aerospace

Article
Aircraft-Type-Specific Impact of Speed Brakes on Lift and Drag
Judith Rosenow * , Thomas Sachwitz, Shumpei Kamo , Gong Chen and Hartmut Fricke

Institute of Logistics and Aviation, Technische Universität Dresden, 01069 Dresden, Germany;
[email protected] (T.S.); [email protected] (S.K.); [email protected] (G.C.);
[email protected] (H.F.)
* Correspondence: [email protected]

Abstract: The increasing influence of current research in air traffic management on daily flight opera-
tions leads to a stronger consideration of individually optimized aircraft trajectories. However, in the
dichotomy between ecological, economic, and safety-based optimization goals, four-dimensionally
optimized trajectories are subject to severe constraints in terms of position and speed. To fully assess
the performance envelope of these trajectories, precise modelling of the influence of secondary control
surfaces on flight performance is necessary. In particular, the use of speed brakes can significantly
influence the descent and speed profile and allows the implementation of different cost indices. In
this study, we present a modelling approach of the influence of extended speed brakes on flight
performance and apply this method in a simulation environment for trajectory modelling of twelve
different aircraft types. In doing so, we can determine an almost linear influence of the additional fuel
requirement from the effective area of the speed brakes. The results can be implemented in any flight
performance model and enable more precise modelling of future aircraft trajectories. Specifically,
optimization targets regarding the required time of arrival, or the cost index and the consideration of
the dynamic impact of atmospheric conditions in the trajectory optimization, only becomes possible
through the calculation of the influence of the speed brake on lift and drag.

Keywords: speed brakes; aircraft trajectory optimization; flight performance calculation

Citation: Rosenow, J.; Sachwitz, T.;
Kamo, S.; Chen, G.; Fricke, H.
Aircraft-Type-Specific Impact of
Speed Brakes on Lift and Drag. 1. Introduction
Aerospace 2022, 9, 263. https://
Due to the ever-increasing volume of air traffic and its concentration, especially in
doi.org/10.3390/aerospace9050263
the airspaces around airports, new methods for increasing capacity are needed. Global
Academic Editor: Eri Itoh crises, impending economic collapses, and the increasing social impact of aviation on global
warming further require reducing the total costs and emissions of air traffic during flight.
Received: 18 March 2022
Accepted: 6 May 2022
In addition to the technical optimization of existing aircraft models and the development
Published: 12 May 2022
of new modes of flying, changes to current flight operation procedures are also necessary
to achieve these goals. An important step here is the change from static waypoint-fixed
Publisher’s Note: MDPI stays neutral routes and approach procedures to individual continuous descent operations with routes
with regard to jurisdictional claims in
adapted to the aircraft type, the atmospheric conditions, and the environment [1]. However,
published maps and institutional affil-
to calculate this most efficient flight path in advance, powerful flight performance models
iations.
are necessary [2–4], which calculate these trajectories depending on weather data, airspace
information, cost requirements and the performance data of the respective aircraft type. In
real operations, however, the atmospheric conditions might deviate from those assumed in
Copyright: © 2022 by the authors.
the flight performance calculations, and the pre-calculated Top of Descent (TOD) position
Licensee MDPI, Basel, Switzerland. might not completely follow the airline-specific optimization target anymore. An accelera-
This article is an open access article tion phase (for reducing or increasing speed) at an inefficient flight level might be necessary.
distributed under the terms and For decelerating, speed brakes can be used with a significant impact on drag and a small
conditions of the Creative Commons impact on the lift. This effect, in turn, may, have a significant impact on fuel burn and is
Attribution (CC BY) license (https:// crucial for trajectory optimization. This impact will be shown in this paper.
creativecommons.org/licenses/by/ The aim of this study is therefore to provide pilots with information on whether or
4.0/). how to return to the optimal flight path in the event of deviations from the flight path

Aerospace 2022, 9, 263. https://doi.org/10.3390/aerospace9050263 https://www.mdpi.com/journal/aerospace

Aerospace 2022, 9, 263 2 of 19

calculated in advance, taking into account the performance characteristics of the speed
brakes. Therefore, the aircraft-type-specific impact of speed brakes on flight behavior must
be estimated by the flight performance model.

1.1. State of the Art

The impact of speed brakes on flight performance (i.e., on lift and drag) depends not
only on aircraft-type-specific aerodynamic properties, but also on the speed and altitude
of the aircraft. Furthermore, there are several different modes to adjust the speed brake
position. Therewith, the degree of freedom increases significantly in flight performance
calculations when considering speed brakes. Probably for this reason, the impact of speed
brakes is often not in the focus [3,5–13] of the relevant models, or is approximated by a
constant factor of ∆c D = 0.02 and ∆c L = 0.00 [14].
Older experimental studies focus on the effect of speed brake deflection on lift coef-
ficient and pitching moment, because of their high relevance for controlling the aircraft.
Although from the aerodynamic and flight performance point of view, the effect on the drag
coefficient is more important, it has often not been measured and evaluated. Furthermore,
experimental data only account for the pressure element of the drag when using absolute
drag coefficients, ignoring the viscous contribution.
In his dissertation, Mclachlan [15] experimentally analyzed steady and unsteady
flow fields generated by a typical two-dimensional transport airfoil-spoiler configuration
without specification of the wing, and reported a decrease in lift coefficient c L up to
∆c L,α=0 = −1.2 for a speed brake deflection angle of δ = 60◦ , and an angle of attack of
α = 0◦ . For an increasing angle of attack up to α = 16◦ , the impact of speed brakes on lift
coefficient decreases to ∆c L,α=16 = −0.8.
Stucky [16] demonstrates the use of a modified step-wise regression technique for
estimating the aerodynamic nonlinearities inherent in lateral spoiler control and claims
that, as compared to conventional control surfaces, wing spoilers exhibit higher-order
characteristics when used in the roll. These features were represented using polynomial
splines as functions of angle-of-attack and spoiler deflection. Unfortunately, the author only
considered the lateral response of the aircraft model and therewith concentrated on the
impact of spoiler deflection angle and angle of attack on rolling c Mx and yawing moment
coefficients. He proved assumed linear relationships between the resultant side-forces and
the spoiler deflection angle of attack with real-flight measurements.
Abdelrahman et al. [17] experimentally studied the effects of spoiler configurations
on the aircraft aerodynamics of a B-747 airplane under wind shear conditions and show
that spoiler deflection enhances the wind shear-induced lift loss, especially at low angles of
attack. They published experimentally gained values of ∆c L,α=16,δ=90 = 1.25, corresponding
to −70 % of c L,α=16,δ=0 and ∆c L,α=4,δ=90 = 0.6 which is −1 % of c L,α=4,δ=0 . Wind shear tends
to amplify the spoiler effect on the side force and yawing moment, but only at high angles
of attack does it alter the resulting rolling moment. Furthermore, Abdelrahman proved
that speed brake deflections, on the other hand, have almost no influence on the aircraft
stability margin, which is generally increased when wind shear is present [17].
Lindsay et al. [18] experimentally proved that at high angles of attack (α > 16◦ ) and
high deflection angles (δ > 25◦ ), the spoiler airfoil on top of a NACA 2412 airfoil created
even less drag and more lift than in a clean configuration. Beyond these extreme conditions,
they gained similar results for the increase in a lift (∆c L = −0.375 and ∆c D = 0.05).
However, the results do not depend on altitude and speed and are not formalized to be
used for other than the plotted values.
Geisbauer [19] published results of fluid dynamic simulations of steady and unsteady
simulations of static and dynamic spoiler deflections in the low-speed regime with an
in-house flow solver TAU, modelling an in-house DLR-F15DS (dynamic spoiler) model
which is an assembly of a 2D wall-to-wall model representing a small aircraft wing. He
also investigated the dynamic impact of speed brake deflection on lift and drag and gained
results in the range of ∆c L,α=0,δ=60 = −0.35 and ∆c D,α=0,δ=60 = 0.05.
Aerospace 2022, 9, 263 3 of 19

