Aerospace Science and Technology 100 (2020) 105808

Contents lists available at ScienceDirect

Aerospace Science and Technology

www.elsevier.com/locate/aescte

Effect of a fuselage boundary layer ingesting propulsor on airframe

forces and moments
Antonio Martínez Fernández ∗ , Howard Smith
Cranfield University, College Road, Cranfield, Bedfordshire, MK43 0AL, UK

a r t i c l e i n f o a b s t r a c t

Article history: The aim of this research project was to investigate the benefits of hybrid-distributed propulsion and
Received 19 June 2019 boundary layer ingestion applied to an airliner similar in size, range and cruise velocity to an Airbus
Received in revised form 16 November 2019 A320. The power system selected consisted of two under-wing mounted conventional turbofans and an
Accepted 20 February 2020
electrically driven boundary layer ingesting fan, located at the rear fuselage. The power required to drive
Available online 21 February 2020
Communicated by Cummings Russell
the fan is extracted from both turbofans. The worked carried out and presented in this paper offers a
new approach to modelling boundary layer ingestion configurations, consisting of using the sliding mesh
Keywords: method to simulate the rotation of the blades within an airflow. The simulations were performed in
Aircraft design two phases. The first stage corresponded to the analysis of an isolated configuration, where fuselage
Boundary layer ingestion and electric fan were considered to be far away from each other such that there is no aerodynamic
Computational fluid dynamics interference between the fuselage and the electric fan. During the second phase, the simulations were
conducted with the electric fan integrated with the airframe, ingesting the boundary layer around the
fuselage. The simulations indicate that ingesting the boundary layer results in remarkable improvements
reducing drag and increasing propulsive force, reaffirming the potential of this technology to reduce fuel
consumption. However, the results also registered a substantial increment in the pitching, rolling and
yawing moments in the integrated configuration with respect to the isolated arrangement. This gain
needs to be accounted for during early stages of the design process to maintain the stability and control
of the aircraft within acceptable margins.
© 2020 Elsevier Masson SAS. All rights reserved.

1. Introduction Therefore, the ram drag is reduced in the former design. However,
the engines are designed to be embedded in the freestream, where
the flow is continuous and uniform. On the other hand, the bound-
Boundary layer ingestion has stood out as a solution to improve
ary layer at the rear of the fuselage is highly turbulent, which
both aerodynamics and propulsion performance [1–3]. The aim of
produces severe and uneven stresses in the fan blades and the na-
ingesting the boundary layer is to reenergize the slower air moving
celle that can reduce engine performance [4,5].
near the surface of a body to reduce the drag due to friction. This
The potential of this technology to reduce drag and fuel con-
can be achieved by placing a propulsor at the rear of a fuselage
sumption has made it a target of numerous aircraft designs such
such that the boundary layer is accelerated passing through it. as the NASA’s D8 and N3-X or the E-Thrust, concept developed by
The benefits of ingesting the boundary layer around a body are Airbus and Rolls-Royce. However, due to the level of sophistication
not limited to the reduction of drag but also to improve engine of these designs, it is necessary first to overcome technological and
performance. The velocity and the total pressure of the airstream certification challenges. Consequently, the entry into service of the
at the intake of the engine in boundary layer ingesting concepts aircraft previously mentioned is predicted for several generations
are lower than the freestream ingested by conventional podded beyond the current air transport technology [6,7]. In contrast, air-
configurations, where the engine is placed separate from the body. craft such as the STARC-ABL, also conceived by NASA, have adopted
more traditional designs, incorporating an electric fan at the rear
of the tail cone, ingesting the boundary layer around the fuselage
in a conventional wing-tube configuration.
* Corresponding author. Between 2010 and 2015, Boeing Research & Technology con-
E-mail address: [email protected] (A. Martínez Fernández). ducted a project called SUGAR (Subsonic Ultra Green Aircraft Re-

https://doi.org/10.1016/j.ast.2020.105808
1270-9638/© 2020 Elsevier Masson SAS. All rights reserved.
2 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808

