IC349 Integrated Manufacturing Project IPD 01

The Hong Kong Polytechnic University










Human Powered Flying Device

Chan Chun Yiu


Chan Yik Ting


Leung Kai Yeung


Mak Ho Ming


Kwok Pui Ki


Hui Shuk Ying


Poon Tsz Wang


Koo Lai Fong














1.  Introduction


     i)  Background


     ii)  Objectives


2.  Project Statement


     i)  Propeller


     ii)  Chain Drive


3.  Customer Requirement & Analysis


      i)  Propeller


      ii)  Transmission System


4.  Design Specification


5.  Conceptual Design


     i)  Brainstorming


     ii)  Possible solution


6.  Detailed design


7.  Final Design


8.  Bill of Materials


     i)  BOM - Overall Design


     ii)  BOM - Transmission System


     iii)  BOM - Bottom Bracket Transmission System


     iv)  BOM - Training Wheels


9.  Project Expense         


10.  Material Selection


11.  Manufacturing Process


      i)  Propeller


      ii)  Fuselage


      iii)  Transmission System


12.  Manufacturing Technique


13.  Manufacturing Process Selection


      i)  Propeller


      ii)  Transmission System


      iii)  Fuselage


14.  Problem Encountered


      i)  During Manufacturing Stage 


      ii)  During Assembly Stage


15.  Project Management


      i)  Work Contribution


      ii)  Communications


      iii)  Timeline and Gantt Chart


16.  Results


17.  Improvement


      i)  On the aircraft


      ii)  On our teamwork     


18.  Further development   


19.  Conclusion


20.  Appendix


1.   Introduction                                 
Prepared by Chan Yik Ting 10561900D


i)  Background


An Aircraft can fly firstly to overcome the drug due to movement in air and the gravity due to the weight of structure materials and the pilot. In other words, flying is basically involved an enough pulling force (thrust) to overcome the drag and a lifting force to overcome the gravity force. In existing artificial power assisted airplanes, they usually equipped with powerful jet engines in order to provide pulling force to cancel out the effect on the weight and the frictional force of air.


The first Human Powered Aircraft (HPA) was developed in 1923. After that, there are many similar projects, which have been carried on over the world in the previous century. The first officially authenticated take-off and landing HPA was made in 1961 by Southampton University.


HPA commonly faces a challenge of insufficient power to take off and sustain flying in the air. To accomplish these challenging issues, the aircraft must be light enough in the beginning, and the lifting force provided by wings must be sufficient to raise the aircraft and the pilot.


Universities in other Countries have studied the topic years ago and the world record is still holding by Kanellos Kanellopoulos, recognized by FAI, which was a straight distance of 115.11km (71.53mi) in 3 hours, 54 minutes. There are many teams from different universities attempted to break the record and many references resources are shared on the platform. Human-Powered Aircraft is a hot topic for engineering student in other countries and we think this project is a good opportunity for us to communicate and share the knowledge with other university students. NASA and other related website are really good platform for us to learn from other past experience.







ii)  Objective of the manufacturing project       Prepared by Chan Yik Ting 10561900D


In course IC349, we are going to complete an integrated manufacturing project, which aims at applying and integrating the engineering knowledge from the previous course IC348 and experience the practical manufacturing process during the course IC349.


During the training period, we have to go through all the stages in the project. The stages are Design and Drafting, Costing, Project Planning and Control, Manufacturing, Assembly, Testing and Evaluation.

The project provides opportunity for student to develop their personal and professional qualities such as leadership, communication skill, co-operative attitude, and co-ordination ability as well as enthusiasm for accepting technical responsibility.

At the very beginning of the project, the existing basic theoretical knowledge of aerodynamics was studied and the most efficient human powered thrusting mechanism was selected. In the following, we can gain the practical experience of aircraft engineering through building prototype of real human powered aircraft for testing and evaluation.

Since we are going to perform on the project “Human Powered Flying Device”, we have to overcome several criteria for flying aircraft such as the gravity from the pilot and the whole aircraft and the frictional force against the air. After the research stage, we come up the main objective of manufacturing a human-powered flying device which is capable to fly with the following requirement:


1.     Flying 1m from land for at least 30 second

2.     Flying with at least 6ms-1 in order to give enough draft power

3.     The weight of the frame doesn’t excess 40kg

4.     Driving force of the transmission 






2.  Project Statement                                     Prepared by Leung Kai Yeung


i)        Propeller


Propeller location

We have designed the propeller located at the front of the aircraft. There are benefits and limitations with these designs from an aerodynamics and mechanical (drive train) point of view.


