis (refer to

Table A1. Enactment of distributed propulsion in the BWB conceptual aircraft.

Ref.	Year	Selected BWB	Engines Location	Thrust Split	Number of Engines (Fans)		Mach Number	Wing Area (m2)	MTOM (kg)	Payload (kg)	Fuel (kg)	Pax	L/D (Cru)	Range (km)
[8]	2011	BWB-450-1	PA	No	4	(4)	0.85	1245	390,920	47,700	142,578	480	23.90	14,353
[8]	2011	BWB-450-2	PA	No	8	(8)	0.85	1258	420,929	47,700	166,007	480	24.23	14,353
[8]	2011	BWB-450-3	BLI	No	8	(8)	0.85	1248	410,432	47,700	159,442	480	24.90	14,353
[8]	2011	BWB-450-4	BLI	No	8	(8)	0.85	1248	409,602	47,700	158,628	480	24.92	14,353
[8]	2011	BWB-450-5	BLI	No	8	(8)	0.85	1249	399,603	47,700	156,947	480	24.54	14,353
[8]	2011	BWB-450-6	BLI	No	8	(8)	0.85	1258	414,230	47,700	162,579	480	24.39	14,353
[8]	2011	BWB-450-7	BLI	No	8	(8)	0.85	1260	420,712	47,700	167,787	480	24.52	14,353
[8]	2011	BWB-450-8	BLI	No	8	(8)	0.85	1243	390,699	47,700	144,412	480	24.28	14,353
[117]	2007	CESTOL-3000 nmi	BLI	No	12	(12)	0.80	444	85,793	18,144	20,003	170	-	5556
[117]	2007	CESTOL-500 nmi	BLI	No	12	(12)	0.80	444	71,610	18,144	5821	170	-	926
[117]	2007	CESTOL-150 nmi	BLI	No	12	(12)	0.80	444	69,325	18,144	3988	170	-	278
[82]	2014	FW-UWE	UW	No	2	(2)	0.85	-	181,437	18,144	84,000	256	20.00	16,000
[118]	2011	BWB-350	BLI	No	18	(18)	0.85	-	232,000	-	-	350	-	11,112
[22]	2016	BW-11 mission	BLI	No	2	(2)	0.85	1391	481,148	65,115	193,821	555	-	14,173
[83]	2001	Hybrid layout-IWB	PA	No	4	(4)	0.85	1588	518,910	103,419	280,000	975	24.5	14,168
[83]	2001	Lifting-body	UW	No	4	(4)	0.80	1670	583,774	121,294	-	940	-	10,000
[35]	2006	SAX-12	BLI	No	4	(4)	0.80	754	154,290	-	74,224	215	21.9	14,816
[36]	2007	SAX-40	BLI	Yes	3	(9)	0.80	836	150,847	23,405	33,253	215	20.1	9260
[119]	2014	BWB	PA	No	2	(2)	0.85	748	186,970	22,680	-	224	-	14,816
[120]	2015	BWB100-02	PA	No	2	(2)	0.78	254	54,295	12,882	9752	105	22.5	926
[120]	2015	BWB100-03	PA	No	2	(2)	0.78	261	54,295	12,882	9752	105	23.3	926
[120]	2015	BWB100-10	PA	No	2	(2)	0.78	236	52,345	12,882	9979	105	21.4	926
[120]	2015	BWB100-12	PA	No	2	(2)	0.78	248	52,345	12,882	9979	105	24.3	926
[120]	2015	BWB300-00	PA	No	2	(2)	0.84	797	348,586	63,957	125,781	310	25.8	11,112
[120]	2015	BWB300-02	PA	No	2	(2)	0.84	892	348,586	63,957	125,781	310	29.3	11,112
[121]	2014	BWB100-1	PA	No	2	(2)	0.78	261	54,975	15,014	13,608	100	23.00	926
[121]	2014	BWB160-1	PA	No	2	(2)	0.79	371	98,566	26,762	21,092	160	26.60	1852
