
[Sponsors] 
January 5, 2005, 00:48 
Converge Problem

#1 
Guest
Posts: n/a

I am using FLUENT to model spray in a big tank, the spray gun has several outlet, the one in the center is for water and the others are for air. I set these openings as velocity_inlet, and the outlet of tank is set to be pressure_outlet. I tried some ways, but it does not converge especially the residule of volume fraction of water( the water speed is 1.9m/s, and the air is 20m/s).
What's more, when I calculate, the warning message "reversed flow in 94 faces on pressure_outlet" appear. Does this means there is something wrong with my codes? This is really a big headache to me, I really wish some one could help me or give me some suggestion, Thanks so much!!! 

January 5, 2005, 00:51 
Re: Converge Problem

#2 
Guest
Posts: n/a

BTW: I am using Multiphase model. I wonder if I should use Discrete Phase Models. The turbulent model is kepsilon


January 5, 2005, 07:49 
Re: Converge Problem

#3 
Guest
Posts: n/a

Are the values you defined for backflow on the pressure outlet reasonable?
Moving your pressure outlet further downstream might be beneficial. You might also try the target mass flow rate option on the pressure outlet. 

January 5, 2005, 12:19 
Re: Converge Problem

#4 
Guest
Posts: n/a

Now, I can get the solution converge without slip velocity, but when I add slip velocity, all the residuals fly to the sky.
Is dicreted phase model a better choice? 

January 6, 2005, 08:22 
Re: Converge Problem

#5 
Guest
Posts: n/a

I think this problem caused due to the precision of ur geometry.
U just open your case file in 2ddp when u have made the model in 2d or in 3ddp when u have made the model in 3d. then try to solve it. see whether it is working . soundar 

January 6, 2005, 11:28 
Re: Converge Problem

#6 
Guest
Posts: n/a

After about 5000 iterations, it converged without slip velocity. But the contours are very strange, and the velocity seems so small. I don't know how to get better results. I never use 3ddp, does it has any difference with 3d?
I am going to put my outlet downstream to see if it will become better. In the inlet boundary condition, I specify turbulent intensity (I) and hydraulic diameter (Dh), but why other parameters do not change, eg. k and epsilon. Is it normal? Thanks a lot!!! 

January 6, 2005, 11:34 
Re: Converge Problem

#7 
Guest
Posts: n/a

Acctually, I don't know what I could specify at the pressure_outlet. I just know the pressure there, nothing else. Do you have some suggestion on how to specify pressure_outlet?
Where can I choose target mass flow rate option? 

January 7, 2005, 11:39 
Re: Converge Problem

#8 
Guest
Posts: n/a

Now the solution is converged. But the reversed flow warning is still exist. I have moved the outlet downstream. Is the warning an error? Any other method I can solve this problem?


January 25, 2005, 01:24 
Re: Converge Problem

#9 
Guest
Posts: n/a

Any body could suggest for my convergence problem.
i'm running at steady state species transport model. while iterating it shows reversed flow in faces 14 at pressureoutlet 5. however it is running for next iterations is it relates with convergence or it may cause err in the result. my outlet is outflow and inlets are mass inflow. waiting for response soundar 

January 25, 2005, 08:29 
Re: Converge Problem

#10 
Guest
Posts: n/a

I'm assuming when you say "my outlet is outflow" that you're talking about the pressureoutlet 5. If it's a pressureoutlet, then you defined some values for reversed flow in the boundary conditions window. "reversed flow in XX faces at XXXX" just tells you that you're getting some flow into your system through what you assumed was your pressure outlet and it's using the reversed flow values you defined. If the reversed flow is real, and your reversed flow settings are true, then it shouldn't mess with convergence or cause an error. If this isn't real, then either it will work itself out during convergence, or your model setup is wrong. One example of where I've seen this: we had a flat plate representing a portion of a fuselage, and a large bluff body sticking out of it. The pressure outlet wasn't far enough downstream, so the recirculation behind the bluff body was pulling air back into the control volume through the pressure outlet. We ended up relocating the B.C. to avoid this, but it wasn't actually causing problems with the results.
Hope this helps, and goodluck Jason 

January 28, 2005, 09:11 
Re: Converge Problem

#11 
Guest
Posts: n/a

Thank you jason for your detail explanation.
I'm doing species transport model with various chemical reactions for gasification. After 400 iterations it has shown inert as 99.99% where as for mole fraction of CO,Co2 it shows very less value of the order 4.0 *10^16. any suggestion please for rectifying this so that reaction will take place inside. thanks from soundar 

January 28, 2005, 15:20 
Re: Converge Problem

#12 
Guest
Posts: n/a

Now you're getting outside my knowledge base. I've never dealt with species transport (I'm strictly an aerodynamics guy).
Sorry, and I hope someone else on here can help. Goodluck, Jason 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Handling cyclic BC from gambit to openfoam for a cascade airfoil problem  OF 1.6  maverick  OpenFOAM Other Meshers: ICEM, Star, Ansys, Pointwise, GridPro, Ansa, ...  2  June 18, 2011 04:36 
Velocity profiles problem behind the elbow (3D problem)  kabat73  FLUENT  8  May 9, 2010 04:26 
converge problem  snowshovel  CFX  13  January 18, 2007 22:05 
Converge problem of 3D free jet flow  Jen  FLUENT  11  January 24, 2005 01:21 
problem with using colocated code  Jack  Main CFD Forum  0  December 15, 2002 01:15 