Class PoissonFACPreconditionerStrategy is an abstract FACPreconditionerStrategy implementing many of the operations required by smoothers for the Poisson equation and related problems. More...

#include </home/runner/work/IBAMR/IBAMR/ibtk/include/ibtk/PoissonFACPreconditionerStrategy.h>

Inheritance diagram for IBTK::PoissonFACPreconditionerStrategy:

[legend]

Public Member Functions
	PoissonFACPreconditionerStrategy (std::string object_name, SAMRAI::tbox::Pointer< SAMRAI::hier::Variable< NDIM > > scratch_var, int ghost_cell_width, SAMRAI::tbox::Pointer< SAMRAI::tbox::Database > input_db, const std::string &default_options_prefix)
	Constructor.

	~PoissonFACPreconditionerStrategy ()
	Destructor.

virtual void	setPoissonSpecifications (const SAMRAI::solv::PoissonSpecifications &poisson_spec)
	Set the SAMRAI::solv::PoissonSpecifications object used to specify the coefficients for the scalar-valued or vector-valued Laplace operator.

virtual void	setPhysicalBcCoef (SAMRAI::solv::RobinBcCoefStrategy< NDIM > *bc_coef)
	Set the SAMRAI::solv::RobinBcCoefStrategy object used to specify physical boundary conditions. More...

virtual void	setPhysicalBcCoefs (const std::vector< SAMRAI::solv::RobinBcCoefStrategy< NDIM > * > &bc_coefs)
	Set the SAMRAI::solv::RobinBcCoefStrategy objects used to specify physical boundary conditions. More...

Functions for configuring the solver.
void	setResetLevels (int coarsest_ln, int finest_ln)
	Specify the levels that need to be reset the next time the operator is re-initialized. More...

virtual void	setSmootherType (const std::string &smoother_type)=0
	Specify the smoother type.

virtual void	setCoarseSolverType (const std::string &coarse_solver_type)=0
	Specify the coarse level solver.

void	setCoarseSolverMaxIterations (int coarse_solver_max_iterations)
	Set the maximum number of iterations for the coarse level solve. More...

void	setCoarseSolverAbsoluteTolerance (double coarse_solver_abs_residual_tol)
	Set the absolute residual tolerance for convergence for coarse level solve. More...

void	setCoarseSolverRelativeTolerance (double coarse_solver_rel_residual_tol)
	Set the relative residual tolerance for convergence for coarse level solve. More...

void	setProlongationMethod (const std::string &prolongation_method)
	Set the name of the prolongation method.

void	setRestrictionMethod (const std::string &restriction_method)
	Set the name of the restriction method.

Public Member Functions inherited from IBTK::FACPreconditionerStrategy
	FACPreconditionerStrategy (std::string object_name, bool homogeneous_bc=false)
	Constructor.

virtual	~FACPreconditionerStrategy ()=default
	Empty virtual desctructor.

const std::string &	getName () const
	Return the object name.

virtual bool	getIsInitialized () const
	Return whether the operator is initialized.

virtual void	setFACPreconditioner (SAMRAI::tbox::ConstPointer< FACPreconditioner > preconditioner)
	Method to allow the FACPreconditioner object to register itself with the concrete FACPreconditionerStrategy.

virtual void	setHomogeneousBc (bool homogeneous_bc)
	Set whether the solver should use homogeneous boundary conditions.

virtual bool	getHomogeneousBc () const
	Return whether the solver is using homogeneous boundary conditions.

virtual void	setSolutionTime (double solution_time)
	Set the time at which the solution is to be evaluated.

virtual double	getSolutionTime () const
	Get the time at which the solution is being evaluated.

virtual void	setTimeInterval (double current_time, double new_time)
	Set the current time interval.

virtual std::pair< double, double >	getTimeInterval () const
	Get the current time interval.

virtual double	getDt () const
	Get the current time step size.

virtual void	smoothError (SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &error, const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &residual, int level_num, int num_sweeps, bool performing_pre_sweeps, bool performing_post_sweeps)=0
	Smooth the error by the specified number of sweeps on the specified level of the patch hierarchy.

virtual bool	solveCoarsestLevel (SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &error, const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &residual, int coarsest_level_num)=0
	Solve the system of equations on the coarsest level of the patch hierarchy. More...

virtual void	computeResidual (SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &residual, const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &solution, const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &rhs, int coarsest_level_num, int finest_level_num)=0
	Compute the composite-grid residual on the specified range of levels of the patch hierarchy.

virtual void	allocateScratchData ()
	Allocate scratch data.

virtual void	deallocateScratchData ()
	Deallocate scratch data.

virtual void	printClassData (std::ostream &stream)
	Print class data to stream.

