Dear all,

I have a domain which contains a rotating impeller. But the domain consists of non-rotating fluid zones before and after the impeller. I used a rotating reference frame attached to the fluid zone. But in the velocity vector plot of the solution, it appears that the velocity of the elements in the non-rotating part of the fluid have components in all directions. How can I fix this problem?

Thank you in advance

I think sliding mesh may be a better choice.

how u can say this . it depends on need and requirements

she might use multiple reference frame, but usually MRF results are no good because:

If you do 5 or 6 MRF simulations of the same geometry but with the rotating parts fixed in a difference position for each simulation: each 5 simulaiton has different result. Try it you will see!

To avoid this you need to average the 5 (or 6) simulations when post-rpocessing or use sliding grid.

B.L.

If you will plot absolute velocity vectors in front of the impeller you should see only axial direction vectors (if you set such inlet boundary condition!), but after the impeller absolute velocity vectors will not be axial if you don't have a stator blade stage to straighten the flow (the impeller will add a circumferential and a radial component for velocity).

On the other way, if you will plot relative velocity vectors, then pretty much all vectors in the flow domain will have all-direction components. It is perfectly normal.

Try to think more deeply about these concepts: absolute and relative velocity!

Razvan

Dear all,

Thank you for your prompt reply.

Let me put my question in this manner. What will happen if I used a rotating reference frame attached to the fluid to simulate a flow through a pipe. Is it possible to do so in fluent theoretically and yet get meaningful result?

In my domain, there is part of the fluid which is outside the influence of the rotor to get a relatively developed flow condition at the rotor inlet.

Best regards

r u sure. can u give some reference or any result u have done ur self .

i am will gratful to u

regards sam

[ingarcia1703]

Hello community

I would like to ask you recommendations in order to solve a single reference frame compressible flow in OpenFOAM. I have been adding some lines to the original SRFModel available, and also create a new compressible solver "SRFrhoSImpleFoam". The issue is that I still receive the following message from units problem at the Coriolis and centrifugal force:
--> FOAM FATAL ERROR:
incompatible dimensions for operation
[U[1 -2 -2 0 0 0 0] ] + [(Fcoriolis+Fcentrifugal)[0 1 -2 0 0 0 0] ]

From function checkMethod(const fvMatrix<Type>&, const GeometricField<Type, fvPatchField, volMesh>&)
in file /opt/openfoam240/src/finiteVolume/lnInclude/fvMatrix.C at line 1319.

FOAM aborting


It is a problem concerning the density. I have to multiply the centrifugal and Coriolis force by the density "rho" in order to get the same magnitude. Nevertheless, the files seem not to be taken into account this product:

Foam::tmp<Foam:imensionedField<Foam::vector, Foam::volMesh> >
Foam::SRF::SRFModel::Fcoriolis() const
{
return tmp<DimensionedField<vector, volMesh> >
(
new DimensionedField<vector, volMesh>
(
IOobject
(
"Fcoriolis",
mesh_.time().timeName(),
mesh_,
IOobject::NO_READ,
IOobject::NO_WRITE
),
rho*2.0*omega_ ^ Urel_
)
);
}


Foam::tmp<Foam:imensionedField<Foam::vector, Foam::volMesh> >
Foam::SRF::SRFModel::Fcentrifugal() const
{
return tmp<DimensionedField<vector, volMesh> >
(
new DimensionedField<vector, volMesh>
(
IOobject
(
"Fcentrifugal",
mesh_.time().timeName(),
mesh_,
IOobject::NO_READ,
IOobject::NO_WRITE
),
rho*omega_ ^ (omega_ ^ (mesh_.C() - origin_))
)
);
}
At following you can see what I have made ar the SRFModel.H and SRFModel.C

I already added some lines to SRFMOdel.C to consider this new value rho:
// * * * * * * * * * * * * * * * * Constructors * * * * * * * * * * * * * * //

Foam::SRF::SRFModel::SRFModel
(
const word& type,
const volVectorField& Urel
const volScalarField& rho
)

And the SRFModel.H:

Public:

const volScalarField& rho;


BUt still appears the issue concerning the units, it is not taken into account the "rho" value.

I appreciate your feedback