Design and Flight Test of a Medium Range UAV for Aerial Photography

IJUSEng 2015, Vol. 3, No. 3, 40-49 http://dx.doi.org/10.14323/ijuseng.2015.12 Research Article Design and Flight Test of a Medium Range UAV for Aerial Photography Dewi Anggraeni, Dony Hidayat, AM Pramutadi,?? Mujtahid and Arifin Rasyadi. National Institute of Aeronautics and Space Aviation Technology Centre, Bogor, Indonesia. Abstract: Anggraeni D, Hidayat D, Pramutadi AM, Mujtahid?? and Rasyadi A. (2015). Design and flight test of a medium range UAV for aerial photography. International Journal of Unmanned Systems Engineering. 3(3): 40-49. UAV-based remote sensing for supporting environmental forestry planning, ecosystem preservation, and agricultural planning has great potential in Indonesia. Wherefore it is necessary to develop a UAV for these applications. The development of a medium range UAV have been conducted at LAPAN (National Institute of Aeronautics and Space). The aircraft was designed with a wing span of 5.5 meters and payload capacity of 30 kg, considered to be an optimum solution for aerial photography. In this study, the design activities included comparisons to an existing aircraft to determine initial sizing for the UAV. A pusher high-wing twin-boom aircraft was chosen because of its enhanced stability. Aerodynamic analysis of the wing was conducted using a four method calculation to obtain the design configuration. Wing structural strength was analyzed, and then an appropriate engine propeller configuration for the UAV power generation was selected. Flight tests were successful and UAV performance was confirmed. Results show that the new medium range UAV concept can be applied to realize long endurance flight for aerial photography missions. Marques Engineering Ltd. Keywords: Aerodynamics Aerial photography Ecosystem preservation UAV Remote sensing I. INTRODUCTION The role of geographic information system technology is important in environmental forestry development planning. One benefit of this technology is the capability to plan for forest and land rehabilitation in order to anticipate deforestation and land degradation. Remote sensing data is effective in determining the parameters of deforestation and land degradation [1]. Remote sensing technology with Correspondence National Institute of Aeronautics & Space Aviation Technology Center Bogor Indonesia dewi.anggraeni@lapan.go.id high spatial resolution is expected to be used for mapping forest areas, specifically locations must be protected from increased development so that ecosystem preservation in coastal areas can be realized [2]. Remote sensing using UAVs must achieve cost-effectiveness, fast production, ease 40 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

of operation by local staff, and good geometry accuracy (sub-meter) during the remote sensing process [3]. For the development of Indonesia it is imperative to develop a UAV for remote sensing [4]. This research was carried out at the Aviation Technology Centre LAPAN, with the aim of designing a Medium Range UAV platform and manufacturing a scale sized UAV prototype. The main objectives were to develop a medium range of UAV capable of carrying 30 kg of payload. On the first year, configuration and sizing were conducted and also the engine and propeller selection process was completed. In the following year, the structure was analyzed and also aerodynamic analysis using wind tunnel testing was carried out to confirm the performance of the design configuration. By the end of 2014 flight tests were conducted successfully. II. DESIGN OF A MEDIUM RANGE OF UAV Goal of the Project The general requirement is to develop a utility unmanned platform that is reliable, robust and safe, that can be operated with minimum support, and which is able to carry out low speed low altitude observation missions carrying on board electro-optic devices. Aircraft Sizing The sizing of the aircraft was carried out using a comparison method. Several existing aircraft (Yabhon-H, Chacal, Skyblade IV, Acturus T-20) were used as a base of comparison to determine the initial sizing for the aircraft. Initial weight sizing was performed with the help of a method described in the literature [5]. The method, with the input of several parameters and iteration, gives a result of 75 kg for the initial weight sizing. There empty weight was 31.37 kg, fuel weight was 13.77 kg and payload weight was 29.99 kg. The geometrical sizing for the aircraft was carried out by variation of wing loading and speed for several lift coefficients (c l); Fig. 1. Fig. 1: Baseline design value for wing loading 41 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