Protected Attributes
Hierarchy-dependent objects.
SAMRAI::tbox::Pointer< SAMRAI::solv::SAMRAIVectorReal< NDIM, double > >	d_solution

SAMRAI::tbox::Pointer< SAMRAI::solv::SAMRAIVectorReal< NDIM, double > >	d_rhs

SAMRAI::tbox::Pointer< SAMRAI::hier::PatchHierarchy< NDIM > >	d_hierarchy

int	d_coarsest_ln = IBTK::invalid_level_number

int	d_finest_ln = IBTK::invalid_level_number

std::vector< SAMRAI::tbox::Pointer< SAMRAI::math::HierarchyDataOpsReal< NDIM, double > > >	d_level_data_ops

std::vector< SAMRAI::tbox::Pointer< IBTK::HierarchyGhostCellInterpolation > >	d_level_bdry_fill_ops

std::vector< SAMRAI::tbox::Pointer< IBTK::HierarchyMathOps > >	d_level_math_ops

bool	d_in_initialize_operator_state = false

int	d_coarsest_reset_ln = IBTK::invalid_level_number

int	d_finest_reset_ln = IBTK::invalid_level_number

Solver configuration variables.
std::string	d_smoother_type = "DEFAULT"

std::string	d_prolongation_method = "DEFAULT"

std::string	d_restriction_method = "DEFAULT"

std::string	d_coarse_solver_type = "DEFAULT"

std::string	d_coarse_solver_default_options_prefix

double	d_coarse_solver_rel_residual_tol = 1.0e-5

double	d_coarse_solver_abs_residual_tol = 1.0e-50

int	d_coarse_solver_max_iterations = 10

Internal context and scratch data.
SAMRAI::tbox::Pointer< SAMRAI::hier::VariableContext >	d_context

int	d_scratch_idx = IBTK::invalid_index

Protected Attributes inherited from IBTK::FACPreconditionerStrategy
SAMRAI::tbox::ConstPointer< IBTK::FACPreconditioner >	d_preconditioner

const std::string	d_object_name

bool	d_is_initialized = false

bool	d_homogeneous_bc

double	d_solution_time = std::numeric_limits<double>::quiet_NaN()

double	d_current_time = std::numeric_limits<double>::quiet_NaN()

double	d_new_time = std::numeric_limits<double>::quiet_NaN()

Methods for executing, caching, and resetting communication
schedules.
SAMRAI::solv::PoissonSpecifications	d_poisson_spec

std::unique_ptr< SAMRAI::solv::RobinBcCoefStrategy< NDIM > >	d_default_bc_coef

std::vector< SAMRAI::solv::RobinBcCoefStrategy< NDIM > * >	d_bc_coefs

SAMRAI::hier::IntVector< NDIM >	d_gcw

void	xeqScheduleProlongation (int dst_idx, int src_idx, int dst_ln)
	Execute a refinement schedule for prolonging data.

void	xeqScheduleRestriction (int dst_idx, int src_idx, int dst_ln)
	Execute schedule for restricting solution or residual to the specified level.

void	xeqScheduleGhostFillNoCoarse (int dst_idx, int dst_ln)
	Execute schedule for filling ghosts on the specified level.

void	xeqScheduleDataSynch (int dst_idx, int dst_ln)
	Execute schedule for synchronizing data on the specified level.

Various refine and coarsen objects.
SAMRAI::tbox::Pointer< RobinPhysBdryPatchStrategy >	d_bc_op

SAMRAI::tbox::Pointer< CoarseFineBoundaryRefinePatchStrategy >	d_cf_bdry_op

SAMRAI::tbox::Pointer< SAMRAI::xfer::VariableFillPattern< NDIM > >	d_op_stencil_fill_pattern

SAMRAI::tbox::Pointer< SAMRAI::xfer::VariableFillPattern< NDIM > >	d_synch_fill_pattern

Partial implementation of FACPreconditionerStrategy interface.
void	setToZero (SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &vec, int level_num) override
	Zero the supplied vector.

void	restrictResidual (const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &src, SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &dst, int dst_ln) override
	Restrict the residual quantity to the specified level from the next finer level. More...

void	prolongError (const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &src, SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &dst, int dst_ln) override
	Prolong the error quantity to the specified level from the next coarser level. More...

