CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > FloEFD, FloWorks & FloTHERM

FloEFD Residuals

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

Reply
 
LinkBack Thread Tools Display Modes
Old   March 17, 2014, 11:02
Question FloEFD Residuals
  #1
Super Moderator
 
-mAx-'s Avatar
 
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,135
Rep Power: 32
-mAx- will become famous soon enough
Hello,
As we used to work with typical residuals curves, we are a little bit lost with those % bar-charts,
So here are the questions:

1- if we choose convergence-criterior for goals as Automatic; how set FloEFD the limit for the convergence?

2- If we want to switch from automatic to manual; how big or how small should be the
limit for having converged, credible and reliable solution?

3- During how many iterations should remain the "Delta" under the specific value untill FloEFD decides that the solution has converged?

Thanks for the explanations
__________________
In memory of my friend Hervé: CFD engineer & freerider
-mAx- is offline   Reply With Quote

Old   March 23, 2014, 10:49
Default
  #2
Senior Member
 
Boris Marovic
Join Date: Jul 2009
Posts: 449
Rep Power: 14
Boris_M is on a distinguished road
Hi Maxime,

By %-Charts you probably have all goals in the goal plot which will automatically use the normalized y-axis for the plot. If you use only similar parameter types when creating the goal plot you should get the absolute value as y-axis. You can also change the type of y-axis by right-click into the graph and change it from normalized to absolute. You will see very large values and very small when looking at pressure in Pa and velocity in m/s as the one values are in the hundrests of thousands and the others are single digit numbers. That's the reason why it normalizes it. Otherwise you will only see flat lines at the bottom of the graph and at the very top.

1. These absolute values mentioned above show the absolute value of the parameter and you can see how much it fluctuates which is a more obvious indication for non numerical experts than residuals.

2. That depends on the acutal value of your goal. If you are looking at velocities around 2-3 m/s the criteria shouldn't be in the same order but much smaller. It wouldn't make sens to look for convergence with 1m/s fluctuations as it would be 30-50% of the final value. I usually consider something around less than 5% as good convergence. Sometimes some values are set automatically very high but others need much longer to converge so that even thought the automatic setting is very high the acutal delta (right click on the top row of the table in the goal plot where the names are and activate "Delta" as a parameter column to display). The criterion might be 1m/s but the actual delta is at 0.0000012532 m/s or something like that and with that way below 1%.

3. The typical setting is half a travel which is good. If you can see the values are very good increasing and then flatteing to a converged value you can also say that is good enough but as long as they are fluctuating a lot I would let it run.

I hope this helps,
Boris
Boris_M is offline   Reply With Quote

Old   March 24, 2014, 01:56
Default
  #3
Super Moderator
 
-mAx-'s Avatar
 
Maxime Perelli
Join Date: Mar 2009
Location: Switzerland
Posts: 3,135
Rep Power: 32
-mAx- will become famous soon enough
Thank you Boris for your replies, 2 points more:
Quote:
Originally Posted by Boris_M View Post
1. These absolute values mentioned above show the absolute value of the parameter and you can see how much it fluctuates which is a more obvious indication for non numerical experts than residuals.
If I read your reply on my second question, then the limit should be set around 5% or something like that, no?

We are talking now about "monitoring" goals-values. But is there a way to observe the real residuals from equations of continuity, Navier-Stokes, energy etc...
__________________
In memory of my friend Hervé: CFD engineer & freerider

Last edited by -mAx-; March 28, 2014 at 01:26.
-mAx- is offline   Reply With Quote

Old   March 27, 2014, 21:46
Default
  #4
Senior Member
 
Boris Marovic
Join Date: Jul 2009
Posts: 449
Rep Power: 14
Boris_M is on a distinguished road
Yes, roughly that's what I do if the general delta value is not already small enough

If I'm not mistaken the residuals can be found in the *.stdout in small tables with the header in the table such as "Target res". But that's not very usefull to look for it in a nice way and I'm not 100% sure. Let me check and I will get back to you.

Boris
Boris_M is offline   Reply With Quote

Reply

Thread Tools
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 On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Tutorial of how to plot residuals ! wolle1982 OpenFOAM Running, Solving & CFD 132 May 20, 2016 09:28
under-relaxation factors -> level of residuals Zigainer FLUENT 17 February 6, 2016 01:12
motorBike Residuals for SST k-omega... and mine JR22 OpenFOAM Running, Solving & CFD 6 August 1, 2013 09:08
judging convergence through residuals MachZero Main CFD Forum 7 December 25, 2012 13:18
Convergence - scaled vs unscaled residuals HS FLUENT 1 November 7, 2005 06:45


All times are GMT -4. The time now is 05:35.