libMesh::FEType Class Reference

class FEType hides (possibly multiple) FEFamily and approximation orders, thereby enabling specialized finite element families. More...

#include <fe_type.h>

Public Member Functions
	FEType (const int o=1, const FEFamily f=LAGRANGE)
	Constructor. More...
	FEType (const int o=1, const FEFamily f=LAGRANGE, const int ro=THIRD, const FEFamily rf=JACOBI_20_00, const InfMapType im=CARTESIAN)
	Constructor. More...
bool	operator== (const FEType &f2) const
	Tests equality. More...
bool	operator!= (const FEType &f2) const
	Tests inequality. More...
bool	operator< (const FEType &f2) const
	An ordering to make FEType useful as a std::map key. More...
Order	default_quadrature_order () const
std::unique_ptr< QBase >	default_quadrature_rule (const unsigned int dim, const int extraorder=0) const
Order	unweighted_quadrature_order () const
std::unique_ptr< QBase >	unweighted_quadrature_rule (const unsigned int dim, const int extraorder=0) const

Public Attributes
OrderWrapper	order
	The approximation order of the element. More...
FEFamily	family
	The type of finite element. More...
OrderWrapper	radial_order
	The approximation order in radial direction of the infinite element. More...
FEFamily	radial_family
	The type of approximation in radial direction. More...
InfMapType	inf_map
	The coordinate mapping type of the infinite element. More...

Detailed Description

class FEType hides (possibly multiple) FEFamily and approximation orders, thereby enabling specialized finite element families.

Author

Benjamin S. Kirk

Date

2002

Manages the family, order, etc. parameters for a given FE.

Definition at line 178 of file fe_type.h.

Constructor & Destructor Documentation

FEType() [1/2]

libMesh::FEType::FEType ( const int  o = 1, const FEFamily  f = LAGRANGE  )

inline

Constructor.

Optionally takes the approximation Order and the finite element family FEFamily

Definition at line 188 of file fe_type.h.

                                      :
    order(o),
    family(f)
  {}

FEType() [2/2]

libMesh::FEType::FEType ( const int  o = 1, const FEFamily  f = LAGRANGE, const int  ro = THIRD, const FEFamily  rf = JACOBI_20_00, const InfMapType  im = CARTESIAN  )

inline

Constructor.

Optionally takes the approximation Order and the finite element family FEFamily.

Note

For non-infinite elements, the order and base order are the same, as with the family and base_family. It must be so, otherwise what we switch on would change when infinite elements are not compiled in.

Definition at line 216 of file fe_type.h.

                                          :
    order(o),
    radial_order(ro),
    family(f),
    radial_family(rf),
    inf_map(im)
  {}

Member Function Documentation

default_quadrature_order()

Order libMesh::FEType::default_quadrature_order ( ) const

inline

Returns

The default quadrature order for this FEType. The default quadrature order is calculated assuming a polynomial of degree order and is based on integrating the mass matrix for such an element exactly on affine elements.

Definition at line 353 of file fe_type.h.

{
  return static_cast<Order>(2*static_cast<unsigned int>(order.get_order()) + 1);
}

References libMesh::OrderWrapper::get_order(), and order.

Referenced by LinearElasticity::assemble(), AssembleOptimization::assemble_A_and_F(), assemble_elasticity(), assemble_ellipticdg(), assemble_poisson(), assemble_shell(), assemble_stokes(), compute_jacobian(), libMesh::FEGenericBase< FEOutputType< T >::type >::compute_periodic_constraints(), libMesh::FEGenericBase< FEOutputType< T >::type >::compute_proj_constraints(), compute_residual(), LinearElasticityWithContact::compute_stresses(), LinearElasticity::compute_stresses(), compute_stresses(), LargeDeformationElasticity::compute_stresses(), default_quadrature_rule(), LargeDeformationElasticity::jacobian(), OverlappingCouplingFunctor::operator()(), LargeDeformationElasticity::residual(), LinearElasticityWithContact::residual_and_jacobian(), InfFERadialTest::testSingleOrder(), and libMesh::Elem::volume().

default_quadrature_rule()

std::unique_ptr< QBase > libMesh::FEType::default_quadrature_rule	(	const unsigned int	dim,
		const int	extraorder = 0
	)		const

Returns

A quadrature rule of appropriate type and order for this FEType. The default quadrature rule is based on integrating the mass matrix for such an element exactly, with an additional power on the basis order to help account for nonlinearities and/or nonuniform coefficients. Higher or lower degree rules can be chosen by changing the extraorder parameter.

