oft_hcurl_operators Module Reference

Detailed Description

H(Curl) FE operator definitions.

• Operator construction
  • MOP: mass matrix \( \int \left( u^T \cdot v \right) dV \)
  • KOP: helicity matrix \( \int \left( \nabla \times u^T \cdot v \right) dV \)
  • WOP: energy matrix \( \int \left( \nabla \times u^T \cdot \nabla \times v \right) dV \)
• Interpolation classes
• Explicit operators
  • Gradient
  • Divergence
  • GradT
  • MC gradient
  • Divergence cleaning
• Boundary conditions
• Multi-Grid setup and operators
• Default preconditioner setup
  • Optimal smoother calculation

Authors

Chris Hansen

Date

August 2011

Data Types
type	oft_hcurl_cinterp
	Interpolate \( \nabla \times \) of a H(Curl) field. More...
type	oft_hcurl_divout
	Clean the divergence from a H(Curl) vector field. More...
type	oft_hcurl_orthog
	Orthogonalize a H(Curl) vector against a library of given modes. More...
type	oft_hcurl_rinterp
	Interpolate a H(Curl) field. More...
type	oft_hcurl_zerob
	Needs docs. More...

Functions/Subroutines
subroutine	hcurl_base_pop (self, acors, afine)
	Transfer a base level H(Curl) vector field to the next MPI level.
subroutine	hcurl_base_push (self, afine, acors)
	Transfer a MPI level H(Curl) vector field to the base level.
subroutine	hcurl_cinterp (self, cell, f, gop, val)
	Reconstruct the curl of a H(Curl) vector field.
subroutine	hcurl_div (hcurl_fe, lag_fe, a, b)
	Apply the divergence operator to a H(Curl) field.
subroutine	hcurl_divout_apply (self, a)
	Remove divergence from a H(Curl) vector field by adding a gradient correction.
subroutine	hcurl_divout_delete (self)
	Clean-up internal storage for a oft_hcurl_divout object.
subroutine	hcurl_divout_setup (self, ml_hcurl_rep, ml_lag_rep, bc, solver)
	Setup matrix and solver with default.
subroutine	hcurl_getjmlb_pre (ml_hcurl_rep, pre, mats, alam, level, nlevels)
	Construct default MG preconditioner for H(Curl)::JMLB.
subroutine	hcurl_getwop_pre (ml_hcurl_rep, pre, mats, level, nlevels)
	Construct default MG preconditioner for H(Curl)::WOP.
subroutine	hcurl_grad (hcurl_fe, a, b)
	Add the gradient of a linear Lagrange scalar field to a H(Curl) field.
subroutine	hcurl_gradtp (hcurl_fe, a, b)
	Apply the transposed gradient operator to a H(Curl) vector field.
subroutine	hcurl_mloptions (ml_hcurl_obj)
	Read-in options for the basic H(Curl) ML preconditioners.
subroutine	hcurl_orthog_apply (self, a)
	Orthogonalize a H(Curl) vector against a library of given modes.
subroutine	hcurl_orthog_delete (self)
	Clean-up internal storage for a oft_hcurl_orthog object.
subroutine	hcurl_rinterp (self, cell, f, gop, val)
	Reconstruct a H(Curl) vector field.
subroutine	hcurl_rinterp_delete (self)
	Destroy temporary internal storage.
subroutine	hcurl_rinterp_setup (self, hcurl_rep)
	Setup interpolator for H(Curl) fields.
subroutine	hcurl_setup_interp (ml_hcurl_rep)
	Construct interpolation matrices on each MG level.
subroutine	hcurl_wop_eigs (ml_hcurl_rep, minlev)
	Compute eigenvalues and smoothing coefficients for the operator H(Curl)::WOP.
subroutine	oft_hcurl_bcurl (fe_rep, bfe_rep, field, x)
	Compute the boundary term for integration by parts of the curl operator using a HCurl basis.
subroutine	oft_hcurl_getjmlb (fe_rep, mat, alam, bc)
	Construct force-free response matrix for a H(Curl) representation.
subroutine	oft_hcurl_getkop (fe_rep, mat, bc)
	Construct helicity matrix for a H(Curl) representation.
subroutine	oft_hcurl_getmop (fe_rep, mat, bc)
	Construct mass matrix for a H(Curl) representation.
subroutine	oft_hcurl_getwop (fe_rep, mat, bc)
	Construct energy matrix for a H(Curl) representation.
subroutine	oft_hcurl_project (fe_rep, field, x)
	Project a vector field onto a H(Curl) basis.
subroutine	zerob_apply (self, a)
	Zero the tangential component of a H(Curl) vector field on the boundary.
subroutine	zerob_delete (self)
	Zero the tangential component of a H(Curl) vector field on the boundary.

