# under relaxation factors

 User Name Remember Me Password
 Register Blogs Members List Search Today's Posts Mark Forums Read

 November 22, 2009, 15:36 under relaxation factors #1 New Member   sateesh Join Date: Oct 2009 Posts: 9 Rep Power: 9 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: Surabaya, Indonesia Posts: 153 Rep Power: 9 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: 9 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: 9 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: 9 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: 9 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: 9
Quote:
 Originally Posted by sateesh 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
Do you mean that, after the residuals decrease to some value (say 10^-2), you reduce the URF and your residuals decrease to some lower value (say 10^-6)?

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: 9 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 Display Modes Linear Mode

 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 OffTrackbacks are On Pingbacks are On Refbacks are On Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post philippose OpenFOAM Running, Solving & CFD 19 March 20, 2014 05:39 Mike Main CFD Forum 7 May 21, 2005 12:41 Tim Phoenics 3 June 30, 2004 02:03 zhujianguo Phoenics 1 August 5, 2003 06:17 zhujianguo Phoenics 1 July 15, 2003 11:11

All times are GMT -4. The time now is 10:14.

 Contact Us - CFD Online - Top