Nomenclature

AR Aspect ratio Fy Lateral force

b Wing span Fz Vertical force
BLI Boundary layer ingestion HPC High performance computer
c Mean Aerodynamic Chord MAC Mean aerodynamic chord
CAD Computer aided design MTOW Max take-off weight
CFD Computational fluid dynamics
Mx Moment with respect to the X axis
C mx Rolling moment coefficient
My Moment with respect to the Y axis
C my Pitching moment coefficient
C mz Yawing moment coefficient Mz Moment with respect to the Z axis
Cx Direction of flight force coefficient OEW Operative empty weight
Cy Lateral force coefficient S re f Reference area
Cz Vertical force coefficient U∞ Freestream velocity
Fx Force along the direction of flight ρ∞ Freestream density

search) [8], issued by NASA, to explore Hybrid electric propulsion. done to model the configurations such as 2D representations or
The project, led by Marty K. Bradley and Cristopher K. Droney, axisymmetric fuselages. Although these simplifications are valid
included a configuration with a boundary layer ingesting aft- for a preliminary study on the effects of boundary layer inges-
propulsor. The configuration also included two LNG (Liquid Natural tion and examine the forces in the direction of flight, it does not
Gas) fuel cell hybrid turbines and a truss braced wing. Due to the capture the ramifications of ingesting the boundary layer in a real
level of technology considered in this airplane the design was cat- 3-dimensional and non-axisymmetric geometry. Therefore, there is
egorized as an N+4 generation (entry into service 2040-2050). The a niche in the current research on BLI assessment regarding how
results of the study indicated a saving of 10 percent in fuel con- this technology affects the other components of the forces acting
sumption with respect to a similar configuration without boundary of the aircraft.
layer ingestion [8,9]. The aim of this project was to investigate the three-dimensional
One year later, based on the SUGAR project, Jason R. Welstead changes, in forces and moments, while ingesting the boundary
and James L. Felder, from the NASA Langley Research Center and layer around the fuselage of a single-aisle airliner, with respect to a
the NASA Glenn Research Center respectively, presented their own conventional configuration, through CFD simulation. The main ad-
BLI concept, the STARC-ABL [10]. This design adopted a more tradi- vantage of using CFD is that it can be applied to an accurate 3D
tional configuration and, excluding the aft-fan, the other elements representation of the elements that compose the aircraft. This is
of the aircraft remained similar to the ones found in today’s air- crucial to examine the different profiles of the boundary layer de-
liners. The less sophisticated technology assumptions reduced the pending not only on the longitudinal component of the fuselage
entry into service to 2035. The STARC-ABL offers a potential 12% but also at different circumferential and radial locations.
mission fuel burn reduction with respect to a conventional airliner
with the same technology assumptions [10]. 2. Concept description and design
The significant potential of this configuration, to improve both
The BLI configuration selected for this research consisted of
aerodynamics and engine performance, motivated NASA to con-
a narrow-body airliner designed to be similar in size, range and
tinue the research in boundary layer ingestion with more detailed
cruise velocity to an Airbus A320 (Fig. 1). The power system in-
analyses. In 2017, Justin S. Gray, from the NASA Glenn Research
cluded two under-wing mounted turbofans and an electric fan
Center, expanded Welstead and Felder’s work by developing a fully
attached to the tail cone, ingesting the boundary layer around the
coupled propulsion-RANS model applied to a boundary layer in-
fuselage. The power required to drive the fan is extracted from two
gestion configuration [11]. Before Gray’s research, the analyses in
generators located in each turbofan (Fig. 2). The main features of
boundary layer ingestion considered the effects of the aerodynam-
the concept are summarized in Table 1.
ics on the propulsor or the effects of the propulsion system on
the aerodynamics but not both simultaneously [11]. Gray devel-
2.1. Aircraft design
oped a reciprocal process using the RANS solver ADflow for the
aerodynamics model and the 1D thermodynamic cycle tool pyCy-
The conceptual design of the aircraft was completed in two dif-
cle to model the propulsion system [11]. Gray concluded that a ferent phases. The first stage included the design of a conventional
boundary layer ingestion configuration produces improvements in airliner configuration, which acted as a baseline to build upon
performance due to both propulsion and aerodynamics equally and during the development of the BLI concept in the second step of
considering only aerodynamics or propulsion ignores half of the
overall effect [11]. The results obtained demonstrated the need of
using a fully coupled method to account for both aerodynamics
and propulsion system simultaneously to fully capture the benefits
of boundary layer ingestion.
Additionally, Mark J. Drela and Aurélien Arntz developed
methodologies based on power balance [12] and exergy-based for-
mulation [13] respectively that highlight the importance of consid-
ering the full integration between the propulsor and the airframe.
The research projects conducted on BLI focused on propulsive
force improvements and aerodynamic enhancement in the direc-
tion of flight and it has been well documented in both fields.
However, in the previous research on BLI, simplifications were Fig. 1. BLI concept render (90 mm width).
 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808 3

Fig. 4. Fuselage CAD parametrization (90 mm width).

Fig. 2. BLI power system arrangement (90 mm width).

Table 1
BLI configuration specifications.

