Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings

Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings Jason T. Cantrell 1, Bradley W. LaCroix 2 and Peter G. Ifju 3 Mechanical and Aerospace Engineering Department, University of Florida, MAE Receiving, 134 MAE-C, Gainesville, Florida 32611, USA 163 ABSTRACT Micro air vehicles with thin membrane wings have the ability to passively adapt to flight conditions improving aircraft stability. This work is concerned with the prop-wash induced asymmetrical loading of a perimeter reinforced wing. A perimeter reinforced membrane wing is evaluated versus a relatively rigid graphite/epoxy wing to determine if this variation improves lateral-directional flight characteristics, mainly adverse roll due to propeller torque. Digital image correlation, in conjunction with loads measurements in a wind tunnel investigation, are implemented to quantify the wing deflections and stability coefficients. Results indicate that the perimeter reinforced wing does display potential benefits for low speed and high angle of attack flight that could be further tailored to enhance roll compensation for future micro air vehicles. NOMENCLATURE AOA = Angle of Attack AR = Aspect Ratio BR = Batten Reinforced CF = Carbon Fiber CG = Center of Gravity C L,C D,C M,C l = Lift, Drag, Pitch, and Roll Moment Coefficients C Lα, C mα = Lift and Pitch Moment Slope Versus Angle of Attack DIC = Digital Image Correlation MAV = Micro Air Vehicle PR = Perimeter Reinforced RCF = Rigid Carbon Fiber RPM = Revolutions Per Minute u,v,w = Chordwise, Spanwise, and Vertical Displacements VI = Virtual Instrument 1. INTRODUCTION The concept of a flexible wing micro air vehicle (MAV) was first introduced by University of Florida researchers in the late 1990s and early 2000s [1, 2]. MAVs are notoriously difficult to fly due to their small scale, yet must be highly maneuverable to navigate challenging environments. The size of MAVs generally dictate the planform be a low aspect ratio wing in order to maximize lifting capability causing handling qualities at low speeds to degrade rapidly. Additionally, the range of statically stable center of gravity (CG) locations is minute for MAVs making the margin of error on proper construction and weight distribution equally small. Flight conditions, including wind gusts on the same order of magnitude as the vehicle flight speed (10-15 m/s) can make maintaining controllable flight difficult as well [3, 4]. Passive shape adaptation through the incorporation of flexible wing structures into MAVs diminishes many of these known issues and provides a larger flight envelope. These adaptations 1 Graduate Research Assistant, jasontcantrell@gmail.com 2 PhD Candidate, bradlacroix@gmail.com 3 Professor, ifju@ufl.edu

164 Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings greatly enhance the controllability of MAVs if they are properly implemented and can be specifically tailored to obtain increased aerodynamic performance [5]. The most common configurations are the batten reinforced (BR) or perimeter reinforced (PR) structures pictured in Figure 1. a) b) Figure 1. Typical flexible wing MAVs constructed by the University of Florida. a) A batten reinforced aircraft. b) A perimeter reinforced aircraft. Both of these configurations consist of a bi-directional carbon fiber composite skeleton affixed with an extensible membrane skin. Latex was used in previous iterations; however, the material of choice for current MAVs is silicone rubber. Latex degrades quickly and reasonable manufacturing repeatability for the latex skin is not easily accomplished. Innovations provided by Abudaram et al. have allowed silicone to become the preferred material for MAVs as it resists degradation and allows for a consistent, repeatable pre-tensioning during manufacturing [6, 7]. Previous work by Albertani et al. has shown that depending upon the nature of the reinforcement, the wing deformation may alleviate flight loads, as in BR wings, or enhance flight loads, as in PR wings [8]. The BR design incorporates uni-directional carbon fiber which is left unconstrained at the trailing edge of the wing resulting in significant adaptive geometric twist being incorporated into the wing. This wing design has been proven to decrease C Lα, C D, permit wind gust rejection, and delay the onset of stall as compared to a solid wing of similar dimensions [9]. The PR design, which is the main focus of this study, aims to enhance flight characteristics via modifying its contour by using external disturbances on the wing. This is accomplished by creating a perimeter of relatively stiff carbon fiber around an membrane and can be thought of as an aerodynamic twist. Passive adaptations using the PR methodology have been extensively studied for their benefits in the longitudinal stability and have been shown that an inflation of the membrane leads to an increase in C L and decrease of C mα allowing the aircraft to self-correct during flight [10]. The ability of the wing to adapt in this manner allows the aircraft to achieve flight characteristics not possible in rigid wing MAV configurations. 2. MOTIVATION The potential of a flexible wing MAV is limitless as they can be tailored to a variety of situations and new innovations are realized every day. The discovery of the phenomena studied in this paper was realized by Albertani while studying the effects of flexible fixed wing MAVs in a wind tunnel setting [11]. Much research has been done characterizing the effects of PR wings with respect to C L, C D, and C M both with and without the effects of a propulsion system. Early analysis by von Mises expressed interest in the effects of propulsion system interaction with the aircraft fuselage and the importance of the size relationship between the propeller and aircraft [12]. Some of the studies in this work were done with propeller to wingspan ratios much larger than that of a typical MAV; however, the importance of this interaction cannot be neglected. More recently modeling as well as simulation of the propellerwing interaction on the MAV scale has documented the effects of the propulsion system on aerodynamic coefficients, vortical structures, and flow fields over the wing [13 15]. Gamble et al. revealed the effects of wing placement with respect to the propeller on a BR MAV finding that the reaction torque seen by the wing can be as high as 45% of the propeller thrust [16, 17]. The use of a PR membrane wing in conjunction with a powered propulsion system on MAVs International Journal of Micro Air Vehicles

