[Diptaru]

I am conducting an aerodynamic analysis of a 2D airfoil, specifically examining the lift and drag coefficients at 10 different angles of attack (AOA), ranging from 0 to 20 degrees. I am using a C-H type computational domain, fixing the airfoil position and changing the velocity components (Sine_AOA and Cosine_AOA) to define the angle of attack direction.

I initially simulated each of the 10 AOA points separately without using a parametric set. However, I also attempted a parametric study by treating Sine_AOA and Cosine_AOA as input parameters (before the first or base calculation is done) and the lift and drag coefficients as output parameters (before the first or base calculation is done). After completing the first simulation, I added the remaining 9 AOA points as multiple design points by providing different Sine_AOA and Cosine_AOA values for these points. However, I noticed that the results for lift and drag obtained from the parametric study differed from those obtained from individual simulations at each AOA point.

My questions are as follows:

Is the procedure I used for the parametric study—varying the sine and cosine components of AOA to calculate lift and drag—correct?

Will I be able to extract contours from different design points (in this case, different AOA points) if I conduct a parametric study?

I need to simulate 28 different airfoil geometries at 10 different AOAs, resulting in a total of 280 individual simulation runs. The parametric study approach could save time, but since it did not yield matching results, is there an alternative approach that would be more reliable?

Note: Since I am using a structured mesh, it is preferable that the airfoil remains fixed, without any rotational changes, when adjusting the AOA.



Please help me out of this issue as early as possible.

This is my contact dipta.sarkar36@gmail.com if any details are required.

[Hamidzoka]

Hi,
- The procedure itself seems ok. Hower, there are items to be checked.How much is the difference between parametric and non-parametric study? If it is small, it is not an issue. But in case this difference is big, a thorough check should be conducted: Is the domain size properly selected so that outlet flow streamlines are parallel with zero pressure gradient? if it is not the case, results are not reliable and vary from one iteration to another. Do you use the same grid size and distribution? are solver settings identical? Check boundary consitions types and values. Check material properties and pressure reference value in both cases. Check if you are applying right parameter values and dimensions. i.e. degree or radian? mm or m? Pa or bar? sometimes in automated processes this happens. Do they have the same iteration number?

- As long as your solution files are saved you can access them later for further postprocessings. It may need a small coding depending on your software.

Best