The tractor configuration allows the aircraft to fly in undisturbed air, so the propeller efficiency is high, and the drag is lower. However, the rest of the aircraft flies in turbulence from the propeller wake. The drive train for a tractor configuration has the potential to be mechanical simple as it can be designed to be vertical, which also minimizes weight. A need for fewer components may also increase the efficiency of the drive train.


A propeller in the pusher configuration has a reduced efficiency as it works with the disturbed airflow from the fuselage, wing and tails. However, the skin friction drag of the aircraft is reduced as it flies in undisturbed air. Die to the location of the propeller, this configuration requires larger tails and a larger rear weight. The propeller shaft is also longer (and therefore heavier) than the tractor propeller shaft. Additionally, there is a risk of the propeller touching the runway during takeoff as the aircraft noses up. This is an important consideration as the propeller radius is very close to the fuselage height. It therefore needs a greater clearance and longer landing gear compared to a tractor and the additional structure for this may increase the weight and drag of aircraft.











Propeller Optimization

Figure 1.3: Thrust and propeller rpm vs. propeller radius for the E193 airfoil with Cl=0.5




Fx 60-100


Cl = 0.5

Efficiency (%)

Thrust (N)





Cl = 0.5

Efficiency (%)

Thrust (N)





Table 2: Maximum propeller efficiency and corresponding thrust for the FX60-100 and E193 airfoils at

        Cl 0.5 and 0.7.


The corresponding chord and twist angle is shown in Figures 1.1 and 1.2. The chord of the propeller is quite narrow, about 9cm at the widest point. This gives a maximum thickness of about 9mm (the thickness of the E193 airfoil is about 10% of the chord as shown in Table 1), which is quite small. This seems too narrow to produce the required thrust so the shape should be verified with a different propeller analysis program.


Since this project is the first iteration of design, this is beyond the scope of the project and the propeller will be left as is. From a structural point of view, a spar is required to pass through the centre of the blade for stiffness and torsion rigidity. This is not possible for the first 0.2m of the radius. However, the inner part of the propeller to about 25% of the radius has a very small contribution to the thrust (Raymer 2006, p. 251), so the propeller chord from r=0 to 0.375m can be ignored, and tailored to be suited to the structural requirements of the blade. In its place, a spinner can be placed over this area to push the air out to where the propeller is more efficient. In piston engines, the spinner is not this size due to the need for an air intake for engine cooling.


For a human powered aircraft there is no engine, so the spinner can cover the full 25% of the propeller radius. However, having a spinner with diameter 0.75m is impractical as the placement of the propeller at the top of the fuselage, in front of the wing will produce a lot of drag. The greater efficiency gained by such spinner is outweighed by the drag it will produce. Hence, a more reasonably sized spinner will be used.




ii)   Chain drive


Chain weight

As we chose to use the chain transmission system, it has a certain amount of weight. In order to reduce the effect cause by the weight of the chain, we have to increase the size of wings or increase the speed of the propeller. However, the powered generated by human has a limitation; the wings size should be increased. There has a technical problem cause by the long wings. During flight, there is up - thrust force acting on the wings. They will bend upward. If the materials used are not tough and strong enough, wings will be broken. In a final decision, we decided to reduce the materials used in fuselage and some other parts such as seat and wheel.


Tension cause by the chain

During rotation of the chain, there is great tensional force acting on the shaft. This is acting the shaft downward. If the welding process is not good enough, the fuselage will be broken during the flight.





3.  Customer Requirement & Analysis    Prepared by Leung Kai Yeung 10194550D


i)  Propeller



Assume :                

rpm = 200

                   forward speed= 8m/s


Rotational Speed= 2∏r x rpm x 60

              = 2∏r x200 x 60


tan Ө = Forward speed/ Rotational speed

   = Ө + 15


Given:   Angle of Attack = 15

Twisting angle of the each cross section


As a fixed wing aircraft was chosen for the overall configuration, a propeller is chosen for thrust generation. The main requirement of the propeller is to produce thrust that is greater than drag of the aircraft. Due to the low amount of power available to the propeller, it should be design to be as efficient as possible. However, the propeller is limited by the diameter, weight and manufacturing constraints.


Number of propeller blades

Theory states that propeller efficiency increase with blade number. However, this does not account for interference between the blades as the blade number increases, so two blade propellers are more efficient due to the less disturbed flow. Additionally, a two bladed propeller will achieve minimum weight, which is desirable. These reasons are probably why all previous successful human powered aircraft have used two bladed propellers. Therefore, a two blades propeller will be used for this design as well.