All references mentioned so far focus on the discussion of experimental studies without
transferring the results into equations and thus, applying them to data other than the
tested input data. However, all experiments show a greater influence of speed brakes
on lift ∆c L = 0.00 and drag ∆c D = 0.02 than assumed by [14]. From this it follows that
the results can be used for validation purposes and comparisons with the theoretically
calculated values.
Omori [20] also presented experimental results of a spoiler as a panel vertically stand-
ing on a flat wing surface, and calculated the lift coefficient decrease due to speed brake
deflection. He used the speed brake height and the speed brake location as dependent
variables and parameterized a reference deflection angle θ0 and an effective height h. The
theoretical results were validated by measurements on a NACA 0009 airfoil. However, due
to unknown profile geometry, speed and altitudes, the results are not applicable to other
aircraft types and flight conditions.
In 1966, Barnes [21] developed an experimental approach for measuring the impact of
speed brakes on lift and pitching moment on RAE100 and RAE102 airfoils. To do so, he
identified dependencies of c L on the boundary layer displacement thickness on the aerofoil
at the position of the speed brake. Therewith, he calculated ∆c L = −0.2. The RAE airfoils,
however, seem to have lower maximum values of c L,max = 1.0. From this, it follows that
the method might not be applicable to large modern aircraft types. Additionally, the method
described in this paper depends on several parameters provided by hand-printed graphs
in low resolution with a high degree of uncertainty.
Kalligas [22] and Lee [23] looked at static, while Consigny et al. [24], Costes et al. [25], and
Nelson et al. [26] considered harmonic oscillation in their experiments, and Yeung et al. [27]
investigated ramp-type spoiler deflections in two dimensions. Consigny et al. [24] discov-
ered that, unlike deflection, quick retraction of a spoiler had no negative consequences.
Kalligas [22] observed that the drag response is slower than the lift response.
Finally, three-dimensional spoilers have been experimentally studied by Jordan et al. [28]
and Scott et al. [29]. However, the experiments were applied to very old mid-wing
electronic-warfare aircraft models [28] with aerodynamics different from today’s civil
aircraft types. Scott et al. [29] only published the drag polar for a NACA airfoil 0012,
without showing the effect of spoiler deflection on a lift and drag individually.
From the literature studied so far, it follows that either unsuitable aircraft types were
considered or important dependencies (such as speed and altitude) are missing in the
experimental setups and published values of ∆c L and ∆c D . The results do not make it
possible to elaborate a general model for the impact of speed brake deflection on lift
and drag.
In 1971, Hanke [30] quantified the impact of spoilers on flight performance (i.e.,
the impact on lift and drag) in a granular manner, for a research project of the National
Aeronautics and Space Administration (NASA) on a modern aircraft type (B747-100). In this
study, this NASA model is transferred to twelve aircraft types.

1.2. Paper Structure

After the introduction, Section 2 provides an overview of forces considered in flight
performance and some theoretical considerations of speed brakes in this system of equa-
tions. Section 3 describes the reference speed brakes model for the B747 and important
geometrical and aerodynamic details of the B747, before the transfer of the reference speed
brake model to other aircraft types is described and discussed. The implementation of the
adapted speed brake model is shown in Section 4. Using a deceleration phase just above
10,000 ft, the influence of speed brakes on the four-dimensional descent profile and the
resulting fuel consumption is analyzed for different aircraft types. In Section 5, the paper
concludes with a summary and conclusion.
Aerospace 2022, 9, 263 4 of 19

2. Flight Performance Calculation

Aircraft movements can be described by forces acting on the aircraft. In the two-
dimensional view, these essentially comprise lift FL (Equation (1)), weight FG (Equation (2)),
drag FD (Equation (3)), and thrust FT (Equation (4)) forces, as shown in Figure 1.
ρ 2
FL = v S cL (1)
2 TAS
FG = m g (2)
ρ
FD = v2TAS S c D (3)
2
FT cos γ − FL sin γ − FD cos γ = m a x (4)
where vTAS is the true air speed [m s−1 ], m the aircraft mass [kg], and a x the acceleration
[m s−2 ] in the horizontal plane x . c D and c L denote drag coefficient [a.u.] and lift coefficient
[a.u.], and ρ describes air density [kg m−3 ]. S [m2 ] describes the total wing area (including
non-deflected control surfaces) and γ [rad] the flight path angle [31]. During the flight,
the forces act on different positions of the aircraft, resulting in additional moments to the
vertical and lateral axes. Only in an equilibrium state of all forces and moments, does
the aircraft fly straight with constant speed. Control surfaces are used to change the
magnitude and the points of application of those forces, producing moments along the
three axes and controlling aircraft rotation.

Thru
st F Lift FL
T

Dra
gF
Weight FG D

Figure 1. Forces acting on an aircraft in the flight-path-oriented coordinate system. Thrust FT and
drag FD form the x-axis.

Depending on the flight attitude, lift and also drag may need to be changed. There-
fore, secondary control surfaces are installed on commercial aircraft (see Figure 2), which
essentially comprise the landing flaps and speed brakes [32]. Speed brakes or spoiler
flaps are movable flaps on the upper wing surface and are not primarily used for aircraft
rotation around the three axes (see Figure 2 for details). In modern commercial aircraft,
spoilers are also used to support the primary control surfaces, which is why they are now
differently referred to as ground spoilers, speed brakes or roll spoilers [33], depending on
their function and use [32]. Extended secondary control surfaces increase the drag on the
wing and thus make the aircraft descend if the magnitude of thrust is maintained. A single
panel can also perform several functionalities, depending on the aircraft’s current attitude
and flight phase. In the following, different uses of secondary control surfaces during the
flight are described.
Some of the secondary control surfaces can be manually extended as speed brakes,
ensuring a symmetrical deflection of the panels. Speed brakes can be extended either
individually or altogether. According to the deflection, speed brakes increase the drag
(which results in a corresponding reduction in speed) and decrease lift [33]. The effect of an
individually extended speed brake i on drag and lift can be described as a change in drag
coefficient ∆c D,i [a.u.] and lift coefficient ∆c L,i [a.u.]. By using the speed brakes, steeper
flight profiles can be flown without exceeding the speed limits of the aircraft type, or the
airspace currently being flown through [32]. The angle of the speed brakes can be set on
all commercial aircraft using a lever in the cockpit. The principle for speed-brake setting
differs not only according to aircraft type and position on the wing, but also depending
on the manufacturer as concerns the type of activation. For example, the manufacturers
Aerospace 2022, 9, 263 5 of 19

Airbus and Bombardier have fixed detents for the various angles on the selection lever
for the speed brakes [34,35], while Boeing and Embraer allow the pilot to adjust them
continuously [36,37].
Most of the secondary control surfaces acting as roll spoilers are used to support the
ailerons to increase roll moment along the longitudinal axis to initiate a turn manoeuvre.
Since the spoiler panels serve to reduce lift, they are extended on one side of the wing
for a turn manoeuvre. The resulting asymmetric lift causes the wing to sink on one
side, which creates a moment about the longitudinal axis. As roll spoilers have two
functionalities, one as spoilers and the other as ailerons, they are also called spoilerons.
Unlike ailerons, however, spoilers can only be extended upwards, and thus only reduce
lift and do not increase it [32]. Since modern flight control systems always combine yaw
and roll momentum to initiate a turn, the spoilers also take over part of the function of the
rudder, so that it does not have to be additionally deflected to initiate the turn [32].
Ground spoilers are extended to their maximum angle during landing to reduce lift as
much as possible. The additional drag caused by the ground spoilers, on the other hand,
plays only a small role during landing [32]. The lift on the wing in the area of the spoilers
is eliminated since the airflow can no longer be applied here. During so-called firm landing
procedures, the landing gear is pressed more strongly onto the ground for increasing the
efficiency of the brakes [32]. After touch-down, it is important to reduce as much lift as
possible as quickly as possible so that all existing spoiler panels are usually extended to
their maximum angle as ground spoilers.
Speedbrakes &
Ground Spoiler

Roll Spoiler
Flaps
Slats
eron

Figure 2. Primary (grey) and secondary (red, green) flight control surfaces on a commercial aircraft.

