Literaturne Review

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/331384978
Further Experimentation and Theoretical Analysis of Thrust Characteristics

using Ducted Fan Theory for the UWE UAV Development Group Project
Thesis · April 2018

DOI: 10.13140/RG.2.2.21070.92483
CITATIONS READS
0 2,351
1 author:
Julian Stephens
University of the West of England, Bristol
2 PUBLICATIONS 0 CITATIONS
SEE PROFILE
All content following this page was uploaded by Julian Stephens on 27 February 2019.
The user has requested enhancement of the downloaded file.

Further Experimentation and Theoretical
Analysis of Thrust Characteristics using
Ducted Fan Theory for the UWE UAV
Development Group Project
Figure 1: Ducted Fan under CFD

Analysis (Avid LLC, 2016)
Julian Stephens
Student Number: 15009425
Award Programme: BEng Aerospace Engineering
Module: UFMFX8-30-3 Individual Project BEng
Project Supervisor: Jason Matthews
April 2018
Word Count: 11624 Page Count: 53 Pages
i
15009425 Further Experimentation and Theoretical Analysis of Thrust
Characteristics using Ducted Fan Theory for the UWE UAV Development Group Project
ABSTRACT
This report will cover the detailed analysis in the experimental methods to find out the best
thrust efficiencies for a 10 inch diameter ducted fan rig. It presents the continuation of the
dissertation by Jack Ingleby who completed his project on the theoretical and experimental
analysis of ducted fans to increase thrust and power efficiencies. (Ingleby, 2017)
The main objective of this project is to look into more theory for ducted fans, devise a better
design for the duct, investigate key parameters such as lip radius, drag and lift forces acted
upon by different design aspects and use stronger and more reliable manufacturing methods
and also experimental wind tunnel data to gain an accurate measurement of how a ducted
fan will perform under environmental conditions.
A detailed literature review was undertaken using appropriate resources to develop further
knowledge about ducted fans and their functionality as well as looking into wider research
topics such as noise emissions, tip vortices, flow properties etc. From conducting a literature
review and examining the research gathered, it was noted that there was a total potential
increase in thrust efficiency of up to 94% compared to un-ducted rotors, and that, although
this value was not achieved, a thrust efficiency increase was still accomplished.
Several 3D Printed 3 inch diameter ducted fans were created with a very basic thrust rig
built to measure the thrust using a pair of scales. The maximum thrust efficiency to be
obtained from these 3D Printed ducts was found to be 27.27%.
The rig was designed on CAD with attention to specific details such as diffusion angle or tip
clearance and constructed for use in the wind tunnel with a manual adjusting for the angle
of attack using a 9 way screw rotor with a bearing to rotate the duct with precision and
accuracy in the wind tunnel.
Thrust was measured from the wind tunnel by the mounting of the rig itself and the forces
calculated to analyse the thrust accordingly. The thrust obtained from the duct from wind
tunnel testing was found to be of a decrease at lower wind speeds and an increase at higher
wind speeds.
It can be concluded that the project was a success as all aims and objectives were met and
a good number of efficient ducted fans were designed, tested and evaluated. The data
gleaned from this project will hopefully help to provide a strong development for further
work to this project.
ii
TABLE OF CONTENTS
ABSTRACT ....................................................................................................................................................... II
TABLE OF CONTENTS...................................................................................................................................... III
TABLE OF FIGURES .......................................................................................................................................... V
TABLE OF TABLES ........................................................................................................................................... VI
1 INTRODUCTION .......................................................................................................................................7
1.1 AIMS AND OBJECTIVES .......................................................................................................................... 8

1.2 SCOPE .................................................................................................................................................... 8
2 LITERATURE REVIEW .............................................................................................................................10
2.1 NOISE EMISSIONS ................................................................................................................................ 12

2.2 LONGITUDINAL STABILITY ................................................................................................................... 13
2.3 BRIEF HISTORY AND OPTIMISATIONS TO SHROUDED PROPELLERS ..................................................... 14
2.4 PEREIRA’S RESEARCH AND DESIGN OPTIMISATION............................................................................. 15
3 METHODOLOGY ....................................................................................................................................17
3.1 EQUIPMENT REQUIRED ....................................................................................................................... 17

3.2 CAD DESIGNS ....................................................................................................................................... 17
3.2.1 THE DUCT ........................................................................................................................................ 18
3.2.2 THE AMENDED DUCTED FAN RIG ................................................................................................... 21
3.2.3 INITIAL DUCTED FAN THRUST RIG................................................................................................... 21
3.2.3.1 The Base ................................................................................................................................................. 22
3.2.3.2 Motor support beam and secondary support beam .............................................................................. 22
3.2.3.3 Supporting Brackets ............................................................................................................................... 23
3.2.3.4 Support L clamps .................................................................................................................................... 23
3.2.3.5 Supporting block .................................................................................................................................... 23
3.2.4 WIND TUNNEL TEST RIG ................................................................................................................. 23
3.2.4.1 The idea ................................................................................................................................................. 23
3.2.4.2 Mounting baseplate ............................................................................................................................... 24
3.2.4.3 Rotation block ........................................................................................................................................ 24
3.2.4.4 Securing duct block ................................................................................................................................ 24
3.2.4.5 Initial Design 1........................................................................................................................................ 25
3.2.4.6 Initial Design 2........................................................................................................................................ 26
3.2.4.7 Design Correction based on Literature Survey ....................................................................................... 27
3.2.4.8 Simplified Design for Manufacture ........................................................................................................ 28
3.2.4.9 Further Problems and Solutions ............................................................................................................ 30
3.2.4.10 Final design ............................................................................................................................................ 32
3.2.4.11 Building of the Rig .................................................................................................................................. 33
3.2.4.11.1 Supporting block .............................................................................................................................. 33
3.2.4.11.2 Rotating wooden block .................................................................................................................... 33
3.2.4.11.3 L Clamp and supporting Clamps ...................................................................................................... 34
3.3 FINAL WIND TUNNEL RIG..................................................................................................................... 36
4 RESULTS ................................................................................................................................................ 39
4.1 RETESTING OF JACK’S DUCT ................................................................................................................ 39

4.2 3D PRINTED DUCT TESTS ..................................................................................................................... 43
4.3 WIND TUNNEL TESTING ....................................................................................................................... 48
4.3.1 Wind Tunnel Test Rig Data.............................................................................................................. 48
iii
4.3.2 Influence of Wind Tunnel Mounting Rig ......................................................................................... 49
5 EVALUATION .........................................................................................................................................50
5.1 PREVIOUS DESIGN IMPROVEMENTS ................................................................................................... 50

5.2 CARBON FIBRE LAYUP.......................................................................................................................... 51
5.3 JACK’S RIG RETESTING CONSIDERATIONS ........................................................................................... 52
5.4 3D PRINTED DUCTS EVALUATION ........................................................................................................ 52
5.5 WIND TUNNEL CONSIDERATIONS ....................................................................................................... 53
6 ANALYSIS AND DISCUSSION ..................................................................................................................54
6.1 COMPARISON OF JACK’S DUCTED FAN RIG WITH THE LITERATURE .................................................... 54
7 CONCLUSION .........................................................................................................................................56
7.1 WIND TUNNEL REMARKS ........................................................................................................................... 56

7.2 PARAMETERS TO MAKE A NOTE OF ............................................................................................................... 56
7.3 READDRESSING AIMS AND OBJECTIVES .......................................................................................................... 57
7.4 FUTURE WORK/RECOMMENDATIONS .......................................................................................................... 58
8 REFERENCES ..........................................................................................................................................60
9 BIBLIOGRAPHY ......................................................................................................................................63
10 APPENDICES ..........................................................................................................................................64
10.1 CARBON FIBRE MOTOR BEAM ............................................................................................................. 64

10.2 TIP CLEARANCES OF THE DUCT ............................................................................................................ 67
10.3 GANTT CHART...................................................................................................................................... 69
10.4 AMENDED GANTT CHART .................................................................................................................... 70
10.5 RISK ASSESSMENT ............................................................................................................................... 71
10.6 RISK MATRIX: (TO GENERATE THE RISK LEVEL). ............................................................................................ 73
10.7 ETHICS CHECKLIST ............................................................................................................................... 75
10.8 DISSERTATION PROJECT RESOURCES FORM........................................................................................ 79
10.9 CONTACTS REGISTER ........................................................................................................................... 81
Disclaimer: I hereby declare that this is my own original work and that this has not been
submitted to any other institution for assessment purposes.
iv
TABLE OF FIGURES
Figure 1: Ducted Fan under CFD Analysis (Avid LLC, 2016) ................................................ i
Figure 2: The Open Propeller Cruise condition vs Static condition (F.M. and Wright Jr, G.E.,
2001) ......................................................................................................................... 11
Figure 3: Ducted fan under cruise condition (F.M. and Wright Jr, G.E., 2001) ................... 11
Figure 4: Duct Parameters, (Pereira, 2008) ................................................................... 16
Figure 5: Duct Parameters, (Pereira, 2008) ................................................................... 18
Figure 6: Initial Ducted fan Configuration based on Pereira's data results......................... 19
Figure 7: Isometric Projection of initial duct design ........................................................ 20
Figure 8: Top view of initial duct design ........................................................................ 20
Figure 9: Isometric view of initial ducted fan rig ............................................................ 21
Figure 10: Front and Rear views of initial ducted fan rig ................................................. 22
Figure 11: Side view of initial ducted fan rig showing secondary support beam ................ 22
Figure 12: Initial measurements for wind tunnel rig ....................................................... 24
Figure 13: Isometric views of initial wind tunnel rig ........................................................ 25
Figure 14: Bottom, Back, and side views of initial wind tunnel rig .................................... 25
Figure 15: Transparent view of initial wind tunnel rig showing mounting of support block . 26
Figure 16: 2nd Initial wind tunnel rig Design with reconfigured support block .................. 26
Figure 17: 3rd initial design for wind tunnel rig showing amended rotor placement........... 28
Figure 18: Ducted fan rig with two arms and ideal duct .................................................. 29
Figure 19: Amended duct design for 10° diffusion angle ................................................. 30
Figure 20: Amended 5° Duct for CNC Cutting ................................................................ 31
Figure 21: CNC Duct complete at a 5° diffusion angle .................................................... 32
Figure 22: Final Design of the wind tunnel ducted fan rig showing isometric, plan and side
views ......................................................................................................................... 32
Figure 23: Cutting of the aluminium supporting block with the band saw ......................... 33
Figure 24: Final Manufactured Ducted fan Rig ............................................................... 35
Figure 25: Wind Tunnel Mounted ducted fan rig ............................................................ 36
Figure 26: Wiring of Wind Tunnel rig ............................................................................ 37
Figure 27: Wind Tunnel Rig mounted and wired ............................................................ 37
Figure 28: Mounted wind tunnel rig powered up ............................................................ 38
Figure 29: Vortex flow observed from Wind Tunnel Mounted rig ..................................... 38
Figure 30: Graph of the comparison for the thrust vs RPM for un-ducted vs ducted fan..... 42
Figure 31: Graph for comparison of RPM vs PWM for un-ducted vs ducted fan ................. 43
Figure 32: Green divergent duct with thrust rig setup and 3 inch propeller ....................... 44
Figure 33: Graph of comparison for PWM vs thrust of un-ducted fans vs 2 different ducted
fans ........................................................................................................................... 47
Figure 34: Extra images from the wind tunnel testing of the ducted fan rig ...................... 64
Figure 35: Process of preparing and laying up of the Carbon Fibre Motor Beam ................ 65
Figure 36: Final product achieved from vacuum forming carbon fibre motor support beam 66
Figure 37: Final cutting of Carbon Fibre Motor Beam ...................................................... 67
Figure 38: Tip clearance of the wind tunnel ducted fan rig.............................................. 67
Figure 39: Tip clearance of the wind tunnel ducted fan rig from a different angle ............. 68
v
TABLE OF TABLES
Table 1: Table of results from the retesting of Jack’s ducted fan test rig .......................... 39
Table 2: Table of un-ducted propeller results using Jack's Rig ......................................... 40
Table 3: Comparison between ducted and un-ducted rotor on Jack's Rig ......................... 40
Table 4: Jack's Open rotor test results .......................................................................... 41
Table 5: Jack's ducted rotor results .............................................................................. 41
Table 6: Motor Specification for the MT 2216 810Kv from http://www.unmannedtech.com 42
Table 7: Un-ducted rotor test on 3° prop thrust rig ........................................................ 44
Table 8: Green divergent duct results test from 0%-100% power input ........................... 45
Table 9: Results for the grey divergent duct tested from 0% to 100% power input........... 45
Table 10: Comparison between un-ducted and ducted rotor for both Green and Grey ducts
................................................................................................................................. 46
Table 11: Table of Wind tunnel results of ducted fan at 0° ............................................. 48
Table 12: Table of Wind Tunnel results of ducted fan at 0° in grammes of thrust ............. 48
Table 13: Results from amount of drag influence by wind tunnel mount .......................... 49
Table 14: Results from amount of drag influence from wind tunnel mount converted to
grammes .................................................................................................................... 49
vi
1 INTRODUCTION
Unmanned aerial vehicles (UAVs) are a constantly growing market and the demand for a
device with the capabilities of producing more thrust and keeping power emissions low is
increasing, with companies constantly trying to find a way to increase the thrust of their
Micro-Autonomous-Vehicles (MAVs) with a reduction in power. One possible approach to
solve this problem is the use of ducted fans. A ducted fan is a device which puts a cylindrical
duct or shrouding around a propeller to prevent thrust losses by reducing the effect of the
tip vortices occurring on the propeller. (Nagpurwala, 2016)
It is projected that ducted fans have the capability of increasing the thrust efficiency by up
to 94% and power efficiency by up to 62% in comparison to regular un-ducted propellers.
(Pereira, 2008) This has led to an increase in the usage of ducted fans and has resulted in
lower costs due to higher thrust efficiency and lower power consumption.
This project aims to continue the theoretical and experimental analysis of the dissertation
project by Jack Ingleby by improving his ducted fan rig, to test different composite materials
on the rig to achieve a higher tensile strength and also to manufacture a rig that is less
likely to fracture or show any deformation from significant bending stresses acting upon it.
(Ingleby, 2017) This project will also look into the usage of the UWE Subsonic wind tunnel
to gain a set of thrust results, including lift, drag and moment forces, thereby allowing
accurate lifelike data to be obtained and analysed, comparing the data against the effects of
a propeller with no duct.
With future use in higher applications such as helicopters or larger drones, ducted fans will
be very useful to implement as they achieve thrust increases at a much larger scale
compared to unshrouded propellers.
The ducted fan project is based on a newly found idea and has not been fully tested by
many aerospace companies throughout the world. To achieve the best possible outcome,
this project will determine the best choice of ducted fan design for obtaining maximum
thrust efficiency and maximum power efficiency.
7
1.1 AIMS AND OBJECTIVES