Variables
real(r8), dimension(fem_max_levels)	df_wop =-1.d99
integer(i4), dimension(fem_max_levels)	nu_jmlb =20
integer(i4), dimension(fem_max_levels)	nu_wop =0
real(r8), dimension(:,:), pointer	oft_hcurl_cop => NULL()
real(r8), dimension(:,:), pointer	oft_hcurl_rop => NULL()

Function/Subroutine Documentation

hcurl_base_pop()

subroutine hcurl_base_pop	(	class(oft_ml_fe_vecspace), intent(inout)	self,
		class(oft_vector), intent(inout)	acors,
		class(oft_vector), intent(inout)	afine )

Transfer a base level H(Curl) vector field to the next MPI level.

Parameters

[in,out]	acors	Vector to transfer
[in,out]	afine	Fine vector from transfer

hcurl_base_push()

subroutine hcurl_base_push	(	class(oft_ml_fe_vecspace), intent(inout)	self,
		class(oft_vector), intent(inout)	afine,
		class(oft_vector), intent(inout)	acors )

Transfer a MPI level H(Curl) vector field to the base level.

Parameters

[in,out]	afine	Vector to transfer
[in,out]	acors	Fine vector from transfer

hcurl_cinterp()

subroutine hcurl_cinterp	(	class(oft_hcurl_cinterp), intent(inout)	self,
		integer(i4), intent(in)	cell,
		real(r8), dimension(:), intent(in)	f,
		real(r8), dimension(3,4), intent(in)	gop,
		real(r8), dimension(:), intent(out)	val )

Reconstruct the curl of a H(Curl) vector field.

Parameters

[in]	cell	Cell for interpolation
[in]	f	Position in cell in logical coord [4]
[in]	gop	Logical gradient vectors at f [3,4]
[out]	val	Reconstructed field at f [3]

hcurl_div()

subroutine hcurl_div	(	class(oft_afem_type), intent(inout)	hcurl_fe,
		class(oft_afem_type), intent(inout)	lag_fe,
		class(oft_vector), intent(inout)	a,
		class(oft_vector), intent(inout)	b )

Apply the divergence operator to a H(Curl) field.

Note

This subroutine computes the divergence of a H(Curl) field as projected on to a linear Lagrange scalar basis.

Parameters

[in,out]	a	Needs docs
[in,out]	b	Needs docs

hcurl_divout_apply()

subroutine hcurl_divout_apply	(	class(oft_hcurl_divout), intent(inout)	self,
		class(oft_vector), intent(inout)	a )

Remove divergence from a H(Curl) vector field by adding a gradient correction.

Parameters

[in,out]

a

Field for divergence cleaning

hcurl_divout_delete()

subroutine hcurl_divout_delete

(

class(oft_hcurl_divout), intent(inout)

self

)

Clean-up internal storage for a oft_hcurl_divout object.

hcurl_divout_setup()

subroutine hcurl_divout_setup	(	class(oft_hcurl_divout), intent(inout)	self,
		class(oft_ml_fem_type), intent(inout), target	ml_hcurl_rep,
		class(oft_ml_fem_type), intent(inout), target	ml_lag_rep,
		character(len=*), intent(in)	bc,
		class(oft_solver), intent(in), optional, target	solver )

Setup matrix and solver with default.

Parameters

[in]

bc

Boundary condition

hcurl_getjmlb_pre()

subroutine hcurl_getjmlb_pre	(	type(oft_ml_fem_type), intent(inout), target	ml_hcurl_rep,
		class(oft_solver), intent(out), pointer	pre,
		type(oft_matrix_ptr), dimension(:), intent(inout), pointer	mats,
		real(r8), intent(in)	alam,
		integer(i4), intent(in), optional	level,
		integer(i4), intent(in), optional	nlevels )

Construct default MG preconditioner for H(Curl)::JMLB.