Parameter Units Value

Range nm 3500
Cruise mach - 0.78
Cruise altitude m 11,790
MTOW kg 73,990 Fig. 5. Tail cone mesh – inflation detail (90 mm width).
OEW kg 38,220
Wing area m2 137.85
2.2. Electric fan design
Wing span m 36
MAC m 4.32
AR - 9.4 The electric fan, displayed in Fig. 3, was outlined based upon
Leading edge sweep deg 30 the “Fuselage Fan” used by Boeing in the SUGAR project [8]. The
Max T-O field length m 2500 convergent nacelle was sized following the same process taken by
Thrust (cruise) kN 40.62
Welstead and Felder during the development of the STARC-ABL,
based on the relation between the deficit of momentum captured
and the percentage of the boundary layer ingested [10]. The size
of the fan’s core was determined taking into account the systems
the design process. In addition, the baseline configuration was also
that need to be allocated inside, following Jansen’s study on sizing
used as a comparison point to evaluate the changes produced by
electric components for a similar configuration [16], and the shape
the ingestion of the boundary layer.
of the aft-fuselage. The fan was designed as a single stage com-
The mission profile was based upon that of the Airbus A320
pressor composed of 22 blades constructed using the methodology
and the performance requirements were constructed through a
presented by Mattingly [17]. The rotational speed of the blades
market study and a parametric analysis taking into account take
was set to 2000 rpm determined to avoid supersonic speeds at the
off, second segment climb, approach and cruise constraints [14,15]. tip of the blades.
The size of the wing for both configurations was determined from
the wing loading and optimized for Mach 0.78 cruise. A T-tail ar- 3. Modelling
rangement was selected for the empennage due to the minimum
aerodynamic interference with the aft-fan that it provides, sized The aircraft was modelled in Catia V5 and the fuselage and the
using the tail volume coefficient method [15]. The fuselage was in- electric fan were parametrized to enable rapid changes in size and
tended to accommodate 160 passengers in a 2-class arrangement, geometry. The CAD models were imported into Ansys Workbench
18 seats were allocated to first class and the remaining 142 were for meshing and simulation setup. The simulations were carried
distributed in Economy Class. out on a 60 Teraflops, 1976 core HPC (High Performance Com-
puter).

Fig. 3. Electric fan CAD render (left) and cross-section (right) (140 mm width).
4 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808

Fig. 6. Electric fan with rounded spinner.

3.1. Isolated configuration Table 2

Reference values.

The isolated configuration corresponds with the baseline air- Parameter Units Value
frame and the electric fan such that both elements interact with ρ∞ kg/m3 0.32
separate flows, and therefore there is no aerodynamic interference U∞ m/s 230.61
S re f m2 137.85
between the fuselage and the electric propulsor. Hence, two sep-
c m 4.32
arate CFD analyses were conducted for the isolated configuration, b m 36
one for the fuselage and one for the propulsor.
The CAD model of the fuselage was parametrized with a se-
ries of points to act as control points allowing the free defor-
mation of the body, aiming to facilitate optimization work. The
main goal of the optimization of the aft fuselage geometry was
to avoid flow separation around the tail cone during cruise con-
ditions. The boundaries of the deformation were constrained with
several fixed points to maintain the volume required for the pay-
load and the take-off and landing clearance angle (Fig. 4). The
CAD model of the fuselage was then imported into Ansys Work-
bench, where a 3D unstructured mesh was generated composed of
25,367,151 nodes and 18,783,158 elements. The aim of the simu-
lations of the fuselage was to determine the pressure distribution
over the airframe as well as capturing the boundary layer profiles
in longitudinal, circumferential and radial directions. Therefore, the
mesh was formed using the feature Inflation Layers on the fuse-
lage’s surfaces (Fig. 5), which refines the mesh locally, creating thin Fig. 7. Reference axis system (90 mm width).
layers near the walls to generate high aspect ratio cells, suitable
for capturing areas where the airflow experiences rapid changes in
critical variables such as within the boundary layer. Transient sim-
ulations were conducted to replicate flight conditions during cruise
using the density solver included in the Ansys Workbench’s mod-
ule Fluent. The boundary conditions were specified as no-slip wall
for the surfaces of the fuselage and pressure-far field for the sur-
faces of a control volume large enough such that the properties
of the flow correspond to the properties of the freestream during
cruise conditions.
For the isolated analysis of the propulsor the fuselage was re-
placed by a rounded spinner, following the same procedure as that
shown in Gray’s research [11]. The modification is shown in Fig. 6.
A parametrized CAD model of the electric fan, was also im-
ported into Ansys Workbench. The CFD simulations were con-
ducted using the sliding mesh technique, which allows the sim-
ulation of flows in multiple moving reference frames. To apply
this technique, two different cell zones were created during the
meshing phase, one to represent the static parts of the model
and the other for the moving parts. The former included the far- Fig. 8. Cx (90 mm width).
field control volume, the nacelle and the engine core and the
latest, included the fan blades and a cylindrical control volume 3.2. Integrated configuration
around the fan blades. Both zones were bounded by all the faces
of the fan control volume creating interface zones that limited For the last set of simulations, the electric fan was integrated
both cell zones. The adjacent cells from different zones are related with the airframe. The fuselage shape was slightly modified to ac-
to each other by a mesh interface, allowing movement relative commodate the fan at the tail cone. The direct coupling of the fan
to each other. The mesh was composed of 17,427,518 nodes and with tail cone produced sharp edges in the connection between
12,562,991 elements. the elements, therefore, it was necessary to adapt the shape of
 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808 5