[121]	2014	BWB220-1	PA	No	2	(2)	0.80	594	196,224	60,872	73,028	220	28.90	5556
[121]	2014	BWB300-1	PA	No	2	(2)	0.84	880	375,030	317,742	138,346	300	30.00	11,112
[121]	2014	LFC100-1	PA	No	2	(2)	0.78	290	53,841	14,969	13,608	100	24.00	926
[121]	2014	LFC160-1	PA	No	2	(2)	0.79	460	59,073	26,308	21,092	160	27.90	1852
[121]	2014	LFC220-1	PA	No	2	(2)	0.80	709	201,577	61,643	73,028	220	30.30	5556
[68]	2018	BWB	PA	No	3	(3)	0.85	900	435,617	59,000	142,622	450	-	15,742
[122]	2016	VELA 3	UW	No	4	(4)	-	-	635,029	124,284	312,072	750	22.10	-
[123]	2013	OREIO	PA	No	3	(3)	0.80	744	475,800	45,359	57,611	-	23.30	12,038
[124]	2011	N2A	PA	No	2	(2)	0.79	859	209,333	46,720	63,458	-	21.61	11,112
[124]	2011	N2B	BLI	Yes	3	(9)	0.80	846	216,545	46,720	66,587	-	21.55	11,112
[124]	2011	N2A-EXTE	PA	No	2	(2)	0.81	925	213,914	46,720	65,493	-	21.42	11,112
[125]	2017	ACFA	PA	No	2	(2)	0.85	-	363,781	45,359	-	450	-	13,334
[126]	2012	BWB	PA	No	4	(4)	0.82	1051	405,900	122,100	151,000	-	25.40	15,403
[127]	2011	ERA-224	PA	No	3	(3)	0.85	772	202,756	45,545	56,246	223	23.53	-
[128]	2012	HWB98	PA	No	2	(2)	0.78	-	45,813	9800	5906	98	20.70	4445
[128]	2012	HWB160	PA	No	2	(2)	0.78	-	75,342	17,128	10,369	160	23.10	5325
[128]	2012	HWB216	PA	No	2	(2)	0.80	-	130,136	20,185	35,349	216	22.30	12,223
[128]	2012	HWB301	PA	No	2	(2)	0.84	-	246,119	53,569	66,506	301	23.50	13,890
[128]	2012	HWB400	PA	No	3	(3)	0.85	-	314,612	67,059	67,064	400	23.70	10,742
[129]	2018	Ascent 1000	BLI	No	2	(2)	0.80	-	46,802	17,091	7983	112	21.60	5926
[84]	2014	Aft-mounted	BLI	No	4	(4)	0.82	1091	406,000	66,400	149,000	400	27.20	14,400
[84]	2014	Wing-mounted	UW	No	4	(4)	0.82	1043	406,000	66,400	157,000	400	23.70	13,600
[84]	2014	Forward-swept	UW	No	4	(4)	0.82	1074	406,000	66,400	183,000	400	22.50	16,500
[4]	2016	Baseline BWB	UW	No	2	(2)	0.85	937	203,675	33,309	64,814	350	-	14,816
[124]	2008	BWB	UW	No	4	(4)	0.85	-	670,000	100,110	132,920	700	-	14,200
[130]	2010	H1	BLI	No	4	(4)	0.83	540	171,287	18,371	25,368	180	22.10	5556
[130]	2010	H2	BLI	No	4	(4)	0.83	778	279,950	44,707	11,126	256	24.40	15,372
[130]	2010	H3	BLI	No	4	(4)	0.83	941	470,566	60,977	126,159	354	24.10	14,075
[131]	2018	BWB150	PA	No	2	(2)	0.78	308	72,000	20,500	13,320	156	18.00	3917
[131]	2018	BWB250	PA	No	2	(2)	0.80	575	170,000	43,800	46,580	248	19.40	7264
[131]	2018	BWB400	PA	No	2	(2)	0.84	818	262,000	64,000	66,555	368	19.10	6700
[132]	2015	baseline HWB	BLI	No	2	(2)	0.84	-	243,253	29,030	-	300	21.60	4066

