[physicosm]

Dear solid4Foamers,


I am trying to implement an effective stress at the Boundary Conditions that will represent a thin layer of a material (e.g. gold) in contact with a bulk of a different material. Thus, the traction will have the form


Teff + T = sigma&n;

with Teff=div_{surface}(sigma)_surface and the index "surface" means that I need the surface divergence. Therefore, an idea is to use Finite Area Calculus (FAC). So, I would like to define an area tensor field and then call it at the boundary conditions. After this, I could make the calculations in order to obtain the desired sigma and get the surface divergence and also, the effective stress.



Thus, my first question is how can someone define the areaTensorField at the main solver (e.g. nonLinGeomTotalLagTotalDispSolid ). Do I need to include new Header files and if yes, which files?


Then, if this problem is solved, I believe that I can call a reference of the areaTensorField at the boundary conditions as


const faPatchField<tensor>& FP = patch().lookupPatchField<areaTensorField,tensor<(" F")

..... Calculations ....

const symmTensorField sigmaS = ...;

// Calculate surface divergence
areaVectorField areaDivSigma = fac::div(areaSigma);

vectorField Teff = areaDivSigma.internalField();



Therefore, I would like to ask if this scheme makes sense and of course, if there is any better way to do it, feedback is always welcome!

Thomas N.

[bigphil]

Hi Thomas,

Following on from our email conversation, you can include the following header file at the top of the file where you are writing your code:

#include "faCFD.H"

This will let you use the finite area classes.

Then you can define a finite area field as, for example:

// Create finite area mesh from patch with patch index ‘patchID;
faMesh aMesh(mesh, patchID);

// Create finite area sigma field (this is only defined on the patch)
areaSymmTensorField areaSigma
(
    IOobject
    (
        “areaSigma",
        runTime().timeName(),
        mesh(),
        IOobject::MUST_READ,
        IOobject::AUTO_WRITE
    ),
    aMesh
);

I don't quite understand why a thin layer of material produces the formula (Teff + T = sigma&n) but I guess you understand this!

Philip