Fig. 9. Mach number and pressure contours around fuselage (Isolated configuration) (140 mm width). (For interpretation of the colours in the figure(s), the reader is referred
to the web version of this article.)

the rear fuselage to provide a smooth aerodynamic surface. Nev- such forces around the center of gravity of the elements (fuse-
ertheless, the modifications were minimum to ensure that the lage and electric fan). Although the moments around the overall
comparison with the isolated configuration was valid. The method center of gravity of the aircraft are what ultimately impacts the
followed to mesh the model was the same as the one created for performance and control of the design, at this stage it was consid-
the fan and the fuselage in the isolated configuration to maintain ered sufficient to analyze only the changes in the moments at the
consistence between the simulations. In this case, the mesh was
center of gravity of the individual components. In the next stages
composed of 31,787,908 nodes and 22,523,639 elements.
of the design process the analysis will expand to account for the
3.3. Simulation environment rest of the elements that compose the aircraft (wings, empennage
and turbofans). Then, an appropriate overall aircraft center of grav-
The CFD analyses were conducted to simulate cruise conditions, ity can be estimated to establish a consistent reference to evaluate
as presented in Table 1. The angle of attack was considered to be the moments. However, it was necessary first to acquire initial data
zero such that the longitudinal axis of the fuselage is aligned with for the development of the design and get a preliminary insight of
the freestream. The rotational speed of the blades was set constant the effects of ingesting the boundary layer around the fuselage at
to 2000 rpm in both analyses. component level.
The forces and moments were transformed to non-dimensional
4. Results
coefficients using expressions (1) to (6) and the reference values
From the simulations were extracted the pressure and viscous shown in Table 2, corresponding to cruise conditions. The conven-
forces integrated over the bodies and the moments induced by tion of signs of the forces are positive as indicated in Fig. 7 and
6 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808

Fig. 10. Mach contours around electric fan (Isolated configuration) (140 mm width).

Fig. 11. Mach number and pressure contours around BLI configuration (140 mm width).
 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808 7

the moments are considered to be positive according to the right

hand grip rule applied to each axis.

2 · Fx
Cx = (1)
ρ∞ · U ∞
2 ·S
re f
2 · Fy
Cy = (2)
ρ∞ · U ∞
2 ·S
re f
2 · Fz
Cz = (3)
ρ∞ · U ∞
2 ·S
re f
2 · Mx
C mx = (4)
ρ∞ · U ∞
2 ·S
re f · b
2 · My
C my = (5)
ρ∞ · U ∞
2 ·S
re f · c
2 · Mz
C mz = (6)
ρ∞ · U ∞
2 ·S
re f · b

