Solving PDEs with OpenFOAM

• The PDEs we wish to solve involve derivatives of tensor fields with

respect to time and space
• The PDEs must be discretized in time and space before we solve
them

• We will start by having a look at algebra of tensors in OpenFOAM

at a single point
• We will then have a look at how to generate tensor fields from
tensors
• Finally we will see how to discretize PDEs and how to set boundary
conditions using high-level coding in OpenFOAM

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 112

 Basic tensor classes in OpenFOAM
• Pre-defined classes for tensors of rank 0-3, but may be extended indefinitely

Rank Common name Basic name Access function

0 Scalar scalar
1 Vector vector x(), y(), z()
2 Tensor tensor xx(), xy(), xz(), ...

Example:
 
11 12 13
A tensor T =  21 22 23  is defined line-by-line:
31 32 33
tensor T( 11, 12, 13, 21, 22, 23, 31, 32, 33);

Info << "Txz = " << T.xz() << endl;

Outputs to the screen:
Txz = 13

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 113

 Algebraic tensor operations in OpenFOAM
• Tensor operations operate on the entire tensor entity instead of a
series of operations on its components
• The OpenFOAM syntax closely mimics the syntax used in written mathematics,
using descriptive functions or symbolic operators

Examples:

Operation Comment Mathematical Description

description in OpenFOAM
Addition a+b a+b
Outer product Rank a, b ≥ 1 ab a*b
Inner product Rank a, b ≥ 1 a · b a&b
Cross product Rank a, b = 1 a × b aˆb
Operations exclusive to tensors of rank 2
Transpose TT T.T()
Determinant detT det(T)
Operations exclusive to scalars
Positive (boolean) s≥0 pos(s)
Hyperbolic arc sine asinh s asinh(s)

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 114

 Examples of the use of some tensor classes
• In $FOAM_APP/test we can find examples of the use of some classes.
• Tensor class examples:
run
cp -r $FOAM_APP/test .
cd test/tensor
wmake
tensorTest >& log
• Have a look inside tensorTest.C to see the high-level code.
• See also vector, symmTensorField, sphericalTensorField
and many other examples.

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 115

 Dimensional units in OpenFOAM
• OpenFOAM checks the dimensional consistency
Declaration of a tensor with dimensions:
dimensionedTensor sigma
(
"sigma",
dimensionSet( 1, -1, -2, 0, 0, 0, 0),
tensor( 1e6, 0, 0, 0, 1e6, 0, 0, 0, 1e6)
);
The values of dimensionSet correspond to the powers of each SI unit:
No. Property Unit Symbol
1 Mass kilogram kg
2 Length metre m
3 Time second s
4 Temperature Kelvin K
5 Quantity moles mol
6 Current ampere A
7 Luminous intensity candela cd
 2

sigma then has the dimension kg/ms

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 116

 Dimensional units in OpenFOAM
• Add the following to tensorTest.C:
Before main():
#include "dimensionedTensor.H"
Before return(0):

dimensionedTensor sigma
(
"sigma",
dimensionSet( 1, -1, -2, 0, 0, 0, 0),
tensor( 1e6, 0, 0, 0, 1e6, 0, 0, 0, 1e6)
);
Info<< "Sigma: " << sigma << endl;

• Compile, run again, and you will get:

Sigma: sigma [1 -1 -2 0 0 0 0] (1e+06 0 0 0 1e+06 0 0 0 1e+06)

You see that the object sigma that belongs to the dimensionedTensor class
contains both the name, the dimensions and values.

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 117

 Dimensional units in OpenFOAM
• Try some member functions of the dimensionedTensor class:

Info<< "Sigma name: " << sigma.name() << endl;

Info<< "Sigma dimensions: " << sigma.dimensions() << endl;
Info<< "Sigma value: " << sigma.value() << endl;

• You now also get:

Sigma name: sigma

Sigma dimensions: [1 -1 -2 0 0 0 0]
Sigma value: (1e+06 0 0 0 1e+06 0 0 0 1e+06)

• Extract one of the values:

Info<< "Sigma yy value: " << sigma.value().yy() << endl;
Note here that the value() member function first converts the expression to a
tensor, which has a yy() member function. The dimensionedTensor class
does not have a yy() member function, so it is not possible to do sigma.yy().