Definition at line 31 of file fe_type.C.

{
  // Clough elements have at least piecewise cubic functions
  if (family == CLOUGH)
    {
      Order o = static_cast<Order>(std::max(static_cast<unsigned int>(this->default_quadrature_order()),
                                            static_cast<unsigned int>(7 + extraorder)));
      return libmesh_make_unique<QClough>(dim, o);
    }
 
  if (family == SUBDIVISION)
    return libmesh_make_unique<QGauss>(dim, static_cast<Order>(1 + extraorder));
 
  return libmesh_make_unique<QGauss>(dim, static_cast<Order>(this->default_quadrature_order() + extraorder));
}

References libMesh::CLOUGH, default_quadrature_order(), dim, family, and libMesh::SUBDIVISION.

Referenced by libMesh::ExactSolution::_compute_error(), libMesh::UniformRefinementEstimator::_estimate_error(), libMesh::RBConstruction::add_scaled_matrix_and_vector(), assemble_biharmonic(), assemble_laplace(), assemble_shell(), libMesh::System::calculate_norm(), libMesh::FEGenericBase< FEOutputType< T >::type >::coarsened_dof_values(), libMesh::ExactErrorEstimator::estimate_error(), libMesh::WeightedPatchRecoveryErrorEstimator::EstimateError::operator()(), libMesh::PatchRecoveryErrorEstimator::EstimateError::operator()(), libMesh::BoundaryProjectSolution::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectEdges::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectSides::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectInteriors::operator()(), Biharmonic::JR::residual_and_jacobian(), libMesh::HPCoarsenTest::select_refinement(), and libMesh::FEMContext::use_default_quadrature_rules().

operator!=()

bool libMesh::FEType::operator!= ( const FEType &  f2 ) const

inline

Tests inequality.

Definition at line 279 of file fe_type.h.

  {
    return !(*this == f2);
  }

operator<()

bool libMesh::FEType::operator< ( const FEType &  f2 ) const

inline

An ordering to make FEType useful as a std::map key.

Definition at line 287 of file fe_type.h.

  {
    if (order != f2.order)
      return (order < f2.order);
    if (family != f2.family)
      return (family < f2.family);
 
#ifdef LIBMESH_ENABLE_INFINITE_ELEMENTS
    if (radial_order != f2.radial_order)
      return (radial_order < f2.radial_order);
    if (radial_family != f2.radial_family)
      return (radial_family < f2.radial_family);
    if (inf_map != f2.inf_map)
      return (inf_map < f2.inf_map);
#endif // ifdef LIBMESH_ENABLE_INFINITE_ELEMENTS
    return false;
  }

References family, inf_map, order, radial_family, and radial_order.

operator==()

bool libMesh::FEType::operator== ( const FEType &  f2 ) const

inline

Tests equality.

Definition at line 264 of file fe_type.h.

  {
    return (order == f2.order
            && family == f2.family
#ifdef LIBMESH_ENABLE_INFINITE_ELEMENTS
            && radial_order == f2.radial_order
            && radial_family == f2.radial_family
            && inf_map == f2.inf_map
#endif // ifdef LIBMESH_ENABLE_INFINITE_ELEMENTS
            );
  }

References family, inf_map, order, radial_family, and radial_order.

unweighted_quadrature_order()

Order libMesh::FEType::unweighted_quadrature_order ( ) const

inline

Returns

The default quadrature order for integrating unweighted basis functions of this FEType. The unweighted quadrature order is calculated assuming a polynomial of degree order and is based on integrating the shape functions for such an element exactly on affine elements.

Definition at line 359 of file fe_type.h.

