How to design a fuel-powered quadcopter with 3D printing

Transcription

1 How to design a fuel-powered quadcopter with 3D printing 1

2 Table of Contents Executive Summary Introduction Rotor Design Concept Design Detailed Design Prototype Results Conclusion References

3 Executive Summary The white paper will describe the process of designing a fuel-powered quadcopter with the use of 3D printing. The document starts off with a short introduction and is followed by four main parts of research, namely the rotor design, concept design, detailed design and prototype results. The rotor design touches on the momentum blade element theory and the beam theory where it is described how the design for the rotors was chosen and why. Concept design talks about the control subsystem, structure subsystem and the power generation subsystem. Detailed design is about choosing other components of the quadcopter, for example the engine and why they were chosen. The prototype results show data for the simulation and real test. Lastly, there is a conclusion along with recommendations for further design. 3

4 1. Introduction This white paper intends to develop a design methodology to manufacture custom model multi-rotorcraft rotors using 3D printing techniques available at Leapfrog 3D Printers. In accordance with the above statement, the following requirements for the design methodology will also be met. Firstly, the resulting design will be printable at Leapfrog 3D Printers. Secondly, the resulting designs will be able to withstand the operating conditions of similar model rotorcraft rotors. Lastly, the design methodology will be easy to adapt to rotors of different sizes. The white paper is divided into four main sections, namely rotor design, concept design, detailed design and prototype results. The section on rotor design explains how the rotors will be designed and why and it also touches on the momentum blade element theory and the beam theory. The concept design breaks down the following components of the quadcopter, namely the control subsystem, structure subsystem and the power generation subsystem. It is explained why the final designs have been chosen and how they influence the quadcopter as a whole. Detailed design touches on other components of the quadcopter, namely the rotor hubs, rotor blades, engine, transmission system and frame layout and why the final design for each component has been chosen. Lastly, the prototype results section shows data about the simulation and test results compared to a bigger quadcopter. 4

5 2. Rotor Design Momentum blade element theory Beam theory The methodology used to determine rotor s characteristics is based on a simplified version of momentum blade element theory by Ruijgrok (2009). In order to start the design process, the wanted thrust, rotor radius, blade count, lift coefficient distribution, airfoil shape and chord distribution have to be specified. Then, the momentum blade element theory can be used to calculate the rotational rate and blade twist distribution necessary to create the wanted thrust. Once these parameters are known, the aerodynamic coefficient distributions over the rotor blades can be calculated. Since the rotational rate is also known, it is possible to calculate the force distribution over the rotor. This makes it possible to calculate the drag moment acting on the rotor blades and calculating the necessary power to achieve the wanted thrust. Furthermore, the beam theory by Hibbeler (2008) can be used to calculate the internal forces acting on each rotor blade. Since the airfoil shape and twist angle distributions are known, the area moment of the blade s inertia can be calculated at every radius. By integrating the aerodynamic and centripetal force distributions, an accurate estimation of the maximum internal stresses on each radial segment of the rotor can be made. This in turn makes it possible to check if the rotors will survive the required operating conditions. Since the resulting airfoils have a curved and twisted shape, it is almost impossible to print them using an FDM technique because they lack a flat side which can be printed against the bottom plate of the printer. In order to solve this problem, we can transpose 5

6 each blade element ever so slightly in order for the top and the leading edge of the rotor blades to align over the entire radius of the blades. Figure 1: Example blade element alignment side views. Left: before correction, right: after correction. In order to test the design, several prototypes were made. The first prototype was unsuccessful. We attempted to print the complete rotor at once on the largest scale possible. A simple rotor hub was designed to connect the three rotor blades together and the rotor was supposed to be printed on the flat upper side. This failed due to the rotor warping upwards and detaching from the printer bed. 6

7 Figure 2: Unfinished rotor prototype. The second prototype was printed in multiple parts, since the first prototype failed. This also made it possible to print the parts on the leading edge. To test what works, two blades were made, one printed on the leading edge down and another one the same way as the first prototype. The rotor printed on the leading edge provided far superior results, which can also be seen in the figures below. This is due to the fact that the upside down prototype requires support material in order to create the correct shape, while the other component can be printed without any support material. A drawback; however, is that it does result in a slightly rougher bottom side of the airfoil since the printing surface is perpendicular to the blade surface. 7

