Rotary Wing Decelerators for Miniature Atmospheric Entry Probes

Transcription

1 20th AIAA Aerodynamic Decelerator Systems Technology Conference and Seminar<BR> 4-7 May 2009, Seattle, Washington AIAA Rotary Wing Decelerators for Miniature Atmospheric Entry Probes First Austin Howard 1 and Donald Elger, Ph.D.,P.E. 2 University of Idaho, Moscow, ID, Rotary wing decelerators were investigated as part of a miniature entry probe feasibility study. Previous technology demonstrations have shown that traditional low speed aerodynamic deceleration systems for this size class of probe can be relatively massive and risky. Rotary wing decelerators offer many potential advantages for miniature entry probes. The objectives of this project were to investigate the potential advantages of the rotary wing decelerator technology, assess aerodynamic modeling tools, and quantify performance of duel wing decelerator concepts using wind tunnel experimentation. Test models included a variety of two wing rotors with diameters of 152 mm or less in low speed air flows less than 20 m/s. The models with the lowest aspect ratio, 0.7 and 1.1, and highest solidity, 0.39 and 0.56, were observed to have the largest average thrust coefficients, 1.13 and 1.32 respectively. The tip speed ratio and the blade pitch were also identified as important parameters related to the magnitude of the Thrust Coefficient. Blade Element Momentum theory demonstrated the ability to predict average rotor thrust within 30% when compared with experimental observation for the majority of the cases considered. Nomenclature A b - Area of the blades A r - Area of the rotor disk C T - Thrust Coefficient c - Chord length D - Rotor diameter N - Number of rotor blades R - Radius of the rotor disk Re b - Reynolds number based on blade geometry Re t - Reynolds number based on rotor geometry s - Span length T Thrust force U - Free stream flow velocity - Angle of attack - Tip speed ratio (or TSR) - Angular velocity of the rotor (in radians per second) - Fluid density σ - Solidity γ - Kinematic viscosity I. Introduction INCE the 1960 s, NASA and other space agencies around the world have been sending instrumented S atmospheric entry probes throughout our solar system as part of an effort to understand planetary evolution and search for signs of life. Each destination in our solar system offers unique engineering challenges, requiring a unique set of design solutions. Parachutes, aeroshells and retro-rockets have been successfully applied as decelerator systems on atmospheric probes. As scientific instruments become smaller and require less power and designers develop new mission architectures to address the next generation of science questions; new technologies must be developed and old technologies adapted to meet the engineering demands of future missions. This project 1 Graduate Student, Mechanical Engineering Dept., AIAA Student Member 2 Professor, Mechanical Engineering Dept. 1 Copyright 2009 by the, Inc. All rights reserved.

