Hallo,

I have solved with icoFOAM a problem into a cylindrical geometry and I need to visualize the radial and angular velocity components.
Please, can somebody tell me how to extract from the cartezian velocity the polar components?

Thanks,
Nicoleta

[msrinath80]

See http://www.opencfd.co.uk/openfoam/do...5-250002.1.5.7

Hi,

thank you pUl| for the link, but my question is not answered.
With "Ucomponents" one can extract the x,y,z components of the velocity field, and I need the U_r component. Is there a function to extract the cylindrical or sperical components of the velocity?
Or my I construct such a function? I need for that the angles of the cell centers ...

Nicoleta

[osn]

Hi,

I wrote a simple dirty program to calc angular velocity.
It is from OpenFOAM-1.3/applications/utilities/postProcessing/velocityFi
eld/magU/magU.C.
So I show diff output of magU.C and my code below.

Thanks.

55a56,59
> tensor t1(0.0, -1.0, 0.0,
> 1.0, 0.0, 0.0,
> 0.0, 0.0, 0.0);
>
78,80c82,88
< Info<< " Calculating magU" << endl;
< volScalarField magU
< (
---
> Info<< " Calculating tanU" << endl;
>
> volVectorField p1 = t1 & mesh.C();
> volVectorField p2 = p1/mag(p1);
>
> volScalarField tanU
> (
83c91
< "magU",
---
> "tanU",
87,89c95,102
< ),
< mag(U)
< );
---
> ),
> U & p2
> );
>
> Info << "tan(U): max: " << max(tanU.internalField())
> << " min: " << min(tanU.internalField()) << endl;
>
> tanU.write();
91,94d103
< Info << "mag(U): max: " << max(magU.internalField())
< << " min: " << min(magU.internalField()) << endl;
<
< magU.write();

Dear Osamu

that's what I needed.
Thank you, it works.

Nicoleta

[hjasak]

A thousand time easier: OpenFOAM has got some coordinate system classes. You will need a cylindrical one - the class is called cylindricalCS (have a search through the source or Doxygen).

So:

1) make yourself a coordinate system

cylindricalCS(
const word & name,
const vector & origin,
const vector & axis,
const vector & direction
)

or any of the other constructors.

2) transform the (internal) field to the new cs using the toLocal function

3) if you want a radial component, use vector::y

Thus (something like this, I'm not compiling):

// This cs would be in the x-y plane, with "x" pointing up and cylindrical axis in the blobal x direction
cylindricalCS ccs
(
"ccs",
vector(0, 0, 0),
vector(1, 0, 0),
vector(0, 1, 0)
);

Info << "Ur: " << ccs.toLocal(U)().component(vector::Y) << endl;

(one line).

Currently, cs does not have a member to convert the whole geometric field, but I can add one for you if you wish.

Hrv

[hani]

I've been trying to learn how to use the cylindricalCS class, and I thought that I might share my experiences. No warranties to what I write though, and I guess that all of it can be made much simpler if you know what you are doing.

The example requires that you have a mesh and a velocity field U, and that you have defined inletPatchID. You also need to do #include "cylindricalCS.H". If your base code writes out the velocity field U you can then visualize the radial, tangential and axial coordinates as U at the inlet patch using for instance paraFoam. The example was developed for OF1.3.

//Test of cylindricalCS

//Find the Cartesian positions at the patch
const fvPatchVectorField& cartFaceCentres = mesh.Cf().boundaryField()[inletPatchID];

//Put the results in U at the patch so that you can visualize the result
fvPatchVectorField& CCSin = U.boundaryField()[inletPatchID];

//Define your cylindrical coordinate system
cylindricalCS ccs
(
"ccs",
vector(0, 0, 0), //center point of ccs
vector(0, 0, 1), //axis of ccs
vector(1, 1, 0) //base axis for cylindrical angle
);

//It doesn't seem to be possible to do the whole field at once, so
//loop through all the patch faces and set the radial, tangential and
//axial position
forAll(CCSin, facei)
{
CCSin[facei] = ccs.toLocal(cartFaceCentres[facei]);
//Make sure that you have only positive angles and that the
//angle is zero at the base axis:
CCSin[facei][1] = CCSin[facei][1] + neg(CCSin[facei][1])*360;
}

//You can also look at a single cylindrical component, here radial:
Info << "ccs.toLocal(cartFaceCentres[0]).component(vector::X)" << endl;
Info << ccs.toLocal(cartFaceCentres[0]).component(vector::X) << endl;

Håkan

ok I finally found that ccs.toLocal(U)().component(vector::Y) has for type Field<double> but the member write() does not work with it.

I guess that I have to convert Uz_cyl into a variable of volScalarField type. i.e (Foam::GeometricField<double,>::GeometricField(Foa m::IOobject, Foam::Field<double>&)) but I dunno how to fill the three first items...

Does anyone knows where is the source that manage the "write" member?

Cheers

pvc

[olesen]

Did you try something like this?

volScalarField Uz
(
IOobject
(
"Uz_cyl",
runTime.timeName(),
mesh,
IOobject::NO_READ
),
ccs.localVector(U)().component(vector::Y)
);

Uz.write();

BTW: in 1.4.1, toLocal() has been eliminated in favour of localPosition() and localVector()

If you are using an older version (with toLocal), be certain that the origin == (0,0,0) or you will have very funny results!

Hi mark,

Thanks for answering...
Actually, I am on OF-1.4 and .localVector() is not working.

I have seen in the User manual (p-32) that it was possible to convert Field<type> into volField<type> using fvc::~~ and mesh.~().

So I guess that converting ccs.toLocal(U)() which is a vectorField into volVectorField should be feasible, but I have not found yet how to make it...

If I am wrong please let me know, otherwise any idea?

Cheers

pvc

[florian_krause]

Hi guys!

the last post in this thread is more then 2 years old and related to the use of cylindrical CS in OF-1.4 .

I am working now with OF-1.6.x on a turbulent pipe flow (RANS & LES) with cylindrical cross section. I also want to visualize my fields in a cylindrical CS, lets say (r,teta,z).

My question - Are there any changes from OF-1.4 to OF-1.6.x in how to use the cylindrical CS or can I just try the way Hakan or Hrv explained ???

thanks!
Florian

[T.D.]

Hi me too, i'd like to know if i set in version 1.7 my coord system to cylindrical CS, then all my values will be automatically in (r, teta, and z) directions?

help please