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 118

 Construction of a tensor field in OpenFOAM
• A tensor field is a list of tensors
• The use of typedef in OpenFOAM yields readable type definitions:
scalarField, vectorField, tensorField, symmTensorField, ...
• Algebraic operations can be performed between different fields,
and between a field and a single tensor, e.g. Field U, scalar 2.0:
U = 2.0 * U;
• Add the following to tensorTest:
Before main():
#include "tensorField.H"
Before return(0):

tensorField tf1(2, tensor::one);

Info<< "tf1: " << tf1 << endl;
tf1[0] = tensor(1, 2, 3, 4, 5, 6, 7, 8, 9);
Info<< "tf1: " << tf1 << endl;
Info<< "2.0*tf1: " << 2.0*tf1 << endl;

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 119

 Discretization of a tensor field in OpenFOAM
• FVM (Finite Volume Method)
• No limitations on the number of faces bounding each cell
• No restriction on the alignment of each face
• The mesh class polyMesh can be used to construct a polyhedral
mesh using the minimum information required
• The fvMesh class extends the polyMesh class to include additional
data needed for the FV discretization (see test/mesh)
• The geometricField class relates a tensor field to an fvMesh (can
also be typedef volField, surfaceField, pointField)
• A geometricField inherits all the tensor algebra of its correspond-
ing field, has dimension checking, and can be subjected to specific
discretization procedures

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 120

 Examine an fvMesh
• Let us examine an fvMesh:
run
rm -rf cavity
cp -r $FOAM_TUTORIALS/incompressible/icoFoam/cavity .
cd cavity
sed -i s/"20 20 1"/"2 2 1"/g constant/polyMesh/blockMeshDict
blockMesh
• Run the test/mesh/meshTest (first compile it!)
• C() gives the center of all cells and boundary faces.
V() gives the volume of all the cells.
Cf() gives the center of all the faces.
• Try also adding in meshTest.C, before return(0):
Info<< mesh.C().internalField()[1][1] << endl;
Info<< mesh.boundaryMesh()[0].name() << endl;

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 121

 Examine a volScalarField
• Read a volScalarField that corresponds to the mesh. Add in meshTest.C,
before return(0):

volScalarField p
(
IOobject
(
"p",
runTime.timeName(),
mesh,
IOobject::MUST_READ,
IOobject::AUTO_WRITE
),
mesh
);
Info<< p << endl;
Info<< p.boundaryField()[0] << endl;

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 122

 Equation discretization in OpenFOAM
• Converts the PDEs into a set of linear algebraic equations, Ax=b, where x and b are volFields
(geometricFields). A is an fvMatrix, which is created by a discretization of a geometricField and
inherits the algebra of its corresponding field, and it supports many of the standard algebraic
matrix operations
• The fvm (Finite Volume Method) and fvc (Finite Volume Calculus) classes contain static func-
tions for the differential operators, and discretize any geometricField. fvm returns an fvMatrix,
and fvc returns a geometricField.
Examples:
Term description Mathematical expression fvm::/fvc:: functions
Laplacian ∇ · Γ∇φ laplacian(Gamma,phi)
Time derivative ∂φ/∂t ddt(phi)
∂ρφ/∂t ddt(rho, phi)
Convection ∇ · (ψ) div(psi, scheme)
∇ · (ψφ) div(psi, phi, word)
div(psi, phi)
Source ρφ Sp(rho, phi)
SuSp(rho, phi)
φ: vol<type>Field, ρ: scalar, volScalarField, ψ: surfaceScalarField

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 123

 Example
The equation
~
∂ρU ~ − ∇ · µ∇U
+ ∇ · φU ~ = −∇p
∂t

has the OpenFOAM representation

solve
(
fvm::ddt(rho, U)
+ fvm::div(phi, U)
- fvm::laplacian(mu, U)
==
- fvc::grad(p)
)

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 124

 Example: laplacianFoam, the source code
Solves ∂T /∂t − ∇ · k∇T = 0
#include "fvCFD.H" // Include the class definitions
int main(int argc, char *argv[])
{
# include "setRootCase.H" // Set the correct path
# include "createTime.H" // Create the time
# include "createMesh.H" // Create the mesh
# include "createFields.H" // Temperature field T and diffusivity DT
while (runTime.loop()) // Time loop
{
# include "readSIMPLEControls.H" // Read solution controls
for (int nonOrth=0; nonOrth$<$=nNonOrthCorr; nonOrth++)
{
solve( fvm::ddt(T) - fvm::laplacian(DT, T) ); // Solve eq.
}
# include "write.H" // Write out results at specified time instances}
}
return 0; // End with ’ok’ signal
}

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 125

Example: laplacianFoam, discretization and boundary conditions
Find this in $FOAM_TUTORIALS/basic/laplacianFoam/flange
Discretization:
dictionary fvSchemes, read from file:
ddtSchemes
{
default Euler;
}

laplacianSchemes
{
default none;
laplacian(DT,T) Gauss linear corrected;
}
Boundary conditions:
Part of class volScalarField object T, read from file:
boundaryField{
patch1{ type zeroGradient;}
patch2{ type fixedValue; value uniform 273;}}

Håkan Nilsson, Chalmers / Applied Mechanics / Fluid Dynamics 126