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

omit some equations withregularity

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

Like Tree5Likes
  • 1 Post By amin.z
  • 1 Post By brunoc
  • 2 Post By LuckyTran
  • 1 Post By LuckyTran

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   October 19, 2015, 03:57
Default omit some equations withregularity
  #1
New Member
 
Mina
Join Date: Jul 2015
Location: Deutschland
Posts: 21
Rep Power: 10
Mina_Mbg is on a distinguished road
Hi,
Does anybody hear about turning on and off some equations in specific time interval during simulating in transient mode?for example every 1S turn off turbulence and Navier-stocke equations,but Energy is always involved in simulation.
I am supposed to do that for my simulation which is dry quenching process in high pressure.
involved equations are:SST K-omega,and Energy besides Navier-Stoks
fluid material is Nitrogen and Helium which are applied of cource separately.Unfortunatelly the density for both are temperature dependent!
Hope the question is clear
thanks in advanced

Last edited by Mina_Mbg; October 21, 2015 at 04:36. Reason: feedback from people who try to answer
Mina_Mbg is offline   Reply With Quote

Old   October 19, 2015, 04:58
Default
  #2
Senior Member
 
amin.z's Avatar
 
Amin
Join Date: Oct 2013
Location: Germany
Posts: 397
Rep Power: 14
amin.z is on a distinguished road
Sounds interesting!
I'm interested in this topic,
Does anybody have any idea?
Mina_Mbg likes this.
amin.z is offline   Reply With Quote

Old   October 20, 2015, 11:14
Default
  #3
Senior Member
 
Bruno
Join Date: Mar 2009
Location: Brazil
Posts: 277
Rep Power: 21
brunoc is on a distinguished road
You could set predefined commands in Calculation Activities > Execute Commands that turn your equations on and off every N iterations/timesteps.

You could also write a journal file with a loop in scheme language that does that.
brunoc is offline   Reply With Quote

Old   October 20, 2015, 12:39
Default
  #4
New Member
 
Mina
Join Date: Jul 2015
Location: Deutschland
Posts: 21
Rep Power: 10
Mina_Mbg is on a distinguished road
Thanks for your feedback.
But I am looking for the logic behind that?Why and under which situation we will do that?
Mina_Mbg is offline   Reply With Quote

Old   October 20, 2015, 12:55
Default
  #5
Senior Member
 
Bruno
Join Date: Mar 2009
Location: Brazil
Posts: 277
Rep Power: 21
brunoc is on a distinguished road
I thought you wanted to use it in your model.

Well, this would be useful anytime the transient thermal behavior is important, but the change in velocities is quasi-static. One example could be the transient thermal behavior of a building. The change in fluid velocities in its interior is small enough that you could consider it to be quasi-static. So solve the energy equation continuously and only update the momentum equations from time to time.
Mina_Mbg likes this.
brunoc is offline   Reply With Quote

Old   October 20, 2015, 13:04
Default
  #6
New Member
 
Mina
Join Date: Jul 2015
Location: Deutschland
Posts: 21
Rep Power: 10
Mina_Mbg is on a distinguished road
Many Thanks for explanaition
Exactly I was told to us that but I havent known the reason.
I use also SST k-omega equation besides Navier-Stoks and Energie.What is your Idea about turning on and off also the turbulent equations?


Quote:
Originally Posted by brunoc View Post
I thought you wanted to use it in your model.

Well, this would be useful anytime the transient thermal behavior is important, but the change in velocities is quasi-static. One example could be the transient thermal behavior of a building. The change in fluid velocities in its interior is small enough that you could consider it to be quasi-static. So solve the energy equation continuously and only update the momentum equations from time to time.
Mina_Mbg is offline   Reply With Quote

Old   October 20, 2015, 20:35
Default
  #7
Senior Member
 
Lucky
Join Date: Apr 2011
Location: Orlando, FL USA
Posts: 5,654
Rep Power: 65
LuckyTran has a spectacular aura aboutLuckyTran has a spectacular aura aboutLuckyTran has a spectacular aura about
Like bruno I also thought that you were asking on how to do it, rather than why it is done.

I get why it is beneficial to do a simulation this way, but it is generally not implemented because it's useful only for constant density simulations. For temperature dependent densities and real densities, this method is not applicable. Since the updated temperature field affects the velocity field.

Any time you are dealing with a passive scalar transport this approach is useful, or when doing constant property simulations (where temperature becomes a passive scalar). The passive scalars cannot converge until the velocity field is converged. However, as the passive scalars are passive, there is no benefit to calculating them every iteration. You only need to start computing them after you have the correct velocity field. So it is possible to save considerably on computational cost and initially solve only the velocity field and no others, and then disable velocity field and compute only the scalars. In constant property LES you can reduce overall computational cost by ~20%.

