See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/310168229

Weight Estimation of a Conceptual Wing for a High Altitude, Solar Powered

Unmanned Aerial Vehicle

Conference Paper · October 2016

CITATIONS READS

8 3,382

2 authors, including:

Ahmad Alsahlani
University of Basrah
26 PUBLICATIONS 127 CITATIONS

SEE PROFILE

All content following this page was uploaded by Ahmad Alsahlani on 24 April 2019.

The user has requested enhancement of the downloaded file.

 Weight Estimation of a Conceptual Wing for a High Altitude, Solar
Powered Unmanned Aerial Vehicle
Ahmad Alsahlani Thurai Rahulan
University of Salford
School of Computing, Science & Engineering
Salford, Greater Manchester, M5 4WT, United Kingdom
a.a.alsahlani@edu.salford.ac.uk t.rahulan@salford.ac.uk

In this paper, an in-house developed pre-conceptual design methodology and a composite structure module have
been used to study the impact of aerodynamic performance, payload requirements and wing thickness to estimate
the weight. In the first part, a pre-conceptual design methodology is introduced to examine the weight estimation
for a conceptual high altitude, long endurance, solar powered aircraft, called ‘SUMER’, operating at 17 km altitude
and 31o latitude (southern Iraq). The impact of aerodynamic performance parameters such as lift, drag and span
efficiencies on the weight estimation are investigated. The study also includes estimation of the payload and the
power consumption. This pre-conceptual design methodology establishes all the relationships between the
components with analytical functions using their characteristics. This approach can directly provide a unique and
optimal design for a given set of mission requirements. In the second part, an in-house developed low order
composite structure module has been used to design and optimise the wing structure in addition to studying the
impact of the maximum thickness of the wing and the spar dimensions on the weight. Detailed buckling
characteristics were also taken into account during the study. The final conceptual design of the SUMER is
introduced as a case study in this investigation.

Introduction
High-altitude aircraft can be characterised by their light weight and high aspect ratio wing which suggest a flexible
airframe. One of the difficulties is the reduction of the lift force due to the low air density. Therefore, the research
objective was to increase the lift coefficient and the platform area, whilst reducing the structural weight to a
minimum with consideration of the flexibility issues. Powering the aircraft by solar energy with the aim of long-
term operation also adds extra challenges to the design and operation, such as power management and the system
requirements. Relatively few high altitude aircraft have been designed which creates demand for more research
activity in this field.

Solar powered high altitude, long endurance, unmanned aerial vehicles (HALE UAVs) use solar irradiance, which
in turn is a function of the geographical parameters and the capability of solar cell panels. In theory, for long
endurance missions, the aircraft can fly continuously if the energy collected during the daytime is adequate to
power the aeroplane day and night [1]. Therefore, energy and mass balance estimates provide a good starting point
for the design. Motors, solar cell panels, fuel cells or batteries, as well as avionic systems are characterised
according to the power requirements. For this purpose, two different approaches from published literature were
adopted to achieve the conceptual design. The first approach is discrete and iterative, and is based on pure
estimation for the first set of components (motors, solar cell panels, fuel cell or batteries, and avionic system).
From their weight, one can estimate the total weight and required power. The latter power estimation is then
compared with the previous estimation, and the process is performed iteratively until a converged solution is found
[2]. The second approach is an analytical and continuous method, which consists of establishing all the
relationships between all the components with analytical functions using the component characteristics. Therefore,
this approach must be supported by recently developed analytical equations updated with statistical data retrieved
from existing relevant aircraft characteristics. The analytical approach can directly provide a unique and optimal
design, but requires a robust mathematical model [2, 3]. Noth [4] presented a design methodology for a small
solar-powered aircraft which was used to design the first prototype of Sky-Sailor. In this paper, the Noth model
will be tailored for high altitude, long endurance UAVs.

Composite materials have been widely used in aeronautical structures. This is due to their high strength to weight
ratio, which is extremely important for these ultra-light aircraft. It is claimed that the weight of a composite wing
could be between 34% to 40% lighter than the equivalent metallic wing [5]. The structural weight can be estimated
conceptually as a function of the wing aspect ratio and the wing area or the span. Other factors such as the load
5th Aircraft Structural Design Conference 2016 (The Royal Aeronautical Society)

factor and the total weight can be implied in the estimation model. The typical wing structure of most HALE
UAVs (such as Helios, Qinetiq Zephyr, and Pathfinder) consists mainly of spars, ribs, and skin. The spar can be
modelled as a rectangular or circular beam section, which can be sized according to the critical load. Few research
studies have investigated the weight estimation for this type of aircraft. Therefore, the purpose of this research is
to study the impact of the selected aerodynamic performance, the payload requirements and the wing thickness on
the weight estimation.

1- The Pre-Conceptual Design Methodology

The starting point of designing a solar powered aircraft is the balance of required energy and the weight of the
components. An analytical and continuous approach is used to estimate the power and the weight of the aircraft.
According to this approach, analytical functions are used to establish the relationships between the weight aircraft
components to provide a unique and optimal design directly. The mass and power of the system components can
be estimated as a constant fraction of either the structural mass or the total mass, or of the power needs [6, 7].

