Flow Past NACA 0012 Airfoil Test

Flow past a NACA airfoil test CFD Project 1 Report

Arthur Saw Sher-Qen 0301339

Bachelor of Engineering (Mechanical) School of Engineering Taylor’s University MEC4513 Computational Fluid Dynamics

14 October 2012

School of Engineering, Taylor’s University

Arthur Saw Sher-Qen(0301339)

Flow past a NACA airfoil test

Contents 1.0 Abstract 2.0 Introduction 2.1 Background 2.2 Project objectives and problem statement 3.0 Methodology 3.1 Geometry 3.2 Meshing 3.3 FLUENT setup 4.0 Results 4.1 Velocity 4.2 Pressure 4.3 Coefficient of lift and drag 5.0 Conclusion 6.0 References 9.0 Appendix

[School of Engineering, Taylor’s University]

MEC4513 Page 3 3 3 3 4 4 5 6 7 7 8 9 9 10 11

Page 2



MEC4513

1.0 Abstract This project main focus is to use computational fluid dynamics as a method of simulation for a study of an airfoil deign with relation to its angle of attack and also to explore the capabilities of the FLUENT software as a CFD tool. To reduce the complexity and computational time of the model, various assumptions are made such as no surface roughness and inviscid flow. The results are lift and drag coefficients which will be compared to published experimental data to validate the results. Also of interest are the velocity and pressure profiles of each angle of attack which will demonstrate the changes in lift produced at different angles of attack. 2.0 Introduction 2.1 Background An airfoil is generally known as the shape of an airplane wing. What happens is when air passes across an airfoil the airfoil generates am effect called lift. When the airfoil slices through the fluid medium known as air, the air is ‘split’ and it flows over the top and the bottom of the airfoil. On a flat plate or a symmetrical airfoil placed horizontally or at zero angle of attack, nothing will happen apart from producing drag due to the resistance between the airfoil and the air particles; the velocity of air across the top and bottom of the airfoil is the same. As such there is no pressure difference to cause any movement or force as according to Bernoulli’s principle (Figure 16) [7]. A force will be generated when there is a pressure difference between the top and bottom section of the airfoil. If the pressure at the bottom is higher than the top due to the velocity of air passing above the top of the airfoil being higher than the bottom; then a force called lift will be generated pushing the airfoil upwards; if the opposite happens, that is the pressure at the top is higher than the bottom then the force called negative lift or downforce will instead push the airfoil downwards [1], [2]. Airfoil design and sections have come a long way since the earliest serious documented work in the late 1800s. Generally, it is known that a flat plane is able to produce lift when placed at an angle; the lift generated differs with the angle of attack; this is because by angling the plane it forces the air to act with more force on one side of the plane thus producing negative or positive lift depending on the angle. However, it was suspected that curvature shapes, which bear some resemblance to one of nature’s ultimate flying ‘machines’, birds are also able to perform the job and more efficiently too. This is because for curvature shapes, even at zero angle of attack, the velocity of the flow is already different between the top and bottom of the airfoil thus producing lift at low angles of attack [1]. 2.2 Project objectives and problem statement In the past understanding of the flow on an airfoil requires construction of expensive prototypes and an expensive wind tunnel. It is cheaper for a small scale design, however to go full scale the cost was enormous and time consuming as a full scale airfoil for example an airplane is huge in design and thus consumes a lot of resources to build and operate.


Page 3



MEC4513

However, with advances in computational technology it has become easier to model and study the design of airfoils with regards to various parameters, angle of attack being one of them. This also forms the basis of this project which is to study the flow, coefficient of lift and subsequently different angles of attack on a fixed airfoil and also to explore and the about the potential applications of computational fluid dynamics in modeling various applications. An angle of attack is defined as the angle where the air at a certain velocity meets the airfoil. The angle is measured from the chord line of the airfoil. As the angle of attack increases, the coefficient of lift increases as well up to a certain point, above it is a phenomenon known as stalling in which the airfoil loses all lift; however, this project will not cover stalling. It is to be noted that with the increase in angle of attack the drag will also increase [3]. In this project, the angle of attack is fixed at 6° and 10° with all other conditions as standard using a NACA 0012 airfoil design. The NACA naming sections represent the camber, max camber and max thickness of the chord line based on the design of a parabola on the chord from the leading edge to the max chord line and back [5].

Figure 1: An airfoil section with airfoil terminology explained [3] 3.0 Methodology 3.1 Geometry The airfoil section basic outline is first outlined in solidworks as it provides batter control of sketches. It is then loaded into the design modeler in Ansys to be created as a surface. The C mesh is also designed in the design modeler module to create an operating boundary for the airfoil. The airfoil is designed in a XY or two dimensional domain for study, this is because we are only interested in the flow at the top and bottom section of the airfoil and also for simplicity.