Jason T. Cantrell, Bradley W. LaCroix and Peter G. Ifju 165 presented a variation in the membrane with respect to the deformation seen on either side of the wing. This variation presented a unique opportunity as previous research by the University of Florida indicated that the membrane wing was assistive in nature for longitudinal stability, but the effects on lateral-directional stability were unknown. MAVs are notorious for their subpar low speed flight handling characteristics, and the opportunity to tailor them for low speed, high AOA flight is a desired trait. The aspiration for a perched take off and landing MAV that is passively assisted in these maneuvers was discussed early on by Crowther and is still a topic of much deliberation today [18]. In order to properly understand the overall performance of PR wings with respect to lateraldirectional stability a study of the aircraft via digital image correlation (DIC) and wind tunnel loads testing would provide valuable insight. A study of the flexible membrane s capability across the MAVs flight envelope is necessary to develop an understanding of the interaction between airspeed and lateral-directional stability. Additionally, a more detailed survey of the low speed characteristics must be documented via DIC to properly quantify the variations in the membrane. The details of the flight analysis and the experimental procedure are explained in the subsequent sections. 3. PRELIMINARY RESULTS Previous research by Albertani provided an initial insight into the magnitude of the expected deformations for a PR wing under wind tunnel conditions [11, 19]. Figure 2 shows the effect of an 80 mm propeller on a 15 cm PR latex wing. The free spinning propeller aircraft shows a slight variation of the wing membrane within the noise level of the measurement. The wing with the powered motor shows a difference between the left and right side membrane of approximately 0.3 mm at maximum deflection. a) b) Figure 2. Displacements in the Z (w) out-of-plane direction in mm for PR Wing at AOA = 4, velocity = 8 m/s. The aircraft is positioned such that the viewer is seeing the upper surface of the wing with the leading edge located at the top of the image. a) A wing with the propeller under windmill (free spinning) conditions, b) Propeller under powered conditions. The propeller s rotation is counter-clockwise when viewed from the rear. Using this as a guide for experimental testing, a 30 cm wingspan Zimmerman planform PR silicone wing was developed, constructed and tested statically to determine the maximum deformation of the membrane. The wing was designed using MAVLAB, a wing design tool developed by Claxton at the University of Florida [20]. Table 1 displays the geometric properties of the wing used on the airframe. Table 1. Geometric properties of the aircraft Parameter Measurement Wing area 0.033 m 2 (51.2 in 2 ) Root chord 0.14 m (5.5 in) Wingspan 0.3 m (12 in) Aspect ratio 2.73 Maximum camber 0.053c Maximum camber location 0.3c Rear of propeller to leading edge of wing distance 0.072 m (2.8 in)

