
[Sponsors] 
May 26, 2009, 12:48 
Submerged fin, Convergence problem

#1 
New Member
Julie
Join Date: May 2009
Posts: 6
Rep Power: 8 
Hello everybody!
I have a convergence problem and it is making me crazy! I am going to explain the situation as clearly as possible. I have to model a fin completely immersed into water to determine the lift, drag and moment coefficient. Here are the flow characteristics: Speed : 1 m/s Temperature : 277.15 (3°C) Reynolds number : 7900 Rho : 1034.53 kg/m3 Viscosity : 0.00167 Pa/s Fin angle of attack= 0° This is how I use Fluent:
First problem : I have the following error "turbulent viscosity limited to 1e5 in 1 cell". Is it an important error? I would say no because it is only in 1 cell, but give me your opinion please. Second problem : CL, Cd and Cm don't stop oscillating! There is no convergence after 5,000 iterations. I tried to decrease the under relaxation factors but it didn't help to converge. Third problem : the continuity residuals remain at 1. But it is not the major problem! the second one is!! I really don't know what to do to solve these problems! I am not used to use FLUENT for "water problems", so maybe I forgot something. Help me, please!! thank you in advance Julie 

May 27, 2009, 03:45 

#2 
Member
Ralf Schmidt
Join Date: Mar 2009
Location: Austria
Posts: 67
Rep Power: 8 
Hi,
there might be something wrong with your grid.... check grid quality and yplus values at your walls.... For oscillation of the residuals and residuals, see oscillation and Oscillations in Results and Residuals Best wishes Ralf
__________________
CFD  nothing but Colourful Fluid Dynamics Last edited by Ralf Schmidt; May 27, 2009 at 05:03. 

May 27, 2009, 05:33 

#3 
New Member
Dimitrios Mylonas
Join Date: Apr 2009
Location: Glasgow, Scotland
Posts: 15
Rep Power: 8 
What are the dimensions of your fin?
I might be wrong, but your Re is relatively low, so you might be in the laminar regime. I'd recommend you make a run with laminar, then try to move on to a turbulence model, it could help you with having an initial idea of what values to expect. Also why are you setting pressure outlet to 1.51e7 Pa? Hope it helps, please anyone feel free to correct me if I'm wrong so I can learn too. Dimo 

May 27, 2009, 06:05 

#4 
Member
Ralf Schmidt
Join Date: Mar 2009
Location: Austria
Posts: 67
Rep Power: 8 
How do you calculate your Re number?
__________________
CFD  nothing but Colourful Fluid Dynamics 

May 27, 2009, 11:28 

#5 
New Member
Julie
Join Date: May 2009
Posts: 6
Rep Power: 8 
I calculate my Reynolds number as follows:
Re= rho* V* L/ mu rho=1034.53 kg/m3 V=1m/s L= mean aerodynamic chord (I have a sweptwing) mu= dynamic viscosity= 1.67e3 Pa.s and the result is 7940. I am running a laminar simulation to have an idea of the values and to check wether it is my grid or my model that is wrong. I'll keep you posted Julie 

May 27, 2009, 11:35 

#6  
New Member
Julie
Join Date: May 2009
Posts: 6
Rep Power: 8 
Quote:
root chord: 0.1515 m tip chord: 0.136 m wingspan : 0.2 m I am setting pressure outlet at 1.51e7 Pa because my fin is at a depth of 1500 m. So, as the pressure is about 151 bars (1.51e7 Pa) at 1500 m, I set this value. Please, tell me if I'm wrong to do that. Julie 

May 27, 2009, 12:14 

#7 
New Member
Dimitrios Mylonas
Join Date: Apr 2009
Location: Glasgow, Scotland
Posts: 15
Rep Power: 8 
Ok, let us know how you get along. Are you modifying the pressure in the Operating Pressure tab?
Dimo 

May 28, 2009, 16:17 

#8 
New Member
Julie
Join Date: May 2009
Posts: 6
Rep Power: 8 
Hi,
To answer your question Dimo, yes, I changed the Operating Pressure and set it to 0. I ran a simulation with the laminar model with an angle of attack of 0° and the results were still oscillating! So, I decided to run a simulation with a higher angle (2°) and it is working. I also ran a simulation with the laminar model ,with a structured grid, at alpha=0° and it is also oscillating. But for alpha=2°, the solution seem to converge (the simulation is not finished yet). I wonder why there is a problem when alpha=0°. This is the simplest case and there should not have any problems... If you have any ideas, please, let me know. I have one question: When are we certain that the solution is converged? At which precision a value is said stable? I mean, my solution looks like CL=3.25e4 +/ 0.1e4. Can I consider that it is converged? I hope my question is understandable! Thanks for your help! I'll let you know the results when I get back from holidays Julie 

June 26, 2009, 10:26 

#9 
New Member
Julie
Join Date: May 2009
Posts: 6
Rep Power: 8 
Hello everybody!
Actually, my boundary conditions were wrong...I set every faces with "Wall" except the inlet and the outlet, so it made my flow unsteady. Now that I have changed them everything is fine, no convergence problem anymore. Thank you for helping me. Bye 

July 6, 2009, 09:45 

#10 
New Member
Dimitrios Mylonas
Join Date: Apr 2009
Location: Glasgow, Scotland
Posts: 15
Rep Power: 8 
What B.C. did you change?


July 6, 2009, 10:47 

#11 
New Member
Julie
Join Date: May 2009
Posts: 6
Rep Power: 8 
My control volume is a 3Drectangle with the fin attached to one vertical face. I made a drawing for you to understand clearly (the file is attached). So, when you look at the drawing, I set the green face with "pressure oulet", the yellow one with "symmetry" and the other faces are set with "velocity inlet".
I set those BC because with only the "wall" condition it seems like the fin is in a closed box, where the flow passes through it. I hope my anwer is clear. Bye Julie[IMG]file:///C:/Users/julie/AppData/Local/Temp/mozscreenshot.jpg[/IMG] 

Tags 
convergence, oscillation, submerged, water 
Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
convergence problem  alexandre  FLUENT  2  June 14, 2007 14:31 
convergence problem  ammi reddy  FLUENT  1  March 28, 2007 17:01 
3D Fluid Flow Convergence problem  Emily  FLUENT  2  March 21, 2007 23:18 
Non Convergence of 3D Heat transfer cfd problem  Balraj  Main CFD Forum  3  December 9, 2004 01:24 
Convergence problem  Praveen  FLUENT  0  September 15, 2000 06:34 