From Fig. 1 a design point is chosen. Therefore, to fly at a speed of 100 kph (28 m/s) and with the assumption for c l of 0.533 the wing loading is about 229 N/m 2. Then, by combining the initial weight and the wing loading parameter, the wing area shall be 3.22 m 2. Aircraft Configuration The configuration chosen for the aircraft is a pusher high-wing twin-boom aircraft. This configuration is common among various UAVs and gives a good clean front area in the aircraft that can be utilized to install various payloads without any interference. A high wing aircraft gives better stability which is needed for an aircraft with the present objective. The placement of the engine in the back of the aircraft with a pusher layout provides better center of gravity placement; whereby, the twin boom is the result of placing the engine in a pusher layout in the middle of the aircraft. The front of the fuselage is reserved for the installation of various devices needed to support the mission. With good center of gravity placement and good stability, the task of observation can be performed effectively. The center of gravity placement also gives the flexibility to interchange devices without major changes to the aircraft. The wing incorporates one pair of ailerons and two pairs of flaps. The flaps are utilized during takeoff and landing. The flaps are also used during low speed observation flight. Both the aileron and flap have a plain configuration. The boom is connected to the mid wing. At the end of the tail boom lies the vertical tail. The horizontal tail is connected to both tail booms. The landing gear is attached to the fuselage in a fixed tricycle layout. III. AERODYNAMICS PERFORMANCE From a wide selection of airfoils analyzed, the NACA 4415 was selected. Selection was based on the needs of cl, ease of fabrication, and a requirement to incorporate high-lift devices in the wings. Wind tunnel testing for various angles of attack (α) was conducted to determine the performance of the wing designed. The results of wind tunnel tests were then compared with the results of simulation using Datcom software and XFLR5 software, and also to analytical calculations based on Raymer s book [5], as shown in Fig. 2. Maximum α of the wind tunnel test results was 12 deg, the same as the results of simulations using Datcom, however in the graphic (Fig. 2) the slope of wind tunnel test result showed that from an angle of 6 deg c l diverges from a linear trend. This indicates that the test model was not a rigid body and caused large deflection at the wing and it will be noted in future research. While the slope of the wind tunnel test data diverges compared to other lines after α of 6 deg, generally all lines showed identical slope. IV. AIRCRAFT STRUCTURE Structural Analysis The UAV was constructed using glass fiber reinforced plastic and carbon fiber reinforced plastic. The frame is shown in Fig. 3. The final weights are shown in Table 1, and the total mass is less than 48 kg. Generally, for this prototype, the overweight was 34% of the total weight [6,7]. This indicates that the assembly process was dominant and should be reduced to improve structural performance in the design of the UAV. Analysis using Finite Element Method (FEM) was conducted to ensure that each airframe structure was capable to withstanding the applied load. 42 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

Fig. 2: Comparison graphic c l calculation with different methods Fig. 3: Design of medium range of UAV 43 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