4.1. Aerodynamic coefficients Fig. 12. Fan face total pressure comparison (90 mm width).

The results of the simulations, shown in Fig. 8, indicate that

the BLI configuration offers a remarkable improvement in fuselage
aerodynamics, reducing the force acting against the direction of
flight, the fuselage drag, by 34.58% with respect to the isolated
configuration.
To gain a better understanding of the reasons behind the im-
provement in fuselage aerodynamics several contours were plotted.
Fig. 9 and Fig. 10 present the Mach contours around the fuselage
and electric fan obtained from the analysis of the isolated configu-
ration. In addition, in Fig. 9 the pressure distribution acting across
the surface of the fuselage are also shown. Similarly, the Mach
number and pressure contours extracted from the simulations of
the integrated configuration are displayed in Fig. 11.
As can be observed in Fig. 11, the presence of the electric fan
at the rear of the tail cone modifies the airflow around the aft
fuselage, making the boundary layer to grow thicker and slowing
down the airspeed. As a consequence, the static pressure acting
on the rear fuselage is higher in the BLI configuration (Fig. 9 and
Fig. 11), which explains the aerodynamic improvement.
Similarly, the simulations predicted an improvement in propul-
sive force of 33.15% in the integrated analysis relative to the iso-
lated configuration (Fig. 8). By definition, the isolated configuration Fig. 13. Cy (90 mm width).
considers that airframe and the electric fan are far away from
each other such that there is no aerodynamic interference between
both components. Hence, the propulsor is immersed in a clean
and uniform airflow (Fig. 10). In contrast, in the BLI configuration
the propulsor is integrated with the airframe at the aft fuselage,
ingesting the boundary layer around the fuselage (Fig. 11). The ve-
locity of the airflow and the total pressure within the boundary
layer are lower than those of the freestream and as a consequence
the ram drag is reduced, explaining the gain in propulsive per-
formance. To confirm this, the mass-averaged total pressure at
the electric fan intake was also examined. The results, shown in
Fig. 12, predicted a lower total pressure in the integrated analysis.
The decrement in total pressure is evidence of a slower airstream
entering the electric fan intake. Furthermore, the nacelle of the
electric fan is also submerged in a slower airstream (Fig. 10 and
Fig. 11), which reduces the drag acting on the nacelle surfaces,
enhancing the gain in propulsor performance. It is important to
notice that the change in propulsive force only is not an indication
of any energy or fuel consumption benefit. In order to evaluate
the efficiency of the electric fan is also necessary to take into ac-
count the power provided to the electric fan, required to maintain
the same rotational speed in both configurations. However, at this
stage, the analysis focused on propulsive force only, postponing the Fig. 14. Cz (90 mm width).
8 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808

Fig. 15. Pressure distribution over tail cone surface (Integrated configuration) (140 mm width).

assessment of the power input until the next phase of the design the surfaces is also axisymmetric, as shown in Fig. 16. This implies
process. that the contribution of the pressure distribution, acting on the
These results are in agreement with Gray’s findings during his nacelle surfaces, to the overall lateral and vertical forces is zero.
research on boundary layer ingestion with a similar configuration Therefore, the lateral and vertical forces predicted by the simula-
[11] and reaffirm not only the potential of BLI to improve both tions are exclusively induced by the rotation of the fan’s blades.
aerodynamics and power system performance but also the impor- In contrast, during the integrated analysis, the propulsor is sub-
tance of using coupled analysis to capture the benefits of BLI. merged in a highly turbulent airflow, ingesting the boundary layer
However, the simulations also predicted large changes in the around the fuselage. Furthermore, due to the upswept geometry
lateral and vertical coefficients, Cy and Cz respectively, of the BLI of the rear fuselage, the airstream is non-axisymmetric (Fig. 11).
concept with respect to the isolated configuration. Fig. 13 and Thus, the boundary layer profile around the rear fuselage varies de-
Fig. 14 show the lateral and vertical coefficients for both isolated pending on the longitudinal, circumferential and radial location. As
and integrated configurations. The results indicate an increment in a consequence, the upper surface of the nacelle faces a faster flow
the lateral and vertical forces acting on the surfaces of the fuse- than the lower surface, which produces a lifting effect, explain-
lage. The fuselage features a non-axisymmetric geometry and as ing the changes in the vertical force experienced by the electric
a consequence a uniform airstream produces an uneven pressure fan. In addition, the distribution of pressure over the fan’s blades
distribution around the surfaces of the fuselage, producing the is also non-axisymmetric in the BLI configuration (Fig. 16), which
vertical force found in the isolated analysis. The lateral force reg- contributes to the overall vertical force and provides a possible ex-
istered during the simulations of the isolated configuration is not planation to the increment in the lateral force.
significant and it was more likely caused by numerical errors. Nev-
ertheless, the presence of the propulsor, located at the very rear 4.2. Aerodynamic moment coefficients
of the tail cone, modifies the flow field around the aft fuselage
(Fig. 9 and Fig. 11). It can be noticed In Fig. 15, where the pressure It has been shown that the integration of the electric fan with
distribution acting on the rear fuselage of the BLI configuration is the airframe induces changes in the flow field at the aft fuselage,
plotted, that the pressure distribution is not uniform across the modifying the aerodynamic response of both the fuselage and the
surface. The airflow, influenced by the electric fan and the rotation propulsor. The different distributions of pressure over the surfaces
of the blades (inflow swirl), induces a higher pressure on the top between both configurations induce different forces. Consequently,
and right sides of the tail cone, which explains the increment in a variation in the moments generated around the center of gravity
the lateral and vertical forces. of the elements is to be expected.
Similarly, the airflow around the electric fan is affected by the To analyze the effects of these variations on the moment coef-
coupling airframe-propulsion system and the lateral and vertical ficients, the aerodynamic moments were also extracted from the
forces, acting on the surfaces of the propulsor, vary between both simulations and transformed to non-dimensional coefficients. The
configurations. The analysis of the isolated configuration assumes results are shown in Fig. 17, Fig. 18 and Fig. 19.
by definition that there is no aerodynamic interference between The rolling moment of the fuselage is enlarged by 2 orders
airframe and electric fan. Hence, the electric fan is immersed in a of magnitude while ingesting the boundary layer (Fig. 17) with
uniform airstream and due to the propulsor’s axisymmetric geom- respect to the isolated analysis. This boost is caused by the in-
etry (with respect to the X axis), the distribution of pressure over crement in the lateral force at the aft-fuselage. It is important to
 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808 9

