|
[Sponsors] |
November 22, 2009, 15:36 |
under relaxation factors
|
#1 |
New Member
sateesh
Join Date: Oct 2009
Posts: 9
Rep Power: 16 |
Dear All,
I am getting convergence for once of my problem only if i reduce the under relaxation factors to ten times in middle of the simulation. can i reduce the under relaxation factor to any fraction? |
|
November 23, 2009, 07:26 |
|
#2 |
Senior Member
teguh hady
Join Date: Aug 2009
Location: Saga, Japan
Posts: 222
Rep Power: 17 |
Hi
I've read in books that decrease alpha will increase convergence. the limitation of value to decrease is must be >= 0.2. maybe it can be used for your limitations to set alpha. teguh
__________________
Now Or Never!!! |
|
November 23, 2009, 17:42 |
|
#3 |
Senior Member
Chris
Join Date: Jul 2009
Location: Ohio, USA
Posts: 169
Rep Power: 16 |
The under-relaxation factors are used to make the solution more stable. They should not affect your final solution, only how many iterations it takes to reach it. Decreasing the URF's will reduce the residuals, but the reduction is artificial. By what measure are you saying that your simulation is converged?
|
|
November 23, 2009, 19:13 |
URF doubts
|
#4 |
Member
R. Roy
Join Date: Mar 2009
Location: India
Posts: 52
Rep Power: 17 |
yes, i also do not understand these URF parameters. for the moment, i feel like i am just trying to achieve conevrgence through trial and error using these URFs. also what would be the connection between these parameters among each other ?
can you refer me any book/articles where I can learn the concepts behind these parameters ? |
|
November 23, 2009, 21:11 |
|
#5 |
Senior Member
Chris
Join Date: Jul 2009
Location: Ohio, USA
Posts: 169
Rep Power: 16 |
In the FLUENT 6.3 User's Guide, see Section 25.4.4. The under-relaxation factors are used to prevent large changes in solution values from one iteration to the next. Large fluctuations of the solution variables are possible in the initial stages of the simulation, so the under-relaxation factors keep it from blowing up during this time. After the flow has settled down and is nicely headed towards steady-state, they shouldn't be necessary.
If you're finding that your simulation will only "converge" if you decrease the under-relaxation factors to a very small amount, then there is probably something wrong. By setting low values of the URF's, you're only allowing the variables to change by a very small amount, so the only reason that it looks like the residuals/surface monitors have stopped changing is because you have forced them to stop changing. |
|
November 24, 2009, 13:02 |
|
#6 |
New Member
sateesh
Join Date: Oct 2009
Posts: 9
Rep Power: 16 |
Hi,
Thanks for your valuable information on URF. I am trying to solving flow, turbulence, energy, species transport equations simultaneously in a irregular 3-D geometry. after long time, there is no change in the residuals. the plot becomes straight line. in this case if i reduce the URF by 10 times, i am getting the required convergence. this is my problem. can you suggest me what should i do? or where is the problem? Thanks in Advance- Sateesh |
|
November 24, 2009, 14:37 |
|
#7 | |
Senior Member
Chris
Join Date: Jul 2009
Location: Ohio, USA
Posts: 169
Rep Power: 16 |
Quote:
If that's true, then your simulation is not necessarily converged. The residuals are only decreasing because you decreased the URF. To monitor convergence, you need to consider other factors besides the residuals. Just because your residuals dropped below some value does not mean that your simulation is converged. You should monitor some other solution values to help you judge convergence. If you are performing a steady-state calculation on an open system, for example, you should monitor the net mass flow rate, which should converge to zero. I would suggest monitoring some parameters during the simulation that are important to you. You can setup the monitors under the solve-->monitors menu. If these monitors converge, then that is the best you can do with that particular grid, and with those particular boundary conditions and solver settings. Changing any of those things may change your solution, and you should investigate what happens when you change them, but your URF should not influence your converged solution. |
||
November 25, 2009, 08:30 |
Shaddow Wall / Help Sought
|
#8 |
New Member
sateesh
Join Date: Oct 2009
Posts: 9
Rep Power: 16 |
Hi,
I have two adjacent zones one is solid (coal) and another one is fluid (Occupaid by Oxygen). A reaction is taking place at the solid surface/interface i.e. coal is getting combusted with oxygen (which is coming from the fluid zone). It is a exothermic reaction. because of this reaction, heat transfer 'll takes place in both the zines. I need to obtain the temperature profile in side the two zones. I need to create a geometry in a way that only heat should get transfer through the interface between the two zones. How can i create the shadow wall for on side interface? or can any body suggest me how can i solve this problem in FLUENT? Thanks in Advance |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Relaxation Factors for Transient solvers | philippose | OpenFOAM Running, Solving & CFD | 19 | March 20, 2014 05:39 |
relaxation factors and time accuracy | Mike | Main CFD Forum | 7 | May 21, 2005 13:41 |
Relaxation Factors | Tim | Phoenics | 3 | June 30, 2004 03:03 |
relaxation factors | zhujianguo | Phoenics | 1 | August 5, 2003 07:17 |
relaxation factors adjust | zhujianguo | Phoenics | 1 | July 15, 2003 12:11 |