ANSYS FLUENT Project
- 1. Flow Over Clean and Loaded Wing Using ANSYS FLUENT Andrew D’Onofrio Unversiti Teknologi PETRONAS Background: Aircrafts generate lift due to a pressure gradient that forms on both sides of the wings as the airfoil deflects the surrounding air. A positive pressure forms on the lower surface of the airfoil while a "suction" pressure forms on the upper surface of the wing. The pressure differences between the upper and lower surface of the airfoil, the pushing of the air down over the trailing edge of the airfoil, as well as changes in velocity are what enable the airfoil to create lift. Airfoils operating at low Reynolds numbers (Re) and high angles of attack (AOA) have been shown to lead to boundary layer separation. The boundary layer separation occurs on the upper surface of the airfoil, which dramatically alters the pressure gradient around the airfoil. This change in pressure gradient has serious repercussions and can even lead to the stalling of the aircraft. An aircraft typically stalls when the angle of attack goes beyond a critical point, resulting in the loss of most if not all of "suction" pressure. This causes the lift of the aircraft to begin to decrease, drag to increase, and the aircraft to become less stable. Large boundary layer separation can be attributed to high angles of attack or due to airfoils with a load such as a fuel tank or missile on the lower surface of the wing. Problem Statement: The loading of a wing with underneath missile significantly changes flow characteristics. Therefore it is problematic to see the actual flow characteristics around the wings. In this project, CFD simulation is to be used to replicate flow over the two types of airfoils: Clean/Unloaded Wing and Loaded Wing. Objectives: - To simulate the flow field structure over clean wing - To assess the change in the separation and wake structure over the wing. - To simulate the flow field structure over loaded wing
- 2. Airfoil Model: The airfoil to be tested is NACA 4412. This high lift-high performance airfoil has a camber that is 4% of the chord, with the max camber located at 4% of the chord. The thickness of the airfoil is 12% of the chord, and the chord has a length of 250 mm. Fig. 1) Airfoil Cross Section The airfoil is to be tested at angles of 0°, 5°,10°, 15°, and 20°, to represent angles typically employed by fixed wing aircraft. Missile Model: The external load to be studied is a missile with a rounded frontal nose and geometry as shown below. Fig. 2) Missile Cross Section Methodology: Before starting my project I needed to learn how to use the software I was assigned. ANSYS Fluent is a Computational Fluid Dynamics (CFD) tool capable of carrying out physical modeling of fluid flow, Chord length, c = 250 mm Location of max camber = 0.4c Max camber = 0.04c y/x = 0.2 Location of max thickness = 0.3c
- 3. turbulence, heat transfer, and chemical reactions for industrial applications. I learned the basics of the software through online tutorials and trial and error. Through the tutorials and further research I created a method to construct the geometry of the airfoil. The method involves using an online airfoil generator to create the coordinates in excel as a .txt file. Fig. 3) Airfoil Generator - NACA 4412 at an AOA of 150 Those coordinates are then imported to FLUENT using the 3D Curve feature. After the coordinates are imported I created the volume fluid sketch as shown below.
- 4. Fig. 4) Volume Fluid @ 100 The shape of the volume fluid was created as per industry standard; the inlet is curved to reflect the curvature of the leading edge of the airfoil. After the airfoil sketch and volume fluid sketch were created the sketches were transformed into lines before using a Boolean operation to subtract the airfoil from the volume fluid. This completes the model, and the next steps involve preparing the model for the mesh step. In order to create a smooth uniform mesh it is required to use projections to divide the volume fluid. This is done by sketching lines on a new plane and then transforming the lines into projections. The geometry step is now finished. Fig. 5) Projections @ 100 The mesh step was quite troublesome for the airfoil and was completed after trial and error. Meshing the airfoil required specifying the number of divisions for each projection as well as the walls of the volume fluid. Finding the optimal number of divisions is important as too few divisions results in an inaccurate mesh and too many divisions takes up too much computing power and time. After dividing up the projections, a bias was added to each projection to focus the mesh around the critical areas. Finally the faces were mapped before generating the mesh.
