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

Jump in residuals under Linux

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

Reply
 
LinkBack Thread Tools Display Modes
Old   July 7, 2005, 07:20
Default Jump in residuals under Linux
  #1
mAx
Guest
 
Posts: n/a
Hi, I have a problem. Maybe has somebody an answer. Ok I describe the problem: Under Linux, I read my case, which has non-conformal interfaces, I iterate, and I write data. If now I re-read my data, and iterate more, all my residuals explode (ie: jump in my residuals). I didn't change any parameters in my .cas file That means, I can't pursue my calculations. Any idea? Thanks mAx
  Reply With Quote

Old   July 8, 2005, 02:03
Default Re: Jump in residuals under Linux
  #2
mAx
Guest
 
Posts: n/a
Hi, I think I know where the problem is. If I stop the calculations, check for instance a force on a piston, then write data.... then reread data, re-check the piston's force... i don't get the same result (and the pressure distribution has been corrupted). Does Anyone have the same problem???? I am running under Linux x86 & Parallel Thank U mAx
  Reply With Quote

Old   July 11, 2005, 11:15
Default Re: Jump in residuals under Linux
  #3
Rafael
Guest
 
Posts: n/a
Do you save only .dat file?
  Reply With Quote

Old   July 12, 2005, 01:53
Default Re: Jump in residuals under Linux
  #4
mAx
Guest
 
Posts: n/a
yes, just like autosave does. But I have no moving mesh, ie: my grid is still the same
  Reply With Quote

Old   July 12, 2005, 08:54
Default Re: Jump in residuals under Linux
  #5
Rafael
Guest
 
Posts: n/a
Are you using the command

if PRINCIPAL_FACE_P(f,t) inside the force loop ?
  Reply With Quote

Old   July 12, 2005, 08:57
Default Re: Jump in residuals under Linux
  #6
mAx
Guest
 
Posts: n/a
No... I am in contact with support, but they never have seen that before...
  Reply With Quote

Old   July 12, 2005, 09:47
Default Re: Jump in residuals under Linux
  #7
Rafael
Guest
 
Posts: n/a
if you want try it:

#if !RP_HOST

ID = IDvd1;

tv = Lookup_Thread( domain,ID );

begin_f_loop(face,tv)

if (PRINCIPAL_FACE_P(face,tv))

{

F_AREA(area,face,tv);

total_force += NV_MAG(area) * F_P(face,tv);

}

end_f_loop(face,tv)

#endif

node_to_host_double_1(total_force);
  Reply With Quote

Old   July 12, 2005, 09:51
Default Re: Jump in residuals under Linux
  #8
Rafael
Guest
 
Posts: n/a
I forgot:

total_force = PRF_GRSUM1(total_force);
  Reply With Quote

Old   July 12, 2005, 09:55
Default Re: Jump in residuals under Linux
  #9
mAx
Guest
 
Posts: n/a
If you get the same kinds of error, I assume it is because of an error PR_GRSUM1(total_force), which is supposed to add the force computing from each node. So normally the force computed shouldn't change if you run your case under 2,4 or 8 CPU. In my case I don't have any UDF...
  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
sudden jump in residuals wlt_1985 FLUENT 8 June 26, 2014 04:51
Jump in Residuals ryzd FLUENT 1 March 1, 2011 21:39
Dual Boot Windows and Linux and Go Open Source andyj Main CFD Forum 2 October 21, 2010 16:49
Convergence - scaled vs unscaled residuals HS FLUENT 1 November 7, 2005 05:45
Fluent Parallel/Serial version..jump in residuals Sandeep Abhyankar FLUENT 4 January 10, 2000 09:57


All times are GMT -4. The time now is 18:14.