Page 4



MEC4513

Figure 2: The relative size and layout of the airfoil in meters. As can be seen from the sketch, the NACA 0012 airfoil is symmetrical in design and thus will only generate lift at an angle. 3.2 Meshing To reduce complexity and size of the project, both 6 and 10 degree angle of attack airfoil will share the same geometry and mesh model in FLUENT. A 2D mesh model is created separated into four sections so that proper bias can be applied. The mesh is a 12.5 meter radius at the C section while the rectangular section also shares the same size. The bias of the mesh is toward the center of the C mesh domain as we are interested in the flow around and on the airfoil, as such by biasing the mesh toward the center it will provide a more accurate result at the area of interest. inlet outlet

airfoil

Figure 3: Mesh As can be seen from the grid, the element size is heavily biased or concentrated toward the airfoil in the middle as it is the point of interest and should provide better results. The mesh size is set at 15000 elements based on Cornell’s tutorial as a guide [6]. While a [School of Engineering, Taylor’s University]

Page 5



MEC4513

more refined mesh (more elements) will produce a more accurate result; however, it will also increase the complexity of the domain and will also increase the computational time. As such to keep it simple, all simulation done in this project will retain this mesh. 3.3 FLUENT Setup FLUENT is selected as the computational software for this project, the mesh is loaded into FLUENT for study and processing. The following settings is changed at the FLUENT launcher to ensure a faster processing time; double precision is used so the results are more accurate and parallel processing is selected with two processes as the machine utilizes a dual core processor so the processing time can be reduced. i)

Solver As the operating medium of the airfoil is air, the type of solver chosen is density based instead of pressure based, as later we are interested to also look at the pressure distribution around the airfoil. The time is set as steady flow and gravity disabled for simplicity. Other settings are left as default.

ii) Models For simplification, the flow is assumed to be inviscid, that is without viscosity. This is assuming an ideal air condition to be able to get the best result. Nothing else is changed as the others are not considered in this study. iii) Materials The fluid medium is air and taken at a constant density of 1kg/m3. This is an ideal rounded up value of the air density and the solid, is left as default aluminum. iv) Boundary Conditions Since the study is only focused on the airflow, no other boundary conditions have been set apart from the velocity components at X and Y direction for the inlet. This is because of the angle of attack; thus the velocity is split into X and Y components with cosine (angle) for X and sinus (angle) for Y. While the outlet is set as a pressure outlet at 0 gage pressure so that there can be a flow from the inlet to the outlet due to pressure difference. The airfoil is set as a wall as that is the solid layer the flow has to pass over and it is assumed ideal, no slip condition. v) Solution The solution method is set to use second order upwind method. While this method may take more computational time but it produces a more accurate result. The convergence of the solution is determined based on the scaled residuals of the continuity and the momentum (velocity of X and Y) equations. The solution convergence is set at 1e-6 as the accuracy level. When all the residuals have at least reached this level together the solution is [School of Engineering, Taylor’s University]

Page 6



MEC4513

considered to be converged. Since the mesh is biased toward the airfoil which is the wall, better results will be obtained at that area. Angle X component (m/s) Y component (m/s) 6 0.9945 0.1045 10 0.9848 0.1736 Table 1: Angle of attack and the relative X and Y components for inlet boundary condition

4.0 Results The aerodynamic results especially the coefficient of lift are compared to the works of Chris C. Critzos, Harry H. Heyson, and Robert W. Boswinkle, Jr [4] on their work of the Aerodynamic characteristics of NACA 0012 airfoil. The other sections of the results such as velocity magnitude, pressure coefficients and etc are compared between the different angles of attack just for study and comparison purposes. 4.1 Velocity

Figure 4: 6° velocity vector

Figure 5: 10° velocity vector As can be seen from the velocity vectors, at 6° angle of attack, the velocity does not have much change between the top and bottom section of the airfoil, as such the lift generated will be lesser than that of the 10° angle of attack. In which it is noticed that that is a significant different between the velocity at the top and bottom section as such there will be more pressure difference to generate lift.


Page 7



MEC4513

4.2 Pressure

Figure 6: 6° pressure contour

Figure 7: 10° pressure contour From Figure 6 and 7 it is more clearly illustrated that the 10° angle of attack is capable of producing a higher angle of attack. As it is noted that the pressure at the bottom of the 2nd airfoil is clearly higher that the top as compared to the 1st airfoil. So Bernoulli’s principle would dictate that at 10° attack angle more lift would be produced due to higher pressure [7]. Although, it is noted that the tip or leading edge of the airfoil the pressure is the highest, this is to be expected as that point interacts with the air first before the other parts of the airfoil.