This project is the continuation of the work of Jack Ingleby who completed his dissertation
on the theory and use of ducted fans for quadcopters in order to increase the thrust and
power efficiency. (Ingleby, 2017)
The objective of this project is to further analyse the use of ducted fans and to conduct
further experimentation on the existing ducted fan rig using and developing on current
theory and knowledge. The specific aims of this project are:
• Using MatLab to programme functions, generating useful graphs and data to

establish thrust and efficiency characteristics
• Identifying and repairing of current ducted fan rig for suitability of retesting
• Working with composites to improve current designs
• Manufacturing of rigs and samples
• Utilising the UWE Subsonic Wind Tunnel to test the rig and obtain speeds and forces
on the ducted fan design
• Designing some small test ducts using the 3D Printer
• Using the Computer Numerical Control (CNC) Machine to automatically cut out the
test ducts accurately and precisely
• Not utilising Computational Fluid Dynamics (CFD) due to time constraints and lack of
current knowledge
• Not using Extensive coding due to lack of knowledge and because it is not required
1.2 SCOPE
The reason for choosing Ducted Fan investigation as a project is to gain a full understanding
of the characteristics and functionality of ducted fans and to determine whether their use on
unmanned aerial vehicles is justified in view of their thrust efficiency.
After completing research on current literature and planning the stages of the project, CAD
Designs will be processed and created from October to mid-January.
These preliminary tasks will include:
• Testing of the previous ducted fan model produced by Jack Ingleby

• Manufacturing a new rig to obtain new results
• Wind tunnel testing of the manufactured rig to obtain lifelike results
• Constructing and testing of smaller 3D printed ducted fan models
8
The completion dates of these individual tasks will vary, but the manufacturing and testing
process should be completed by the end of February to ensure that enough time is left for
the final write-up.
This project will involve further analysis of creating and improving on current designs,
including the design of a larger range of ducted fans using Computer Aided Design software
such as SolidWorks.
To make the experiment more effective, a sample test rig will be manufactured using a
range of composite materials, such as carbon fibre or glass fibre, and also epoxy resin for
adhesive purposes. The rig will be an improved version of the previous ducted fan rig, being
stronger and having a higher tensile strength. Therefore it will be less likely to fracture or
bend under testing conditions.
The UWE subsonic wind tunnel will be used to gain accurate and lifelike results regarding
the lift, drag and moment forces at various angles of attack acting on the rig. In this way a
complete set of data can be gathered and a different set of experimental data for the ducted
fan rig can be obtained.
9
2 LITERATURE REVIEW
Ducted fans, or shrouded propellers, are potential devices for high static thrust propulsion
systems.
The first characteristic of ducted fans is that they are expected to produce a greater thrust
efficiency in comparison to propellers of exactly the same diameter and loading. (NASA,
2002) According to Pereira, for Unmanned Aerial Vehicles (UAVs) or Micro Air Vehicles
(MAVs), the shrouded rotors have shown an increase in thrust of up to 94% and reductions
in power by up to 62% with the same thrust, compared to open rotor configurations.
(Pereira, 2008)
A second characteristic of ducted fans is their ability to provide a supplementary safety

feature due to the addition of the duct’s enclosure which not only is a better option for
people interested in the use of unmanned aerial vehicles, but is also useful to protect the
propellers themselves from scathing or fatigue. (NASA, 2002) Ducted fans also provide a
safer approach to specific areas of interest due to the shrouding of the propeller, thereby
leading to reduced risk from injury from the blades or propeller. If the shrouding has an
optimal design this also leads to a large reduction in noise. (Gelhausen, 2015)
The cross section of a duct is like an aerofoil where the resultant lift vectors of the duct are
strongly canted towards the centre of the duct. These vertical lift components provide
additional duct lift. The air divides at a stagnation point after it enters the duct and at this
point the duct gets closer to the flow. Due to higher angles of attack, the lift vector is
greatly increased in the upstream component, as opposed to the downstream lift vector
which is reduced due to lower angles of attack. NASA also states that, although the
combined vectors result in a net increase of the lift vector, there is a resultant pitching
moment acting on the duct’s leading edge, referred to as duct translational drag. (NASA,
2002)
This translational drag can result in a loss of energy, as the flow after a rotor also tends to
be rotational. Straightening this flow by using vanes may recover some of this energy,
however the effectiveness of this procedure will most likely depend on the vane design and
will introduce elements of drag into the flow downstream of the rotor. (Stickle and Crigler,
1940)
Flow has three regions including the inflow into the system, slipstream behind the propeller
and the free stream which flows outside of the configuration. The second part of Figure 2
10
shows the configuration with a completely different flow regime. This demonstrates that
there is no longer a free stream flow regime and that it is now part of the inflow. Air
particles have to travel through the propeller disk and through the slipstream, however due
to air particles having mass and the propellers being loaded at the tips, the flow is reversed
around the propeller tip, which negates the propeller’s disk area and the amount of air that
can flow through the propeller is reduced. (F.M. and Wright Jr, G.E., 2001)
Figure 2: The Open Propeller Cruise condition vs Static condition (F.M. and Wright Jr, G.E., 2001)
With the addition of a shroud around the propeller and curved leading edge, a boundary is
created between the propeller of the two flow regimes. Due to the addition of this curved
leading edge, the air particles can flow around the duct lip with less difficulty resulting in a
greater airflow intake in the slipstream.
The effect of this additional device is a higher operating efficiency, as the effective propeller
diameter is larger than its physical diameter due to the slipstream diameter being identical
to the diameter of the duct exit. This resulting larger diameter means that the airflow has a
Figure 3: Ducted fan under cruise condition (F.M. and

Wright Jr, G.E., 2001)
11
smaller change in velocity and therefore a greater mass flow, leading to a larger amount of
thrust produced by the duct. (See Figure 3)
With careful attention to the correct design configuration, the secondary force acting on the
duct can also produce thrust. This has the effect that the ducted fan produces more thrust
than it would if the thrust was generated without a duct. (F.M. and Wright Jr, G.E., 2001)
2.1 NOISE EMISSIONS
Ducted fans also produce a lower noise emission than normal un-ducted fans due to the
shrouding around them. For aircraft that are operated under a certain hovering and low
speed flying condition, a duct is a good option as it produces a much higher thrust to weight
ratio than a standard propeller on its own with no shrouding. (Jian-bo, L, 2005)
Ducted Fans offer several advantages, which is why they are implemented by aerospace
companies including the advantage of noise reduction which may not seem like much of a
physical difference, but which is a key component in aerospace engineering as the quieter
the rotor, the less energy is transferred to noise emission, and also the lower the cost due
to reducing noise levels
According to Jang, Furukawa & Inoue, the control of tip vortices on the fan rotor using
ducted fans was able to provide a reduction in noise emissions from the fan rotor. To
understand this, further investigation has to be made into the vortical flow acting on the tips
and mid chord section of the propeller. Jang used laser Doppler velocimetry to investigate
the three dimensional vortical flow structures. He deduced that the tip vortex formed at the
mid-chord of the rotor is developed in the tangential direction and that the tip vortex
induces an inward radial flow near the leading edge of the rotor, and thus a reverse flow
between rotor tip and duct. This effect causes what is called a blockage on the main flow in
the rotor. It is due to this blockage that noise emissions occur as the air is trapped and
cannot escape efficiently. This section contains high velocity fluctuations which produce the
noise effect. Adding a sufficient duct around the propeller will diminish the reverse flow
between rotor tip and duct and thus reduce this blockage effect caused by the vortices
occurring at the tip of the propeller. The end result is a reduction in the noise produced by
the rotor. (Jang, Furukawa & Inoue, 2001)
Many other sources agree with Jang’s statements about acoustic effects generated by the
ducted fans and the noise reduction associated with them.
12
A greater amount of aerodynamic lift is achieved by the use of ducted fans. A by-product of
this is the result of low noise due to the physical shrouding that the duct properties utilises
when compared to un-ducted or simple propeller rotors. (Li & Gao, 2005)
2.2 LONGITUDINAL STABILITY
The axial forces of a rotor propeller can be reduced by the addition of a duct and are
determined by a combination of downwash and interference effects. (Parlett, 1960) These
effects result in the forward assembly inducing a downwash, and incoming air is as a result
more aligned with the duct axis than in the case of a duct setup in the forward
configuration. According to Parlett, if the forces on each duct assembly are considered as
being resolved into an axial force acting along the axis of rotation, a normal force
perpendicular and a pitching moment about intersection of these two force vectors, then
rear duct assembly will experience less axial force.
Li & Gao have also found by gathering their test data and analysing the pitching moment
that ducted fans are only suitable for aircrafts operated and flying at low speed conditions.
They have also ascertained that the centre of gravity of the aircraft would result in being
greater than the resultant drag if the ducted fan were to be used as a lifting device in
forward flight. (Li & Gao, 2005)
NASA performed hover and forward flight performance characteristics for a ducted fan with
a 4-foot diameter. In order to alleviate the longitudinal and control problems associated with
the ducted fan, the exit vanes and horizontal stabilisers of the duct were studied including
relevant performance effects of change in blade angle. (Mort & Yaggy, 1962), (Mort, 1965),
(Abrego & Bulaga, 2002)
According to Mendenhall & Spangler, if the pitching moment of a ducted fan rotates the
nose upwards, it can be taken as a positive moment. Following this, the centre of moments
acting on a ducted fan can be taken at the mid-chord section and the total pitching moment
is a sum of normal force which is distributed along the duct’s chord. This is therefore an
addition to the total moment and to the thrust type forces which add a moment on the duct
when considered over the entire duct, although these forces cancel each other out.
(Mendenhall, M. & Spangler, S, 1970)
The pitching moment coefficient can be defined as:
𝑀𝑀𝐷𝐷
𝐶𝐶𝑀𝑀𝐷𝐷 = 𝑞𝑞𝑞𝑞𝑞𝑞 (1)
13
Where q is the dynamic pressure and equal to:
1
𝑞𝑞 = 𝜌𝜌𝑈𝑈 2 (2)
2
Where AR is the aspect ratio of the wings as shown in Equation 3:
𝑏𝑏2
𝐴𝐴𝐴𝐴 = 𝑆𝑆 (3)
𝑤𝑤
Where b is the wing span and Sw is the wing area.
2.3 BRIEF HISTORY AND OPTIMISATIONS TO SHROUDED PROPELLERS
Ducted fan experiments have been conducted since 1931 and documentation regarding
these experiments was published in Italy in 1931 by Luigi Stipa who conducted tests using
the wind tunnel on "intubed propellers" which were found to improve the propeller rotor for
static operation and low speed flight. (Stipa, 1931) As the results were encouraging the
Italian government built an aeroplane the fuselage of which contained the ducted propeller.
As a result of this, the aeroplane’s manoeuvrability was increased and its landing speed
reduced. (Stipa, 1934)
In 1944, further research by a German Engineer by the name of Krüger was conducted
which continued the progress of Stipa’s experiments and concluded that energy losses due
to the propeller could be successfully reduced by the use of shrouded profiles of small chord
and thickness. This was shown by the experimental testing. The maximum efficiency
achieved with the shrouded propellers was 0.71. Krüger stated that since the propeller is
heavily loaded a stator is absolutely necessary in order to avoid the high rotational losses
and the large pressure drag of the nacelle due to suction of the afterbody. He also
suggested that the use of a nose split ring could increase the static thrust on a shrouded
profile of small chord and thickness. (Krüger, W, 1949)
In theoretical analysis, many assumptions are made which are listed below:
• Flow is assumed to be incompressible.
• Flow is assumed to be inviscid, so viscosity effects of flow are ignored/negligible.