The propeller should be optimized to produce the greatest thrust for the given input power watts at the design flying speed of 11.9m/s. the final propeller design requires the determination of a radius, rpm, twist and airfoil profile for the design point.


Two potential airfoils are investigated: the E193 and the FX60-100. The E193 air-foil has been used for the propeller for the Monarch and the Gossamer Albatross. The FX60-100 was designed specially for human power and was used on the Velair. Thus, it can be assumed that the properties of these airfoils at low rotational and forward speeds are best suited to human powered aircraft. The characteristic of the airfoil is summarized in Table 1.











LE radius



TE angle



Table 1: Airfoil characteristics of the E193 and FX60-100

   (Barnhart et al. 2004, p. 30).


The minimum induced loss propeller design for human powered vehicles by E. Eugene Larrabee (Larrabee 1984, p. 9-11) is used to determine these optimum parameters. The theory has been implemented in a design excel sheet provided by the Royal Aeronautical society. This spreadsheet requires the input of propeller radius, blade number, flying speed, rpm, power and airfoil data (Cl, Cd and Cl0), and give the performance, optimum chord along the radius and propeller twist. Since there are three unknown inputs (propeller radius, rpm and airfoil data) an iterative approach was taken to determine the best propeller blade.


The airfoil section data was determined from Javafoil, for a range of Reynolds numbers. In propellers, the design lift coefficient is usually around 0.5 (Raymer 2006, p. 379) so this value is initially chosen. The Monarch, Goassamer Albatross and Velair aircraft had a design lift coefficient of 0.7, and the Daedalus had a lift coefficient of 0.8, so an optimization for a lift coefficient of 0.7 was also performed.


For each airfoil, the optimum rpm for a fixed radius was determined, and the corresponding efficiency and thrust was recorded. It was ensured that the appropriate airfoil section data was used to be similar to the reference Reynolds number (the Reynolds number at 75% of the radius of the propeller). The optimum rpm decreases with increasing propeller radius. This trend is also observed with the FX 60-100 airfoil. It can be seen that the greatest thrust occurs at a radius of about 3 meters which is impractical for the final design. The thrust produced increases with increasing radius prior to the optimum 3 meters mark, so the largest radius allowable should be chosen. Therefore, a propeller radius of 1.5 meters was chosen for analysis for the E193 with Cl=0.7 and for the FX60-100.


For the fixed radius of 1.5m, the airfoil with lift coefficient of 0.7 had better performance than the airfoil with lift coefficient of 0.5 for both airfoils. The desired lift coefficient is achieved by placing the blade at the corresponding angle of attack. The E193 airfoil with Cl=0.7 exhibited the best efficiency as seen in Table 11. Thus, this airfoil is chosen for the propeller.



The tip speed of the propeller should be below sonic speed. For the design point of 180 rpm, the tip speed is V tip = Пx n x d=60 = 28.27m/s, which is clearly less than the speed of sound (340 m/s). Thus, the propeller tip will always be kept below sonic speed.

Figure 1.1: Optimal propeller chord along the propeller radius.


Figure 1.2: Optimal propeller twist along the propeller radius.


The helical tip speed of a propeller should be limited to about 700fps to reduce noise. For the human powered aircraft, V tip helical =√(V 2 tip + V 2) = 30.68 m/s = 100.6 fps which is within the specified limits


ii)  Transmission system


There are three main methods of transmitting power between rotating bodies: belt drives, chain drives and gear systems. Finally, we chose to use chain drive.


Chain drive

Chain drives are used over a shorter distance than belt drives and must generally transmit power in the same plane {Hence, no twisting is possible. Although chain drives typically have a mechanical efficiency greater than 97%, they must not only be oriented in the same plane, the driven and driving sprockets should not be vertically aligned. These limitations mean that it is not practical to fit a chain drive to the pilot's pedals.



 A twisted chain drive is chosen to transmit the pilot's power to the propeller. The lower weight of a chain drive is considered to be a greater benefit than the slightly higher mechanical efficiency of a gear driven system. A chain is extended from a driver pulley of radius R (attached directly to the pedal shaft) and twisted through 90o to turn the driven pulley (attached directly to the propeller shaft). In order to achieve the desired speed 200 rpm, the driven pulley must have radius 0.5R. From Funk (1996, Table 31.3) the smallest standard pulley sizes for at-belt drives are chosen to meet this radius relation.