Parameters

[out]	pre	Needs docs
[in,out]	mats	Needs docs
[in]	alam	Needs docs
[in]	level	Needs docs
[in]	nlevels	Needs docs

hcurl_getwop_pre()

subroutine hcurl_getwop_pre	(	type(oft_ml_fem_type), intent(inout), target	ml_hcurl_rep,
		class(oft_solver), intent(out), pointer	pre,
		type(oft_matrix_ptr), dimension(:), intent(inout), pointer	mats,
		integer(i4), intent(in), optional	level,
		integer(i4), intent(in), optional	nlevels )

Construct default MG preconditioner for H(Curl)::WOP.

Parameters

[out]	pre	Needs docs
[in,out]	mats	Needs docs
[in]	level	Needs docs
[in]	nlevels	Needs docs

hcurl_grad()

subroutine hcurl_grad	(	class(oft_afem_type), intent(inout)	hcurl_fe,
		class(oft_vector), intent(inout)	a,
		class(oft_vector), intent(inout)	b )

Add the gradient of a linear Lagrange scalar field to a H(Curl) field.

Parameters

[in,out]	a	Needs docs
[in,out]	b	Needs docs

hcurl_gradtp()

subroutine hcurl_gradtp	(	class(oft_afem_type), intent(inout)	hcurl_fe,
		class(oft_vector), intent(inout)	a,
		class(oft_vector), intent(inout)	b )

Apply the transposed gradient operator to a H(Curl) vector field.

Note

Only the 0th order component is computed from the discrete gradient operator

Parameters

[in,out]	a	Input field
[in,out]	b	\( G^{T} a \)

hcurl_mloptions()

subroutine hcurl_mloptions

(

class(oft_ml_fem_type), intent(inout)

ml_hcurl_obj

)

Read-in options for the basic H(Curl) ML preconditioners.

hcurl_orthog_apply()

subroutine hcurl_orthog_apply	(	class(oft_hcurl_orthog), intent(inout)	self,
		class(oft_vector), intent(inout)	a )

Orthogonalize a H(Curl) vector against a library of given modes.

Note

Used as a member function of oft_hcurl_orthog only

Parameters

[in,out]

a

Field to orthogonalize

hcurl_orthog_delete()

subroutine hcurl_orthog_delete

(

class(oft_hcurl_orthog), intent(inout)

self

)

Clean-up internal storage for a oft_hcurl_orthog object.

Note

Used as a member function of oft_hcurl_orthog only

hcurl_rinterp()

subroutine hcurl_rinterp	(	class(oft_hcurl_rinterp), intent(inout)	self,
		integer(i4), intent(in)	cell,
		real(r8), dimension(:), intent(in)	f,
		real(r8), dimension(3,4), intent(in)	gop,
		real(r8), dimension(:), intent(out)	val )

Reconstruct a H(Curl) vector field.

Parameters

[in]	cell	Cell for interpolation
[in]	f	Position in cell in logical coord [4]
[in]	gop	Logical gradient vectors at f [3,4]
[out]	val	Reconstructed field at f [3]

hcurl_rinterp_delete()

subroutine hcurl_rinterp_delete

(

class(oft_hcurl_rinterp), intent(inout)

self

)

Destroy temporary internal storage.

hcurl_rinterp_setup()

subroutine hcurl_rinterp_setup	(	class(oft_hcurl_rinterp), intent(inout)	self,
		class(oft_afem_type), intent(inout), target	hcurl_rep )

Setup interpolator for H(Curl) fields.

Fetches local representation used for interpolation from vector object

hcurl_setup_interp()

subroutine hcurl_setup_interp

(

class(oft_ml_fem_type), intent(inout)

ml_hcurl_rep

)

Construct interpolation matrices on each MG level.

hcurl_wop_eigs()

subroutine hcurl_wop_eigs	(	type(oft_ml_fem_type), intent(inout), target	ml_hcurl_rep,
		integer(i4), intent(in)	minlev )

Compute eigenvalues and smoothing coefficients for the operator H(Curl)::WOP.