166 Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings Figure 3 features a three dimensional top view of the aircraft wing, the polyhedral contour from wing tip, to wing tip and the airfoil profile used for the wing. a) b) Figure 3. Visualizations of the MAVLAB developed wing including a) The wingspan and polyhedral contour and b) airfoil shape for the wing. An 8x8 as well as an 8x4 propeller were outfitted on the test airframe and photographed via DIC at a constant rotational speed of 4500 RPMs. The propeller s rotation is clockwise when viewed from the rear of the aircraft. An 8x4 propeller corresponds to an 8 inch (20.3 cm) propeller with a pitch of 4 inches (10.2 cm). This means that if the propeller was immersed in a conceptual jelly-like substance, it would traverse 4 inches for each revolution. Therefore, a propeller with a higher value for pitch has more curvature and would advance further in a single revolution while also producing more torque. A very small pitch would correspond to a nearly flat propeller. Figure 4 illustrates the dissimilarity between the left and right sides of the aircraft for each propeller. The left side of the wing had a maximum deformation of 1.13 mm while the right deformed a maximum of 0.512 mm, a variation of 0.618 mm between the surfaces. Similar results were seen for the 8x4 propeller as the left side of the wing deformed a maximum of 0.785 mm while the right deformed a maximum of 0.115 mm. As expected the deformations were slightly decreased due to the decreased pitch and thus diminished International Journal of Micro Air Vehicles

Jason T. Cantrell, Bradley W. LaCroix and Peter G. Ifju 167 thrust of the propeller. Having established these theoretical maxima at static, further analysis focused on determining how these variations in the PR wing affect the aircraft s roll moment during flight conditions. b) a) c) Figure 4. Displacement in the out-of-plane direction for a PR wing under loading from a propeller at 4500 RPMs. a) Observed deflection for an 8x8 propeller, b) an 8x4 propeller c) and the DIC set-up used for preliminary testing. 4. EXPERIMENTAL SET-UP 4.1 Silicone Wing Fabrication The following is a step-by-step procedure of the fabrication process used to create the silicone rubber perimeter reinforced membrane wings used in this study. Step 1. A computer numerical control milled wing mold is used to lay-up the wing. A light coat of spray glue is applied. Step 2. A layer of Teflon film is laid on the mold for easy separation of the wing from the mold after the curing process. Step 3. Three layers of [0/90 ] oriented bi-directional prepreg carbon fiber are cut out utilizing a template. Step 4. Silicone sheets are cut for the perimeter reinforced sections and corona treated to insure bonding of the pre-preg epoxy during cure [6]. Step 5. The perimeter of the silicone is sandwiched between the layers of the carbon fiber composite. Step 6. The entire assembly is covered with an additional Teflon layer, vacuum bagged, and cured in a convection oven at 130 C for four hours. A similar process is used for the solid carbon fiber wing fabrication; however, Steps 4 and 5 are neglected because of the absence of a silicone layer. Figure 5 illustrates the step-by-step process used during the construction of these wings.

168 Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings Rigid Carbon Fiber (RCF) Perimeter Reinforced Silicone Membrane (PR) Figure 5. Methodology for the construction of a silicone and carbon fiber composite PR wing. The finished wings are shown at the bottom in the final image. 4.2 Wind Tunnel Set-Up The primary equipment used for testing was the closed-loop wind tunnel with a 0.84 m by 0.84 m test section. The tunnel is capable of producing airspeeds from 2 m/s to 45 m/s and driven by a twostage axial fan. The speed controller is operated by a custom LabVIEW virtual instrument (VI) and verified via a Heise model PM differential pressure transducer connected to pitot-static tube mounted within the test section. A U-shaped model arm extends from the sidewall of the test section and holds the 6 degree-of-freedom sting balance on which the airframe is attached. As seen in Figure 6 the model Figure 6. A schematic of the full wind tunnel setup with important components indicated. International Journal of Micro Air Vehicles

