Airfoil Tools 0012

[Sara Legath]

HelloI have recently done a study to compare my simulated results with my calculated results on the naca 0012 airfoil. This study was to have a highly studied airfoil and simulate it and compare it with my calculated results. I have come across a few interesting finds along the way too.

I have used an airfoil add-on in Fusion 360 to create the airfoil profile, extruded it, and then imported it as a .STEP file. I ran a simulation once with the standard mesher and then ran 2 simulations with the hex dominant meshes

For the results, the expected results were 0 N for lift and got -0.01 N for the standard mesher, I think this is pretty good. For the hex dominant, I got 0.1 N of lift which is a little less accurate but not too bad(I have not been able why there is a big difference like this, my guess is that when you have a 3d airfoil it is different than inputting the CL and CD in the lift and drag equation, maybe someone else has an idea why this is happening?).

The present paper showcases the predicting ability of an in-house 2D/ Quasi-3D steady state Ice Crystal Accretion Tool (ICAT) applicable for both heated and un-heated surfaces. The previously existing code for unheated surfaces, has been extended to heated scenarios with the inclusion of: 1) coupling with solid conduction model 2) inclusion of advanced models for crystal melting, water film modeling, sticking and erosion. The results obtained from ICAT are verified against the experimental results of heated NACA0012 airfoil, conducted in the icing wind tunnel of TU Braunschweig as part of MUSIC-haic project. ICAT predictions are found to be well in agreement with the ICI physics, which is proven with the various parameters addressed in this paper, such as tunnel temperature, ice crystal temperature, inlet melt ratio, heating power, etc.

In this step, we will import the coordinates of the airfoil and create the geometry we will use for the simulation. Begin by downloading this file here and saving it somewhere convenient. This file contains the points of a NACA 0012 airfoil.

Before we launch the design modeler, we need to specify the problem as a 2D problem. Right click and select Properties . In the Properties of Schematic A2: Geometry Window, select Analysis Type > 2D . Now, double click to launch the Design Modeler. When prompted, select Meters as the unit of measurement.

First, we will create the geometry of the airfoil. In the menu bar, go to Concept > 3D Curve. In the Details View window, click Coordinates File and select the ellipsis to browse to a file. Browse to and select the geometry file you downloaded earlier. Once you have selected the desired geometry file, click to create the curve. Click to get a better look at the curve.

Next, we need to create a surface from the curve we just generated. Go to Concepts > Surfaces from Edges. Click anywhere on the curve you just created, and select Edges > Apply in the Details View Window. Click to create the surface.

Now that the airfoil has been generated, we need to create the meshable surface we will use once we begin to specify boundary conditions. We will begin by creating a coordinate system at the tail of the airfoil - this will help us create the geometry for the C-mesh domain. Click to create a new coordinate system. In the Details View window, select Type > From Coordinates . For FD11, Point X , enter 1.

Click to generate the new coordinate system. In the Tree Outline Window, select the new coordinate system you created (defaulted to Plane 4 ), then click to create a new sketch. This will create a sketching plane on the XY plane with the tail of the airfoil as the origin. At the bottom of the Tree Outline Window, click the Sketching tab to bring up the sketching window.

The first action we will take is create the arc of the C-Mesh domain. Click . The first click selects the center of the arc, and the next two clicks determine the end points of the arc. We want the center of the arc to be at the tail of the airfoil. Click on the origin of the sketch, making sure the P symbol is showing

For the end points of the arc, first select a point on the vertical axis above the origin (a C symbol will show), then select a point on the vertical axis below the origin. You should end up with the following:

To create the right side of the C-Mesh donain, click . Click the following points to create the rectangle in this order - where the arc meets the positive vertical axis, where the arc meets the negative vertical axis, then anywhere in the right half plane. The final result should look like this:

Now, we need to get rid of necessary lines created by the rectangle. Select Modify in the Sketching Toolboxes window, then select . Click the lines of the rectangle the are collinear with the positive and negative vertical axises. Now, select the Dimensions toolbox to dimension the C-Mesh domain. Click , followed by the arc to dimension the arc. Assign the arc a value of 12.5. Next, select . Click the vertical axis and the vertical portion of the rectangle in the right half plane. Also assign the horizontal dimension a value of 12.5.

Next, we need to create a surface from this sketch. To accomplish this, go to Concept > Surface From Sketches. Click anywehere on the sketch, and select Base Objects > Apply in the Details View Window. Also, select Operation > Add Frozen . Once you have the correct settings, click . The final step of creating the C-Mesh is creating a surface between the boundary and the airfoil. To do this, go to Create > Boolean. In the Details View window, select Operation > Subtract . Next, select Target Bodies > Not selected , select the large C-Mesh domain surface, then click Apply . Repeat the same process to select the airfoil as the Tool Body . When you have selected the bodies, click

Selecting the Airfoil Body

Because the C-Mesh domain and the airfoil overlap, once you click in the vicinity of the airfoil ANSYS will select the C-Mesh domain but give you the option of selecting multiple layers Select the layer that corresponds to the airfoil and the airfoil will be highlighted.