8 Figure 3: Rotor blade example, printed upside down. Figure 4: Rotor blade example, printed standing on the leading edge. For the final prototype a simple joint was designed to fit all blades together. 8

9 Figure 5: Final prototype rotor example. 9

10 3. Concept Design Control subsystem Structure subsystem Power generation subsystem In order to create the most optimal design, it is important to investigate multiple technology directions and to do this, three different concepts will be investigated. It is important to split up the design into several, mostly independent subsystems since this will provide a good overview of the design space available for the concept generation. For the model multi-rotorcraft, three different subsystems can be distinguished, namely the control subsystem, the power subsystem and the structure subsystem. In terms of the control system, there are several possible options. Traditionally, the control of the model rotorcraft is achieved by altering the lift generated by the different rotors, which is done by changing their rotational rate. A more recent option is control by pitch adjustment. More fine-tuned control of the rotor is possible by using more complex variable pitch rotors (Cutler, 2010). The third option is controlling the rotorcraft by using deflection vanes right below the rotors. This is less effective than the two previously mentioned systems due to the limited torque generation. The other side of the control system exists in the controller itself. Even though variation is possible here, almost all commercial systems use the same approach. This uses a 6-Degree Of Freedom Inertial Measurement Unit (6-DOF IMU) in combination with a Global Navigation Satellite System (GNSS) like GPS and altitude measurement. By using simple controllers, a system like this allows complete and accurate control of the rotorcraft. 10

11 The structure subsystem controls the layout of the rotorcraft itself. With this system, there is a lot of variation possible between different rotor configurations, materials, and construction. The rotor configuration is one of the most defining aspects of a rotorcraft design. The most common choices are the X and H frame configurations. In former configuration, the rotors are located at the ends of several booms which extend from a main central body. This configuration is one of the most versatile since it can handle any amount of rotors. However, there are drawbacks in the limited space available in the center of the rotorcraft. This configuration is more interesting for systems, which have significant amounts of parts in the main body of the rotorcraft. While the center beam is free for mounting miscellaneous components, the rotors are located on the tips of the cross beams. The main drawback of this configuration is that it tends to be heavier than comparable X frames due to the amount of material, which is needed to keep the center frame sturdy. The final subsystem of the rotorcraft is the power generation subsystem, which can be further divided into energy storage, power generation and power transmission. Since almost all model multi-rotorcrafts are electricity based, this system traditionally consists of high-energy density batteries. These batteries can store and discharge energy very efficiently; however, their main drawback is the relatively low amount of energy they can store compared to systems, which rely on combustion. For endurance systems, this component tends to be responsible for majority of the rotorcraft weight. The power generation system is responsible for transforming stored energy into mechanical power. With electricity-based systems this happens through the use of lightweight brushless DC motors, which are controlled by Electronic Speed Controllers (ESCs). Another approach is using a small combustion engine to generate mechanical power directly from the combustion of the fuel. 11

12 Finally, a hybrid solution exists by using a fuel-powered engine coupled with a generator to power an electrical system. The transmission system is responsible for bringing this mechanical power to the rotors of the rotorcraft. Many models have the rotors mounted directly on top of the motors powering them; however, it can be an advantage to power all rotors from one central power source. In this case, a combination of shafts and belts can be used to connect the rotors directly to the motor. The layout of this system is closely intertwined with the chosen structure layout. Based on these possibilities, three different concept designs are chosen to be further investigated. The choice in the power generation system is considered to be the most important difference due to its impact on the rotorcraft s endurance. The first concept design is mainly a reference design, featuring a fully electrical drive system with one motor per rotor. This design is the closest to the current commercially available rotorcrafts, and is the simplest option to develop. The second concept design features a single combustion engine as the power generation system. This system features a mechanical transmission to power all the rotors, which means the control system cannot rely on the speed control to control the rotorcraft. To compensate for this, pitch control can be used to stabilize the rotorcraft. The final concept design features the hybrid power generation system. This system combines the simplicity of commercial model rotorcraft control systems with the endurance of fuel-based systems. Its major drawback; however, is the weight required for complex power generation system. In order to make an accurate comparison between different concept designs, accurate theory for sizing different rotorcraft models must be established. In terms of the electricity-based design, an estimation of the rotorcraft endurance can be made by solving the equations for total mass and power consumption using data of motor efficiency and battery energy density. Firstly, data collected from T-motor (2013) is 12