Jason T. Cantrell, Bradley W. LaCroix and Peter G. Ifju 169 arm aligns the aircraft at a positive AOA; however, upon start up of the wind tunnel this model arm rotates the balance/airframe assembly to a 0 AOA. Angle of attack is monitored throughout the AOA sweep through the LabVIEW VI and is frequently verified via a digital protractor accurate to ±0.1. The sting balance, an Aerolab model 01-15, is capable of sensing loads on the order of 0.01 N and provides all force and moment data in an easily readable format through the LabVIEW VI interface [21]. The model assembly consists of the previously assembled silicone wing mounted at three locations on a two layer bi-directional carbon fiber airframe. The airframe is cut to accommodate the sting balance, electronics, and motor. Both an APC 8x4 and APC 8x8 thin electric propeller were used in this set of experiments. These propellers were mounted on an E-flite 400, 740 kv motor and coupled with a Phoenix-45 brushless motor controller. The system is controlled externally via an Astroflight Inc. servo tester and powered by a constant 12V power source. The propeller RPMs are measured using a DT-2234C+ digital tachometer with a resolution of 1 RPM. Figure 7-a shows the model assembly in the wind tunnel while Figure 7-b displays all of the parts used in testing the MAV wings. a) b) Figure 7. a) A close up picture of the aircraft assembly on the model arm. b) An image of all parts used in the testing. These parts are: 1. E-flight 400 motor, 2. Fuselage, 3. Phoenix-45 controller, 4. Servo Tester, 5. Digital Tachometer, 6. 8x4 propeller, and 7. 8x8 propeller. 4.3 Digital Image Correlation (DIC) Set-up Determining the deformation and shape of the wing during testing is the primary objective for this study. This is done through the use of the DIC system, a non-contact full-field shape and deformation technique, developed by researchers at the University of South Carolina [22]. The system uses two Point Grey Research 5-megapixel grayscale cameras to simultaneously capture images of a random speckle pattern painted on the wing. The speckle pattern is achieved by first covering the black silicone with masking tape and painting the exposed carbon fiber entirely with flat black spray paint. The masking tape is then removed and the random speckle pattern is sprayed on using flat white spray paint. The cameras are calibrated with calibration grid composed of a high contrast dot pattern of known diameter and spacing. In this set-up, it was a 9x9 grid of points with a separation of 10 mm. Once calibrated, the system is ready to photograph the wing in the wind tunnel and determine the wing deformation under flight conditions. Reference images of the undeformed wing at various angles of attack are taken and contrasted against images taken of the aircraft under load in the wind tunnel. Images are captured via VIC Snap 2009 and processed via VIC-3D 2009 to determine the deformation of the wing surface. Figure 8 demonstrates the DIC camera setup used in the wind tunnel analysis.

170 Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings Figure 8. Experimental setup for DIC analysis in the wind tunnel. 5. RESULTS 5.1 Full Flight Range Study Roll moment coefficient (C l ) was measured through a range of AOA sweeps from 0 to 20 at 6, 10, and 14 m/s for the aircraft assembly. At each speed the aircraft was tested using two propellers (8x4 and 8x8) and throttle settings (4500 and 6000 RPMs) for the PR wing as well as the rigid carbon fiber (RCF) wing. Roll moment coefficient data was collected in 2 AOA increments and plotted for each scenario. Figures 9 and 10 show some of the general trends seen in this full flight investigation while Table 2 shows the statistical analysis for the aircraft at 6 m/s and each propulsion configuration. The data clearly shows a reduction in C l at the same flight conditions for the PR wing in comparison to the RCF wing. Figure 9 shows C l of the PR silicone wing maintains a lower value versus the RCF wing when equipped with either an 8x4 or 8x8 propeller at 4500 RPMs. An anomaly is seen in the data from 12-16 AOA for the 8x4 propeller as the roll moment coefficient for the PR wing falls below that of RCF wing. This is thought to be a product of the coupling between moments in the sting balance at these conditions as this is not seen at any other point during testing. The 8x4 propeller enhances the effect on the PR silicone wing further than the 8x8 propeller due to the additional torque produced by the 8x8 propeller. This torque results in a larger roll moment on the aircraft which the silicone membrane appears to be unable to compensate for. Increasing the propeller speed results in a similar outcome as both configurations overall effectiveness are reduced with the 8x8 propeller seeing a significantly larger reduction. Table 2. Average C l decrease and percentage difference between the PR and RCF wings. International Journal of Micro Air Vehicles

