
[Sponsors] 
February 28, 2024, 07:02 
Difference steady state / unsteady

#1 
New Member
Daniel
Join Date: Feb 2024
Location: Austria
Posts: 5
Rep Power: 2 
Hei, im getting used to work with openFoam but Im having a hard time understanding the difference between steady state solver like simple and unsteady like pimple/piso. a Flow is called steady state when it doesn't change over time. However I found a YouTube tutorial where simple is used vor a vortex shed problem (https://www.youtube.com/watch?v=Udt3...ist=WL&index=1)
The same problem is listed in tutorials > incompressible > PimpleFoam>LES>Vortexshed ) The results look nearly the same. I don't understand in general what openFoam writes as solution. Like in Simple you have to insert a TimeStep but this is the Step for the iterations of the solving process. So each step the solution gets more precise and your only interested in the last solution (in best case it converges before the last iteration step) bc this is the most precise solution for your case? However in unsteady solver  are these written solution time steps or also just the iterations and the higher the number of the written solution the more precise it is? Im just wondering because I wanna visualize how a vortex like in vortex shed behaves over time. How its moves, maybe it splits or it just dissipates. How can this be done? I came across the word transient simulation, but is transient not equal to unsteady state? Why is it possible in an unsteady solver to visualize the flow over time, like how it porpergates through the domain. Wen uniformfield is (0 0 0) and inlet boundary condition is for example (5 0 0) the velocity needs some time to hit an object (if the mesh is big enough) but the solution is always a given field which has already been fully developed. How can I see how the flow develops over time? And what is an example for steady state and unsteady? I always thought turbulence is an unsteady event. But why can you solve it then with steady state solver? Im confused af 

March 8, 2024, 16:04 

#2 
Senior Member
Josh Williams
Join Date: Feb 2021
Location: Scotland
Posts: 113
Rep Power: 5 
SimpleFOAM is a steadystate solver. Although it is solved with a timestepping method, the timesteps do not represent physical time. They just represent each iteration of your simulation. So the OpenFOAM output fields e.g. 100, 200, 300, are only really useful for understanding how your solver is progressing (not for extracting any physical information). For e.g. flow behind a cylinder, this may look similar to a transient simulation, it just means the solver is not really converging because it is not built for what you are using it for. For flows that are expected to be unsteady, such as vortex streets, use pimpleFoam.
With unsteady solvers such as pimpleFoam, the OpenFOAM output directories are converged results for each step in physical time. So they can show the transient behaviour of a flow. Regarding your last point on turbulence. Yes turbulence is unsteady, so any steady solver such as simpleFoam cannot be used to study turbulence except in an average sense (e.g. using RANS turbulence models such as kepsilon). You may consider reviewing a textbook such as Pope (2000). RANS turbulence models can also be used in unsteady flows (e.g. with pimpleFoam), giving socalled 'unsteady RANS'. Although the RANS closures are very empirical, so any errors in applying them to unsteady flows are likely compounded to produce completely incorrect results. For unsteady turbulence, I would recommend largeeddy simulation. 

March 9, 2024, 09:23 

#3 
Member
Shravan
Join Date: Mar 2017
Posts: 63
Rep Power: 9 
Hello,
To add to Josh’s explanation, Vortex shedding can simply get triggered because of round off or truncation errors in your simulation. So, you running a steady simulation has nothing to do with you observing the vortex shedding phenomenon. You can also do a simple test, by decreasing your tolerance values in fvSolution file, to let’s say 1e12. You will observe that the errors are too small the it will take a very long time for you to observe vortex shedding when compared to a case where the tolerances are 1e5 or 1e6. I have done this in the past and I had to simulate for a really long time to see the symmetry break and for the onset of vortex shedding (even with icoFoam). The question is if your results are accurate. If you need accurate results you have to use an unsteady solver in OpenFOAM. For more information on the same you can also check out the following paper and links: 1) Vortex Shedding for Flow Past Circular Cylinder:Effects of Initial Conditions (see the 2nd paragraph in the Inroduction section  https://www.papersciences.com/Laarou...ol320153.pdf) 2) Why does vortex shedding occur? 3) Can simpleFoam calculate accurate lift coefficient for vortex shedding of cylinder? Thanks 

Tags 
openfoam, steady state simulations, unsteady cfd 
Thread Tools  Search this Thread 
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Using steady state results to initialize flow in unsteady simulation  sabesj_  STARCCM+  3  February 8, 2023 05:03 
Steady State vs Unsteady  Zaktatir  FLUENT  2  May 11, 2016 05:00 
Solver for transonic flow?  Martin Hegedus  OpenFOAM Running, Solving & CFD  22  December 16, 2015 04:59 
is it possible to predict how long it takes to reach steady state solution in unstead  Alimohamadi_nasr  CFX  4  November 11, 2013 06:11 
About the difference between steady and unsteady problems  Lisa  Main CFD Forum  11  July 5, 2000 14:37 