The propeller shaft pulley is to have a diameter of 10 mm and the driver pulley a diameter of 17 mm. To keep the weight of the system down, wooden pulleys will be used and a mineral-tanned leather belt will be used to maximize the coefficient of friction (Blazewicz, p. 50). The pedals will rotate at a radius of 180mm (based on a survey of bicycle pedals) so the given pilot power of 225W implies a torque of 23.9Nm. A mechanical efficiency of 97% combined with the pulley ratio of 1:2 means that the propeller shaft will be turned with torque of 11.6Nm. This is sufficient to drive the chosen propeller.


Power efficiency - Available power

The human muscle converts chemical energy obtained from the oxidation of carbohydrates and glycogen and the conversion of fatty acids into CO2 and water, into mechanical energy at an efficiency of approximately 20-25% (Wilkie, 1960). This means that for a total metabolic work of 100 watts, approximately 76 watts go to body heat, leaving 24 watts of usable mechanical power (Bussolari & Nadel, 1989). Figure 1.31 shows that for longer durations of exercise, the constant power output levels out. It was found that for the long duration human powered fight of the Daedalus, the maximum power per kg of human was approximately 3.5 W/kg. Basing the design on a 68kg pilot, and allowing for a reduction in total output power due to the required maneuverability, it was found through the matching diagram that the design W/P ratio is 0.44 kg/W, resulting in 225 Watts of power produced.


  Figure 1.31: Maximum output of mechanical power VS total duration of exercise for various types of exercise: x- cycling, + rowing, - Δrunning uphill, *- cycling and turning a hand crank (the around a symbol indicates performance by a champion athlete) (Wilkie, 1960)


The pedal cycle configuration is a mechanically simple design which can be adjusted for different durations, and can achieve the best use of kinetic energy of the moving legs. Whilst power is only obtained through leg movement, the muscle mass of the legs is large enough to utilize all of the oxygen which can be absorbed therefore making it more efficient over longer durations than exercise configurations which use both the arms and legs such as rowing, or cycling and turning a crank shaft.


The rowing configuration involves a sliding seat, and utilizes both arms and legs to create mechanical work. This is shown to be a good method of creating mechanical work for durations longer than 3 minutes, however it requires the use of a larger muscle mass for a similar power output to cycling, and therefore has a lower mechanical conversion efficiency. In a human powered aircraft it is also required that the pilots have their arms available for controlling the steering of the aircraft around the turns and for controlling level fight stability. The Muscular projects required that the pilot be almost completely immobile above the hips to maintain precise control of the aircraft (Schoberl), therefore the rowing configuration would be ineffective for the required fight maneuverability. The chosen configuration, therefore, is the pedal cycle for converting human mechanical power.



Power efficiency - Recumbent vs upright pedal

Design of the Marathon Eagle (Bliesner, 1994) has suggested that the upright position is more efficient for obtaining peak power levels, however the recumbent cycling position is desirable for minimum aerodynamic drag and the lower aspect ratio is less prone to losses due to sideslip. The recumbent position also allows for more upper body freedom of movement to use the fight controls (Bussolari and Nadel, 1989). Investigations into the on road cycling configurations of human powered vehicles have compared the standard upright position to the aerodynamic recumbent position. Figure 1.32 indicates that in order to achieve the same on road speed, the recumbent human powered vehicle required a much lower power/weight output than the conventional cycle designs, whilst also providing greater back support and therefore increased ride comfort for longer duration use.

Figure 1.32: Power requirements for touring bicycle, upright and crouched racing bicycles and

            Vector recumbent human powered vehicle (Hennekam, 1990)












4.   Design Specification                Prepared by Chan Yik Ting 10561900D


Upon the research we made in the previous stage and we come out the design specification for the human powered flying device as follow:




Power Source

Human Powered

Aircraft Weight

Below 50kg

Propeller Efficiency

Approx. 20N

Ground Speed



Aluminium, Teflon, wood, thin plastic sheet


The human  powered aircraft should be made of  light, stiff and common materials in order to make it possible to fly above the ground. Aluminium doesn’t break easily and it has certain high strength, which is ideal material for making light and inexpensive components and structure of the whole aircraft. Teflon  is another light material that is excellent for making spacers due to its high thermal expansion and low frictional coefficient. Wood, thin plastic sheet were used to manufacture the propeller. The output driving force generated by the propeller should be around 20N in order to gain enough forward force to raise 50kg aircraft and 50kg pilot to the sky by forwards speed not lower than 6ms-1.

