1.2 Power Management

The lift force generated by the aircraft at flight level must be equal to its weight; similarly, the drag must be
equal to the thrust. The required power for level flight (𝑃𝑙𝑒𝑣 ) can be calculated by [1, 8]:

𝐶𝐷 2𝐴𝑅 𝑔3 𝑚1.5
𝑃𝑙𝑒𝑣 = √
𝐶𝐿1.5 𝜌 𝑏

𝐶𝐿1.5
𝐶𝐷 = 𝐶𝐷𝑜 +
𝑒 𝜋 𝐴𝑅
where 𝑆 is the planform wing area, 𝜌 the air density, AR the aspect ratio, e is the span efficiency factor, b is the
span, 𝐶𝐿 is the design reference lift coefficient of the aircraft, 𝑚 is the total mass, and 𝐶𝐷𝑜 is the zero lift drag
coefficient.

The power losses in the gearbox, propeller, motors, and the controller can be considered by taking into account
their efficiencies 𝜂𝑔𝑟𝑏 , 𝜂𝑝𝑙𝑟 , 𝜂𝑚𝑜𝑡 , 𝜂𝑐𝑡𝑟𝑙 respectively. The required power for a given payload (𝑃𝑝𝑙𝑑 ), avionics
systems (𝑃𝑎𝑣 ) and the voltage converter efficiency (𝜂𝑏𝑒𝑐 ) also need to be considered. Thus, the total electric power
consumption (𝑃𝑒𝑙𝑒𝑐 𝑡𝑜𝑡 ) can be given by:
1 1
𝑃𝑒𝑙𝑒𝑐 𝑡𝑜𝑡 = 𝑃𝑙𝑒𝑣 + (𝑃𝑎𝑣 + 𝑃𝑝𝑙𝑑 )
𝜂𝑐𝑡𝑟𝑙 𝜂𝑚𝑜𝑡 𝜂𝑔𝑟𝑏 𝜂𝑝𝑙𝑟 𝜂𝑏𝑒𝑐

The power source for these long endurance aircraft is the solar energy during daytime (𝑇𝑑𝑎𝑦 ) and must be enough
to operate the aircraft over a 24 hour period (𝑇𝑑𝑎𝑦 +𝑇𝑛𝑖𝑔ℎ𝑡 ). The solar energy can be saved using efficient storage
schemes such as a fuel cell or batteries. The performance of the power storage capability during the charging and
the discharging cycle may not be ideal and hence this has to be taken into account by employing the charging and
the discharging efficiencies 𝜂𝑐ℎ𝑟𝑔 and 𝜂𝑑𝑐ℎ𝑟𝑔 . The total energy required (𝐸𝑒𝑙𝑒𝑐 𝑡𝑜𝑡 ) over a 24 hour period for level
flight can be evaluated by:

𝑇𝑛𝑖𝑔ℎ𝑡
𝐸𝑒𝑙𝑒𝑐 𝑡𝑜𝑡 = 𝑃𝑒𝑙𝑒𝑐 𝑡𝑜𝑡 (𝑇𝑑𝑎𝑦 + )
𝜂𝑐ℎ𝑟𝑔 𝜂𝑑𝑐ℎ𝑟𝑔

The available solar energy inside the atmosphere depends on several factors, such as altitude, latitude, curvature
of the solar panel, number of daylight hours and the time of the year [9, 10]. A simple model can be used to
estimate the solar power consumed by a flat solar cell associated with the maximum irradiance and the daytime
period of the operating latitude [2]. This model assumes that the daytime solar irradiance can be simplified by a
sinusoidal function. Therefore, the available power will represent the total area under the daily irradiance curve
[4, 6, 10, 11]:

𝐼𝑚𝑎𝑥 𝑇𝑑𝑎𝑦
𝐸𝑑𝑎𝑦 𝑑𝑒𝑛𝑠𝑖𝑡𝑦 =
𝜋/2

where 𝐼𝑚𝑎𝑥 is the maximum solar irradiance.

Weight Estimation of a Conceptual Wing for a High Altitude, Solar Powered Unmanned Aerial Vehicle. Ahmad Alsahlani & Thurai Rahulan

The effects of the curvature of the panel, solar cell efficiency, power adapter efficiency and the influence of the
weather condition can be employed in the energy equation by their efficiencies 𝜂𝑐𝑏𝑟 , 𝜂𝑠𝑐 , 𝜂𝑚𝑝𝑝𝑡 , 𝑎𝑛𝑑 𝜂𝑤𝑡ℎ𝑟
respectively:

𝐼𝑚𝑎𝑥 𝑇𝑑𝑎𝑦
𝐸𝑒𝑙𝑒𝑐 𝑡𝑜𝑡 = 𝐴𝑠𝑐 𝜂𝑐𝑏𝑟 𝜂𝑠𝑐 𝜂𝑚𝑝𝑝𝑡 𝜂𝑤𝑡ℎ𝑟
𝜋/2

where 𝐴𝑠𝑐 is the total solar cell area.

1.3 Mass Estimations

Empirical equations based on data published for existing solar UAVs have been used to estimate the weight of the
aircraft systems. The mass and power of each aircraft component can be estimated as a constant fraction of either
the structural mass or of the total mass or of the total power. These fractions were evaluated statistically from
existing solar powered UAV data which are likely to improve with future technological developments. These
equations will now be presented briefly:

1. The structural weight (𝑚𝑎𝑓 ) is usually estimated as a function of the span length (b) and the wing aspect ratio
(AR) as:
𝑚𝑎𝑓 = 𝐾𝑎𝑓 𝑏 𝑥1 𝐴𝑅𝑥2

where 𝑥1 𝑎𝑛𝑑 𝑥2 are constant.

