Namespace containing the basic application framework used to loop over cells and create objects passed to FuelCellShop::Equation objects. More...

Classes
class	ApplicationBase
	Base class for applications. More...

class	ApplicationData
	Here we handle general data of applications. More...

class	ApplicationWrapper
	This class implements either iterative or time-stepping wrapper of applications. More...

class	BlockMatrixApplication
	Application handling matrices and assembling linear systems of equations. More...

class	DoFApplication
	Base class for all linear applications, i.e., all applications requiring Triangulation and DoFHandler classes. More...

class	Event
	Objects of this kind are used to notify interior applications of changes provoked by an outer loop. More...

class	FEVectors
	The data type used in function calls of Application. More...

struct	MatrixBlock
	A wrapper around a MATRIX object, storing the coordinates in the global block matrix as well. More...

struct	BlockInfo
	A small structure collecting the different `BlockIndices` of `FEVector` vectors (for instance, solution) involved in the computations. More...

class	DoFInfo
	Very basic info class only containing information on geometry and degrees of freedom on a mesh entity. More...

class	IntegrationInfo
	This class is created for the objects handed to the mesh loops. More...

class	OptimizationBlockMatrixApplication
	Application handling matrices and assembling the linear system to solve the sensitivity equations. More...

class	DAESolver
	This class provides an interface to the Fortran 77 code COLDAE. More...

class	DAEWrapper
	This class is to be used as a wrapper for the functions needed in the DAESolver. More...

class	AdaptiveRefinement
	This class is initialized with an application that describes the linearization of the problem that we would like to solve and the nonlinear solver that drives the process (usually a Newton loop). More...

class	newtonBase
	Base class for all classes performing Newton's iteration. More...

class	NewtonBasic
	This class performs basic Newton iterations with a constant weight. More...

class	Newton3pp
	Application class performing Newton's iteration. More...

class	NewtonLineSearch
	Application class performing a Newton's iteration as described in newtonBase. More...

class	Picard
	Application class performing a Picard iteration as described in the PicardBase. More...

class	PicardBase
	Base class for all classes performing Picard iterations. More...

class	ThetaScheme

class	TransientBase
	This is a base class for transient solvers. More...

Typedefs
typedef BlockVector< double >	FEVector
	The vector class used by applications. More...

typedef std::vector < MatrixBlock< FullMatrix < double > > >	MatrixVector
	The matrix vector used in the mesh loops. More...

typedef std::vector< Vector < double > >	VectorVector
	The std::vector of dealii::Vectors used in the mesh loops. More...

typedef void(*	fsub_ptr )(double &, double[], double[], double[])

typedef void(*	dfsub_ptr )(double &, double[], double[], double[])

typedef void(*	gsub_ptr )(int &, double[], double &)

typedef void(*	dgsub_ptr )(int &, double[], double[])

typedef void(*	guess_ptr )(double &, double[], double[], double[])

Enumerations
enum	LinearSolver { LinearSolver::UMFPACK, LinearSolver::CG, LinearSolver::ILU_GMRES, LinearSolver::MUMPS, LinearSolver::BICGSTAB }
	Enumeration class for Linear solvers. More...

enum	NonLinearSolver { NonLinearSolver::NONE, NonLinearSolver::NEWTONBASIC, NonLinearSolver::NEWTON3PP, NonLinearSolver::NEWTONLINESEARCH, PICARD }
	Enumeration class for Non-linear solvers. More...

enum	RefinementSolver { RefinementSolver::ADAPTIVE }
	Enumeration class for refinement types. More...

enum	SolverName { noSolver = 0, LINEAR, NEWTON, PICARD, PICARD }

Functions
template<int dim, typename TYPE >
void	fill_data (const FEValuesBase< dim > &fe_values, const FEVector &fe_vector, const std::vector< unsigned int > &local_dof_indices, unsigned int first_index, unsigned int n_indices, std::vector< std::vector< TYPE > > &result)
	Helper functions computing the desired data in each quadrature point of a mesh entity by calling `FEValuesBase::get_function_values()`, `FEValuesBase::get_function_grads()`, and `FEValuesBase::get_function_hessians()`. More...

template<int dim>
void	fill_data (const FEValuesBase< dim > &fe_values, const FEVector &fe_vector, const std::vector< unsigned int > &local_dof_indices, unsigned int first_index, unsigned int n_indices, std::vector< std::vector< double > > &result)

template<int dim>
void	fill_data (const FEValuesBase< dim > &fe_values, const FEVector &fe_vector, const std::vector< unsigned int > &local_dof_indices, unsigned int first_index, unsigned int n_indices, std::vector< std::vector< Tensor< 1, dim > > > &result)

template<int dim>
void	fill_data (const FEValuesBase< dim > &fe_values, const FEVector &fe_vector, const std::vector< unsigned int > &local_dof_indices, unsigned int first_index, unsigned int n_indices, std::vector< std::vector< Tensor< 2, dim > > > &result)

void	DAE_dummy_guess (double &x, double z[], double y[], double df[])
	A dummy guess function to be provided to COLDAE when a user wishes to provide none. More...

void	for_to_c_matrix (int rows, int cols, double fmat, double *cmat)
	Converts a FORTRAN 2D array to a C/C++ 2D array. More...

void	c_to_for_matrix (int rows, int cols, double *cmat, double fmat)
	Converts a C/C++ 2D array to a Fortran 2D array. More...

Detailed Description

Namespace containing the basic application framework used to loop over cells and create objects passed to FuelCellShop::Equation objects.