Table 1: Weight of components No Component Breakdown Mass (design) [kg] Mass (manfacture) [kg] 1 Fuselage 3.27 7.30 2 Fuselage Cap 0.44 0.55 3 Inner Wing + insert 5.00 11.34 4 Outer Wing-left 5.00 5.90 5 Outer Wing-right 5.00 6.00 6 Tail Boom-right 1.68 1.05 7 Tail Boom-left 1.68 1.05 8 Right Vertical Tail 1.00 1.40 9 Left Vertical Tail 1.00 1.30 10 Horizontal Tail 3.00 3.25 11 Main Landing Gear 2.82 2.55 12 Right Wheel 1.61 2.24 13 Left Wheel 1.61 2.24 14 Nose Landing Gear 0.64 0.35 15 Nose Wheel 1.61 1.15 Total 35.35 47.67 The wing skin was constructed using WR 185 E-glass fiber and epoxy resin Bakelite EPR 174 with fiber directions at ± 45 deg. The spar was built from UD 300 carbon fiber material and resin Bakelite EPR 174. From the tensile test performed, the strength of fiber composite material E-glass was 119 MPa and 771 MPa for carbon fiber. The analysis found that the stresses that occur in every airframe structural components are still under the material failure strenght (Fig. 4). This shows that the component considered is capable of withstanding the maximum force of varying stress. V. ENGINE AND PROPELLER SELECTION Choosing an engine involves several considerations such as twin tail boom pusher configuration, propeller fixed pitch effect, propeller ground clearance, and engine operation at mission profile. A thrust requirement was established to select suitable engine and propeller. From the design configuration data (Table 2) the thrust needed is 294 N and the ground clearance of the aircraft is suitable for a propeller of 28 inch to preserve it from sweeping the ground. Engine power shaft is determined to generate minimum power required by the UAV to select the proper engine available in the market. It is necessary to define the thrust to weight ratio coefficient for low loading and high loading, then calculate the thrust force and thrust coefficient for both loading conditions and, lastly, determine the total efficiency of the propeller to deliver the engine power shaft. See Table 2 for the UAV s engine shaft power calculation data input and Table 3 for the UAV s engine shaft power data output. From the three-step process of determining and calculation, the required power to turn the propeller with 28 inch x 12 inch dimension at heavy loading condition is 18 44 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

horsepower. At a light loading condition of 11.2 hp, the propeller met the requirement of weight to power ratio at cruise. Therefore, we could conclude that the UAV needs an engine that can supply a minimum power of 18 hp [8]. Fig. 4: Finite element results for the wing Table 2. UAV s engine shaft power calculation - Input data INPUT Value Units Max Take Off Weight (MTOW) 75 kg Max Cruising Speed 33.3 (65) m/s (kts) Target : Weight to Power (W/P) Ratio 65.76 at static sea level cruise Propeller Radius (Dimension: 30 x 12) 38.1 cm Propeller Efficiency 0.85 literature (T/W) Ratio at Take Off 0.4 (T/W) Ratio at Cruise 0.3 Thrust coefficient at Take-Off 0.94 Thrust coefficient at Cruise 0.68 Viscous Efficiency at Take-Off 0.35 Viscous Efficiency at Cruise 0.85 literature Thrust (T) at Take Off 294 N Thrust (T) at Cruise 212 N Density of Air (sea level) 1.225 kg/m³ Max Cruising Speed (Airspeed) 33.43 m/s 45 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

Table 3: UAV s engine shaft power calculation - Output data OUTPUT Value Units Power at Heavy Loading 13.358 Watts 18 hp Power at Light Loading 8356 Watts 11.2 hp Weight to Power (W/P) Ratio 65.65 VI. FLIGHT TESTS Flight tests were conducted succesfully in the morning of a sunny day by the end of the year 2014, precisely on 22 nd December 2014. Flight tests were conducted on a grass type runway or airfield at Pameungpeuk Beach, Garut, West Java. The flight was controlled by a pilot on the ground using a remote wireless control system. The flight data was transmitted by onboard autopilot module to the ground and was displayed and recorded by the ground station. Fig. 5 shows the prediction of take-off distance and maximum take-off weight with variations of flap configuration [9,10]. A flight test was conducted to measure various parameters of ground distance and loiter speed performance. The flight route of the UAV was restricted to a small range, the radius was about 150 m. The measurement method of ground distance at take-off consisted of using onboard instrumentation from the autopilot module. According to data transmitted and recorded by the onboard autopilot module, the altitude was mostly varied in the range of 150 230 m, and the speed ranged between 25 38 m/s, the average speed was 30 m/s. Fig. 6 shows the measurements of altitude and speed at take-off. The calculated ground distance is 143 m. The takeoff profile (Fig. 6) agreed with the prediction for take off performance (Fig. 5). Fig. 5: Take-off performance prediction 46 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