- 5. Fig. 6) Mesh @ 00
- 6. After the mesh was created I specified the different areas of the volume fluid into: inlet, outlet, airfoil-wall, and wall. Fig. 7) Zoomed in View of Mesh @ 00 Fig. 8) Named Selections of Model @ 00
- 7. After generating the mesh, I repeated this procedure for angles of attack of 0°, 5°,10°, 15°, and 20°. This same procedure was used in creating the model with load. In the setup stage the inlet velocity was specified as 7.3 m/s, the fluid as air, and the simulation was then ran for 500 iterations to generate the results. i.) Projections on Airfoil w/ LoadFig. 9) Projections on Airfoil w/ Load
- 8. Results: Fig. 10) Set-Up: Velocity Contours around 20° Fig. 13) Pressure Contour - 0° Fig. 11) Velocity Streamline - 0° Fig. 12) Velocity Streamline - 0° Fig. 14) Pressure Contour - 0°
- 9. Fig. 16) Velocity Streamline - 10° Fig. 15) Velocity Streamline -10° Fig. 16) Pressure Contour - 10° Fig. 17) Pressure Contour - 10°
- 10. Fig. 19) Velocity Streamline -15° Fig. 20) Velocity Streamline -15° Fig. 21) Pressure Contour - 15° Fig. 22) Pressure Contour - 15°
- 11. Fig. 23) Velocity Streamline -20° Fig. 25) Pressure Contour - 20° Fig. 26) Pressure Contour - 20° Fig. 24) Velocity Streamline -20°
- 12. Discussion: When undertaking this project meshing the airfoils proved to be a very challenging problem. ANSYS Fluent is very sensitive and in order to get a precise and accurate mesh each model had to be projected differently. From finding the optimal number of divisions, to determining the correct bias to use, to refining the mesh with sizing and mapping, a great deal of time was spent working on the meshing stage. Due to meshing issues Imran and I decided to split the work up in a fair way as we were working as a team. Since I had already spent most of my time trying to perfect the meshes I took on that responsibility while Imran worked on generating the missile geometry attached to the airfoil at various angles of attack. We completed our tasks- I successfully meshed the clean airfoils and Imran attached the missile to the airfoils. These being the most difficult parts, we figured it would be simple to replicate the meshes on the loaded airfoils. Unfortunately, after working on the loaded airfoils for a number of days we encountered a number of unforeseen problems which did not arise during the prior models. I believe that due to our different methods of creating the geometry, further methods at later stages seemed to be incompatible. This caused many issues and delays, which ultimately hindered me from being able to extract results from the loaded wing. The models are created, but as has been the reoccurring problem throughout this project the mesh stage simply would not work as desired. Imran and I predicted that the addition of the load would result in increased forces upon the airfoil and would lower the angle of attack at which boundary layer separation would occur. Fig. 27) Loaded Airfoil Mesh @ 0°
- 13. Results: My data shows that increasing the angle of attack from 0° to 20° in increments of 5° creates favorable lift conditions in terms of an increase in velocity along the upper surface of the airfoil as well as increased pressure gradients along both the upper and lower surfaces. An AOA of 15° shows signs of boundary level separation starting to occur at the trailing edge of the airfoil. This separation becomes worse at an AOA of 20° with lines of vortices starting to occur otherwise known as a vortex sheet. This vortex sheet indicates that fluid is in opposite directions which results is massive energy losses and it is very likely the aircraft will stall if the angle is increased much more. Fig. 28) Loaded Airfoil Mesh @ 10°
- 14. References: 1) Al-Kayiem, Hussain H. Visualization of the Flow Field and Wake over Clean and Under-loaded NACA4412 Airfoil. Rep. Tronoh, Perak: Universiti Teknologi PETRONAS, n.d. Web. 17 July 2014. 2) Hu, Hui, and Masatoshi Tamai. "Bioinspired Corrugated Airfoil at Low Reynolds Numbers." Journal of Aircraft 45.6 (2008): 2068-077. Web. 17 July 2014. 3) Richards, Scott, Keith Martin, and John C. Cimbala,. ANSYS Workbench Tutorial – Flow Over an Airfoil. Rep. Penn State University, 17 Jan. 2011. Web. 17 July 2014. 4) Sleigh, Andrew, Dr. "Boundary Layers." An Itntroduction to Fluid Mechanics. University of Leeds, n.d. Web. 17 July 2014. 5) "Airfoils and Lift." Airfoils and Lift. The Aviation History On-Line Museum, n.d. Web. 16 July 2014. 6) Flow over an Airfoil - Part 1 - Ansys Fluent 14 Tutorial. Dir. Pavan Mehta. Perf. Pavan Mehta. Youtube. N.p., 9 Nov. 2013. Web. 17 July 2014. <http://www.youtube.com/watch?v=LAIB7qK-9pE>. 7) Flow over an Airfoil - Part 2 - Ansys Fluent 14 Tutorial. Dir. Pavan Mehta. Perf. Pavan Mehta. Youtube. N.p., 9 Nov. 2013. Web. 17 July 2014. <http://www.youtube.com/watch?v=FQK51-cb-78>.