
[Sponsors] 
June 15, 2014, 19:02 
offsetCylinder tutorial, steady state,

#1 
New Member
Riccardo
Join Date: May 2014
Posts: 10
Rep Power: 4 
Greetings all, I am trying to run the offsetCylinder "tutorial" with different nonnewtonian fluids at flow velocities such that the regimen stays laminar. So, by modifying only the viscosity parameters I am able to run the simulation and to obtain a visual representation of a state that appears to be laminar and steady. However from the log of the simulation only one component of velocity appears to have converged after the simulation. This means that after a certain timestep, only that component of velocity shows 0 iterations.
I ran the simulation for a long time so I don't think the simulation duration is of any importance anymore. 1Is this a normal behavior of the nonNewtonianIcoFoam solver? 2If yes, without performing new simulations with steady state solvers, how can I understand at what approximate timestep a steady state is acceptably reached (merely for presentation purpouses)? 3Unfortunately C syntax is still misterious to me so I am not really aware of what I'm doing most of the time. Could someone sum up what is going on when running nonNewtonianIcoFoam, I mean from a mathematical/numerical point of view? Thank you Last edited by tabsquare; June 15, 2014 at 20:02. 

June 16, 2014, 11:08 

#2 
New Member
Riccardo
Join Date: May 2014
Posts: 10
Rep Power: 4 
I attach a plot of the residuals. I am not really sure how to interpret the results. Indeed this is a case where both components of velocity converge. But what about pressure?


June 18, 2014, 12:27 

#3 
Senior Member
Joachim Herb
Join Date: Sep 2010
Posts: 373
Rep Power: 10 
The pressure equations is solved twice (or more often, probably 10 times) per solution of the other equations because you have set (in fvSolutions) either nCorrectors bigger than 1 or nNonOrthogonalCorrectors bigger than 0.


July 15, 2014, 12:45 

#4 
New Member
Riccardo
Join Date: May 2014
Posts: 10
Rep Power: 4 
OK still I have no good knowledge of the math behind this, but I get what you meant to some extent.
I have another question. I attached a plot of velocity residuals as a function of time. Pressure residuals show a similar behavior. Final residuals of velocity decrease until "tolerance" is reached, then the curve of final residuals merges with the Initial residuals one and both keep decreasing until the end of the simulation(manually stopped) is reached. To what solver is this second decrease due to? I am using the same files contained in the incompressible/nonNewtonianIcoFoam/offsetCylinder/ tutorial. I only made some modification to transportProperties and controlDict. 

Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Time step dependence of convergence behavior of steady state simulations in CFX  Chander  Main CFD Forum  5  December 23, 2013 06:31 
Constant velocity of the material  Sas  CFX  15  July 13, 2010 08:56 
false time step implementation for steady state  Kushagra  CFX  1  June 22, 2008 19:06 
About the difference between steady and unsteady problems  Lisa  Main CFD Forum  11  July 5, 2000 14:37 
Steady state formulation of turbulence ??  Jitendra  Main CFD Forum  1  June 27, 2000 18:49 