Fig. 6: Take-off profile VII. CONCLUSIONS This paper presents the design and flight test of a medium range UAV. The configuration was pusher twin-tail boom, high-wing, with fixed tricycle landing gear. The UAV was mainly constructed with glass fiber reinforced plastic and carbon fiber reinforced plastic, and was analyzed using FEM to achieve an aircraft that withstands maximum force of varying stresses. The NACA 4415 airfoil was selected and the characteristic c l /α were analyzed using Datcom software, XFLR5 software, analytical calculations based on Raymer s book [5], and also wind tunnel tests result. The results show that c l /α curves from different method were identical. The propeller was determined to use 28 inch x 12 inch, and the engine required about 18 hp minimum of power. According the flight test result of take-off distance about 143 m was similar with the take-off design consideration of 150 m in clean configuration. It was shown that the UAV have the capability for long endurance flight. It can be concluded that the weight of the preliminary UAV may be further reduced and the 47 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

aerodynamic performance may be improved to obtain better flight performance. VIII. ACKNOWLEDGEMENT The authors wish to thank Mr. Gunawan Setyo Prabowo and Mr. Ari Sugeng B. for the financial support for this research. Thank you also to all group members who have worked hard to complete this research program. IX. REFERENCES 1. Rusdiyatmoko A. (2015). Forest and land rehabilitation based on watershed management by using remote sensing data and geographic information system (Case study: watershed Kahayan, Central Kalimantan province). VI Congress of Indonesia Remote Sensing Professional Community. Bogor Institute of Agriculture. P5-2. 2. Anggraini N and Hasyim B. (2015). Landsat data utilization for the analysis of spatial northern coast of Jakarta. VI Congress of Indonesia Remote Sensing Professional Community. Bogor Institute of Agriculture. P7-2. 3. Laily R. (2015). LAPAN: Drone more efficient for remote sensing. Antara. http://www.antaranews.com. 4. Rokhmana CA. (2015). The potential of UAV-based remote sensing for supporting precision agriculture in Indonesia. Procedia Environmental Sciences. 24: 245-253. 5. Raymer DP. (2012). Aircraft design: A conceptual approach. (5 th ed.). American Institute of Aeronautics and Astronautics. Reston, VA. 6. Abdurrohman K, Wandono FA and Hidayat D. (2014). Stress and analysis of LSU 05 twin tail boom using FEM. International Seminar on Aerospace Science and Technology I. LAPAN, Tangerang. 7: 51-60. 7. Wandono FA, Ardiansyah R and Hidayat D. (2014). Failure analysis on main landing gear frame structure LSU-05 based on Tsai-Hill failure criterion. Research and Scientific Thought Concerning Aircraft Technology. 5: 70-85. 8. Anggraeni D, Sumaryanti AR, Sumarna E and Rahmadi A. (2014). Engine and propeller selection for propulsion system LAPAN Surveillance UAV 05 (LSU-05) using analytic and experimental test. Proceedings International Seminar of Aerospace Science and Technology II. Tangerang. 6: 41 50. 9. Ruijgrok GJJ. (2009). Elements of airplane performance. Delft Academic Press. Delft. The Netherlands. 10. Roskam J and Lan CT. (1997). Airplane aerodynamics and performance. DAR Corporation. Lawrence. Kansas. AR C C l CD CDO E GPS H L/D LAPAN LLT Aspect ratio Chord length (m) Lift coefficient Drag coefficient Zero-lift drag coefficient Wing efficiency factor Global Positioning System Altitude (ft) X. NOTATION Lift-to-drag ratio Indonesian National Institute of Aeronautics and Space Lifting Line Theory 48 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49

MAC MPa MTOW NACA RC Mean aerodynamic chord (m) Mega Pascal Maximum take-off weight National Administrative Committee for Aeronautics Radio control Re Reynolds number RPA Remotely piloted aircraft S Wing area (m 2 ) T/W Thrust to weight ratio UAV Unmanned aerial vehicle UD Unidirectional VLM Vortex Lattice Method W/P Weight to power ratio W/S Wing loading (kg/m 2 ) WR Woven roofing X, Y Cartesian coordinates α Angle of attack ( ) ρ Air density (kg/m 3 ) Copyright of IJUSEng is the property of Marques Engineering Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. 49 www.ijuseng.com IJUSEng - 2015, Vol. 3, No. 3, 40-49