Factory for creating AMRProjectionOp's. More...

#include <AMRProjectionOp.h>

Inheritance diagram for AMRProjectionOpFactory:

Public Member Functions
void	define (const ProblemDomain &a_coarseDomain, const Vector< DisjointBoxLayout > &a_grids, const Vector< int > &a_refRatios, const Real &a_coarsedx, BCHolder a_bc, const Real &a_alpha, Vector< RefCountedPtr< LevelData< FArrayBox > > > &a_aCoef, const Real &a_beta, Vector< RefCountedPtr< LevelData< FluxBox > > > &a_bCoef, int a_averaging_type=CoarseAverage::arithmetic)

void	define (const ProblemDomain &a_coarseDomain, const Vector< DisjointBoxLayout > &a_grids, const Vector< int > &a_refRatios, const Real &a_coarsedx, BCHolder a_bc, const IntVect &a_ghostVect, int a_averaging_type=CoarseAverage::arithmetic)

virtual MGLevelOp< LevelData< FArrayBox > > *	MGnewOp (const ProblemDomain &a_FineindexSpace, int a_depth, bool a_homoOnly=true)
	Make a new multigrid operator for a given domain.

virtual AMRLevelOp< LevelData< FArrayBox > > *	AMRnewOp (const ProblemDomain &a_indexSpace)
	Make a new AMR multirid operator.

virtual int	refToFiner (const ProblemDomain &a_domain) const
	Returns the refinement ratio to the next finer level.

Public Member Functions inherited from AMRLevelOpFactory< LevelData< FArrayBox > >
	MGLevelOpFactory ()

Public Attributes
int	m_coefficient_average_type
	How to do coefficient averaging (arithmetic, harmonic, etc. )

int	m_relaxMode
	How to do relaxation (gauss-seidel, jacobi, etc.)

Detailed Description

Factory for creating AMRProjectionOp's.

Would have liked to base this off VCAMRPoissonOp2Factory, but that's all private so we have to do it ourselves

Member Function Documentation

◆ define() [1/2]

void AMRProjectionOpFactory::define	(	const ProblemDomain &	a_coarseDomain,
		const Vector< DisjointBoxLayout > &	a_grids,
		const Vector< int > &	a_refRatios,
		const Real &	a_coarsedx,
		BCHolder	a_bc,
		const Real &	a_alpha,
		Vector< RefCountedPtr< LevelData< FArrayBox > > > &	a_aCoef,
		const Real &	a_beta,
		Vector< RefCountedPtr< LevelData< FluxBox > > > &	a_bCoef,
		int	a_averaging_type = `CoarseAverage::arithmetic`
	)

a_coarseDomain is the domain at the coarsest level. a_grids is the AMR hierarchy. a_refRatios are the refinement ratios between levels. The ratio lives with the coarser level so a_refRatios[ilev] is the ratio between ilev and ilev+1 a_coarseDx is the grid spacing at the coarsest level. a_bc holds the boundary conditions. a_alpha is the identity constant coefficient a_beta is the laplacian constant coefficient. a_aCoef is the identity spatially varying coefficient a_bCoef is the laplacian spatially varying coefficient a_averaging_type is the method to use for averaging cell to face centred variables

◆ define() [2/2]

void AMRProjectionOpFactory::define	(	const ProblemDomain &	a_coarseDomain,
		const Vector< DisjointBoxLayout > &	a_grids,
		const Vector< int > &	a_refRatios,
		const Real &	a_coarsedx,
		BCHolder	a_bc,
		const IntVect &	a_ghostVect,
		int	a_averaging_type = `CoarseAverage::arithmetic`
	)

Defines a factory for VCAMRPoissonOp2 which allows the operators to allocate their own coefficient data. $\alpha$ and $\beta$ coefficients are initialized to 1.

Parameters

a_coarseDomain	The domain at the coarsest level.
a_grids	The disjoint box layouts for the various refinement levels.
a_refRatios	The refinement ratios between levels.
a_coarsedx	The grid spacing at the coarsest level.
a_bc	The boundary condition imposed on the solution.
a_ghostVect	The ghost stencil to use in the created coefficient data.
a_averaging_type	The method to use for averaging cell to face centred variables

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

/home/parkinsonjl/mushy-layer/src/AMRProjectionOp.h
/home/parkinsonjl/mushy-layer/src/AMRProjectionOp.cpp

Public Member Functions

Public Attributes

Detailed Description

Member Function Documentation

◆ define() [1/2]

◆ define() [2/2]