LTSinterfoam-calculating several speeds without snappyhexmesh each time

Ingenieur5 · October 8, 2015, 11:38

Hi there,
as being mor a naval architect than being a software guy I configured the tutorials to suit my needs for cfd calculations. This includes that I prepare a case for each speed and run the ./Allrun command to finalise my calculations, which works really fine.
But I guess I could save some time if I do this for the first speed and then use the produced mesh also for the next series of speeds without doing the preprocessing for every speed.
What I assume to do is to copy the casefolder after the first calculation and rename the folder and then

?
What do I have to delete and how to start the next calculation with new speed?

thanks for help,
Ingenieur5

tomf · October 9, 2015, 04:32

Hi,

I would suggest the following approach:

1. Run the first speed case until you are happy with the results
2. Clone your case using the foamCloneCase script
3. Modify your new Case U file in the zero folder
4. Run your new case.
5. Repeat 2-4 as many times as you like

For part two the syntax is as follows:

Code:

/pathToCases/FirstCase> cd ..
/pathToCases> foamCloneCase FirstCase SecondCase
/pathToCases> cd SecondCase

If you want to keep using the Allrun script you can also copy that script into your new folder. Inside the Allrun script in your new folder comment the line about snappyHexMesh (put a # at the start of it).

This should be the general recipe. You probably need to adapt is for your particular set-up (for instance maybe change your turbulence parameters to match the new conditions). An inbetween step may be to use mapFields to start with the results of the previous speed in order to speedup convergence for your new case.

Good luck,
Tom

Ingenieur5 · October 12, 2015, 15:09

Hi Tom,
thanx for the quick answer. Actually beinig outside the office I cannot do next calcuölations before the weekend.
Will your proposal mean that in the cloned case the calculation use the 'old' mesh and just recalculate the new speed updated in the 0.org file?
Will it than still use the available processors in parallel?

I will try on the weekend but maybe you can give me a hint in advance if necessary

Best
Ingenieur5

tomf · October 13, 2015, 03:36

foamCloneCase simply copies the 0 constant and system folders into a new one. There will not be a 0.org folder copied. You can just copy that one to the new folder as well and than run the Allrun script. If this is unclear I would suggest that you read up on folder structures in Linux and carefully read the OpenFOAM user guide again.

Regards,
Tom

October 8, 2015, 11:38	LTSinterfoam-calculating several speeds without snappyhexmesh each time	#1
Ingenieur5 New Member sreibisch Join Date: Sep 2013 Posts: 10 Rep Power: 12	Hi there, as being mor a naval architect than being a software guy I configured the tutorials to suit my needs for cfd calculations. This includes that I prepare a case for each speed and run the ./Allrun command to finalise my calculations, which works really fine. But I guess I could save some time if I do this for the first speed and then use the produced mesh also for the next series of speeds without doing the preprocessing for every speed. What I assume to do is to copy the casefolder after the first calculation and rename the folder and then? What do I have to delete and how to start the next calculation with new speed? thanks for help, Ingenieur5

October 9, 2015, 04:32		#2
tomf Senior Member Tom Fahner Join Date: Mar 2009 Location: Breda, Netherlands Posts: 637 Rep Power: 32	Hi, I would suggest the following approach: 1. Run the first speed case until you are happy with the results 2. Clone your case using the foamCloneCase script 3. Modify your new Case U file in the zero folder 4. Run your new case. 5. Repeat 2-4 as many times as you like For part two the syntax is as follows: Code: /pathToCases/FirstCase> cd .. /pathToCases> foamCloneCase FirstCase SecondCase /pathToCases> cd SecondCase If you want to keep using the Allrun script you can also copy that script into your new folder. Inside the Allrun script in your new folder comment the line about snappyHexMesh (put a # at the start of it). This should be the general recipe. You probably need to adapt is for your particular set-up (for instance maybe change your turbulence parameters to match the new conditions). An inbetween step may be to use mapFields to start with the results of the previous speed in order to speedup convergence for your new case. Good luck, Tom

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[solidMechanics] solidMechanics gear contact in rotation	nlc	OpenFOAM CC Toolkits for Fluid-Structure Interaction	3	January 11, 2015 06:41
Help for the small implementation in turbulence model	shipman	OpenFOAM Programming & Development	25	March 19, 2014 10:08
Unstabil Simulation with chtMultiRegionFoam	mbay101	OpenFOAM Running, Solving & CFD	13	December 28, 2013 13:12
dynamic Mesh is faster than MRF????	sharonyue	OpenFOAM Running, Solving & CFD	14	August 26, 2013 07:47
pisoFoam with k-epsilon turb blows up - Some questions	Heroic	OpenFOAM Running, Solving & CFD	26	December 17, 2012 03:34

October 12, 2015, 15:09		#3
Ingenieur5 New Member sreibisch Join Date: Sep 2013 Posts: 10 Rep Power: 12	Hi Tom, thanx for the quick answer. Actually beinig outside the office I cannot do next calcuölations before the weekend. Will your proposal mean that in the cloned case the calculation use the 'old' mesh and just recalculate the new speed updated in the 0.org file? Will it than still use the available processors in parallel? I will try on the weekend but maybe you can give me a hint in advance if necessary Best Ingenieur5

October 13, 2015, 03:36		#4
tomf Senior Member Tom Fahner Join Date: Mar 2009 Location: Breda, Netherlands Posts: 637 Rep Power: 32	foamCloneCase simply copies the 0 constant and system folders into a new one. There will not be a 0.org folder copied. You can just copy that one to the new folder as well and than run the Allrun script. If this is unclear I would suggest that you read up on folder structures in Linux and carefully read the OpenFOAM user guide again. Regards, Tom