|
[Sponsors] |
May 15, 2022, 10:58 |
Not going for second design
|
#1 |
New Member
Zahra
Join Date: May 2022
Posts: 5
Rep Power: 4 |
Hi everyone,
I am learning optimisation with SU2, and for the first step, I ran the tutorial. It finally gave me 35 designs. But when I substitute my own mesh for the tutorial's mesh, I'd just have one design, and after completing the first design, the optimisation will be terminated. I'd appreciate it if someone could give me any advice about this matter. |
|
May 17, 2022, 07:26 |
|
#2 |
Senior Member
bigfoot
Join Date: Dec 2011
Location: Netherlands
Posts: 605
Rep Power: 18 |
Hi,
Do you use the new FADO framework for optimization? example: https://su2code.github.io/tutorials/Species_Transport/ There are many things that can cause the adjoint optimization to fail. First, you have to make sure that the primal is converged. Is this the case? Then the adjoint needs to be converged. Ten the deformation of the mesh should lead to a valid new mesh, if the deformation is too large, the mesh will become invalid. Do you know where in this process you run into problems? For instance, do you get a new, deformed geometry after the first design iteration or not? |
|
May 18, 2022, 10:25 |
Problem with second design
|
#3 |
New Member
Zahra
Join Date: May 2022
Posts: 5
Rep Power: 4 |
Hello,
Thank you so much for your answers. Yes, exactly I initially used the Constrained Shaped Design of a Turbulent transonic airfoil which is here: https://su2code.github.io/tutorials/...ained_RAE2822/. Then without changing anything in the configuration file, I just substitute my own mesh. The deformation file is not generated but it just gives me one successful and converged flow simulation and thereafter it stops and does not continue and go for the second design. And in the file "DESIGNS>DSN_001>DIRECT>log_Direct.out." I have this section: ----------------------------- Solver Exit ------------------------------- All convergence criteria satisfied. +-----------------------------------------------------------------------+ | Convergence Field | Value | Criterion | Converged | +-----------------------------------------------------------------------+ | rms[Rho] | -5.63552 | < -3 | Yes| | Cauchy[CL] | 9.6812e-07 | < 1e-06 | Yes| +-----------------------------------------------------------------------+ -------------------------------------------------------------------------- Many thanks again for your answer and attention. Last edited by Zahra_m; May 18, 2022 at 18:45. |
|
May 18, 2022, 14:06 |
|
#4 |
New Member
Zahra
Join Date: May 2022
Posts: 5
Rep Power: 4 |
Hello,
Thank you so much for your answers. Yes, exactly I initially used the Constrained Shaped Design of a Turbulent transonic airfoil which is here: https://su2code.github.io/tutorials/...ained_RAE2822/. Then without changing anything in the configuration file, I just substitute my own mesh. It seems that it is not failing. It just gives me one successful and valid mesh, which has a tiny difference from the original mesh, and thereafter it stops and does not continue and goes for the second design. And in the file "DESIGNS>DSN_001>DIRECT>log_Direct.out." I have this section: ----------------------------- Solver Exit ------------------------------- All convergence criteria satisfied. +-----------------------------------------------------------------------+ | Convergence Field | Value | Criterion | Converged | +-----------------------------------------------------------------------+ | rms[Rho] | -5.63552 | < -3 | Yes| | Cauchy[CL] | 9.6812e-07 | < 1e-06 | Yes| +-----------------------------------------------------------------------+ -------------------------------------------------------------------------- Many thanks again for your answer and attention. |
|
November 9, 2022, 02:50 |
|
#5 |
New Member
Join Date: Oct 2022
Posts: 1
Rep Power: 0 |
Hello,
I have encountered the same problem. How can I solve it? Design variable (FFD_CONTROL_POINT_2D) number 0. DRAG gradient : 2.66888e-10 ------------------------------------------------------------------------- Design variable (FFD_CONTROL_POINT_2D) number 1. DRAG gradient : 2.66888e-10 ------------------------------------------------------------------------- Design variable (FFD_CONTROL_POINT_2D) number 2. DRAG gradient : 2.66888e-10 ------------------------------------------------------------------------- Design variable (FFD_CONTROL_POINT_2D) number 3. DRAG gradient : 0 ------------------------------------------------------------------------- Design variable (FFD_CONTROL_POINT_2D) number 4. DRAG gradient : 0 ----------------------------- Solver Exit ------------------------------- All convergence criteria satisfied. +-----------------------------------------------------------------------+ | Convergence Field | Value | Criterion | Converged | +-----------------------------------------------------------------------+ | rms[A_Rho]| -12.0357| < -12| Yes| +-----------------------------------------------------------------------+ To adjust the mesh deformation size, adjust DV_ VALUE? At the end of my configuration file run, no DEFORM file was generated and the SU2_DEF program was not called. I would be grateful if anyone could give me any advice on this matter. |
|
Tags |
optimization, su2, tutorial |
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
CFX problems I got in Workbench | Anna Tian | CFX | 13 | July 3, 2017 01:49 |
Sensitivity information for custom design variables | dghate | SU2 | 0 | March 20, 2017 06:39 |
JOB: Sr. Thermal/Mechanical Design Engineer | Dee Dee Dial | Main CFD Forum | 0 | April 1, 2003 18:49 |
Info: Short Course On Thermal Design of Electronic Equipment | Arnold Free | Main CFD Forum | 0 | August 10, 1999 10:18 |
CFD - Trends and Perspectives | Jonas Larsson | Main CFD Forum | 16 | August 7, 1998 16:27 |