Modern transport aircraft use flap spoilers: a flap spoiler is a panel affixed to the
wing upper surface trailing edge region that, when deflected upwards (the panel rotating
about its leading edge), causes the flow to separate over the wing surface in a controlled
manner, resulting in a decrease in lift and an increase in drag. Due to the current inability
to simulate separated flows, speed brake aerodynamic properties are the most challenging
of the airplane control surfaces to anticipate. Speed brakes have a number of characteristics
that make them ideal for lateral control in aircraft: large rolling moments are produced by
speed brakes, and speed brakes are an alternative to ailerons for full roll control, allowing
full span flaps to be used, with apparent STOL uses and benefits. Ailerons provide an
undesirable yawing moment, whereas speed brakes produce a favourable yawing moment.
Furthermore, speed brakes, as opposed to ailerons, are usually more effective at high speeds
and are less prone to suffer from aeroelastic effects [15].
Unfortunately, several of the aerodynamic characteristics of spoilers increase the
complexity of predicting the unsteady flow field generated by airfoils with deflected speed
brakes. The speed brake control effectiveness is non-linear: the lift decrease is a non-linear
function of the speed brake deflection. This non-linearity is most noticeable when speed
brakes are employed in conjunction with a deflected flap. This requires the speed brakes to
be integrated with other control surfaces (such as ailerons) in order to give linear control
(necessary to satisfy the pilot and autopilot functions). Historically, this feature has made
it difficult to achieve the use of speed brakes for full roll control in the presence of full
span flaps. When the speed brake is deflected, the resulting turbulent wake is extremely
unsteady. The wing interacts with the horizontal tailor and buffets (i.e., aerodynamics-
induced vibrations) can be caused by themselves. Furthermore, the time delay between the
deflection of the spoiler and the reduction in lift causes a delay in the aircraft’s response to
Aerospace 2022, 9, 263 6 of 19

speed brake deflection. Finally, the change in wing pitching-moment with spoiler deflection,
as well as the influence of the spoiler wake on the horizontal tail, can generate unacceptable
pitching moments.

Preliminary Considerations of Speed Brakes

The extension of the individual panels causes an increase in drag, as the surfaces
are pressed into the air flowing around them. The impact of speed brakes on forces can
be reduced to that on the coefficients of the forces. Mainly, speed brakes influence flight
performance through four different coefficients: the lift coefficient c L , drag coefficient c D ,
pitch moment coefficient c My , and the roll moment coefficient c Mx .
The total change in drag ∆FD [N] is the sum of the impact of each control surface,
where n denotes the number of speed-brake panels
n
∆FD = ∑ ∆FD,i . (5)
i =1

Assuming constant speed v [m s−1 ], wing area S [m2 ], and air density ρ [kg m−3 ],
the change in drag also results in a change in the drag coefficient c D [38]. Since all parame-
ters except for the drag force are constant on the entire aircraft, the change in the total drag
coefficient can also be calculated by

2 ∆FD 2 ∑in=1 ∆FD,i

∆c D = 2
= . (6)
v ρA v2 ρ A

The extension of speed brakes also causes a reduction of the lift at the position of the
extended speed brakes, since in these areas of the wing the flow no longer completely flows
around the profile. The lift is even omitted with a correspondingly large extension angle.
The sum of the individual changes in lift forces ∆FL,i [N] results in the total change in lift
∆FL [N]
n
∆FL = ∑ ∆FL,i (7)
i =1

As with drag, changes in the lift coefficient c L can be calculated from the lift change.
Since the speed, air density, and wing area can be assumed to be uniform for the entire
aircraft, the change in the total lift coefficient can also be calculated in this way [38]

2 ∆FL 2 ∑in=1 ∆FL,i

∆c L = = . (8)
v2 ρ A v2 ρ A

The moment around the pitch axis is also influenced by the use of the speed brakes.
An increase in drag and a reduction in lift at the location of the speed brake (with longitudi-
nal distance dy,i [m] to the point of action of the force on the i-th control surface) generates
a moment about the average point of application of lift on the aircraft. Thus, the change in
pitch moment ∆My can be calculated by [32]
n
∆My = ∑ (∆FL,i dy,i ) (9)
i =1

The change in the pitch moment coefficient ∆c My depends on the reference wing depth
lµ [m]
2 ∆My 2 ∑in=1 ( FL,i dy,i )
∆c My = 2 = . (10)
v ρ A lµ v2 ρ A l µ
The roll moment Mx only affects the spoiler panels, which are used as roll spoilers.
By reducing the lift at the location of the spoiler, the wing sinks on that side, which provides
a moment along the longitudinal axis of the aircraft. The change in roll moment ∆Mx can
Aerospace 2022, 9, 263 7 of 19

be calculated according to the same principle as the change in pitch moment by multiplying
the change in the lift force ∆FL,i by the respective distance d x,i to the longitudinal axis
n
∆Mx = ∑ (∆FL,i dx,i ). (11)
i =1

The change in the roll moment coefficient ∆c Mx again depends on the reference wing
depth lµ [m]
2 ∆Mx 2 ∑in=1 ( FL,i d x,i )
∆c Mx = 2 = . (12)
v ρ A lµ v2 ρ A l µ

3. Methodology
3.1. Speed Brakes on NASA Model B747
The B747 consists of six speed brakes per wing (see Figure 3). Six speed brakes per
wing are extended synchronously, whereby the inner sped brakes can only be extended
by 20◦ compared to the outer speed brakes with a maximum angle of 45◦ (see Table 1). In
addition to the theoretical calculation already explained, Hanke [30] calculates the modified
coefficients for each speed brake panel i separately using Boeing-specific performance
coefficients. These coefficients depend on speed brake angle δi , altitude, angle of attack α,
flap configuration, and speed. The values are digitized based on the graphic representation
by Hanke. We interpolate these coefficients between provided discretized values with
increments of 10,000 feet for altitude, 5◦ for speed brake angle, and 5◦ for the angle of
attack, as well as for Mach numbers (0.5 < Mach < 0.85) with ∆ Mach = 0.1.

6
5

4
3
2
1

Figure 3. Assumed number and position of spoilers on a single B747 wing.

The change in the lift coefficient for the i-th panel ∆c L,i is expressed as
 
(c ) M L E,i
∆c L,i = Kδ,i (∆c L,i )45 L,i F (13)
(c L,i ) M0 L R,i L,GE

where Kδ,i determines speed-brake-specific effectiveness factors and (∆c L,i )45 describes the
change in lift coefficient due to a speed brake extension to 45◦ . Speed effects are considered
(c L,i ) M
with (c L,i ) M0
as change in the lift coefficient due to Mach number M compared to the change
in the lift coefficient at M = 0 (symbol M0 in Equation (13)). Since there is no change in
 at M
the lift coefficient  = 0, the ratio takes the value 1 at M = 0. The aeroelastic effect is
L E,i B only effects ∆c
considered with L . The ground effect factor KGE L,i below an altitude of
R,i
B = 0. Below 90 feet we assume F
90 feet and therefore we assume KGE L,GE = 1 [30].
With KGE = 0, the change in drag coefficient ∆c D,i depends on the change of the drag
B

coefficient by extending the speed brakes at an angle of attack of 4◦ (∆c D,i (α = 4)) and on
d(∆c D,i )
the rate of change of drag coefficient with angle of attack dα [30].

d(∆c D,i )
 
(c D,i ) M
∆c D,i = ∆c D,i (α = 4) + ( α − 4) (14)
dα (c D,i ) M0
Aerospace 2022, 9, 263 8 of 19

m,GE = 1, the change of the pitch moment coefficient ∆c My,i

Assuming KGEB = 0 and F

due to extended speed brake i is defined as [30]

 
(cm,i ) M ME,i
∆c My,i = (Kδ,i )m (∆cm,i )45 (15)
(cm,i ) M0 MR,i

where (Kδ,i )m describes the effectiveness of the respective speed brake panel and (∆cm,i )45
the change in the basic pitching moment, with extended speed brakes by 45◦ . Again,
(cm,i ) M
the ratio refers to the effect of the change in Mach number. The aeroelastic effect is
(cm,i ) M0
 
M
considered by the ratio M E,i for each speed brake i.
R,i
The same procedure is used for the change in the roll moment coefficient ∆c Mx,i with
extended speed brake i [30]:
 
(c ) M R E,i
∆c Mx,i = (Kδ,i )l (∆cl,i )45 l,i (16)
(cl,i ) M0 R R,i

∆c Mx,i depends on the effectiveness of the individual speed brake panel (Kδ,i )l and on
the change in the roll moment coefficient (∆cl,i )45 by extending speed brake i panel to 45◦ .
(cl,i ) M
The effect of Mach number is described by the component and the aeroelastic effect
(cl,i ) M0
 
