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

How to get initial residuals for U components during runtime?

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

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   April 11, 2011, 12:49
Default How to get initial residuals for U components during runtime?
  #1
Senior Member
 
Florian Krause
Join Date: Mar 2009
Location: Munich
Posts: 103
Rep Power: 17
florian_krause is on a distinguished road
Dear all,
I use a modified simpleFoam in OpenFOAM-1.7.x. I simulate a laminar pipe flow with uniform fixed value inlet boundary conditons and zeroGradient outlet boundary conditions. I've set up the case as a wedge using wedge boundary conditions. so far so good...

Now I want to stop the simulations if a pre-defined convergence criterion is satisfied for pressure p and the axial (x) and wall-normal (y) velocity component.

If I use

tmp<fvVectorMatrix> UEqn
(
fvm::div(phi, U)
+ turbulence->divDevReff(U)
);

UEqn().relax();

eqnResidualU = solve
(
UEqn() == -fvc::grad(p)
).initialResidual();

maxResidualU = max(eqnResidualU, maxResidualU);


then I get the initial residuals for all three velocity components. And using

if (maxResidualU < convergenceCriterionU && maxResidualP < convergenceCriterionP)
{
Info<< "reached convergence criterion for U and p"<< endl;
runTime.writeAndEnd();
Info<< "latestTime = " << runTime.timeName() << endl;
}


I perform the actual convergence check on all three velocity components.

As mentioned above, for this wedge case, I want to stop the calculation if only the x and y component of U satisfy the pre-defined convergence criterion.

Does anyone know how to do that? As always, any help is appreciated!

Best,
Florian
florian_krause is offline   Reply With Quote

Old   May 5, 2011, 19:22
Default
  #2
Member
 
Logan Page
Join Date: Sep 2010
Posts: 38
Rep Power: 15
Logan Page is on a distinguished road
Try modifying the solver along the lines of this

int StopItter = 1;
while (runTime.loop() && StopItter)
{
.
.
.
#include "UEqn.H"
#include "TEqn.H"
#include "pEqn.H"
.
.
.
}



then where you do the convergence check try this

if (maxResidualU < convergenceCriterionU && maxResidualP < convergenceCriterionP)
{
Info<< "reached convergence criterion for U and p"<< endl;
runTime.writeAndEnd();
Info<< "latestTime = " << runTime.timeName() << endl;
StopItter = 0;
}
Logan Page is offline   Reply With Quote

Old   May 6, 2011, 04:10
Default
  #3
Senior Member
 
Florian Krause
Join Date: Mar 2009
Location: Munich
Posts: 103
Rep Power: 17
florian_krause is on a distinguished road
thanks for your reply, but if I understand your suggestions correctly, then it will basically do the same as before using
if (maxResidualU < convergenceCriterionU && maxResidualP < convergenceCriterionP)
{
runTime.writeAndEnd();
}
It will stop the calculation when the convergence criteria are satisfied. My concern is
eqnResidualU = solve
(
UEqn() == -fvc::grad(p)
).initialResidual();
maxResidualU = max(eqnResidualU, maxResidualU);
If I compare maxResidualU with my convergence criterion, it will always check if the residuals of all velocity components satisfy the convergence criterion. What I need in my specific case is to stop the calculation if only the residuals of Ux and Uy satisfy my convergence criterion.
Best,
Florian
florian_krause is offline   Reply With Quote

Old   May 6, 2011, 09:54
Default
  #4
Member
 
Logan Page
Join Date: Sep 2010
Posts: 38
Rep Power: 15
Logan Page is on a distinguished road
Hmm yeah sorry i didtnt look at the code properly before replying :P

I actually have no idea how you would split the Ux and Uy residuals in OpenFOAM itself.

How i ended up terminating a simulation upon reaching steady-state was by controlling it with an octave script
1) run OpenFoam solver for n time steps (say 200) from previous time step
2) access the log file and pull out the initial residuals
3) check convergence criterion
4) if convergence reached stop else (1)

not very efficient i know, but for what i am doing i had problems with the Ux and Uy residuals, they never actually reach the convergence criterion but rather approached a flat line so i had to specify
(1-UxRes(@200)/UxRes(@0)) < 1e-4 for example

sorry i couldnt be of more help.
Logan Page 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
transsonic nozzle with rhoSimpleFoam Unseen OpenFOAM Running, Solving & CFD 8 July 1, 2022 07:54
Moving mesh Niklas Wikstrom (Wikstrom) OpenFOAM Running, Solving & CFD 122 June 15, 2014 07:20
How to write k and epsilon before the abnormal end xiuying OpenFOAM Running, Solving & CFD 8 August 27, 2013 16:33
Differences between serial and parallel runs carsten OpenFOAM Bugs 11 September 12, 2008 12:16
Unknown error sivakumar OpenFOAM Pre-Processing 9 September 9, 2008 13:53


All times are GMT -4. The time now is 00:59.