Rizzo [6] proposed a structure mass estimation model which was obtained from data published for NASA
prototypes. This model can be written in terms of the span length and the aspect ratio:

𝑚𝑎𝑓 = 1.548 𝑏1.312 𝐴𝑅−0.0046

2. The solar panel usually covers the upper wing surface; hence it should be limited by the planform wing area
during the design iterations. Basically, the efficiency of the solar cell per its mass density is desired to be as
high as possible. Its area can be designed to cover a percentage of the planform area. The mass of the solar
cells can be evaluated by:
𝑚𝑠𝑐 = 𝐴𝑠𝑐 ( 𝐾𝑠𝑐 + 𝐾𝑒𝑛𝑐 )

where 𝐾𝑠𝑐 , 𝐾𝑒𝑛𝑐 are the mass density of solar cells and the mass density of encapsulation.

3. The mass of the voltage adapter (maximum power point tracker 𝑚𝑚𝑝𝑝𝑡 ) can be assessed according to the
solar power consumed 𝑃𝑠𝑜𝑙 𝑚𝑎𝑥 by [12]:.
𝑚𝑚𝑝𝑝𝑡 = 𝐾𝑚𝑝𝑝𝑡 𝑃𝑠𝑜𝑙 𝑚𝑎𝑥 = 𝐾𝑚𝑝𝑝𝑡 𝐴𝑠𝑐 𝐼𝑚𝑎𝑥 𝜂𝑐𝑏𝑟 𝜂𝑠𝑐 𝜂𝑚𝑝𝑝𝑡

where ( 𝐾𝑚𝑝𝑝𝑡 ) is the mass density of the maximum power point tracker.

4. The mass of the fuel cells ( 𝑚𝑓𝑐 ) can be estimated from the equation:
𝑚𝑓𝑐 = 𝐾𝑓𝑐 𝑃𝑒𝑙𝑒𝑐 𝑡𝑜𝑡

where 𝐾𝑓𝑐 is the energy density of the storage system including the fuel cell [7].

5. The entire weight of the propulsion system 𝑚𝑝𝑟𝑜𝑝 which includes the electric motors, propellers, gearboxes,
and controllers, can be scaled linearly with the shaft power ( 𝑃𝑙𝑒𝑣 ) [13]:
𝑚𝑝𝑟𝑜𝑝 = 𝐾𝑝𝑟𝑜𝑝 𝑃𝑙𝑒𝑣

where ( 𝐾𝑝𝑟𝑜𝑝 ) is the weight density of the propulsion group.

6. The masses of the avionics systems and the landing gear ( 𝑚𝑎𝑣 and 𝑚𝑙𝑔 ) can be estimated as a constant ratio
of the total aircraft weight 𝑚 [12, 14]:
𝑚𝑎𝑣 = 𝐾𝑎𝑣 𝑚

𝑚𝑙𝑔 = 𝐾𝑔𝑙 𝑚

where ( 𝐾𝑎𝑣 , 𝐾𝑔𝑙 ) are the avionics systems and landing gear weight fractions.
5th Aircraft Structural Design Conference 2016 (The Royal Aeronautical Society)

7. The required power of the avionics systems can be defined as a constant fraction of the mass of the avionics
[12]:
𝑃𝑎𝑣 = 𝐾𝑃𝑎𝑣 𝑚𝑎𝑣

where 𝐾𝑃𝑎𝑣 is the power-to-mass factor of the avionics systems.

The total mass m of the aircraft is the sum of all the components weights:

𝑚 = 𝑚𝑝𝑎𝑦𝑙𝑜𝑎𝑑 + 𝑚𝑎𝑓 + 𝑚𝑠𝑐 + 𝑚𝑚𝑝𝑝𝑡 + 𝑚𝑎𝑣 + 𝑚𝑔𝑙 + 𝑚𝑓𝑐 + 𝑚𝑝𝑟𝑜𝑝

As in the Noth methodology, the entire mass can be represented by a schematic as shown in Fig. 1. The equation
of the total mass can be simplified to:
𝑎12 3 𝑎13
𝑚 ⏞ 2
𝑚=𝑚 ⏞
[(𝑎5 + 𝑎6 )𝑎10 𝑎11 𝑎9 + 𝑎7 𝑎10 𝑎11 + 𝑎10 ] + [(𝑎5 + 𝑎6 )𝑎0 𝑎1 𝑎9 + 𝑎7 𝑎0 𝑎1 + 𝑎8 𝑎0 𝑎1 ]
𝑎
𝑏
14

+ [𝑏 𝑥1 𝑎4 ⏞
+𝑎3 + (𝑎5 + 𝑎6 ) 𝑎2 𝑎9 + 𝑎7 𝑎2 ]

where 𝒂𝟏−−𝟏𝟏 represent the factors shown inside each box in Fig1. This leads to the equation:
3
𝑚2
𝑚(1 − 𝑎12 ) − 𝑎 − (𝑏 𝑥1 𝑎4 + 𝑎14 ) = 0
𝑏 13
This equation can be solved to find the maximum positive real root with different span lengths (b) and aspect ratios
(AR). The solution must be constrained so that the value of the solar cell area is less than the planform area. The
fraction values of the components are assumed constant during the design process which are shown in Table 1. In
this methodology, the design space will be illustrated in order to conclude the minimum aircraft weight, its
corresponding span length and the wing aspect ratio. The final choice can lead to the other component weight
estimates such as the weight and power of the propulsion, the solar cells, the fuel cells, the avionics systems, the
maximum power tracker and all other elements.