Are you asking particularly about the transient case? It sounds weird to turn equations on and off for transient simulations, but this is commonplace in steady state simulations for the same reasons that it is beneficial in transient simulations.
devesh.baghel and Mina_Mbg like this.
LuckyTran is offline   Reply With Quote

Old   October 21, 2015, 04:41
Default
  #8
New Member
 
Mina
Join Date: Jul 2015
Location: Deutschland
Posts: 21
Rep Power: 10
Mina_Mbg is on a distinguished road
Thanks for your description and feedback.
yes in transient because my simulation goes very slow .I cannot understand wat to you mean by passiv scalar.
I am doubt now about turning off turbulent model especially when the eddies produced.
!



Quote:
Originally Posted by LuckyTran View Post
Like bruno I also thought that you were asking on how to do it, rather than why it is done.

I get why it is beneficial to do a simulation this way, but it is generally not implemented because it's useful only for constant density simulations. For temperature dependent densities and real densities, this method is not applicable. Since the updated temperature field affects the velocity field.

Any time you are dealing with a passive scalar transport this approach is useful, or when doing constant property simulations (where temperature becomes a passive scalar). The passive scalars cannot converge until the velocity field is converged. However, as the passive scalars are passive, there is no benefit to calculating them every iteration. You only need to start computing them after you have the correct velocity field. So it is possible to save considerably on computational cost and initially solve only the velocity field and no others, and then disable velocity field and compute only the scalars. In constant property LES you can reduce overall computational cost by ~20%.

Are you asking particularly about the transient case? It sounds weird to turn equations on and off for transient simulations, but this is commonplace in steady state simulations for the same reasons that it is beneficial in transient simulations.
Mina_Mbg is offline   Reply With Quote

Old   October 28, 2015, 14:17
Default
  #9
Senior Member
 
amin.z's Avatar
 
Amin
Join Date: Oct 2013
Location: Germany
Posts: 397
Rep Power: 14
amin.z is on a distinguished road
Quote:
Originally Posted by brunoc View Post
You could set predefined commands in Calculation Activities > Execute Commands that turn your equations on and off every N iterations/timesteps.

You could also write a journal file with a loop in scheme language that does that.
thanks dear Bruno for your answer,
could you please share the command which you use for it in Calculation Activities tab?
amin.z is offline   Reply With Quote

Old   October 28, 2015, 14:26
Default
  #10
Senior Member
 
amin.z's Avatar
 
Amin
Join Date: Oct 2013
Location: Germany
Posts: 397
Rep Power: 14
amin.z is on a distinguished road
Quote:
Originally Posted by LuckyTran View Post
Like bruno I also thought that you were asking on how to do it, rather than why it is done.

I get why it is beneficial to do a simulation this way, but it is generally not implemented because it's useful only for constant density simulations. For temperature dependent densities and real densities, this method is not applicable. Since the updated temperature field affects the velocity field.

Any time you are dealing with a passive scalar transport this approach is useful, or when doing constant property simulations (where temperature becomes a passive scalar). The passive scalars cannot converge until the velocity field is converged. However, as the passive scalars are passive, there is no benefit to calculating them every iteration. You only need to start computing them after you have the correct velocity field. So it is possible to save considerably on computational cost and initially solve only the velocity field and no others, and then disable velocity field and compute only the scalars. In constant property LES you can reduce overall computational cost by ~20%.

Are you asking particularly about the transient case? It sounds weird to turn equations on and off for transient simulations, but this is commonplace in steady state simulations for the same reasons that it is beneficial in transient simulations.
Thanks Lucky for sharing your knowledge,
Is it only applicable for energy equation?
Is it logical to use it for turbulence equation?
For isothermal domains, is it logical to ignore mass and momentum cons. equ. for some iterations in each time step and just solve the turbulence equations?
amin.z is offline   Reply With Quote

Old   October 28, 2015, 16:05
Default
  #11
Senior Member
 
Lucky
Join Date: Apr 2011
Location: Orlando, FL USA
Posts: 5,654
Rep Power: 65
LuckyTran has a spectacular aura aboutLuckyTran has a spectacular aura aboutLuckyTran has a spectacular aura about
There may be some uses for systematically disabling equations, but can you please tell us what you are trying to do? This is a very tough conversation where we are trying to guess the question to your answer.

For a general problem, I do not think turbulence models should ever be disabled separately from the velocity (flow). The purpose of the turbulence modelling is to model the Reynolds shear stresses which appear explicitly in the momentum equation. Hence, if the turbulence has not converged then neither can the velocity.

It is logical to disable as many equations as you like, after confirming that they have converged. I would agree with disabling flow and turbulence together but not separately.