R
on the rolling moment coefficient is described by the component R E,i .
R,i
After the impact of an extended speed brake i on flight performance has been obtained
for the B747, we determine the corresponding effective panel area of the B747 for different
aircraft types (using Equation (18)) and transfer the model, assuming a similar impact on
lift, drag, and moment for a similar effective panel area. The aerodynamic properties of
the wing and thus the properties of the spoiler are determined by the airfoil shape, i.e., the
depth and thickness of the wing. Since the maximum lift of an airfoil shape is mainly
determined by the maximum thickness and camber of an airfoil [39], the maximum airfoil
thickness of the wing is chosen as the reference value for comparing speed brake panels
as an averaged value in the middle of the speed brake width lw,i [m]. This assumption is
underlined by the test series of NASA on the flow influence of speed brakes [40]. Here
it was found that an extended speed brake does not influence the pressure ratio on the
airfoil surface in front of the speed brake [40]. Thus, it can be assumed that the forward
airfoil continues to maintain its effectiveness in terms of maximum lift. Here, as in other
tests [41], there is only a reduction in the maximum possible lift. Further tests by NASA
in the subsonic range up to a Mach number of 0.94 result in a speed brake-induced loss
of lift between ∆c L = 0.0015 and ∆c L = 0.0025 for small and large extension angles,
respectively [42]. Considering the definition of the Reynolds number Re

ρvl
Re = (17)
η

(where ρ, v, l and η are the fluid density [kg m−3 ], the fluid speed [m s−1 ], the characteristic
linear dimension [m], and the dynamic viscosity of the fluid [N s m−2 ], respectively) it can
be concluded that when comparing different speed brake panels, the wing depth at this
point has little effect on the speed brake impact and the variation of the speed brake angle
is more relevant [42]. Note that this fact is only valid for laminar flows adjacent to the
profile in the subsonic range, a constant angle of attack α, and a clean flap configuration.
Furthermore, the maximum speed must be well below the speed of sound (below Mach
0.94) [30].
To comparatively determine the speed brake of a different aircraft type, the geometric
dimensions of the speed brake panel, as well as the profile thickness, are determined.
Aerospace 2022, 9, 263 9 of 19

Subsequently, the speed brake angle δi,B747 is calculated for the respective B747 reference
panel with which it reaches the same effective area Ai [m2 ]

Ai = bi li sin δi,B747 (18)

where bi and li are the speed brake width [m] and speed brake length [m], respectively.
Using these angles, the variables for the coefficients in Equations (6), (13), (15) and (16)
are calculated.

3.2. Geometric Data of NASA Model B747

For the B747, the following geometric data with impact on speed-brake-specific flight
performance is assumed.

Table 1. Assumed geometric dimensions of the B747-100 speed brakes (Data collected from [43]).
The Panels are numbered from outside to inside (see Figure 3).

Panel Length Width Max. Angle Max. Area Mean Wing Depth
Number li [m] bi [m] δi [◦ ] Ai [m2 ] lw,i [m]
1 1.109 1.905 45 1.495 0.37
2 1.109 1.905 45 1.495 0.377
3 1.109 1.905 45 1.495 0.488
4 1.109 1.905 45 1.495 0.559
5 1.397 2.286 20 1.092 0.936
6 1.397 2.286 20 1.092 1.187

3.3. Transfer of the B747 Model to Other Aircraft Types

The following aircraft types (and data sources) are considered in this study: the Airbus
models A310 [44], A319 [44,45], A320 [44,45], A321 [44,45], A330 [44,46] and A380 [44,47];
the Boeing models B737 [48], B767 [49,50], B777 [36,51–53]; Bombardier CRJ900 [54,55];
and Embraer E170 [37] and E190 [56]. Geometric information for panel sizes is taken
from [44], whereas operational information (i.e., possible configurations for speed brake
and maximum speed brake angle) are taken from aircraft operational manuals. Because the
E190 and the E170 have identical performance data, wing and speed brake dimensions
(see [56]), there should be no differences in speed brakes on lift and drag. During the
literature research, a few special features stood out. In comparison to Boeing aircraft types,
where speed brake angles δi can take arbitrary values between δi = 0◦ and δi,max , Airbus,
Bombardier and Embraer only enable discrete values of δi in full, three quarter, half and
quarter positions [34], except for A310, which has 11 positions [35]. Note that only those
panels used as speed brakes are listed in Tables 2 and 3.
In Tables 2 and 3, aircraft-specific speed brake information is summarized. Due to
the geometry, we limited the list to a single wing side. The speed brakes are recorded
according to the manufacturer-specific numbering. Airbus numbers the speed brake from
inside to outside, whereas Boeing, Bombardier, and Embraer number the speed brake
from outside to inside. Besides the number of speed brakes per wing, the mean wing
depth lw,1 at the position of each speed brake allows the allocation of each speed brake
to a corresponding B747 speed brake. The individual aircraft speed brake length li and
width bi are used to calculate the individual speed brake area Ai . The individual speed
brake area in turn is used to estimate the corresponding B747 speed brake angle δi,B747 .
The described speed brake panel in Tables 2 and 3 are limited to those spoilers with speed
brake function. For example, the inner and outer panels No. 1 and No. 5 of A319 are not
used as speed brakes [45] and panel 4 of B767 is only used at low speeds and not relevant
for this paper [50].
Aerospace 2022, 9, 263 10 of 19

Table 2. Transfer of Airbus-specific speed brakes and their mean wing depth lw , i to the speed brakes
of the B747 (No, B747) parameterized by NASA (numbered from outside to inside). Additionally,
Airbus speed brake length li , width bi , area Ai , maximum speed brake angles δi,max , and the resulting
maximum speed brake angles of the corresponding B747 speed brake δi [◦ ] are given.

Panel bi li δi Ai No lw,i δi
max max B747 B747
[m] [m] [◦ ] [◦ ] [m] [◦ ]
A310
1 1.62 0.743 35 0.69 5 1.006 12.485
2 1.69 0.904 35 0.876 4 0.746 24.488
3 1.87 0.684 35 0.734 2 0.428 20.31
4 1.862 0.684 35 0.731 2 0.41 20.277
5 1.526 0.549 0 0.286 2 0.388 7.786
6 1.529 0.549 0 0.287 1 0.377 7.805
7 1.478 0.549 0 0.277 1 0.289 7.542
A319
1 1.763 0.626 17.5 0.633 2 0.381 17.434
2 1.661 0.625 25 0.596 1 0.297 16.37
3 1.528 0.625 25 0.548 1 0.343 15.016
4 1.531 0.627 0 0.55 1 0.251 15.096
A320
1 1.763 0.626 12.5 0.633 2 0.381 17.434
2 1.661 0.625 25 0.596 1 0.343 16.37
3 1.528 0.625 25 0.548 1 0.297 15.016
4 1.531 0.627 0 0.55 1 0.251 15.096
A321
1 1.763 0.626 25 0.633 2 0.381 17.434
2 1.661 0.625 25 0.438 1 0.343 11.986
3 1.528 0.625 25 0.548 1 0.297 15.016
4 1.531 0.627 0 0.55 1 0.251 15.096
A330
1 2.43 0.693 25 0.711 5 1.054 12.871
2 2.12 0.705 30 0.857 4 0.682 23.913
3 2.163 0.707 30 0.878 4 0.597 24.531
4 2.165 0.709 30 0.88 3 0.541 24.614
5 2.193 0.71 30 0.893 3 0.49 24.998
6 2.209 0.712 30 0.902 2 0.445 25.254
A380
1 2.43 1.296 20 1.077 6 1.655 19.713
2 2.374 1.229 20 0.998 6 1.34 18.211
3 1.972 1.117 20 0.754 5 1.154 13.649
4 2.195 1.095 20 0.882 5 1.042 14.919
5 2.084 1.056 20 0.752 5 0.985 13.629
6 2.782 0.883 45 1.736 4 0.907 55.241
7 2.503 0.737 45 1.305 4 0.878 38.13
8 2.35 0.581 45 1.329 4 0.788 38.956

The determination of the change in the roll moment coefficient ∆c Mx,i is made by
taking into account the corresponding maximum B747 deflection angle δi,max to support
the rolling function. This may differ from the maximum available angle when used as a
speed brake.
Aerospace 2022, 9, 263 11 of 19

Table 3. Transfer of aircraft-specific speed brakes and their mean wing depth lw,i to the speed
brakes of B747 (No, B747) parameterized by NASA (numbered from outside to inside). Additionally,
individual aircraft speed brake length li , width bi , area Ai , maximum speed brake angles δi,max , and
the resulting maximum speed brake angles of the corresponding B747 speed brake δi [◦ ] are given.
Note that the E190 and the E170 have the same performance data as well as wing and speed brake
dimensions [56], so no differences in speed brakes on lift and drag can be expected.