Jason T. Cantrell, Bradley W. LaCroix and Peter G. Ifju 171 Figure 9. C l vs AOA for 6 m/s flight conditions at a propeller speed of 4500 RPMs. Cl C l Figure 10. C l vs AOA for 6 m/s flight conditions at a propeller speed of 6000 RPMs. When the velocity of the wind tunnel is increased to 10 m/s and then again to 14 m/s similar results are found. Faster flight speeds reduce the percentage of roll compensation for a given roll moment. Therefore, the phenomena seen at low speeds is reduced and the larger pitched propeller negates the advantage far more quickly, since it is producing far more torque. Figures 11 and 12 display this trend for the 8x4 propeller throughout its flight regime. The results show that the ability of the PR wing to compensate for increased propeller induced roll moments is significantly reduced at 10 m/s and is unmeasureable at 14 m/s.

172 Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings Figure 11. C l vs AOA for 6, 10, and 14 m/s flight conditions at a propeller (8x4) speed of 4500 RPMs. Figure 12. C l vs AOA for 6, 10, and 14 m/s flight conditions at a propeller (8x4) speed of 6000 RPMs. 5.2 Detailed Low Speed Study After evaluating the interaction between the silicone membrane and flight speed, further investigation at low speed was necessary. The wind tunnel velocity was reduced to 5 m/s and only the 8x4 propeller was used at 4 different motor speeds (3000, 4500, 6000, and 7500 RPMs). DIC data was recorded for each motor speed at 0, 10, and 20 AOA to visualize the deformation asymmetry under flight conditions. The same procedure was utilized from the full flight study to record C l for every AOA sweep. Figure 13 shows the full investigation and again the data shows the PR wing compensating for the propeller induced roll moment at this velocity. All four propeller speeds yielded similar results which are reported in Table 3. The trend revealed among these findings is that as the aircraft increases its angle of attack the compensation of roll moment actually increased significantly. At 0 AOA the first three speeds were nearly identical; however, as the AOA is increased the PR wing positively differentiates itself from the RCF wing. The average decrease in C l developed a linear pattern at these low speeds and across the spectrum of RPMs reduces C l by an average of 14.7%. International Journal of Micro Air Vehicles

Jason T. Cantrell, Bradley W. LaCroix and Peter G. Ifju 173 Table 3. Average Cl decrease and percentage difference between the PR and RCF wings with an 8x4 propeller at 5 m/s. Figure 13. Cl vs AOA for 5 m/s flight conditions at propeller speeds of 3000-7500 RPMs. In addition to the discoveries revealed by analyzing Cl, DIC data reveals the same results. Figure 14 displays the images compiled from the first three propeller speeds and the PR wing with the propeller a) b) c) d) Figure 14. DIC imagery of the out-of-plane displacement for the aircraft at 10 AOA for a) no propeller, b) 3000 RPMs, c) 4500 RPMs, and d) 6000 RPMs.

174 Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings removed at 10 AOA. DIC reveals that at 10 with no propeller the wing is symmetrical with a maximum deflection of 1.7 mm. When the propeller is powered up, an obvious asymmetry can be spotted on the images, confirming the results of the wind tunnel investigation. At 3000 RPMs the asymmetry between the right and left sides of the wing was 0.5 mm with a maximum deflection of 2.1 mm. This increased to a differential of 0.8 mm and maximum deflection of 2.7 mm at 4500 RPMs. A difference of 1.3 mm with a maximum deflection of 3.5 mm was seen at 6000 RPMs. Further DIC data was collected at 0, 10, and 20 for all four propeller speeds in the low speed realm. Multiple images were collected and averaged via MATLAB post-processing to determine the average deflection at each stage of flight. In order to accurately quantify the deflections, a section view on each side of the wing was selected equidistant from the center to compare. Figure 15 shows where the sections views were taken on the PR wing. Figure 15. Schematic of the section views taken to determine deflection differential. The deflections for each flight condition were averaged across ten images, then the left wing deflections subtracted from the right to quantify the deflection differential. Figure 16 illustrates the DIC data for all ten images taken for a particular trial illustrating it with respect to wing camber as well as purely membrane deflection. The left and right sections of the wing were both averaged and then subtracted from one another to produce a chordwise differential for each angle of attack while varying propeller RPMs. Figure 17 displays the differential along the chord at a selected angle of attack while varying the angle RPMs. In Figure 18, the maximum deflection differential for each data set is shown staggered about each respective data point on the x-axis for improved visibility. A trend is apparent as there is a statistical difference in deflection when varying RPM and holding the PR wing at a set angle of attack. This confirms the results seen in Figure 13 as a larger differential will result in a greater reduction in propeller induced roll moment. Figure 18 shows that as the MAV increases in angle of attack the differential decreases by a statistically insignificant amount. The decrease or steadiness in the differential with respect to angle of attack can be explained by the elastic nature of the silicone rubber membrane. The load on the silicone membrane increases in a nonlinear fashion as the displacement is increased. Therefore, incrementally larger loads are required to displace the membrane further once the membrane reaches a certain displacement. This explains why the margin between the left and right side of the aircraft decreases while the overall deflection is greater as illustrated previously in Figure 14. An increase in overall deflection means a greater force is required as the deflection increases further. Thus, at higher angles of attack, larger forces are required for differentials that appear smaller than those at lower angles of attack. These findings show that there is definitive evidence of the silicone PR wing reducing C l in MAVs at low speeds. International Journal of Micro Air Vehicles

