MAE 571 Applications of CFD Problem Set #7

2D NACA 63-3-418 Airfoil Lift and Drag Applications of CFD concepts learned in class

Due date: 4/16/2015 Your objective for this assignment is to construct a 2D mesh using the strategies discussed in class to solve the flow around a NACA 63-3-418 airfoil. The airfoil geometry has been provided, as well as the parameters for setting up the far-field domain.

At Re = 3.0*10^6 the lift and drag curves for this airfoil are given in the table and figures below. This data is taken from Abbotts Theory of Wing Sections 1959. You must build a CFD model to address the following:

1. Iterative Convergence: Iterative convergence indicates that as the discrete equations are iterated, some computed quantities of interest approach a fixed value. For example for the flow around an air foil iterative convergence will be obtained if properties like lift or drag coefficient does not change with further iteration.

a. Develop a monitoring strategy that demonstrates full iterative convergence for each angle of attack. This means monitoring Lift and Drag at a minimum, other choices are Moment, shear stress at the surface, continuity, mass flux through the domain, etc.

b. How does the convergence behavior change with varying angle of attack, particularly approaching stall?

2. Turbulence Model: The choice of turbulence model can significantly impact the models ability to predict parameters that depend strongly on viscous and turbulent stresses in the flow. For an airfoil problem, this means that the models drag prediction will vary significantly with turbulence model.

a. Evaluate both the standard k-epsilon model and the Spallart-Allmaras (S-A) model when applied to this problem.

b. What are the strengths and weaknesses of each model as applied to this problem? c. How does the choice of turbulence model impact the time required to solve the

problem? 3. Consistency with Experimental Data: Experimental data is the observation of the "real

world" in some controlled manner. By comparing the CFD results to experimental data, one hopes that there is a good agreement, which increases confidence that the physical system can be accurately modeled by the CFD code.

a. Evaluate the performance of the CFD at each data point shown in the table below using each of the two turbulence models.

b. How accurately does the model predict lift? c. How accurately does the model predict drag? d. How accurately can you predict flow separation and stall?

Angle of Attack Section cl Section cd -12 -1.03 0.019 -9 -.85 0.0133 -6 -.4 0.0087 0 .4 0.006 6 1.0 0.009 12 1.4 ~0.021 20 1.3 Fully stalled

Assumptions and Problem Setup 1. The test data corresponds to Re = 3*10^6. 2. The airfoil chord is 1.5 meters, you will need to scale the mesh accordingly once

imported into Star-CCM+. 3. Also you will need to compute the inlet velocity.

Challenges and expectations:

4. It is very difficult to predict drag accurately. An accurate drag prediction depends strongly upon the turbulence model and the mesh, and even then significant errors are to be expected (on the order of 50% of the measured value). Be careful when setting up the problem.

5. Lift predictions will be very accurate will not depend strongly on the turbulence model. 6. Stall predictions will depend on both the mesh quality, particularly in the boundary layer,

and the turbulence model setup. Again, take care when setting this up. 7. Wall y+ is critical when using the S-A turbulence model; it must be less than 1

everywhere.