void	prolongErrorAndCorrect (const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &src, SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &dst, int dst_ln) override
	Prolong the error quantity to the specified level from the next coarser level and apply the correction to the fine-level error. More...

void	initializeOperatorState (const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &solution, const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &rhs) override
	Compute hierarchy-dependent data. More...

void	deallocateOperatorState () override
	Remove all hierarchy-dependent data. More...

virtual void	initializeOperatorStateSpecialized (const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &solution, const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &rhs, int coarsest_reset_ln, int finest_reset_ln)=0
	Compute implementation-specific hierarchy-dependent data.

virtual void	deallocateOperatorStateSpecialized (int coarsest_reset_ln, int finest_reset_ln)=0
	Remove implementation-specific hierarchy-dependent data.

Additional Inherited Members
Protected Member Functions inherited from IBTK::FACPreconditionerStrategy
virtual SAMRAI::tbox::Pointer< SAMRAI::solv::SAMRAIVectorReal< NDIM, double > >	getLevelSAMRAIVectorReal (const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &vec, int level_num) const
	Return a SAMRAIVectorReal object that corresponds to the given object but restricted to a single level of the patch hierarchy.

Detailed Description

Class PoissonFACPreconditionerStrategy is an abstract FACPreconditionerStrategy implementing many of the operations required by smoothers for the Poisson equation and related problems.

The parameter ghost_cell_width is the necessary ghost cell width for all operations required by this class and derived classes. Such operations includes restricting, prolongation, computing residuals, and smoothing error. The patch indices in the RHS vector must have ghost cell widths that are equal to this parameter.

Sample parameters for initialization from database (and their default values):

smoother_type = "DEFAULT"                    // see setSmootherType()
prolongation_method = "DEFAULT"              // see setProlongationMethod()
restriction_method = "DEFAULT"               // see setRestrictionMethod()
coarse_solver_type = "DEFAULT"               // see setCoarseSolverType()
coarse_solver_rel_residual_tol = 1.0e-5      // see setCoarseSolverRelativeTolerance()
coarse_solver_abs_residual_tol = 1.0e-50     // see setCoarseSolverAbsoluteTolerance()
coarse_solver_max_iterations = 10            // see setCoarseSolverMaxIterations()

Member Function Documentation

◆ deallocateOperatorState()

void IBTK::PoissonFACPreconditionerStrategy::deallocateOperatorState ( )

overridevirtual

Remove all hierarchy-dependent data.

Remove all hierarchy-dependent data set by initializeOperatorState().

See also: initializeOperatorState

Reimplemented from IBTK::FACPreconditionerStrategy.

◆ initializeOperatorState()

void IBTK::PoissonFACPreconditionerStrategy::initializeOperatorState	(	const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &	solution,
		const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &	rhs
	)

overridevirtual

Compute hierarchy-dependent data.

Note that although the vector arguments given to other methods in this class may not necessarily be the same as those given to this method, there will be similarities, including:

hierarchy configuration (hierarchy pointer and level range)
number, type and alignment of vector component data
ghost cell width of data in the solution (or solution-like) vector

An unrecoverable error will occur if the rhs vector does not have consistent ghost cell width as that provided in the constructor.

Parameters

solution	solution vector u
rhs	right hand side vector f

Reimplemented from IBTK::FACPreconditionerStrategy.

◆ prolongError()

void IBTK::PoissonFACPreconditionerStrategy::prolongError	(	const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &	src,
		SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &	dst,
		int	dst_ln
	)

overridevirtual

Prolong the error quantity to the specified level from the next coarser level.

Parameters

src	source error vector
dst	destination error vector
dst_ln	destination level number of data transfer

Implements IBTK::FACPreconditionerStrategy.

◆ prolongErrorAndCorrect()

void IBTK::PoissonFACPreconditionerStrategy::prolongErrorAndCorrect	(	const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &	src,
		SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &	dst,
		int	dst_ln
	)

overridevirtual

Prolong the error quantity to the specified level from the next coarser level and apply the correction to the fine-level error.

Parameters

src	source error vector
dst	destination error vector
dst_ln	destination level number of data transfer

Implements IBTK::FACPreconditionerStrategy.

◆ restrictResidual()

void IBTK::PoissonFACPreconditionerStrategy::restrictResidual	(	const SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &	src,
		SAMRAI::solv::SAMRAIVectorReal< NDIM, double > &	dst,
		int	dst_ln
	)

overridevirtual

Restrict the residual quantity to the specified level from the next finer level.