2 considers the rotary wing decelerator as one of the potential enabling and/or enhancing technologies for future science architectures, including miniature atmospheric entry probes, lander networks, and precision targeted landers. For the purpose of this study, a rotary wing decelerator is defined as one or more rotating wings that retard the vertical velocity of the free falling payload. Unpowered rotary wing decelerators achieve steady rotation rates through a condition where net aerodynamic loads acting on a body result in a rotating state of equilibrium. Rotary wing decelerators have notable advantages when compared with parachute technology including lower sensitivity to lateral winds 1, mitigation of deployment problems 2, ability to execute soft landings 3, self scanning instrument platform 4, and higher falling velocities 4. A wind tunnel experiment was designed that could provide data on the feasibility of rotary wing decelerators for small probes, geometric influences on rotor performance and reference data for aerodynamic math model validation and development. II. Background Many studies and experiments have been conducted beginning in the 1960 s that investigate the feasibility and performance of rotary wing decelerators for vehicles ranging from the Apollo entry capsule to payloads less than 1kg in mass. NASA Researchers in the late 1960 s and early 1970 s researched a decelerator concept for entry vehicles that featured rotary wings that would perform as a decelerator during entry and later act as helicopter rotors close to the ground. This design could give astronauts the ability to maneuver and land the entry vehicle safely. Analysis and wind tunnel test were performed for subsonic, transonic and supersonic flow regimes 3. A research team, Nadal-Mora, Sanz-Andrés and Cuerva, is currently developing a rotary wing sounder for the purpose of measuring atmospheric properties around airports. Figure 1 shows a sketch of the rotary wing decelerator configuration as tested in vertical and horizontal wind tunnels. The unique feature in this design that makes modeling a challenge is the low aspect ratio of the wings, which is less than one. The wings are designed to wrap around the body and this requirement limits the wing dimensions. Body dimensions are limited by gyroscopic stability in rotation. Nadal-Mora, Sanz-Andrés and Cuerva developed a semi empirical aerodynamic model for their low aspect ratio, low Reynolds number rotor in autorotation 5, 6. The most effective blade pitch angles were found to be between 2 to 8 degrees. The team also developed a stability Figure 1. Rotary Wing Decelerator model for a rotary wing decelerator 7. Another research team, Bartz and Miklosovic, performed work to develop a decelerator technology capable of delivering supplies to troops on the ground in a variety of weather conditions and surface topologies. Six 247.7mm model helicopter blades were mounted to a rotating hub and placed in a horizontal wind tunnel. Pitch on the blades was adjusted as well as the angle of the trailing edge flaps. The flaps effectively added additional camber to the airfoil profile. This additional camber was shown to greatly impact the performance of the rotor. In one case the thrust performance was increased by 193% while in the same case the advance ratio was decreased by 67% (decrease in advance ratio corresponds to increase in rotational velocity). Bartz and Miklosovic reported thrust coefficients of up to 2.8 for wings with 2% camber. A relationship between thrust coefficient and rotational velocity was also noted; in the experiments, the higher the angular velocity of the rotor resulted in higher thrust coefficients 1. III. Experiment Design The experimentation for this research project was performed in the University of Idaho horizontal low speed wind tunnel. The models were mounted on a shaft with a bearing which allowed free rotation of the model. A force balance was used to measure lift and drag forces, an infrared tachometer was used to measure rotation rates and a hot film anemometer was used to measure air velocity. A stroboscope was also used as a redundant method to measure rotor velocity and to aid in flow visualization. See Figure 2 for a summary of the test setup. A control panel allowed the operator to control the velocity in the wind tunnel which induced rotation of the model attached to the force balance. The free stream velocity was gradually increased until the rotor began to spin. Velocity was increased by finite increments (1 or 0.5 m/s). Once 10 steps or a rotor rotation rate 6000 RPM was reached (whichever came first), the process was repeated in reverse until the rotor stopped spinning. Yaw angles 2

3 were fixed at 0 o for all experiments. A limited number of flow and model configurations were selected for photography and flow visualization. Figure 2. Experiment overview (looking from the side) A. Experiment Models A total of six dual wing rotor models were constructed for the wind tunnel experiment. The span-wise dimensions were limited by a test section blockage requirement of less than 0.10 and a preferred blockage ratio of 0.05 as suggested by Barlow, Rae, & Pope 8. However, the wings of model 2 did not meet this requirement with a test section blockage ratio of Each wing was constructed from mm thick, flat brass plates. a) b) c) d) Figure 3. Overview of wings 1, 2, 3 and 4 3

