Base class for all standard finite volume or finite element problems. More...

#include <dumux/common/problem.hh>

Inheritance diagram for Dumux::Experimental::Problem< TypeTag >:

Description

template<class TypeTag>
class Dumux::Experimental::Problem< TypeTag >

Note: All quantities (regarding the units) are specified assuming a three-dimensional world. Problems discretized using 2D grids are assumed to be extruded by \(1 m\) and 1D grids are assumed to have a cross section of \(1m \times 1m\).

Classes
struct	Traits
	export traits of this problem More...

Public Member Functions
	Problem (std::shared_ptr< const GridDiscretization > gridDiscretization, const std::string &paramGroup="")
	Constructor.
const std::string &	name () const
	The problem name.
void	setName (const std::string &newName)
	Set the problem name.

Boundary conditions and sources defining the problem
BoundaryTypes	boundaryTypes (const ElementDiscretization &elemDisc, const typename ElementDiscretization::BoundaryFace &boundaryFace) const
	Specifies which kind of boundary condition should be used for which equation on a given boundary face.
BoundaryTypes	boundaryTypesAtPos (const GlobalPosition &globalPos) const
	Specifies which kind of boundary condition should be used for which equation on a given boundary segment.
template<class ResidualVector, class ElementVariables, class BoundaryTypes>
void	addFEBoundaryFluxIntegral (ResidualVector &residual, const ElementDiscretization &elemDisc, const ElementVariables &elemVars, const typename ElementDiscretization::BoundaryFace &boundaryFace, const BoundaryTypes &bcTypes) const
	Evaluates finite-element boundary flux integral contributions for a given boundary face.
template<class ResidualVector, class ElementVariables, class BoundaryTypes>
void	addFVBoundaryFluxIntegral (ResidualVector &residual, const ElementDiscretization &elemDisc, const ElementVariables &elemVars, const typename ElementDiscretization::SubControlVolumeFace &scvf, const BoundaryTypes &bcTypes) const
	Evaluates finite-volume boundary flux integral contributions for a given scvf.
template<class ElementVariables, class FaceIpData>
NumEqVector	boundaryFlux (const ElementDiscretization &elemDisc, const ElementVariables &elemVars, const FaceIpData &faceIpData) const
	Evaluate the boundary conditions for a neumann boundary segment.
NumEqVector	boundaryFluxAtPos (const GlobalPosition &globalPos) const
	Evaluate the boundary flux at a given position.
template<class ElementVariables, class IpData>
NumEqVector	source (const ElementDiscretization &elemDisc, const ElementVariables &elemVars, const IpData &ipData) const
	Evaluate the source term at a given interpolation point.
NumEqVector	sourceAtPos (const GlobalPosition &globalPos) const
	Evaluate the source term for all phases at a given position.
template<class MatrixBlock, class Variables, class IpData>
void	addSourceDerivatives (MatrixBlock &block, const Element &element, const ElementDiscretization &elemDisc, const Variables &volVars, const IpData &ipData) const
	Add source term derivative to the Jacobian.
EmptyPointSources	pointSources () const
	Return the point sources for this problem.
const GridDiscretization &	gridGeometry () const
	The grid discretization.
const GridDiscretization &	gridDiscretization () const
	The grid discretization.
const std::string &	paramGroup () const
	The parameter group in which to retrieve runtime parameters.
Implementation &	asImp_ ()
	Returns the implementation of the problem (i.e. static polymorphism).
const Implementation &	asImp_ () const
	Returns the implementation of the problem (i.e. static polymorphism).