Jason T. Cantrell, Bradley W. LaCroix and Peter G. Ifju 175 a) b) c) d) Figure 16. Chordwise deflection and wing camber for all 10 DIC images on the PR wing at 10 degrees AOA and 6000 RPMs for the a) left wing section deflection, b) right wing section deflection, c) left wing section camber, and d) right wing section camber. Figure 17. Chordwise deflection differential of the PR wing at 10 degrees AOA.

176 Passive Roll Compensation on Micro Air Vehicles with Perimeter Reinforced Membrane Wings a) b) Figure 18. Maximum chordwise deflection differential of the PR wing while varying a) RPMs, and b) AOA. Measurements at the same RPM and angle of attack are artificially staggered to aid the reader. 6. CONCLUSIONS A series of tests were conducted in a wind tunnel setup to determine the effects of a perimeter reinforced membrane wing on flight characteristics of micro air vehicles at a low speed. Preliminary results from previous experiments provided the basis for this study and presented the idea of improvement of lateral-directional flight characteristics through passive means. Digital image correlation and a 6 degree-of-freedom sting balance in a wind tunnel set-up were employed to find the membrane deformation and roll moment coefficients respectively. The following conclusions were drawn from the data collected: 1. In the experiments conducted, an aircraft with a perimeter reinforced membrane wing experienced a reduced propeller induced roll moment compared to its rigid carbon fiber wing counterpart when subjected to the same flight conditions. 2. The reduction in roll moment is dependent upon propeller pitch, flight speed, and propeller RPMs. Larger propeller pitches and increasing flight speeds decrease the roll reduction capability of the wing. Increasing the propeller speed increases the wings roll reduction ability at low flight speeds. 3. The perimeter reinforced wing increased the roll moment reduction capability as the angle of attack increased. The average reduction of the perimeter reinforced wing over the rigid carbon fiber wing at 5 m/s was 14.7% through angle of attack sweep. 4. The wing membranes increase in overall deflection with increasing angle of attack and RPMs. 5. DIC established that the vertical differential between the left and right side surfaces of the membrane wing showed an increase with propeller RPMs while holding angle of attack constant. The inverse is true when holding propeller RPMs constant while varying angle of attack. Perimeter reinforced wings show great potential for low speed, high angle of attack flight on micro air vehicles. There are countless modifications that can be made to the wing and membrane structure that can be implemented to further develop the ability of perimeter reinforced wings to assist in roll compensation of propeller torque. A fundamental understanding of how the propulsion system and aircraft interact has now been established. REFERENCES [1] Ifju PG, Jenkins DA, Waszak MR, et al. (2002) Flexible-Wing-Based Micro Air Vehicles. American Institute of Aeronautics and Astronautics, pp 1 13 [2] Ifju P, Ettinger S, Jenkins D, Martinez L (2001) Composite materials for micro air vehicles. Proceeding for the SAMPE annual conference [3] Jenkins DA, Ifju PG, Abdulrahim M, Olipra S Assessment of Controllability of Micro Air Vehicles. Bristol International RPV/UAV Conference International Journal of Micro Air Vehicles