Parameters

[in]

minlev

Needs docs

oft_hcurl_bcurl()

subroutine oft_hcurl_bcurl	(	class(oft_afem_type), intent(inout), target	fe_rep,
		class(oft_afem_type), intent(inout), target	bfe_rep,
		class(fem_interp), intent(inout)	field,
		class(oft_vector), intent(inout)	x )

Compute the boundary term for integration by parts of the curl operator using a HCurl basis.

Parameters

[in,out]	field	Vector field for projection
[in,out]	x	\( \int \left( \textbf{u}^T \times \textbf{F} \right) \cdot \textbf{dS} \)

oft_hcurl_getjmlb()

subroutine oft_hcurl_getjmlb	(	class(oft_afem_type), intent(inout), target	fe_rep,
		class(oft_matrix), intent(inout), pointer	mat,
		real(r8), intent(in)	alam,
		character(len=*), intent(in)	bc )

Construct force-free response matrix for a H(Curl) representation.

Supported boundary conditions

• 'none' Full matrix
• 'zerob' Dirichlet for all boundary DOF

  Parameters

  [in,out]	mat	Matrix object
  [in]	alam	Lambda for response
  [in]	bc	Boundary condition

oft_hcurl_getkop()

subroutine oft_hcurl_getkop	(	class(oft_afem_type), intent(inout), target	fe_rep,
		class(oft_matrix), intent(inout), pointer	mat,
		character(len=*), intent(in)	bc )

Construct helicity matrix for a H(Curl) representation.

Supported boundary conditions

• 'none' Full matrix
• 'zerob' Dirichlet for all boundary DOF

  Parameters

  [in,out]	mat	Matrix object
  [in]	bc	Boundary condition

oft_hcurl_getmop()

subroutine oft_hcurl_getmop	(	class(oft_afem_type), intent(inout), target	fe_rep,
		class(oft_matrix), intent(inout), pointer	mat,
		character(len=*), intent(in)	bc )

Construct mass matrix for a H(Curl) representation.

Supported boundary conditions

• 'none' Full matrix
• 'zerob' Dirichlet for all boundary DOF

  Parameters

  [in,out]	mat	Matrix object
  [in]	bc	Boundary condition

oft_hcurl_getwop()

subroutine oft_hcurl_getwop	(	class(oft_afem_type), intent(inout), target	fe_rep,
		class(oft_matrix), intent(inout), pointer	mat,
		character(len=*), intent(in)	bc )

Construct energy matrix for a H(Curl) representation.

Supported boundary conditions

• 'none' Full matrix
• 'zerob' Dirichlet for all boundary DOF

  Parameters

  [in,out]	mat	Matrix object
  [in]	bc	Boundary condition

oft_hcurl_project()

subroutine oft_hcurl_project	(	class(oft_afem_type), intent(inout), target	fe_rep,
		class(fem_interp), intent(inout)	field,
		class(oft_vector), intent(inout)	x )

Project a vector field onto a H(Curl) basis.

Note

This subroutine only performs the integration of the field with test functions for a H(Curl) basis. To retrieve the correct projection the result must be multiplied by the inverse of H(Curl)::MOP

Parameters

[in,out]	field	Vector field for projection
[in,out]	x	Field projected onto H(Curl) basis

zerob_apply()

subroutine zerob_apply	(	class(oft_hcurl_zerob), intent(inout)	self,
		class(oft_vector), intent(inout)	a )

Zero the tangential component of a H(Curl) vector field on the boundary.

Parameters

[in,out]

a

Field to be zeroed

zerob_delete()

subroutine zerob_delete

(

class(oft_hcurl_zerob), intent(inout)

self

)

Zero the tangential component of a H(Curl) vector field on the boundary.

Variable Documentation

df_wop

real(r8), dimension(fem_max_levels) df_wop =-1.d99

nu_jmlb

integer(i4), dimension(fem_max_levels) nu_jmlb =20

nu_wop

integer(i4), dimension(fem_max_levels) nu_wop =0

oft_hcurl_cop

real(r8), dimension(:,:), pointer oft_hcurl_cop => NULL()

oft_hcurl_rop

real(r8), dimension(:,:), pointer oft_hcurl_rop => NULL()