Experimental Testing of Low Reynolds Number Airfoils for Unmanned Aerial Vehicles by Leon Li A thesis submitted in conformity with the requirements for the degree of Master of Applied Science Graduate Department of Institute for Aerospace Studies University of Toronto Copyright (cid:13)c 2013 by Leon Li Abstract Experimental Testing of Low Reynolds Number Airfoils for Unmanned Aerial Vehicles Leon Li Master of Applied Science Graduate Department of Institute for Aerospace Studies University of Toronto 2013 This work is focused on the aerodynamics for a proprietary laminar flow airfoil for Un- manned Aerial Vehicle (UAV) applications. The two main focuses are (1) aerodynamic performance at Reynolds number on the order of 10,000, (2) the effect of a conventional hot-wire probe on laminar separation bubbles. For aerodynamic performance, pressure and wake velocity distributions were measured at Re = 40,000 and 60,000 for a range of angles of attack. The airfoil performed poorly for these Reynolds numbers due to laminar boundary layer separation. 2-D boundary layer trips significantly improved the lift-to-drag ratio. For probe effects, three Reynolds numbers were investigated (Re = 100,000, 150,000, and 200,000), with three angles of attack for each. Pressure and sur- face shear distributions were measured. Flow upstream of the probe tip was not affected. Transition was promoted downstream due to the additional disturbances in the separated shear layer. ii Dedication To my maternal grandfather, who has always wanted to have a scientist in the family. iii Acknowledgements I would like to thank Professor Philippe Lavoie for his mentorship over the past four years. He has taught me and fostered my interest in experimental studies ever since my undergraduate days. Without his guidance, this work would not have been possible. I must also thank Professor James Gottlieb for his prompt review and valuable feedback for this document. I very much appreciate the friendship and support of my colleagues in the FCET group. I must thank Ronnie Hanson and Jason Hearst for all of their help in teaching me the intricacies of wind tunnel tests. They’ve taught me to appreciate the amount of work and dedication required to obtain good experimental data. A special thanks to Jason for always going to Thai Express with me, five times a week. A special thanks to Luke Osmokrovich for showing me how to take hot-wire data. I would like to thank Dr. Arash Naghib-Lahouti for getting me started on the project, and for his timely and valuable inputs throughout. I would like to thank Rafik Chekiri for all the help he gave in my course works and RAC reports. A special thanks to Heather Clark and David Sutton for providing valuable feedbacks on my thesis draft. Finally, I’ll have to apologize to everyone in the group for forbidding them to turn on the lights in the lab during oil film tests. A special thanks to Jason Liang, who helped with all my CAD draftings. Lastly, I must thank my family and friends for their love and understanding over the many long nights in the past two years. Without their support, I wouldn’t have made it. iv Contents 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 Literature Review 6 2.1 Laminar Separation Bubbles . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.1.1 Formation and Dynamic Behaviour . . . . . . . . . . . . . . . . . 6 2.1.2 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.1.3 Transition Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . 10 2.2 Low Reynolds Number Airfoils . . . . . . . . . . . . . . . . . . . . . . . . 11 2.2.1 History and Categorization . . . . . . . . . . . . . . . . . . . . . . 11 2.2.2 Flow Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.2.3 Flow Control Schemes . . . . . . . . . . . . . . . . . . . . . . . . 15 2.2.4 Experimental Difficulties . . . . . . . . . . . . . . . . . . . . . . . 17 2.3 Oil Film Interferometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.3.1 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.3.2 Practical Considerations . . . . . . . . . . . . . . . . . . . . . . . 25 v 3 Experiment Details 27 3.1 Facilities and Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 3.2 Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2.1 Aerodynamic Force Tests . . . . . . . . . . . . . . . . . . . . . . . 31 3.2.2 Hot-Wire Effect Tests . . . . . . . . . . . . . . . . . . . . . . . . 33 3.2.3 Oil Film Interferometry Tests . . . . . . . . . . . . . . . . . . . . 35 4 Results 37 4.1 Aerodynamic Properties at Low Re . . . . . . . . . . . . . . . . . . . . . 37 4.1.1 Surface Pressure and Wake Velocity Distribution . . . . . . . . . 37 4.1.2 Aerodynamic Coefficients . . . . . . . . . . . . . . . . . . . . . . 45 4.2 Hot-Wire Probe Effects on LSBs . . . . . . . . . . . . . . . . . . . . . . . 49 4.2.1 Pressure Distribution . . . . . . . . . . . . . . . . . . . . . . . . . 49 4.2.2 Separation and Reattachment Locations . . . . . . . . . . . . . . 62 4.2.3 Surface Shear Stress Distribution . . . . . . . . . . . . . . . . . . 62 5 Conclusions and Future Works 66 5.1 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 5.2 Future Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Bibliography 71 vi List of Tables 2.1 Characteristics for surfaces used in OFI, reproduced from Naughton and Sheplak [47] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 3.1 Test matrix for very low Reynolds number performance . . . . . . . . . . 