13 analyzed by correcting the produced lift and efficiency of the motor weight and the matching rotor. The result can be seen below. Figure 6: Net motor lift compared to motor efficiency of several brushless DC motors for model rotorcraft. Rotors used are the largest rotors described in the engine model datasheets. From this graph, we can conclude that the highest currently possible efficiency by using COTS technology is 14g/W. Further research shows that the maximum storage density of commercially available batteries has a maximum of approximately 0.2 Wh/g. If we assume there is constant power consumption in hover, we can use these two variables to calculate the maximum possible quadcopter endurance by solving the following equation: It is important to note that in this equation, the thrust is units of mass because the thrust data is provided in grams at sea level. By substituting the mass of the batteries and the thrust by power and endurance, we can write the following equation: 13

14 The final unknown is the mass of the frame and the payload over the mass of the batteries. If we assume the weight of the frame is negligible, we can calculate a maximum theoretical flight time of 2.8 hours independent of mass. This flight time quickly decreases as the frame and payload mass fraction increases. A practical number can be derived from the fact that the frame mass of model rotorcraft is often approximately equal to the mass of the batteries, which yields a practical flight time of 1.4 hours. The sizing of the second concept design is more difficult. Due to the motor and rotor performance data not readily available in this sector, estimations of performance will have to be made using mostly theoretical analysis of the rotorcraft. Firstly, it is necessary to generate a performance estimation of the variable pitch rotors this concept requires. By using the momentum blade element theory adapted from Ruijgrok (2009), we can provide an estimation of the thrust and power coefficients of rotors similar in size to the rotors used by the motors from the first concept. These rotors have an estimated radius of 350mm, a constant chord of 33 mm and a NACA0009 airfoil similar to the rotors of model helicopters (Cutler, 2010). By using XFOIL (Drela, 2013), the aerodynamic characteristics of this airfoil were derived at a tip Mach number of 0.3 and a Reynolds number of , which were estimated from operating conditions of similarly-sized helicopter rotors. By using the momentum blade element theory and assuming a two-bladed rotor with a 50mm radius rotor hub, this data is enough to calculate the values of the rotor power and thrust coefficients depending on the blade pitch angle and assuming hover conditions. With performance characteristics of the rotors being known, it is then possible to estimate the endurance of the resulting rotorcraft design itself. By substituting the thrust required by the weight of the vehicle as well as accounting for the presence of multiple rotors, we gain the following relation: 14

15 This relation indicates that for minimum power consumption, the ratio maximized. For the used rotors, the ratio is equal to must be We can then calculate the endurance of the vehicle by solving the mass of the vehicle. Because the mass change of the vehicle is caused by the fuel being burned and accounting for the efficiency of the power conversion, we can write the following equation: This is the first order nonlinear differential equation and can easily be solved. The boundary conditions of the equation have to be calculated first in order to solve the system. The first boundary condition can be derived from the fact that at the end of the flight, the mass of the system equals the empty mass of the system. The other boundary condition follows from the maximum power that can be provided by the system. For rotorcraft control, it is essential there is more power available than is being used since control happens through thrust adjustment. For concept one, the motors are operating at 50% power, and to keep the comparison fair, the same useful power fraction will be used for this concept. These boundary conditions can then be written down as the following relation: 15

16 The system can now be completely solved, which results in the following equation for the endurance of the fuel-powered quadcopter: As can be seen from the equation above, an estimation of the fuel-powered rotorcraft s performance requires knowledge about the empty mass of the rotorcraft, the maximum power output of the engine, the efficiency of the system, and the amount of rotors present on the rotorcraft. These factors will be estimated for the requirement of one kilogram of payload. For simplicity of the transmission system, we will assume the rotorcraft has four rotors. Research shows the smallest commercially available engines produce around 2 kw while weighing approximately 1 kilogram including peripherals (DLE Engines, 2015; O.S. Engines, 2015). To keep the comparison fair, we will use the empty mass of the first concept as an indication of the empty mass of this concept. Since it will run at 50% power, it is able to weigh 14 kg due to the efficiency of 14 g/w. It is assumed that half the weight of the system are batteries, therefore, the result will be an empty weight of 7 kg including payload. The final, unknown, parameter is the engine efficiency. While gasoline engines can reach an average efficiency of 20%, it is not expected this efficiency will be reached on this rotorcraft because of the drawbacks of downscaling a gasoline engine. Fuel consumption data on this engine size is not well documented, therefore, an assumption of efficiency is made at 10% as a safety factor because the choice and the performance of the engine are not yet known. As for fuel energy density, it is assumed gasoline has an energy density of 44 MJ/kg. 16

