Public Member Functions | Private Member Functions | Private Attributes | List of all members
FluxPoissonMGProblem< ELEMENT, MESH > Class Template Reference
+ Inheritance diagram for FluxPoissonMGProblem< ELEMENT, MESH >:

Public Member Functions

 FluxPoissonMGProblem (PoissonEquations< 2 >::PoissonSourceFctPt source_fct_pt)
 Constructor: Pass a pointer to the source function. More...
 
 ~FluxPoissonMGProblem ()
 Destructor. More...
 
void actions_before_adapt ()
 Actions before adapt: Wipe the mesh of prescribed flux elements. More...
 
void actions_after_adapt ()
 Actions after adapt: Rebuild the mesh of prescribed flux elements. More...
 
void actions_before_newton_solve ()
 
void actions_after_newton_solve ()
 Update the problem specs after solve (empty) More...
 
void doc_solution (DocInfo &doc_info)
 Document the solution: doc_info contains labels/output directory, etc. More...
 
 FluxPoissonMGProblem (PoissonEquations< 2 >::PoissonSourceFctPt source_fct_pt)
 Constructor: Pass pointer to source function. More...
 
 ~FluxPoissonMGProblem ()
 Destructor. More...
 
void doc_solution ()
 Doc the solution: doc_info contains labels/output directory etc. More...
 
void actions_before_adapt ()
 Actions before adapt: Wipe the mesh of prescribed flux elements. More...
 
void actions_after_adapt ()
 Actions after adapt: Rebuild the mesh of prescribed flux elements. More...
 
TreeBasedRefineableMeshBase * mg_bulk_mesh_pt ()
 
void actions_before_newton_solve ()
 
void actions_after_newton_solve ()
 Update the problem specs after solve (empty) More...
 
void set_multigrid_solver ()
 Build and set multgrid solver. More...
 
- Public Member Functions inherited from oomph::MGProblem
 MGProblem ()
 Constructor. Initialise pointers to coarser and finer levels. More...
 
virtual ~MGProblem ()
 Destructor (empty) More...
 
- Public Member Functions inherited from oomph::Problem
virtual void debug_hook_fct (const unsigned &i)
 
void set_analytic_dparameter (double *const &parameter_pt)
 
void unset_analytic_dparameter (double *const &parameter_pt)
 
bool is_dparameter_calculated_analytically (double *const &parameter_pt)
 
void set_analytic_hessian_products ()
 
void unset_analytic_hessian_products ()
 
bool are_hessian_products_calculated_analytically ()
 
void set_pinned_values_to_zero ()
 
bool distributed () const
 
OomphCommunicator * communicator_pt ()
 access function to the oomph-lib communicator More...
 
const OomphCommunicator * communicator_pt () const
 access function to the oomph-lib communicator, const version More...
 
 Problem ()
 
 Problem (const Problem &dummy)=delete
 Broken copy constructor. More...
 
void operator= (const Problem &)=delete
 Broken assignment operator. More...
 
virtual ~Problem ()
 Virtual destructor to clean up memory. More...
 
Mesh *& mesh_pt ()
 Return a pointer to the global mesh. More...
 
Mesh *const & mesh_pt () const
 Return a pointer to the global mesh (const version) More...
 
Mesh *& mesh_pt (const unsigned &imesh)
 
Mesh *const & mesh_pt (const unsigned &imesh) const
 Return a pointer to the i-th submesh (const version) More...
 
unsigned nsub_mesh () const
 Return number of submeshes. More...
 
unsigned add_sub_mesh (Mesh *const &mesh_pt)
 
void flush_sub_meshes ()
 
void build_global_mesh ()
 
void rebuild_global_mesh ()
 
LinearSolver *& linear_solver_pt ()
 Return a pointer to the linear solver object. More...
 
LinearSolver *const & linear_solver_pt () const
 Return a pointer to the linear solver object (const version) More...
 
LinearSolver *& mass_matrix_solver_for_explicit_timestepper_pt ()
 
LinearSolver * mass_matrix_solver_for_explicit_timestepper_pt () const
 
EigenSolver *& eigen_solver_pt ()
 Return a pointer to the eigen solver object. More...
 
EigenSolver *const & eigen_solver_pt () const
 Return a pointer to the eigen solver object (const version) More...
 
Time *& time_pt ()
 Return a pointer to the global time object. More...
 
Time * time_pt () const
 Return a pointer to the global time object (const version). More...
 
double & time ()
 Return the current value of continuous time. More...
 
double time () const
 Return the current value of continuous time (const version) More...
 
TimeStepper *& time_stepper_pt ()
 
const TimeStepper * time_stepper_pt () const
 
TimeStepper *& time_stepper_pt (const unsigned &i)
 Return a pointer to the i-th timestepper. More...
 
ExplicitTimeStepper *& explicit_time_stepper_pt ()
 Return a pointer to the explicit timestepper. More...
 
unsigned long set_timestepper_for_all_data (TimeStepper *const &time_stepper_pt, const bool &preserve_existing_data=false)
 
virtual void shift_time_values ()
 Shift all values along to prepare for next timestep. More...
 
AssemblyHandler *& assembly_handler_pt ()
 Return a pointer to the assembly handler object. More...
 
AssemblyHandler *const & assembly_handler_pt () const
 Return a pointer to the assembly handler object (const version) More...
 
double & minimum_dt ()
 Access function to min timestep in adaptive timestepping. More...
 
double & maximum_dt ()
 Access function to max timestep in adaptive timestepping. More...
 
unsigned & max_newton_iterations ()
 Access function to max Newton iterations before giving up. More...
 
void problem_is_nonlinear (const bool &prob_lin)
 Access function to Problem_is_nonlinear. More...
 
double & max_residuals ()
 
bool & time_adaptive_newton_crash_on_solve_fail ()
 Access function for Time_adaptive_newton_crash_on_solve_fail. More...
 
double & newton_solver_tolerance ()
 
void add_time_stepper_pt (TimeStepper *const &time_stepper_pt)
 
void set_explicit_time_stepper_pt (ExplicitTimeStepper *const &explicit_time_stepper_pt)
 
void initialise_dt (const double &dt)
 
void initialise_dt (const Vector< double > &dt)
 
Data *& global_data_pt (const unsigned &i)
 Return a pointer to the the i-th global data object. More...
 
void add_global_data (Data *const &global_data_pt)
 
void flush_global_data ()
 
LinearAlgebraDistribution *const & dof_distribution_pt () const
 Return the pointer to the dof distribution (read-only) More...
 
unsigned long ndof () const
 Return the number of dofs. More...
 
unsigned ntime_stepper () const
 Return the number of time steppers. More...
 
unsigned nglobal_data () const
 Return the number of global data values. More...
 
unsigned self_test ()
 Self-test: Check meshes and global data. Return 0 for OK. More...
 
void enable_store_local_dof_pt_in_elements ()
 
void disable_store_local_dof_pt_in_elements ()
 
unsigned long assign_eqn_numbers (const bool &assign_local_eqn_numbers=true)
 
void describe_dofs (std::ostream &out= *(oomph_info.stream_pt())) const
 
void enable_discontinuous_formulation ()
 
void disable_discontinuous_formulation ()
 
void get_dofs (DoubleVector &dofs) const
 
void get_dofs (const unsigned &t, DoubleVector &dofs) const
 Return vector of the t'th history value of all dofs. More...
 
void set_dofs (const DoubleVector &dofs)
 Set the values of the dofs. More...
 
void set_dofs (const unsigned &t, DoubleVector &dofs)
 Set the history values of the dofs. More...
 