). Configurations with fewer samples, such as jet wings, have been excluded. The wing areas of these BWBs are from 236∼1670 m². Three types of propulsor locations are included. They are BLI, pylon-above-wing (PA), and under-wing (UW). The numbers of cases are 20, 32, and 7, respectively. Among them, there are two thrust-split samples, both of which are driven by three engines and nine fans. The number of propulsors is from 2∼18, and the Mach number of each propulsor for each BWB is systematic. In all cases, it is less than or equal to 0.85.

Figure 4 shows numerous scatterplot matrices about the integration of BWB and DP systems. Orthogonal correlation plots show that BLI’s performance is closely linked to PA and UW. However, PA and UW are only weakly correlated with each other. The figure shows that each sample has pros and cons. None is much better than the others. The DP location, which has provided a long flight range, high L/D ratio, and bearing capacity, requires more fuel, a higher Mach number, and a larger wing area.

4.2. Wake Ingestion and Propulsion Technologies

The operation mode of BLI technology starts when the low-energy boundary layer flow through the fuselage is absorbed into the engines of the DP with BLI. BLI technology mainly absorbs and reuses the stream generated by the trailing edge to reduce the energy loss from the wake and then produces the same thrust with less fuel consumption in the same flight condition. Applying BLI technology in some specific conditions can reduce fuel consumption by 5.5–10% [69,70,71,72]. When the BLI technology is integrated with ultra-high bypass turbofan engines, it can improve fuel efficiency by min. $3\%$ to max. 9% [67,73,74]. The performance of this technology (BLI) can be raised dramatically by 25–30% when the propulsors are placed at the trailing edge [75,76]. The above studies are from a partial research study based on BLI. An independent BLI could fully cover its drag to achieve high efficiency and energy-saving. However, when multidisciplinarity is considered, the expected results cannot be achieved. When the DP is disabled to cover the wake seamlessly and achieve the thrust–drag balance, these BLIs would continuously disperse propulsion and energies. Such uncoordinated thrust–drag is unable to achieve the expected favorable advancement effect (as shown in Figure 5).

On the other hand, when the airflow into the engine becomes distorted, the performance of the blades is degraded under additional pressure and a fatigue scenario [69]. The aerodynamic blockage related to the boundary layer in the body is much larger than that caused by the channel boundary layer [70]. The non-uniformity circumstance affects the thrust momentum at the exit of the nozzle, which is much greater than with a conventional propulsion system [77]. In the BLI outlet, the most important objective is to arrange the propulsor position until the thrust achieves a balance on the horizon of the trailing edge. When part of the thrust is converted from energy into “anergy”, this leads to a dilemma, which has a negative impact on the propulsion operation [41]. This is no longer a problem that can be solved by unilateral research and operation. The trailing edge must be precisely designed and calculated. This is to ensure the balance of thrust and drag.

Nonetheless, a limited number of propulsors could not compensate for the wake caused by excessive wing area. These conflicts will lead to the stagnation of DP technologies in fuel efficiency, energy conservation, and carbon emission reduction. A summary of the propulsive layout that affects DP performance is as follows:

(i)

Propulsion layout: In terms of the wingspan samples, distances between propulsors are all arranged in an even manner. Their thrust distributions are not arranged by considering the drag on each domain of an aircraft. This arrangement would cause a high drag domain (the central axis of an aircraft) to propel with low thrust and a low drag domain (the domain farther from the center axis) to propel with high thrust. If propulsor location is not considered according to the magnitude of the drag, the thrust might be dispersed and inadequate. It would cause the trailing edge to fail to reach the state of thrust balance, further causing the wake to dissipate. Ultimately, more propulsors would be requested, leading to higher energy consumption;

(ii)

Dilemmas in aero-propulsive integration: From the individual BLI local optimization to the aero-propulsive overall optimization, the difficulty of simulating and optimizing the performance of an overall aircraft is still sustained. The first reason is that local optimization results would mismatch the local optimization results of the overall optimization results. The second is that the computational methods and tools used to execute the overall aircraft optimization are more complex than the process for local optimization. In addition, the performance analysis of multidisciplinary objects is inherently flawed, such as the flight range, maximum take-off mass (MTOM), and fuel usage. The designer could adjust these at their discretion. Raising one factor’s performance to be advantaged and lowering the other two, or maximizing two of them and lowering one, can generate an illusion for readers that performance has been improved. Most proposal data are adjusted by optimization, in which the essential problem has not been solved, unless three of those performances are significantly better than others. Otherwise, advanced algorithms or MDAO must be performed on the displayed data.

4.3. Propulsion Controlled Aircraft

To achieve complete wake filling, almost the entire control surface (rudder, aileron, elevator) of an aircraft (BWB) should be replaced by a DP system. This would make control operations very difficult. A thrust modulation control system, namely, propulsion controlled aircraft (PCA), has to be applied to manipulate the magnitude of forwarding drive forces acting on the aircraft to achieve the effect of propulsion control. PCA could also manipulate the thrust vector on each propulsor to produce the required moments and accelerations. As the number of propulsors increases, the force and moment of distributed PCA acting on the aircraft also increases [78].