4 Wing 1 was considered the baseline design in its 10 o pitch configuration. It had an aspect ratio of 2 and was tested in blade pitch configurations of 5 o, 10 o, and 20 o. See Figure 3a). Wing 2 was designed to compare the effects of wing aspect ratio/solidity on thrust performance. The aspect ratio was set at 3 and the pitch was set at 10 o. See Figure 3b). Wing 3 was also designed to compare the effects of wing aspect ratio. The aspect ratio was set to 0.71 which is 5, 6, 7, 9 similar to the aspect ratio of the wings used in the research conducted by Mora et al. Wing 3 was tested in a 10 o pitch configuration. See Figure 3c). Wing 4 was designed to study the effect of a flexible trailing edge. A piece of paper was taped to the trailing edge of wing 1 which nearly doubled the chord length of the wing resulting in an aspect ratio of The rigid component of the wing was set to a pitch of 5 o during testing. See Figure 3d). 1. Performance Parameters During the wind tunnel experiment, velocity, drag force and rotor rotation rates were recorded. measurements were reduced to two representative dimensionless values: thrust coefficient and tip speed ratio. The thrust coefficient is defined by Eqn. (1). All (1) Note that the thrust force for this research has the opposite polarity when compared to propeller thrust force. The term thrust is used to describe the total drag of the rotor to avoid confusion with the drag of the rotor blades and to achieve consistency with standard mathematical modeling notation. The tip speed ratio is defined by Eqn.(2). (2) Another important parameter is the solidity, σ, of the rotor disk and is defined by Eqn. (3). (3) Reynolds number is another dimensionless constant that is commonly used to describe characteristic flow. For this experiment two Reynolds number definitions can be used. One defined by rotor s swept area and the other defined by the rotor blade geometry. All results in this paper are referenced to the wind tunnel test section velocity instead of Reynolds number because of the small range of Reynolds numbers observed. The range of Reynolds numbers tested, based on the swept area of the rotor was 2.36x10 4 to 1.37x10 5. The range of Reynolds numbers tested, based on the rotor blades, was 1.43x10 4 to 1.21x10 5. The Reynolds number based on the rotor geometry is defined by Eqn. (4) and the Reynolds number based on the rotor wing is defined by Eqn. (5). (4) (5) 2. Notation Each test configuration was assigned a designation that was used as a reference. For example w1p10 refers to wing 1 with a pitch of 10 o. Similarly w2p10 refers to wing 2 with a pitch of 10 o. 4

5 IV. Experiment Results 1. Thrust Data was collected primarily in the range of 5 to15 meters per second. A very wide range of thrust coefficients were observed ranging from 0.19 to The magnitude of the thrust coefficient of w1p20, w1p5, and w3p10 decreased with increasing velocity while the opposite behavior was observed for w2p10. The thrust coefficient of w4p5 was observed to be relatively constant throughout the velocity ranges tested. Wing 4 with a pitch of 5 o showed the highest average thrust coefficient of See Figure 4 for a comparison of thrust coefficients. The two wings with the highest average thrust coefficient, w4p5 and w3p10, also had the largest chord lengths. The high values of chord length resulted in the highest solidity and lowest rotor aspect ratios of the samples considered. Due to apparatus design, the test fixture was only capable of two wing geometries and therefore solidity and aspect ratio were varied together. Figure 4. Summary of rotor thrust coefficient versus free stream velocity The lowest thrust was produced by w1p20 which had the lowest average thrust coefficient of 0.34 and minimum thrust coefficient of See Table 1 for a comparison of mean, min and max thrust coefficients for all configurations. Table 1. Ranking of thrust performance Description Average Thrust Max Thrust Min Thrust Coefficient Coefficient Coefficient w4p w3p w1p w2p w1p w1p

6 2. Tip Speed Ratio Another performance parameter of interest is the tip speed ratio. It is a ratio of the tangential and normal velocity components at the tip of a rotor blade. The tip speed ratio can be used to determine the angle of incidence of the flow. The angle of attack and local airfoil lift and drag can be estimated by using knowledge of the rotor geometry and the local blade pitch. High values of tip speed ratio indicate a lower angle of incidence and contribute to a lower angle of attack. The tip speed ratio is defined by Eqn. (6). (6) Wing 2 with a pitch of 10 o had the highest average tip speed ratio of 5.42, and maximum tip speed ratio of The lowest value of tip speed ratio was encountered by w3p10 with an average value of The tip speed ratio versus wind tunnel velocity is plotted in Figure 5. See Table 2 for a comparison of recorded tip speed ratios for all configurations. Upon comparison of measured tip speed ratios to the thrust performance summarized in the previous section, no global trends were observed. A plot of thrust coefficient versus tip speed ratio is shown in Figure 6. Description Figure 5. Summary of rotor tip speed ratio versus wind tunnel velocity Table 2 Ranking of average tip speed ratio performance Average Tip Speed Ratio Max Tip Speed Ratio Min Tip Speed Ratio w2p w1p w4p w1p w1p w3p

