[vidyadhar]

Hi All,

I am running fluid(water) flow through a square channel at different Re.

For small values of Re, I can observe that the streamlines are not changing with time (after certain duration of simulation) and I conclude that the flow has reached steady state. [Is it correct?]

But, for higher values of Re, the streamlines are varying with time even after certain duration of simulation run. For how long should I run the simulation, at higher Re, to verify if the flow reaches steady state or if it remains unsteady?

Or is there a way to decide that the flow will be in unsteady state or periodic (after certain Re).

The mesh contains nearly 5 million cells.

Regards,
vidyadhar

[piu58]

> But, for higher values of Re, the streamlines are varying with time even after certain duration of simulation run. For how long should I run the simulation, at higher Re, to verify if the flow reaches steady state or if it remains unsteady?

First, this depends on your solver. If you use simple (which is a steady solver) you get a steady flow sometime. This requires that the physical reality of this Re is steady. If you try simple at high unsteady flows the solver may crash.

In a pipe flow you get eddies which evolves form the boundary an then move in the free stream. This is unsteady by nature. If you use a correct (unsteady) solver like piso or pimple you don't get a steady solution, which is physical correct.

I recommend to follow the course of a suitable flow variable and wait until it's frequency comes more or less to a steady value. You may try the pressure at the outlet for this or forces at the wall.

[vidyadhar]

Thank you Pilz for the reply.

I am using a transient solver.
If the flow becomes unsteady at higher Re, you are asking me to monitor the frequency of a flow variable. But, in my case, the pressure at outlet is equal to ambient and fixed. Can I try to plot the frequency of the flow velocity.
I do not know how to plot the forces at the wall.

Thank you.

[piu58]

> I do not know how to plot the forces at the wall.


Put in system/controlDict

functions
{
    forces
    {
        type            forceCoeffs;
        functionObjectLibs ( "libforces.so" );
        outputControl   timeStep;
        outputInterval  1;
        patches         ( nameOfMyWall );
        pName           p;
        UName           U;
        rhoName         rhoInf;
        log             true;
        rhoInf          1;
        liftDir         (0 1 0);
        dragDir         (1 0 0);
        CofR            (-0.5 0 0);
        pitchAxis       (0 1 0);
        magUInf         1;
        lRef            1;
        Aref            1;
    }
}

You have to adapt the name(s) of your wall(s) and the direction of the main stream which results in dragDir and liftDir. Beside this, the functions needs to now the free stream velocity magUInf and some other values.