Figure 1. Schematic Representation of the Pre-conceptual Design Methodology

Weight Estimation of a Conceptual Wing for a High Altitude, Solar Powered Unmanned Aerial Vehicle. Ahmad Alsahlani & Thurai Rahulan

Table 1. The Constant Parameters of the Design

Parameter Value Units Description
Kfc 550*3600 J/kg Energy density of fuel cell
2
Ksc 0.25 kg/m Mass density of solar cells
Kenc 0.0* kg/m2 Mass density of encapsulation
Kmppt 0.00047 kg/W Mass to power ratio of max power point tracker
Kprop 0.0045 kg/W Mass to power ratio of propulsion unit
1.312
Kaf 1.548 kg/m Structural mass constant
ƞcbr 1* - Efficiency of the curved solar panels
ƞbec 1 *
- Efficiency of step-down converter
ƞsc 0.31 - Efficiency of solar cells (Triple-Junction ELO Tabbed Solar Cell (MicroLink))
ƞchrg 1 *
- Efficiency of fuel cell charge
ƞctrl 0.95 - Efficiency of motor controller
ƞdchrg 0.6 - Efficiency of fuel cell discharge
ƞgrb 1* - Efficiency of gearbox
ƞmot 0.95 - Efficiency of motor
ƞmppt 1* - Efficiency of max power point tracker
ƞplr 0.85 - Efficiency of propeller
𝒙𝟏 1.312 - Structural mass area exponent
𝒙𝟐 -0.0046 - Structural mass aspect ratio exponent
Kav 0.03 - Avionic weight fraction
Klg 0.00725 - Landing gear weight fraction
KPav 6 W/kg Power-to-mass ratio of avionics
*
Kwthr 1 - Irradiance margin factor

 Ignored in the present work

1.4 Case Study - SUMER HALE UAV

As a case study, a new conceptual design of a high-altitude, long endurance, and solar powered UAV has been
designed using this methodology. The aircraft mission is to operate in southern Iraq, particularly above the Iraqi
marshes close to the ancient city (Sumer), for surveillance purposes and carrying a payload of 100 kg between 1st
March and 1st October at an altitude of about 17 km. The aerodynamic performance is assumed constant during
the sizing process which is detailed with the other associated mission requirements in Table 2. Applying this
methodology to the given mission leads to the possible design space shown in Fig.2. From the gross weight-span
curves, one can conclude that the minimum weight is at the span length of about 69 m with a wing aspect ratio of
about 20. This choice can be used to determine the weight of the corresponding elements in Fig.3 as concluded
and detailed in Table 3.
Table 2 Mission Requirement of the Case Study (SUMER)
Parameter Value Units Description
CL 0.85 - Aeroplane lift coefficient
Cd 0.012 - Aerofoil drag coefficient
e 0.95 - Oswald’s efficiency factor of the wing
2
Imax 974 W/m Maximum irradiance (minimum (Imax) between 1st March and 1st October) at latitude 31o
(southern Iraq) [15]
Tday 12 h Daytime duration
mpayload 100 kg Payload mass
Ppayload 1250 W Payload power consumption
ρ 0.1382 kg/m 3
Air density (at 17 k m)
5th Aircraft Structural Design Conference 2016 (The Royal Aeronautical Society)

AR

Figure 2. Design Space for the Given Mission Requirements

Weight Estimation of a Conceptual Wing for a High Altitude, Solar Powered Unmanned Aerial Vehicle. Ahmad Alsahlani & Thurai Rahulan

Table 3. The Main Characteristics of SUMER

Parameter Value Units Description
AR 20 - Aspect ratio
b 69 m Wingspan
m 1111.01 kg Total mass
mavionic 33.34 kg Avionic weight
mfc 454.6 kg Fuel cell weight
mprop 50.65 kg Propulsion system weight
maf 394.8 kg Structure mass
msc 44.8 kg Mass of solar cells
mmppt 24.64 kg Mass of maximum power point tracker
S 238.1 m2 Platform Area
cav 3.45 m Mean chord
V∞ 27.92 m/s Airspeed

1.4.1 Aerodynamic Performance study

In the pre-conceptual design stage, the aerodynamic performance is usually assumed to be constant at the cruise
condition. An accurate estimate of this performance is required to reduce the effort and time needed to conduct the
calculations. A brief study is conducted to outline the impact of the reference aerodynamic performance parameters
such as lift & drag coefficients and the span efficiency on the estimation of the gross weight during the pre-
conceptual design stage. This study has been conducted as represented in Figs. 3 to 5 using the same mission
requirements shown in Table 3 except the wing performance parameters which were varied.

1- Figure 3 shows that the minimum gross weight and the corresponding span length are decreasing with
increasing reference lift coefficient. The required aspect ratio is increasing with increasing reference lift
coefficient to reduce the induced drag.
2- Figure 4 shows that the minimum gross weight and the corresponding span length are increasing with
increasing zero-lift drag coefficient. Increasing the drag means high energy is needed for the propulsion
system. Therefore a larger wing area is required to accommodate the solar cells. Hence the span efficiency
is fixed in this particular study, the enlarging of the wing area would mostly be in the chordwise direction,
leading to reduced aspect ratio.
3- Figure 5 shows that the minimum gross weight and the corresponding span length and the aspect ratio are
decreasing with increasing span efficiency.