17 With these parameters, it is possible to make an estimation of the rotorcraft s performance, which results in an expected flight duration of 7.7 hours with a take-off mass of 11.4 kg. The sizing of the third design involves a mixture of techniques from both designs. The new components necessary for this design include a high power alternator and rectifier. If we assume the engine size is the same as for concept two, commercial components of this power rating would weigh at least 2 kg at a maximal 92% efficiency (Sullivan Products, 2015). By using the mass differential equation from the second concept and the rotor efficiency approach from the first concept, we can then derive a new mass differential equation for this concept: This equation has the following boundary conditions: Solving these equations then yields: Due to the craft s mass being variable, it is impossible to always fly at top performance of the fixed pitch rotors. This variation is rather small thus we can see that it is still possible to fly with an average efficiency of 13.5 g/w. If we then assume the empty mass is equal to the empty mass of the second concept with the added mass of the required generator systems on top, we can calculate the total flight time of 4.9 hours. We can also further calculate that the concept s take-off mass is kg. Both concept two and three have the possibility of providing a much better performance than what is possible with the electricity-based concept one. While both concepts have a lower maximum take-off mass at the same power generation compared to the first concept, they make up for this by being able to carry significantly more energy. 17

18 Both these concepts have the potential to meet the flight time requirement, however, the expected flight time of concept two provides a significantly higher design safety margin than concept three. The improved efficiency due to fixed pitch rotors in concept three seems to not make up for the mass and losses added by the generator subsystem. With current technology, it is theoretically impossible for the first concept to meet the required flight time so it will not be discussed any further. A big problem with concept two is the system s complexity. Both concept one and three have relatively small moving parts; however, concept two requires a complicated transmission system in order to power the rotors. This limits the freedom of structure s design and requires a strong frame to function well. Furthermore, it requires variable pitch rotors which bring significant complexity to the rotor mounting itself. The development time for the rotorcraft is very limited thus it could pose a risk for the project. The variable pitch rotors do; however, have several advantages. The large rotors which are necessary to make concept three efficient have significant inertia and can make control of the vehicle difficult because of the delayed response. Meanwhile the response of the variable pitch rotors of concept two can react faster than any fixed pitch design and can also provide negative lift. Furthermore, no energy is wasted during maneuvering since it is not necessary to slow down and accelerate the rotors. The variable pitch rotors also allow for a safety measure to be in place, in case the engine of the rotorcraft runs out of fuel. In concept three, a separate battery-powered system will have to be created in order to allow safe landing; however, concept two has the innate possibility of using autorotation to safely descend. Finally, concept two is a better fit for the scalability requirement. The fixed pitch rotors of concept three cannot be scaled up efficiently because of their increasing inertia. This can be solved by adding more rotors instead of scaling up rotors; however, this results in reduced efficiency and an overall more complicated system. Meanwhile, the structure of concept two can be scaled up while still having functional rotors. 18

19 The first concept was quickly discarded because it did not meet the requirements. The second concept easily meets the expected performance and the third concept can be adapted to meet the goals. Since it was found out that commercial parts are easily available to create the complicated transmission of concept two and this removed the largest risk, it was decided to continue the preliminary design with concept two. Once the concept design was settled on, we had to make a choice about the engine type. For a rotorcraft of this scale, there are three major engine types, namely nitro engines, gas engines and model turbine engines. Nitro engines are the smallest engines available on the market. Their main feature is that they run on a special nitro fuel mixture, which allows them to run at very high rotational speeds and this enables very lightweight engine designs to still produce considerable power. Their main drawbacks exist in how expensive their fuel is and the very oil-rich exhaust they produce which can cause damage to the systems they are integrated in. The second option is a miniature twostroke gasoline engine. Even though it cannot deliver the extreme power to weight ratio as a nitro engine can, they do provide significantly more efficiency and can operate on normal gasoline. One drawback; however, is that they need a spark-plug based ignition system, but recent developments have made this less of an issue because of the rise of electronic ignition systems. The final option exists in model turbine engines. These engines can run on almost any fuel; however, they require a special controller to regulate the engine. While they can produce incredible power to weight ratio, they tend to be inefficient (Salt, 2015). The choice was made to go with the two-stroke gasoline engine because the project s goal is to design a long-range quadcopter. Even though the engine is slightly heavier, the motor s efficiency easily weighs up to the slightly increased mass. The largest issue in terms of the physical size is the sourcing of the variable pitch rotor hubs required for the design to function. These components are very complex and to produce them to order is impractical due to project s deadlines, which means these parts will have to be sourced commercially. The largest commercial availability of small scale 19