Parameters

src	source residual
dst	destination residual
dst_ln	destination level number

Implements IBTK::FACPreconditionerStrategy.

Reimplemented in IBTK::VCSCViscousOpPointRelaxationFACOperator.

◆ setCoarseSolverAbsoluteTolerance()

void IBTK::PoissonFACPreconditionerStrategy::setCoarseSolverAbsoluteTolerance ( double coarse_solver_abs_residual_tol )

Set the absolute residual tolerance for convergence for coarse level solve.

If the coarse level solver uses a absolute convergence tolerance parameter, the specified value is used. If the coarse level solver does not use such a stopping parameter, implementations are free to ignore this value.

◆ setCoarseSolverMaxIterations()

void IBTK::PoissonFACPreconditionerStrategy::setCoarseSolverMaxIterations ( int coarse_solver_max_iterations )

Set the maximum number of iterations for the coarse level solve.

If the coarse level solver uses a maximum number of iterations parameter, the specified value is used. If the coarse level solver does not use such a stopping parameter, implementations are free to ignore this value.

◆ setCoarseSolverRelativeTolerance()

void IBTK::PoissonFACPreconditionerStrategy::setCoarseSolverRelativeTolerance ( double coarse_solver_rel_residual_tol )

Set the relative residual tolerance for convergence for coarse level solve.

If the coarse level solver uses a relative convergence tolerance parameter, the specified value is used. If the coarse level solver does not use such a stopping parameter, implementations are free to ignore this value.

◆ setPhysicalBcCoef()

void IBTK::PoissonFACPreconditionerStrategy::setPhysicalBcCoef ( SAMRAI::solv::RobinBcCoefStrategy< NDIM > * bc_coef )

virtual

Set the SAMRAI::solv::RobinBcCoefStrategy object used to specify physical boundary conditions.

Note: bc_coef may be NULL. In this case, default boundary conditions (as supplied to the class constructor) are employed.

Parameters

bc_coef Pointer to an object that can set the Robin boundary condition coefficients

◆ setPhysicalBcCoefs()

void IBTK::PoissonFACPreconditionerStrategy::setPhysicalBcCoefs ( const std::vector< SAMRAI::solv::RobinBcCoefStrategy< NDIM > * > & bc_coefs )

virtual

Set the SAMRAI::solv::RobinBcCoefStrategy objects used to specify physical boundary conditions.

Note: Any of the elements of bc_coefs may be NULL. In this case, default boundary conditions (as supplied to the class constructor) are employed for that data depth.

Parameters

bc_coefs Vector of pointers to objects that can set the Robin boundary condition coefficients

◆ setResetLevels()

void IBTK::PoissonFACPreconditionerStrategy::setResetLevels	(	int	coarsest_ln,
		int	finest_ln
	)

Specify the levels that need to be reset the next time the operator is re-initialized.

When the operator is initialized, then only the specified range of levels are reset in the operator state the next time that the operator is initialized. If the operator is not initialized, this method has no effect.

To ensure the range of levels that is reset includes all levels in the patch hierarchy, use coarsest_ln = finest_ln = -1.

Note: This function is used to save some unnecessary computations when the hierarchy is regridded. The range of levels specified must include all levels which need to be reset by SAMRAI::mesh::StandardTagAndInitStrategy::resetHierarchyConfiguration(). Any data residing outside of this range of levels will not be reset. This is not what you want to have happen if, for instance, the Poisson specifications changes.

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

/home/runner/work/IBAMR/IBAMR/ibtk/include/ibtk/PoissonFACPreconditionerStrategy.h
/home/runner/work/IBAMR/IBAMR/ibtk/src/solvers/impls/PoissonFACPreconditionerStrategy.cpp

Public Member Functions

Protected Attributes

Methods for executing, caching, and resetting communication

Various refine and coarsen objects.

Partial implementation of FACPreconditionerStrategy interface.

Additional Inherited Members

Detailed Description

Member Function Documentation

◆ deallocateOperatorState()

◆ initializeOperatorState()

◆ prolongError()

◆ prolongErrorAndCorrect()

◆ restrictResidual()

◆ setCoarseSolverAbsoluteTolerance()

◆ setCoarseSolverMaxIterations()

◆ setCoarseSolverRelativeTolerance()

◆ setPhysicalBcCoef()

◆ setPhysicalBcCoefs()

◆ setResetLevels()