Fig. 16. Electric fan pressure distribution extracted from the isolated analysis (left) and from the integrated analysis (right) (140 mm width).

notice that the tail cone, due to the upswept geometry of the aft tail cone (Fig. 15), which induces a clockwise pitching moment to
fuselage, is located above the center of gravity and the airstream, the fuselage.
disturbed by the presence of the electric fan, generates a higher The fuselage is symmetric across the x-z plane and thus there
pressure on the right side of the tail cone (Fig. 15), which induces is no moment around the z axis (yawing) in the isolated analy-
a clockwise moment around the X axis to the fuselage. sis. Notice that the yawing moment of the isolated configuration
On the other hand, the fuselage pitching moment is alleviated shown in Fig. 19 is effectively insignificant and it is most likely
due to numerical errors. Nevertheless, the simulations predicted
in the BLI configuration (Fig. 18). The existence of a pitching mo-
an increment in the yawing moment in the BLI analysis relative
ment during the isolated configuration analysis is due to the non-
to the isolated case. This increment in the yawing moment is due
axisymmetric geometry of the fuselage, which, submerged in an
to a stronger pressure on the right-side surface of the tail cone
airflow, produces a different distribution of pressure on the upper (Fig. 15), which causes a clockwise moment with respect to the Z
and lower surfaces and as a consequence a moment with respect axis.
to the Y axis is generated. In the geometry analyzed, the resultant Accordingly, the moment coefficients of the electric fan suffer
pitching moment was found to be anti-clockwise, pushing the nose a large variation due to the airframe-propulsor coupling. The re-
of the fuselage downwards. The simulations performed on the BLI sults yielded a reduction of the rolling moment produced on the
configuration indicate a higher pressure in the top surface of the electric fan in the BLI configuration relative to the isolated analysis
10 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808

ing the forces, this boost was to be expected. The nacelle of the
fan generates lift in the BLI model due to different velocity gradi-
ents at the top and the bottom surfaces (Fig. 11), which induces a
pitching moment in the electric fan.
Lastly, the yawing moment of the electric fan is largely in-
creased in the BLI configuration analysis (Fig. 19). The increment
in yawing moment in the electric fan is a consequence of the in-
creasing lateral force generated by the distortion on the fan blades
and the non-uniform distribution of pressure around the nacelle
(Fig. 16).
It may be that the results of the simulations are penalized by
a non-optimized design and in future work, when more refined
models are available, the trends may change or be alleviated. Nev-
ertheless, the results highlight the importance of implementing 3D
studies in boundary layer ingestion taking into account the inte-
gration of the electric fan into the airframe. The highly turbulent
and non-axisymmetric flow found in the boundary layer at the aft
fuselage precipitates changes in the behaviour of the electric fan,
Fig. 17. Cmx (90 mm width). changes that are further enhanced by the mere presence of the
propulsor, which modifies the flow field upstream. The distribu-
tion of forces is uneven across the surfaces of the bodies, and as
a consequence, moments are generated. This represents the most
significant finding of this work and can only be noticed by per-
forming the analysis on a 3D geometry, considering the actual
non-axisymmetric shape of the fuselage with features such as the
upswept tail cone. The particular moments generated in BLI con-
cepts need to be accounted for in the early stages of the design
process due to the direct impact of such moments on the perfor-
mance and control of the aircraft.

