[tomding0107]

Hi all,

I have some problems with the testcase "rae2822". The original case is the case 6 of RAE2822 and it works well.

But when I am trying to study the case 9 by only modifying the mach number and the angle of attack (0.734 for Mach and 2.79 for angle of attack), the solution can be poorly convergent and not stable at all.

Can anyone give me some advise ? Any helps are appreciated.

[JPBLourenco]

Originally Posted by tomding0107

Hi all,

I have some problems with the testcase "rae2822". The original case is the case 6 of RAE2822 and it works well.

But when I am trying to study the case 9 by only modifying the mach number and the angle of attack (0.734 for Mach and 2.79 for angle of attack), the solution can be poorly convergent and not stable at all.

Can anyone give me some advise ? Any helps are appreciated.

I have the same problem too. Anyone can help?

Thanks,
Joao

[whizkid]

Originally Posted by JPBLourenco

I have the same problem too. Anyone can help?

Thanks,
Joao

You may change Reynolds Number, too. Also I suggest using Roe instead of JST. Are the y+ values below 1?

[JPBLourenco]

Originally Posted by whizkid

You may change Reynolds Number, too. Also I suggest using Roe instead of JST. Are the y+ values below 1?

Hi,

According to AGARD paper, and other papers, the Reynolds Number is almost the some. In case 6, I'm using a mach number of 0.729 and an angle of attack of 2.31 and, for case 9, a mach of 0.734 and angle 2.79.

My y+ is 2.6 (above 1) and the turbulence model is Spalart-Allmaras. I attached my configure file.

thanks for the help.
Joao

Turb_SA_RAE2822.txt

[whizkid]

Originally Posted by JPBLourenco

Hi,

According to AGARD paper, and other papers, the Reynolds Number is almost the some. In case 6, I'm using a mach number of 0.729 and an angle of attack of 2.31 and, for case 9, a mach of 0.734 and angle 2.79.

My y+ is 2.6 (above 1) and the turbulence model is Spalart-Allmaras. I attached my configure file.

thanks for the help.
Joao

Attachment 59800

If you can, I suggest generating a better mesh around the airfoil such that y+ values are below 1 on the surface. There is no wall function for SA in SU2.

[JPBLourenco]

Originally Posted by whizkid

If you can, I suggest generating a better mesh around the airfoil such that y+ values are below 1 on the surface. There is no wall function for SA in SU2.

Yes, I can and I'll do it. Can you explain me better why the value needs to be below 1?

Thanks
Joao

[whizkid]

Originally Posted by JPBLourenco

Yes, I can and I'll do it. Can you explain me better why the value needs to be below 1?

Thanks
Joao

here a discussion on the matter https://goo.gl/ESAKcK

[JPBLourenco]

Originally Posted by whizkid

here a discussion on the matter https://goo.gl/ESAKcK

Thanks for your help. I will check.

Joao

[JPBLourenco]

Originally Posted by whizkid

here a discussion on the matter https://goo.gl/ESAKcK

I have done your suggestion and the solution converges. Now I`m trying to do a shape optimization using continues adjoint but the adjoint solution doesn`t stop and take a long time. Can you help me?

Thank
Joao

[whizkid]

Originally Posted by JPBLourenco

I have done your suggestion and the solution converges. Now I`m trying to do a shape optimization using continues adjoint but the adjoint solution doesn`t stop and take a long time. Can you help me?

Thank
Joao

I suggest looking at some tutorials for that. I believe there are some airfoil optimization cases in the Test Cases repo.