Panel bi li δi Ai No lw,i δi
max max B747 B747
[m] [m] [◦ ] [◦ ] [m] [◦ ]
B737
2 1.05 0.567 19.5 0.199 3 0.499 4.495
3 1.5 0.594 19.5 0.297 3 0.509 6.732
4 1.012 0.612 24.5 0.257 3 0.538 5.807
5 1.008 0.632 24.5 0.264 3 0.529 5.973
B767
1 1.532 0.736 45 0.797 3 0.489 18.318
2 1.532 0.736 45 0.797 3 0.537 18.318
3 1.532 0.736 45 0.797 4 0.58 18.318
5 1.768 0.952 17 0.492 4 0.93 11.192
6 1.684 0.952 17 0.496 5 1.059 8.04
B777
1 2.032 0.642 60 1.13 4 0.559 32.329
2 1.95 0.641 60 1.082 4 0.62 30.795
3 1.992 0.655 60 1.13 4 0.681 32.305
4 1.91 0.661 45 0.893 4 0.729 24.991
5 1.958 0.667 0 1.131 4 0.802 32.347
6 1.995 1.011 60 1.747 6 1.252 33.172
7 1.995 1.011 60 1.747 6 1.562 33.172
CRJ900
1 0.815 0.199 50 0.124 1 0.3 3.373
2 0.85 0.199 50 0.13 1 0.268 3.519
E170/E190
3 1.433 0.712 30 0.521 1 0.34 9.864
4 1.309 0.655 30 0.439 1 0.317 8.291
5 1.122 0.616 30 0.353 1 0.274 6.658

Figure 4 depicts the change in the lift coefficient ∆c L,i and the drag coefficient ∆c D,i for
an Airbus A310 aircraft. depending on speed brake angle, at 30,000 feet altitude and Mach
0.82, with the flaps in clean configuration and the angle of attack at α = 0◦ . As expected,
the total lift coefficient decreases and the drag coefficient slightly increases with increasing
speed brake angle. Since Airbus numbers speed brakes from inside to outside, the effect of
inner speed brakes is larger than the effect of outer speed brakes. Furthermore, the effect of
extended speed brake on ∆c L,i and ∆c D,i increases with increasing speed brake angle δi .
For the twelve aircraft types mentioned, it was possible to determine the change in
drag coefficients and lift coefficients, which can be classified in the respective performance
field of the aircraft types and thus allow correspondingly good comparison possibilities of
the individual types (see Figure 5 for maximum values).
Aerospace 2022, 9, 263 12 of 19

0.000
−0.002

cL i, ΔcD i
,
−0.004

,
−0.006

Δ
−0.008
0 10 20 30
Spoiler angle δi

Figure 4. Change in lift coefficient ∆c L,i (blue) and drag coefficient ∆c D,i (orange) for different speed
brake angles of an Airbus A310 aircraft. A clean flap configuration, a Mach number of 0.82, an angle
of attack of α = 0◦ , and an altitude of 30,000 ft are assumed. A310 speed brakes are numbered from
inside to outside (scatter markers).

0.03 A310
max. ΔcD [a.u.]

A319
0.02 A320
A321
A330
0.01 A380
B737
0.00 B767
B777
max. ΔcL [a.u.]

−0.05 CRJ9
E170
−0.10 E190
B747
−0.15

0 5 10 15
max. Speedbrake Area [m2]

Figure 5. Change in maximum total drag coefficient ∆c D (top) and maximum total lift coefficient ∆c L
(bottom) depending on maximum speed brake area A for thirteen different aircraft types assuming
a clean flap configuration, a Mach number of 0.82, an angle of attack of α = 0◦ , and an altitude of
30,000 ft. Note the different scales on the y-axes.

4. Trajectory Modelling with Speed Brakes

The extension of speed brakes and their effects on the descent phase were then inves-
tigated using the modelled trajectories. For this purpose, the flight performance model
SOPHIA (Sophisticated Aircraft Performance Model) was used. SOPHIA applies the
methodology published in [2], but obtains data for the drag polar and the maximum avail-
able thrust as a function of altitude and speed from the open-source flight performance
model OpenAp [3]. As a reference, a uniform flight plan was set up for all aircraft types,
to be compared in order to carry out an analysis limited to the speed brake systems. A
flight from cruise to Munich Airport (specifically, the approach point MAGAT 5000 ft,
which serves the approach to runway 08L) was set up as a reference flight. For reasons
of energy efficiency, the descent should be conducted as long as possible as a continuous
descent with idle thrust. However, in controlled airspace, a maximum speed of 250 kt
is permitted below an altitude of 10,000 ft. Therefore, before reaching 10,000 ft, a flight
segment is usually flown during which the aircraft reduces its descent speed [57]. Since the
on-board computers always fly at a speed that is 10 kt below this speed limit [57], a speed
of 240 kt is set as the deceleration target speed. This deceleration phase with a maximum
flight path angle (γ = 0.5◦ ) and idle thrust was modelled with SOPHIA at 10,000 ft altitude.
To reduce speed in this phase, speed brakes are extended at different speed brake angles δi .
Per scenario, all speed brakes per wing are extended at the same angle. Figures 6 and 7
show the impact of speed brake angle δi in this deceleration phase on the distance and time
of the descent phase on a B777 aircraft. The path angle was calculated from the equilibrium
of forces and restricted to 0.5◦ in the deceleration phase during descent.
As with the reference B747, the speed brakes can be driven to any angle between 0
and the respective maximum angle (see Table 3) when used as speed brakes. An effect
Aerospace 2022, 9, 263 13 of 19

that cannot be reproduced by SOPHIA is the so-called “blowdown effect”, in which the
speed brakes cannot be extended to these angles despite being set higher, because the
air pressure on the surfaces pushes them downwards [58]. To enable a calculation by
SOPHIA nevertheless, it is assumed that the individual speed brakes can be set to the
maximum values.

25000 0%
25%
20000 50%
75%
Altitude [ft]

100%
15000

10000

5000
−600 −500 −400 −300 −200 −100 0
Time to go [s]

Figure 6. Descent profiles of a B777 aircraft as a function of time with different speed brake angles δi
(indicated by color). At 25,000 ft, the B777 flies with Mach 0.75; below 10,000 ft 240 kt are mandatory.
A maximum deceleration path angle of 0.5◦ and 9 t payload are considered.

Figure 6 indicates that B777 benefits from its large speed brake area and that it thus
achieves a maximum reduction in descent time of 265 s. This results in a reduction of the
descent distance of 50.81 km between retracted flaps (orange) and maximum flap position
(black). The increased effectiveness of a larger speed brake angle δi induces a displacement
of the top of descent towards a later initiation of the descent (see Figure 7).

25000 0%
25%
20000 50%
75%
Altitude [ft]

100%
15000

10000

5000
−120 −100 −80 −60 −40 −20 0
Distance to go [NM]

Figure 7. Descent profiles of a B777 aircraft as a function of distance at different speed brake angles
(indicated by color). A cruising altitude of 25,000 ft, a speed of Mach 0.75 and below 10,000 ft, 240 kt
are calculated. A maximum deceleration path angle of 0.5◦ and a payload of 9 t are considered.

