|
[Sponsors] |
August 3, 2016, 13:59 |
Restarting unsteady from steady
|
#1 |
New Member
Anonymous
Join Date: Aug 2013
Location: Europe
Posts: 24
Rep Power: 12 |
I am trying to restart an unsteady simulation from a steady simulation by taking the restart file from the steady simulation and using it as input for the unsteady simulation. That all works.
Unfortunately there seems to be a mismatch between the dimensions of the problem (even though everything in the two .cfg files seems to indicate that their nondimensionalization should be the same). Note that I'm using SU2 v4.0.0. In a different forum post one developer said that capability would be added to restart unsteady from a steady result automatically, has that capability been implemented yet? My main question is: why do the restart files have different scales for the two problems? First line from steady restart file: Code:
"PointID" "x" "y" "Conservative_1" "Conservative_2" "Conservative_3" "Conservative_4" "Pressure" "Temperature" "Pressure_Coefficient" "Mach" 0 9.997500181200000e-01 -3.632896519016437e-05 1.162994388336792e+00 6.778368570881975e-01 9.485600937438987e-02 3.292141898109963e+00 1.236295671066837e+00 1.063028062271972e+00 6.001159900110158e-01 4.824168537699385e-01 Code:
"PointID" "x" "y" "Conservative_1" "Conservative_2" "Conservative_3" "Conservative_4" "Grid_Velx" "Grid_Vely" "Pressure" "Temperature" "Pressure_Coefficient" "Mach" 0 9.997500181200000e-01 -3.632896519016437e-05 1.354554021822148e+00 2.374579822981118e+02 3.107994697341827e+01 3.140370645716729e+05 -6.832973826618208e-05 -1.413937383320905e+00 1.171467609841514e+05 3.012758361261317e+02 2.753944274370143e-01 5.080985192194752e-01 |
|
June 7, 2017, 05:04 |
|
#2 |
New Member
Osman Mirza Demircan
Join Date: May 2017
Location: Ankara, Türkiye
Posts: 29
Rep Power: 8 |
Hi, I realize this is a very old post. However, I have recently started using SU2 myself and I am trying to restart using a converged steady run, like you mentioned.
The converged steady solution has a "solution_flow.dat" file but the unsteady run requires "solution_flow_00000.dat" so I renamed it accordingly. This time the code gives the following error: Initialize Jacobian structure (Navier-Stokes). MG level: 0. Initialize Jacobian structure (SST model). Requesting a negative iteration number for the restart file!! I have checked my config file for a "negative iteration number" that I may have entered. However, I have entered the following and I don't see the problem: % Iteration number to begin unsteady restarts UNST_RESTART_ITER= 1 And if I try to restart another unsteady case directly where it is stopped, it asks for a turbulent restart file which is odd since the code didn't output any in the first run. I would really appreciate it if you could show how you managed to restart and get the iteration going. This is a general comment but anyone who knows how to run unsteady simulations properly using SU2 should write tutorials about unsteady cases for all the SU2 community because it is quite confusing for anyone who works on them for the first time and there are hardly any information on the forums. |
|
June 7, 2017, 18:22 |
|
#3 |
Super Moderator
Tim Albring
Join Date: Sep 2015
Posts: 195
Rep Power: 10 |
Hi Osman,
for 2nd order dual-time stepping you'll need two restart files. Copy "solution_flow.dat" also to "solution_flow_00001.dat" and set UNST_RESTART_ITER=2. Tim |
|
June 7, 2017, 18:51 |
|
#4 |
New Member
Osman Mirza Demircan
Join Date: May 2017
Location: Ankara, Türkiye
Posts: 29
Rep Power: 8 |
Actually, I haven't tried dual time-stepping yet. I used the standard time-stepping for my simulations.
|
|
June 9, 2017, 02:21 |
|
#5 | |
New Member
jit
Join Date: Jun 2015
Location: bangalore
Posts: 25
Rep Power: 10 |
Quote:
I want to use steady solution file for restarting FSI case. I am using DUAL_TIME_STEPPING_2ND_ORDER. UNST_RESTART_ITER= 2 DYN_RESTART_ITER= 2 So, I am providing 3 files(solution_flow_0_00000.dat, solution_flow_0_00001.dat & solution_beam_1_00001.dat) But SU2 asks for one more file "solution_flow_00001.dat" ( if I am using any turbulence model). It takes only 3 files if no turbulence model is used. I just want to understand why it is asking for that one extra file Any suggestions or hints are greatly appreciated. |
||
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Simulating an unsteady phenomenon with a steady model | JochemGrietens | Main CFD Forum | 2 | April 13, 2016 12:51 |
Steady and Unsteady nature | Natali_Beginner | FLUENT | 15 | April 21, 2013 13:43 |
conversion steady to unsteady | Clementhuon | Siemens | 7 | January 26, 2012 05:33 |
Steady needs unsteady. | nico | Main CFD Forum | 0 | September 21, 2007 05:50 |
steady or unsteady? (in dpm) | winnie | FLUENT | 1 | April 28, 2003 12:30 |