CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   error iteration in FLUENT--difficult to take data (http://www.cfd-online.com/Forums/fluent/75035-error-iteration-fluent-difficult-take-data.html)

rosalina April 14, 2010 22:20

error iteration in FLUENT--difficult to take data
 
i have a model that consist of few volume in GAMBIT..while i mesh it all at once by Tet/Hybrid-TGrid, it was not completed..so I split the volumes and mesh one by one by mesh volume Tet/Hybrid-TGrid..

While running at FLUENT, error floating number shown..and the ENERGY LINE of the residual is about 10e+22..what a big number, so it wasn't convergent till now..

what should I do to make the convergence faster to take the data?
what was the problems actually?

Thank you..

Chris D April 14, 2010 23:01

If your energy residual is that large, then you definitely have not converged the problem. Without knowing anything else, I would start by checking your grid, because you mentioned that you had trouble generating it in the first place. A few things to check:

Make sure your geometry doesn't have any sharp edges. An infinitely sharp edge is non-physical, and could blow out your solution.

Make sure you don't have highly skewed elements in a region of large flow gradients (i.e., boundary layers.)

Make sure you don't have a rapid change in cell volume from one cell to the next. This can also blow out you solution.

To get more help, you might want to provide more information about what your doing. Pictures are always good! :)

rosalina April 14, 2010 23:53

1 Attachment(s)
thank you for your prompt reply...
actually,yes..I have seen comment about highly skewed elements when i mesh my object..but truly I don't really know about what was that mean..

my object is a reactor tank which consist of 116 cylinder of fuel on it..the initial problem is the grid of fuel always not completed, so i split the tank volume and fuels volumes..then i mesh it separately by mesh volume..when it works, if i'm not mistaken, comment about highly skewed elements shown..

[IMG]file:///C:/Users/Oca/AppData/Local/Temp/moz-screenshot.png[/IMG][IMG]file:///C:/Users/Oca/AppData/Local/Temp/moz-screenshot-1.png[/IMG]

-mAx- April 15, 2010 01:11

* highly skewed elements can provide convergence issue.
* disable energy equations and re-iterate. Just to be sure that your problem doesn't come from your thermal's set-up.
Anyway I would go back to gambit and check how many cells do you have with skewness > 0.9
Then correct your mesh

Philipov April 15, 2010 02:27

Hi rosalina,
do you want me to try meshing your model?

rosalina April 15, 2010 08:52

Quote:

Originally Posted by -mAx- (Post 254688)
* highly skewed elements can provide convergence issue.
* disable energy equations and re-iterate. Just to be sure that your problem doesn't come from your thermal's set-up.
Anyway I would go back to gambit and check how many cells do you have with skewness > 0.9
Then correct your mesh


i choose the energy equations because i want to insert thermal parameter in the boundary condition..can i do your suggestion while the thermal parameter was i inserted and still running the iteration?

rosalina April 15, 2010 08:55

Quote:

Originally Posted by Philipov (Post 254695)
Hi rosalina,
do you want me to try meshing your model?


are you serious?how could it be then?

-mAx- April 15, 2010 09:07

Quote:

Originally Posted by rosalina (Post 254777)
i choose the energy equations because i want to insert thermal parameter in the boundary condition..can i do your suggestion while the thermal parameter was i inserted and still running the iteration?

You have to restart your computation
Go to define/models/energy and disable energy.
Then initialize and iterate.

Philipov April 16, 2010 01:58

send the model to my e-mail

rosalina April 18, 2010 22:09

Quote:

Originally Posted by -mAx- (Post 254783)
You have to restart your computation
Go to define/models/energy and disable energy.
Then initialize and iterate.


if i would take the data of temperature (later when it was already converged)while the energy equation is disable..can i?

-mAx- April 19, 2010 01:51

This trick is just to dismiss your energy's set-up as source of error.
If you don't get problem anymore while disabling energy equation, then your problem comes from there.
You are in debugging phasis: once you find the source of your problem, you may re-compute your model properly.
It doesn't cost lot's of effort... just disabling energy model and re-iterate: either you don't have problem anymore, and you know that something is going wrong with your energy set-up, or you still have the problem, and you have to find in another direction (like you mesh for instance)

rosalina April 19, 2010 22:57

Quote:

Originally Posted by -mAx- (Post 255170)
This trick is just to dismiss your energy's set-up as source of error.
If you don't get problem anymore while disabling energy equation, then your problem comes from there.
You are in debugging phasis: once you find the source of your problem, you may re-compute your model properly.
It doesn't cost lot's of effort... just disabling energy model and re-iterate: either you don't have problem anymore, and you know that something is going wrong with your energy set-up, or you still have the problem, and you have to find in another direction (like you mesh for instance)



yap..seems my problems came from "energy"..i've done your suggestion, and not for long time iteration..it converged..but i couldn't get the temperature data..when i enable again the energy equation and initialize..i should iterate again to have the temperature data right?

rosalina April 19, 2010 23:21

Quote:

Originally Posted by Philipov (Post 254878)
send the model to my e-mail


still trying attach my file to your e-mail..by the way, thank you in advanced..

-mAx- April 20, 2010 00:46

Quote:

Originally Posted by rosalina (Post 255358)
yap..seems my problems came from "energy"..i've done your suggestion, and not for long time iteration..it converged..but i couldn't get the temperature data..when i enable again the energy equation and initialize..i should iterate again to have the temperature data right?

Then something isn't properly set in the problem set-up (energy)
In parallel a mesh without highly skewed cells would help

rosalina April 21, 2010 10:12

Quote:

Originally Posted by -mAx- (Post 255364)
Then something isn't properly set in the problem set-up (energy)
In parallel a mesh without highly skewed cells would help



What was it means'and what should i try then?sorry, actually i am still in early stage learning CFD...by the way, do you knoe how and where i could get the spesific Fluent training?

-mAx- April 22, 2010 00:59

Check your parameters involved in the energy equation, also your BC
Tutorials are first step for self-learning


All times are GMT -4. The time now is 16:54.