20 variable pitch rotor hubs exists in the market of model helicopter s tail rotors. Even though the main rotors used by these helicopters are unsuitable because of their full coaxial control, their tail rotors are an excellent fit in our case since the assemblies often already include a right angle connection. Helicopter tail rotors normally support significantly smaller blades thus it is important to find the largest commercially available tail rotor hub since the efficiency of the rotorcraft increases with the size of the rotors. This led to the Align T-rex 800E tail rotor hub being chosen as the rotor hub assembly for this project. It is commercially available and can fit rotor blades as large as 325 mm in length, like for example the Align T-rex 325 mm carbon fiber blades (van Natter, 2014). Larger blades are more efficient thus this blade size was taken as the final rotor size for the design. Next, is choosing the frame design. This choice will have a large impact on the transmission design because it determines how axes and belts can be used to power different rotors. The chosen rotor hub components allow an axis connection directly perpendicular to the rotor hub so it was decided to use an H-frame with four rotors due to the simplicity of the resulting transmission system. Furthermore, the rotors will be located on top of the frame arms in order to prevent problems due to rotor flexibility and will also simplify the landing gear. With the last remaining core design parameters fixed, we can make a full estimation of the rotorcraft design. The data below lists all data used for this design. Airfoil (profile estimated based on the thickness of the chosen rotor blades) NACA 0013 Amount of blades per rotor 2 Amount of rotors 4 Rotor radius [mm] 350 Rotor hub radius [mm] 50 20

21 Rotor chord [mm] 33 Rotor shape Straight, no twist Engine efficiency [%] 10 Fuel energy density [MJ/kg] 44.4 Engine maximum shaft power [kw] 2 Steady state maximum power usage [%] 50 Engine efficiency [%] 10 Empty mass (no payload) [kg] 6 Payload mass [kg] 1 Figure 7: data used to estimate the performance of the rotorcraft during preliminary design. Rotor airfoil characteristics were derived using XFOIL at M=0.3, Re= With this data, a relation between maximum flight time and payload mass can be developed. Total fuel mass [kg] 5.33 Maximum flight duration [h] 9.94 Rotor rotational rate at start of flight [rot/min] Rotor rotational rate at end of flight [rot/min] Rotor rotational rate at maximum thrust [rot/min] Power usage at the end of flight [W] Figure 8: Calculated performance characteristics of the rotorcraft specified in table 7. 21

22 Figure 9: Expected maximum flight time for variable payload mass of the rotorcraft design specified in table 7. Interestingly, it turns out the rotor design used in this performance estimation is significantly better than the rotor design previously assumed. Furthermore, in the above figure, it can be seen that the resulting rotorcraft is not only able to lift 1 kilogram of payload for a significant amount of time, but it can also lift the payload close to the rotorcraft s weight for shorter amounts of time. 22