The use of DP to control the direction and thrust magnitude of aircraft to substitute conventional control surfaces would increase energy consumption. Additionally, the stability and controllability of airworthiness must be satisfied, particularly in the case of BWB, which initially has stability problems. It is widely accepted that redundant propulsors would help. They would address stability issues in the BWB design and may improve control with thrust vectoring. However, the interaction between this coupling system and the aerodynamics of the aircraft configuration would increase sophistication. It requires more robust intelligent engine control systems with adaptive thrust modulation and thrust vector to provide sufficient stability and controllability [79]. As of today, no one has studied propulsion control as an MDO factor in the BWB-DP system. It is conceivable that the level of complexity is the current difficulty.

4.4. Frameworks of Multidisciplinary Optimization

The concept of MDO is to integrate the knowledge of various disciplines in the whole process of complex system design, to fully consider the interaction and coupling between disciplines for obtaining the overall optimal solution of the system [77]. Generally, the design constraints of MDO refer to the mission requirements of the subject, that is, the final target of an aircraft, such as total energy consumption, flight range, etc. The design variables refer to item values that could affect the development of the indicator results, such as the area of an aircraft or the number of propulsors [80,81]. Figure 6 shows the different MDO frameworks in conventional and reform (regional/distributed) propulsion systems in TWB and BWB.

In analytical aspects (reference collection from Table A1), a small number of sample cases completed the scheme configuration by experimental methods [82,83]. Most of the cases are designed using the parametric model. After the propulsion-airframe scheme is established for these samples, the gradient-based optimization method is used to obtain the optimal model. In optimization (reference collection from Table A1), the models with more energy-saving and higher flight range have a MTOM between 400,000 and 440,000 kg. Prakasha et al. arranged three propulsors in the PA position. The configuration scheme is designed by using a general parametric aircraft configuration scheme (design space exploration). The model-based agile framework is used as an MDO tool. The results show that the fuel/range ratio is 9.06 kg/km [68]. Van Dommelen et al. installed four propulsors at the UW position. The parametric model is used to create the initial configuration, and the multi-model generator module is applied to build the final model. Their best fuel/range ratio result is 11.09 kg/km [84]. L. Leifsson applied eight propulsors at the BLI location. A parametric model with a few design parameters is drawn for the initial configurations. The MDO framework is built with a hierarchical strategy, and the final model is calculated through normal-boundary intersection. The optimal result of fuel/range ratio is 11.69 kg/km [8]. Most of these cases listed in Table A1 belong to multi-variate optimization. Its analysis algorithms are generally Reynolds-averaged Navier–Stokes equations, the vortex-lattice method, positive normal-boundary intersection, etc. These algorithms can obtain more significant optimization results. Comparatively, the demand for computing capacity is relatively big and requires a long computing time.

Since many disciplines are involved in the multidisciplinary optimization of DP aircraft, its performance evaluation has considerable complexity. The integration of aero-propulsive systems involves the coupling of aerodynamics, propulsion, structure, weight, stability, and control. These disciplines make multidisciplinary coupled analysis and collaborative optimization the core of DP aircraft design. Although research works on DP aircraft have been carried out, most of them are focused on configuration design. The exploration of the related energy-saving in integrated status and their outcomes, such as thrust distribution criterion, theoretical modeling, algorithms of large scale, and multi-level nesting MDO for DP aircraft configuration design, are still rare.

5. Research Challenges and Future Trends

By means of the comprehensive description and analysis of DP integrated with aircraft, it can be found that some functional conflicts need to be resolved in terms of energy-saving and propulsion control operations. Since most components interact with each other, they might produce coherent negative coupling effects. In multidisciplinary optimization, it is a challenge to distinguish between factors that can be processed independently and those that require coupled processing. Therefore, the problem has to be solved through more complicated MDO processing methods. The following sections first describe the conceptual and scheme issues to be solved before dealing with optimization problems. The methods and challenges that may reduce DP energy consumption are illustrated hereafter. Then, the impacts and future trends of independent and coupling processing components in MDO are discussed.

