Opm::TpsaLinearizer< TypeTag > Class Template Reference

Linearizes TPSA equations and generates system matrix and residual for linear solver. More...

#include <tpsalinearizer.hpp>

Public Member Functions
	TpsaLinearizer ()
	Constructor.
void	init (Simulator &simulator)
	Initialize the linearizer.
void	eraseMatrix ()
	Causes the Jacobian matrix to be recreated from scratch before the next iteration.
void	finalize ()
	Finalize creation of Jacobian matrix and make ready for linear solver.
template<class SubDomainType>
void	resetSystem_ (const SubDomainType &domain)
	Initializing and/or reset residual and Jacobian.
void	linearize ()
	Linearize the non-linear system.
void	linearizeDomain ()
	Linearize the non-linear system for the spatial domain.
template<class SubDomainType>
void	linearizeDomain (const SubDomainType &domain)
	Linearize the non-linear system for the spatial domain.
void	linearizeAuxiliaryEquations ()
	Linearize auxillary equation.
void	setResAndJacobi (VectorBlock &res, MatrixBlock &bMat, const ADVectorBlock &resid) const
	Extract local residuals and sub-block Jacobians from locally computed AD residual.
void	updateBoundaryConditionData ()
	Update boundary condition information.
const SparseMatrixAdapter &	jacobian () const
	Get Jacobian matrix.
SparseMatrixAdapter &	jacobian ()
	Get Jacobian matrix.
const GlobalEqVector &	residual () const
	Get residual vector.
GlobalEqVector &	residual ()
	Get residual vector.
const LinearizationType &	getLinearizationType () const
	Get linearization type.
std::map< unsigned, Constraints >	constraintsMap () const
	Get constraints map.
void	setLinearizationType (LinearizationType linearizationType)
	Set linearization type.

Detailed Description

template<class TypeTag>
class Opm::TpsaLinearizer< TypeTag >

Linearizes TPSA equations and generates system matrix and residual for linear solver.

Member Function Documentation

constraintsMap()

template<class TypeTag>

std::map< unsigned, Constraints > Opm::TpsaLinearizer< TypeTag >::constraintsMap ( ) const

inline

Get constraints map.

Returns

Constraints map

Note

No constraints implemented in TPSA

eraseMatrix()

template<class TypeTag>

void Opm::TpsaLinearizer< TypeTag >::eraseMatrix ( )

inline

Causes the Jacobian matrix to be recreated from scratch before the next iteration.

This method is usually called if the sparsity pattern has changed for some reason. (e.g. by modifications of the grid or changes of the auxiliary equations.)

getLinearizationType()

template<class TypeTag>

const LinearizationType & Opm::TpsaLinearizer< TypeTag >::getLinearizationType ( ) const

inline

Get linearization type.

Returns

Reference to the linearization object

The LinearizationType controls the scheme used and the focus time index. The default is fully implicit scheme, and focus index equal to 0, i.e. current time (end of step).

init()

template<class TypeTag>

void Opm::TpsaLinearizer< TypeTag >::init ( Simulator & simulator )

inline

Initialize the linearizer.

Parameters

simulator

Simulator object

At this point we can assume that all objects in the simulator have been allocated. We cannot assume that they are fully initialized, though.

jacobian() [1/2]

template<class TypeTag>

SparseMatrixAdapter & Opm::TpsaLinearizer< TypeTag >::jacobian ( )

inline

Get Jacobian matrix.

Returns

Reference to (sparse) Jacobian matrix object

jacobian() [2/2]

template<class TypeTag>

const SparseMatrixAdapter & Opm::TpsaLinearizer< TypeTag >::jacobian ( ) const

inline

Get Jacobian matrix.

Returns

Reference to (sparse) Jacobian matrix object

linearizeDomain() [1/2]

template<class TypeTag>

void Opm::TpsaLinearizer< TypeTag >::linearizeDomain ( )

inline

Linearize the non-linear system for the spatial domain.

The global Jacobian of the residual is assembled and the residual is evaluated for the current solution. The current state of affairs (esp. the previous and the current solutions) is represented by the model object.

linearizeDomain() [2/2]

template<class TypeTag>

template<class SubDomainType>

void Opm::TpsaLinearizer< TypeTag >::linearizeDomain ( const SubDomainType & domain )

inline

Linearize the non-linear system for the spatial domain.

Parameters

domain

(Sub-)domain to linearize

The global Jacobian of the residual is assembled and the residual is evaluated for the current solution. The current state of affairs (esp. the previous and the current solutions) is represented by the model object.

resetSystem_()

template<class TypeTag>

template<class SubDomainType>

void Opm::TpsaLinearizer< TypeTag >::resetSystem_ ( const SubDomainType & domain )

inline

Initializing and/or reset residual and Jacobian.

Parameters

domain

(Sub-)domain to reset

residual() [1/2]

template<class TypeTag>

GlobalEqVector & Opm::TpsaLinearizer< TypeTag >::residual ( )

inline

Get residual vector.

Returns

Reference to the residual

residual() [2/2]

template<class TypeTag>

const GlobalEqVector & Opm::TpsaLinearizer< TypeTag >::residual ( ) const

inline

Get residual vector.

Returns

Reference to the residual

setLinearizationType()

template<class TypeTag>

void Opm::TpsaLinearizer< TypeTag >::setLinearizationType ( LinearizationType linearizationType )

inline

Set linearization type.

Parameters

linearizationType

Linearization object

setResAndJacobi()

template<class TypeTag>

void Opm::TpsaLinearizer< TypeTag >::setResAndJacobi ( VectorBlock & res, MatrixBlock & bMat, const ADVectorBlock & resid ) const

inline

Extract local residuals and sub-block Jacobians from locally computed AD residual.

Parameters

res	Residual to set
bMat	Block matrix from Jacobian to set
resid	Computed AD residual

---

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

• opm/models/discretization/common/tpsalinearizer.hpp