(Mendenhall & Spangler, 1970)
• Flow is assumed to be uniform with no losses in drag or frictional drag.
• Pressure far upstream and far downstream of the disk are assumed to be equal to
atmospheric/ambient pressure, patm
14
• The propeller is assumed to be a disc that is infinitesimally thin with an infinite

number of blades.
• The propeller offers no air resistance to air passing through it
The energy received by the disk is assumed to be in the form of a differential pressure jump
across the actuator disk and is uniformly distributed across the disk’s surface. (Bouferrouk,
2018)
2.4 PEREIRA’S RESEARCH AND DESIGN OPTIMISATION
The research conducted by Pereira was focused on optimising the ducted fan design for the
use of MAVs or UAVs. This was necessary because the research conducted by Stipa (1932)
and Krüger (1949) was not applicable to UAV applications as the difference in Reynolds
number was too great, and therefore a separate testing process was required.
Pereira performed hover tests on 17 models with a nominal rotor diameter of 16cm (6.3 in)
which were tested on key parameters such as diffuser expansion angle, diffuser length, inlet
lip radius and blade tip clearance, at different angles.
Pereira’s research and testing was able to show that there is a potential to nearly double the
thrust using ducted fans, compared to open rotor configurations, by increasing thrust by up
to 94%. The fact that Pereira was using a representative rotor to reduce uncertainty
compared to open rotor suggests, however, that the increase in thrust should be accurate
when compared to the UWE UAV when using the ideal duct as shown below.
The ideal rotor conditions and geometry for ducted fans were found to be the following:
Blade tip Clearance: 𝛿𝛿𝑡𝑡𝑡𝑡𝑡𝑡 = 0.1%𝐷𝐷𝑡𝑡
Inlet Lip Radius: 𝑟𝑟𝑙𝑙𝑙𝑙𝑙𝑙 = 13%𝐷𝐷𝑡𝑡
Diffuser angle: 𝜃𝜃𝑑𝑑 = 10°
Diffuser Length: 𝐿𝐿𝑑𝑑 = 50%𝐷𝐷𝑡𝑡 𝑡𝑡𝑡𝑡 72%𝐷𝐷𝑡𝑡
15
Using these dimensions for the 10 inch ducted fan as required by the UWE UAV
development testing team, the parameters are as follows:
Figure 4: Duct Parameters, (Pereira, 2008)
𝐷𝐷𝑡𝑡 = 254𝑚𝑚𝑚𝑚 𝑟𝑟𝑙𝑙𝑙𝑙𝑙𝑙 33.02𝑚𝑚𝑚𝑚 𝛿𝛿𝑡𝑡𝑡𝑡𝑡𝑡 = 0.254𝑚𝑚𝑚𝑚 𝐿𝐿𝑑𝑑 = 127𝑚𝑚𝑚𝑚 𝑡𝑡𝑡𝑡 182.88𝑚𝑚𝑚𝑚
16
3 METHODOLOGY
3.1 EQUIPMENT REQUIRED
To begin the first task, the right equipment had to be located and bought including
component cables such as electronic speed controllers (ESCs), a motor which could support
the 10 inch propellers that were going to be used on the ducted fan, the various bullet
connectors required to connect the motor to the ESCs and the Arduino responsible for
controlling the ESC and to turn the motor on at certain intervals throughout the testing
period.
EQUIPMENT:
• 10 inch propellers with 4.5 pitch for use with the motor
• Unmanned Tech Motor MT 2216 810Kv
• Electronic speed controller (ESC)
• 3 cell 22mAh Lipo fully charged Battery pack
• Arduino
• Necessary wires to connect components together
• XT60 Yellow Connectors
• Jack Ingleby’s Ducted Fan Rig
The first task was to test out the previously made ducted fan test rig to see if the same
results as Jack’s could be ascertained. The rig was set up in a classroom on top of a stable
table and a few blocks of wood were used to hold the ducted fan in place and to dampen
out any resonance occurring to the wooden rig frame.
The rig was set up the same way that Jack’s was set up and then the same amount of
power and amps was put through the motor to turn the propeller to equal Jack’s testing
conditions or at least be a close match to them.
Figure 3 shows how the structure of the experiment was expected to be set up.
3.2 CAD DESIGNS
To improve on the quality and efficiency of the current ducted fan rig, several design tests
were carried out and an approach was taken which would reduce the weight of the rig and
simplify it in terms of design so that it was easier to manufacture while still being strong and
reliable.
17
3.2.1 THE DUCT

The duct was created using Pereira’s system based on the many ducted fan trials produced
by him which analysed the blade tip clearance, inlet lip radius, diffuser angle, diffuser length
and the diffuser expansion ratio. All these parameters were used in creating the ideal ducted
fan which, according to Pereira, “increases in thrust over the open rotor by up to 94% were
observed, or, conversely, up to 49% reductions in power at the same thrust…”. (Pereira,
2008) The initial dimensioning of the parameters was based on the correlating data
developed by Pereira which are listed below:
Blade tip Clearance: 𝛿𝛿𝑡𝑡𝑡𝑡𝑡𝑡 = 0.1%𝐷𝐷𝑡𝑡
Inlet Lip Radius: 𝑟𝑟𝑙𝑙𝑙𝑙𝑙𝑙 = 13%𝐷𝐷𝑡𝑡
Diffuser angle: 𝜃𝜃𝑑𝑑 = 10°
Diffuser Length: 𝐿𝐿𝑑𝑑 = 50%𝐷𝐷𝑡𝑡 𝑡𝑡𝑡𝑡 72%𝐷𝐷𝑡𝑡
Using these dimensions for the 10 inch ducted fan as required by the UWE UAV
development testing team, the parameters are as follows:
Figure 5: Duct Parameters, (Pereira, 2008)
𝐷𝐷𝑡𝑡 = 254𝑚𝑚𝑚𝑚 𝑟𝑟𝑙𝑙𝑙𝑙𝑙𝑙 33.02𝑚𝑚𝑚𝑚 𝛿𝛿𝑡𝑡𝑡𝑡𝑡𝑡 = 0.254𝑚𝑚𝑚𝑚 𝐿𝐿𝑑𝑑 = 127𝑚𝑚𝑚𝑚 𝑡𝑡𝑡𝑡 182.88𝑚𝑚𝑚𝑚
18
The presented ducted fan design based on the exact parameters listed above was drawn up
in CAD and is shown below in Figure 6:
Figure 6: Initial Ducted fan Configuration based on Pereira's data results
As can be seen from this duct, the maximum height was chosen as 182.88mm as this would
produce the best duct for the 10 inch propeller rotor scaling that will be tested upon this
duct.
The duct is simple in design with the profile drawn from a straightforward 2D sketch as
shown in Figure 6 with certain dimensions labelling the distance between half of the duct
rotor diameter, which is 254/2 = 127mm. Subsequently a construction line was drawn at
this point vertically to revolve the sketch and produce a 3D Duct as shown in Figure 7 &
Figure 8.
19
Figure 7: Isometric Projection of initial duct design
Figure 8: Top view of initial duct design

This ducted fan design is simple to create in CAD. As it could, however, be difficult to
manufacture the lip and the duct together simultaneously, in real manufacture these parts
would most likely have to be made separately and put together afterwards, paying attention
to meet the fitting requirements to ensure that both components are flush when melded
together.
20
3.2.2 THE AMENDED DUCTED FAN RIG

As the previous ducted fan rig had several slight discrepancies in the actual life design such
as lack of stability, a new rig was created on SolidWorks 2017 using a series of supporting
stands and large bases with cut outs to fit the mounts so that the rig would hold the large
duct and rotor assembly firmly in place.
3.2.3 INITIAL DUCTED FAN THRUST RIG

The initial rig was constructed using simple designs with a wide enough base for support, a
strong supporting stand made of carbon fibre supporting the weight of the shroud and rotor
mounting system and also several support rods to hold the rig in position. This rig is shown
in Figure 9 & Figure 10.
Figure 9: Isometric view of initial ducted fan rig
21
Figure 10: Front and Rear views of initial ducted fan rig
3.2.3.1 The Base
The base of the thrust rig was designed to be constructed of medium density fibreboard
which, although it is cheap, would be able to withstand the stress and strain that the duct
and rotor would exert against the base.
3.2.3.2 Motor support beam and secondary support beam
This part of the structure was designed to hold the key component which is the motor and
rotor rig responsible for producing the thrust. The best choice for this was to use 5mm thick
Figure 11: Side view of initial ducted fan rig showing secondary support beam
22
carbon fibre as this would be able to withstand the strong forces that the duct would exert
on the beam. A secondary support beam was added on and is fixed on the inside of the
duct, thereby reducing the amount of force that the first beam had to withstand. This is
shown in Figure 11.
3.2.3.3 Supporting Brackets
Two supporting brackets were used to hold the rig firmly in place so that the rig was able to
withstand forces from both sides and would not topple due to the beams being too thin to
withstand the side forces acting upon it. These brackets can be seen in Figure 9 & Figure
10.
3.2.3.4 Support L clamps
These clamps were attached to hold the carbon fibre support beam firmly in place, with four
holes inserted into the beam on both sides, holding the beam tightly together with screws.
The clamps are also secured to the flat plate base which is held in securely by the MDF base
with glue and screws, as appropriate.
3.2.3.5 Supporting block
At the top of the duct a precision made block was attached holding the carbon fibre motor
support beam as well as the duct securely in place to avoid any oscillations during testing
and minimalize the rubbing or movement of parts.
3.2.4 WIND TUNNEL TEST RIG

Due to time constraints, it was decided at this point that it would be more beneficial to
design the ducted fan thrust rig in a way that would be used exclusively for testing in the
wind tunnel.
3.2.4.1 The idea
The initial idea was to make the wind tunnel rig in a way that the ducted fan could be
rotated at any angle of attack, using some kind of secure rotation method.
At first, the idea was to create a support rig that could accommodate servos for the rotation
method as they can be programmed easily to rotate at set angles of attack based on what
the user sets in an Arduino code breakdown.
A theorised CAD model was attempted for using servos as an accurate method of measuring
angles of attack, but it was found to be too time-consuming and difficult to accommodate.
23
Therefore another method was utilised which incorporated a rotation block that could be set
to 8 different angles of attack using 9 screw holes set in a 8 hole pattern.
Although the initial idea of experimenting with angles of attack could not be completed with
servos, ducts are quite inefficient at angles indirect to the flow anyway. Therefore the idea
of using manual rotation methods did not seem to be too much of a drawback as it would
be possible to measure the difference in thrust produced for these angles of attack, as
required.
Subsequently a new CAD design was devised and presented based on the initial dimensions
given by Zac Kanaa as shown in Figure 12.
Figure 12: Initial measurements for wind tunnel rig
3.2.4.2 Mounting baseplate
In order to mount the duct onto the wind tunnel, a rig had to be built which was able to
surround the duct and secure the arms of the duct to the mounting holes of the wind
tunnel. An initial baseplate was drawn up and designed based on the specific dimensions
shown in Figure 12, including the hole separation distance of 315mm between the mounting
brackets. This was arguably the most important piece to model and construct as it had to
accurately fit the wind tunnel itself.
3.2.4.3 Rotation block
To model and define different angles of attack, a rotation block was designed incorporating
the 8 hole pattern with a centre hole which could be used as a bearing application as this
hole was just needed for rotation purposes. This block was made of Laser cut wood as it is
strong and easy to manufacture.
3.2.4.4 Securing duct block
24
To ensure that the duct was able to be held securely without affecting the lip of the duct a
securing block was modelled based upon the already designed supporting block from the
previous thrust rig design. This block had 3 holes extruded out for the purpose of rotation
which is needed to secure and measure the angles of attack in 45° increments, as designed.
The initial wind tunnel rig can be seen in the top view, side view, bottom view and isometric
view configurations in Figure 13 & Figure 14 below:
3.2.4.5 Initial Design 1
Figure 13: Isometric views of initial wind tunnel rig