5. Conclusions

The conceptual design of a narrow-body airliner with a fuse-

lage boundary layer ingesting fan was developed for this work. The
fuselage and the electric fan were analyzed through 3D transient
CFD simulation, using the sliding mesh technique to represent the
rotation of the blades within an airflow. The simulations were
performed on the fuselage and electric fan across two different
Fig. 18. Cmy (90 mm width). configurations: isolated and integrated. The isolated configuration
corresponded to the analysis of the fuselage and electric fan in-
teracting with separate flows such that there is no aerodynamic
interference between both elements. The integrated configuration
represented the BLI configuration, where the electric fan was inte-
grated with the airframe at the very rear of the fuselage’s tail cone.
The aim of the isolated configuration was to act as a comparison
point to evaluate the benefits and penalties of the boundary layer
configuration.
The pressure and viscous forces integrated over the bodies
were extracted from the simulations. Furthermore, the moments
induced by such forces were also recorded during the analyses.
The results show that the BLI concept outperforms the isolated
configuration in terms of propulsive performance and reduction of
fuselage drag. The fuselage drag is reduced by 34.58% in the inte-
grated configuration due to the changes in the airflow induced by
the presence of the electric fan at the aft fuselage. The propulsor
modifies the airstream slowing it down and making the bound-
ary layer to grow thicker, increasing the static pressure acting on
the surface of the fuselage, which precipitates the aerodynamic im-
Fig. 19. Cmz (90 mm width). provement. The electric fan propulsive force is increased by 33.15%
in the integrated configuration. This gain is due to a slower flow
(Fig. 17). The rolling moment of the electric fan is mainly gener- field ingested by the propulsor and a lower total pressure at the
ated by the rotation of the blades and its relaxation is associated fan-face, which decrease the ram drag in the propulsor inlet and
with a lower total pressure at the fan-face location while ingesting the drag along the surfaces of the nacelle.
the boundary layer (Fig. 12). However, the results also predicted large variations in the
Contrarily, there is an increment in the pitching moment of the vertical and lateral forces between both configurations, induc-
propulsor in the BLI configuration (Fig. 18). However, after analyz- ing changes in the moment coefficients of both the fuselage and
 A. Martínez Fernández, H. Smith / Aerospace Science and Technology 100 (2020) 105808 11