1400 90 26
80 25
1200 24
70
Gross weight [kg]

1000 23
60 22
Span [m]

800 50 21
AR

20
600 40 19
30 18
400 17
20
200 16
10 15
0 0 14
0.75 0.8 0.85
0.9 0.95 1 0.75 0.8 0.85 0.9 0.95 1 0.75 0.8 0.85 0.9 0.95 1
CL CL
CL
Figure 3. Effect of The Reference Lift Coefficient on The Gross Weight and Corresponding Aspect Ratio and Span of the Wing.
CDo=0.012, Span efficiency=0.95, Payload=100kg, Payload Power=1250 W
5th Aircraft Structural Design Conference 2016 (The Royal Aeronautical Society)

1800 120 25
1600 24
100 23
1400
22
Gross weight [kg]

1200 80 21

Span [m]
1000 20

AR
60
800 19
40 18
600
17
400 20 16
200 15
0 0 14
0.01 0.012 0.014 0.016 0.018 0.01 0.012 0.014 0.016 0.018 0.01 0.012 0.014 0.016 0.018
CDo CDo CDo
Figure 4. Effect of The Zero Lift Drag Coefficient on Gross Weight and Corresponding Aspect Ratio and Span of The Wing.
CLreference=0.85, Span efficiency=0.95, Payload=100kg, Payload Power=1250 W
1400 90 30
80
1200 25
70
Gross weight [kg]

1000 60 20
Span [m]

800 50

AR
15
600 40
30 10
400
20 5
200 10
0 0
0
0.8 0.85 0.9 0.95 1 0.8 0.85 0.9 0.95 1
0.85 0.8
0.9 0.95 1
Span effeciency Span effeciency Span effeciency
Figure 5. Effect of The span efficiency on The Gross Weight and Corresponding Aspect Ratio and Span of The Wing. CLreference=0.85,
CDo=0.012, Payload=100 kg, Payload Power=1250 W
1.4.2 Payload Requirement Study
The influences of the payload weight and its required power on the weight estimation have been studied. These
two factors can have an intense effect on the gross weight estimation for this type of aircraft as shown in Fig. 6 &
7. Increasing the payload weight and its required power necessitates higher required lift and energy, which in turn
need a larger wing area leading to an increase in the span, thus increasing the weight.

1200 80 30
28
1000 70
26
Gross weight [kg]

800 60 24
Span [m]

22
50
AR

600 20
18
400 40
16
200 30 14
12
0 20 10
0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160
Payload [kg] Payload [kg] Payload [kg]
Figure 6. Effect of The Payload Weight on The Gross Weight and Corresponding Aspect Ratio and Span of The Wing. CLreference=0.85,
CDo=0.012, Span efficiency=0.95, Power of Payload=0
1200 72 30
28
1000 70
26
Gross weight [kg]

68 24
800
Span [m]

66 22
600
AR

20
64
400 18
62 16
200 14
60
0 12
58 10
0 0.3 0.6 0.9 1.2 1.5 0 0.5 1 1.5 0 0.5 1 1.5
Power of Payload [kW] Power of Payload [kW] Power of Payload [kW]
Figure 7. Effect of The Payload Power on The Gross Weight and Corresponding Aspect Ratio and Span of The Wing. CLreference=0.85,
CDo=0.012, Span efficiency=0.95, Payload=100 kg
Weight Estimation of a Conceptual Wing for a High Altitude, Solar Powered Unmanned Aerial Vehicle. Ahmad Alsahlani & Thurai Rahulan

2- Structure Analysis
High Altitude UAVs operate at low air density, so the need is raised to extend the planform wing area to generate
a required lift. Therefore, these aircraft are characterised by their high wing aspect ratio to reduce induce drag.
Composite materials are largely used in aircraft structures due to their higher stiffness to weight ratio. For instance,
the Helios, Pathfinder, Qinetiq Zephyr and X-HALE and most other existing HALEUAV are of lightweight carbon
fibre construction [16]. In a companion paper [17], a composite structure model was developed to design and
estimate the wing weight. In addition, they used a linear finite beam elements model to evaluate the wing
deflection. This model is adopted here and will be used to optimise the orientations of the plies in the wing box
(spar) in order to study the effects of the maximum wing thickness on the weight estimation and on the maximum
wing deflection.

2.1 Composite Structure Model

The typical wing structure of the high altitude aircraft can be divided into non-spar elements and spars as shown
in Fig. 8. The non-spar elements can be estimated using empirical equations that were published by NASA Langley
Research Centre and Lockheed corporation for high altitude solar powered platforms [18]. These elements can be
estimated according to the aircraft aspect ratio (𝐴𝑅), wing area (𝑆), and wetted area (𝑆𝑤𝑒𝑡 ) as in the following
equations given by [18, 19]:

Trailing edge mass: 𝑚𝑡𝑒 = 0.0998 (𝐴𝑅. 𝑆)0.5

Covering mass: 𝑚𝑐𝑜𝑣 = (0.2055 + 0.0028 (𝐴𝑅/ 𝑆)0.5 )𝑆𝑤𝑒𝑡

Ribs mass: 𝑚𝑟𝑖𝑝 = 1.033 𝑆 0.6

Control systems mass: 𝑚𝑐𝑜𝑛𝑡 = 0.3006 𝑆 /𝐴𝑅0.5

The wing-box (spar) is the main structure element and might constitute half the structural weight for this kind of
aircraft. In this model, the spar is assumed to be a rectangular thin walled beam with a height equal to 0.8 of the
maximum wing section profile (see Fig. 8). Its width can be assumed to be a proportion of the height and should
be selected carefully for reasons that will be explained in the next sections. The spar can be sized by finding the
required number of layers in each of its sides (flanges and webs) to withstand the critical load with a safety margin.
Once the required plies are found, the spar weight can be evaluated and this value, which is effectively inertia
relief, will be added to the overall loading condition. The calculation will be repeated until a weight convergence
is achieved. A linear finite beam element is used to evaluate the wing-box deflection under the total load while
only a quasi-static equilibrium is considered during the sizing process. The overall wing design methodology is
shown in Fig. 9.