32 3.2 Test matrix for hot-wire probe effects on pressure distribution . . . . . . 34 3.3 Test matrix for oil film interferometry . . . . . . . . . . . . . . . . . . . . 36 4.1 Max. L/D ratio of the different airfoil configurations at Re = 40,000 and 60,000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 vii List of Figures 1.1 Flight Reynolds number spectrum, taken from Lissaman [31] . . . . . . . 3 1.2 Low Reynolds number airfoil L/D, taken from Lissaman [31] . . . . . . . 4 2.1 Structure of a laminar separation bubble, taken from Ha¨ggmark et al. [23] 7 2.2 Effects of long and short bubbles on C distributions. “S” and “R” denote p separation and reattachment respectively. Taken from Gaster [20] . . . . 8 2.3 Anexampleofthe“zig-zag”patternseenindragpolaratReynoldsnumber on the order of 104; taken from Lyon et al. [33] . . . . . . . . . . . . . . . 15 2.4 Controlvolumeusedforthin-oil-filmderivation; reproducedfromNaughton and Sheplak [47] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.5 A basic schematic diagram of an OFI experiment setup. The essential components shown are: model, oil droplet, light source, and camera. This is reproduced from Stimson and Naguib [57]. . . . . . . . . . . . . . . . . 21 2.6 A schematic of the formation of interference patterns due to phase shift between two beams of reflected light; taken from Naughton et al. [46] . . 22 2.7 An original image showing fringe line formation near the attachment point of a laminar separation bubble on an airfoil . . . . . . . . . . . . . . . . . 22 2.8 Exteriorandinteriorparametersforphotogrammetryanalysis, reproduced from Naughton and Liu [45]. . . . . . . . . . . . . . . . . . . . . . . . . . 25 viii 2.9 3-D camera interior parameter calibration block, similar to the one de- scribed in Naughton [44]. . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 3.1 Test section of the wind tunnel . . . . . . . . . . . . . . . . . . . . . . . 28 3.2 Test model with hot-wire traversing system . . . . . . . . . . . . . . . . . 31 3.3 Diagram of oil film test setup . . . . . . . . . . . . . . . . . . . . . . . . 31 3.4 Schematic diagram of the cross section geometries of the two boundary layer trips used. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.5 Size comparison of the two strips to a penny. . . . . . . . . . . . . . . . . 33 3.6 Schematic of the measurement locations for hot-wire pressure distribution effect tests. Each boundary layer profile contains 40 data points; pressure distribution was measured for every three points. . . . . . . . . . . . . . 34 4.1 C distribution for Re = 40,000 without roughness elements . . . . . . . 38 p 4.2 C distribution for Re = 60,000 without roughness elements . . . . . . . 38 p 4.3 Wake velocity distribution for Re = 40,000 without roughness elements . 39 4.4 Wake velocity distribution for Re = 60,000 without roughness elements . 39 4.5 Additional C distribution for Re = 60,000 at 6.5◦ ≤ α ≤ 7.5◦ without p roughness elements; “higher” and “lower” Re denotes slight deviations from the nominal Reynolds number . . . . . . . . . . . . . . . . . . . . . 40 4.6 Additional wake velocity distribution for Re = 60,000 at 6.5◦ ≤ α ≤ 7.5◦ without roughness elements; “higher” and “lower” Re denotes slight deviations from the nominal Reynolds number . . . . . . . . . . . . . . . 40 4.7 C distribution for α = 7.5◦ for three tests at slightly deviated Reynolds p numbers from Re = 60,000 . . . . . . . . . . . . . . . . . . . . . . . . . . 40 ix 4.8 C distribution for Re = 40,000 with rectangular roughness element; the p black dots denote the pressure port downstream of the roughness element; the trailing edge of the roughness element is located 1 mm upstream . . . 41 4.9 C distribution for Re = 60,000 with rectangular roughness element; the p black dots denote the pressure port downstream of the roughness element; the trailing edge of the roughness element is located 1 mm upstream . . . 41 4.10 Wake velocity distribution for Re = 40,000 with rectangular roughness element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 4.11 Wake velocity distribution for Re = 60,000 with rectangular roughness element . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 4.12 C distribution for Re = 40,000 with half-round roughness element; the p black dots denote the pressure port downstream of the roughness element; the trailing edge of the roughness element is located 1 mm upstream . . . 43 4.13 C distribution for Re = 60,000 with half-round roughness element; the p black dots denote the pressure port downstream of the roughness element; the trailing edge of the roughness element is located 1 mm upstream . . . 43 4.14 Wake velocity distribution for Re = 40,000 with half-round roughness el- ement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.15 Wake velocity distribution for Re = 60,000 with half-round roughness el- ement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.16 Trailing edge C vs. C for all Re = 40,000 test cases . . . . . . . . . . . 44 p l 4.17 Trailing edge C vs. C for all Re = 60,000 test cases . . . . . . . . . . . 44 p l 4.18 C at Re = 40,000 for baseline and roughness element cases . . . . . . . . 45 l 4.19 C at Re = 60,000 for baseline and roughness element cases . . . . . . . . 45 l 4.20 C at Re = 40,000 for baseline and roughness element cases . . . . . . . 47 d 4.21 C at Re = 60,000 for baseline and roughness element cases . . . . . . . 47 d x