[ClaudioC]

Hello Foamers!
I'm currently trying to implement a few lines of code to be inserted in fvOptions to generate a non uniform heat source. The geometry i'm working with is a tube along z direction that is running the chtMultiRegionFoam solver on 4 parallel processors.
I'm currently stuck due to the implementation of the following lines of code:

energySource
{
type scalarCodedSource;
active true;
name generation;
selectionMode cellSet;
cellSet tube;
fields (e);

scalarCodedSourceCoeffs
{
selectionMode cellSet;
cellSet tube;
fields (e);
name generation;


codeInclude
#{
#include <IFstream.H>
#include <stdio.h>
#include <iostream>
#};

codeCorrect
#{
Pout<< "**codeCorrect**" << endl;
#};

codeAddSup
#{
const Time& time = mesh().time();
const scalarField& V = mesh_.V();

scalarField& heSource = eqn.source();
// Retrieve the x component of the cell centres
const scalarField& cellx = mesh().C().component(0);
// Retrieve the y component of the cell centres
const scalarField& celly = mesh().C().component(1);

const scalarField& cell10 = mesh().C().component(10);
const volVectorField& cell = mesh().C();

cout<<"The Value of 'cellx' is "<<cellx[10000];
cout<<"The Value of 'cell10' is "<<cell10 [10000];
cout<<"The Value of 'celly' is "<<celly [10000];
cout<<" cell 0 is "<< cell[10000][0];
cout<<" cell 1 is "<< cell[10000][1];
cout<<" cell 2 is "<< cell[10000][2];
cout<<" cell 3 is "<< cell[10000][3];
cout<<" cell 10 is "<< cell[10000][10];

}#
codeContrain
#{
Pout<< "**codeConstrain**" << endl;
#};
codeSetValue
#{
Pout<< "**codeSetValue**" << endl;
#};

I would expect cellx and celly to have different values, but any selected point posses the same coordinates in x and y. In addition, I would expect mesh().C() to possess only 3 components (the x,y and z coordinates of the cell centers) but I'm obtaining the same results for any component selected (in the results below the component(10) is shown)

The Value of 'cellx' is -0.0259017The Value of 'cell10' is -0.0259017The Value of 'celly' is -0.0259017 cell 0 is 0.0501623 cell 1 is -0.0258483 cell 2 is 0.807492 cell 3 is 0.0528021 cell 10 is -0.0259017[3] **codeSetValue**
The Value of 'cellx' is 0.038028The Value of 'cell10' is 0.038028The Value of 'celly' is 0.038028 cell 0 is 0.0410684 cell 1 is 0.0380552 cell 2 is 0.054488 cell 3 is 0.0380504 cell 10 is 0.038028[0] **codeSetValue**
The Value of 'cellx' is 0.00781293The Value of 'cell10' is 0.00781293The Value of 'celly' is 0.00781293 cell 0 is 0.0583116 cell 1 is 0.00139837 cell 2 is 0.306497 cell 3 is 0.0554213 cell 10 is 0.00781293[1] **codeSetValue**
The Value of 'cellx' is 0.00773194The Value of 'cell10' is 0.00773194The Value of 'celly' is 0.00773194 cell 0 is -0.0558103 cell 1 is 0.00779338 cell 2 is 0.558712 cell 3 is -0.0582097 cell 10 is 0.00773194[2] **codeSetValue**

Any ideas on what I could do to solve this?