void set_dofs (const unsigned &t, Vector< double * > &dof_pt)
 
void add_to_dofs (const double &lambda, const DoubleVector &increment_dofs)
 Add lambda x incremenet_dofs[l] to the l-th dof. More...
 
double * global_dof_pt (const unsigned &i)
 
double & dof (const unsigned &i)
 i-th dof in the problem More...
 
double dof (const unsigned &i) const
 i-th dof in the problem (const version) More...
 
double *& dof_pt (const unsigned &i)
 Pointer to i-th dof in the problem. More...
 
double * dof_pt (const unsigned &i) const
 Pointer to i-th dof in the problem (const version) More...
 
virtual void get_inverse_mass_matrix_times_residuals (DoubleVector &Mres)
 
virtual void get_dvaluesdt (DoubleVector &f)
 
virtual void get_residuals (DoubleVector &residuals)
 Get the total residuals Vector for the problem. More...
 
virtual void get_jacobian (DoubleVector &residuals, DenseDoubleMatrix &jacobian)
 
virtual void get_jacobian (DoubleVector &residuals, CRDoubleMatrix &jacobian)
 
virtual void get_jacobian (DoubleVector &residuals, CCDoubleMatrix &jacobian)
 
virtual void get_jacobian (DoubleVector &residuals, SumOfMatrices &jacobian)
 
void get_fd_jacobian (DoubleVector &residuals, DenseMatrix< double > &jacobian)
 Get the full Jacobian by finite differencing. More...
 
void get_derivative_wrt_global_parameter (double *const &parameter_pt, DoubleVector &result)
 
void get_hessian_vector_products (DoubleVectorWithHaloEntries const &Y, Vector< DoubleVectorWithHaloEntries > const &C, Vector< DoubleVectorWithHaloEntries > &product)
 
void solve_eigenproblem (const unsigned &n_eval, Vector< std::complex< double >> &eigenvalue, Vector< DoubleVector > &eigenvector, const bool &steady=true)
 Solve the eigenproblem. More...
 
void solve_eigenproblem (const unsigned &n_eval, Vector< std::complex< double >> &eigenvalue, const bool &steady=true)
 
virtual void get_eigenproblem_matrices (CRDoubleMatrix &mass_matrix, CRDoubleMatrix &main_matrix, const double &shift=0.0)
 
void assign_eigenvector_to_dofs (DoubleVector &eigenvector)
 Assign the eigenvector passed to the function to the dofs. More...
 
void add_eigenvector_to_dofs (const double &epsilon, const DoubleVector &eigenvector)
 
void store_current_dof_values ()
 Store the current values of the degrees of freedom. More...
 
void restore_dof_values ()
 Restore the stored values of the degrees of freedom. More...
 
void enable_jacobian_reuse ()
 
void disable_jacobian_reuse ()
 Disable recycling of Jacobian in Newton iteration. More...
 
bool jacobian_reuse_is_enabled ()
 Is recycling of Jacobian in Newton iteration enabled? More...
 
bool & use_predictor_values_as_initial_guess ()
 
void newton_solve ()
 Use Newton method to solve the problem. More...
 
void enable_globally_convergent_newton_method ()
 enable globally convergent Newton method More...
 
void disable_globally_convergent_newton_method ()
 disable globally convergent Newton method More...
 
void newton_solve (unsigned const &max_adapt)
 
void steady_newton_solve (unsigned const &max_adapt=0)
 
void copy (Problem *orig_problem_pt)
 
virtual Problem * make_copy ()
 
virtual void read (std::ifstream &restart_file, bool &unsteady_restart)
 
virtual void read (std::ifstream &restart_file)
 
virtual void dump (std::ofstream &dump_file) const
 
void dump (const std::string &dump_file_name) const
 
void delete_all_external_storage ()
 
virtual void symmetrise_eigenfunction_for_adaptive_pitchfork_tracking ()
 
double * bifurcation_parameter_pt () const
 
void get_bifurcation_eigenfunction (Vector< DoubleVector > &eigenfunction)
 
void activate_fold_tracking (double *const &parameter_pt, const bool &block_solve=true)
 
void activate_bifurcation_tracking (double *const &parameter_pt, const DoubleVector &eigenvector, const bool &block_solve=true)
 
void activate_bifurcation_tracking (double *const &parameter_pt, const DoubleVector &eigenvector, const DoubleVector &normalisation, const bool &block_solve=true)
 
void activate_pitchfork_tracking (double *const &parameter_pt, const DoubleVector &symmetry_vector, const bool &block_solve=true)
 
void activate_hopf_tracking (double *const &parameter_pt, const bool &block_solve=true)
 
void activate_hopf_tracking (double *const &parameter_pt, const double &omega, const DoubleVector &null_real, const DoubleVector &null_imag, const bool &block_solve=true)
 
void deactivate_bifurcation_tracking ()
 
void reset_assembly_handler_to_default ()
 Reset the system to the standard non-augemented state. More...
 
double arc_length_step_solve (double *const &parameter_pt, const double &ds, const unsigned &max_adapt=0)
 
double arc_length_step_solve (Data *const &data_pt, const unsigned &data_index, const double &ds, const unsigned &max_adapt=0)
 
void reset_arc_length_parameters ()
 
int & sign_of_jacobian ()
 
void explicit_timestep (const double &dt, const bool &shift_values=true)
 Take an explicit timestep of size dt. More...
 
void unsteady_newton_solve (const double &dt)
 
void unsteady_newton_solve (const double &dt, const bool &shift_values)
 
void unsteady_newton_solve (const double &dt, const unsigned &max_adapt, const bool &first, const bool &shift=true)
 
double doubly_adaptive_unsteady_newton_solve (const double &dt, const double &epsilon, const unsigned &max_adapt, const bool &first, const bool &shift=true)
 
double doubly_adaptive_unsteady_newton_solve (const double &dt, const double &epsilon, const unsigned &max_adapt, const unsigned &suppress_resolve_after_spatial_adapt_flag, const bool &first, const bool &shift=true)
 
double adaptive_unsteady_newton_solve (const double &dt_desired, const double &epsilon)
 
double adaptive_unsteady_newton_solve (const double &dt_desired, const double &epsilon, const bool &shift_values)
 
void assign_initial_values_impulsive ()
 
void assign_initial_values_impulsive (const double &dt)
 
void calculate_predictions ()
 Calculate predictions. More...
 
void enable_mass_matrix_reuse ()
 
void disable_mass_matrix_reuse ()
 
bool mass_matrix_reuse_is_enabled ()
 Return whether the mass matrix is being reused. More...
 
void refine_uniformly (const Vector< unsigned > &nrefine_for_mesh)
 
void refine_uniformly (const Vector< unsigned > &nrefine_for_mesh, DocInfo &doc_info)
 
void refine_uniformly_and_prune (const Vector< unsigned > &nrefine_for_mesh)
 
void refine_uniformly_and_prune (const Vector< unsigned > &nrefine_for_mesh, DocInfo &doc_info)
 
void refine_uniformly (DocInfo &doc_info)
 
void refine_uniformly_and_prune (DocInfo &doc_info)
 
void refine_uniformly ()
 
void refine_uniformly (const unsigned &i_mesh, DocInfo &doc_info)
 Do uniform refinement for submesh i_mesh with documentation. More...
 
void refine_uniformly (const unsigned &i_mesh)
 Do uniform refinement for submesh i_mesh without documentation. More...
 
