[P]arallel [Hi]gh-order [Li]brary for [P]DEs  Latest
Parallel High-Order Library for PDEs through hp-adaptive Discontinuous Galerkin methods
PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real > Class Template Referenceabstract

Large Eddy Simulation equations. Derived from Navier-Stokes for modifying the stress tensor and heat flux, which is derived from PhysicsBase. More...

#include <large_eddy_simulation.h>

Inheritance diagram for PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >:
Collaboration diagram for PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >:

Public Types

using thermal_boundary_condition_enum = Parameters::NavierStokesParam::ThermalBoundaryCondition
 
using two_point_num_flux_enum = Parameters::AllParameters::TwoPointNumericalFlux
 

Public Member Functions

 LargeEddySimulationBase (const Parameters::AllParameters *const parameters_input, const double ref_length, const double gamma_gas, const double mach_inf, const double angle_of_attack, const double side_slip_angle, const double prandtl_number, const double reynolds_number_inf, const bool use_constant_viscosity, const double constant_viscosity, const double temperature_inf, const double turbulent_prandtl_number, const double ratio_of_filter_width_to_cell_size, const double isothermal_wall_temperature=1.0, const thermal_boundary_condition_enum thermal_boundary_condition_type=thermal_boundary_condition_enum::adiabatic, std::shared_ptr< ManufacturedSolutionFunction< dim, real > > manufactured_solution_function=nullptr, const two_point_num_flux_enum two_point_num_flux_type=two_point_num_flux_enum::KG)
 Constructor.
 
std::array< dealii::Tensor< 1, dim, real >, nstate > convective_flux (const std::array< real, nstate > &conservative_soln) const
 Convective flux: \( \mathbf{F}_{conv} \).
 
std::array< dealii::Tensor< 1, dim, real >, nstate > dissipative_flux (const std::array< real, nstate > &conservative_soln, const std::array< dealii::Tensor< 1, dim, real >, nstate > &solution_gradient, const dealii::types::global_dof_index cell_index) const
 Dissipative (i.e. viscous) flux: \( \mathbf{F}_{diss} \).
 
std::array< real, nstate > convective_eigenvalues (const std::array< real, nstate > &, const dealii::Tensor< 1, dim, real > &) const override
 Convective eigenvalues of the additional models' PDEs. More...
 
real max_convective_eigenvalue (const std::array< real, nstate > &soln) const
 Maximum convective eigenvalue of the additional models' PDEs. More...
 
real max_convective_normal_eigenvalue (const std::array< real, nstate > &soln, const dealii::Tensor< 1, dim, real > &normal) const
 Maximum convective normal eigenvalue (used in Lax-Friedrichs) of the additional models' PDEs. More...
 
std::array< real, nstate > source_term (const dealii::Point< dim, real > &pos, const std::array< real, nstate > &solution, const real current_time, const dealii::types::global_dof_index cell_index) const
 Source term for manufactured solution functions.
 
double get_filter_width (const dealii::types::global_dof_index cell_index) const
 Compute the nondimensionalized filter width used by the SGS model given a cell index. More...
 
virtual dealii::Tensor< 2, dim, real > compute_SGS_stress_tensor (const std::array< real, nstate > &primitive_soln, const std::array< dealii::Tensor< 1, dim, real >, nstate > &primitive_soln_gradient, const dealii::types::global_dof_index cell_index) const =0
 Nondimensionalized sub-grid scale (SGS) stress tensor, (tau^sgs)*.
 
virtual dealii::Tensor< 1, dim, real > compute_SGS_heat_flux (const std::array< real, nstate > &primitive_soln, const std::array< dealii::Tensor< 1, dim, real >, nstate > &primitive_soln_gradient, const dealii::types::global_dof_index cell_index) const =0
 Nondimensionalized sub-grid scale (SGS) heat flux, (q^sgs)*.
 
virtual dealii::Tensor< 2, dim, FadTypecompute_SGS_stress_tensor_fad (const std::array< FadType, nstate > &primitive_soln, const std::array< dealii::Tensor< 1, dim, FadType >, nstate > &primitive_soln_gradient, const dealii::types::global_dof_index cell_index) const =0
 Nondimensionalized sub-grid scale (SGS) stress tensor, (tau^sgs)* (Automatic Differentiation Type: FadType)
 
virtual dealii::Tensor< 1, dim, FadTypecompute_SGS_heat_flux_fad (const std::array< FadType, nstate > &primitive_soln, const std::array< dealii::Tensor< 1, dim, FadType >, nstate > &primitive_soln_gradient, const dealii::types::global_dof_index cell_index) const =0
 Nondimensionalized sub-grid scale (SGS) heat flux, (q^sgs)* (Automatic Differentiation Type: FadType)
 