23 4. Detailed Design Firstly, it is important to investigate the most critical components of the design. Several subsystems have significant impact on other components of the system. These subsystems are the rotor hubs, the rotor blades, the engine, the transmission system and the frame layout. While the rotor hub and blade subsystems have already been decided on, the rest of the power systems still has to be designed. The first choice to make is choosing the engine. The requirements for this component come from the necessary rotational rate range, power range and the rotational range on which the clutch engages. Due to the fact that gas helicopter engines require a clutch for smooth start-up behavior, the availability of this component will be researched first. Market research shows that all lightweight clutches for the wanted engine size are socalled centrifugal clutches. They operate in a concentric fashion where the inner flywheel expands because of the rotational rate of the flywheel until it locks against the clutch bell at a set rotational rate. Due to the size and rotational rate of the wanted clutch, the only commercially available models engage at 6000 RPM. This means a fast motor is necessary, followed by a stepdown conversion in the transmission system. The only engine available that meets these requirements is the OS GT15HZ engine model. It has an effective range of 2000 up to RPM, a 4:5:1 stepdown in the transmission system so it provides the rotational rates required to fly the rotors. With the choice of the engine settled on, the transmission system can be designed. To ease the design process, a clutch is selected that is known to be compatible with the engine model. This clutch already comes with a 14 teeth pinion gear thus a matching spur gear with 64 teeth is selected which results in a step down conversion of 4.57:1, which meets the engine requirements. The frame layout is decided to be an H-frame so 23

24 this gear will power an axis which extends through the central beam of the frame. Similar axes will be mounted between the rotor hubs and by using miter gears where these axes meet the central axis, the correct rotation of rotors can be ensured. In order to keep friction in the transmission system low, it will be suspended on bearings rated for both the rotational rate of the rotors and the axial loads placed on the axis due to the gear joints. An overview of the transmission system can be seen below. Figure 10: Layout of the complete transmission system including the engine. With the critical subsystems designed, it is possible to continue with the general design of the rotorcraft. Firstly, we will discuss the general frame layout and after that the different subsystems. One of the core requirements is that parts of the frame will be 3D printed. While the 3D printing technology does not offer many advantages for the central 24

25 beam of the frame, the multitude of different interfaces present on the rotorcraft arms make these components excellent for 3D printing. A big problem with these arms is that they are significantly larger than the maximum size that can be printed with the printers available at Leapfrog 3D Printers. Therefore, it was decided to split the arm into multiple pieces that can be glued together. Meanwhile, the frame of the center beam poses a different problem. It has to provide support for the transmission system, and also house the components of the other subsystems while remaining easy to produce as well as lightweight. Keeping the requirement of upgradability in mind, it was decided to construct this component from sparse folded aluminum sheets. This allows easy mounting of different components and provides the strength required to house the transmission system while remaining extremely lightweight. The engine, clutch and central transmission shafts can then be supported through several milled pieces of aluminum mounted inside the structure. Since the size of the central beam structure is known, it is time to focus on the joint between the central beam and the arms supporting the rotors. Keeping maintainability in mind, we decided this joint should be easy to unlock in order to inspect the gearing system which will be located at this joint. A big issue; however, is accurately attaching the manufactured aluminum structure to the rough 3D printed structure of the arm frames. A solution to this problem is putting the 3D printed part in between the metal parts of the joint and a small aluminum plate on the other side of the frame. An iterative process is performed to design the 3D printed arm frames. Because of the way 3D printed parts are manufactured, the stresses on the part will be almost completely carried by the outside of the part, while the inside of the part contains a simple supporting structure that prevents the outside skin from collapsing. Due to the distribution of the 25

26 loads, a simple design is made featuring a tapered stretched cylinder design which will not significantly interfere with the airflow of the rotors. Testing shows that an optimal result is reached by combining a part wall thickness of 0.8 mm with a supporting structure with 10% infill. The resulting part structure can be seen below: Additional features include the presence of two extrusions at the arm tips which act as landing gear, and a pocket in which servos can be mounted in order to control the variable pitch rotors. Another interesting point is that the arm frame is asymmetrical. This is caused by the asymmetry of the used rotor hubs, which means that the servos will always be located on the same side of the outer rotor arm components. It also means that both outer rotor arm parts are identical, which simplifies the production of the parts. Figure 11: Unfinished test print of the central section of the arm frame. 26

27 Figure 12: Overview of the resulting arm frame design. 27

28 Figure 13: Overview of the center frame design. The frame design is completed so the control subsystem can now be focused on. This system can be divided into different subsystems, namely measurement, controller and actuation. The first system we will look at is the actuation subsystem. The rotorcraft features four pitch-controlled motors, and one engine which can be throttle controlled. All of these components receive their input through the movement of a mechanical lever and therefore, a mechanical actuator is necessary. Due to the relatively high speed control loop of the model multi-rotorcraft, these actuators are able to quickly apply feedback. In order to achieve this, servo motors which update at 330 Hz were chosen for this system. 28