In the final step of creating the geometry, we will break up the new surface into 4 quadrants; this will be useful for when we want to mesh the geometry. To begin, select Plane 4 in the Tree Outline Window, and click . Open the sketching menu, and select . Draw a line on the vertical axis that intersects the entire C mesh. Trim away the lines that are beyond the C-Mesh, and you should be left with this:

Next, go to Concepts > Lines from Sketches. Select the line you just drew and click Base Objects > Apply , followed by . Now that you have created a vertical line, create a new sketch and repeat the process for a horizontal line that is collinear to horizontal axis and bisects the geometry.

Once you click , you'll notice that the geometry is now composed of two surfaces split by the line we selected. Repeat this process to create 2 more projections: one projection the line left of the origin onto the left surface, and one projecting the right line on the right surface. When you're finished, the geometry should be split into 4 parts.

GRIDSNote: significantly finer grids for this NACA 0012 case can be found from the:Numerical Analysis of 2D NACA 0012 Airfoil Validation Case page.The quantities of interest for comparison are as follows:Lift coefficient (CL) vs. angle of attack (alpha) (for 0 Drag coefficient (CD) vs. CL for 0 Surface pressure coefficient (Cp) vs. x/c (for alpha = 0, 10, 15)Surface skin friction coefficient (Cf) vs. x/c (for alpha = 0, 10, 15)where amax is the maximum angle of attack at which the CFD yields steady-stateresults with no force oscillations.There are experimental data available for validation, but it should be recognizedthat two-dimensional experiments are extremely difficult to achieve,particularly at higher angles of attack approaching stall.Therefore, the experimental data provided here should be used with that in mind.There are no known surface skin friction data available for this case for validation;however, CFD predictions of this quantity are still of interest.Experimental curves for comparison at Re=6 million (essentially incompressible conditions)are given below. As can be seen, there are some differences between various experimental results.In particular, these differences occur near stall where the experiments areno doubt very far from being two-dimensional any more.A useful paper that discusses many NACA 0012 experiments conducted over the years is:McCroskey, W. J., "A Critical Assessment of Wind Tunnel Results for the NACA 0012Airfoil," AGARD CP-429, July 1988; also NASA TM 100019, October 1987 particular, it is important to note that experimental drag coefficient levels aregreatly affected by tripping the boundary layer at Reynolds numbers in this range. Forcomparing with "fully turbulent" CFD drag results, tripped experimental data aremore appropriate than untripped.

The Abbott and von Doenhoff data (Abbott, I. H. and von Doenhoff, A. E., "Theoryof Wing Sections," Dover Publications, New York, 1959) were not tripped.The Gregory and O'Reilly data (Gregory, N. and O'Reilly, C. L., "Low-SpeedAerodynamic Characteristics of NACA 0012 Aerofoil Sections, including the Effectsof Upper-Surface Roughness Simulation Hoar Frost," R&M 3726, Jan 1970)were tripped, but were at a lower Re of 3 million. Lift data are notaffected too significantly between 3 million and 6 million, but drag data are (e.g.,according to McCroskey, tripped CD,0 at Re=3 million is about 10% higherthan tripped CD,0 at Re=6 million).The Ladson tripped data (Ladson, C. L., "Effects of Independent Variation of Mach andReynolds Numbers on the Low-Speed Aerodynamic Characteristics of the NACA 0012Airfoil Section," NASA TM 4074, October 1988, ) appear to be the most appropriateof these data sets for comparison with fully turbulent CFD forces at Re=6 million. For comparing with surface pressure coefficients, data of Ladson et al (Ladson, C. L., Hill, A. S., and Johnson, Jr., W. G., "PressureDistributions from High Reynolds Number Transonic Tests of an NACA 0012 Airfoil inthe Langley 0.3-Meter Transonic Cryogenic Tunnel," NASA TM 100526, December 1987, )do not appear to resolve the leading edge upper surface pressurepeak well. Gregory and O'Reilly CP data (at Re=3 million) appear to be better resolved.The Gregory and O'Reilly data also show some noticeable differencesfrom the Ladson et al pressure data levels over the front half of the airfoilat alpha=10 and 15.It is believed that the Gregory data are likely more two-dimensional and hence more appropriatefor CFD validation of surface pressures.

3a8082e126