- Public Member Functions inherited from PHiLiP::Physics::ModelBase< dim, nstate, real >
 ModelBase (std::shared_ptr< ManufacturedSolutionFunction< dim, real > > manufactured_solution_function_input=nullptr)
 Constructor.
 
virtual ~ModelBase ()=default
 Virtual destructor required for abstract classes.
 
virtual std::array< real, nstate > physical_source_term (const dealii::Point< dim, real > &pos, const std::array< real, nstate > &solution, const std::array< dealii::Tensor< 1, dim, real >, nstate > &solution_gradient, const dealii::types::global_dof_index cell_index) const
 Physical source terms additional to the baseline physics (including physical source terms in additional PDEs of model)
 
void boundary_face_values (const int boundary_type, const dealii::Point< dim, real > &pos, const dealii::Tensor< 1, dim, real > &normal, const std::array< real, nstate > &soln_int, const std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_int, std::array< real, nstate > &soln_bc, std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_bc) const
 Boundary condition handler.
 
virtual dealii::Vector< double > post_compute_derived_quantities_vector (const dealii::Vector< double > &uh, const std::vector< dealii::Tensor< 1, dim > > &, const std::vector< dealii::Tensor< 2, dim > > &, const dealii::Tensor< 1, dim > &, const dealii::Point< dim > &) const
 Returns current vector solution to be used by PhysicsPostprocessor to output current solution. More...
 
virtual std::vector< dealii::DataComponentInterpretation::DataComponentInterpretation > post_get_data_component_interpretation () const
 Returns DataComponentInterpretation of the solution to be used by PhysicsPostprocessor to output current solution. More...
 
virtual std::vector< std::string > post_get_names () const
 Returns names of the solution to be used by PhysicsPostprocessor to output current solution. More...
 

Public Attributes

const double turbulent_prandtl_number
 Turbulent Prandtl number.
 
const double ratio_of_filter_width_to_cell_size
 Ratio of filter width to cell size.
 
std::unique_ptr< NavierStokes< dim, nstate, real > > navier_stokes_physics
 Pointer to Navier-Stokes physics object.
 
- Public Attributes inherited from PHiLiP::Physics::ModelBase< dim, nstate, real >
std::shared_ptr< ManufacturedSolutionFunction< dim, real > > manufactured_solution_function
 Manufactured solution function.
 
dealii::LinearAlgebra::distributed::Vector< int > cellwise_poly_degree
 Cellwise polynomial degree.
 
dealii::LinearAlgebra::distributed::Vector< double > cellwise_volume
 

Protected Member Functions

template<typename real2 >
real2 get_tensor_magnitude_sqr (const dealii::Tensor< 2, dim, real2 > &tensor) const
 Returns the square of the magnitude of the tensor (i.e. the double dot product of a tensor with itself)
 
template<typename real2 >
std::array< dealii::Tensor< 1, dim, real2 >, nstate > dissipative_flux_templated (const std::array< real2, nstate > &conservative_soln, const std::array< dealii::Tensor< 1, dim, real2 >, nstate > &solution_gradient, const dealii::types::global_dof_index cell_index) const
 Templated dissipative (i.e. viscous) flux: \( \mathbf{F}_{diss} \).
 
dealii::Tensor< 2, nstate, real > dissipative_flux_directional_jacobian (const std::array< real, nstate > &conservative_soln, const std::array< dealii::Tensor< 1, dim, real >, nstate > &solution_gradient, const dealii::Tensor< 1, dim, real > &normal, const dealii::types::global_dof_index cell_index) const
 
dealii::Tensor< 2, nstate, real > dissipative_flux_directional_jacobian_wrt_gradient_component (const std::array< real, nstate > &conservative_soln, const std::array< dealii::Tensor< 1, dim, real >, nstate > &solution_gradient, const dealii::Tensor< 1, dim, real > &normal, const int d_gradient, const dealii::types::global_dof_index cell_index) const
 
std::array< real, nstate > get_manufactured_solution_value (const dealii::Point< dim, real > &pos) const
 Get manufactured solution value (repeated from Euler)
 
std::array< dealii::Tensor< 1, dim, real >, nstate > get_manufactured_solution_gradient (const dealii::Point< dim, real > &pos) const
 Get manufactured solution value (repeated from Euler)
 
std::array< real, nstate > dissipative_source_term (const dealii::Point< dim, real > &pos, const dealii::types::global_dof_index cell_index) const
 Dissipative flux contribution to the source term (repeated from NavierStokes) More...
 
