# higher grid resolution --> later convergence?

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

 October 31, 2013, 06:35 higher grid resolution --> later convergence? #1 New Member   Udo Join Date: Oct 2013 Location: Munich Posts: 3 Rep Power: 4 Hello, I'm new here and in the CFD-World. At the moment I'm writing my bachelor's thesis about a fluid flow problem. My question: Can anybody give a brief explanation about the fact, that if you calculate the same case for different meshes (for a grid study) it needs more iterations to converge the higher the grid resolution respectively the cell number is? Thank you very much in advance!

October 31, 2013, 07:33
#2
Senior Member

Filippo Maria Denaro
Join Date: Jul 2010
Posts: 2,614
Rep Power: 33
Quote:
 Originally Posted by cfd-noob89 Hello, I'm new here and in the CFD-World. At the moment I'm writing my bachelor's thesis about a fluid flow problem. My question: Can anybody give a brief explanation about the fact, that if you calculate the same case for different meshes (for a grid study) it needs more iterations to converge the higher the grid resolution respectively the cell number is? Thank you very much in advance!

your question is not very clear... convergence in what? What kind of formulation and solver are you using and for solving what? What about your starting vector?

Just speaking in general, if you solve a linear system with an interative method, the rate of convergence of the method depends on the spectral radius of the iteration matrix. Thus, while changing h, you should check if the spectral radius changes.

 October 31, 2013, 09:43 #3 New Member   Udo Join Date: Oct 2013 Location: Munich Posts: 3 Rep Power: 4 Excuse me for not expressing clearly and thanks for your answer! Im not deep into theoretical backgrounds yet. And my advisor at the university is neither. I have three grids for a combustion furnace with inlet and outlet. Im using FLUENT with steady-state flow and pressure-based solver (SIMPLE). The residuals of the coarsest grid are converging earlier than they're doing with finer meshes. What is the reason? Why doest it need not only a longer computing time, but also more iterations till the flow is fully developed? I'm kindly asking for a quite simply answer, if there is one. I hope I made myself sufficiently clear enough. Thank you in advance.

October 31, 2013, 09:46
#4
Senior Member

Join Date: May 2012
Posts: 156
Rep Power: 6
Quote:
 Originally Posted by cfd-noob89 Hello, I'm new here and in the CFD-World. At the moment I'm writing my bachelor's thesis about a fluid flow problem. My question: Can anybody give a brief explanation about the fact, that if you calculate the same case for different meshes (for a grid study) it needs more iterations to converge the higher the grid resolution respectively the cell number is? Thank you very much in advance!
I don't know about the spektral radius, but if you have a line of people (representing the domain points) and each person can only talk to people next to him/her in order to transmit a message, then it will take more time to transmit that message from one end to the other with increasing number of people, right? If you can talk to the 2 closest people etc., then it will go a little bit faster, but the talking will take more time.

 October 31, 2013, 10:47 #5 New Member   Udo Join Date: Oct 2013 Location: Munich Posts: 3 Rep Power: 4 Thanks. That sounds plausible ;-)

 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 dialolema CFX 1 June 14, 2010 18:33 George Main CFD Forum 0 September 4, 2007 17:38 Ben Main CFD Forum 16 April 7, 2007 11:45 Chandra Shekhar Main CFD Forum 9 January 27, 2005 17:31 Ridwan Setiadi Arrizar Main CFD Forum 1 January 30, 1999 18:43

All times are GMT -4. The time now is 12:43.