Constructor & Destructor Documentation

◆ Problem()

template<class TypeTag>

Dumux::Experimental::Problem< TypeTag >::Problem	(	std::shared_ptr< const GridDiscretization >	gridDiscretization,
		const std::string &	paramGroup = "" )

inline

Parameters

gridDiscretization	The grid discretization
paramGroup	The parameter group in which to look for runtime parameters first (default is "")

Member Function Documentation

◆ addFEBoundaryFluxIntegral()

template<class TypeTag>

template<class ResidualVector, class ElementVariables, class BoundaryTypes>

void Dumux::Experimental::Problem< TypeTag >::addFEBoundaryFluxIntegral	(	ResidualVector &	residual,
		const ElementDiscretization &	elemDisc,
		const ElementVariables &	elemVars,
		const typename ElementDiscretization::BoundaryFace &	boundaryFace,
		const BoundaryTypes &	bcTypes ) const

inline

Parameters

residual	The residual vector to which the boundary flux contributions are added
elemDisc	The element discretization
elemVars	All variables for the element
boundaryFace	The boundary face for which the boundary flux integral is evaluated
bcTypes	The boundary types

◆ addFVBoundaryFluxIntegral()

template<class TypeTag>

template<class ResidualVector, class ElementVariables, class BoundaryTypes>

void Dumux::Experimental::Problem< TypeTag >::addFVBoundaryFluxIntegral	(	ResidualVector &	residual,
		const ElementDiscretization &	elemDisc,
		const ElementVariables &	elemVars,
		const typename ElementDiscretization::SubControlVolumeFace &	scvf,
		const BoundaryTypes &	bcTypes ) const

inline

Parameters

residual	The residual vector to which the boundary flux contributions are added
elemDisc	The element discretization
elemVars	All variables for the element
scvf	The scvf for which the boundary flux integral is evaluated
bcTypes	The boundary types

◆ addSourceDerivatives()

template<class TypeTag>

template<class MatrixBlock, class Variables, class IpData>

void Dumux::Experimental::Problem< TypeTag >::addSourceDerivatives	(	MatrixBlock &	block,
		const Element &	element,
		const ElementDiscretization &	elemDisc,
		const Variables &	volVars,
		const IpData &	ipData ) const

inline

Note: Only needed in case of analytic differentiation and solution dependent sources

◆ asImp_() [1/2]

template<class TypeTag>

Implementation & Dumux::Experimental::Problem< TypeTag >::asImp_ ( )

inlineprotected

◆ asImp_() [2/2]

template<class TypeTag>

const Implementation & Dumux::Experimental::Problem< TypeTag >::asImp_ ( ) const

inlineprotected

◆ boundaryFlux()

template<class TypeTag>

template<class ElementVariables, class FaceIpData>

NumEqVector Dumux::Experimental::Problem< TypeTag >::boundaryFlux	(	const ElementDiscretization &	elemDisc,
		const ElementVariables &	elemVars,
		const FaceIpData &	faceIpData ) const

inline

This is the method for the case where the Neumann condition is potentially solution dependent

Parameters

elemDisc	The element discretization
elemVars	All variables for the element
faceIpData	Face interpolation point data

Negative values mean influx. E.g. for the mass balance that would be the mass flux in \( [ kg / (m^2 \cdot s)] \).

◆ boundaryFluxAtPos()

template<class TypeTag>

NumEqVector Dumux::Experimental::Problem< TypeTag >::boundaryFluxAtPos ( const GlobalPosition & globalPos ) const

inline

Parameters

globalPos The position of the boundary face's integration point in global coordinates

Negative values mean influx. E.g. for the mass balance that would be the mass flux in \( [ kg / (m^2 \cdot s)] \).

As a default, i.e. if the user's problem does not overload any boundary flux method return no-flux boundary conditions at all boundaries

◆ boundaryTypes()

template<class TypeTag>

BoundaryTypes Dumux::Experimental::Problem< TypeTag >::boundaryTypes	(	const ElementDiscretization &	elemDisc,
		const typename ElementDiscretization::BoundaryFace &	boundaryFace ) const

inline

Parameters

elemDisc	The element discretization
boundaryFace	The boundary face

◆ boundaryTypesAtPos()

template<class TypeTag>

BoundaryTypes Dumux::Experimental::Problem< TypeTag >::boundaryTypesAtPos ( const GlobalPosition & globalPos ) const

inline

Parameters

globalPos The position in global coordinates

As a default, i.e. if the user's problem does not overload any boundaryTypes method we set no-flux everywhere.