void p_refine_uniformly (const Vector< unsigned > &nrefine_for_mesh)
 
void p_refine_uniformly (const Vector< unsigned > &nrefine_for_mesh, DocInfo &doc_info)
 
void p_refine_uniformly_and_prune (const Vector< unsigned > &nrefine_for_mesh)
 
void p_refine_uniformly_and_prune (const Vector< unsigned > &nrefine_for_mesh, DocInfo &doc_info)
 
void p_refine_uniformly (DocInfo &doc_info)
 
void p_refine_uniformly_and_prune (DocInfo &doc_info)
 
void p_refine_uniformly ()
 
void p_refine_uniformly (const unsigned &i_mesh, DocInfo &doc_info)
 Do uniform p-refinement for submesh i_mesh with documentation. More...
 
void p_refine_uniformly (const unsigned &i_mesh)
 Do uniform p-refinement for submesh i_mesh without documentation. More...
 
void refine_selected_elements (const Vector< unsigned > &elements_to_be_refined)
 
void refine_selected_elements (const Vector< RefineableElement * > &elements_to_be_refined_pt)
 
void refine_selected_elements (const unsigned &i_mesh, const Vector< unsigned > &elements_to_be_refined)
 
void refine_selected_elements (const unsigned &i_mesh, const Vector< RefineableElement * > &elements_to_be_refined_pt)
 
void refine_selected_elements (const Vector< Vector< unsigned >> &elements_to_be_refined)
 
void refine_selected_elements (const Vector< Vector< RefineableElement * >> &elements_to_be_refined_pt)
 
void p_refine_selected_elements (const Vector< unsigned > &elements_to_be_refined)
 
void p_refine_selected_elements (const Vector< PRefineableElement * > &elements_to_be_refined_pt)
 
void p_refine_selected_elements (const unsigned &i_mesh, const Vector< unsigned > &elements_to_be_refined)
 
void p_refine_selected_elements (const unsigned &i_mesh, const Vector< PRefineableElement * > &elements_to_be_refined_pt)
 
void p_refine_selected_elements (const Vector< Vector< unsigned >> &elements_to_be_refined)
 
void p_refine_selected_elements (const Vector< Vector< PRefineableElement * >> &elements_to_be_refined_pt)
 
unsigned unrefine_uniformly ()
 
unsigned unrefine_uniformly (const unsigned &i_mesh)
 
void p_unrefine_uniformly (DocInfo &doc_info)
 
void p_unrefine_uniformly (const unsigned &i_mesh, DocInfo &doc_info)
 Do uniform p-unrefinement for submesh i_mesh without documentation. More...
 
void adapt (unsigned &n_refined, unsigned &n_unrefined)
 
void adapt ()
 
void p_adapt (unsigned &n_refined, unsigned &n_unrefined)
 
void p_adapt ()
 
void adapt_based_on_error_estimates (unsigned &n_refined, unsigned &n_unrefined, Vector< Vector< double >> &elemental_error)
 
void adapt_based_on_error_estimates (Vector< Vector< double >> &elemental_error)
 
void get_all_error_estimates (Vector< Vector< double >> &elemental_error)
 
void doc_errors (DocInfo &doc_info)
 Get max and min error for all elements in submeshes. More...
 
void doc_errors ()
 Get max and min error for all elements in submeshes. More...
 
void enable_info_in_newton_solve ()
 
void disable_info_in_newton_solve ()
 Disable the output of information when in the newton solver. More...
 
- Public Member Functions inherited from oomph::ExplicitTimeSteppableObject
 ExplicitTimeSteppableObject ()
 Empty constructor. More...
 
 ExplicitTimeSteppableObject (const ExplicitTimeSteppableObject &)=delete
 Broken copy constructor. More...
 
void operator= (const ExplicitTimeSteppableObject &)=delete
 Broken assignment operator. More...
 
virtual ~ExplicitTimeSteppableObject ()
 Empty destructor. More...
 
virtual void actions_before_explicit_stage ()
 
virtual void actions_after_explicit_stage ()
 

Private Member Functions

void create_flux_elements (const unsigned &b)
 Create Poisson Flux Elements on the boundary 1. More...
 
void delete_flux_elements ()
 Delete Poisson flux elements and wipe the surface mesh. More...
 
void set_multigrid_solver ()
 Build and set multgrid solver. More...
 
TreeBasedRefineableMeshBase * mg_bulk_mesh_pt ()
 
MGProblem * make_new_problem ()
 Return a pointer to a new instance of the same problem. More...
 
MESH * mesh_pt ()
 Access function for the mesh. More...
 
MGProblem * make_new_problem ()
 Return a pointer to a new instance of the same problem. More...
 
void create_flux_elements (const unsigned &b, MESH *const &bulk_mesh_pt, Mesh *const &surface_mesh_pt)
 Create Poisson Flux Elements on the b-th boundary of the Mesh. More...
 
void delete_flux_elements (Mesh *const &surface_mesh_pt)
 Delete Poisson flux elements and wipe the surface mesh. More...
 

Private Attributes

MESH * Bulk_mesh_pt
 Pointer to the "bulk" mesh. More...
 
Mesh * Surface_mesh_pt
 Pointer to the "surface" mesh. More...
 
PoissonEquations< 2 >::PoissonSourceFctPt Source_fct_pt
 Pointer to source function. More...
 

Additional Inherited Members

- Public Types inherited from oomph::Problem
typedef void(* SpatialErrorEstimatorFctPt) (Mesh *&mesh_pt, Vector< double > &elemental_error)
 Function pointer for spatial error estimator. More...
 
typedef void(* SpatialErrorEstimatorWithDocFctPt) (Mesh *&mesh_pt, Vector< double > &elemental_error, DocInfo &doc_info)
 Function pointer for spatial error estimator with doc. More...
 
- Public Attributes inherited from oomph::Problem
bool Shut_up_in_newton_solve
 
- Static Public Attributes inherited from oomph::Problem
static bool Suppress_warning_about_actions_before_read_unstructured_meshes
 
- Protected Types inherited from oomph::Problem
enum  Assembly_method {
  Perform_assembly_using_vectors_of_pairs , Perform_assembly_using_two_vectors , Perform_assembly_using_maps , Perform_assembly_using_lists ,
  Perform_assembly_using_two_arrays
}
 Enumerated flags to determine which sparse assembly method is used. More...
 
- Protected Member Functions inherited from oomph::Problem
unsigned setup_element_count_per_dof ()
 
virtual void sparse_assemble_row_or_column_compressed (Vector< int * > &column_or_row_index, Vector< int * > &row_or_column_start, Vector< double * > &value, Vector< unsigned > &nnz, Vector< double * > &residual, bool compressed_row_flag)
 
virtual void actions_before_newton_convergence_check ()
 
virtual void actions_before_newton_step ()
 
virtual void actions_after_newton_step ()
 
virtual void actions_before_implicit_timestep ()
 
virtual void actions_after_implicit_timestep ()
 
virtual void actions_after_implicit_timestep_and_error_estimation ()
 
virtual void actions_before_explicit_timestep ()
 Actions that should be performed before each explicit time step. More...
 
virtual void actions_after_explicit_timestep ()
 Actions that should be performed after each explicit time step. More...
 
virtual void actions_before_read_unstructured_meshes ()
 
virtual void actions_after_read_unstructured_meshes ()
 
virtual void actions_after_change_in_global_parameter (double *const &parameter_pt)
 
virtual void actions_after_change_in_bifurcation_parameter ()
 
virtual void actions_after_parameter_increase (double *const &parameter_pt)
 
double & dof_derivative (const unsigned &i)
 
double & dof_current (const unsigned &i)
 
virtual void set_initial_condition ()
 
virtual double global_temporal_error_norm ()
 
unsigned newton_solve_continuation (double *const &parameter_pt)
 
unsigned newton_solve_continuation (double *const &parameter_pt, DoubleVector &z)
 
void calculate_continuation_derivatives (double *const &parameter_pt)
 
void calculate_continuation_derivatives (const DoubleVector &z)
 
void calculate_continuation_derivatives_fd (double *const &parameter_pt)
 
bool does_pointer_correspond_to_problem_data (double *const &parameter_pt)
 
void set_consistent_pinned_values_for_continuation ()
 
- Protected Attributes inherited from oomph::Problem
Vector< Problem * > Copy_of_problem_pt
 
std::map< double *, bool > Calculate_dparameter_analytic
 
bool Calculate_hessian_products_analytic
 
LinearAlgebraDistribution * Dof_distribution_pt
 
Vector< double * > Dof_pt
 Vector of pointers to dofs. More...
 