29 The controller subsystem has to drive these actuators. In order to keep the development time short, a COTS system was selected for the controller subsystem. This system can be customized to fit the needs of the rotorcraft and has the capability to power all servos. It also includes a pressure sensor and a six degrees of freedom inertial measurement unit. The measurement system is responsible for gathering data for the control subsystem. In order to fulfil the requirements, several sensors are selected next to the sensors already present in the control system. The first sensor is an ultrasonic distance sensor which can be used to accurately measure the distance to the ground for soft landings. Furthermore, a Hall Effect switch sensor will be included which can be used to measure the rotational rate of the rotors due to a set of magnets mounted on the clutch. This is necessary since the performance of the rotors significantly varies with this rotational rate, therefore, the controller needs to be able to effectively regulate the engine. These systems run on electrical power, the same as the electrical ignition used by the engine. While it is possible to generate this power on board, it makes the design unnecessarily complex. Due to this, it is decided to power the electrical systems of this prototype rotorcraft from batteries. The current consumption of the electrical ignition is rated at 750 ma at RPM; however, the current consumption of the control system is not well known because of the variable power draw of the servos. To compensate for this, the current sensor will be integrated in the servo power supply to estimate the servo current draw. Until this performance is tested, two 10 Ah two-cell Lithium polymer batteries will be used as the energy source. The combined capacity of these batteries should be able to handle a current draw of 4A for the required flight duration, providing a safe power budget of 0.65A per servo. The current draw of the controller and measurement subsystems is negligible compared to this and is therefore not included in the analysis. 29

30 The final subsystem to be designed is the energy storage subsystem. This rotorcraft design incorporates a gasoline engine, therefore this means a large amount of gasoline has to be stored on board. If the full 5.3 kg of fuel are stored in one large tank, this can cause issues due to the fuel sloshing around in the tank, making it impossible to stabilize the rotorcraft. Another issue lies in the change of the center of gravity of the rotorcraft due to the fuel being depleted over the duration of the light. These issues are solved by storing the fuel in multiple small tanks located in the center of the craft, at the sides of the center frame. Since the fuel cannot move freely around large distances, the impact of sloshing on the stability of the rotorcraft is reduced. Furthermore, the fuel is stored very closely to the center of gravity, which reduces the influence of the fuel mass on the movement of the center of gravity. As an added benefit, this also allows the rotorcraft to remain relatively flat since there is plenty of space not covered by the rotors in this area. 30

31 5. Prototype Results The following estimation is made for the prototype: Airfoil (characteristics derived NACA 0009 using XFOIL) Amount of blades per rotor 2 Amount of rotors 4 Rotor radius [mm] 350 Rotor hub radius [mm] 50 Rotor chord [mm] 33 Rotor shape Straight, no twist Engine efficiency [%] 15 Fuel energy density [MJ/kg] 44.4 Engine maximum shaft power 2 [kw] Steady state maximum power 50 usage [%] Engine efficiency [%] 10 Engine mass [kg] 1 Rotor and rotor hub mass [kg] 0.25 Frame mass [kg] 3 Figure 14: Prototype estimation. Based on this information, the simulation gives the following results: Maximum fuel mass [kg] 7.3 Maximum flight duration 19.0 [h] Rotor RPM when full 2603 [rot/min] 31

32 Rotor RPM when empty [rot/min] 1554 Figure 15: Simulation results. This helps establish the following relation between the payload weight and flight duration: Figure 16: Relation between the payload weight and flight duration. Then, we performed these same tests on a larger 12 kw rotorcraft and these are the results: Airfoil (characteristics derived NACA 0009 using XFOIL) Amount of blades per rotor 2 Amount of rotors 4 Rotor radius [mm]

33 Rotor hub radius [mm] 120 Rotor chord [mm] 81 Rotor shape Straight, no twist Engine efficiency [%] 22 Fuel energy density [MJ/kg] 46 Engine maximum shaft power 12 [kw] Steady state maximum power 50 usage [%] Engine mass [kg] 10 Rotor and rotor hub mass [kg] 1.5 Frame mass [kg] 18 Figure 17: 12 kw rotorcraft test results. The simulation gave the following results: Maximum fuel mass [kg] 34.5 Maximum flight duration 26.9 [h] Rotor RPM when full 1058 [rot/min] Rotor RPM when empty 745 [rot/min] Figure 18: 12 kw rotorcraft simulation results. 33