[Tobermory]

Instead of the components method, just use the index, eg:

    func1
    {
        libs            ( "libutilityFunctionObjects.so" );
        type            coded;
        name            test;
        writeControl    timeStep;
        writeInterval   1;
        codeWrite       #{
            float  value(-0.4);
            int intValue(value);
            bool boolValue(value);
            bool boolIntValue(intValue);
            Info << "Cell[100] C  = " << mesh().C()[100]
                 << "  Cell[100] Cx = " << mesh().C()[100][0]
                 << "  Cell[100] Cy = " << mesh().C()[100][1] << endl;
        #};
    }

works nicely:

Cell[100] C = (0.125 -4.625 -4.875) Cell[100] Cx = 0.125 Cell[100] Cy = -4.625

[Tobermory]

... as does:

    func1
    {
        libs            ( "libutilityFunctionObjects.so" );
        type            coded;
        name            test;
        writeControl    timeStep;
        writeInterval   1;
        codeWrite       #{
            float  value(-0.4);
            int intValue(value);
            bool boolValue(value);
            bool boolIntValue(intValue);
            Info << "Cell[100] C  = " << mesh().C()[100]
                 << ",  Cell[100] Cx = " << mesh().C()[100][0]
                 << ",  Cell[100] Cy = " << mesh().C()[100][1] << endl;
            Info << "Cell[100] C  = " << mesh().C()[100]
                 << ",  Cell[100] Cx = " << mesh().C()[100].component(0)
                 << ",  Cell[100] Cy = " << mesh().C()[100].component(1) << endl;
        #};
    }

Cell[100] C = (0.125 -4.625 -4.875), Cell[100] Cx = 0.125, Cell[100] Cy = -4.625
Cell[100] C = (0.125 -4.625 -4.875), Cell[100] Cx = 0.125, Cell[100] Cy = -4.625

i.e. check your syntax. Good luck!

[olesen]

Originally Posted by ClaudioC

Hello Foamers!
I'm currently trying to implement a few lines of code to be inserted in fvOptions to generate a non uniform heat source. The geometry i'm working with is a tube along z direction that is running the chtMultiRegionFoam solver on 4 parallel processors.
I'm currently stuck due to the implementation of the following lines of code:






I would expect cellx and celly to have different values, but any selected point posses the same coordinates in x and y. In addition, I would expect mesh().C() to possess only 3 components (the x,y and z coordinates of the cell centers) but I'm obtaining the same results for any component selected (in the results below the component(10) is shown)






Any ideas on what I could do to solve this?

Your syntax is so completely wrong in so many places. Look at what Tobermory suggested. With your [] syntax for obtaining component 10?? of the vector. There are only three components, so you are slicing into some other memory location.

[olesen]

Originally Posted by Tobermory

... as does:

    func1
    {
        libs            ( "libutilityFunctionObjects.so" );
        type            coded;
        name            test;
        writeControl    timeStep;
        writeInterval   1;
        codeWrite       #{
            float  value(-0.4);
            int intValue(value);
            bool boolValue(value);
            bool boolIntValue(intValue);
            Info << "Cell[100] C  = " << mesh().C()[100]
                 << ",  Cell[100] Cx = " << mesh().C()[100][0]
                 << ",  Cell[100] Cy = " << mesh().C()[100][1] << endl;
            Info << "Cell[100] C  = " << mesh().C()[100]
                 << ",  Cell[100] Cx = " << mesh().C()[100].component(0)
                 << ",  Cell[100] Cy = " << mesh().C()[100].component(1) << endl;
        #};
    }

i.e. check your syntax. Good luck!

Good solution. For additional clarity I would always use the vector accessors x(), y(), z() - which also happen to be shorter to type.
People should also be aware that the components method on the volume field will return a tmp scalarField. So binding to a const-reference is dubious. And unless you need these as an entire field, also wasteful of memory and slower than directly accessing the components of individual elements.
Slightly reworked:

    func1
    {
        name test;
        type coded;
        libs ( "libutilityFunctionObjects.so" );

        codeWrite
       #{
            const auto& cc = mesh().C();

            Info << "Cell[100] C  = " << cc[100]
                 << ",  Cell[100] Cx = " << cc[100].x()
                 << ",  Cell[100] Cy = " << cc[100].y() << nl;
        #};
    }

[ClaudioC]

Thank you very much Tobermory. I was actually thinking to employ the components as entire fields, as Olesen suggested, but i managed to rewrite my code so that i can employ a single

mesh().C()[i].component(0)

in a for cycle and obtain the result i needed.

Thank you all guys