[mushtime]

Dear Foamers:
I install the LEMOS-2.3.x in openfoam2.3.1 and use the 'inflowGenerator' for producing the turbulent inlet of the jet flame.
My use the case of flameC which is similar to flameD (sandia data), with mean velocity of : JET(29.7m/s),PILOT(6.8m/s) and COFLOW(0.9m/s).
My simulation domain is cylindrical domain with outer diameter of :JET(7.2mm),PILOT(18.2mm) and COFLOW(72mm) and axial length of 350mm.
For the first 0.04s, I use groovy B.Cs to produce inlet velocity profile of JET part. And it all goes fine.
But after I add the ''decayingTurbulenceInflowGenerator'' patch of the LEMOS to generate the turbulent inlet (with groovy BCs as the velocity profile), strange velocity profile just occurs.
My U field of 0.04s are listed below:

boundaryField
{
    JET
    {
        type            decayingTurbulenceInflowGenerator;
        LField          uniform 0.003;
        RField          uniform (0.05 0 0 0.05 0 0.05);
        direction       1;
        refField        nonuniform List<vector> 
        ...(the groovy B.C profile of jet velocity)
        value           nonuniform List<vector>
        ...(the groovy B.C profile of jet velocity)

Which pilot and coflow use turbulentInlet patch:

    PILOT
    {
        type            turbulentInlet;
        fluctuationScale (0.05 0.03 0.03);
        referenceField  uniform (6.8 0 0);
        alpha           0.1;
        value           nonuniform List<vector> 
        ...
}

    COFLOW
    {
        type            turbulentInlet;
        fluctuationScale (0.01 0 0);
        referenceField  uniform (0.9 0 0);
        alpha           0.1;
        value           nonuniform List<vector> 
        ...
}

At 0.04s, the velocity field is like Screenshot from 2016-05-23 23^%16^%22.jpg
After adding the LEMOS turbulent inlet BCs, and calculate for about 0.0032s the velocity field:
Screenshot from 2016-05-23 23^%19^%31.jpg
To adjust the velocity showing scale and the U_x,U_y and P field is listed below:
Screenshot from 2016-05-23 23^%20^%03.jpg
U_y.png

[mushtime]

Can anybody give some suggestions that whether I set the boundary of JET right? Thanks very much!!

[tomf]

Hi,

I would not be sure if the LEMOS is to blame for this. There seems to be something going on near your outlet. Just to make sure I would advise you to keep running without the LEMOS part and see if the same instability occurs. If so you probably need to have different boundary conditions at the outlet or move your outlet further away (and maybe coarsen the mesh there as well) to prohibit backflow from the outlet.

Regards,
Tom

[mushtime]

Thank you Tom,
I will give it try according to your suggestion and reply the result to the thread.

[mushtime]

Hello, Tom
I think I should mention the patch here that:
The outlet patch I use for p is wavetransmissive：

OUT
    {
        type            waveTransmissive;
        gamma           1.4;
        fieldInf        100000;
        lInf            0.17;
        value           uniform 100000;
    }

And my velocity for OUT is set like below.

    OUT
    {
        type            pressureInletOutletVelocity;
        value           uniform (0.9 0 0);
    }

Actually, I have tried fixedValue for p and InletOutlet for velocity. Still it does not work.
It seems that, the flow just "accumulates" at the outlet so some unphysical things happen.
I find some people use advective patch boundary, but it seems that the fraction of fixedValue and zeroGradient is fixed. So I don't not know whether it is good to try advective path.
Can you suggest me some proper B.Cs for the pressure and velocity to solve the "accumulation" problem？ And I will first try to lengthen the simulation domain. Thank you ahead of time.

[tomf]

Hi,

I have sucessfully used the advective boundary condition recently indeed. Also wavetransmissive is derived from advective if I am not mistaken. But I did coarsen the mesh closer to the outlet, which may have helped.

You may want to try the combination of totalPressure on p with pressureInletOutlet on U, but I am not sure if this will solve the issue.

Regards,
Tom