CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > ANSYS > CFX

Problems with initialization of new runs

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

Like Tree1Likes
  • 1 Post By Gert-Jan

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   March 21, 2023, 03:38
Default Problems with initialization of new runs
  #1
New Member
 
Lukas Bellmann
Join Date: Mar 2023
Posts: 8
Rep Power: 2
LBellmann is on a distinguished road
Hello there

I am simulation a Fan-Stage with Nacelle in a farfield using Ansys CFX. My goal is to simulate it under an angle of attack. What I did is I first did a run with no angle of attack and then wanted to initialize the next runs (with AOA) with the results of the first run. Unfortunately the next run took the boundary conditions of the one I tried to initialize with and not the ones set in CFX-Pre so there was no angle of attack seen in the Post-results.
I am running my simulations on a cluster and I start them via batch files. I also tried to change boundary conditions using CCL-files but nothing helps. Even the solver manager says that the boundary conditions are set as I want to but as I said there is nothing seen in the results.
By the way, the AOA is big enough so it should definitely make a difference.

Thanks for your time. I hope someone got an idea
LBellmann is offline   Reply With Quote

Old   March 21, 2023, 04:27
Default
  #2
Senior Member
 
Gert-Jan
Join Date: Oct 2012
Location: Europe
Posts: 1,742
Rep Power: 25
Gert-Jan will become famous soon enough
Please share the output file. To upload it, use 'Go Advanced' below in your screen.
Gert-Jan is offline   Reply With Quote

Old   March 21, 2023, 04:49
Default
  #3
New Member
 
Lukas Bellmann
Join Date: Mar 2023
Posts: 8
Rep Power: 2
LBellmann is on a distinguished road
I've deleted the convergence history in the out-file to match the maximum data size.
In the Out-file you can see that velocity components in x and z direction were defined to create an angle of attack.
Attached Files
File Type: txt CFX_Gesamtsetup_ccl_15m_s_15deg_std_04.txt (181.4 KB, 6 views)
LBellmann is offline   Reply With Quote

Old   March 21, 2023, 05:04
Default
  #4
Senior Member
 
Gert-Jan
Join Date: Oct 2012
Location: Europe
Posts: 1,742
Rep Power: 25
Gert-Jan will become famous soon enough
From an output file it is difficult to completely understand what you are doing.

But I see an angle on OUTLET_FF
And on INLET_FF I read

FLOW DIRECTION:
Option = Normal to Boundary Condition
END

Shouldn't that have an angle of attack?
Bottomline: How does your file CCL_15deg.ccl look like?
Opaque likes this.
Gert-Jan is offline   Reply With Quote

Old   March 22, 2023, 01:44
Default
  #5
New Member
 
Lukas Bellmann
Join Date: Mar 2023
Posts: 8
Rep Power: 2
LBellmann is on a distinguished road
I've set an Total Pressure boundary condition at the inlet to make the convergence more stable. In previous simulations I've seen that the incoming flow in the farfield quickly turns over to an angle so that shouldn't be the problem.

The CCL files looks like:

FLOW: Flow Analysis 1
DOMAIN: FERNFELD
&replace BOUNDARY: OUTLET_FF
Boundary Type = OUTLET
Location = OUTLET_FF,OUTLET_FF 2
BOUNDARY CONDITIONS:
FLOW REGIME:
Option = Subsonic
END
MASS AND MOMENTUM:
Option = Cartesian Velocity Components
U = 14.489 [m s^-1]
V = 0 [m s^-1]
W = 3.882 [m s^-1]
END
END
END
END
END

FLOW: Flow Analysis 1
DOMAIN: TONNE
&replace BOUNDARY: OUTLET_TONNE
Boundary Type = OUTLET
Location = TONNE_OUTLET,TONNE_OUTLET 2
BOUNDARY CONDITIONS:
FLOW REGIME:
Option = Subsonic
END
MASS AND MOMENTUM:
Option = Cartesian Velocity Components
U = 14.489 [m s^-1]
V = 0 [m s^-1]
W = 3.882 [m s^-1]
END
END
END
END
END