7 Figure 6. Thrust coefficient vs. tip speed ratio 3. Summary of Results Table 3 summarizes the geometric, thrust and speed parameters. The two configurations with the highest solidity (w3p10 and w4p5) also had the lowest aspect ratio and the largest thrust coefficients. Description Chord (m) Table 3. Summary of geometry, thrust coefficient and tip speed ratio Span (m) Dia (m) Solidity Aspect Ratio Ave. Ang. of Attack ( ) Ave. Thrust Coefficient Ave. RPM Ave. Tip Speed Ratio w1p w1p w1p w2p w3p w4p

8 4. Flow Visualization Flow visualization was performed during the testing of three configurations: w1p20, w2p10, and w3p10. The objective of the flow visualization effort is to observe flow separation and span-wise flow in the boundary layer of the rotor blades. Thread streamers were taped to the top and bottom of the rotary wings with thin pieces of transparent tape. A stroboscope and a flash camera were used to capture the behavior of the threads in the boundary layer of the wing. a) b) c) Figure 7. w1p RPM, w2p RPM, w3p RPM Wing 2 with a pitch of 10 o was observed to have the most uniform flow pattern over the entire wing. Six threads were distributed evenly over the chord and span of the wing. Each thread oriented itself at nearly the same angle with the flow. See Figure 7. Wing 1 with a pitch of 20 o was observed to have a different flow pattern. The threads taped to the leading edge were observed to orient themselves almost parallel to the radial direction of the wing. This suggests that the centrifugal forces are dominant and that flow separation may be occurring in this region. The threads taped to the trailing edge of the wing show similar behavior to that of the threads on wing 2. See Figure 7. The third configuration, w3p10, had similar behavior on the leading edge of the wing. Along the middle of the chord of the wing, the flow behaved similar to the trailing edge of w1p20. The threads on the trailing edge of w3p10 were almost completely aligned with the chord-wise direction. See Figure 7. V. Comparison of Experimental Data with Predictive Math Models Two classic rotor aerodynamic modeling techniques were applied and compared to the experimental results in order to understand the applicability of the modeling techniques for the rotor geometries and flow regimes considered. The two techniques applied were Blade Element Theory (BET) and Blade Element Momentum Theory (BEM). Both theories are described in several text books 10, 11, 12. A custom Matlab script was developed for the BET model. For the BEM model, a previously developed wind turbine performance code was applied. The BEM software used was WT_Perf v3.10 which was developed by the National Renewable Energy Laboratory under the U.S Department of Energy. The software is available on the web at the address: last accessed July The airfoil 8