Figure 8. Sketch of the Structural Wing Model and Wing-box Idealization

The spar can be discretised into several partitions; each partition is sized according to the maximum exerted load.
In each individual partition sizing, an optimisation loop (using a genetic algorithm) will be performed to find
suitable stacking patterns for each spar side (flanges and webs) to achieve the minimum number of layers
5th Aircraft Structural Design Conference 2016 (The Royal Aeronautical Society)

(minimum weight) required to satisfy the failure & buckling criteria. These steps are outlined in Fig.10 and detailed
in ref. [17]. The calculation of the failure criteria, maximum principal stresses, and mechanical properties of the
wing-box section are detailed in references [17, 20-22]. In this paper, the optimal stacking pattern of each flange
and web in each partition will be found so that it achieves the minimum spar weight, given the following
constraints:
1- Each web and flange have symmetric laminate about their mid-plane.
2- The webs are identical.
3- The flanges are identical.
4- Each laminate is balanced (containing an equal number of ±45 degrees).
5- Each laminate must have at least 10 % of plies of the orientations (-45, +45, 90, 0). This is applied to
every subsequent ply up to 10 [5, 23].
The Tornado Vortex Lattice code has been used to evaluate the aerodynamic load, such as lift, drag, and pitching
moment. The outer dimensions and external forces (lift, torsion, and other internal weight) are given by the wing
configuration and its critical aerodynamic performance. Once the spar weight is calculated, the total structural
mass can be evaluated as the summation of the weights of the non-spar elements and the spar itself:

𝑚𝑠𝑡 = 𝑚𝑠𝑝𝑎𝑟 + 𝑚𝑛𝑜𝑛_𝑠𝑝𝑎𝑟

Wing Geometry, critical flight state

Estimate Non-Spar elements weight

Critical load

Aerodynamic solver
(Tornado VLM) Set initial number and
stacking pattern of layers
Critical aerodynamic load
And internal weight Stacking patterns
Stack the plies in the webs
Discretize wing-box into stations and the flanges according to
the suggested patterns
Find bending moment, shear force and
torque at each station (see Fig.10) Evaluate the cross-section
properties

Size the cross-section of each spar station

Check each lamina against
the failure
Add the structure weight to the load Increase the
number of layer
Check each web and flange
at the failed
against the buckling
No element
Has the weight
converged?

Yes Yes Have buckling or

failure detected?
Evaluate the wing box deflection

No

Has the deflection Yes Optimisation Loop Section weight

converged? Minimise (weight)

No
Plies stacking, section weight,
Modify the wing geometry mechanical properties

Structure weight
Figure 9. Wing design methodology. Figure 10. Cross-section Design Methodology
Weight Estimation of a Conceptual Wing for a High Altitude, Solar Powered Unmanned Aerial Vehicle. Ahmad Alsahlani & Thurai Rahulan

2.2 Structural Weight of SUMER

The composite structure model has been used to design the structure of the SUMER, which is described in section
1.4. Its main characteristics are detailed in Table 3. A rectangular straight flying wing configuration is adopted to
carry the payload at root chord section whilst six fuel cells and eight electrical motors (as propulsion system) are
used and distributed along the wing span as shown in Fig.11. Its aerofoil section is similar to NASA Pathfinder
(LZA2575), which has a maximum thickness-to-chord ratio of 13.7%. The summation of the non-spar element
weights is 153.6 kg as calculated in Table 4. Hence the structural weight estimated from the pre-conceptual design
is 394.8 kg (see Table 3). The spar weight should be no more than 241.2 kg. The target now is to design a spar
which must satisfy the failure criterion. In addition, its weight should not exceed the expected value. The semi-
spar is discretised into ten partitions. The length of each partition equals 3.5 m except the last one at the wing tip,
which is 3 m. In order to study the influence of the maximum wing thickness on the weight prediction and the
maximum wing deflection, several heights to width ratios of the spar dimension are studied.

Payload Propulsion unite

Fuel cell

Figure 11. The Initial Configuration of SUMER

Table 4. Structure Weight Distribution

Trailing edge mass 6.89 kg
Covering mass 103.16 kg
Ribs mass 27.55 kg
Control systems mass 16.0 kg
Non-Spar elements weight (sum all above weights) 153.6 kg
Conceptual spar weight (structure – non spar weights) 394.8-153.6=241.2 kg