5.1. Scheme of Aero-Propulsive Configuration

Optimization cannot be made out of nothing;an idea that does not exist cannot be optimized. In the case of satisfying the node requirements, optimization can modify a square to a more suitable shape (circle) for rolling, but it is unable to turn a circle into a donut. The actual mechanism of DP would be limited if the general propulsor is composed and conceived as a DP system. Nowadays, DP is regarded as a scheme of dividing a large propulsor into several propulsors [8]. In fact, DP should not be limited by quantity. As long as the wake can be completely filled by thrust without generating excess wake, it can be a perfect DP. An ideal DP is one filling the wake at the trailing edge of an aircraft by nearly $100\%$ [85]. However, when thrust is activated, a residual called “anergy” is leftover. This phenomenon causes the launch energy to not be fully converted into thrust. Furthermore, the circular propulsors are not in contact with each other when juxtaposed to form a DP [41,86,87]. Thus, achieving perfect thrust–drag balance with standard propulsors is unrealistic.

The cross-flow fan proposed by FanWing Ltd could produce airflow and evenly cover the entire trailing edge. This configuration does not present discontinuous thrust distribution. Its shape and energy method also completely subverts the conventional form [39]. Currently, there are two difficulties in implementing cross-flow fans on BWBs. One is that it has not yet met the thrust requirements of large size BWBs. The second is that its thrust can only be output monotonically and uniformly through the elongated cross-flow fan, which does not solve the problem of the thrust–drag balance between the shape and DP of a BWB. From the perspective of a conventional DP layout, the position and distance of the propulsors, exhaust nozzles, fan distribution, jet flow, and control systems affect the thrust magnitude and thrust–drag balance in the different areas of an aircraft. A feasible approach that can improve these variables to improve propulsion efficiency, noise reduction, and flight range is to adjust the thrust–drag balance using MDO.

5.2. Flow Distortion Optimization

The BLI airflow inlet determines the smoothness of thrust. Excessive flow distortion may cause energy loss and compressor instability, which leads to propulsor malfunction. This problem can be solved by carefully designing the S-duct of intake flow ahead of the compressor. In appropriate circumstances, the use of a straight duct could avoid flow distortion problems more efficiently. It can also be addressed by active or passive flow control with boundary layer excitation to avoid the corresponding separation loss and flow distortion while reducing the adverse effects [72,88,89]. Flow control technology can change a fluid’s flow. It does this by applying force, mass, heat, or electromagnetic energy to the moving fluid. This alters its force or motion. It can improve the flow distortion in BLI intake flow and shorten its length to reduce the surface area and skin friction drag, reducing the structural weight accordingly.

Generally, there are four control methods to improve the airflow distortion of the engine inlet, namely, the vortex generator, pulsed jet, synthetic jet, and oscillating surface. These methods are distributed on the inner wall of the intake duct to change boundary layer flow and overall flow in the duct, thereby improving the airflow distortion of the flow before duct compression (refer to Figure 7) [72,90,91]. Currently, a control method combines these four physical quantities. It is the cross-flow fan. This method uses electric power and has not been practically studied as a BWB propulsion system. Its uninterrupted propulsion area possesses the potential to cover wake losses. However, the extent of its thrust and energy consumption still needs further improvement. Combinations of these methods are complex and varied.

The current optimization method for the duct geometry is generally to optimize a duct after screening out the effective scheme for the designed aircraft, such as using an adjoint method to optimize and reduce the flow distortion of an S-duct geometry [92]. However, as mentioned in the previous sections, the results of composite local optimization do not meet the needs of the overall situation of an aircraft. A high fidelity (MACH) MDO framework for aircraft configuration is used in Kenway et al. works [49]. They adjusted and optimized the straight duct in an integral TWB (STARC-ABL). The result shows that the optimal distortion for a circumferential duct of STARC-ABL is approximately double that of the duct. These examples are all cases of local duct optimization or straight duct optimization. A framework for targeting the S-duct of an unconventional propulsor (elongated cross-flow fan) of BWB-DP as a large-dimensional MDO objective has not yet been revealed.

5.3. Application of Energy Methods

Balancing thrust and drag at an aircraft’s trailing edge could reduce energy loss, improve the L/D ratio, and shorten take-off distance. Other than reforming the propulsion system to achieve the purpose of thrust–drag balance. A choice of energy system could also indirectly change the propulsion configuration.