9 aerodynamics used in the performance predictions were documented by Rosen and Seter 13, who researched low speed airfoil aerodynamics as part of a study of Samara wing aerodynamics. For many of the wind tunnel configurations analyzed, the BEM model produced less than 20% average error. The cases where the BEM failed to produce predictions with lower error than measurement uncertainty were the cases with the lowest thrust coefficients. The custom BET code did not produce representative predictions for most of the rotor models considered. Table 4 includes a summary of model error and measurement uncertainty for all the test configurations. Each of the aerodynamic models listed in Table 4 used 100 rotor segments in the simulation. Ave. Thrust Coeff. Table 4. Summary of analysis Measur. Uncertainty Ave. % BEM % Error (WT_Perf ) Ave.% Max% BET % Error Ave.% Max% w1p w1p w1p w2p w3p w4p VI. Conclusion A. Influence of Geometric Features on Rotor Thrust The following list represents a summary of the observed effects that the geometric features of the various configurations had on the rotor performance. The following observations are referenced to the baseline configuration, w1p10 (except for no. 4 below which was referenced to w1p5). 1. Both increasing and decreasing pitch from the nominal value of 10 o for wing 1 reduced the thrust coefficient and the tip speed ratio. a. Increasing the blade pitch to 20 o reduced the thrust coefficient from 0.87 to 0.34 and the tip speed ratio from 4.06 to b. Reducing the blade pitch to 5 o reduced the thrust coefficient from 0.87 to 0.76 and the tip speed ratio from 4.06 to Increasing aspect ratio/reducing solidity reduced the magnitude of the thrust coefficient. 3. Decreasing aspect ratio/increasing solidity increased the magnitude of the thrust coefficient and reduced the magnitude of the tip speed ratio. 4. Adding flexible flaps to the trailing edge of w1p5 (referred to as w4p5) increased the average thrust coefficient from 0.76 to 1.32 and increased the average tip speed ratio from 1.80 to B. Assessment of Rotary Wing Modeling Techniques Two modeling techniques were investigated and applied: Blade Element Theory, and Blade Element Momentum Theory. Out of the modeling techniques BEM provided the most accurate predictions. Most models using BEM predicted thrust performance with less than 30% error with exception of w1p20 which produced predictions with 39% error. It was observed that BET does not produce accurate performance predictions for the rotor configurations tested. 9

10 Acknowledgments A special thanks to Tony Colaprete (NASA Ames Research Center) who provided the opportunity for this research project and who inspired this study, and to the Idaho Space Grant Consortium for their consistent support, advice, and connections to the people and resources need to complete this study. References 1 Bartz, J., & Miklosovic, D. S., An Experimental Analysis of Camber Effects of a 6-Bladed Flapped Autorotational Aerodynamic Decelerator, 17th AIAA Aerodynamic Decelerator Systems Technology Confrence. Monterey, CA Karlsen, L., Borgstrom, D., & Paulsson, L., Aerodynamics of a Rotating Body Descending From the Separation Position of an Artillery Munition Shell, 11th AIAA Aerodynamic Decelerator Systems Technology Conference Levin, A. D., & Smith, R. C. An Analytical Investigation of the Aerodynamic and Performance Characteristics of an Unpowered Rotor Entry Vehicle, NASA TN-D4537, Crimi, P. Analysis of Samara Wing Decelerator Steady State Characteristics, Journal of Aircraft, Vol. 25, Pg , Nadal-Mora, V., Sanz-Andrés, Á., & Cuerva, Á., Experimental Investigation of an Autorotating-Wing Aerodynamic Decelerator System, 18th AIAA Aerodynamic Decelerator Systems Technology Conference and Seminar, Nadal-Mora, V., Sanz-Andrés, Á., & Cuerva, Á., Model of the Aerodynamic Behavior of a Pararotor, AIAA Journal of Aircraft, Vol. 43, Pg , Nadal-Mora, V., & Sanz-Andres, A., Stability Analysis of a Free-Falling Pararotor, AIAA Journal of Aircraft, Vol. 43 No. 4, Pg , Barlow, J., Rae, W., & Pope, A., Low-Speed Wind Tunnel Testing. Wiley, Mora, V. N., Aerodynamic behavior of Atmospheric Soundings in Airport Surroundings, Doctoral Thesis. Madrid, Spain: Polytechnical University of Madrid, Johnson, W. Helicopter Theory, Princeton University Press, Stepniewski, W. Z., & Keys, C., Rotary Wing Aerodynamics, Dover, Burton, T., Sharpe, D., Jenkins, N., & Bossanyi, E., Wind Energy Handbook. Wiley., Rosen, A., & Seter, D., Vertical Autorotation of a Single-Winged Samara, Transactions of the ASME, Vol. 58, Pg ,