DoubleVectorWithHaloEntries Element_count_per_dof
 
double Relaxation_factor
 
double Newton_solver_tolerance
 
unsigned Max_newton_iterations
 Maximum number of Newton iterations. More...
 
unsigned Nnewton_iter_taken
 
Vector< double > Max_res
 Maximum residuals at start and after each newton iteration. More...
 
double Max_residuals
 
bool Time_adaptive_newton_crash_on_solve_fail
 
bool Jacobian_reuse_is_enabled
 Is re-use of Jacobian in Newton iteration enabled? Default: false. More...
 
bool Jacobian_has_been_computed
 
bool Problem_is_nonlinear
 
bool Pause_at_end_of_sparse_assembly
 
bool Doc_time_in_distribute
 
unsigned Sparse_assembly_method
 
unsigned Sparse_assemble_with_arrays_initial_allocation
 
unsigned Sparse_assemble_with_arrays_allocation_increment
 
Vector< Vector< unsigned > > Sparse_assemble_with_arrays_previous_allocation
 
double Numerical_zero_for_sparse_assembly
 
double FD_step_used_in_get_hessian_vector_products
 
bool Mass_matrix_reuse_is_enabled
 
bool Mass_matrix_has_been_computed
 
bool Discontinuous_element_formulation
 
double Minimum_dt
 Minimum desired dt: if dt falls below this value, exit. More...
 
double Maximum_dt
 Maximum desired dt. More...
 
double DTSF_max_increase
 
double DTSF_min_decrease
 
double Minimum_dt_but_still_proceed
 
bool Scale_arc_length
 Boolean to control whether arc-length should be scaled. More...
 
double Desired_proportion_of_arc_length
 Proportion of the arc-length to taken by the parameter. More...
 
double Theta_squared
 
int Sign_of_jacobian
 Storage for the sign of the global Jacobian. More...
 
double Continuation_direction
 
double Parameter_derivative
 Storage for the derivative of the global parameter wrt arc-length. More...
 
double Parameter_current
 Storage for the present value of the global parameter. More...
 
bool Use_continuation_timestepper
 Boolean to control original or new storage of dof stuff. More...
 
unsigned Dof_derivative_offset
 
unsigned Dof_current_offset
 
Vector< double > Dof_derivative
 Storage for the derivative of the problem variables wrt arc-length. More...
 
Vector< double > Dof_current
 Storage for the present values of the variables. More...
 
double Ds_current
 Storage for the current step value. More...
 
unsigned Desired_newton_iterations_ds
 
double Minimum_ds
 Minimum desired value of arc-length. More...
 
bool Bifurcation_detection
 Boolean to control bifurcation detection via determinant of Jacobian. More...
 
bool Bisect_to_find_bifurcation
 Boolean to control wheter bisection is used to located bifurcation. More...
 
bool First_jacobian_sign_change
 Boolean to indicate whether a sign change has occured in the Jacobian. More...
 
bool Arc_length_step_taken
 Boolean to indicate whether an arc-length step has been taken. More...
 
bool Use_finite_differences_for_continuation_derivatives
 
OomphCommunicator * Communicator_pt
 The communicator for this problem. More...
 
bool Always_take_one_newton_step
 
double Timestep_reduction_factor_after_nonconvergence
 
bool Keep_temporal_error_below_tolerance
 
- Static Protected Attributes inherited from oomph::Problem
static ContinuationStorageScheme Continuation_time_stepper
 Storage for the single static continuation timestorage object. More...
 

Detailed Description

template<class ELEMENT, class MESH>
class FluxPoissonMGProblem< ELEMENT, MESH >

/////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////////////////////// 2D Poisson problem on rectangular domain, discretised with 2D QPoisson elements. Flux boundary conditions are applied along boundary 1 (the boundary where x=1). The specific type of element and mesh is specified via the template parameter.

/////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////////////////////// 2D Poisson problem on rectangular domain, discretised with 2D QPoisson elements. Flux boundary conditions are applied along boundary 1 (the boundary where x=L). The specific type of element and mesh is specified via the template parameter.

Constructor & Destructor Documentation

◆ FluxPoissonMGProblem() [1/2]

template<class ELEMENT , class MESH >
FluxPoissonMGProblem< ELEMENT, MESH >::FluxPoissonMGProblem ( PoissonEquations< 2 >::PoissonSourceFctPt  source_fct_pt)

Constructor: Pass a pointer to the source function.

Constructor for Poisson problem: Pass pointer to source function.