This rig would be fairly easy to manufacture and would also withstand well the forces from
the wind tunnel approaching it.
Figure 14: Bottom, Back, and side views of initial wind tunnel rig
25
The other side of the carbon fibre motor mount beam would be secured well with a
machined block to match the geometry of the curvature of the duct, held in place by the
carbon fibre beam and glued together in place with epoxy resin. This can be seen in Figure
15:
Figure 15: Transparent view of initial wind tunnel rig showing mounting of
support block
3.2.4.6 Initial Design 2
This design was the improved version of Design 1. In order to simplify the manufacturing
process, the support block was changed by employing a flat machined block rather than a
geometry specific support block. This alteration also strengthened the adhesion between the
Figure 16: 2nd Initial wind tunnel rig Design with

reconfigured support block
26
support block and the duct. For this purpose, a small indented square matching the
dimensions of the support block was cut out of the duct to accommodate the block. This is
shown in Figure 16.
This block would be slightly inserted into the duct and glued onto both the duct and the
beam with epoxy resin to make the component extra secure.
3.2.4.7 Design Correction based on Literature Survey
The initial design of the duct and the cut-outs of the duct for the fastening of the carbon
fibre motor beam was flawed. The designation of the propeller tips was incorrect and when
comparing it to relevant literature especially Pereira’s findings, it was discovered that the
best efficiency for the shroud would be obtained when the rotor is placed at the lip of the
duct. (Pereira, 2008)
This would have been a major design flaw as the pressure difference between the rotors
would have been much greater due to the offset distance of the rotor. Furthermore, under
natural harmonic conditions, the rotor tips would have worn into the duct itself, thereby
creating noise and reducing the efficiency significantly.
27
To rectify this problem, the duct and the motor beam were redesigned to change the length
and the geometry of the duct where the actual duct lip diameter was 254mm, or 10 inches.
This is shown in Figure 17.
Figure 17: 3rd initial design for wind tunnel rig showing amended rotor
placement
There were a few drawbacks to this amended design.
• Firstly, the duct could swing out of rotation and produce a large bending moment on
the only supporting arm holding it in place in the wind tunnel.
• The supporting block holding the duct on the left would have very specific geometry
and would be therefore hard to match perfectly with the duct’s curved surface.
• The weight of the duct would be likely to weigh down the supporting arm and
therefore bank the structure slightly to a certain direction. This would also be
amplified under wind tunnel testing conditions.
3.2.4.8 Simplified Design for Manufacture
A meeting with the supervisor, Jason Matthews, took place on the 12th February 2018 to
discuss the CAD design of the proposed wind tunnel test rig.
28
In this meeting the CAD design was reviewed and it was decided that the wind tunnel rig
should have a second supporting arm matching that of the opposite side to give the rig a
better structure and to provide more rigidity and robustness in order to hold the duct
together.
The design was iterated with the following changes made from the comments discussed in
the meeting above as seen in Figure 18.
Figure 18: Ducted fan rig with two arms and ideal duct
A second meeting with Jason Matthews was held on 19th February 2018 to discuss suitable
manufacturing methods for the duct.
Also present at this meeting was the Thermolab and CNC Machine supervisor, Geoff Rowley,
who was approached for advice about the design and manufacturing of the duct.
Geoff Rowley specified that only the lower part of the duct could actually be made whilst the
lip of the duct would have to be produced with other manufacturing methods.
Geoff Rowley also advised that it would be too complicated to produce the outer slanted
curved profile of the duct, and that it should be designed in such a way that it has no
diffusion angle on the outside of the duct.
As the diffusion angle would not have made any significant difference to the duct’s thrust
performance capabilities anyway, this was not a huge problem and the design was changed
accordingly. The duct’s new profile with the flatter outside is shown in Figure 19.
29
Figure 19: Amended duct design for 10° diffusion angle

To produce the duct shown in Figure 19 with the CNC Hot Wire Cutter the following
dimensions were required:
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑡𝑡𝑡𝑡𝑡𝑡 𝑑𝑑𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = 386.11𝑚𝑚𝑚𝑚

𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑡𝑡𝑡𝑡𝑡𝑡 𝑑𝑑𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 254𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑑𝑑𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = 386.11𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑑𝑑𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 317.48𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 ℎ𝑒𝑒𝑒𝑒𝑒𝑒ℎ𝑡𝑡 ℎ𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = 179.92𝑚𝑚𝑚𝑚
However, due to the precision of the CNC Machine the dimensions were rounded to the
nearest 0.5mm. Thus the final dimensions for the 10° diffusion angle were as follows:
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑡𝑡𝑡𝑡𝑡𝑡 𝑑𝑑𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = 386𝑚𝑚𝑚𝑚

𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑡𝑡𝑡𝑡𝑡𝑡 𝑑𝑑𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 254𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑑𝑑𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = 386𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑑𝑑𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 317.5𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 ℎ𝑒𝑒𝑒𝑒𝑒𝑒ℎ𝑡𝑡 ℎ𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = 180𝑚𝑚𝑚𝑚
3.2.4.9 Further Problems and Solutions
On 20th February 2018 a problem with the CNC Machine occurred as it was not able to cut a
diffusion angle of 10° which had to be changed to a 5° diffusion angle. This was due to 5°
being the maximum angle that the CNC Machine is able to cut.
30
The amended design accommodating this change is shown in Figure 20 with the modified
dimensions given below.
Figure 20: Amended 5° Duct for CNC Cutting
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑡𝑡𝑡𝑡𝑡𝑡 𝑑𝑑𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = 386.11𝑚𝑚𝑚𝑚 → 386𝑚𝑚𝑚𝑚

𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑡𝑡𝑡𝑡𝑡𝑡 𝑑𝑑𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 254𝑚𝑚𝑚𝑚 → 254𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑑𝑑𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜 = 386.11𝑚𝑚𝑚𝑚 → 386𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 𝑑𝑑𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 285.63𝑚𝑚𝑚𝑚 → 286𝑚𝑚𝑚𝑚
𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷 ℎ𝑒𝑒𝑒𝑒𝑒𝑒ℎ𝑡𝑡 ℎ𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 = 179.92𝑚𝑚𝑚𝑚 → 180𝑚𝑚𝑚𝑚
On Tuesday 27th February 2018, the duct was finally produced by the CNC machine and was
collected from Geoff Rowley in order to be put together and sanded down for further
testing.
Due to the sheer size of the duct, it had to be cut out in two individual halves to ensure it
could be machined properly. This was necessary because the hot wire cutter was not able to
cut the circular profile without actually going inside of the ducted profile itself. As the duct
was large, it was cut in two separate pieces as the wire cutter was only able to cut the 5°
angle in one semi-section at a time.
31
The two halves of the duct were then glued together and the gaps were filled in with
Polyfilla™ and set to dry for 30 minutes. This was to ensure that there would be no
unwanted discrepancies in the testing process. The duct can be seen in Figure 21:
Figure 21: CNC Duct complete at a 5° diffusion angle

3.2.4.10 Final design
After many iterations, the design of the ducted fan rig was finally complete and can be seen
showing the isometric, plan and side views in Figure 22.
Figure 22: Final Design of the wind tunnel ducted fan rig showing isometric, plan and side views
32
3.2.4.11 Building of the Rig
3.2.4.11.1 Supporting block
The support block was manufactured from a large aluminium block that was cut down to the
required dimensions of 50x40x32 mm3 using the guillotine band saw which can be seen in
Figure 23.
Figure 23: Cutting of the aluminium supporting block with the band saw
These blocks were then machined to the correct dimensions using the milling machine which
was capable of taking off 1.25mm of material off the block at a time. Each cut of the block
was made using the automatic feed control and took around 5 minutes per cut or 10
minutes if a finer finish was desired. This process took multiple days to complete because it
was decided to produce as fine a finish as possible, which required slower cuts.
The next task was to drill holes into the aluminium blocks using the milling machine and a
drill bit as well as the quill feed to drill a few millimetres at a time into the material.
This process was completed twice in order to achieve the two supporting blocks required for
the rig.
3.2.4.11.2 Rotating wooden block
Initially it was planned to produce the wooden blocks from aluminium. Due to the complex
positioning of the holes on the piece, however, it was later decided that it would be easier
to print these parts using the laser cutter.
33
The Technician in charge of laser cutting, Geoff Sims, was contacted and the following
requirements for the laser cut were given:
• Material: Plywood
• Thickness: 3mm
• Piece number: 10
As the plywood could only be cut to 3mm thickness per piece, 10 pieces were required to be
printed, 5 per block, totalling the 15mm depth as required.
These blocks were subsequently glued together using a 2 stage glue process with a sealing
spray and a super glue, Mitre bond fast adhesive, which sealed the pieces together in 10
seconds with a tight bonding seal.
Once all the strips had been glued together to form two blocks consisting of 5 pieces each
forming the 15mm depth as required, they were dried overnight to achieve the maximum
adhesion. These blocks were then drilled into by using the same hole diameter that had
been used to cut them out in order to ensure that the holes were flush and that the screws
could easily fit into the components.
3.2.4.11.3 L Clamp and supporting Clamps
The clamps that hold the rotating blocks onto the main baseplate were made out of a long
thin piece of U-shaped aluminium. They were cut down using the guillotine band saw and
then machined to the correct dimensions using the milling machine.
The holes were subsequently drilled into the clamps using the milling machine and the quill
feed respectively. 4 holes were drilled into the piece so that the clamp would have a secure
fit onto the block and also be able to withstand the forces in the wind tunnel.
34
The final rig can be seen in Figure 24.
Figure 24: Final Manufactured Ducted fan Rig
35
3.3 FINAL WIND TUNNEL RIG
The test rig was set up as shown in Figure 25.
Figure 25: Wind Tunnel Mounted ducted fan rig

The Electronic Speed Controllers [ESCs], battery pack and receiver were then connected to
the duct so that the motor could be controlled wirelessly with a controller. This can be seen
in Figure 26 & Figure 27.
The Motor was then powered up wirelessly to full power and can be seen in the collection of
images in Figure 27 & Figure 28.
An additional image was taken showing the air flow coming out of the ducted fan. This can
be seen in Figure 29.
36
Figure 26: Wiring of Wind Tunnel rig
Figure 27: Wind Tunnel Rig mounted and wired
37
Figure 28: Mounted wind tunnel rig powered up
Figure 29: Vortex flow observed from Wind Tunnel Mounted rig
38
4 RESULTS
4.1 RETESTING OF JACK’S DUCT
To measure the thrust obtained from the duct, a moment arm cancelling method was used.
The duct’s centre at the propeller hub centre was measured exactly 10 inches from the
pivoting point of the moving rig arm. The scales to measure the force being thrust were
placed exactly 10 inches away from the pivot point as well. This careful positioning allowed
for any moment arms to be cancelled out and therefore resulting in the force or thrust
output by the duct being exactly equal to the force measured on the scales on the rig.
The rig was tested after the lipo battery was fully charged to ensure the best possible
results could occur and the discrepancies could be eliminated in case of an unreliable power
supply.
The results obtained from the duct are shown in Table 1
duct test of jack's rig

Power input PWM Signal g's of thrust Amps Voltage RPM
0% 1000
10% 1100 17 0.1 12.25 940
20% 1200 63 0.5 12.24 1580
30% 1300 113 1 12.24 2040 Natural Harmonic
40% 1400 177 1.5 12.24 2480
50% 1500 234 2.2 12.24 2820
60% 1600 329 3 12.23 3160
70% 1700 446 4.4 12.2 3640
80% 1800 587 6.3 12.17 4120
90% 1900 748 8.6 12.17 4580
100% 2000 926 11.5 12.17 5060
Table 1: Table of results from the retesting of Jack’s ducted fan test rig
In order to ascertain if the duct was actually producing more thrust and therefore a useful
addition to the normally un-ducted propeller, the duct was removed from the rig so that just
the propeller could be tested on its own to see how much thrust could be obtained from the
setup in this way.
The un-ducted propeller results are shown in Table 2 below:
39
no duct test of jack's rig

Power input PWM Signal g's of thrust Amperes Voltage RPM
0% 1000
10% 1100
20% 1200 70 0.5 12.3 1380
30% 1300 137 1 12.29 2060
40% 1400 188 1.5 12.26 2420
50% 1500 244 2.2 12.25 2780
60% 1600 317 3.1 12.24 3140
70% 1700 413 4.5 12.23 3620
80% 1800 542 6.4 12.17 4080
90% 1900 680 8.8 12.17 4500
100% 2000 860 11.9 12.1 4960
Table 2: Table of un-ducted propeller results using Jack's Rig
As can be seen from the results above, a natural harmonic occurred on the duct at around
the 30% power input which may have caused the results to fluctuate and therefore produce
an anomalous output in the results.
Table 3 has been devised to show the differences in grammes of thrust achieved as a
comparison of both methods with the duct and without the duct.
Power input PWM Signal Actual difference (g) % increase for ducted vs un-ducted
0% 1000 N/A N/A
10% 1100 N/A N/A
20% 1200 -7 -10.00%
30% 1300 -24 -17.52%
40% 1400 -11 -5.85%
50% 1500 -10 -4.10%
60% 1600 12 3.79%
70% 1700 33 7.99%
80% 1800 45 8.30%
90% 1900 68 10.00%
100% 2000 66 7.67%
Table 3: Comparison between ducted and un-ducted rotor on Jack's Rig
This table shows that the actual thrust increase and % increase is not applicable to the
lower power inputs from 0% to 10%. It can also be observed that the duct actually
performed more poorly at lower power inputs from 20% - 50% where the % increase
40
achieved was negative. It can therefore be concluded that the un-ducted rotor was the
better rig to use at lower speeds.
Although there was a large discrepancy in the results as described above, from 60%
onwards the duct did achieve a larger thrust intake than the un-ducted rotor reaching to
levels of maximum 10.00% greater at a power rating of 90%.
Even though this is a significant thrust improvement compared to the un-ducted rotor test,
this is still nowhere near the results that Jack Ingleby was obtaining from his test results
which are shown in Table 4 & Table 5 below.
The rating to be used for a 10 inch propeller is given in Table 6 which gives a representation
of the theoretically best thrust quantities that the motor can withstand.
Table 4: Jack's Open rotor test results
Table 5: Jack's ducted rotor results
41
Table 6: Motor Specification for the MT 2216 810Kv from http://www.unmannedtech.com
Several Graphs were drawn up based on the results attained and can be seen in Figure 30 &
Figure 31.
Figure 30 shows the comparison of thrust vs RPM for Jack’s thrust rig under both ducted
and un-ducted conditions.
Comparison of Thrust vs RPM for ducted fan vs

