CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   steady/unsteady state (http://www.cfd-online.com/Forums/fluent/49667-steady-unsteady-state.html)

 richard October 31, 2008 06:05

steady/unsteady state

Dear all

As a beginner I just want to know when a steady and unsteady state simulations should be conducted. I assume that unsteady state means that there is vortex shedding. If I run a steady state analysis does this mean that there is no vortex shedding. Then how do I know if the simulation is steady or not

confused Richard

 mAx October 31, 2008 07:56

Re: steady/unsteady state

if all your BC aren't time dependant, then iterate with steady state solver. If you get good convergence, then you have the steady state solution If you don't get convergence, then it could be a unsteady phenomenon in your domain. You can monitor one variable (for instance static pressure on a specific body) while iterating. If you get a constant solution for this monitoring--> steady, if it's oscillate --> unsteady

 kospan November 2, 2008 14:14

Re: steady/unsteady state

Ok. but when i solute a problem which is steady with the stady state, i take different reults form unstedy state. what happen?

How operate the iteration number?

 mAx November 3, 2008 01:50

Re: steady/unsteady state

If your cas is steady, and if you compute it with unsteady solver, your first results won't be steady. You have an delta-t during which, your solution will change till the steady-one

 kospan November 3, 2008 04:34

Re: steady/unsteady state

ok. So, if i have a proble (i dont know if it is steady or unsteady), what procedure i follow for a really computational solution?

 mAx November 3, 2008 05:17

Re: steady/unsteady state

I would solve with steady solver, and I would monitor one target (for instance force on a body) and observing if the value remains constant or not with the iterations. if you get good convergence and your target is also well converged, I would say that you have an steady state solution. If not, switch on unsteady solver

 kospan November 3, 2008 08:49

Re: steady/unsteady state

thank you... i understand.

with other words. All problems i must solute,first, with the steady state and i look if the solute converge during the iteration... And if it converge the solute mean that is steady problem and i solute after with the steady state (with proper iteration number for which obtain the convergement), else the broblem is unsteady and i sollute it with unsteady mothod. I say right?

Now if the problem is unsteady, what parametres i use? 1.Iteation number on unstedy problem? 2.step time on implicit method?

I quest this because i have a hydrofoil problem and when i go to sollute it with unsteady state with step time(for example 1e-03 sec), i take importand different solution from i sullute with step time (1e-10 sec). why occur this?

 mAx November 3, 2008 09:18

Re: steady/unsteady state

Also with an unsteady problem, you can start with a steady "solution" and use it as initialization for the unsteady problem.

dt=1e-10 is a quite small time step, I would say far too small.

enable the Adaptive Time Stepping in Fluent, but you can easily have a coarse order for you time step, by dividing your reference's length by your reference's velocity.

 kospan November 3, 2008 10:09

Re: steady/unsteady state

I try to solute my hydrofoil problem but when i run the steady state with 5 or 10 or 30 iterations the sulotion which i take is wrong (it shows some turbulent flow front of hydrofoil and nothing other). Thus i dont take the common distribution of pressure up and down of hydrofoil. Why occur that? Do you know anythinh about this?

 mAx November 3, 2008 10:40

Re: steady/unsteady state

let the steady calculation well converge

 All times are GMT -4. The time now is 18:09.