188  : Source_fct_pt(source_fct_pt)
189 {
190  // Set up the multigrid solver
191  set_multigrid_solver();
192 
193  // Build the mesh:
194  //----------------
195  // Number of elements in the x-direction
196  unsigned n_x=4;
197 
198  // Number of elements in the y-direction
199  unsigned n_y=4;
200 
201  // Domain length in the x-direction
202  double l_x=1.0;
203 
204  // Domain length in the y-direction
205  double l_y=2.0;
206 
207  // Build the "bulk" mesh
208  Bulk_mesh_pt=new MESH(n_x,n_y,l_x,l_y);
209 
210  // Create/set error estimator
211  Bulk_mesh_pt->spatial_error_estimator_pt()=new Z2ErrorEstimator;
212 
213  // Create "surface mesh" that will contain only the prescribed-flux
214  // elements. The constructor just creates the mesh without giving it
215  // any elements, nodes, etc.
216  Surface_mesh_pt=new Mesh;
217 
218  // Set boundary conditions:
219  //-------------------------
220  // Create prescribed-flux elements from all elements that are adjacent to
221  // boundary 1, but add them to a separate mesh
222  create_flux_elements(1);
223 
224  // Find out how many boundaries the "bulk" mesh possesses
225  unsigned n_bound=Bulk_mesh_pt->nboundary();
226 
227  // Loop over the boundaries of the "bulk" mesh
228  for(unsigned b=0;b<n_bound;b++)
229  {
230  // Leave nodes on boundary 1 free
231  if(b!=1)
232  {
233  // Find the number of nodes on the b-th boundary
234  unsigned n_node=Bulk_mesh_pt->nboundary_node(b);
235 
236  // Loop over the boundary nodes
237  for (unsigned n=0;n<n_node;n++)
238  {
239  // All nodes are free by default -- we just need to pin the ones that
240  // have Dirichlet conditions here
241  Bulk_mesh_pt->boundary_node_pt(b,n)->pin(0);
242  }
243  } // if(b!=1)
244  } // for(unsigned b=0;b<n_bound;b++)
245 
246  // Build the global mesh from its submeshes:
247  //------------------------------------------
248  // Add the "bulk" mesh to the Problem's submeshes
249  add_sub_mesh(Bulk_mesh_pt);
250 
251  // Add the mesh containing flux elements to the Problem's submeshes
252  add_sub_mesh(Surface_mesh_pt);
253 
254  // Rebuild the Problem's global mesh from its various sub-meshes
255  build_global_mesh();
256 
257  // Complete the build of all elements so they are fully functional:
258  //-----------------------------------------------------------------
259  // Loop over the "bulk" mesh elements and pass in the source function pointer
260  for(unsigned e=0;e<Bulk_mesh_pt->nelement();e++)
261  {
262  // Upcast from GeneralisedElement to the element type specified by the
263  // template argument
264  ELEMENT* el_pt=dynamic_cast<ELEMENT*>(Bulk_mesh_pt->element_pt(e));
265 
266  // Set the source function pointer
267  el_pt->source_fct_pt()=Source_fct_pt;
268  }
269 
270  // Set up the equation numbering scheme
271  assign_eqn_numbers();
272 } // End of FluxPoissonMGProblem constructor
n
const unsigned n
Definition: CG3DPackingUnitTest.cpp:11
e
Array< double, 1, 3 > e(1./3., 0.5, 2.)
b
Scalar * b
Definition: benchVecAdd.cpp:17
FluxPoissonMGProblem::Surface_mesh_pt
Mesh * Surface_mesh_pt
Pointer to the "surface" mesh.
Definition: two_d_poisson_tanh_flux_bc.cc:176
FluxPoissonMGProblem::create_flux_elements
void create_flux_elements(const unsigned &b)
Create Poisson Flux Elements on the boundary 1.
Definition: two_d_poisson_tanh_flux_bc.cc:388
FluxPoissonMGProblem::set_multigrid_solver
void set_multigrid_solver()
Build and set multgrid solver.
Definition: two_d_poisson_tanh_flux_bc.cc:450
FluxPoissonMGProblem::Source_fct_pt
PoissonEquations< 2 >::PoissonSourceFctPt Source_fct_pt
Pointer to source function.
Definition: two_d_poisson_tanh_flux_bc.cc:179
FluxPoissonMGProblem::Bulk_mesh_pt
MESH * Bulk_mesh_pt
Pointer to the "bulk" mesh.
Definition: two_d_poisson_tanh_flux_bc.cc:173
oomph::Mesh
Definition: mesh.h:67
oomph::Problem::add_sub_mesh
unsigned add_sub_mesh(Mesh *const &mesh_pt)
Definition: problem.h:1330
oomph::Problem::build_global_mesh
void build_global_mesh()
Definition: problem.cc:1493
oomph::Problem::assign_eqn_numbers
unsigned long assign_eqn_numbers(const bool &assign_local_eqn_numbers=true)
Definition: problem.cc:1989
oomph::Z2ErrorEstimator
Definition: error_estimator.h:266
oomph::source_fct_pt
GeneralisedAxisymAdvectionDiffusionSourceFctPt & source_fct_pt()
Access function: Pointer to source function.
Definition: gen_axisym_advection_diffusion_elements.h:229

References oomph::Problem::add_sub_mesh(), oomph::Problem::assign_eqn_numbers(), b, oomph::Problem::build_global_mesh(), FluxPoissonMGProblem< ELEMENT, MESH >::Bulk_mesh_pt, FluxPoissonMGProblem< ELEMENT, MESH >::create_flux_elements(), e(), n, FluxPoissonMGProblem< ELEMENT, MESH >::set_multigrid_solver(), FluxPoissonMGProblem< ELEMENT, MESH >::Source_fct_pt, and FluxPoissonMGProblem< ELEMENT, MESH >::Surface_mesh_pt.

◆ ~FluxPoissonMGProblem() [1/2]

template<class ELEMENT , class MESH >
FluxPoissonMGProblem< ELEMENT, MESH >::~FluxPoissonMGProblem

Destructor.

Destructor for Poisson problem.

279 {
280  // Delete the MG solver pointer
281  delete linear_solver_pt();
282 
283  // Set the MG solver pointer to null
284  linear_solver_pt()=0;
285 
286  // Delete the error estimator
287  delete Bulk_mesh_pt->spatial_error_estimator_pt();
288 
289  // Set the error estimator pointer to null
290  Bulk_mesh_pt->spatial_error_estimator_pt()=0;
291 
292  // Delete the "bulk" mesh
293  delete Bulk_mesh_pt;
294 
295  // Set the "bulk" mesh pointer to null
296  Bulk_mesh_pt=0;
297 
298  // Delete the mesh containing flux elements
299  delete Surface_mesh_pt;
300 
301  // Set the surface mesh pointer to null
302  Surface_mesh_pt=0;
303 } // End of destructor
oomph::Problem::linear_solver_pt
LinearSolver *& linear_solver_pt()
Return a pointer to the linear solver object.
Definition: problem.h:1466

◆ FluxPoissonMGProblem() [2/2]

template<class ELEMENT , class MESH >
FluxPoissonMGProblem< ELEMENT, MESH >::FluxPoissonMGProblem ( PoissonEquations< 2 >::PoissonSourceFctPt  source_fct_pt)

Constructor: Pass pointer to source function.

◆ ~FluxPoissonMGProblem() [2/2]

template<class ELEMENT , class MESH >
FluxPoissonMGProblem< ELEMENT, MESH >::~FluxPoissonMGProblem ( )

Destructor.

Member Function Documentation

◆ actions_after_adapt() [1/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::actions_after_adapt
virtual

Actions after adapt: Rebuild the mesh of prescribed flux elements.

Reimplemented from oomph::Problem.

329 {
330  // Create prescribed-flux elements from all elements that are
331  // adjacent to boundary 1 and add them to surfac mesh
332  create_flux_elements(1);
333 
334  // Add the flux mesh back into the global mesh
335  add_sub_mesh(Surface_mesh_pt);
336 
337  // Rebuild the Problem's global mesh from its various sub-meshes
338  rebuild_global_mesh();
339 } // End of actions_after_adapt
oomph::Problem::rebuild_global_mesh
void rebuild_global_mesh()
Definition: problem.cc:1533

◆ actions_after_adapt() [2/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::actions_after_adapt ( )
virtual

Actions after adapt: Rebuild the mesh of prescribed flux elements.

Reimplemented from oomph::Problem.

◆ actions_after_newton_solve() [1/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::actions_after_newton_solve ( )
inlinevirtual

Update the problem specs after solve (empty)

Reimplemented from oomph::Problem.

135 {}

◆ actions_after_newton_solve() [2/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::actions_after_newton_solve ( )
inlinevirtual

Update the problem specs after solve (empty)

Reimplemented from oomph::Problem.

255  {
256  // Document the solution
257  doc_solution();
258  }
FluxPoissonMGProblem::doc_solution
void doc_solution()
Doc the solution: doc_info contains labels/output directory etc.
Definition: two_d_poisson_tanh_flux_bc_validate.cc:625

◆ actions_before_adapt() [1/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::actions_before_adapt
virtual

Actions before adapt: Wipe the mesh of prescribed flux elements.

Reimplemented from oomph::Problem.

310 {
311  // Kill the flux elements and wipe surface mesh
312  delete_flux_elements();
313 
314  // Flush the submeshes
315  flush_sub_meshes();
316 
317  // Add the bulk mesh back in
318  add_sub_mesh(Bulk_mesh_pt);
319 
320  // Rebuild the Problem's global mesh from its various sub-meshes
321  rebuild_global_mesh();
322 } // End of actions_before_adapt
FluxPoissonMGProblem::delete_flux_elements
void delete_flux_elements()
Delete Poisson flux elements and wipe the surface mesh.
Definition: two_d_poisson_tanh_flux_bc.cc:430
oomph::Problem::flush_sub_meshes
void flush_sub_meshes()
Definition: problem.h:1339

◆ actions_before_adapt() [2/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::actions_before_adapt ( )
virtual

Actions before adapt: Wipe the mesh of prescribed flux elements.

