Learning Forum

[Rushi_7]

Hello,

I am simulating an external flow around a rocket to find the drag and lift coefficients against the angle of attack. I have a few questions:

1. Should I use SST k-w or Spalart Allmaras model? My free stream velocity is 15m/s
2. As I want the variance of drag/lift coefficients versus the angle of attack, should I keep the reference values of the geometry same for all design points?

[Keyur Kanade]

Use sst kw. Please see mesh requirements for sst kw. You can keep reference values same.
Please search this forum and you will find many posts related to turbulence model and drag/lift.
Regards Keyur
How to access Ansys Online Help Document
How to show full resolution image
Guidelines on the Student Community
How to use Google to search within Ansys Student Community

[Rushi_7]

I tried the simulation with both SST k-w and Spalart Allmaras for 0 angle of attack. Ideally the lift should be as close to 0 as possible. The result from the Spalart Allmaras is more close to zero than that of SST k-w. Hence the confusion in selecting the model
As the angle of attack changes, the frontal area would also change. So won't we get a wrong drag coefficient if we keep the reference area same?

[aitor.amatriain]

Are both simulations fully converged? Have you performed a mesh-independence analysis?
I agree with,SST k-w is the best turbulence model for 99% of the cases. The other 1% is just for flows with strong curvature or swirl (Reynolds Stress Model may perform better in these cases, even though it is usually numerically unstable) or if the user has reference results with a specific model tuned and additional simulations have to be performed for comparison.
The reference area is always the same and equal to the one at zero angle of attack.

[Rushi_7]

Thank you for your reply So I should use the frontal area for both lift and drag coefficients? Or frontal area for drag and some other reference area (Probably the one projected perpendicular to the flow) for lift?

[aitor.amatriain]

Frontal area for both

[Amine Ben Hadj Ali]

NASA Glenn website has a good reference for the reference areas which shall be used. For wings I general use chord length time depth for other I usually use projected area or frontal area. In all cases I always present the raw data and then then refer to the coefficients and highlight which area I am referring to!

[Amine Ben Hadj Ali]

Thanks for the valuable efforts!

[aitor.amatriain]

Thank you too!

[Rushi_7]

If I am using SSTk-w, what y+ value should I target? Note that my free stream velocity is 15m/s

[aitor.amatriain]

If possible, y+<1 for maximum accuracy. If not, y+ around 25 is acceptable

[Rushi_7]

How much would be the difference in the result for these y+ cases?

[Amine Ben Hadj Ali]

Run and quantify the difference :) Always good to keep low yplus but important here to keep the BL well resolved with at least 15 or more cells! Also the streamwise resolution of the boundary is of great relevance for evaluating drag and lift forces ( I called it sometimes pixelation)