2.2.1 Spar Dimension Effect on the weight estimation

In this section, the spar height (which is 0.8 of the maximum wing thickness) and the spar width will be examined
against the weight and the maximum wing deflection at the critical load condition (load factor of 3 and safety of
factor 1.5). The wing-box is modelled as a rectangular thin walled beam whose centre coincides with the quarter
chord axis. The aerodynamic performance and the non-spar element weights are assumed constant during this
study. Each partition is designed and optimised individually to be joined by an ideal Aluminium joiner with other
neighbouring partitions such as that adopted in ref. [24]. The consideration of buckling is employed in this analysis.
However, the contribution by the ribs is ignored because it is difficult to employ at the preliminary design stage
when not much detail is available about the structure. The shear and bending buckling are considered in the webs
whilst only the bending buckling is considered in the flanges, assuming that each side of each spar partition can
buckle as a long plate [22]. Moreover, this study is conducted with and without the buckling consideration. Figure
3 shows that increasing the wing thickness leads to a heavier structure and a lower wing tip deflection. For this
type of loading, an increase in the width of the flanges leads to an increase in the structural weight, as well as a
decrease in the maximum deflection due to increasing the stiffness. The same study is repeated without considering
the buckling failure as revealed in Fig. 4, in which increasing the wing thickness shows a conflicting influence on
the weight estimation. Likewise, the width of the spar also shows the same influence on the weight estimation at
only the higher wing thickness. However, the maximum wing deflection is lightly influenced by changing the
width. The conclusions of these results can be outlined as follows:

1- The buckling consideration leads to a heavier structure.

2- The width of the spar can be manipulated to decrease the weight.
3- The maximum wing deflection can be reduced by increasing the spar height (wing thickness).
4- Increasing the buckling resistance without increasing the number of plies could result in reducing the
weight. For instance, stiffeners or internal lightweight foam can be used to increase the buckling
resistance [24].
5- If the buckling considerations are ignored, the influence of the wing thickness can also lead to an
increasing in weight, but the trend is not uniform throughout the range of the thicknesses considered.
5th Aircraft Structural Design Conference 2016 (The Royal Aeronautical Society)

700 20
w=0.2 h w=0.2 h

Maximum Deflection [m]

600 w=0.4 h 16 w=0.4 h
w=0.6 h w=0.6 h
Spar Weight [kg]

500
12
400
8
300
4
200

100 0
0.12 0.1
0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
Maximum Aerofoil Thickness Ratio Maximum Aerofoil Thickness Ratio
Figure 13 Effect of the aerofoil maximum thickness on the maximum deflection and the structure weight with different spar width to height
ratio, with buckling consideration

250 20
w=0.2 h w=0.2 h
w=0.4 h 16 w=0.4 h

Maximum Deflection [m]

220
w=0.6 h w=0.6 h
Spar Weight [kg]

w=0.8 h w=0.8 h
190 12

160 8

130 4

100 0
0.1 0.12 0.14 0.16 0.18 0.2 0.1 0.12 0.14 0.16 0.18 0.2
Maximum Aerofoil Thickness Ratio Maximum Aerofoil Thickness Ratio
Figure 14 Effect of the aerofoil maximum thickness on the maximum deflection and the structure weight with different spar width to height ratio,
without buckling consideration
2.2.2 Spar sizing for SUMER
The maximum thickness of the aerofoil selected for SUMER is 0.137 %. When considering buckling failure, it is
apparent from Fig. 13 that a spar width of about 0.25 of the spar height results in a weight of about 241 kg (see
Table 4). Therefore, the tool has been used to size and optimise the spar to withstand the critical load and not
exceed the expected weight. Table 5 details the optimal orientation and stacking of the plies for each flange and
web in each spar partition while Fig.15 shows the spar deflection and the critical load distribution. The
corresponding spar weight which is obtained is 237.32 kg.

Table 3. The optimal plies stacking in each spar station, where the spar width equal to 0.25 of the spar height. Buckling is considered.

Station Stacking at flanges Stacking at webs

1 [04/-45/45/02/90/ 04/45/ 45/ 02/-45/90/ 45/ 0]s [-45/45/02/90/05/-45/45/02/90/05/-45/45/02/-45/90/45/0]s
2 [0 /-45/45/ 04/90/0/-45/45/04 /-45/ 90/45/0]s [90/02/-45/45/02/90/02/-45/45/02/90/02/-45/45/02/-45/90/45/0]s
3 [02 /90/ 0/ -45/ 45/ 07 /-45/ 90/ 45/ 0]s [-45/45/0/90/04/-45/45/0/90/ 04 /-45/ 45/ 0/ -45/ 90/ 45/ 0]s
4 [90/ 02 /-45/ 45/ 06 /-45/ 90/ 45/ 0]s [-45/ 45/ 05/90/ 02 /-45/45/ 05 /-45/90/ 45/ 0]s
5 [-45/ 45/ 07 /-45/ 90/ 45/ 0]s [-45/45/ 02 /90/ 06 /-45/45/ 02 /-45/90/ 45/ 0]s
6 [90/ 04 /-45/ 45/ 0/-45/ 90/ 45/ 0]s [02 /-45/ 45/0/90/ 04 /-45/45/ 0 /-45/ 90/ 45/ 0]s
7 [-45/ 45/ 07 /-45/ 90/ 45/ 0]s [90/ 0/ -45/ 45/ 05 /45/ 90/ -45/ 0]s
8 [-45/ 45/04 /-45/ 90/ 45/ 0]s [04 /-45/ 45/03 /-45/ 90/ 45/ 0]s
9 [-45/ 45/ 02 /-45/ 90/ 45/ 0]s [-45/ 45/ 02/-45/ 90/ 45/ 0]s
10 [-45/ 90/ 45/ 0]s [02 /-45/ 90/ 45/ 0]
Weight Estimation of a Conceptual Wing for a High Altitude, Solar Powered Unmanned Aerial Vehicle. Ahmad Alsahlani & Thurai Rahulan

Spar Deflection at the Critical Load

Figure 15. The Total Critical Load Distributions and Spar Deflection

Conclusion
In this paper, two design models are introduced to design a pre-conceptual high altitude, long endurance, solar
powered UAV. The aerodynamic performance and the payload impacts are studied for a particular design case.
Furthermore, a composite structure model was used to investigate the impacts of spar dimensions and buckling
failure on the weight estimation. Even when considering the buckling issue, the weight obtained can satisfy the
pre-conceptually estimated weight. The maximum wing deflection was estimated but was not taken into account
in the optimisation loops. Considerably more work will need to be done to investigate an alternative stiffener
material, which can be used to increase the buckling resistance without increasing the number of plies to lighten
the structure.