un-ducted fan
1000
900
800
700
Thrust (grammes)
600
500
400
300
200
100
0
0 1000 2000 3000 4000 5000 6000
RPM (Revolutions Per Minute)
no duct duct
Figure 30: Graph of the comparison for the thrust vs RPM for un-ducted vs ducted fan
42
Figure 31 shows the comparison of RPM vs PWM for Jack’s thrust rig under both ducted and
un-ducted conditions.
Comparison of RPM vs PWM for ducted fan vs un-ducted

fan
6000
5000
RPM (Revolutions Per Minute)
4000
3000
2000
1000
0
0 500 1000 1500 2000 2500
PWM (Pulse Width Modulation)
no duct duct
Figure 31: Graph for comparison of RPM vs PWM for un-ducted vs ducted fan
4.2 3D PRINTED DUCT TESTS
To test the efficiency of the ducts in a quick and appropriate manner, it was initially planned
to use 3D Printing facilities to manufacture the ducts as this would be the easiest way to
produce the propeller mount and complex geometry such as the lip and the diffusion angle.
2 different ducts were designed and printed each of which had slightly different geometry to
it.
The 2 ducts tested were as follows:
• One Grey duct which had a divergence and diffusion angle of 10°
• One Green duct which had a divergence and a diffusion angle of 5°
• Open rotor test with a 4 bladed prop of 3”x3” specification
The first test to be conducted was the open rotor test in order to determine the thrust that
could be generated from the 3 inch propeller so that accurate thrust improvements could be
measured from this base data.
The un-ducted results are presented in Table 7.
43
No Duct
PWM Signal g's of thrust Amperes Voltage RPM
0% 1000
10% 1100 12 0.3 12.37
20% 1200 40 1.2 12.36
30% 1300 62 1.9 12.36
40% 1400 83 2.9 12.37
50% 1500 103 3.8 12.34
60% 1600 123 4.7 12.33
70% 1700 137 5.5 12.32
80% 1800 157 6.6 12.3
90% 1900 187 8.5 12.3
100% 2000 203 9 12.33
Table 7: Un-ducted rotor test on 3° prop thrust rig
The RPM was left blank because it was not possible to measure with the RPM reader as the
difference between the air and the propeller blade could not be determined by the reader
resulting in the reader displaying zero.
The first duct to be tested was the Green divergent duct. The setup can be seen in Figure
32.
Figure 32: Green divergent duct with thrust rig setup and 3 inch propeller
44
The results obtained from this duct can be seen in Table 8
Green Divergent Duct

0% 1000
10% 1100 13 0.3 12.05 4230
20% 1200 41 1.7 12.04 7710
30% 1300 68 1.9 12.03 8370
40% 1400 94 2.9 12.02 11460
50% 1500 118 3.8 12.01 5070
60% 1600 137 4.7 11.98 1530
70% 1700 156 5.5 11.99 2490
80% 1800 186 6.6 11.95 5520
90% 1900 216 8.4 11.97 4470
100% 2000 230 8.8 11.97 8310
Table 8: Green divergent duct results test from 0%-100% power input
As can be seen straight away, the thrust increased when the duct was attached to the rotor.
This can be observed when comparing the thrust values of Table 7 with Table 8. The thrust
increased with every increase in power input. However, the lowest reading ascertained was
the second result at 41 grammes of thrust for the duct compared to 40 grammes of thrust
for un-ducted rotor.
The next test to be conducted was on the Grey divergent duct which was carried out in the
same conditions as the previous tests. The same 4-bladed 3-inch rotor, same battery and
Arduino code were used.
The table of results for the Grey divergent duct can be seen in Table 9.
Grey Divergent Duct
0% 1000
10% 1100 14 0.2 12.24
20% 1200 36 1 12.22 natural harmonic
30% 1300 73 1.9 12.17
40% 1400 100 3.9 12.15
50% 1500 124 3.9 12.17
60% 1600 148 4.8 12.15
70% 1700 170 5.6 12.14
80% 1800 198 6.8 12.11
90% 1900 238 8.6 12.1
100% 2000 256 9.2 12.17
Table 9: Results for the grey divergent duct tested from 0% to 100% power input
45
The above table clearly shows that the thrust achieved for the second power input was less
than that of the un-ducted rotor. This means that the duct was a worse performer than the
un-ducted rotor and was actually producing more drag than thrust increase. It can therefore
be concluded that it is not suitable to use on the UWE UAV project.
The full comparison table of results can be seen in Table 10
% Improvement
Power Input Green Grey
10% 8.333333333 16.6666667
20% 2.5 -10
30% 9.677419355 17.7419355
40% 13.25301205 20.4819277
50% 14.5631068 20.3883495
60% 11.38211382 20.3252033
70% 13.86861314 24.0875912
80% 18.47133758 26.1146497
90% 15.50802139 27.2727273
100% 13.30049261 26.1083744
Average 13.75301459 22.8150948

Table 10: Comparison between un-ducted and ducted rotor for both
Green and Grey ducts
The resulting graphs can be seen in Figure 33 which shows the comparison of thrust vs
PWM for the un-ducted and ducted tests.
46
Comparison of un-ducted fan vs 2 different ducted fans for

PWM vs Thrust
300
250
Thrust (grammes)
200
150
100
50
0
0 500 1000 1500 2000 2500
PWM (Pulse Width Modulation)
Green Divergent Duct Grey Divergent Duct No Duct
Figure 33: Graph of comparison for PWM vs thrust of un-ducted fans vs 2 different ducted fans
47
4.3 WIND TUNNEL TESTING
4.3.1 Wind Tunnel Test Rig Data

The results were tested in a way to compare the drag or thrust produced without the motor
running, and then the thrust with the motor running. Subsequently the differences were
compared to see how much extra thrust the motor produces on top of the duct’s thrust
generation.
The results are shown in Table 11.
Speed (ms-1) Drag (thrust) without motor (N) Drag (thrust) with motor (N)
0 0 -6.8
5 2.03 -2.1
8 5 1.31
10 7.42 5
15 18.1 15.56
20 31.12 29.54
Table 11: Table of Wind tunnel results of ducted fan at 0°
As all the other tests were completed in grammes of thrust, the units were converted by
dividing the acceleration due to gravity which can be assumed as 9.8077 ms-2 on Earth
(Taylor & Thompson, 2008).
Table 12 shows the thrust obtained from the ducted fan at 0° angle of attack in grammes of
thrust to match the other result data attained.
Speed (ms-1) Drag (thrust) without motor (g) Drag (thrust) with motor (g)
0 0 -693.4070248
5 207.0023912 -214.1404047
8 509.8581065 133.5828239
10 756.62943 509.8581065
15 1845.686345 1586.678427
20 3173.356855 3012.241693
Table 12: Table of Wind Tunnel results of ducted fan at 0° in grammes of thrust
48
4.3.2 Influence of Wind Tunnel Mounting Rig

To find out how much drag the actual rig used to support the ducted fan was causing, a
further test result was taken of just the mounting rig. The results from this test can be seen
in Table 13.
Speed (ms-1) Lift (N) Drag (thrust) (N) Pitch (Nm)
0 0 0 0
5 -0.5 0.4 -0.05
8 -0.57 2.56 -1.5
10 -0.53 3.56 -0.12
15 -0.99 7.74 -0.17
20 -1.86 13.54 -0.2
Table 13: Results from amount of drag influence by wind tunnel mount
These results were converted to grammes to match the results obtained from the other test
result data as shown in Table 14
Speed (ms-1) Lift (g) Drag (thrust) (g) Pitch (gm)