34 This helps establish the following relation between the payload weight and flight duration: Figure 19: Relation between the payload weight and flight duration. These results show that quadcopters that are fuel-powered have the potential of having a much longer flight time when compared to an electrical multi-rotorcraft since those are limited to a flight time of maximum two hours. This is due to an increased amount of energy that can be stored in the fuel as well as the sharp rise in flight time at low payload mass implies there is another effect present. This rise can be made more clearly visible by looking at the power consumption of the craft during the flight. 34

35 Figure 20: Relation between the rotorcraft s power use and flight time. The above graph shows that the rotorcraft s power use drops significantly as the flight time increases. Power consumption of electrical multi-rotorcraft is, on the other hand, constant. The explanation for this is that fuel-powered rotorcraft loses mass while flying, while electrical rotorcraft has to carry the batteries even when they are empty. This, therefore, doubles the efficiency of the rotorcraft near the end of its flight time, which allows these rotorcrafts to have an extremely good endurance. Another element to consider is the rotorcraft s maximum range. However, due to the aerodynamics of the multi-rotorcraft this is hard to estimate because of the complex interactions between the multiple rotors and the rotorcraft body itself. It is possible to give an estimation of the range at a low speed by comparing it to other quadcopters of the same size; however, the effects of quadcopters forward flight on their power consumption is not very well understood. Due to the variable pitch rotors, the rotorcraft 35

36 should be able to maintain reasonable efficiency at higher forward velocities, however, it is hard to estimate the effect it will have on the power consumption and the stability of the rotorcraft. If looking at the performance of a normal fixed-pitch rotorcraft, it is definitely possible to fly 50 kilometers per hour without experiencing major drawbacks in power consumption. This makes it possible for the prototype design to have a range of nearly 800 kilometers while the large scale design could travel over 1000 kilometers. These numbers would; however, decrease because of the payload in the same way as the maximum flight time decreases. 36

37 6. Conclusion This white paper has proven it is possible to construct a rotorcraft that meets all the requirements mentioned at the beginning of the report. The designed rotorcraft has the potential to fly with one kilogram of payload for more than nine hours using power provided by a gasoline engine. To ensure the project results in an operational rotorcraft that meet the requirements, it is essential the recommendations below are followed. These recommendations are based on problems encountered during the construction of rotorcraft s parts, as well as concerns raised during the design of the rotorcraft. In terms of structural integrity, it is important to analyze the expected deformation of the frame under maximum load. Next, it is important to investigate the effects of continuous operation on the 3D printed components of the frame. Furthermore, different components of the transmission system have to be further researched. When it comes to vibration, the possibilities of damping between the frame and the controller have to be researched. It is also important to measure vibrations caused by the engine and the rotors. Finally, a vibrational analysis on the frame has to be performed. In terms of performance, it is important to analyze the thrust losses due to downwash on the frame, investigate the use of an expansion chamber on the engine, analyze the actual efficiency of the engine and investigate the effects of forward flight on fuel usage. With control, there has to be an investigation of different controller possibilities for regulating engine performance and it is important to integrate GNSS or remote control into the design. 37

38 When it comes to enhancements, an aerodynamic cover for the center beam has to be designed as well as an inbuilt starter instead of requiring the use of a manual starter for the engine. The aim of these recommendations is to aid future development based on this prototype. 38

39 7. References DLE Engines (2015). Small gasoline engines. Ger J. J. Ruijgrok (2009). Elements of airplane performance. VSSD J. Salt (2015). Choosing the best RC helicopter engine. M. Drela. (2013). Xfoil- an analysis and design system for low Reynolds number airfoils. M. J. Cutler (2010). Design and Control of an Autonomous Variable-Pitch Quadrotor Helicopter. Massachusetts Institute of Technology O.S. Engines (2015). Heli engines. R. C. Hibbeler (2008). Mechanics of Materials. Prentice Hall S. van Natter (2014). Gas Powered Single Engine Variable Pitch Quadcopter. Sullivan Products (2015). Alternators and Regulators for Unmanned Vehicles. T-motor (2013). Consulted on the 28th of May