Reimplemented from oomph::Problem.

◆ actions_before_newton_solve() [1/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::actions_before_newton_solve
virtual

Update the problem specs before solve: Reset boundary conditions to the values from the exact solution.

Reimplemented from oomph::Problem.

347 {
348  // How many boundaries are there?
349  unsigned n_bound=Bulk_mesh_pt->nboundary();
350 
351  // Loop over the boundaries
352  for(unsigned i=0;i<n_bound;i++)
353  {
354  // Only update Dirichlet nodes
355  if (i!=1)
356  {
357  // How many nodes are there on this boundary?
358  unsigned n_node=Bulk_mesh_pt->nboundary_node(i);
359 
360  // Loop over the nodes on boundary
361  for (unsigned n=0;n<n_node;n++)
362  {
363  // Get pointer to node
364  Node* nod_pt=Bulk_mesh_pt->boundary_node_pt(i,n);
365 
366  // Extract nodal coordinates from node:
367  Vector<double> x(2);
368  x[0]=nod_pt->x(0);
369  x[1]=nod_pt->x(1);
370 
371  // Create storage for the solution
372  Vector<double> u(1);
373 
374  // Compute the value of the exact solution at the nodal point
375  TanhSolnForPoisson::get_exact_u(x,u);
376 
377  // Assign the value to the one (and only) nodal value at this node
378  nod_pt->set_value(0,u[0]);
379  }
380  } // if (i!=1)
381  } // for(unsigned i=0;i<n_bound;i++)
382 } // End of actions_before_newton_solve
i
int i
Definition: BiCGSTAB_step_by_step.cpp:9
oomph::Data::set_value
void set_value(const unsigned &i, const double &value_)
Definition: nodes.h:271
oomph::Node
Definition: nodes.h:906
oomph::Node::x
double & x(const unsigned &i)
Return the i-th nodal coordinate.
Definition: nodes.h:1060
TanhSolnForPoisson::get_exact_u
void get_exact_u(const Vector< double > &x, Vector< double > &u)
Exact solution as a Vector.
Definition: extrude_with_macro_element_representation.cc:206
plotDoE.x
list x
Definition: plotDoE.py:28

References TanhSolnForPoisson::get_exact_u(), i, n, oomph::Data::set_value(), plotDoE::x, and oomph::Node::x().

◆ actions_before_newton_solve() [2/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::actions_before_newton_solve ( )
virtual

Update the problem specs before solve: Reset boundary conditions to the values from the exact solution.

Reimplemented from oomph::Problem.

◆ create_flux_elements() [1/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::create_flux_elements ( const unsigned &  b)
private

Create Poisson Flux Elements on the boundary 1.

Create Poisson flux elements on the b-th boundary of the problem's mesh

389 {
390  // How many bulk elements are adjacent to boundary b?
391  unsigned n_belement=Bulk_mesh_pt->nboundary_element(b);
392 
393  // Loop over the bulk elements adjacent to boundary b?
394  for(unsigned e=0;e<n_belement;e++)
395  {
396  // Get pointer to the bulk element that is adjacent to boundary b
397  ELEMENT* bulk_elem_pt=dynamic_cast<ELEMENT*>
398  (Bulk_mesh_pt->boundary_element_pt(b,e));
399 
400  // Find the index of the face of the bulk element at the boundary
401  int face_index=Bulk_mesh_pt->face_index_at_boundary(b,e);
402 
403  // Build the corresponding prescribed-flux element
404  PoissonFluxElement<ELEMENT>* flux_element_pt=
405  new PoissonFluxElement<ELEMENT>(bulk_elem_pt,face_index);
406 
407  // Add the prescribed flux element to the mesh
408  Surface_mesh_pt->add_element_pt(flux_element_pt);
409  }
410 
411  // Find out how many flux elements there are on boundary 1
412  unsigned n_selement=Surface_mesh_pt->nelement();
413 
414  // Loop over the flux elements to pass pointer to prescribed flux function
415  for(unsigned e=0;e<n_selement;e++)
416  {
417  // Upcast from GeneralisedElement to Poisson flux element
418  PoissonFluxElement<ELEMENT>* el_pt=
419  dynamic_cast< PoissonFluxElement<ELEMENT>*>(Surface_mesh_pt->element_pt(e));
420 
421  // Set the pointer to the prescribed flux function
422  el_pt->flux_fct_pt()=&TanhSolnForPoisson::prescribed_flux_on_fixed_x_boundary;
423  }
424 } // End of create_flux_elements
oomph::Mesh::element_pt
GeneralisedElement *& element_pt(const unsigned long &e)
Return pointer to element e.
Definition: mesh.h:448
oomph::Mesh::add_element_pt
void add_element_pt(GeneralisedElement *const &element_pt)
Add a (pointer to) an element to the mesh.
Definition: mesh.h:617
oomph::Mesh::nelement
unsigned long nelement() const
Return number of elements in the mesh.
Definition: mesh.h:590
oomph::PoissonFluxElement
Definition: poisson_flux_elements.h:51
oomph::PoissonFluxElement::flux_fct_pt
PoissonPrescribedFluxFctPt & flux_fct_pt()
Access function for the prescribed-flux function pointer.
Definition: poisson_flux_elements.h:89
TanhSolnForPoisson::prescribed_flux_on_fixed_x_boundary
void prescribed_flux_on_fixed_x_boundary(const Vector< double > &x, double &flux)
Flux required by the exact solution on a boundary on which x is fixed.
Definition: mpi/distribution/two_d_poisson_flux_bc_adapt/two_d_poisson_flux_bc_adapt.cc:120

References b, e(), oomph::PoissonFluxElement< ELEMENT >::flux_fct_pt(), and TanhSolnForPoisson::prescribed_flux_on_fixed_x_boundary().

Referenced by FluxPoissonMGProblem< ELEMENT, MESH >::FluxPoissonMGProblem().

◆ create_flux_elements() [2/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::create_flux_elements ( const unsigned &  b,
MESH *const &  bulk_mesh_pt,
Mesh *const &  surface_mesh_pt 
)
private

Create Poisson Flux Elements on the b-th boundary of the Mesh.

Create Poisson flux elements on the b-th boundary of the problem's mesh

487 {
488  // How many bulk elements are adjacent to boundary b?
489  unsigned n_belement=bulk_mesh_pt->nboundary_element(b);
490 
491  // Loop over the bulk elements adjacent to boundary b?
492  for(unsigned e=0;e<n_belement;e++)
493  {
494  // Get pointer to the bulk element that is adjacent to boundary b
495  ELEMENT* bulk_elem_pt = dynamic_cast<ELEMENT*>
496  (bulk_mesh_pt->boundary_element_pt(b,e));
497 
498  // Find the index of the face of the bulk element at the boundary
499  int face_index = bulk_mesh_pt->face_index_at_boundary(b,e);
500 
501  // Build the corresponding prescribed-flux element
502  PoissonFluxElement<ELEMENT>* flux_element_pt = new
503  PoissonFluxElement<ELEMENT>(bulk_elem_pt,face_index);
504 
505  // Add the prescribed-flux element to the mesh
506  surface_mesh_pt->add_element_pt(flux_element_pt);
507  } // end of loop over bulk elements adjacent to boundary b
508 
509  // Loop over the flux elements to pass pointer to prescribed flux function
510  unsigned n_selement=Surface_mesh_pt->nelement();
511  for(unsigned e=0;e<n_selement;e++)
512  {
513  // Upcast from GeneralisedElement to Poisson flux element
514  PoissonFluxElement<ELEMENT> *el_pt =
515  dynamic_cast< PoissonFluxElement<ELEMENT>*>(Surface_mesh_pt->element_pt(e));
516 
517  // Set the pointer to the prescribed flux function
518  el_pt->flux_fct_pt()=&SolnForPoisson::prescribed_flux_on_fixed_x_boundary;
519  }
520 } // End of create_flux_elements
SolnForPoisson::prescribed_flux_on_fixed_x_boundary
void prescribed_flux_on_fixed_x_boundary(const Vector< double > &x, double &flux)
Flux required by the exact solution on a boundary on which x is fixed.
Definition: two_d_poisson_tanh_flux_bc_validate.cc:195

References oomph::Mesh::add_element_pt(), b, e(), oomph::PoissonFluxElement< ELEMENT >::flux_fct_pt(), and SolnForPoisson::prescribed_flux_on_fixed_x_boundary().

