CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > General Forums > Main CFD Forum

SU2 Transonic Flow simulations bad results

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   May 14, 2014, 15:29
Default SU2 Transonic Flow simulations bad results
  #1
New Member
 
Spyros
Join Date: Mar 2014
Posts: 20
Rep Power: 7
S.Kontogiannis is on a distinguished road
Hello everyone,

As a part of my master thesis i will be using SU2 to simulate transonic flow around a wing and a complete aircraft configuration with long term goal to use the adjoint optimizer for optimizing the airfoil shape (and maybe the wing planform).

As a starting validation case I am considering the RAE 2822 case described in http://www.grc.nasa.gov/WWW/wind/val.../raetaf04.html (M=0.729, a=2.31 deg). However my results are not good.

I have tried JST method for the convective fluxes as well as the 2nd order ROE Method and the 2nd order HLLC Method (S-A used for modelling turbulence). All of these methods produce bad quality results in both structured grids I have used. I have used a coarse grid of y+~10 and another intermediate grid of y+~1.

HLLC is diverging and ROE and JST converge in non physical results after many iterations. I should also point out that when I use multi-grid every method diverges. Another "strange" fact is that the coarse mesh produces slightly better results.

I have uploaded a photo of my intemediate grid and some figures showing the results and a typical convergence behaviour.

I would appreciate any help, thanks.
Spyros
Attached Images
File Type: jpg structured_mesh.jpg (95.6 KB, 17 views)
File Type: jpg Coarse_grid_Comparison.jpg (67.7 KB, 22 views)
File Type: jpg Convergence.jpg (49.0 KB, 18 views)
File Type: jpg JST_Comparison.jpg (67.8 KB, 15 views)
Attached Files
File Type: txt turb_SA_RAE2822_2ndUpwind.txt (15.0 KB, 6 views)
S.Kontogiannis is offline   Reply With Quote

Old   May 14, 2014, 15:52
Default
  #2
Senior Member
 
Martin Hegedus
Join Date: Feb 2011
Posts: 500
Rep Power: 15
Martin Hegedus is on a distinguished road
I agree the results do look odd.

That being said, be careful about the experimental results. The tunnel will have slots, or something like it, that will noticeably affect the shock location.

You can also try using Aero Troll (disclaimer, I wrote it) to double check the results. Aero Troll is 2D at this time so it will not get you where you want to go, but it may help along the way.

http://www.hegedusaero.com/software.html

The code is free and is coupled to a Java user interface for building the grids and monitoring the runs.
Martin Hegedus is offline   Reply With Quote

Old   May 14, 2014, 15:57
Default
  #3
Senior Member
 
Martin Hegedus
Join Date: Feb 2011
Posts: 500
Rep Power: 15
Martin Hegedus is on a distinguished road
Oh, are you sure the SU2 results have converged enough? I would suggest that you converge a few runs to as close to machine zero as possible.
Martin Hegedus is offline   Reply With Quote

Old   May 14, 2014, 17:16
Default
  #4
New Member
 
Spyros
Join Date: Mar 2014
Posts: 20
Rep Power: 7
S.Kontogiannis is on a distinguished road
Dear Mr Hegedus,

Thank you very much for your reply. I will check Aero Troll to see if I can be helped. As for the SU2 results, the convergence criteria I used was the default it was supposed to work. I will try to further reduce the convergence criteria, but the weird thing is that as the convergence progressed the results seemed to get worse.

Thank you very much,
Spyros
S.Kontogiannis is offline   Reply With Quote

Old   May 14, 2014, 18:37
Default
  #5
Senior Member
 
Martin Hegedus
Join Date: Feb 2011
Posts: 500
Rep Power: 15
Martin Hegedus is on a distinguished road
I'm not an SU2 expert, so consider my follow commits as being general.

Try running a case without a shock.

In regards to a shock, initially the Euler (inviscid) terms will converge and then the boundary layer develops over time. The interaction of the shock and the boundary layer will possibly create unsteady flow. The unsteady flow will exist until the turbulence model kicks in. So, initially the residual converges, then diverges a little, and then converges again. The SA turbulence model uses vorticity as it's forcing metric. So, vorticity needs to build up before the turbulence model begins to ramp up.

I noticed that your CFL number is 2.5. That's on the slow side. The problem may take about 40K to 100K for just the loads to converge with that setting. So once you are satisfied that the solution is behaving as you expect, try pushing up the number.

I also noticed that you are not using a slope limiter. This may cause problems for flows with a shock (discontinuities). You might need to set a limiter.

Good luck.
Martin Hegedus is offline   Reply With Quote

Old   May 15, 2014, 07:18
Default
  #6
New Member
 
Spyros
Join Date: Mar 2014
Posts: 20
Rep Power: 7
S.Kontogiannis is on a distinguished road
Dear Mr Heredus,

Thank you very much for the advice. The low CFL number was to ensure stability because as I mentioned before in some cases the code diverged. Therefore my first aim is to ensure convergence. The use of the limiter could really improve the result, thank you. I will re-try the case and hopefully i will have better results.

Thank you very much,
Spyros
S.Kontogiannis is offline   Reply With Quote

Old   May 16, 2014, 12:22
Default
  #7
New Member
 
Spyros
Join Date: Mar 2014
Posts: 20
Rep Power: 7
S.Kontogiannis is on a distinguished road
Unfortunately the use of the limiters does not improve the results. In fact using the venkatakrishnan slope limiter and using a CFL in the order of 5-10 (starting from 2.5 and increasing by a factor of 1.1 every 100 iterations) leads to divergence after 1000+ iterations.
S.Kontogiannis is offline   Reply With Quote

Old   May 16, 2014, 13:07
Default
  #8
Senior Member
 
Martin Hegedus
Join Date: Feb 2011
Posts: 500
Rep Power: 15
Martin Hegedus is on a distinguished road
Sorry, other than running a lower Mach number case to gain insight, I'm out of ideas.

You should also post this on the SU2 forum. http://www.cfd-online.com/Forums/su2/

I was thinking that this thread would have been moved there, but it did not happen.
Martin Hegedus is offline   Reply With Quote

Old   May 16, 2014, 13:22
Default
  #9
New Member
 
Spyros
Join Date: Mar 2014
Posts: 20
Rep Power: 7
S.Kontogiannis is on a distinguished road
Dear Mr Hegedus,

Thank you for the advice, i thought that the thread would be automatically set to the SU2 forum. I just moved it.
http://www.cfd-online.com/Forums/su2...tml#post492411
S.Kontogiannis is offline   Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Solver for transonic flow? Martin Hegedus OpenFOAM Running, Solving & CFD 22 December 16, 2015 05:59
combustion simulations from cmcPimpleFoam results openfoammaofnepo OpenFOAM 0 July 9, 2013 08:05
Periodic boundary conditions in 3D Eulerian granular flow simulations dsm FLUENT 4 March 2, 2012 20:04
Bad Cd results applying k-ε models to hydrofoil spk FLUENT 1 December 10, 2011 09:27
How to do simulations for pressible flow Gaurav FLUENT 3 October 9, 2003 17:20


All times are GMT -4. The time now is 01:35.