ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
-
- Posts: 4
- Joined: 17 Jan 2020, 07:59
- Antispam: Yes
ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
So, I did everything as it is told in the tutorial (several times), and everytime I get the same solution, which I think is wrong because it is really different from the one presented in images at the end of the tutorial.
This is what I should get, as it is presented in the tutorial
This is what I obtain when I open the post file in ParaView
As you may see, the displacement of the obstacle is in the opposite direction it should be.
Any help is appreciated.
I attached the .sif file, and below is the URL to the PDF with the tutorial
http://www.nic.funet.fi/pub/sci/physics ... orials.pdf
This is what I should get, as it is presented in the tutorial
This is what I obtain when I open the post file in ParaView
As you may see, the displacement of the obstacle is in the opposite direction it should be.
Any help is appreciated.
I attached the .sif file, and below is the URL to the PDF with the tutorial
http://www.nic.funet.fi/pub/sci/physics ... orials.pdf
- Attachments
-
- case.sif
- (4.45 KiB) Downloaded 287 times
-
- Posts: 2237
- Joined: 25 Jan 2019, 01:28
- Antispam: Yes
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
I received the same thing. I added a save scalers solver to equation 2 to print out the displacement and it is positive in the printed output. The error appear to be with the writing or reading of the vtu file.
Solver 4
Exec Solver = After Timestep
Equation = SaveScalars
Procedure = "SaveData" "SaveScalars"
Filename = "displacement.dat"
Variable 1 = Displacement
Operator 1 = min
Operator 2 = max
End
Equation 1
Name = "Flow and mesh"
Active Solvers(2) = 3 1
End
Equation 2
Name = "Elasticity"
Active Solvers(2) = 2 4
End
Solver 4
Exec Solver = After Timestep
Equation = SaveScalars
Procedure = "SaveData" "SaveScalars"
Filename = "displacement.dat"
Variable 1 = Displacement
Operator 1 = min
Operator 2 = max
End
Equation 1
Name = "Flow and mesh"
Active Solvers(2) = 3 1
End
Equation 2
Name = "Elasticity"
Active Solvers(2) = 2 4
End
-
- Posts: 2237
- Joined: 25 Jan 2019, 01:28
- Antispam: Yes
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
when I use save data along a line on the beam, the displacements report out negative.
Solver 4
Exec Solver = After Timestep
Equation = SaveLine
Procedure = "SaveData" "SaveLine"
Filename = "line.dat"
Polyline Coordinates(2,2) = 2.481 0.0 2.481 1.14
Polyline Divisions(2) = 10 10
End
Solver 4
Exec Solver = After Timestep
Equation = SaveLine
Procedure = "SaveData" "SaveLine"
Filename = "line.dat"
Polyline Coordinates(2,2) = 2.481 0.0 2.481 1.14
Polyline Divisions(2) = 10 10
End
-
- Posts: 2237
- Joined: 25 Jan 2019, 01:28
- Antispam: Yes
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
I used an older version of Elmer that has not been updated in a while and the tutorial worked.
-
- Posts: 2237
- Joined: 25 Jan 2019, 01:28
- Antispam: Yes
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
I used Elmer Version 8.2 on Windows and the tutorial worked
I used Elmer Version 8.3 on The Elmer Virtual Machine (VM) and the tutorial worked
I used Elmer Version 8.4 latest update on Linux and the displacements are negative instead of Positive.
I used Elmer Version 8.3 on The Elmer Virtual Machine (VM) and the tutorial worked
I used Elmer Version 8.4 latest update on Linux and the displacements are negative instead of Positive.
-
- Posts: 4
- Joined: 17 Jan 2020, 07:59
- Antispam: Yes
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
Thanks kevinarden for your help and willingness to try so many options. I did the same as you, adding Solver 4, and I also got positive/negative results in the first/second run. I'm thinking in going after the updates in these versions to see if I can find the reason to this error.
Thanks again, and I hope to come back with good news.
Thanks again, and I hope to come back with good news.
-
- Posts: 4
- Joined: 17 Jan 2020, 07:59
- Antispam: Yes
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
Coming here again just to let you know that updating Elmer worked for me.
I'm using it in Linux, so I did
elmer-csc was updated and now the results are just what I expected.
Thanks again.
I'm using it in Linux, so I did
Code: Select all
sudo apt-get update
Code: Select all
sudo apt-get upgrade
Thanks again.
-
- Posts: 2237
- Joined: 25 Jan 2019, 01:28
- Antispam: Yes
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
Not so for me. I am on Linux Mint 19.3 which is based on Ubuntu 18.04
-
- Site Admin
- Posts: 4812
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
Hi
Thanx again for your diligent work to locate the problem. Unfortunately there was some ambiguity in the code that seems to have dependent on some randomness. So a bug.
The question is how the normal should be oriented. There has been the keyword "Normal Target Body" that has been used in some fsi test cases. For example, until today the test case "fsi_beam" (which is very close to tutorial 12) has this keyword.
On the interface between two bodies the direction of the normal otherwise followed a convention that was affected by numbering of bodies. Hence if you had
the sign was wrong, but if you exchange 1's and 2's in the arguments the normal becomes the opposite and things work correctly.
Well, this is now changed it the code such that the direction of the normal is checked rigorously and there is no need to define it.
Also the 2nd method of setting the FSI BCs is now better demonstrated in new test case fsi_beam_nodalforce. Is the name implies we use here nodal forces and the coupling is performed an discrete level.
All the changes are in devel and hopefully in the nightly builds after few hours.
Thanx for reporting!
-Peter
Thanx again for your diligent work to locate the problem. Unfortunately there was some ambiguity in the code that seems to have dependent on some randomness. So a bug.
The question is how the normal should be oriented. There has been the keyword "Normal Target Body" that has been used in some fsi test cases. For example, until today the test case "fsi_beam" (which is very close to tutorial 12) has this keyword.
On the interface between two bodies the direction of the normal otherwise followed a convention that was affected by numbering of bodies. Hence if you had
Code: Select all
Body 1
Target Bodies(1) = 2
Name = "BodyA"
Equation = 2
Material = 2
End
Body 2
Target Bodies(1) = 1
Name = "BodyB"
Equation = 1
Material = 1
End
Well, this is now changed it the code such that the direction of the normal is checked rigorously and there is no need to define it.
Also the 2nd method of setting the FSI BCs is now better demonstrated in new test case fsi_beam_nodalforce. Is the name implies we use here nodal forces and the coupling is performed an discrete level.
All the changes are in devel and hopefully in the nightly builds after few hours.
Thanx for reporting!
-Peter
-
- Posts: 2237
- Joined: 25 Jan 2019, 01:28
- Antispam: Yes
Re: ElmerGUI Tutorial 12 - Interaction between fluid flow and elastic obstacle
Many thanks. I expected that was the case since the pressures were correct. I will test it once I update.
Kevin
Kevin