FLOW: Flow Analysis 1
&replace SOLVER CONTROL:
Turbulence Numerics = High Resolution
ADVECTION SCHEME:
Option = High Resolution
END
CONVERGENCE CONTROL:
Local Timescale Factor = Timescale
Maximum Number of Iterations = 4000
Minimum Number of Iterations = 1
Timescale Control = Local Timescale Factor
END
CONVERGENCE CRITERIA:
Residual Target = 1e-06
Residual Type = RMS
END
DYNAMIC MODEL CONTROL:
Global Dynamic Model Control = On
END
END
END
LBellmann is offline   Reply With Quote

Old   March 22, 2023, 12:13
Default
  #6
Senior Member
 
Gert-Jan
Join Date: Oct 2012
Location: Europe
Posts: 1,742
Rep Power: 25
Gert-Jan will become famous soon enough
I don't know how your geometry looks like. Therefore I can't tell if things work the same on different simulations. So I think the normal direction on your inlet is the problem.
Gert-Jan is offline   Reply With Quote

Old   March 22, 2023, 13:01
Default
  #7
Senior Member
 
Join Date: Jun 2009
Posts: 1,688
Rep Power: 29
Opaque will become famous soon enough
I will second the Gert-Jan, from the settings w/o seeing the geometry of your model, the red flag seems to be the "Normal to Boundary Condition" setting.

From your output file, something else called my attention and it is the use of Local Timescale Factor. I have used Ansys CFX for 20+ years, and it is rare to use Local Timescale Factor except for some extreme cases. Wonder what kind of flow condition/mesh issues you are dealing with
__________________
Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.
Opaque is online now   Reply With Quote

Old   March 24, 2023, 11:44
Default
  #8
New Member
 
Lukas Bellmann
Join Date: Mar 2023
Posts: 8
Rep Power: 2
LBellmann is on a distinguished road
The Total Pressure condition is defined not the problem. Im dealing with the problem that a simulation which was initiated by anothers result file (same geometry/ mesh) does not run with specified boundary conditions (specified via def-file or CCL-file). It runs with exactly the same conditions as the previous run which generated the result-file Im initializing with. Even if I change the Inlet form one surface to another one it does not overwrite the boundary conditions of the old run.
LBellmann is offline   Reply With Quote

Old   March 24, 2023, 11:47
Default
  #9
New Member
 
Lukas Bellmann
Join Date: Mar 2023
Posts: 8
Rep Power: 2
LBellmann is on a distinguished road
I‘m simulating a fan-stage with all passages. In order to do that I chose local timescale because Autotimescale and physical timescale did not lead to a stable convergence. The variable „Timescale“ which is called in the CCL-file is just a user-function I created in CFC-Pre
To be honest I can’t really tell why it’s running on local timescale the best
LBellmann is offline   Reply With Quote

Old   March 24, 2023, 16:10
Default
  #10
Senior Member
 
Join Date: Jun 2009
Posts: 1,688
Rep Power: 29
Opaque will become famous soon enough
I have reused previous calculations as the initial guess for new calculations and I have never seen this problem. Wonder if you have frozen something.

Auto Timescale is an auto-computed physical timescale; however, the Local Timescale Factor is NOT a physical timescale (note it is dimensionless), and it can lead to frozen convergence and unphysical solutions.

I would restart the solution with standard Auto Timescale and verify it is working, i.e. using the new conditions. Then, you can try your approach to make converge at your will.
__________________
Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.
Opaque is online now   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
Patankar 1980 (Diffusion Convection Problems) hhandoko System Analysis 1 January 25, 2018 07:56
UDF initialization problems in 2D rarnaunot Fluent UDF and Scheme Programming 0 March 7, 2017 11:13
Simulation problems donfrulli STAR-CCM+ 5 September 13, 2016 09:49
Needed Benchmark Problems for FSI Mechstud Main CFD Forum 4 July 26, 2011 12:13
Some problems with Star CD Micha Siemens 0 August 6, 2003 13:55


All times are GMT -4. The time now is 18:46.