0 0 0 0
5 -50.98581065 40.78864852 -5.098581065
8 -58.12382414 261.0473505 -152.9574319
10 -54.04495929 363.0189718 -12.23659456
15 -100.9519051 789.2603488 -17.33517562
20 -189.6672156 1380.695752 -20.39432426
Table 14: Results from amount of drag influence from wind tunnel mount converted to grammes
49
5 EVALUATION
5.1 PREVIOUS DESIGN IMPROVEMENTS
There were several things that presented a disadvantage in the previous design created by
Jack Ingleby. These are listed below as follows:
Jack’s Rig used dovetail joints for the mounting of the duct on top of the supporting rig, so
that the top could be easily disassembled and replaced with other ducts of the same
diameter but different parameters, such as diffuser angle or lip radius.
• Although using dovetail joints to secure the duct in place may have seemed a good
idea due to the sheer weight of the top of the rig, including the Computer
Numerically Controlled (CNC) lip of the duct, this eventually caused the dovetail joint
to wear out. Therefore a rocking motion could be applied to it which had the
propensity to interfere with the results gathered from testing.
• In order to solve this problem, a simpler design was used which utilised a carbon
fibre beam of 5mm width which was sturdy enough to mount onto the rig but unable
to bend under high amounts of force. This ensured that the results obtained from
the rig would not be disrupted. This carbon fibre beam is bound onto another beam
which is integrated into the duct with epoxy resin glue to hold it firmly in place.
Jack’s rig was produced from several pieces of laser cut wood and assembled together to
make a final design which used cancellation of bending moments to measure the thrust
achieved from the duct when in operation.
• Measuring the thrust with bending moment cancellation is a good way of measuring
the thrust. However, as the material density of laser cut wood is not consistent,
there could be several bending moments applied along the wooden beam due to the
length of the beam being 10 inches each way.
• To resolve this problem, strain gauges can be used to measure the slight fluctuations
in the deformation of the carbon fibre beam. These can be applied to the front, side
and back of the beam to accurately obtain the thrust values accurate to 6 decimal
places.
Only one arm is used for the rotation of the ducted fan as shown in Figure 13 because this
will be easier to manufacture and also reduce the time taken to build it while still
maintaining sufficient strength against the strong aerodynamic forces in the wind tunnel.
50
Due to the complication in the angle of the duct, the securing duct block will only rotate 45°
at a time.
Another advantage of using one arm to hold the duct is the weight reduction as one arm will
be less heavy than two supporting arms. Furthermore it will be cheaper to manufacture and
design, which means that more ducts can be produced using this configuration at a reduced
cost.
5.2 CARBON FIBRE LAYUP
The carbon fibre layup encountered a few problems during the manufacturing process. The
carbon fibre weighed roughly 140g and so the ratio of 100:29 was taken as previously
stated. However, due to the 40 minute time period before the resin hardened, it was difficult
to lay up each sheet of carbon fibre cut-out accurately. The further time pushed on, the
quicker the sheets had to be laid down. As the sheets had to be laid down with tension at
each corner to achieve the perfect fibre density, this procedure was difficult to maintain. As
a result of this, some delamination occurred in the final layup. At these sections, the resin
had not mixed or cured properly with the carbon fibre sheets. This was probably due to a
single strand of carbon fibre being mixed in with the flat layup, therefore destroying the
lamination at that certain section. Although this was a problem, it did not affect the final
result too much as the resin layup covered the majority of the sheet anyway. Therefore a
complete seal and curing process was successfully achieved. As the entire layup is 20 layers,
this was more than enough to withstand the forces that the wind-tunnel generated and also
reduced or completely dampened out any possible oscillations in the motor beam, thereby
minimizing vibrational effects, which, according to the literature review, can have a
substantial effect on the efficiency of the ducted fan.
51
5.3 JACK’S RIG RETESTING CONSIDERATIONS
During the testing it was noticed that the blade and tips of the propeller oscillated at certain
speeds and RPMs. This seemed to occur at the point of roughly 60%-70% input power
where the tips of the propeller were actually scraping or rubbing against the shrouding of
the inner duct. This occurrence led to the results being far lower that what should have
been obtained from the testing.
It can be observed that the maximum thrust output in Table 6 at the 6000+ RPM mark
generates roughly 600-670 grams of thrust for a standard un-ducted rotor propeller of
specification 1045, meaning 10 inches with a 4.5 pitch. Therefore when comparing to Jack’s
results of the open rotor in Table 4 the maximum thrust obtained in the experiment was
found to be 496.1 grammes at an RPM of 7180. This discrepancy suggests that Jack’s
results are somewhat different to the ideal motor tested specification for the standard 1045
inch propeller. The reason for this could be due to the fact that the propeller was trimmed
at the tips which could be responsible for reducing the amount of thrust that the open rotor
propeller was able produce. However, compared to the results obtained in this project, the
outputs for thrust vary severely and the total extra thrust gained is only amounting to
around 10% while for Jack’s results the total thrust increase was seen to be around 76%.
Therefore it can be concluded that the experiment performed by Jack either had unlevelled
scales, a badly trimmed propeller or that the motor or ESC was not functioning properly,
maybe because of the age of the motor or ESC.
To combat this, new ESCs and motors were purchased in order to obtain the maximum
efficiency using the best possible equipment.
5.4 3D PRINTED DUCTS EVALUATION
As can be gleaned from the 3D printed duct results, the highest thrust achieved was seen to
be 203 grammes, which is just under half the amount of thrust obtained from a 10 inch un-
ducted rotor. These results suggest that scaling up the size of the rotor does not necessarily
lead to a linear increase in thrust efficiency. Therefore separate testing has to be done to
find the curve at which the thrust efficiency lies on for each representative rotor dimension.
52
5.5 WIND TUNNEL CONSIDERATIONS
To conduct wind tunnel testing, a session had to be booked in advance. As the time taken
to design and manufacture the wind tunnel ducted fan rig was unknown, a date could not
be booked. Therefore the plan was to complete the manufacturing process before booking a
session.
Unfortunately, it proved to be rather difficult to obtain a wind tunnel session after the
manufacturing was underway. The manufacturing process started mid-February 2018 with a
carbon fibre motor mount plate and continued for another month up until 28th March 2018,
over-running the estimated time of manufacture by 1 month and 14 days.
Fortunately, a student who had booked the wind tunnel for Wednesday 28th March 2018 for
the whole day did not require the full day’s use of the wind tunnel, which allowed a small
slot to be booked for the thrust testing of the ducted fan.
Therefore, the wind tunnel testing was completed on the ducted fan on time and accurate
results were obtained from this.
As can be seen from Table 12, the thrust of the ducted fan rig with the motor running is less
strong than the thrust without the motor running. This means that the ducted fan actually
produces more drag than lift under wind tunnel testing conditions.
This could be due to the way the battery and wiring were attached to the rig as it required a
large amount of tape to stick them onto the duct, which could have influenced the amount
of drag obtained as a result.
Comparing these data to the data relating to the retesting of Jack’s ducted fan rig, however,
shows that ducted fans provide a good addition of thrust and can therefore be used for the
application of UAV’s/MAV’s as they provide not only extra thrust but also reduce the power
consumption needed to provide the thrust required for UAVs.
53
6 ANALYSIS AND DISCUSSION
6.1 COMPARISON OF JACK’S DUCTED FAN RIG WITH THE LITERATURE
The results obtained from the retesting of Jack’s ducted fan thrust rig were seen to be much
lower than those acquired during Jack’s initial experimentation of the rig in his
experimentation process. The biggest discrepancy in the results obtained can be seen in
Table 1, Table 2 & Table 3, where the thrust in grammes obtained from the rotor was seen
to be 496 grammes of thrust compared to the value that was obtained when the experiment
was repeated, namely 860 grammes of thrust for open rotor.
The data from the actual motor, which was chosen to be an MT 2216 810 kV displayed
thrust values of 670 grammes of thrust at maximum power at an RPM of 6670. The
comparison of these data shows that Jack’s results for an un-ducted rotor are considerably
lower than the actual motor specification. In reality, the value that Jack should have
obtained should be near the value of that of the data relating to the motor testing
specification. As this is not the case, this could mean that most of Jack’s results have
discrepancies across the data sheets acquired.
For the ducted rotor, the obtained thrust for Jack’s experiment was 873.7 grammes of thrust
compared to the value obtained when the experiment was retested of 926 grammes of
thrust at a PWM of 2000. Comparing these results with each other, it can be seen that
Jack’s results obtained from the un-ducted rotor seemed to achieve an increase of thrust of
76.11%. This would be a close match to the literature review which suggests that the
performance in thrust efficiency by the use of ducted fans can increase by up to 97%
compared to the performance produced by an un-ducted rotor configuration.
When the thrust rig was retested, the values achieved for the un-ducted rotor were found to
be 860 grammes of thrust and 926 grammes of thrust for the ducted rotor at maximum
power at 2000 PWM. This represents an increase in thrust of just 7.67% which, although it
is an increase and shows that the duct is an extra generator of thrust, is comparatively low
when analysed against Jack’s experimental results. There could be a discrepancy, however,
in this particular result as 860 grammes of thrust was achieved on the open rotor, whereas
the actual motor’s specification notes that the thrust achieved at 6620 RPM is 670 grammes.
If the open rotor results were to match the motor specification results for a 1045 propeller
as required, the total thrust percentage increase would be 38.21% which is a much greater
54
thrust increase although it is not even close to the 97% that the literature suggests can be
achieved at maximum or the 76.11% achieved by Jack’s experiment.
It can be seen from Table 12 that the ducted fan is able to provide 693.41 grammes of
thrust when the motor is running at full power. This is lower than the result that Jack
Ingleby achieved, which was 873.7 grammes of thrust. It should be taken into
consideration, however, that this test was produced on a newly made duct and that the
results will therefore differ.
Compared to the results obtained by the retesting of Jack’s duct, which achieved a total
thrust of 926 grammes, this figure seems rather low, which could be due to the fact that a
significant amount of weight was being hindered by the actual mounting rig part of the duct.
This could have affected the total thrust performance that the duct could achieve and
therefore lowered the efficiency of the ducted fan.
The thrust obtained by the ducted fan when run at 5 ms-1 was not very efficient as it only
obtained a total thrust of 214.14 grammes. This presented a problem as this result means
that the ducted fan achieved less than a quarter of thrust that un-ducted rotor achieved,
which Table 1 shows as 926 grammes of thrust for the retest and 873.7 grammes of thrust
for Jack’s results.
This setup would therefore not be a suitable option to use as the wind tunnel ducted fan is
extremely inefficient at low speeds. The ducted fan was also seen to be inefficient at other
speeds such as 8 ms-1 and 10ms-1 where the thrust achieved was 133.58 grammes and
509.86 grammes respectively. This means that the total thrust for the 8ms-1 run dropped
further below the 214.14 grammes for the 5ms-1 run. It can therefore be concluded that at
low speeds the duct is very inefficient and hence should be used at higher speeds to achieve
maximum efficiency.
This can be shown where the thrust of the ducted fan rapidly increases to 1586.68 grammes
of thrust at the 15ms-1 run and as much as 3012.24 grammes of thrust for the 20ms-1 run.
This increase in thrust shows that the ducted fan is capable of achieving a large increase in
efficiency of thrust just by increasing the speed of the airflow.
55
7 CONCLUSION
It can be concluded that the testing of the ducted fans was an overall success as a higher
efficiency in thrust was achieved.
7.1 Wind Tunnel remarks
From the wind tunnel testing it can be seen that the ducted fan does not perform well at
low airspeeds and is therefore very inefficient. Only a quarter of additional efficiency was
achieved in this way compared to the ducted fan results from Jack’s rig experiments and the
retesting of Jack’s rig experiments.
Although this is a substantial decrease in the efficiency of the ducted fan, the thrust increase
rose significantly at higher airspeeds such as 15ms-1 and 20ms-1 to a substantial 1586.68
grammes of thrust for the 15ms-1 run and as much as 3012.24 grammes of thrust for the
20ms-1 run.
If this duct was used in a quadcopter this would mean that the total thrust from the rotors
would be as much as 12,000 grammes of thrust. Considering the fact that the maximum
weight for a quadcopter is 7,000 grammes, this represents a substantial amount of thrust.
With a larger and more efficient motor, the thrust could be even higher.
As shown in the wind tunnel data in Table 12, the thrust achieved by the motor was lower
than the thrust achieved when the motor was not running. This means that the motor used
was not the strongest one as it was not able to overcome the thrust intake of the air
blowing through it. To resolve this problem, a stronger motor capable of producing a higher
amount of thrust and power should be used so that the thrust generated will be able to
overcome the aerodynamic forces present in the tunnel.
7.2 Parameters to make a note of
From testing the ducted fans with various geometries and manufacturing methods, it can be
concluded that one of the most important parameters in achieving thrust efficiency is the tip
clearance, as referenced closely by Pereira’s work, as changes in half millimetres can affect
thrust efficiency massively, reducing efficiency by up to 25% just from a slight change in the
clearance. It is therefore most important to keep this geometric feature perfect and exactly
the right size by matching the shrouding of the duct’s geometry to the propeller’s tip
diameter.
56
Another important parameter in achieving thrust efficiency through testing was found to be
the diffusion angle of the duct. As the CNC Machine hot wire cutter was not able to produce
a duct diffusion angle of 10° as requested, some of the potential efficiency increase might
have been reduced, affecting the duct’s thrust performance.
The diffusion angle of the 3D Printed ducted fans was also negatively impacted by not
having the best geometry due to the lack of precision of the printer and the imprecise way it
handled the curves during printing. As a consequence, the diffuser of the duct had to be
sanded down resulting in a loss of thrust efficiency due to the changes in tip clearance
endured.
A factor that did not seem to have a major effect on the thrust efficiency was the duct
height. The test results showed that the duct height could be varied quite a bit without
affecting thrust efficiency greatly. For instance for the 10 inch ducted fan, the total variance
in height amounts to 55.88mm which is quite negligible. It must not be forgotten, however,
that the greater the height of the duct, the more air is allowed to pass through the rotor
generating more thrust as a result. Furthermore, the amount of space below the rotor could
have a negative impact on thrust efficiency due to the change in pressure difference across
the propeller affecting how much airflow is going through the fan.
7.3 Readdressing aims and objectives
Looking back at the aims and objectives stated at the beginning of this report, it can be
seen that most of the aims and objectives were met. This included objective number 5, the
wind tunnel testing, which was one of the main milestones of this project, as it gave
different results to the ones produced by Jack Ingleby and was not just a retesting of his
ducted fan rig, or other ducted fan tests.
The first objective, which was the use of MatLab to programme functions and to generate
graphs, was not met. The reason for the decision not to meet this objective was that the
programme was not required after all, as all calculations and graph plotting were completed
on Microsoft Excel instead because it was easier to format and produce clear data from by
using this application.
The second objective was met, however, by the retesting of Jack Ingleby’s ducted fan rig
(Ingleby, 2017). For this purpose, Jack Ingleby’s ducted fan rig had to be reassembled, a
new Arduino code had to be created and the tests had to be taken under the same
conditions as the ones used by Jack. This retesting procedure turned out to be very
57
important indeed as different results were obtained, even though the same rig, the same
motor, same battery etc. were used. Assumptions were made about what may have caused
this difference of the results in the 5.3 Evaluation section.
The third objective of using composites in the final design and manufacture was also met.
For this purpose, carbon fibre was used for the motor support beam in the final design and
implemented to increase the amount of force the previous beam was able to withstand. This
proved to be very successful as the beam was solid when it was fitted into the duct and was
difficult to bend under loading.
Objectives 4 and 7 were met by designing, manufacturing, and assembling a ducted fan rig
and using the CNC machine to cut the duct’s profile. The duct which was produced in this
way ended up being very smooth in cut, and as a result no extra sanding had to be done on
the duct as the tip clearance was 0mm - 0.025mm. Although this rig was supposed to be a
thrust rig initially, due to time constraints a wind tunnel rig was produced instead which was
able to obtain lifelike results. Fortunately, this wind tunnel rig had an advantage over the
thrust rig, as it was able to achieve results that were not completed by Jack beforehand.
Objective 6 was met by designing several of the ducts and then 3D printing them to achieve
quick results whilst maintaining a sufficient level of accuracy in the manufacture. A simple
thrust rig using moment cancellation was constructed for the small 3 inch diameter ducts
and useful data was attained as shown in the results section in chapter 4.2. This helped to
gain some idea of how the ducts would perform on a smaller scale.
A comment that is required to be made about using the 3D printed ducted fans in the
pursuit of achieving objective 6 is that due to high differences in the Reynolds Number for
geometrically different ducts, separate tests had to be conducted for the ducts to attain
accurate results, as it was impossible to scale them up. (Pereira, 2008) This can be shown in
the Results section where the thrust achieved for the 3D Printed ducts of 3 inch diameter
was incomparable to the thrust produced by the 10 inch duct which generated a significantly
lower amount of thrust than was to be expected. This required a new scale model to be
built for the ducted fans, which is why a 10 inch duct had to be made independently of the
3D printed ducts to achieve results specific to that size of duct.
7.4 Future Work/Recommendations
Future work to be conducted would be to analyse more data and obtain more results from
wind tunnel tests. In order to keep to a proper time schedule and not fall behind the Gantt
58
chart plan, a large portion of testing and manufacturing will be conducted during the
summer period. This will allow sufficient time for other work to be completed throughout the
year and will avoid becoming lost in the timescale devised.
Conducting more wind tunnel tests will allow more lifelike data to be achieved and will allow
more results to be obtained from these tests. It will also provide the opportunity to compare
the obtained results more accurately with the literature presented and researched for this
project.
A new literature review will also be conducted involving a more in depth study of various
aspects relating to the use of ducted fans. This will include further research into important
aspects such as noise emissions which, although complex, is a subject that has attracted a
considerable amount of research in current literature. Further research will also go into the
designation of a more suitable motor as the one used for this project seemed to lack the
power to generate enough thrust to overcome the incoming air into the duct and the
aerodynamic forces associated with this. More detailed research will also be required on the
adverse effects of tip clearance modification in order to achieve the most geometrically
precise duct during the designing and manufacturing process.
Further work that could be done on this project might include the research and
implementation of the use of Computational Fluid Dynamics (CFD) to obtain lifelike values
for specific ducted fan structures. This would be useful as the implementation of CFD would
mean that the construction of a ducted fan would be rendered unnecessary for achieving
lifelike results. The proficient use of CFD could open the door to a vastly improved way of
testing multiple different geometries and their effect on thrust performance by simulation.
59
8 REFERENCES
Abrego, A.I. and Bulaga, R.W. (2002) Performance Study of a Ducted Fan System. [online].
1st edition. San Francisco: American Helicopter Society International. [Accessed 28 January
2018].
Avid LLC (2016) Why Ducted Fan UAV? Available from:

https://static1.squarespace.com/static/54f8af0ce4b087f9c65cf093/t/5509d107e4b0d53e642
dc467/1426706697402/ductedfancfd.png?format=1500w [Accessed 15 November 2017].
Bouferrouk, A. (2018) Lecture 2: Application of the Momentum Equation to aircraft

