CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > OpenFOAM Post-Processing

How to visualize the difference of fields btween two simulation

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

Like Tree3Likes

Reply
 
LinkBack Thread Tools Display Modes
Old   September 1, 2013, 09:05
Default
  #21
Senior Member
 
immortality's Avatar
 
Ehsan
Join Date: Oct 2012
Location: Iran
Posts: 2,186
Rep Power: 16
immortality is on a distinguished road
Hi Bruno,thanks
I used parallel run,now I have to reconstruct specified times?(the equivalent times in two sequential times) and then how use that command?!
__________________
Injustice Anywhere is a Threat for Justice Everywhere.Martin Luther King.
To Be or Not To Be,Thats the Question!
The Only Stupid Question Is the One that Goes Unasked.
immortality is offline   Reply With Quote

Old   September 1, 2013, 09:13
Default
  #22
Super Moderator
 
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 8,258
Blog Entries: 34
Rep Power: 84
wyldckat is just really nicewyldckat is just really nicewyldckat is just really nicewyldckat is just really nice
Hi Ehsan,

Quote:
Originally Posted by immortality View Post
I used parallel run,now I have to reconstruct specified times?(the equivalent times in two sequential times) and then how use that command?!
Sorry, I don't understand what you are specifically asking about
Please provide more details.

Best regards,
Bruno
wyldckat is offline   Reply With Quote

Old   September 1, 2013, 09:30
Default
  #23
Senior Member
 
immortality's Avatar
 
Ehsan
Join Date: Oct 2012
Location: Iran
Posts: 2,186
Rep Power: 16
immortality is on a distinguished road
Bruno I don't know how to use the command when the fields are decomposed in processor folders! I have been mixed up!
__________________
Injustice Anywhere is a Threat for Justice Everywhere.Martin Luther King.
To Be or Not To Be,Thats the Question!
The Only Stupid Question Is the One that Goes Unasked.
immortality is offline   Reply With Quote

Old   September 1, 2013, 09:38
Default
  #24
Super Moderator
 
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 8,258
Blog Entries: 34
Rep Power: 84
wyldckat is just really nicewyldckat is just really nicewyldckat is just really nicewyldckat is just really nice
Why don't you use the reconstructPar with the "-time" option?
And why don't you manually copy only the fields that you want to look at, form one time folder to the other?
wyldckat is offline   Reply With Quote

Old   September 1, 2013, 10:33
Default
  #25
Senior Member
 
immortality's Avatar
 
Ehsan
Join Date: Oct 2012
Location: Iran
Posts: 2,186
Rep Power: 16
immortality is on a distinguished road
I thought it have to be done through -parallel option.
do you know how can get rid of those red cells and have a more smooth result?
they are both in p field and subtract field.
Attached Images
File Type: jpg p_subtract.jpg (13.5 KB, 11 views)
File Type: jpg Screenshot at 2013-09-01 18:58:21.jpg (48.5 KB, 13 views)
__________________
Injustice Anywhere is a Threat for Justice Everywhere.Martin Luther King.
To Be or Not To Be,Thats the Question!
The Only Stupid Question Is the One that Goes Unasked.
immortality is offline   Reply With Quote

Old   September 1, 2013, 10:57
Default
  #26
Super Moderator
 
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 8,258
Blog Entries: 34
Rep Power: 84
wyldckat is just really nicewyldckat is just really nicewyldckat is just really nicewyldckat is just really nice
A few advices:
  • Whenever the results look suspicious, check if the values in parallel are identical to the reconstructed ones. It's always possible that some point/face/cell numbering isn't synced in parallel, but it should be when reconstructed.
  • Whenever there is such a large difference of values between contiguous cells, that might indicate that you do not have enough mesh resolution for describing properly what happened at the location of the cell with the largest differences in values.
  • A form of smoothing is the "vertex/point" representation.
  • The results are what they are. Many times we have to live with how imperfect they are, because better results are not achievable in the time frame for the project at hand.
  • Any excessive form of post-processing will only reduce the validity of your results, unless it was previously proven that for your case, the interpolation/smoothing process has a minimal error in comparison.
  • edit: You can use Gimp for reducing the size and quality of images before uploading them.

Last edited by wyldckat; September 1, 2013 at 10:58. Reason: see "edit:"
wyldckat is offline   Reply With Quote

Old   September 1, 2013, 11:36
Default
  #27
Senior Member
 
immortality's Avatar
 
Ehsan
Join Date: Oct 2012
Location: Iran
Posts: 2,186
Rep Power: 16
immortality is on a distinguished road
may it be a characrestic of rhoCentralFoam?
its an image from near,are cells few in your opinion?
I didn't grasp totally about writePrecision and its role,I have set it to 16,but these results existed when it was only 6,and I don't think time step be responsible for that because it was in lower time steps too(maxCo=.1 till maxCo=.2 that I've tried)
Attached Images
File Type: jpg suspicious.jpg (21.0 KB, 5 views)
__________________
Injustice Anywhere is a Threat for Justice Everywhere.Martin Luther King.
To Be or Not To Be,Thats the Question!
The Only Stupid Question Is the One that Goes Unasked.
immortality is offline   Reply With Quote

Old   September 1, 2013, 12:53
Default
  #28
Super Moderator
 
Bruno Santos
Join Date: Mar 2009
Location: Lisbon, Portugal
Posts: 8,258
Blog Entries: 34
Rep Power: 84
wyldckat is just really nicewyldckat is just really nicewyldckat is just really nicewyldckat is just really nice
From the looks of it and from what I can remember from this test case...
  1. From the previous pictures, it looked like there was only one big red cell surrounded by green cells. That's why I wrote my other comment on insufficient resolution.
  2. But looking at this zoomed in image and remembering the case, what we're seeing is the effect of pressure shock-waves crossing each other's paths. Therefore, more refinement will probably only mean that you'll see with more resolution this crossing of the shock-waves.
    • You could use the trick that is explained in the very first tutorial of the User Guide, namely to use a time step from X seconds before and mapping its fields into a simulation with 2x the mesh resolution, to see how much it would change with the higher resolution.
  3. If you are using a precision of 16 digits, it would be preferable to use "binary" coding instead of "ascii". I say this because it's 16 to 18 bytes per value in "ascii", versus 8 bytes in "binary", therefore your files could be around 50% in size of what they currently are.
immortality likes this.
wyldckat is offline   Reply With Quote

Old   September 1, 2013, 15:25
Default
  #29
Senior Member
 
immortality's Avatar
 
Ehsan
Join Date: Oct 2012
Location: Iran
Posts: 2,186
Rep Power: 16
immortality is on a distinguished road
yes you are true,they are the points that originated oblique shocks hit together and make pressure to be higher than neighbor cells.It was odd for me that had forgotten about it and ignored this problem,thank you for saving me
__________________
Injustice Anywhere is a Threat for Justice Everywhere.Martin Luther King.
To Be or Not To Be,Thats the Question!
The Only Stupid Question Is the One that Goes Unasked.
immortality 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
Overflow problem in steady simulation ReeKo CFX 11 October 8, 2008 17:57
Difference of temporally and spatially simulation? leaf Main CFD Forum 0 May 14, 2006 21:35
Unsteady vs Steady simulation - any difference? Sally Main CFD Forum 16 December 16, 2005 10:15
Fire simulation using FDS from NIST Jens Main CFD Forum 1 January 22, 2004 02:53
3-D Contaminant Dispersal Simulation Apple L S Chan Main CFD Forum 1 December 23, 1998 11:06


All times are GMT -4. The time now is 12:57.