◆ delete_flux_elements() [1/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::delete_flux_elements
private

Delete Poisson flux elements and wipe the surface mesh.

431 {
432  // Find out how many surface elements there are in the surface mesh
433  unsigned n_element=Surface_mesh_pt->nelement();
434 
435  // Loop over the surface elements
436  for(unsigned e=0;e<n_element;e++)
437  {
438  // Kill the e-th surface element
439  delete Surface_mesh_pt->element_pt(e);
440  }
441 
442  // Wipe the mesh
443  Surface_mesh_pt->flush_element_and_node_storage();
444 } // End of delete_flux_elements
oomph::Mesh::flush_element_and_node_storage
void flush_element_and_node_storage()
Definition: mesh.h:407

References e().

◆ delete_flux_elements() [2/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::delete_flux_elements ( Mesh *const &  surface_mesh_pt)
private

Delete Poisson flux elements and wipe the surface mesh.

Delete Poisson Flux Elements and wipe the surface mesh.

529 {
530  // Find out how many surface elements there are in the surface mesh
531  unsigned n_element=surface_mesh_pt->nelement();
532 
533  // Loop over the surface elements
534  for(unsigned e=0;e<n_element;e++)
535  {
536  // Kill surface element
537  delete surface_mesh_pt->element_pt(e);
538  }
539 
540  // Wipe the mesh
541  surface_mesh_pt->flush_element_and_node_storage();
542 } // End of delete_flux_elements

References e(), oomph::Mesh::element_pt(), oomph::Mesh::flush_element_and_node_storage(), and oomph::Mesh::nelement().

◆ doc_solution() [1/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::doc_solution

Doc the solution: doc_info contains labels/output directory etc.

Doc the solution. DocInfo object stores flags/labels for where the output gets written to

626 {
627  // Output file stream
628  ofstream some_file;
629  char filename[100];
630 
631  // Number of plot points
632  unsigned npts;
633  npts=5;
634 
635  // Output solution
636  //-----------------
637  sprintf(filename,"%s/soln%i.dat",
638  Global_Parameters::Doc_info.directory().c_str(),
639  Global_Parameters::Doc_info.number());
640  some_file.open(filename);
641  Bulk_mesh_pt->output(some_file,npts);
642  some_file.close();
643 
644  // Output exact solution
645  //----------------------
646  sprintf(filename,"%s/exact_soln%i.dat",
647  Global_Parameters::Doc_info.directory().c_str(),
648  Global_Parameters::Doc_info.number());
649  some_file.open(filename);
650  Bulk_mesh_pt->output_fct(some_file,npts,SolnForPoisson::get_exact_u);
651  some_file.close();
652 
653  // Doc error and return of the square of the L2 error. Note:
654  //----------------------------------------------------------
655  // this is only calculated over the bulk mesh since we cannot
656  //-----------------------------------------------------------
657  // properly deal with two different types of elements
658  //---------------------------------------------------
659  double error,norm;
660  sprintf(filename,"%s/error%i.dat",
661  Global_Parameters::Doc_info.directory().c_str(),
662  Global_Parameters::Doc_info.number());
663  some_file.open(filename);
664  Bulk_mesh_pt->compute_error(some_file,SolnForPoisson::get_exact_u,error,norm);
665  some_file.close();
666 
667  // Document the L2 norm of the error and the L2 norm of the solution
668  oomph_info << "Norm of error : " << sqrt(error)
669  << "\nNorm of solution: " << sqrt(norm) << std::endl;
670 
671  // Increment the documentation counter
672  Global_Parameters::Doc_info.number()++;
673 } // End of doc_solution
sqrt
AnnoyingScalar sqrt(const AnnoyingScalar &x)
Definition: AnnoyingScalar.h:134
oomph::DocInfo::number
unsigned & number()
Number used (e.g.) for labeling output files.
Definition: oomph_utilities.h:554
Global_Parameters::Doc_info
DocInfo Doc_info
DocInfo object used for documentation of the solution.
Definition: two_d_poisson_tanh_flux_bc_validate.cc:97
MergeRestartFiles.filename
string filename
Definition: MergeRestartFiles.py:39
SolnForPoisson::get_exact_u
void get_exact_u(const Vector< double > &x, Vector< double > &u)
Exact solution as a Vector.
Definition: two_d_poisson_tanh_flux_bc_validate.cc:180
calibrate.error
int error
Definition: calibrate.py:297
oomph::oomph_info
OomphInfo oomph_info
Definition: oomph_definitions.cc:319

References Global_Parameters::Doc_info, calibrate::error, MergeRestartFiles::filename, SolnForPoisson::get_exact_u(), oomph::DocInfo::number(), oomph::oomph_info, and sqrt().

◆ doc_solution() [2/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::doc_solution ( DocInfo &  doc_info)

Document the solution: doc_info contains labels/output directory, etc.

Doc the solution. DocInfo object stores flags/labels for where the output gets written to

475 {
476  // Output file stream
477  ofstream some_file;
478 
479  // Create an array of characters to store the filename
480  char filename[100];
481 
482  // Number of plot points
483  unsigned npts=3;
484 
485  // Output solution:
486  //-----------------
487  // Create the filename
488  sprintf(filename,"%s/soln%i.dat",
489  doc_info.directory().c_str(),
490  doc_info.number());
491 
492  // Open a file with the filename given above
493  some_file.open(filename);
494 
495  // Document the computed solution
496  Bulk_mesh_pt->output(some_file,npts);
497 
498  // Close the file
499  some_file.close();
500 
501  // Output exact solution:
502  //-----------------------
503  // Create the filename
504  sprintf(filename,"%s/exact_soln%i.dat",
505  doc_info.directory().c_str(),
506  doc_info.number());
507 
508  // Open a file with the filename given above
509  some_file.open(filename);
510 
511  // Document the exact solution
512  Bulk_mesh_pt->output_fct(some_file,npts,TanhSolnForPoisson::get_exact_u);
513 
514  // Close the file
515  some_file.close();
516 
517  // Document the error:
518  //--------------------
519  // Note: this is only calculated over the bulk mesh as we cannot properly
520  // deal with two different types of elements
521 
522  // Create storage for the L2 norm of the error
523  double error=0.0;
524 
525  // Create storage for the L2 norm of the solution
526  double soln=0.0;
527 
528  // Create the filename
529  sprintf(filename,"%s/error%i.dat",
530  doc_info.directory().c_str(),
531  doc_info.number());
532 
533  // Open a file with the filename given above
534  some_file.open(filename);
535 
536  // Compute the error and document it
537  Bulk_mesh_pt->compute_error(some_file,
538  TanhSolnForPoisson::get_exact_u,
539  error,soln);
540 
541  // Close the file
542  some_file.close();
543 
544  // Output the L2 norm of the error and solution and the relative error:
545  //---------------------------------------------------------------------
546  oomph_info << "Norm of error : " << sqrt(error) << std::endl;
547  oomph_info << "Norm of solution : " << sqrt(soln) << std::endl;
548  oomph_info << "Relative error : " << sqrt(error)/sqrt(soln) << std::endl;
549 
550  // Increment the documentation counter
551  doc_info.number()++;
552 } // End of doc_solution
oomph::DocInfo::directory
std::string directory() const
Output directory.
Definition: oomph_utilities.h:524

References oomph::DocInfo::directory(), calibrate::error, MergeRestartFiles::filename, TanhSolnForPoisson::get_exact_u(), oomph::DocInfo::number(), oomph::oomph_info, and sqrt().