So far, the energy types available for aircraft include mechanical, pneumatic, electrical, and hybrid power systems (see Table 3) [93]. The configuration and layout of the DP changes as the energy system changes. Mechanical and pneumatic systems are common energy systems that have more emission products. Despite that, new liquid hydrogen pneumatic energy systems are very eco-friendly and energy-saving. An advantage of using an electrical system is that there is no need to generate power through combustion. In principle, this propulsion system does not need to conform to the conventional configuration. It means a seamless elongated cross-flow fan could be constituted. However, the disadvantage of using an electrical system is having a complex configuration and the need to invert gravimetric energy from low to high density (DC to AC) [94]. This series of operations ends up providing only a small amount of power [95]. Furthermore, the emission products of the power system depend on the power generation resources. Chemical resources with low emissions have the potential to produce substances with significant environmental hazards.

A novel and promising approach to energy is the electrohydrodynamic system, which offers advantages in terms of emissions and configuration. It is a method by which propulsion can be generated in a fluid. Specifically, its propulsion is generated by high-frequency, high-voltage initiation of electrons in the surrounding air (Corona discharge), ionizing the air and accelerating the ions by Coulomb force. This method presents a continuous propulsion system with no transfer components. Due to the thinness of the energy system, it is not complicated to configure different output powers in different portions of a “body" to achieve an effect of thrust–drag balance. At present, the thrust and flight speed of this system is relatively weak and low. It is now only available in a limited lightweight aircraft model [96]. The choice and implementation of a novel energy system on large aircraft is a research challenge that would lead to the future development trend of the propulsion system.

Table 3. Current and future promising propulsion energy systems.

Ref.	Energy Sources	Consumables	Configuration	Gravimetric Energy (MJ/kg)	Composition Product	Emission Product
[97]	Mechanical	Gasoline	Jet fuel, power	$46.8$	$2C_8{H}_{18}+25O_2$, $N_2$	$71\%C{O}_2$+ $28\%H_{2}O$
[98]			turbine, fan.		$+O_2$	$+N{O}_x+S{O}_x+CO+$
						$UNC+Particulate$.
[99]	Pneumatic	Liquefied Natural	Gas Generator, jet	$55.5$	$C{H}_4+2O_2$	$C{O}_2+2H_{2}O$
[100]		Gas (Methane)	fuel, power turbine,
		Liquid hydrogen	gearbox, fan.	142	$2H_2+O_2$	$2H_{2}O\left(Zeroemission\right)$
[101]	Electrical	Lithium-ion	Battery, DC/AC	$0.54\to 18.72$	$2Li+O_2$	$L{i}_2+O_2$
	(All electric)		inverter, electric
			motor, gearbox, fan
[102]	Hybrid-electric	Liquefied Natural	Gas Generator, jet	$55.5\to 0.54\to 18.72$	$C{H}_4+2O_2$, $2Li+O_2$	$C{O}_2+2H_{2}O$, $L{i}_2+O_2$
[103]	(Serial hybrid	Gas, lithium-ion	fuel, power turbine,
[104]	electric propulsion)		electric generator,
			AC/DC rectifier,
			Battery, DC/AC
			inverter, Electric
			motor, fan.
[105]	Electroaerodynamic	Lithium-ion	Polymer cells	$2.4$	$e^{-}$, $O_2$	$O_2\left(Zeroemission\right)$
[106]		polymer cells

Table 3. Current and future promising propulsion energy systems.

Ref.	Energy Sources	Consumables	Configuration	Gravimetric Energy (MJ/kg)	Composition Product	Emission Product
[97]	Mechanical	Gasoline	Jet fuel, power	$46.8$	$2C_8{H}_{18}+25O_2$, $N_2$	$71\%C{O}_2$+ $28\%H_{2}O$
[98]			turbine, fan.		$+O_2$	$+N{O}_x+S{O}_x+CO+$
						$UNC+Particulate$.
[99]	Pneumatic	Liquefied Natural	Gas Generator, jet	$55.5$	$C{H}_4+2O_2$	$C{O}_2+2H_{2}O$
[100]		Gas (Methane)	fuel, power turbine,
		Liquid hydrogen	gearbox, fan.	142	$2H_2+O_2$	$2H_{2}O\left(Zeroemission\right)$
[101]	Electrical	Lithium-ion	Battery, DC/AC	$0.54\to 18.72$	$2Li+O_2$	$L{i}_2+O_2$
	(All electric)		inverter, electric
			motor, gearbox, fan
[102]	Hybrid-electric	Liquefied Natural	Gas Generator, jet	$55.5\to 0.54\to 18.72$	$C{H}_4+2O_2$, $2Li+O_2$	$C{O}_2+2H_{2}O$, $L{i}_2+O_2$
[103]	(Serial hybrid	Gas, lithium-ion	fuel, power turbine,
[104]	electric propulsion)		electric generator,
			AC/DC rectifier,
			Battery, DC/AC
			inverter, Electric
			motor, fan.
[105]	Electroaerodynamic	Lithium-ion	Polymer cells	$2.4$	$e^{-}$, $O_2$	$O_2\left(Zeroemission\right)$
[106]		polymer cells