Propellers and Rotors. [Accessed 28 January 2018].
De Piolenc, F.M. and Wright Jr, G.E. (2001) Ducted Fan Design, Volume 1: Volume 1 -
Propulsion Physics and Design of Fans and Long-Chord Ducts [online]. 1st ed. : Marc de
Piolenc. [Accessed 11 January 2018].
Gelhausen, P (2015) The Case for Ducted Fan UAV. Available from:
http://www.avidaerospace.com/blog/2015/3/24/the-case-for-ducted-fan-uavs [Accessed 13
November 2017].
Ingleby, J. (2017) Exploring the Enhancement of Thrust Characteristics Using Ducted Fan
Theory for a UWE UAV Development Group Project. BEng Aerospace Engineering. University
Of the West Of England.
Jang, C.M., Inoue, M. and Furukawa, M. (2001) Noise Reduction by Controlling Tip Vortex in
a Propeller Fan. Jsme International Journal Series B Fluids and Thermal
Engineering [online]. 44 (4), pp. 748-755. [Accessed 30 January 2018].
Krüger, W. (1949) Technical Memorandum on Wind Tunnel Tests and Computations

concerning the Problem of Shrouded Propellers [online]. Report number: NACA-TM-1202.
Washington, National Advisory Committee for Aeronautics. Available from:
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20050241797.pdf [Accessed 24 March
2018].
Li, J.B. and Gao, Z. (2005) Aerodynamical characteristics analysis of ducted fan. Journal of
Nanjing University of Aeronautics & Astronautics, 37 (6), pp. 680-684. [Accessed 8
November 2017].
60
Mendenhall, M. & Spangler, S. (1970) Theoretical study of ducted fan performance. [online].
1st edition. Washington, DC: National Aeronautics and Space Administration. [Accessed 28
January 2018].
Mort, K. & Yaggy, P. (1962) Aerodynamic characteristics of a 4-foot-diameter ducted fan

mounted on the tip of a semispan wing. [online]. 1st edition. Washington, DC: National
Aeronautics and Space Administration. [Accessed 28 January 2018].
Mort, K. & Gamse, B. (1967) A wind-tunnel investigation of a 7-foot-diameter ducted

propeller. [online]. 1st edition. Washington, D.C.: National Aeronautics and Space
Administration. [Accessed 28 January 2018].
Mort, K. (1965) Performance characteristics of a 4-foot-diameter ducted fan at zero angle of

attack for several fan blade angles. [online]. 1st edition. Washington, D.C.: National
Aeronautics and Space Administration. [Accessed 28 January 2018].
Nagpurwala, Q. (2016) Ducted Fans and Propellers. [PowerPoint presentation]. No Date.

Available from: http://164.100.133.129:81/eCONTENT/Uploads/09-
%20Ducted%20Fans%20and%20Propellers%20%5BCompatibility%20Mode%5D.pdf
[Accessed 13 November 2017].
Parlett, L.P. (1960) Wind-tunnel investigation of a small-scale model of an aerial vehicle

supported by ducted fans [online]. Langley Field, Va: National Aeronautics and Space
Pereira,J. (2008) Hover and Wind -Tunnel Testing of Shrouded Rotors for Improved Micro
Air Vehicle Design [online] PhD, University of Maryland. Available from:
https://drive.google.com/file/d/0BzZ213vBQVkTOHFnbUxkRVRZOGc/view [Accessed 1
November 2016].
SACKS, A. & BURNELL, J. (1959) Ducted propellers. A critical review of the state of the art.
[online]. Ft. Belvoir: Defense Technical Information Center. [Accessed 28 January 2018].
Smith, C. (1960) Wind-tunnel investigation of a small-scale model of an aerial vehicle

supported by tilting ducted fans.[online]. Washington, D.C.: National Aeronautics and Space
Stickle,W and Crigler,J. (1940) Propeller Analysis from Experimental Data [online] Report
number: 712. Langley Field: Langley Memorial Aeronautical Laboratory. Available from:
61
http://naca.central.cranfield.ac.uk/reports/1941/naca-report-712.pdf [Accessed 13
November 2017].
Stipa, L. (1931) Experiments with Intubed Propellers.[online]. L'Aerotecnica ppo 923-953.

Translated by Dwight M. Miner, NACA. NACA TM 655, January, 1932. [Accessed 28 January
2018].
Stipa, L. (1934) Stipa monoplane with Venturi fuselage.[online]. Washington, DC: National
Advisory Committee for Aeronautics. [Accessed 28 January 2018].
Taylor, B. & Thompson, A. (2008) The international system of units (SI). Gaithersburg:
National Institute of Standards and Technology. [Accessed 28 January 2018].
Van Niekerk, C. (1956) Measurement of the Noise of Ducted Fans. [online]. 28th edition.
Pretoria: Acoustical Society of America. [Accessed 28 January 2018].
62
9 BIBLIOGRAPHY
MathWorks (2016) MATLAB (R2017a) [Computer Programme]. Available from:
https://uk.mathworks.com/ [Accessed 7 November 2017]
Microsoft (2016) Excel (2016) [Computer Programme]. Available from:

https://products.office.com/en-gb/excel [Accessed 20 November 2016].
Solidworks (2016) Solidworks Standard (2016) [Computer Programme] Available from:

http://www.solidworks.co.uk/sw/products/3d-cad/solidworks-standard.htm [Accessed 13
November 2017].
63
10 APPENDICES
Figure 34: Extra images from the wind tunnel testing of the ducted fan rig
10.1 CARBON FIBRE MOTOR BEAM

The first part of the thrust rig to be manufactured was the motor support beam, which was
designated to be made from 5mm thick carbon fibre.
There are many different manufacturing methods in making the 5mm thick carbon fibre
sheet.
The methods are:
• Wet lay-up procedure
• Vacuum Forming moulding procedure
• Heat Pressing with Pre-impregnated carbon fibre sheets with a 2/2 Twill
• Autoclave manufacturing with heat addition and pressure forming
As an autoclave is not available at the university and the cost to use an autoclave machine
for such a small piece of equipment is substantial, it was decided that it would be best not
to use this manufacturing method for the carbon fibre motor beam
The heat pressing with pre-impregnated carbon fibre was a good idea as the carbon fibre is
already impregnated with epoxy and just requires the addition of a heat and pressure press.
As the university has one of these, this method seemed feasible, however, the heat press
can only cover a 250mmx250mm area and therefore as the motor beam is too big, this will
not suffice and will result in the inside of the beam being rigid and well cured whilst the
outside will be uncured and delamination will occur.
64
A discussion was held with university Technician, Daniel Cole, to ask whether the stress
analysis press could be used to cure the carbon fibre sheet. He confirmed that it could be
done, however a pressing sheet made of metal would have to be ordered in order to fit the
dimensions of the beam as required.
In the end it was decided that a vacuum formed carbon fibre layup would be the best setup
to use as it is fast and the materials are readily available in the workshop. To begin with, a
drawing was made on CAD of the motor beam with the important dimensions listed, and this
was printed off on A3 paper with a 1:1 scaling to ensure that it could be directly copied
when cutting up the carbon fibre sheets.
20 sheets of carbon fibre of 0.25mm thickness each were required to achieve the 5mm
thickness desired for the motor beam. Once this was done, the correct amount of epoxy:
hardener ratio had to be determined which was found to be a ratio of 100:29 where the 100
is the epoxy, and the 29 is the hardener.
As the epoxy and hardener mix takes a certain time before it completely cures and hardens,
the layers had to be laid up quickly in order to achieve the best layup possible. Before the
layers were laid down, a surface was cleaned with wet and dry wax mould 5 times to ensure
that the carbon fibre could be laid flat with no bumps and therefore avoiding delamination
where possible. Between each layer the epoxy/resin mix was dabbed on top of the sheet,
covering all areas of it to make sure that the next layer was able to stick to it well. Each
layer was stretched lightly and placed on top and coated with resin, finally achieving all
layers coated as shown in Figure 35.
Figure 35: Process of preparing and laying up of the Carbon Fibre Motor Beam
Once this was done, the carbon fibre layup was covered in a blue perforated film. Perforated
film allows the resin to be absorbed by the vacuum onto the upper surface of it which is
placed just below the bleed blanket. The bleed blanket allows the resin to be vacuumed
65
upwards, which is spread out across the surface area which it covers, and the vacuum
covers this area equally due to the presence of the bleed blanket.
The vacuum bag acts like a chamber to prevent the air from escaping and is sealed together
with tacky tape.
As the seal was airtight, it presented no problems. Therefore the vacuum seal was good and
the noise of the vacuum pump sounded good as it did not sputter like it would if the seal
was not defect. This formation was left overnight with the pump active for another
additional 4 hours until the resin had cured.
Figure 36: Final product achieved from

vacuum forming carbon fibre motor
support beam
The next day, the final product was made, and the bag was stripped off to achieve Figure
36.
The carbon fibre layup shown above was found to be too small as it was made for the initial
duct which had a lip cut out of the design to start with as shown in Figure 6. As a result of
this the process had to be completed again, but due to having already produced the 20
layer carbon fibre beam before, the process was easier and was finished quicker than the
previous beam, although the new motor beam specifications had to be increased to fit the
newly designed duct.
66
The newly designed carbon fibre motor beam is shown in Figure 37.
Figure 37: Final cutting of Carbon Fibre Motor Beam
10.2 TIP CLEARANCES OF THE DUCT

Figure 38 & Figure 39 show the tip clearances of the wind tunnel ducted fan rig.
Figure 38: Tip clearance of the wind tunnel ducted fan rig
67
Figure 39: Tip clearance of the wind tunnel ducted fan rig from a different angle
68
15009425 Further Experimentation and Theoretical Analysis of Thrust Characteristics using Ducted Fan Theory for the UWE UAV Development Group Project
10.3 GANTT CHART
69
10.4 AMENDED GANTT CHART
70
10.5 RISK ASSESSMENT
EPOXY RESIN & FIBRE REINFORCEMENT (500g)

ACTIVITY RISK ASSESSMENT
Describe the activity being assessed: Epoxy Resin & Fibre Reinforcement. Assessed by: Doug Nash Endorsed by:
The mixing of two liquid chemicals a)Epoxy resin, b)Hardener to produce a hard thermoset
plastic. Quantities not greater than 500g of mixed resin to be used.
Using Carbon, Kevlar, Glass (and other) fibre mat or cloth to reinforce the plastic.
Who might be harmed: Technician, Academic, Students. Date of Assessment: January 7 2013 Review date(s): January 2014
25
How many exposed to risk:
Hazards Identified Existing Control Measures S L Risk Additional Control Measures S L Risk By whom and Date
Level by when completed
(state the potential harm) Level
Inhalation. Overexposure to Adopted ventilated work space. 1 3 3 Proprietry ventilated work space constructed 1 1 1
vapours from resin or for composite use.
hardener may cause Training notes and instruction information given
headache, nausea. prior to activities.
Skin contact. Personal Protective Equipment PPE to be worn at 2 2 4

all times during activities.
Resin
These include Lab coat or apron, nitrile gloves,
May cause sensitization by eye goggles.
skin contact
Training notes and instruction information given
Eye contact Eye irritant upon prior to activities.
direct contact.
71
Ingestion. Swallowing these Training notes and instruction information given 2 2 4

chemicals presents a health prior to activities.
hazard.
Cutting reinforcement fibre Use the correct shears for the material being cut. 2 2 4 Use a down draft table when cutting 1 1 1
cloth and mat. composite fibre.
Refer to Trimming Knives and Scalpels training
Cut fingers. notes.
Wear a fine filter mast when cutting composite

materials.
72
10.6 RISK MATRIX: (To generate the risk level).

