|
[Sponsors] |
June 21, 2011, 14:40 |
Reason for oscillating residuals
|
#1 |
New Member
Fayaaz Hussain
Join Date: Dec 2010
Posts: 12
Rep Power: 15 |
Dear all,
I would like to know the reason behind oscillating residuals. I am trying to simulate flow through a CD nozzle. I had mesh which gave me a converged solution for a pressure ratio of 6. The same mesh was useless with other pressure ratios. I got a error code 255 or error code 1 in ansys cfx. I remeshed by making a little coarser mesh. It works fine with pressure ratio 5. But my residuals started oscillating at the end but somehow managed to converge. The same mesh is being used for a pressure ratio 0f 4 but my residuals are oscillating badly. I heard that a very fine mesh could be one of the reasons for oscillations but my mesh seems to be coarse in my point of view. Can anybody help me in understanding the reasons behind the oscillations? Thanks in advance |
|
June 21, 2011, 19:56 |
|
#2 |
Senior Member
|
Hi Fayaaz,
There are large numbers of reasons why residuals may oscillate, below are a couple of ideas/reasons: 1) The physic of the flow dictates that there is no steady-state solution, only transient ones. Therefore once the solution is within the transient behavior, the residuals oscillates. Coarsening of the mesh may help with this problem. The "appropriate" way is to conduct a transient simulation and time-average it - but it is time expensive; 2) The code implementation (some random if statement) means that the flow is switching between 2 (or more) different models/limiters and the like. This creates oscillation in residuals. You should look at convergence beyond residuals, i.e. monitoring of the flow field and variable of interest, and look at the area of large residuals (by plotting the residuals distribution). Regards, Julien
__________________
--- Julien de Charentenay |
|
June 27, 2011, 13:51 |
|
#3 |
Member
Casey
Join Date: Jun 2009
Posts: 98
Rep Power: 17 |
lower your under relaxation facrors (URFs). Osciallting residuals (genearlly) means the solution is changing too much between iterations and overshooting the solution.
|
|
August 21, 2014, 11:31 |
|
#4 | |
Member
Marcel Jay
Join Date: Jun 2014
Location: Berlin
Posts: 52
Rep Power: 12 |
Quote:
My mesh for example has about 700k elements in a quite simple geometry. Would that be too fine? Cheers |
||
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[Tutorials] Tutorial of how to plot residuals ! | wolle1982 | OpenFOAM Community Contributions | 171 | February 20, 2024 03:55 |
what to monitor besides residuals? | franzdrs | FLUENT | 5 | March 21, 2013 04:59 |
Discussion: Reason of Turbulence!! | Wen Long | Main CFD Forum | 3 | May 15, 2009 10:52 |
Mesh size and solver residuals... | Scott | CFX | 5 | December 15, 2008 18:10 |
Convergence - scaled vs unscaled residuals | HS | FLUENT | 1 | November 7, 2005 06:45 |