5.4. Algorithm of Multidisciplinary Optimization

Numerous MDO frameworks have been developed to enable users to easily use optimization algorithms and couple multiple disciplines via graphical user interfaces. These frameworks include ModelCenter, Isight, VisualDoc, Optimus, ModeFrontier, AML Suite, Sorcer, and so on. However, their numerical methods for optimization problems and convergent multidisciplinary analyses are generally not the most recent ones [107]. These frameworks typically use fixed-point iterations to converge. When gradient-based optimizers require derivatives, finite-difference approximations are used instead of the more precise analytical derivatives. Furthermore, these conventional design optimizations require many iterations to obtain the final design. Therefore, the convergence speed is especially slow, particularly in topology optimization [108]. A program that allows users to explore the coupling system themselves can continuously develop, improve, and rapidly update the optimization tools. The user could add and improve the required modules from their corresponding field, which has become one of the current trends in the development of MDO [109].

The current problems faced by MDO are undoubtedly the challenges for ultra-high-dimensional optimization systems. Ultra-high-dimensional optimization objectives and design variables can be divided into system-level, subsystem-level, discipline-level, and intra-disciplinary. The design variables of these objectives and constraints are intertwined, coupled, and interdependent. There is only a classification of natural attributes between the target at the upper level and the multiple targets at the lower level, and there is no distinction between primary and secondary [110]. The same design variable is associated with the performance functions of multiple disciplines, and system-level design variables are associated with discipline levels and even the optimization target of the intra-disciplinary field. There is an intense mutual interference between the optimization variables, which severely affects the convergence speed of the optimization process. It greatly increases the difficulty of establishing and developing a multidisciplinary optimization theoretical model and efficient algorithms for DP aircraft.

After analyzing the variable induction method of MDO, it is notable that in the problems of propulsion integrated aircraft in MDO, some variables belong to different disciplines, and some belong at both the system level and discipline level. These variables interfere with each other and severely affect the convergence of the optimization procedure [111]. In addition, local and global variables with different properties occasionally couple together, resulting in high-frequency and low-frequency perturbations in the flow field. These circumstances greatly increase the difficulty of MDO modeling and efficient algorithms if multi-objective evolutionary algorithms (MOEAs) based on Pareto cooperative equilibrium are used to solve the huge-scale multi-level nested multi-disciplinary optimization problem. If all design variables and objectives are treated equally as one piece, the hierarchical characteristics between the information transmitted by the objectives and design variables are inevitably ignored. It is difficult to avoid the equivalence between the final optimization result and the original optimization problem [112]. Furthermore, MDO cannot make parallel decisions when all optimization objectives are considered to be an integration (Refer to Figure 8) [113].

Apart from the coupling problems, the exploitation of rapid parametric modeling for a complex object degrades the optimization of aircraft shape design. It is also one of the technical bottlenecks for the current MDO. In the current optimization process, we can easily detect robust and efficient parametric shapes and meshes. The shape geometry generated by the mesh needs to be smooth, feasible, and good quality [114]. Concise shape parametrization can prevent useless variables from entering the optimization system, but it also increases the difficulty of optimization and reduces optimization efficiency. As a consequence of shape changes during optimization, mesh regeneration and deformation techniques are inevitably required. At this phase, the conventional mesh regeneration technology should be used instead of the automatic deformation technology. This solves the problems of inconsistent cell numbers in the flow-sensitive area and insufficient mesh quality control caused by mesh regeneration, improving the optimization efficiency [115]. Automatic mesh deformation includes creating mesh data structures, coordinating editing across processes, and balancing loads. Additionally, it establishes a geometric technique for optimizing surface mesh reconstruction. This technique is an automatic generation of an aircraft shape realized according to the sawtooth of the surface. For large-scale local mesh regeneration, we must establish the adjoint equation. Then, we must solve the flow field to obtain its tensor magnitude. A unified unstructured mesh, boundary layer mesh, and surface mesh adaptive regeneration system can be established through tensor magnitude and mesh editing operations [116].