Jason T. Cantrell, Bradley W. LaCroix and Peter G. Ifju 177 [4] Lian Y, Shyy W (2007) Laminar-Turbulent Transition of a Low Reynolds Number Rigid or Flexible Airfoil. AIAA Journal 45:1501 1513. doi: 10.2514/1.25812 [5] Stanford BK (2008) Aeroelastic Analysis and Optimization of Membrane Micro Air Vehicle Wings. 1 179. [6] Abudaram Y, Rohde S, Hubner J, Ifju P (2013) A Novel Method to Attach Membranes Uniformly on MAV Wings. Composite Materials and Joining Technologies for Composites 7:71 79. [7] Abudaram Y, Ifju P, Hubner JP, Ukeiley L (2012) Controlling Pre-Tension of Silicone Membranes on Micro Air Vehicle Flexible Wings. 50th AIAA Sciences Meeting [8] Albertani R, Stanford B, Hubner JP, Ifju PG (2007) Aerodynamic Coefficients and Deformation Measurements on Flexible Micro Air Vehicle Wings. Experimental Mechanics 47:625 635. doi: 10.1007/s11340-006-9025-5 [9] Waszak MR, Jenkins LN, Ifju P (2001) Stability and Control Properties of an Aeroelastic Fixed Wing Micro Aerial Vehicle. AIAA Atmospheric Flight Mechanics Conference. American Institute of Aeronautics and Astronautics, Montreal, Canada, pp 1 12 [10] Stanford B, Ifju P (2009) Membrane Micro Air Vehicles with Adaptive Aerodynamic Twist: Numerical Modeling. Journal of Aerospace Engineering Vol. 22, N:173 184. [11] Albertani R (2005) Experimental Aerodynamic And Static Elastic Deformation Characterization of Low Aspect Ratio Flexible Fixed Wings Applied To Micro Aerial Vehicles. 1 340. [12] Mises R Von (1959) Theory of Flight. Dover Publications Inc., New York, NY [13] Ageev ND (2011) Numerical investigation of disc-wing MAV with propeller in a wing slot. International Micro Air Vehicles Conference. t Harde Netherlands, pp 27 31 [14] Deng S, Oudheusden BW Van, Xiao T, Bijl H (2012) A Computational Study on the Aerodynamic Influence of a Propeller on an MAV by Unstructured Overset Grid Technique and Low Mach Number Preconditioning. Open Aerospace Engineering Journal 5:11 21. [15] Shkarayev S, Moschetta J-M, Bataille B (2007) Aerodynamic Design of VTOL Micro Air Vehicles. 3rd US-European Competition and Workshop on Micro Air Vehicle Systems (MAV07) & European Micro Air Vehicle Conference and Flight Competition (EMAV2007). Toulouse, France, pp 1 18 [16] Gamble BJ, Reeder MF Experimental Analysis of Propeller Interactions with a Flexible Wing Micro Air Vehicle. 36th AIAA Fluid Dynamics Conference and Exhibit [17] Gamble BJ, Reeder MF (2009) Experimental Analysis of Propeller-Wing Interactions for a Micro Air Vehicle. Journal of Aircraft 46:65 73. doi: 10.2514/1.34520 [18] Crowther WJ (2000) Perched Landing and Takeoff for Fixed Wing UAVs. RTO AVT Symposium. Ankara, Turkey, pp 9 13 [19] Mueller TJ, Kellogg JC, Ifju PG, Shkarayev S V. (2007) Introduction to the Design of Fixed-Wing Micro Air Vehicles: Including Three Case Studies. 289. [20] Claxton D (2007) Development of a Parametric Software Tool for the Design and Manufacturing of Micro Air Vehicles. University of Florida [21] Sytsma MJ (2006) Aerodynamic Flow Characterization of Micro Air Vehicles Using Flow Visualization Methods. 1 152. [22] Sutton MA, Turner JL, Bruck HA, Chae TA (1991) Full-field Representation of Discretely Sampled Surface Deformation for Displacement and Strain Analysis. Experimental Mechanics 31:168 177.