To determine the possible impact of speed brakes on fuel consumption, flights from
Dresden starting at a cruising altitude of 25,000 ft are examined in Figure 8. By shortening
the descent section, the aircraft remains at the cruising altitude for longer and thus flies
with a thrust setting for cruising flight for longer, which is above the idle setting during
descent. As a result, the total amount of fuel required for the flight increases. It follows
that the use of the speed brakes leads to a lengthening of the cruise segment and this,
in turn, leads to a larger amount of fuel required for the entire flight. At the same time,
however, a shortening of the descent time becomes possible. The time savings are very
small compared to the total flight duration. The reason for this are the steep angles that
can be flown by using the speed brakes, especially in the long-haul models. Due to these
steep angles, the descent phases have a low horizontal speed component, which is why
less distance is covered in these phases. While this also leads to a lengthening of the cruise
flight with a greater horizontal speed component, it reduces the savings potential of the
total flight duration.
Aerospace 2022, 9, 263 14 of 19

A310

Time sa%ings [s] Additional fuel [kg]

600 A319
A320
400 A321
A330
200 A380
B737
0 B767
B777
CRJ9
−20 E170
E190
−40 B747

−60
20 40 60 80 100
S oiler extension [%]

Figure 8. Change in fuel consumption (top) and time savings (bottom) of the entire flight as a
function of the speed brake position. (Speed at 25,000 ft: 0.75 Ma; at 10,000 ft: 240 kt; deceleration
path angle: 0.5◦ ; payload: 9 t).

The additional fuel consumption for long-haul aircraft (A330 and B777) increases
almost linearly with the flap angle used and the resulting larger effective area of the speed
brakes (see Figure 9). From this follows an almost linear correlation between change in
drag and flap angle for large aircraft. For smaller aircraft models such as the A320 and E170,
there is no such linear relationship for large flap angles. Airbus A320 requires additional
fuel for the same speed brake area compared to Embraer E170.
Additional fuel [kg]

200

100 A320
A330
B777
E170
0
0.0 2.5 5.0 7.5 10.0 12.5
Speedbrake area [m2]

Figure 9. Comparison of four Aircraft types regarding the increase in additional fuel consumption
depending on effective speed brake area. For initial conditions, see Figure 8.

Validation
In addition to flight performance calculations, the model was implemented in an
aircraft trajectory optimizer using the robust optimal control theory for continuous descent
optimization under real weather conditions [59]. After the new speed brake model was
implemented in the flight performance model SOPHIA and in [59], it needed to be vali-
dated by modelling a real flight with SOPHIA and comparing the dependent variables,
for example, fuel flow, with the real values. Additionally, the aim of validation should
have been to prove the benefit in the precision of flight performance modelling, compared
to ∆c D = 0.02 as used in other studies [14]. However, the speed brake angle is not part
of usual data sets describing a flight (e.g., ADS-B data). Only Flight Operational Data
(FODA) contain this information. FODA remain in the possession of the pilot flying and
are actually only released for safety assessments. For this reason, the authors are only in
possession of a single data set with extended speed brakes during descent. This flight is by
a B748F, which is not significantly different to the B747-100 experimental aircraft, so that
the successful transition of the reference aircraft B747-100 to other aircraft types can only
be shown to a limited degree. However, since this is the only available data set, it is used
here. The authors are happy to validate the model with another data set provided. For
this purpose, we modelled a real Boeing B748F flight from Frankfurt, Germany to Boston,
USA, operated in March 2018 with speed brakes used in the descent phase. This flight is
provided as FODA with wind information (i.e., the difference between ground speed and
true airspeed), fuel flow, and speed brake angle, amongst others. Since the study focuses
on flight performance with speed brake deflection, we only model the descent phase. This
Aerospace 2022, 9, 263 15 of 19

simplifies the mass estimation of the aircraft, which is not given in FODA. Based on the
fuel flow in cruise flight, we determined a payload of 77,705 kg. For the entire flight, we
were able to determine a trip fuel of 70,000 kg and thus set the reserve fuel at 10,000 kg
(10% contingency plus 3000 kg for holding). For the remaining descent phase, the fuel
burn was about 13,500 kg. Thus the total mass of the aircraft is 101,205 kg plus operating
empty weight. According to the FODA, speed brakes were deflected at 11,000 ft altitude
for 733 s (12.2 min), indicated by speed brake handle up to δ = 35◦ . Thereby, vTAS was
reduced by 79 kt from 248 to 169 kt. Afterwards, the speed brakes were set to δ = 4.5◦ for
the remaining flight.
As is usual in FODA, thrust is only provided as thrust lever angle [◦ ]. For this reason,
for each time step, we calculated the thrust required from the equilibrium of the equations
of motion (Equation (4)). Since SOPHIA uses the vTAS as a controlled variable, the aim is to
calculate a profile that, despite the lack of lift and drag information in the FODA, reproduces
the real FODA flight in terms of temporal and spatial altitude variation as well as fuel
consumption. We applied the given speed brake angle δ equally and synchronously to all
speed brakes, whereby the inner speed brake can be extended to a maximum of δ = 20◦ .
Additionally, we modelled the flight in the same way, except for the calculation of the lift
and drag coefficients, which were replaced by the common assumption of ∆c D = 0.02 and
∆c L = 0.00 [14].This parameter setting is highlighted in orange in Figures 10–12.

30000 real
ΔcD=const.
25000 modelled

20000
Altitude [ft]

15000

10000

5000

0
Speed brake handle [°]

30
20
10
0
−2000 −1750 −1500 −1250 −1000 −750 −500 −250 0
Time to go [s]

Figure 10. Validation of the developed speed brake model: Comparison of altitude over time of a
real B748F flight (black) with modelled flights. Blue: using the aircraft-type-specific speed brake
model; orange: assuming ∆c D = 0.02 and ∆c L = 0.00. Bottom: the speed brake handle provided by
the real flight as FODA data. Vertical lines indicate a time frame (12 min) with significantly deflected
speed brakes.

With the aircraft-type-specific speed brake model, depending on the angle of attack
and flap position, the aircraft model was able to correctly react on speed brake deflection,
and follows the real flight profile more closely than the simplified model usually used
in flight performance modelling [14] (see Figures 10 and 11), although the profile is not
perfectly matched. Specifically, the spontaneous switch from δ = 27◦ to δ = 4.5◦ (indicated
by the second vertical line in Figures 10 and 12) induces deviations from the real flight
profile, perhaps due to an overestimation of the controller’s inertia in reaching the new
target speed, or due to an underestimation of resistance at low speed brake angles. In both
modelled profiles (orange and blue), the speed at this point is higher than in real flight, and
converges to the real speed at second 1836.
Aerospace 2022, 9, 263 16 of 19

400
ΔcD=const.
200 modelled

Difference in altitude [ft]

−200
−400
−600
−800
−1000
−1200
−2000 −1750 −1500 −1250 −1000 −750 −500 −250 0
Time to go [s]

Figure 11. Differences in altitude between a real B748F flight and two modelled flights as a function
of time. Blue: modelled with the aircraft-type-specific speed brake model; orange: modelled with
∆c D = 0.02 and ∆c L = 0.00. Vertical lines indicate a time frame of 12 min with significantly deflected
speed brakes.

Besides the vertical difference between the real data and the modelled flights with
differences up to 1150 ft in the case of ∆c D = 0.02 and ∆c L = 0.00 (orange), Figure 11 shows
the ability of SOPHIA’s controller to reach a target altitude which is defined every second.
Furthermore, the improvement of using the aircraft-type-specific speed brake model (blue)
can by quantified to a maximum difference in altitude of 800 ft (Figure 11).
Figure 12 clearly indicates that our model (blue) calculates larger values of c D as soon
as speed brakes are deflected, compared to [14] (orange). Furthermore, the modelled impact
on the lift coefficient results in slightly lower values of c L (mean c D = 0.476) compared
to neglecting the speed brake influence on the lift coefficient (orange, mean c D = 0.479).
This impact is surprisingly small, maybe due to the short time with strongly deflected
speed brakes.
1.0
0.07 cD wi h ΔcD=cons . 0.9 cL wi h ΔcL=0
cD modelled cL modelled
0.06 0.8
Drag coefficien cD [a.u.]

Lif coeffien cL [a.u.]

0.7
0.05
0.6
0.04 0.5
0.03 0.4
0.3
0.02
0.2
0.01 0.1
−1750 −1500 −1250 −1000 −750 −500 −250 0 −1750 −1500 −1250 −1000 −750 −500 −250 0
Time o go [s] Time o go [s]

Figure 12. Comparison of drag coefficient (left) and lift coefficient (right) during the B748F vali-
dation flight. Blue: modelled with the aircraft-type-specific speed brake model; orange: assuming
∆c D = 0.02 and ∆c L = 0.00. The aircraft-type-specific speed brake model results in increased drag
and a slightly decreased lift when speed brakes are deflected. Vertical lines indicate time frame with
significantly deflected speedbrakes.