6. Conclusions

We discuss propulsive configurations. They could improve the thrust–drag balance. This might provide flight benefits, such as energy savings. A series of potential aero-propulsive options and samples of future civil aircraft are analyzed. A series of synergistic problems that exist in the integration of large civil aircraft are discussed. These problems involve the configuration schemes selection and multidisciplinary optimization (MDO) of the integration aircraft. They are either in a situation of mutual interference or unable to achieve the effect of a standalone runtime. To this end, the future challenges and trends for various propulsive technologies and MDO approaches that could handle cross-integration are listed. Specific disclosures include the following:

• Propulsion-airframe: The choice of configuration coupling objects lead to a huge performance difference in the aero-propulsive layout. An incorrect pairing potentially loses the layout benefits from an original configuration. When DP is mounted on the TWB wing, a regional propulsor has to be embedded simultaneously in the aft to rectify the problem of wake losses in the aft. Meanwhile, the larger and more concentrated thrust at the aft either fills or even penetrates the wake. These circumstances cause the DP layout to become invalid and redundant on TWB. Contrarily, the coupling relation between DP and BWB can maintain the original thrust characteristics of DP. It can add some extra benefits to aircraft configuration, such as omitting the area and weight of the rudder, improving the flight performance in the L/D ratio, shortening takeoff distance under specific conditions, etc. However, studies in boundary layer ingestion (BLI) in regional propulsion have shown that TWB has a propulsion efficiency advantage over BWB. The reason is that the current distributed method of BLI is unable to fill the wake loss in BWB completely. Since the drag of BWB is not concentrated on the central axis but is diffused, it makes the uniform thrust of DP unable to achieve the effect of thrust–drag balance after being integrated with BWB. To achieve complete distributed advancement, the distribution of thrust should not depend on the number of propulsors, but should take into account the drag on the aircraft;
• Scheme and design: The layouts and frameworks of 59 aero-propulsive cases from different research units are analyzed. Most layout ideas are constructed and based on the parametric model of the predecessors. The propulsor positions are mostly those predicted to be likely to have high propulsion efficiency. Then, various commercial MDO software is used to optimize the target performance, such as flight range, load, fuel quantity, etc. In fact, scheme selection, evaluation, and analysis of aero-propulsion should be conducted before constructing the parametric model, because conducting the MDO with the existing conventional propulsors and DP layout will not obtain a fresh and effective scheme. From an essential viewpoint of DP, DP does not refer to the distribution of propulsors but the distribution of thrust. These physical limitations lead to the neglect of elongated cross-flow fans that can fully cover the trailing edge of the BWB. Additionally, energy or more advanced propulsion systems that could assist the elongated cross-flow fan are also ignored. Thus, a thrust–drag analysis and calculation model framework for the aircraft’s trailing edge is urgently established such that thrust corresponding to the drag can be distributed and coordinated in each trailing part of the aircraft. This configuration could improve layout efficiency and effectiveness to promptly accomplish momentum balance and reduce the energy consumption of flight;
• MDO framework: In terms of multidisciplinary optimization, it can be found that the current goal of disciplinary integration is growing larger. The integrated disciplines include aerodynamics, propulsion, structure, weight, stability, control, and additional duct systems. In the wake of functional coordination and coupling, the problem of mutual interference also happened. In general, splitting or local computing methods are unable to solve these problems. Since there are too many interdisciplinary subjects, large-scale multi-level nesting multidisciplinary optimization modeling and efficient parallel algorithms that possess coupled performance analysis have to deal with these problems. Among them, the path design of system level and attribute classification can be used to improve scale precision and enhance the reliability of parallel algorithms. On the other hand, the current commercial MDO framework allows users to use optimization algorithms easily. However, they are mostly optimization programs with extremely slow convergence speeds. In addition, the codes of these algorithms cannot be modified in order to compare their efficiency. An optimization system that allows users to explore the coupling system and develop, improve, and update personally could accelerate the development of optimization methods. It has become one of the current trends of MDO development.