◆ gridDiscretization()

template<class TypeTag>

const GridDiscretization & Dumux::Experimental::Problem< TypeTag >::gridDiscretization ( ) const

inline

◆ gridGeometry()

template<class TypeTag>

const GridDiscretization & Dumux::Experimental::Problem< TypeTag >::gridGeometry ( ) const

inline

◆ name()

template<class TypeTag>

const std::string & Dumux::Experimental::Problem< TypeTag >::name ( ) const

inline

This is used as a prefix for files generated by the simulation. It could be either overwritten by the problem files, or simply declared over the setName() function in the application file.

◆ paramGroup()

template<class TypeTag>

const std::string & Dumux::Experimental::Problem< TypeTag >::paramGroup ( ) const

inline

◆ pointSources()

template<class TypeTag>

EmptyPointSources Dumux::Experimental::Problem< TypeTag >::pointSources ( ) const

inline

By default, no point sources are present (returns EmptyPointSources for which empty() returns true). Derived classes can overwrite this function to return a PointSources object containing point sources with contexts and an evaluation function.

The returned object must provide the following interface:

bool empty() const Returns true if no point sources are registered; otherwise false.
For finite-volume schemes:
auto contexts(const ElementDiscretization& elemDisc,

const SubControlVolume& scv) const;

Returns a range of contexts related to the scv.
For finite-element schemes:
auto contexts(const ElementDiscretization& elemDisc,

const Element& element) const;

Returns a range of contexts related to the element.
auto eval(const ElementDiscretization& elemDisc, const ElementVariables& elemVars, const Context& context) const Evaluates the point source for the given context.

◆ setName()

template<class TypeTag>

void Dumux::Experimental::Problem< TypeTag >::setName ( const std::string & newName )

inline

Parameters

newName The problem's name

◆ source()

template<class TypeTag>

template<class ElementVariables, class IpData>

NumEqVector Dumux::Experimental::Problem< TypeTag >::source	(	const ElementDiscretization &	elemDisc,
		const ElementVariables &	elemVars,
		const IpData &	ipData ) const

inline

This is the method for the case where the source term is potentially solution dependent and requires some quantities that are specific to the fully-implicit method.

Parameters

elemDisc	The element discretization
elemVars	All variables for the element
ipData	Interpolation point data

For this method, the return parameter stores the conserved quantity rate generated or annihilate per volume unit. Positive values mean that the conserved quantity is created, negative ones mean that it vanishes.

◆ sourceAtPos()

template<class TypeTag>

NumEqVector Dumux::Experimental::Problem< TypeTag >::sourceAtPos ( const GlobalPosition & globalPos ) const

inline

Parameters

globalPos The position in global coordinates

For this method, the values parameter stores the conserved quantity rate generated or annihilate per volume unit. Positive values mean that the conserved quantity is created, negative ones mean that it vanishes. E.g. for the mass balance that would be a mass rate in \( [ kg / (m^3 \cdot s)] \).

As a default, i.e. if the user's problem does not overload any source method return 0.0 (no source terms)

The documentation for this class was generated from the following file:

common/problem.hh

Description

Classes

Public Member Functions

Boundary conditions and sources defining the problem

Constructor & Destructor Documentation

◆ Problem()

Member Function Documentation

◆ addFEBoundaryFluxIntegral()

◆ addFVBoundaryFluxIntegral()

◆ addSourceDerivatives()

◆ asImp_() [1/2]

◆ asImp_() [2/2]

◆ boundaryFlux()

◆ boundaryFluxAtPos()

◆ boundaryTypes()

◆ boundaryTypesAtPos()

◆ gridDiscretization()

◆ gridGeometry()

◆ name()

◆ paramGroup()

◆ pointSources()

◆ setName()

◆ source()

◆ sourceAtPos()