{
  return order;
}

References order.

Referenced by unweighted_quadrature_rule().

unweighted_quadrature_rule()

std::unique_ptr< QBase > libMesh::FEType::unweighted_quadrature_rule	(	const unsigned int	dim,
		const int	extraorder = 0
	)		const

Returns

A quadrature rule of appropriate type and order for unweighted integration of this FEType. The default quadrature rule is based on integrating the shape functions on an affine element exactly. Higher or lower degree rules can be chosen by changing the extraorder parameter.

Definition at line 50 of file fe_type.C.

{
  // Clough elements have at least piecewise cubic functions
  if (family == CLOUGH)
    {
      Order o = static_cast<Order>(std::max(static_cast<unsigned int>(this->unweighted_quadrature_order()),
                                            static_cast<unsigned int>(3 + extraorder)));
      return libmesh_make_unique<QClough>(dim, o);
    }
 
  if (family == SUBDIVISION)
    return libmesh_make_unique<QGauss>(dim, static_cast<Order>(1 + extraorder));
 
  return libmesh_make_unique<QGauss>(dim, static_cast<Order>(this->unweighted_quadrature_order() + extraorder));
}

References libMesh::CLOUGH, dim, family, libMesh::SUBDIVISION, and unweighted_quadrature_order().

Referenced by libMesh::FEMContext::use_unweighted_quadrature_rules().

Member Data Documentation

family

FEFamily libMesh::FEType::family

The type of finite element.

For InfFE, family contains the radial shape family, while base_family contains the approximation type in circumferential direction.

Valid types are LAGRANGE, HIERARCHIC, etc...

Definition at line 203 of file fe_type.h.

Referenced by libMesh::DofMap::_dof_indices(), libMesh::FEMSystem::assembly(), libMesh::FETransformationBase< OutputShape >::build(), libMesh::FEMap::build(), libMesh::FEAbstract::build(), libMesh::FEGenericBase< FEOutputType< T >::type >::build(), libMesh::FEMContext::build_new_fe(), libMesh::FEInterface::compute_constraints(), libMesh::GMVIO::copy_nodal_solution(), default_quadrature_rule(), libMesh::DofMap::distribute_local_dofs_node_major(), libMesh::DofMap::distribute_local_dofs_var_major(), libMesh::DofMap::dof_indices(), libMesh::JumpErrorEstimator::estimate_error(), libMesh::FEInterface::extra_hanging_dofs(), libMesh::FEInterface::field_type(), libMesh::FEMContext::find_hardest_fe_type(), libMesh::FEInterface::get_continuity(), libMesh::FEAbstract::get_family(), libMesh::System::get_info(), libMesh::FEMContext::init_internal_data(), libMesh::FEInterface::max_order(), libMesh::Variable::n_components(), libMesh::FEInterface::n_vec_dim(), libMesh::DifferentiablePhysics::nonlocal_mass_residual(), libMesh::DofMap::old_dof_indices(), libMesh::BoundaryProjectSolution::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::SortAndCopy::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectVertices::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectEdges::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectSides::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectInteriors::operator()(), operator<(), operator==(), libMesh::System::read_header(), libMesh::System::read_parallel_data(), libMesh::DofMap::reinit(), libMesh::PetscDMWrapper::set_point_range_in_section(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::SubFunctor::SubFunctor(), libMesh::MeshFunctionSolutionTransfer::transfer(), unweighted_quadrature_rule(), libMesh::System::write_header(), libMesh::Nemesis_IO_Helper::write_nodal_solution(), and libMesh::System::write_parallel_data().

inf_map

InfMapType libMesh::FEType::inf_map

The coordinate mapping type of the infinite element.

When the infinite elements are defined over a surface with a separable coordinate system (sphere, spheroid, ellipsoid), the infinite elements may take advantage of this fact.

Definition at line 257 of file fe_type.h.