Acknowledgement
Alsahlani would like to thank the Higher Committee for Education Development in Iraq (HCED-Iraq) for
supporting his research at University of Salford.
 5th Aircraft Structural Design Conference 2016 (The Royal Aeronautical Society)

References
1. Gao, X.-Z., et al., Energy management strategy for solar-powered high-altitude long-endurance aircraft.
Energy Conversion and Management, 2013. 70: p. 20-30.
2. NOTH, A., Design of Solar Powered Airplanes for Continuous Flight. 2008, PhD thesis, ETH ZÜRICH.
3. Mattos, B.S.d., N.R. Secco, and E.F. Salles, Optimal Design of a High-altitude Solar-Powered Unmanned
Airplane. Journal of Aerospace Technology and Management, 2013. 5(3): p. 349-361.
4. Noth, A., R. Siegwart, and W. Engel, Design of solar powered airplanes for continuous flight. 2008.
5. Kennedy, G.J. and J. Martins. A comparison of metallic and composite aircraft wings using aerostructural
design optimization. in 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference,
Indianapolis. 2012.
6. Rizzo, E. and A. Frediani, A model for solar powered aircraft preliminary design. THE
AERONAUTICAL JOURNAL, 2008(FEBRUARY 2008): p. 57-87.
7. Cestino, E., G. Frulla, and G. Romeo, Design of a High-Altitude Long-Endurance Solar-Powered
Unmanned Air Vehicle for Multi-Payload and Operations. Proceedings of the Institution of Mechanical
Engineers, Part G: Journal of Aerospace Engineering, 2007. 221(2): p. 199-216.
8. Manuel, H., Design, Construction and Test of the Propulsion System of a Solar UAV Thesis, in Aerospace
Engineering. 2013, Tecnico Lisboa.
9. Zhu, X., et al., How High Can Solar-Powered Airplanes Fly. JOURNAL OFAIRCRAFT, 2014. 51(No.
5, September–October 2014): p. 1653-1658.
10. P. Guarino, et al., Design of Solar Powered Ultra-light Aircrafts Realization of a Model and its Validation.
INTERNATIONAL JOURNAL OF ENERGY, 2014. 8.
11. Duffie, J.A. and W.A. Beckman, Solar engineering of thermal processes. Vol. 3. 1980: Wiley New York
etc.
12. Min Chang, et al., A GENERAL DESIGN METHODOLOGY FOR YEAR-ROUND SOLAR-
POWERED STRATOSPHERIC UAVS FROM LOW TO MIDDLE LATITUDES. 2014.
13. Guarino, G.C., et al., Multiobjective Optimization for Electric Drives Design in Solar-Powered Ultralight
Aircrafts. Recent Advances in Systems Science, ISBN, 2013: p. 978-960.
14. Montagnier, O. and L. Bovet. Optimisation of a solar-powered high altitude long endurance UAV. in
Proceedings of the 27th international congress of the aeronautical sciences, Nice, France. 2010.
15. M. T. Y. Tadros, M. A. M. Mustafa, and M. Abdel-Wahab, Estimation of the Global Horizontal Solar
Radiation in Iraq. International Journal of Emerging Technology and Advanced Engineering, 2014.
Volume 4(8).
16. Noll, T.E., et al., Investigation of the Helios prototype aircraft mishap volume I mishap report.
Downloaded on, 2004. 9: p. 2004.
17. Ahmad Alsahlani, Thurai Rahulan, and Nathera Abdulhassan, Composite Structural Analysis of a High
Altitude, Solar Powered Unmanned Aerial Vehicle. International Journal of Mechanical Engineering and
Robotics Research, IJMERR, 2016.
18. Colozza, A.J., Effect of Power System Technology and Mission Requirements on High Altitude Long
Endurance Aircraft. Sverdup Technology Inc., NASA Lewis Group, 1993.
19. Hall, D.W. and S.A. Hall, Structural sizing of a solar powered aircraft, N.C.R. Lockheed Missiles and
Space Company, April 1984., Editor. 1984.
20. Ali, E.O., et al. Structural Analysis of Composite Laminated Box-Beams under Various Types of
Loading. in International Journal of Engineering Research and Technology. 2014. ESRSA Publications.
21. Datoo, M.H., Mechanics of fibrous composites. 2012: Springer Science & Business Media.
22. Tarjan, G., A. Sapkas, and L. Kollar, Local Web Buckling of Composite (FRP) Beams. Journal of
Reinforced Plastics and Composites, 2009. 29: p. 1451-1462.
23. Gay, D., S.V. Hoa, and S.W. Tsai, Composite materials: design and applications. 2003: CRC press.
24. Cesnik, C.E., et al., X-HALE: a very flexible UAV for nonlinear aeroelastic tests. Paper No. AIAA-2010-
2715, Proceedings of (2010), 2010.

View publication stats