How to initialize a simulation from RANS to URANS

AS_Aero · September 26, 2018, 03:51

Hi

I am trying to restart a simulation from RANS to URANS

I changed my physics model for the .def file and then in that what should I do with the intialization tab for the domains ?

Then I have the RANS results as .res file.

I know somewhere I have to upload it, but I am running it in cluster so I have a script for it there I normally to restart the case I give
-def flename.def -ini filename.res -ccl filename.ccl

so will this way it works for RANS to URANS as well ? If yes what about the intialization tab in the CFX Pre ?

Gert-Jan · September 26, 2018, 04:39

Sure it will work.
In Pre, you only get a warning that a transient run requires an initial guess. Therefore it would like you to set some values. But you can safely ignore this. Pre doesn't know yet that you have the best initial guess available: your steady state run.

AS_Aero · September 26, 2018, 04:43

Perfect Gert Jan !!

But lets say I want to run an LES case with very refined mesh, in that case if I need to have an intial guess, I need to run the same fine mesh a RANS simulation and then only I can use it for the intialization right ? I cant use another coarse mesh results and use that to initialize the LES simulation ?

Gert-Jan · September 26, 2018, 04:51

You can use any mesh you like for initialization. You can even interpolate a car on a ship, or a pump on your windmill. CFX doesn't care. You are in control, you should make a clever choice.

Please read the output file. There a lot of info is given on the interpolation/initialisation process. It will tell you if the volumes of the different meshes fit, how many nodes are on the same spot, etc, etc.

AS_Aero · September 26, 2018, 04:57

Thanks again. No I was just asking if its a good practise to do like that ? to use different mesh results for intialization of another simulation.

ghorrocks · September 26, 2018, 07:28

If the initial condition mesh is the same as the continuing mesh then you don't need to do interpolation and that is an advantage. But the interpolator is rarely a problem so do not get too concerned if your meshes are different.

What is more important is that the flow described in the initial conditions are actually representative of what you are trying to do.

September 26, 2018, 03:51	How to initialize a simulation from RANS to URANS	#1
AS_Aero Senior Member Join Date: Mar 2013 Location: Germany Posts: 357 Rep Power: 14	Hi I am trying to restart a simulation from RANS to URANS I changed my physics model for the .def file and then in that what should I do with the intialization tab for the domains ? Then I have the RANS results as .res file. I know somewhere I have to upload it, but I am running it in cluster so I have a script for it there I normally to restart the case I give -def flename.def -ini filename.res -ccl filename.ccl so will this way it works for RANS to URANS as well ? If yes what about the intialization tab in the CFX Pre ?

September 26, 2018, 07:28		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	If the initial condition mesh is the same as the continuing mesh then you don't need to do interpolation and that is an advantage. But the interpolator is rarely a problem so do not get too concerned if your meshes are different. What is more important is that the flow described in the initial conditions are actually representative of what you are trying to do. __________________ Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Utility to initialize LES Simulation	ScarFace	OpenFOAM Pre-Processing	3	October 24, 2018 18:24
Transient RANS Simulation	Bazinga	Main CFD Forum	5	June 21, 2017 11:21
Simulation RANS Explode	fredicenci	OpenFOAM Running, Solving & CFD	4	May 23, 2017 13:29
Difference Between RANS And URANS	Divyaprakash	OpenFOAM Running, Solving & CFD	0	January 19, 2016 09:37
Simple question: Time-Dependent RANS vs URANS	c.alanis	Main CFD Forum	3	January 28, 2015 12:10

September 26, 2018, 04:39		#2
Gert-Jan Senior Member Gert-Jan Join Date: Oct 2012 Location: Europe Posts: 1,827 Rep Power: 27	Sure it will work. In Pre, you only get a warning that a transient run requires an initial guess. Therefore it would like you to set some values. But you can safely ignore this. Pre doesn't know yet that you have the best initial guess available: your steady state run.

September 26, 2018, 04:43		#3
AS_Aero Senior Member Join Date: Mar 2013 Location: Germany Posts: 357 Rep Power: 14	Perfect Gert Jan !! But lets say I want to run an LES case with very refined mesh, in that case if I need to have an intial guess, I need to run the same fine mesh a RANS simulation and then only I can use it for the intialization right ? I cant use another coarse mesh results and use that to initialize the LES simulation ?

September 26, 2018, 04:51		#4
Gert-Jan Senior Member Gert-Jan Join Date: Oct 2012 Location: Europe Posts: 1,827 Rep Power: 27	You can use any mesh you like for initialization. You can even interpolate a car on a ship, or a pump on your windmill. CFX doesn't care. You are in control, you should make a clever choice. Please read the output file. There a lot of info is given on the interpolation/initialisation process. It will tell you if the volumes of the different meshes fit, how many nodes are on the same spot, etc, etc.

September 26, 2018, 04:57		#5
AS_Aero Senior Member Join Date: Mar 2013 Location: Germany Posts: 357 Rep Power: 14	Thanks again. No I was just asking if its a good practise to do like that ? to use different mesh results for intialization of another simulation.