Referenced by libMesh::FEGenericBase< FEOutputType< T >::type >::build_InfFE(), libMesh::FEInterface::ifem_inverse_map(), libMesh::FEInterface::ifem_map(), libMesh::FEInterface::ifem_nodal_soln(), libMesh::InfFE< Dim, T_radial, T_map >::InfFE(), operator<(), operator==(), libMesh::System::read_header(), and libMesh::System::write_header().

order

OrderWrapper libMesh::FEType::order

The approximation order of the element.

The approximation order in the base of the infinite element.

Definition at line 197 of file fe_type.h.

Referenced by libMesh::DofMap::_dof_indices(), libMesh::DofMap::_node_dof_indices(), libMesh::PetscDMWrapper::add_dofs_to_section(), libMesh::FEMContext::build_new_fe(), libMesh::FEGenericBase< FEOutputType< T >::type >::coarsened_dof_values(), libMesh::FEGenericBase< FEOutputType< T >::type >::compute_proj_constraints(), libMesh::DofMap::constrain_p_dofs(), libMesh::GMVIO::copy_nodal_solution(), default_quadrature_order(), libMesh::DofMap::distribute_dofs(), libMesh::DofMap::distribute_local_dofs_node_major(), libMesh::DofMap::distribute_local_dofs_var_major(), libMesh::DofMap::dof_indices(), libMesh::FEInterface::dofs_on_edge(), libMesh::FEInterface::dofs_on_side(), libMesh::FEMContext::find_hardest_fe_type(), libMesh::FEAbstract::get_order(), libMesh::FESubdivision::init_shape_functions(), libMesh::Variable::n_components(), libMesh::FEInterface::n_dofs(), libMesh::FEInterface::n_dofs_at_node(), libMesh::FEInterface::n_dofs_per_elem(), libMesh::FEInterface::n_shape_functions(), libMesh::FEInterface::nodal_soln(), libMesh::DofMap::old_dof_indices(), libMesh::WeightedPatchRecoveryErrorEstimator::EstimateError::operator()(), libMesh::PatchRecoveryErrorEstimator::EstimateError::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::SortAndCopy::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectEdges::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectSides::operator()(), libMesh::GenericProjector< FFunctor, GFunctor, FValue, ProjectionAction >::ProjectInteriors::operator()(), operator<(), operator==(), libMesh::System::read_header(), libMesh::System::read_SCALAR_dofs(), libMesh::DofMap::reinit(), libMesh::DofMap::SCALAR_dof_indices(), libMesh::HPCoarsenTest::select_refinement(), libMesh::FEAbstract::set_fe_order(), libMesh::FEInterface::shape(), libMesh::FEInterface::shape_deriv(), libMesh::FEInterface::shape_second_deriv(), and unweighted_quadrature_order().

radial_family

FEFamily libMesh::FEType::radial_family

The type of approximation in radial direction.

Valid types are JACOBI_20_00, JACOBI_30_00, etc...

Definition at line 249 of file fe_type.h.

Referenced by libMesh::FEGenericBase< FEOutputType< T >::type >::build_InfFE(), libMesh::FEInterface::ifem_compute_data(), libMesh::FEInterface::ifem_nodal_soln(), libMesh::FEInterface::ifem_shape(), libMesh::InfFE< Dim, T_radial, T_map >::InfFE(), operator<(), operator==(), libMesh::System::read_header(), and libMesh::System::write_header().

radial_order

OrderWrapper libMesh::FEType::radial_order

The approximation order in radial direction of the infinite element.

Definition at line 236 of file fe_type.h.

Referenced by libMesh::InfFE< Dim, T_radial, T_map >::compute_data(), libMesh::InfFE< Dim, T_radial, T_map >::compute_shape_indices(), libMesh::InfFE< Dim, T_radial, T_map >::n_dofs(), libMesh::InfFE< Dim, T_radial, T_map >::n_dofs_at_node(), libMesh::InfFE< Dim, T_radial, T_map >::n_dofs_per_elem(), operator<(), operator==(), libMesh::System::read_header(), libMesh::InfFE< Dim, T_radial, T_map >::shape(), and libMesh::InfFE< Dim, T_radial, T_map >::shape_deriv().

---

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

• include/fe/fe_type.h
• src/fe/fe_type.C