This namespace encapsulates all the routines that are used to generate an application. ApplicationBase and ApplicationWrapper are declared here. All applications inherit from the objects developed here.

Note: This namespace should only be modified by advanced users as all applications depend on these routines

Typedef Documentation

typedef void(* FuelCell::ApplicationCore::dfsub_ptr)(double &, double[], double[], double[])

typedef void(* FuelCell::ApplicationCore::dgsub_ptr)(int &, double[], double[])

typedef BlockVector< double > FuelCell::ApplicationCore::FEVector

The vector class used by applications.

typedef void(* FuelCell::ApplicationCore::fsub_ptr)(double &, double[], double[], double[])

typedef void(* FuelCell::ApplicationCore::gsub_ptr)(int &, double[], double &)

typedef void(* FuelCell::ApplicationCore::guess_ptr)(double &, double[], double[], double[])

typedef std::vector< MatrixBlock< FullMatrix<double> > > FuelCell::ApplicationCore::MatrixVector

The matrix vector used in the mesh loops.

typedef std::vector< Vector<double> > FuelCell::ApplicationCore::VectorVector

The std::vector of dealii::Vectors used in the mesh loops.

Enumeration Type Documentation

enum FuelCell::ApplicationCore::LinearSolver

strong

Enumeration class for Linear solvers.

Enumerator
UMFPACK
CG
ILU_GMRES
MUMPS
BICGSTAB

enum FuelCell::ApplicationCore::NonLinearSolver

strong

Enumeration class for Non-linear solvers.

Enumerator
NONE
NEWTONBASIC
NEWTON3PP
NEWTONLINESEARCH
PICARD

enum FuelCell::ApplicationCore::RefinementSolver

strong

Enumeration class for refinement types.

Enumerator
ADAPTIVE

enum FuelCell::ApplicationCore::SolverName

Enumerator
noSolver
LINEAR
NEWTON
PICARD
PICARD

Function Documentation

void FuelCell::ApplicationCore::c_to_for_matrix	(	int	rows,
		int	cols,
		double **	cmat,
		double *	fmat
	)

Converts a C/C++ 2D array to a Fortran 2D array.

Parameters

rows	is the number of rows in the 2D array.
cols	is the number of columns in the 2D array.
cmat	is the 2D C/C++ matrix.
fmat	is the returned 2D FORTRAN matrix.

void FuelCell::ApplicationCore::DAE_dummy_guess	(	double &	x,
		double	z[],
		double	y[],
		double	df[]
	)

A dummy guess function to be provided to COLDAE when a user wishes to provide none.

Parameters

x	is a value between a <= x <= b.
z	is the initial guess for z(u(x)).
y	is the initial guess for y(x).
df	is array that contains the $m_{i}$ derivative of u(x)

template<int dim, typename TYPE >

void FuelCell::ApplicationCore::fill_data	(	const FEValuesBase< dim > &	fe_values,
		const FEVector &	fe_vector,
		const std::vector< unsigned int > &	local_dof_indices,
		unsigned int	first_index,
		unsigned int	n_indices,
		std::vector< std::vector< TYPE > > &	result
	)

inline

Helper functions computing the desired data in each quadrature point of a mesh entity by calling FEValuesBase::get_function_values(), FEValuesBase::get_function_grads(), and FEValuesBase::get_function_hessians().

Parameters

fe_values,:	The `FEValues` object.
fe_vector,:	The global finite element function in the form of a global nodal `FEVector`.
local_dof_indices,:	The local DoF indices associated with the current mesh entity.
first_index,:	The first index in `local_dof_indices` to be used.
n_indices,:	The number of indices in `local_dof_indices` to be used.
result,:	The result.

The access to ith component in qth quadrature point is result[i][q].

Author: Guido Kanschat

Referenced by FuelCell::ApplicationCore::IntegrationInfo< dim, FEVALUESBASE >::fill_local_data().

Here is the caller graph for this function:

template<int dim>

void FuelCell::ApplicationCore::fill_data	(	const FEValuesBase< dim > &	fe_values,
		const FEVector &	fe_vector,
		const std::vector< unsigned int > &	local_dof_indices,
		unsigned int	first_index,
		unsigned int	n_indices,
		std::vector< std::vector< double > > &	result
	)

inline

template<int dim>

void FuelCell::ApplicationCore::fill_data	(	const FEValuesBase< dim > &	fe_values,
		const FEVector &	fe_vector,
		const std::vector< unsigned int > &	local_dof_indices,
		unsigned int	first_index,
		unsigned int	n_indices,
		std::vector< std::vector< Tensor< 1, dim > > > &	result
	)

inline

template<int dim>

void FuelCell::ApplicationCore::fill_data	(	const FEValuesBase< dim > &	fe_values,
		const FEVector &	fe_vector,
		const std::vector< unsigned int > &	local_dof_indices,
		unsigned int	first_index,
		unsigned int	n_indices,
		std::vector< std::vector< Tensor< 2, dim > > > &	result
	)

inline

void FuelCell::ApplicationCore::for_to_c_matrix	(	int	rows,
		int	cols,
		double *	fmat,
		double **	cmat
	)

Converts a FORTRAN 2D array to a C/C++ 2D array.

Parameters

rows	is the number of rows in the 2D array.
cols	is the number of columns in the 2D array.
fmat	is the 2D FORTRAN matrix.
cmat	is the returned 2D C/C++ matrix.

Classes

Typedefs

Enumerations

Functions

Detailed Description

Typedef Documentation

Enumeration Type Documentation

Function Documentation