|
[Sponsors] |
October 12, 2005, 03:57 |
continuity convergence
|
#1 |
Guest
Posts: n/a
|
Hi all! What is happend if in my model all the residuals falls down under the imposed value and become flat, only the continuity residual rest flat at 0.5! I've tried many meshing schemes but the result is the same!
Thank you! |
|
October 12, 2005, 08:00 |
Re: continuity convergence
|
#2 |
Guest
Posts: n/a
|
well this probably means you have some mesh elements with very high skew, if it is not a multiphase calculation
|
|
October 12, 2005, 08:13 |
Re: continuity convergence
|
#3 |
Guest
Posts: n/a
|
Is a single phase flow. I think the mesh is OK. The model is 2D and mesh is quad (and all cells are perfect rectangulaire).
|
|
October 12, 2005, 08:22 |
Re: continuity convergence
|
#4 |
Guest
Posts: n/a
|
Try to underrelax factors PRESSURE or DENSITY
Andrea |
|
October 12, 2005, 09:09 |
Re: continuity convergence
|
#5 |
Guest
Posts: n/a
|
Hello Cristion,
If it is a 2D analysis, try solving your model in 2d,dp. This should give you better convergence. Hope this helps and good luck. Miqdad |
|
October 12, 2005, 09:32 |
Re: continuity convergence
|
#6 |
Guest
Posts: n/a
|
Thanks Miqdad, but I already tried this. The continuity residual evolution is the same!
|
|
October 12, 2005, 09:50 |
Re: continuity convergence
|
#7 |
Guest
Posts: n/a
|
Hi Cristion,
Do you mind sending me your model so that I can have a look? You can email it to me if it is not a large file. Many thanks, Miqdad |
|
October 12, 2005, 10:04 |
Re: continuity convergence
|
#8 |
Guest
Posts: n/a
|
Are you monitoring more than just the residuals? You should be using other monitors to judge convergence, not just the residuals.
It could be related to your solver settings. What are you modeling, and which solver did you chose? It sounds like you're on the right track with performing a mesh sensitivity and trying 2ddp. The only other thing might be the solver settings. Let us know some more about your model and what you have for settings. IMHO the residuals are only good for telling you how your model is behaving. I don't think they are good for judging convergence. Part of the reason is that they are scaled based on your first iteration. Another reason is that some models tend to "drift". The residuals will be small, but since the residuals are related to the change for that iteration, a lot of small changes accumulated over several iterations can add up to a decent change. So I've seen models where the continuity residuals don't drop 1e-03, but the model really is converged, and I've seen models drop 1e-06, but the monitors show that the results really aren't converged. It all depends on your model and your initialization. Hope this helps, and good luck, Jason |
|
July 27, 2011, 02:01 |
|
#9 |
New Member
neville
Join Date: May 2011
Posts: 12
Rep Power: 14 |
Hi , I am facing a similar problem. I am modeling 2D single phase flow in a pipe. The residuals drop and the solution converges but when I include Gravity effects ,the continuity residual doesnt drop while the others do.
tried refining the mesh and also changing the UDF's and the solver but no success. Thanks in advance. |
|
July 27, 2011, 08:10 |
|
#10 |
Member
Neil
Join Date: Jun 2009
Location: The Moon
Posts: 42
Rep Power: 16 |
Have you tried changing both pressure and momentum discretisation methods to 2nd order or even QUICK for momentum? The continuity residual is normally the most troublesome one to get converged adequately but I generally find that 2nd order pressure discretisation will fix it.
|
|
July 29, 2011, 07:46 |
|
#11 |
New Member
neville
Join Date: May 2011
Posts: 12
Rep Power: 14 |
Hello, Thanks for your reply. I did try 2nd order upwind scheme for pressure and QUICK for momentum but continuity didn't converge (10^-2).
I also tried using density based solver but that didn't work either. I am simulating a pipe flow using a nist-real-gas-models with gravity and buoyancy effects. my BC's are mass flow inlet, pressure outlet and const wall heat flux. when i solved it without the gravity effects and heat flux it did converge but no luck after that. Is there any other way to approach this problem ? |
|
May 6, 2014, 04:26 |
|
#12 |
New Member
mitra
Join Date: Dec 2013
Posts: 2
Rep Power: 0 |
hi
did you solve your problem , i have the same problem but dont know how solve it ,please help me |
|
May 6, 2014, 04:51 |
|
#13 |
New Member
neville
Join Date: May 2011
Posts: 12
Rep Power: 14 |
Hi,
From your message I get it that you are having convergence problems in the continuity. The residuals are a good indication of convergence but it is not necessary for the residuals to drop below a specified value for all problems. It depends on the flow phenomena and on how complicated your domain is. Check Reports-fluxes-mass flow rate If mass balance is achieved you can say that your solution is converged. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
How to write k and epsilon before the abnormal end | xiuying | OpenFOAM Running, Solving & CFD | 8 | August 27, 2013 15:33 |
Force can not converge | colopolo | CFX | 13 | October 4, 2011 22:03 |
Upgraded from Karmic Koala 9.10 to Lucid Lynx10.04.3 | bookie56 | OpenFOAM Installation | 8 | August 13, 2011 04:03 |
IcoFoam parallel woes | msrinath80 | OpenFOAM Running, Solving & CFD | 9 | July 22, 2007 02:58 |
Could anybody help me see this error and give help | liugx212 | OpenFOAM Running, Solving & CFD | 3 | January 4, 2006 18:07 |