By disabling flow and solving turbulence you are effectively segregating the solver even further than it already is. e.g. you get two iterations of the turbulence equations for each iteration of the flow equation. What this also does is it gives you a lower effective under-relaxation factor for each disabled equation. But is this what you are trying to do?
Mina_Mbg likes this.
LuckyTran is offline   Reply With Quote

Old   October 29, 2015, 03:07
Default
  #12
New Member
 
Mina
Join Date: Jul 2015
Location: Deutschland
Posts: 21
Rep Power: 10
Mina_Mbg is on a distinguished road
Hello LuckyTran,
I am supposed to do that in order to accelerate simulation.But it seems to me ambiguous to ignore some equations in some time step regularly.
It is couple scheme problem.

Quote:
Originally Posted by LuckyTran View Post
There may be some uses for systematically disabling equations, but can you please tell us what you are trying to do? This is a very tough conversation where we are trying to guess the question to your answer.

For a general problem, I do not think turbulence models should ever be disabled separately from the velocity (flow). The purpose of the turbulence modelling is to model the Reynolds shear stresses which appear explicitly in the momentum equation. Hence, if the turbulence has not converged then neither can the velocity.

It is logical to disable as many equations as you like, after confirming that they have converged. I would agree with disabling flow and turbulence together but not separately.

By disabling flow and solving turbulence you are effectively segregating the solver even further than it already is. e.g. you get two iterations of the turbulence equations for each iteration of the flow equation. What this also does is it gives you a lower effective under-relaxation factor for each disabled equation. But is this what you are trying to do?
Mina_Mbg is offline   Reply With Quote

Old   December 2, 2015, 14:32
Default
  #13
Senior Member
 
Bruno
Join Date: Mar 2009
Location: Brazil
Posts: 277
Rep Power: 21
brunoc is on a distinguished road
Quote:
I don't know do you remember or not, but we had a short discussion in a threat about turning equations on/off:

omit some equations withregularity

and you suggested us to use "execute command" to do this, even though this tab is really helpful, but unfortunately there's no source about the commands we can write there,
and my project had been stuck for this command,
so would you mind sharing the command we use to turn some equations on and off every N iterations/timesteps?

Thanks
Hi Amin,

I'm answering you here in case someone in the future has the same question.

The commands expected by Fluent are the same ones used on its TUI (text user interface) and they change depending on what you have enabled. To find out exactly what "chain" of commands you need, just press [ENTER] on Fluent's TUI and see what comes out. For instance, to find out how to enable the energy equation, you could do something like this (everything in bold is my input, [ENTER] shows the options within that level):

Code:
> [ENTER]
adapt/                  file/                   report/
define/                 mesh/                   solve/
display/                parallel/               surface/
exit                    plot/                   views/

> define

/define> [ENTER]
beta-feature-access                             models/
boundary-conditions/                            operating-conditions/
custom-field-functions/                         parameters/
dynamic-mesh/                                   profiles/
enable-mesh-morpher-optimizer?                  set-unit-system
injections/                                     solution-strategy/
materials/                                      units
mesh-interfaces/                                user-defined/
mixing-planes/

/define> models

/define/models> [ENTER]
acoustics/                                      solidification-melting?
addon-module                                    solver/
axisymmetric?                                   species/
crevice-model?                                  steady?
dpm/                                            unsteady-1st-order?
energy?                                         unsteady-2nd-order-bounded?
multiphase/                                     unsteady-2nd-order?
radiation/                                      viscous/

/define/models> energy
Enable energy model? [no] yes
Compute viscous energy dissipation? [no] no
include pressure work in energy equation? [no] no
include kinetic energy in energy equation? [no] no
Include diffusion at inlets? [yes] yes

/define/models>
You'll see that the TUI commands sort of follow the GUI commands through the menu:



So, to enable the energy equation in one go, in this specific condition, you could use:
Code:
> define models energy yes no no no yes
You'll need to find out the commands for what you need to do by yourself because, again, the options shown to the user will vary depending on what you have enabled.

Cheers,
Bruno
brunoc is offline   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
Guide: Writing Equations in LaTeX on the CFD Online Forums pete Site Help, Feedback & Discussions 27 May 19, 2022 04:19
SIMPLE Algorithm Finite Difference Equations: how to discretize and solve? DA6righthand Main CFD Forum 0 August 3, 2015 13:12
how to implement k-epsilon equations in ADI method kiamey Main CFD Forum 2 July 17, 2014 18:13
modelling Differential equations in a udf RikardMNorén Fluent UDF and Scheme Programming 2 October 1, 2013 04:36
CFD governing equations m.gos Main CFD Forum 0 April 30, 2011 15:21


All times are GMT -4. The time now is 08:46.