the electric fan. The non-axisymmetric fuselage implies different coupled models that account for the integration of the electric fan
boundary layer profiles depending on the longitudinal, circumfer- into the airframe to capture the influence of the propulsor in the
ential and radial location, creating a highly turbulent flow that boundary layer around the tail cone.
generates an uneven distribution of pressure along the surface of This work represents a first attempt to investigate the effects
the fuselage. As a consequence, moments are induced around the of ingesting the boundary layer beyond drag and propulsive force;
center of gravity of the airframe. Furthermore, the presence of the introducing a preliminary insight to all the forces and moments re-
electric fan in the integrated configuration, placed at the very rear sulting from the coupling between the fuselage and the propulsor.
of the tail cone, disturbs the flow field and as a consequence mod- Boundary layer ingestion configurations possess a large potential
ifies the boundary layer around the aft fuselage. This perturbation to improve both fuselage aerodynamics and power system perfor-
of the airstream is further enhanced by the whirl of the fan’s mance. However, such concepts may be penalized by the genera-
blades, which induces rotation to the airflow upstream from the tion of substantial lateral and vertical forces and their associated
propulsor inlet. All these effects combined precipitate changes in moments that may affect the performance and control of the air-
the moment coefficients of the airframe in the integrated configu- craft and hence, needs to be accounted for in early stages of the
ration with respect to the isolated analysis. Similarly, the moment design process.
coefficients of the electric fan are also affected by the coupling
airframe-propulsor. The fan is submerged in a clean and uniform Declaration of competing interest
airflow during the analysis of the isolated configuration. Therefore,
the pressure distribution around the nacelle is axisymmetric and There is no conflict of interest.
the only force that contributes to the generation of moments is
that induced by the rotation of the blades. On the other hand, References
the electric fan is embedded in the fuselage’s boundary layer dur-
ing the integrated analysis. This implies that the blades of the fan [1] J. Welstead, J.L. Felder, M.D. Guynn, W. Haller, M.T. Tong, S. Jones, I. Ordaz, J.
Quinlan, B. Mason, Overview of the NASA STARC-ABL (Rev. B) Advanced Con-
whirl within a turbulent flow that stresses each blade differently
cept, in: One Boeing NASA Electr. Aircr. Work., 2017, pp. 1–14.
and hence moments are generated. In addition, the boundary layer [2] C. Goldberg, D. Nalianda, D. MacManus, P. Pilidis, J. Felder, Method for sim-
around the tail cone is non-axisymmetric due to the upswept ge- ulating the performance of a boundary layer ingesting propulsion system
ometry of the fuselage. Hence, the distribution of pressure across at design and off-design, Aerosp. Sci. Technol. 78 (2018) 312–319, https://
the surface of the nacelle is non-axisymmetric, contributing to the doi.org/10.1016/j.ast.2018.04.026.
[3] A. Abbas, J. De Vicente, E. Valero, Aerodynamic technologies to improve aircraft
generation of uneven forces in the longitudinal, lateral and vertical performance, Aerosp. Sci. Technol. 28 (2013) 100–132, https://doi.org/10.1016/
directions and the associated moments. j.ast.2012.10.008.
The effects that ingesting the boundary layer causes in the lat- [4] G. Tanguy, D.G. MacManus, E. Garnier, P.G. Martin, Characteristics of unsteady
eral and vertical forces as well as in the moments have not been total pressure distortion for a complex aero-engine intake duct, Aerosp. Sci.
Technol. 78 (2018) 297–311, https://doi.org/10.1016/j.ast.2018.04.031.
captured before in any studies. Therefore, there was no possibil-
[5] A. Plas, D. Crichton, M. Sargeant, T. Hynes, E. Greitzer, C. Hall, V. Madani, Per-
ity of comparing the results as in the case of the forces in the formance of a Boundary Layer Ingesting (BLI) propulsion system, in: 45th AIAA
flight direction. It may be that in subsequent work, when more Aerosp. Sci. Meet. Exhib., 2007.
accurate and optimized designs are available, the impact of these [6] H. Kim, M.F. Liou, Flow simulation and drag decomposition study of N3-X hy-
forces and moments are found to be less significant. Most impor- brid wing-body configuration, Aerosp. Sci. Technol. 85 (2019) 24–39, https://
doi.org/10.1016/j.ast.2018.11.047.
tantly, the nacelle of the electric fan requires optimization work to [7] H. Kim, M.S. Liou, Flow simulation and optimal shape design of N3-X hybrid
improve the aerodynamics at the intake and alleviate the strong wing body configuration using a body force method, Aerosp. Sci. Technol. 71
shock generated at the top surface. In addition, the optimization (2017) 661–674, https://doi.org/10.1016/j.ast.2017.09.046.
of the aft fuselage’s geometry should aim to minimize the induced [8] M.K. Bradley, C.K. Droney, Subsonic Ultra Green Aircraft Research: Phase
II. N+4 Advanced Concept Development, May 2012, https://dx.doi.org/2060/
moments, which may constitute a design penalty. Nevertheless, the
20120009038, 207.
work presented in this paper represents the first iteration of the [9] M.K. Bradley, C.K. Droney, Subsonic Ultra Green Aircraft Research: Phase II –
design process and the data gathered during the analyses is essen- Volume II – Hybrid Electric Design Exploration, Nasa/Cr–2015-218704. II, 2011,
tial, as a baseline, for the next phases to build upon. https://dx.doi.org/2060/20150017039, 207.
In future work, the analysis must be extended to the complete [10] J. Welstead, J.L. Felder, Conceptual design of a single-aisle turboelectric com-
mercial transport with fuselage boundary layer ingestion, in: 54th AIAA Aerosp.
aircraft, including wings, turbofans and empennage. Only then can
Sci. Meet., 2016, pp. 1–17.
the overall impact of the moments variation on the aircraft perfor- [11] J.S. Gray, C.A. Mader, G.K. Kenway, J. Martins, Approach to modeling bound-
mance be assessed. The resulting moments need to be compared ary layer ingestion using a fully coupled propulsion-RANS model, in: 58th
to the control power of the rudder and tail plane/elevator to eval- AIAA/ASCE/AHS/ASC Struct. Struct. Dyn. Mater. Conf., 2017, pp. 1–16.
[12] M. Drela, Power balance in aerodynamic flows, AIAA J. 47 (2009) 1761–1771,
uate if substantial deflections of the control surfaces are required
https://doi.org/10.2514/1.42409.
for steady, trimmed flight. Furthermore, the wing downwash has [13] A. Arntz, O. Atinault, A. Merlen, Exergy-based formulation for aircraft aero-
a strong influence on the airflow seen by the rear of the aircraft, propulsive performance assessment: theoretical development, AIAA J. 53 (2015)
which may completely modify the trends in the moments shown 1627–1639, https://doi.org/10.2514/1.J053467.
in this work. Thus, the wing downwash needs to be accounted [14] D. Howe, Aircraft conceptual design synthesis, Prof. Eng. (2000), https://doi.org/
10.1002/9781118352700.
for in future researches on boundary layer configurations. Lastly,
[15] D.P. Raymer, Aircraft Design: A Conceptual Approach, fifth edition, American
the studies must broaden to include other critical flight conditions Institute of Aeronautics and Astronautics, Inc., Blacksburg, Virginia, 1992.
to fully determine the impact of this type of technology on the [16] R. Jansen, C. Bowman, A. Jankovsky, Sizing power components of an electrically
moments generated. However, the findings of this study highlight driven tail cone thruster and a range extender, in: 16th AIAA Aviat. Technol.
Integr. Oper. Conf., 2016, pp. 1–9.
the importance of using three-dimensional analyses, considering
[17] J.D. Mattingly, Aircraft Engine Design, 2002, http://books.google.com.tr/books?
the real shape of the elements, to capture the effects of a turbu- id=2Wy5rpdm3DMC.
lent and non-axisymmetric airstream on the pressure distribution.
Moreover, the results also reaffirmed the relevance of employing