◆ make_new_problem() [1/2]

template<class ELEMENT , class MESH >
MGProblem* FluxPoissonMGProblem< ELEMENT, MESH >::make_new_problem ( )
inlineprivatevirtual

Return a pointer to a new instance of the same problem.

Implements oomph::MGProblem.

164  {
165  // Make new problem of the FluxPoissonMGProblem class whose template
166  // parameters are specified by the template parameters of the current
167  // problem
168  return new FluxPoissonMGProblem<ELEMENT,MESH>
169  (&TanhSolnForPoisson::source_function);
170  } // End of make_new_problem
FluxPoissonMGProblem
Definition: two_d_poisson_tanh_flux_bc.cc:115
TanhSolnForPoisson::source_function
void source_function(const Vector< double > &x, double &source)
Source function required to make the solution above an exact solution.
Definition: HypreSolver_test.cc:60

References TanhSolnForPoisson::source_function().

◆ make_new_problem() [2/2]

template<class ELEMENT , class MESH >
MGProblem* FluxPoissonMGProblem< ELEMENT, MESH >::make_new_problem ( )
inlineprivatevirtual

Return a pointer to a new instance of the same problem.

Implements oomph::MGProblem.

273  {
274  // Make new problem of the FluxPoissonMGProblem class whose template
275  // parameters are specified by the template parameters of the current
276  // problem
277  return new FluxPoissonMGProblem<ELEMENT,MESH>
278  (&SolnForPoisson::source_function);
279  }
SolnForPoisson::source_function
void source_function(const Vector< double > &x, double &source)
Source function required to make the solution above an exact solution.
Definition: two_d_poisson_tanh_flux_bc_validate.cc:186

References SolnForPoisson::source_function().

◆ mesh_pt()

template<class ELEMENT , class MESH >
MESH* FluxPoissonMGProblem< ELEMENT, MESH >::mesh_pt ( )
inlineprivate

Access function for the mesh.

267  {
268  return dynamic_cast<MESH*>(Problem::mesh_pt());
269  }

◆ mg_bulk_mesh_pt() [1/2]

template<class ELEMENT , class MESH >
TreeBasedRefineableMeshBase* FluxPoissonMGProblem< ELEMENT, MESH >::mg_bulk_mesh_pt ( )
inlineprivatevirtual

Pointer to the bulk mesh. Overloads the pure virtual function in the abstract base class, MGProblem. Must be refineable to allow the use of refine_base_mesh_as_in_reference_mesh_minus_one() in make_copy()

Implements oomph::MGProblem.

157  {
158  // Return the pointer to the bulk mesh
159  return Bulk_mesh_pt;
160  } // End of mg_bulk_mesh_pt

◆ mg_bulk_mesh_pt() [2/2]

template<class ELEMENT , class MESH >
TreeBasedRefineableMeshBase* FluxPoissonMGProblem< ELEMENT, MESH >::mg_bulk_mesh_pt ( )
inlinevirtual

Pointer to the bulk mesh. Overloads the pure virtual function in the abstract base class MGProblem. Must be refineable to allow the use of refine_base_mesh_as_in_reference_mesh() in make_copy()

Implements oomph::MGProblem.

244  {
245  // Return the pointer to the bulk mesh
246  return Bulk_mesh_pt;
247  }

◆ set_multigrid_solver() [1/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::set_multigrid_solver
private

Build and set multgrid solver.

Build and set multigrid solver.

Build and set the multigrid solver.

451 {
452  // Make an object of the MGSolver class and get the pointer to it
453  MGSolver<2>* mg_solver_pt=new MGSolver<2>(this);
454 
455  // Switch solver to MG
456  linear_solver_pt()=mg_solver_pt;
457 
458  // Set the pre-smoother factory function
459  mg_solver_pt->set_pre_smoother_factory_function
460  (Smoother_Factory_Function_Helper::set_smoother);
461 
462  // Set the post-smoother factory function
463  mg_solver_pt->set_post_smoother_factory_function
464  (Smoother_Factory_Function_Helper::set_smoother);
465 
466  // Create a file to record the convergence history
467  mg_solver_pt->open_convergence_history_file_stream("RESLT/conv.dat");
468 } // End of set_multigrid_solver
oomph::IterativeLinearSolver::open_convergence_history_file_stream
void open_convergence_history_file_stream(const std::string &file_name, const std::string &zone_title="")
Definition: iterative_linear_solver.h:138
oomph::MGSolver
Definition: geometric_multigrid.h:90
oomph::MGSolver::set_pre_smoother_factory_function
void set_pre_smoother_factory_function(PreSmootherFactoryFctPt pre_smoother_fn)
Access function to set the pre-smoother creation function.
Definition: geometric_multigrid.h:101
oomph::MGSolver::set_post_smoother_factory_function
void set_post_smoother_factory_function(PostSmootherFactoryFctPt post_smoother_fn)
Access function to set the post-smoother creation function.
Definition: geometric_multigrid.h:109
Smoother_Factory_Function_Helper::set_smoother
Smoother * set_smoother()
Definition: two_d_poisson_tanh_flux_bc.cc:96

References oomph::IterativeLinearSolver::open_convergence_history_file_stream(), oomph::MGSolver< DIM >::set_post_smoother_factory_function(), oomph::MGSolver< DIM >::set_pre_smoother_factory_function(), and Smoother_Factory_Function_Helper::set_smoother().

Referenced by FluxPoissonMGProblem< ELEMENT, MESH >::FluxPoissonMGProblem().

◆ set_multigrid_solver() [2/2]

template<class ELEMENT , class MESH >
void FluxPoissonMGProblem< ELEMENT, MESH >::set_multigrid_solver ( )

Build and set multgrid solver.

Member Data Documentation

◆ Bulk_mesh_pt

template<class ELEMENT , class MESH >
MESH * FluxPoissonMGProblem< ELEMENT, MESH >::Bulk_mesh_pt
private

Pointer to the "bulk" mesh.

Referenced by FluxPoissonMGProblem< ELEMENT, MESH >::FluxPoissonMGProblem().

◆ Source_fct_pt

template<class ELEMENT , class MESH >
PoissonEquations< 2 >::PoissonSourceFctPt FluxPoissonMGProblem< ELEMENT, MESH >::Source_fct_pt
private

Pointer to source function.

Referenced by FluxPoissonMGProblem< ELEMENT, MESH >::FluxPoissonMGProblem().

◆ Surface_mesh_pt

template<class ELEMENT , class MESH >
Mesh * FluxPoissonMGProblem< ELEMENT, MESH >::Surface_mesh_pt
private

Pointer to the "surface" mesh.

Referenced by FluxPoissonMGProblem< ELEMENT, MESH >::FluxPoissonMGProblem().

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