Very likely 5 10 15 20 25
5
Likely 4 8 12 16 20
4
Possible 3 6 9 12 15
3
Unlikely 2 4 6 8 10
2
Extremely unlikely 1 2 3 4 5
1
Likelihood (L) Minor injury – No first aid Minor injury – Requires First Injury - requires GP treatment Major Injury Fatality
treatment required Aid Treatment or Hospital attendance
1 2 3
Severity (S) 4 5
ACTION LEVEL: (To identify what action needs to be taken).
73
POINTS: RISK LEVEL: ACTION:
1–2 NEGLIGIBLE No further action is necessary.
3–5 TOLERABLE Where possible, reduce the risk further
6 - 12 MODERATE Additional control measures are required
15 – 16 HIGH Immediate action is necessary
20 - 25 INTOLERABLE Stop the activity/ do not start the activity
74
10.7 ETHICS CHECKLIST
Faculty of Environment & Technology

Faculty Research Ethics Committee (FREC)
Ethical Review Checklist for Undergraduate and
Postgraduate Modules
Staff and PG research students must not use this form, but should instead, if
appropriate, submit a full application for ethical approval to the Faculty Research
Ethics Committee (FREC).
Please provide project details and complete the checklist below.
Project Details:
Module name Individual Project BEng 17sep_1
Module code UFMFX8-30-3
Module leader Wendy Fowles-Sweet
Project Supervisor Jason Matthews
Proposed project title Further Experimentation and Theoretical Analysis of

Thrust Characteristics using Ducted Fan Theory for the
UWE UAV Development Group Project
Applicant Details:
Name of Student Julian Stephens
Student Number 15009425
Student’s email address Julian2.Stephens@live.uwe.ac.uk
75
CHECKLIST QUESTIONS Yes/N Explanation

o
1. Does the proposed project involve No If the answer to this is ‘No’ then no
human tissue, human further checks in the list need to
participants, animals, be considered.
environmental damage, or the
NHS.
Are human participants involved? If No

so, then answers question 2, 3 and
4.
2. How will a participant be asked to

give consent to take part in the
research and informed about how
data collected in the research will
be used?
3. How will a participant be informed

of their right to withdraw at any
time (prior to a point of “no return”
in the use of their data).
4. How will you explain the measures

in place to provide confidentiality
for participants and ensure secure
management and disposal of data
collected from them?
5. Does the study involve people who No
are particularly vulnerable or unable

to give informed consent (eg,
children or people with learning
difficulties)?
76
CHECKLIST QUESTIONS Yes/N Explanation

o
6. Could your research cause stress, No
physical or psychological harm to

humans or animals, or
environmental damage?
7. Could any aspects of the research No
lead to unethical behaviour by

participants or researchers (eg,
invasion of privacy, deceit,
coercion, fraud, abuse)?
8. Does the research involve the NHS No
or collection or storage of human

tissue (includes anything containing
human cells, such as saliva and
urine)?
Your explanations should indicate briefly for Qs 2-4 how these requirements will be
met, and for Qs 5-8 what the pertinent concerns are.
• Minimal Risk: If Q 1 is answered ‘No’, then no ethics approval is needed.
• Low Risk: If Q1 is answered ‘Yes’ and Qs 5-8 are answered ‘No’, then no
approval is needed from the Faculty Research Ethics Committee (FREC). However,
your supervisor must approve (a) your information and consent forms (Qs 2 & 3)
and (b) your measures for participant confidentiality and secure data management
(Q4).
• High Risk: If any of Qs 5-8 are answered ‘Yes’, then you must submit an
application for full ethics approval before the project can start. This can take up to
6 weeks. Consult your supervisor about how to apply for full ethics approval.
Risk Assessment: Separate guidance on risk assessment can be found on UWE’s

Health and Safety forms webpage at https://go.uwe.ac.uk/RiskAssessment. If needed,
77
you must complete a Risk Assessment form. This must also be attached to your
application for full ethics approval if your project is High Risk.
Your supervisor must check your responses above before you submit
this form.
Submit this completed form via the Assignm ents area in Blackboard (or
elsewhere if so directed by the module leader or your supervisor).
After you have uploaded this form, your supervisor will confirm it has been
correctly completed by “marking” it as Passed/100% via the My Grades link on the
Blackboard.
Further research ethics guidance is available at

http://www1.uwe.ac.uk/research/researchethics
78
10.8 DISSERTATION PROJECT RESOURCES FORM
Name: Julian Stephens Student ID: 15009425

Supervisor: Jason Matthews
[If you are carrying out a desk-based study and/or are only using the library or ITS facilities available
at UWE, the faculty’s risk assessment procedures will sufficiently cover your activities. In these
circumstances, you should identify this on your risk assessment form, which you should attach -
along with this form - to your Interim Research Proposal.]
Please sign below to confirm your request for resources.
Student: ……… ……………………………………………
Supervisor: …………………………………………
Further Resources – what else will you need?
R-Block workshop □ Laser cutting □
N-Block workshop □ Large format printing □
3D printing □ Miscellaneous □
Labs:
Structures □ Engine test bays □
Materials □ Robotics □
Composites □ Dynamics □
Hydraulics □
Wind tunnel/aero □
Car/flight simulator □
79
If you have ticked any of the above, please fill out a risk assessment of your activities.
Have you spoken to a technician? Yes □ Not yet □
Materials - outline the materials required for your project:

(Material) (Est. quantity)
Composite Material Few sheets
Carbon fibre Few sheets
Fibre glass Couple boards
Epoxy Resin About 500ml
80
10.9 CONTACTS REGISTER
Individual Engineering Projects

Contact Register
This register should be used whenever contact takes place between student and supervisor. This will ensure
all discussions are recorded, and actions clearly identified. Signatures can be added at meetings for records
made of phone calls, etc.
It is the student’s responsibility to keep this register up to date.
The Register should be attached as an appendix to the:
Interim Report (if required for the student’s module)
Final Report
(Extend the boxes as required per contact, and copy further boxes if the number of contacts exceeds the boxes
shown here).
Contact Type Date: Time: Place: Key Actions Agreed:
(Meeting, Phone 23/03/17 1:00pm – 1N22 Spoke to Jason about possible
Call, etc): 1:30pm dissertations relating to aerospace
Meeting and design, came up with idea to
talk with Steve about ducted fans
project.
Student Signature: Supervisor Signature:

(Meeting, Phone 22/04/17 6:40pm N/A Asked Jason about the type of
Call, etc): research I needed to gather over
Email the summer.

(Meeting, Phone 25/04/17 8:54pm N/A Jason emailed me to say that the
Call, etc): project has evolved and would like
Email to book another meeting with me
81
Julian Stephens 15009425

(Meeting, Phone 23/05/17 12:00pm – 1N65 Discussed further plans for project,
Call, etc): 12:30pm where to take the project, what I
Meeting will be doing, wind tunnel testing,
rig making etc.

(Meeting, Phone 19/06/17 7:58pm N/A Jason sent Jack Ingleby’s
Call, etc): dissertation to have a look at over
Email the summer and get a better
understanding of the project.

(Meeting, Phone 28/09/17 10:30am 1N65 Met up with Jason to discuss
Call, etc): further details about what to write
Meeting for the project proposal.

(Meeting, Phone 22/10/17 9:17pm N/A Jason asked how interim report
Call, etc): was going and discussed further
Email meetings.
82

(Meeting, Phone 09/11/17 12:30pm – 1N22 Met up and took notes on what to
Call, etc): 13:00pm put in the interim report, and the
Meeting aims and objectives to include.

(Meeting, Phone 06/12/17 13:20pm N/A Jason emailed back having
Call, etc): discussed the ducted fan project
Email with Zac, stating that a new rig and
mount would have to be
constructed as Jack’s design needs
to be used for future projects

(Meeting, Phone 11/12/17 19:05pm N/A Jason emailed back saying that a
Call, etc): new pivot point mount needs to
Email be designed for use in the wind
tunnel..

(Meeting, Phone 30/01/18 12:00pm – 1N22 The proposed initial wind tunnel
Call, etc): 13:30pm ducted fan rig was shown to Jason
Meeting and key things were noted of what
changes should be made including
the pivoting point and the
materials the rig should be made
from.
83

(Meeting, Phone 14/02/18 12:00pm – 1N22 Met up with Jason to discuss
Call, etc): 13:00pm manufacturing methods and the
Meeting design.

(Meeting, Phone 19/02/18 13:00pm – 1N56 Met up with Geoff Rowley to
Call, etc): 13:30pm discuss how the duct should be
Meeting made, and gave the dimensions to
him to make it.

(Meeting, Phone 20/02/18 13:19pm N/A Jason contacted to say that the
Call, etc): CNC Hot wire Cutter was unable to
Email cut the 10° required, and had to
be dropped down to 5°.

(Meeting, Phone 23/02/18 11:30am – 2R017 Met up with Jason to discuss
Call, etc): 12:15pm progress of the project, and to
Meeting confirm that everything is on track.
Also discussed that Geoff should
be notified of the new design for
the duct so it could be
manufactured.
84

(Meeting, Phone 23/02/18 11:30am – 1N56 Met up with Geoff Rowley to
Call, etc): 12:15pm discuss the new duct for
Meeting manufacture. Geoff said it will be
done and to check back on
Tuesday.

(Meeting, Phone 28/02/18 23:01pm N/A Jason was notified of the printed
Call, etc): Duct model and that it was in
Email 1N56. Also the final refined CAD
Model was sent to Jason so that
he could see the design.

(Meeting, Phone 20/03/18 1:00pm – 1N22 The manufacturing process was
Call, etc): 2:00pm discussed with Jason and the parts
Meeting that had already been made were
shown. The agreement was that a
wind tunnel session be booked as
soon as the manufacturing was
complete..

(Meeting, Phone 29/03/18 2:00pm – 1N22 The wind tunnel results and the rig
Call, etc): 2:45pm were shown to Jason and the
Meeting report outline was discussed. Key
pinpoints for the sections scientific
argument and
85
evaluations/conclusions were
discussed.
86
View publication stats

Literaturne Review

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Literaturne Review

Uploaded by

Copyright:

Available Formats

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

Further Experimentation and Theoretical Analysis of Thrust Characteristics

Thesis · April 2018

The user has requested enhancement of the downloaded file.

Figure 1: Ducted Fan under CFD

Student Number: 15009425

Award Programme: BEng Aerospace Engineering

Module: UFMFX8-30-3 Individual Project BEng

Project Supervisor: Jason Matthews

Word Count: 11624 Page Count: 53 Pages

TABLE OF CONTENTS...................................................................................................................................... III

TABLE OF FIGURES .......................................................................................................................................... V

TABLE OF TABLES ........................................................................................................................................... VI

1.1 AIMS AND OBJECTIVES .......................................................................................................................... 8

2 LITERATURE REVIEW .............................................................................................................................10

2.1 NOISE EMISSIONS ................................................................................................................................ 12

3.1 EQUIPMENT REQUIRED ....................................................................................................................... 17

4.1 RETESTING OF JACK’S DUCT ................................................................................................................ 39

4.3.2 Influence of Wind Tunnel Mounting Rig ......................................................................................... 49

5.1 PREVIOUS DESIGN IMPROVEMENTS ................................................................................................... 50

6 ANALYSIS AND DISCUSSION ..................................................................................................................54

7.1 WIND TUNNEL REMARKS ........................................................................................................................... 56

10.1 CARBON FIBRE MOTOR BEAM ............................................................................................................. 64

1.1 AIMS AND OBJECTIVES

• Using MatLab to programme functions, generating useful graphs and data to

These preliminary tasks will include:

• Testing of the previous ducted fan model produced by Jack Ingleby

A second characteristic of ducted fans is their ability to provide a supplementary safety

Figure 3: Ducted fan under cruise condition (F.M. and

2.1 NOISE EMISSIONS

2.2 LONGITUDINAL STABILITY

The pitching moment coefficient can be defined as:

Where q is the dynamic pressure and equal to:

Where AR is the aspect ratio of the wings as shown in Equation 3:

Where b is the wing span and Sw is the wing area.

2.3 BRIEF HISTORY AND OPTIMISATIONS TO SHROUDED PROPELLERS

• Flow is assumed to be incompressible.

• Flow is assumed to be inviscid, so viscosity effects of flow are ignored/negligible.

• Flow is assumed to be uniform with no losses in drag or frictional drag.

• The propeller is assumed to be a disc that is infinitesimally thin with an infinite

• The propeller offers no air resistance to air passing through it

2.4 PEREIRA’S RESEARCH AND DESIGN OPTIMISATION

Blade tip Clearance: 𝛿𝛿𝑡𝑡𝑡𝑡𝑡𝑡 = 0.1%𝐷𝐷𝑡𝑡

Inlet Lip Radius: 𝑟𝑟𝑙𝑙𝑙𝑙𝑙𝑙 = 13%𝐷𝐷𝑡𝑡

Diffuser angle: 𝜃𝜃𝑑𝑑 = 10°

Diffuser Length: 𝐿𝐿𝑑𝑑 = 50%𝐷𝐷𝑡𝑡 𝑡𝑡𝑡𝑡 72%𝐷𝐷𝑡𝑡

Figure 4: Duct Parameters, (Pereira, 2008)

3.1 EQUIPMENT REQUIRED

3.2 CAD DESIGNS

3.2.1 THE DUCT

Blade tip Clearance: 𝛿𝛿𝑡𝑡𝑡𝑡𝑡𝑡 = 0.1%𝐷𝐷𝑡𝑡

Inlet Lip Radius: 𝑟𝑟𝑙𝑙𝑙𝑙𝑙𝑙 = 13%𝐷𝐷𝑡𝑡

Diffuser angle: 𝜃𝜃𝑑𝑑 = 10°

Diffuser Length: 𝐿𝐿𝑑𝑑 = 50%𝐷𝐷𝑡𝑡 𝑡𝑡𝑡𝑡 72%𝐷𝐷𝑡𝑡

Figure 5: Duct Parameters, (Pereira, 2008)

Figure 6: Initial Ducted fan Configuration based on Pereira's data results

Figure 7: Isometric Projection of initial duct design

Figure 8: Top view of initial duct design

3.2.2 THE AMENDED DUCTED FAN RIG

3.2.3 INITIAL DUCTED FAN THRUST RIG

Figure 9: Isometric view of initial ducted fan rig

3.2.3.2 Motor support beam and secondary support beam

3.2.3.3 Supporting Brackets

3.2.3.4 Support L clamps