5. Conclusions
In this study, a rule for the calculation of the effect of speed brakes on flight perfor-
mance was made possible for twelve aircraft types and applied in trajectory modelling with
the flight performance model SOPHIA. For this purpose, a model developed in 1971 for
the B747-100 within the framework of a NASA research project was transferred to the other
aircraft types. Graphical parameterizations of this speed brake model of the B747 were
digitized. With known dimensions of the speed brakes of any aircraft and known profile
depth of the wing at the position of the speed brakes, as well as taking into account the
possible speed brake angles, corresponding speed brakes and angles on the NASA model
B747 could be determined, and these reflect the different operational and performance
areas of the respective aircraft models well.
Using Computer-Aided Design (CAD) drawings from aircraft manufacturers, the ge-
ometric information of the speed brakes of twelve different aircraft types could be deter-
mined with a high degree of reliability. For the transfer of the NASA model to the other
Aerospace 2022, 9, 263 17 of 19

aircraft types, however, restrictions had to first be made to achieve minimal environmental
influences on the performance data. Since these limitations are typical for Continuous
Descent Operations (CDO) procedures, the accuracy of the results was not significantly
affected. However, a closer look at the digitized parameters of the NASA B747 model shows
a loss of accuracy, which should be reflected in the determined drag and lift coefficients.
In validation, we modelled a real descent profile with known speed brake deflection
during the descent phase and compared the resultant profile following our modelled
impact on c L and c D with the modelled profile assuming a constant factor of c D = 0.02 for
maximum speed brake deflection and c L = 0. Due to unknown information of aircraft mass,
inertia, and thrust, we could not accurately follow the real profile, but gained improvements
compared to the usual model approach. However, missing real flight performance data
including speed brake handle information of other aircraft types hampers our ability to
completely validate the model.
The implementation of the effect of speed brakes on lift and drag in the flight per-
formance model SOPHIA now allows a dynamic trajectory optimization, especially in
the descent phase. Dynamic changes of both external boundary conditions, such as the
influence of the weather or the required arrival time, and internal parameters, such as
the cost index, can now be taken into account in the trajectory optimization, even if such
changes induce unexpected level flight segments or acceleration phases. We have thus
extended the selection of objective functions in SOPHIA and can now devote ourselves to
new research questions.

Author Contributions: Conceptualization, J.R.; methodology, J.R. and T.S.; software, J.R.; validation,
J.R. and T.S.; formal analysis, J.R. and T.S.; investigation, J.R. and G.C.; resources, H.F.; data curation,
J.R. and T.S.; writing—original draft preparation, J.R.; writing—review and editing, J.R. and S.K.;
visualization, J.R.; supervision, J.R. and H.F.; project administration, J.R.; funding acquisition, J.R.
and H.F. All authors have read and agreed to the published version of the manuscript.
Funding: This work is a part of the project “Optimized CDO under Uncertain Environmental and
Mission Conditions” (project number: 327114631) financed by German Research Foundation (DFG).
Conflicts of Interest: The authors declare no conflict of interest.

References
1. ICAO. Continuous Descent Operations (CDO) Manual—Doc 9931/AN/476; ICAO: Montreal, QC, Canada, 2010; Volume 1.
2. Rosenow, J.; Förster, S.; Lindner, M.; Fricke, H. Multicriteria-Optimized Trajectories Impacting Today’s Air Traffic Density,
Efficiency, and Environmental Compatibility. J. Air Transp. 2018, 27, 8–15. [CrossRef]
3. Sun, J.; Hoekstra, J.M.; Ellerbroek, J. OpenAP: An Open-Source Aircraft Performance Model for Air Transportation Studies and
Simulations. Aerospace 2020, 7, 104. [CrossRef]
4. Kamo, S.; Rosenow, J.; Fricke, H.; Soler, M. Robust CDO Trajectory Planning under Uncertainties in Weather Prediction. In
Proceedings of the 14th USA/Europe Air Traffic Management Research and Development Seminar (ATM Seminar), Virtual Event,
20–23 September 2021.
5. Kaiser, M.; Rosenow, J.; Fricke, H.; Schultz, M. Tradeoff between optimum altitude and contrail layer to ensure maximum
ecological en-route performance using the Enhanced Trajectory Prediction Model (ETPM). In Proceedings of the 2nd Inter-
national Conference on Application and Theory of Automation in Command and Control Systems (ATACCS), London, UK,
29–31 May 2012.
6. Serafino, G. Multi-objective Aircraft Trajectory Optimization for Weather Avoidance and Emissions Reduction. In Proceedings of
the Modelling and Simulation for Autonomous Systems, Prague, Czech Republic, 29–30 April 2015; Springer: Berlin/Heidelberg,
Germany, 2015; pp. 226–239.
7. Ng, H.K.; Sridhar, B.; Grabbe, S.; Chen, N. Cross-polar aircraft trajectory optimization and the potential climate impact. In
Proceedings of the 2011 IEEE/AIAA 30th Digital Avionics Systems Conference (DASC), Seattle, WA, USA, 16–20 October 2011;
pp. 3D4-1–3D4-15.
8. Wickramasinghe, N.K.; Miyamoto, Y.; Harada, A.; Kozuka, T.; Shigetomi, S.; Miyazawa, Y.; Brown, M.; Fukuda, Y. Flight
Trajectory Optimization for Operational Performance Analysis of Jet Passenger Aircraft. Trans. Jpn. Soc. Aeronaut. Space Sci.
Aerosp. Technol. Jpn. 2014, 12, a17–a25. [CrossRef]
9. Matthes, S.; Grewe, V.; Lee, D.; Linke, F.; Shine, K.; Stromatas, S. ATM4E: A concept for environmentally-optimized aircraft
trajectories. In Proceedings of the Greener Aviation, Brussels, Belgium, 11–13 October 2016.
Aerospace 2022, 9, 263 18 of 19