Figure 8: Graph of pressure coefficient at 6° angle of attack


Page 8



MEC4513

Figure 9: Graph of pressure coefficient at10° angle of attack 4.3 Coefficient of lift and drag Angle 6 10 Table 1: Lift coefficient

Project value -0.005 0.629

Experimental value 0.48 0.91

Angle 6 10 Table 2: Drag coefficient

Project value 0.041 0.054

Experimental value 0.02 0.05

From the comparison between the project value and the experimental value, it is clear that lift coefficients are quite far off the mark while the drag coefficients are closer to the experimental values. These errors can be attributed to the test conditions that were applied. The experimental values were obtained from a wind tunnel testing with surface roughness and low turbulence wind tunnel system. While this project assumed ideal conditions; inviscid flow and no surface roughness or gravity applied. Also a finer mesh system could produce a more accurate result. While perhaps mistakes in setting up the boundary conditions or mesh could also lead to errors in the simulation. However, as can be seen from the data, as the angle of attack increases, the accuracy increases as well. 5.0 Conclusion In conclusion, based on the original objectives of this project to explore the use of computational fluid dynamics to model various applications and to study about the aerodynamic characteristics of an airfoil; the objective has been met. By the use of the software FLUENT, it is possible to model the airfoil in various angles of attack and determine its lift, drag, flow/ velocity around the airfoil. This proves the usefulness of CFD modeling in the aerospace applications allowing engineers to test any airfoil in any condition without resorting to expensive prototypes and wind tunnel testing at least in the initial stage.


Page 9



MEC4513

This project also proves the assumption that an increase in the angle of attack of an airfoil its coefficient of lift will also increase as well as the drag partly due to more surface area being exposed to the ‘attacking’ air. Also if accuracy in relative to real world applications is of a concern, then it is recommended not to include so many ideal assumptions as the test result may vary quite a lot depending on the condition. As such, for example it is better to use laminar flow instead of inviscid and set the viscosity according to the Reynolds number of the testing condition. 6.0 References 1. Adg.stanford.edu (n.d.) Airfoil History. [online] Available at: http://adg.stanford.edu/aa241/airfoils/airfoilhistory.html [Accessed: 10 Oct 2012]. 2. Boeing.com (n.d.) What is an airfoil?. [online] Available at: http://www.boeing.com/companyoffices/aboutus/wonder_of_flight/airfoil.html [Accessed: 11 Oct 2012]. 3. Centennialofflight.gov (n.d.) Angle of Attack. [online] Available at: http://www.centennialofflight.gov/essay/Dictionary/angle_of_attack/DI5.htm [Accessed: 12 Oct 2012]. 4. Critzos, C. et al. (1955) AERODYNAMIC CHARACTERISTICS OF NACA 0012 AIRFOIL SECTION AT ANGLES OF ATTACK FROM 0° TO 180°. [report] Washington: NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS, p.23. 5. Desktop.aero (n.d.) Airfoil Geometry. [online] Available at: http://www.desktop.aero/appliedaero/airfoils1/airfoilgeometry.html [Accessed: 12 Oct 2012]. 6. Confluence.cornell.edu (2011) ANSYS WB - Airfoil - Problem Specification Simulation - Confluence. [online] Available at: https://confluence.cornell.edu/display/SIMULATION/ANSYS+WB+-+Airfoil++Problem+Specification [Accessed: 09 Oct 2012]. 7. Hyperphysics.phy-astr.gsu.edu (1998) Pressure. [online] Available at: http://hyperphysics.phy-astr.gsu.edu/hbase/pber.html [Accessed: 14 Oct 2012].


Page 10



MEC4513

7.0 Appendix

Figure 10: 6° angle of attack solution convergence

Figure 11: 10 ° angle of attack solution convergence

Figure 12: 6° angle of attack full velocity vector

Figure 13: 10° angle of attack full velocity vector


Page 11



MEC4513

Figure 14: 6° angle of attack full pressure coefficient

Figure 15: 10° angle of attack full pressure coefficient

Figure 16: Bernoulli’s principle


Page 12



MEC4513

Graph 1: Graph of angle of attack vs coefficient of lift [4]


Page 13



MEC4513

Graph 2: Graph of angle of attack vs coefficient of drag [4]


Page 14

Flow Past NACA 0012 Airfoil Test

Recommend Documents