[kmandar]

Dear SU2 developers,

Congratulations for the new release and many thanks for making SU2 available as open source.

I have been using SU2 for my shape sensitivity research. In this regard, I wanted to make sure that the flow solution obtained by using a second order upwind flux scheme indeed gives second order results. After a lot of test-case-hunting, I got the reference by Vassberg and Jameson (2010), who have created high quality NACA0012 grids, specifically to perform grid convergence study.

I have used the same grids to run SU2 and to get the lift and drag coefficients. I am not getting a second order rate of convergence (p) for the lift coeffcient (see the attached figures) although the rate of convergence for the drag coeffient is slightly better than 2. I have used 2nd order Roe's scheme. The analysis is for Mach 0.5 at an an angle of 1.25 degrees. I have also attached the input file which gives other details of the analysis. These results are compared to the same analysis done in Fluent and FLO82 code used by Vassberg and Jameson.

I have used SU2 Release 4.1.1 and the direct_differentiation.py script to generate results.

Was such as study done to verify the second order scheme in SU2? Can you please share the results and details?

Can the convergence be better if I tune any parameters in the input file?

Thanks,
Mandar