10. Ye, B.; Wang, Z.; Tian, Y.; Wan, L. Aircraft-specific trajectory optimization of continuous descent approach for fuel savings. In
Proceedings of the SII 2017—2017 IEEE/SICE International Symposium on System Integration, Taipei, Taiwan, 11–14 December
2017; pp. 751–756. [CrossRef]
11. Luchinskiy, D.G.; Schuet, S.; Brenton, J.; Timucin, D.; Smith, D.; Kaneshige, J. Fast optimization for aircraft descent and approach
trajectory. In Proceedings of the Annual Conference of the Prognostics and Health Management Society, PHM, St. Petersburg, FL,
USA, 2–5 October 2017; pp. 80–90.
12. Lim, Y.; Gardi, A.; Sabatini, R. Energy Efficient 4D Trajectories for Terminal Descent Operations. In Proceedings of the
International Symposium on Sustainable Aviation (ISSA) 2018, Rome, Italy, 1 July 2018.
13. Jia, Y.; Cai, K. The Trade-off Between Trajectory Predictability and Potential Fuel Savings for Continuous Descent Operations. In
Proceedings of the 2018 IEEE/AIAA 37th Digital Avionics Systems Conference (DASC), London, UK, 23–27 September 2018;
pp. 1–6.
14. Sáez, R.; Prats, X. Comparison of Fuel Consumption of Continuous Descent Operations with Required Times of Arrival: Path
Stretching vs. Powered Descents. In Proceedings of the 9th International Conference on Research in Air Transportation (ICRAT
2020), Tampa, FL, USA, 23–26 June 2020.
15. Mclachlan, B.G.; Karamcheti, K. An Experimental Study of Airfoil-Spoiler Aerodynamics; National Aeronautics and Space Adminis-
tration, NASA Ames Research Center: Washington, DC, USA, 1985.
16. Stuckey, R. Identification of Nonlinear Aircraft Spoiler Parameters. IFAC Proc. Vol. 1993, 26, 809–814. [CrossRef]
17. Abdelrahman, M.; Al-Bahi, A.; Ghazi, M.; Olwi, I. Aircraft spoiler effects under wind shear. J. Aircr. 1994, 31, 154–160. [CrossRef]
18. Lindsay, S.; Walsh, P. Experimental Investigation of Spoiler Deployment on Wing Stall. Open J. Fluid Dyn. 2018, 8, 308–320.
[CrossRef]
19. Geisbauer, S. Numerical Simulation and Validation of the Aerodynamics of Static and Dynamic Spoilers. In Proceedings of the
Virtual AIAA Aviation Forum 2020, Online, 15–19 June 2020.
20. Omori, Y. Spoiler. J. Jpn. Soc. Aeronaut. Space Sci. 1963, 11, 108.
21. Barnes, C.S. A Developed Theory of Spoilers on Aerofoils; Ministry of Aviation Aeronautical Research Council: London, UK, 1966.
[CrossRef]
22. Kalligas, K. The Dynamic Characteristics of Two-Dimensional Spoilers at Low Speeds. Ph.D. Thesis, Department of Aeronautical
Engineering, University of Bristol, Bristol, UK, 1986.
23. Lee, C.S.; Bodapati, S. Experimental Investigations of the Flowfield of an Airfoil with Spoiler. AIAA J. 1987, 25, 1411–1416.
[CrossRef]
24. Consigny, H.; Gravelle, A.; Molinaro, R. Aerodynamic Characteristics of a Moving Two-Dimensional Spoiler in Subsonic and
Transonic Flow. J. Aircr. 1984, 21, 687–693. [CrossRef]
25. Costes, M.; Gravelle, A.; Philippe, J.J.; Vogel, S.; Triebstein, H. Investigation of Unsteady Subsonic Spoiler and Flap Aerodynamics.
J. Aircr. 1987, 24, 629–637. [CrossRef]
26. Nelson, C.F.; Koga, D.J.; Eaton, J.K. Unsteady, Separated Flow Behind an Oscillating, Two-Dimensional Spoiler. AIAA J. 1990,
28, 845–852. [CrossRef]
27. Yeung, W.W.H.; Xu, C.; Gu, W. Reduction of Transient Adverse Spoiler Effects. J. Aircr. 1997, 34, 479–484. [CrossRef]
28. Jordan, F.L., Jr.; Gato, W.; Masiello, M.F.; O’Rourke, M.J.; White, E.R. Experimental Investigation of Unsteady Aerodynamics
on a Flap Element Induced by Rapid Spoiler Deflection. In Proceedings of the 12th AIAA Applied Aerodynamics Conference,
Orlando, FL, USA, 23–26 June 1994.
29. Scott, R.C.; Hoadley, S.T.; Wieseman, C.D.; Durham, M.H. The Benchmark Active Controls Technology model aerodynamic data.
In Proceedings of the 35th AIAA Aerospace Sciences and Meeting Exhibit, Reno, NV, USA, 6–9 January 1997.
30. Hanke, C. The Simulation of a Jumbo Jet Transport Aircraft—Volume I: Mathematical Model; National Aeronautics and Space
Administration: Washington, DC, USA, 1971; pp. 1–65.
31. Scheiderer, J. Angewandte Flugleistung; Springer: Berlin/Heidelberg, Germany, 2008.
32. Rossow, C.C.; Wolf, K.; Horst, P. Handbuch der Luftfahrzeugtechnik; Carl Hanser Verlag: München, Germany, 2014.
33. Kundu, A.K. Aircraft Design. Aeronaut. J. 2011, 115, 66. [CrossRef]
34. Bombardier Inc. Bombardier CRJ200 CBT—Flight Controls—Ailerons and Spoilerons; Bombardier Inc.: Montreal, QC, Canada, 2000.
35. AIRBUS S.A.S. A310—Flight Crew Training Manual; REV 0398; AIRBUS S.A.S.: Leiden, The Netherlands, 2008.
36. Boeing Commercial Airplane Company. 777-200 Flight Crew Operations Manual; The Boeing Company: Renton, WA, USA, 1999.
37. Embraer S.A. Embraer 170 FCOM Part B—Volume 2; Embraer S.A.: Sao Paulo, Brazil, 2005.
38. Anderson, G.; Clough, S.; Kneizys, F.; Chetwynd, J.; Shettle, E. AFGL Atmospheric Constituent Profiles (0–120 km); Technical Report
Afgl-tr-86-0110; Air Force Geophys. Lab.: Hanscom Air Force Base, MA, USA, 1986.
39. Roskam, J.; Lan, C.T.E. Airplane Aerodynamics and Performance; DARcorporation: Lawrence, KS, USA, 1997.
40. Lord, D.; Czarnecki, K.R. Pressure Distributions and Aerodynamic Characteristics of several Spoiler-Type Controls on a Trapezoidal Wing
at Mach Numbers of 1.61 and 2.01; National Advisory Committee for Aeronautics: Washington, DC, USA, 1956.
41. Wentz, W.H. Wind Tunnel Tests of the GA-(W)-2 Airfoil with 20% Aileron, 25% Slotted Flap, 30% Fowler Flap and 10% Slot-Lip Spoiler;
National Technical Information Service: Wichita, KS, USA, 1976.
42. Mineck, R. Reynolds Number Effects on the Performance of Ailerons and Spoilers; American Institute of Aeronautics and Astronautics:
Reston, VA, USA, 2001.
Aerospace 2022, 9, 263 19 of 19

43. Boeing Commercial Airplane Company. CAD 3-View Drawing for Airport Planning Purposes; Boeing Commercial Airplane
Company: Chicago, IL, USA, 2021.
44. AIRBUS S.A.S. Airport Operataions—AutoCAD 3 View Aircraft Drawings; AIRBUS S.A.S.: Leiden, The Netherlands, 2012.
45. AIRBUS S.A.S. A319, A320, A321—Flight Crew Operating Manual 1; System Description; AIRBUS S.A.S.: Leiden, The Nether-
lands, 2008.
46. AIRBUS S.A.S. A330—Flight Crew Operating Manual 1; System Description; AIRBUS S.A.S.: Leiden, The Netherlands, 2003.
47. AIRBUS S.A.S. A380—Flight Crew Operating Manual 1; Airbus S.A.S.: Blagnac Cedex, France, 2011.
48. Bulfer, B. 737 Cockpit Companion (-300/-400/-500/-600/-700/-800/-900/BBJ/BBJ 2); Leading Edge Libraries: Kingwood, TX, USA, 2003.
49. Bowers, P.M. Boeing Aircraft Since 1916; Naval Institute Press: Annapolis, MD, USA, 1989.
50. The Boeing Company. B767 Student Book ATA 27—Flight Controls; TechOps Training: Chicago, IL, USA, 2013.
51. The Boeing Company. Boeing 777 Specs, What Makes This Giant Twin Work? The Boeing Company: Chicago, IL, USA, 2020.
52. Eitel, J.M. Flight Standardization Board Report, BOEING B-777-200/-200ER/-200LR/-200F, B-777-300/-300ER; U.S. Department of
Transportation Federal Aviation Administration: Washington, DC, USA, 1998.
53. Petitt, K. Flight to Success. Available online: https://karlenepetitt.blogspot.com/2017/10/b777-roll-control.html (accessed on 5
March 2022).
54. CRJ Regional Jet. Title: Flight Crew Operating Manual CSP C–013–067. Available online: https://www.smartcockpit.com/docs/
CRJ-00_and_00-Flight_Controls.pdf (accessed on 17 March 2022).
55. Bombardier Inc. Airport Planning Manual CRJ900; Bombardier Inc.: Montreal, QC, Canada, 2015.
56. Embraer S.A. Embraer 190 FCOM—Flight Controls; Embraer S.A.: Sao Paulo, Brazil, 2003.
57. Buisson, A. Aircraft Energy management in descent phase using FMS, Airbus fleet performance. In Proceedings of the 3rd
Eurocontrol CDO Workshop, Brussels, Belgium, 18 March 2013; pp. 1–20.
58. Snyder, C.; Snyder, C.A.; Sankar, C. Use of information Technologies in the process of building the Boeing 777. J. Inf. Technol.
Manag. 1998, 9, 31–42. [CrossRef]
59. Kamo, S.; Rosenow, J.; Fricke, H.; Soler, M. Fundamental Framework to Plan 4D Robust Descent Trajectories for Uncertainties in
Weather Prediction. Aerospace 2022, 9, 109. [CrossRef]