5.  Conceptual Design


i)   Brainstorming                                          Prepared by Leung Kai Yeung

We have totally 6 parts of consideration. Half of they are hardware and the other parts are the improvements. Hardware includes propeller, the fuselage and the transmission system. The rest of the parts are the manufacturing method, materials used and consideration.


- Propeller

- Spinner rotation

- No. of blades

- Cross section of blades

- Sixe

- Twisting angle



- Frame (outlook)

- Dimension

- Orientation of seat

- Weight

- Wheels - no. of wheels

        - no. of teeth of wheel



- Aluminium

- Plastic

- Steel

- Wood


Transmission system

- gear - no. of gears

     - gear ratio

- Shaft system

- Chain system

- Belt system

- Power efficiency

Manufacturing method

- welding

- Lathe

- Milling

- RP

- CNC - punching

      - Milling and lathing


- Light in weight

- Safety

- Budget< $1500

- Strength

- Durabilit



ii)   Possible solution                       Prepared by Leung Kai Yeung, 10194550D




Conceptual design 1


The whole frame size is large and hence we decided to use aluminum to reduce the weight. As aluminum is light enough which can counter the large size frame. The pilot posture is upright position. In order to have a more comfort travelling, upright position is the most comfortable posture. However it has the lowest aspect ratio, which would increase the energy lost.


Conceptual design 2


The whole frame is made of carbon fiber which is light in weight and strong. The reason of using carbon fiber is due to the recumbent cycling position. Recumbent position could generate the greatest speed and hence carbon fiber could tolerate the greater speed of the aircraft.

Transmission system


Chain drive


Figure 1 chain drive sketch


Chain drives are used over a shorter distance than belt drives and must generally transmit power in the same plane {Hence, no twisting is possible. Although chain drives typically has a mechanical efficiency greater than 97%, they must not only be oriented in the same plane, the driven and driving sprockets should not be vertically aligned. These limitations mean that it is not practical to fit a chain drive to the pilot's pedals


Belt drive


Figure 2 belt drive sketch


Belt drives can transmit power across the greatest distance and some angle of twist can be introduced into their motion, allowing power transmission between non-parallel axes. They typically have a mechanical efficiency of around 97%. A belt drive driven by the pedals could extend to the top of the fuselage, twisting by 90o to an orientation where it can then turn the propeller shaft (as sketched in Figure 4.22). This design has the advantage of simplicity and, therefore, light weight.


Shaft gear drive


Figure 3 shaft gear drive sketch


Gear drives have the highest mechanical efficiency (around 99%) but driven and driver gear must be touching. This means that to transmit the pilot's power from the pedals to the propeller shaft, for example by an intermediate hypoid bevel gear as shown in Figure 4.23, the intervening space must be bridged by a shaft. A gear drive will, therefore, be heavier than either a belt or chain drive, as will the fastenings required to hold the shafts in position.




DAE 51


6.  Detailed Design                         Prepared by Hui Shuk Ying 10563584D


We have to think many designs in the design stage; we picked three designs to use the screening method to help us to choose the better one to manufacture out.

The screening is showed below:

     Design 1                                         Design 2                                    Design 3

Base on the above screening of the conceptual design, the main criteria of our aircraft are light in weight and the safety. As we can see, the design 3 is fair in these two criteria; design 1 and design 2 are good in the safety criteria. In the light in weight item, the design 1 is good but the design 2 is bad.


Therefore, the screening method shows that the design 1 is the better design in those three designs.


Then we confirmed the design, we need to consider how to manufacture it by using 3D CAD software - Solidworks to modify the final design.

The following are the CAD models:


The following figure1 is our first design drew by Solidworks. As we can see, there have propeller in the front of the aircraft and create three wheels that to balance the fuselage. Two wheels are in the front and one at the back which have two advantages:

1.         Improved aerodynamics

2.         Enables light in weight of a small car powerplant readily





Figure 1








The transmission system is joined to the top front of the frame by using four pairs joints to connect caused we have sprocket at the back of the system can’t link to the frame directly.


We changed the design of the propeller due to the first design (figure 2) of it was too thin and we have to adjust the degree of the propeller. Hence, we have varied the design which shows in the figure 3.


The figure 3 designed wider and thicker than figure 2 that can along the wind and let it revolves smoothly. Also, let the propeller can simplifier the step to manufacture it.










We also re- design the wheels from three wheels to become two wheels which shows in the figure 5. Due to we need to reduce the load from the aircraft, we try to design those wheels like a chain bicycle’s wheels.

Besides, we can reduce the friction between the wheels and the floor and reduce the cost by changed design.