- Protected Member Functions inherited from PHiLiP::Physics::ModelBase< dim, nstate, real >
virtual void boundary_manufactured_solution (const dealii::Point< dim, real > &pos, const dealii::Tensor< 1, dim, real > &normal_int, const std::array< real, nstate > &soln_int, const std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_int, std::array< real, nstate > &soln_bc, std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_bc) const
 Evaluate the manufactured solution boundary conditions.
 
virtual void boundary_wall (std::array< real, nstate > &soln_bc, std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_bc) const
 Wall boundary condition.
 
virtual void boundary_outflow (const std::array< real, nstate > &soln_int, const std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_int, std::array< real, nstate > &soln_bc, std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_bc) const
 Outflow Boundary Condition.
 
virtual void boundary_inflow (const std::array< real, nstate > &soln_int, const std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_int, std::array< real, nstate > &soln_bc, std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_bc) const
 Inflow boundary conditions.
 
virtual void boundary_farfield (std::array< real, nstate > &soln_bc) const
 Farfield boundary conditions based on freestream values.
 
virtual void boundary_riemann (const dealii::Tensor< 1, dim, real > &normal_int, const std::array< real, nstate > &soln_int, std::array< real, nstate > &soln_bc) const
 
virtual void boundary_slip_wall (const dealii::Tensor< 1, dim, real > &normal_int, const std::array< real, nstate > &soln_int, const std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_int, std::array< real, nstate > &soln_bc, std::array< dealii::Tensor< 1, dim, real >, nstate > &soln_grad_bc) const
 

Additional Inherited Members

- Protected Attributes inherited from PHiLiP::Physics::ModelBase< dim, nstate, real >
const MPI_Comm mpi_communicator
 MPI communicator.
 
dealii::ConditionalOStream pcout
 Parallel std::cout that only outputs on mpi_rank==0.
 

Detailed Description

template<int dim, int nstate, typename real>
class PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >

Large Eddy Simulation equations. Derived from Navier-Stokes for modifying the stress tensor and heat flux, which is derived from PhysicsBase.

Definition at line 13 of file large_eddy_simulation.h.

Member Function Documentation

◆ convective_eigenvalues()

template<int dim, int nstate, typename real >
std::array< real, nstate > PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >::convective_eigenvalues ( const std::array< real, nstate > &  ,
const dealii::Tensor< 1, dim, real > &   
) const
overridevirtual

Convective eigenvalues of the additional models' PDEs.

For LES model, all entries are assigned to be zero

Implements PHiLiP::Physics::ModelBase< dim, nstate, real >.

Definition at line 137 of file large_eddy_simulation.cpp.

◆ dissipative_flux_directional_jacobian()

template<int dim, int nstate, typename real >
dealii::Tensor< 2, nstate, real > PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >::dissipative_flux_directional_jacobian ( const std::array< real, nstate > &  conservative_soln,
const std::array< dealii::Tensor< 1, dim, real >, nstate > &  solution_gradient,
const dealii::Tensor< 1, dim, real > &  normal,
const dealii::types::global_dof_index  cell_index 
) const
protected

Dissipative flux Jacobian (repeated from NavierStokes) Note: Only used for computing the manufactured solution source term; computed using automatic differentiation

Definition at line 221 of file large_eddy_simulation.cpp.

◆ dissipative_flux_directional_jacobian_wrt_gradient_component()

template<int dim, int nstate, typename real >
dealii::Tensor< 2, nstate, real > PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >::dissipative_flux_directional_jacobian_wrt_gradient_component ( const std::array< real, nstate > &  conservative_soln,
const std::array< dealii::Tensor< 1, dim, real >, nstate > &  solution_gradient,
const dealii::Tensor< 1, dim, real > &  normal,
const int  d_gradient,
const dealii::types::global_dof_index  cell_index 
) const
protected

Dissipative flux Jacobian wrt gradient component (repeated from NavierStokes) Note: Only used for computing the manufactured solution source term; computed using automatic differentiation

Definition at line 260 of file large_eddy_simulation.cpp.

◆ dissipative_source_term()

template<int dim, int nstate, typename real >
std::array< real, nstate > PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >::dissipative_source_term ( const dealii::Point< dim, real > &  pos,
const dealii::types::global_dof_index  cell_index 
) const
protected

Dissipative flux contribution to the source term (repeated from NavierStokes)

This function is same as the one in NavierStokes; TO DO: Reduce this code repetition in the future by moving it elsewhere

Definition at line 336 of file large_eddy_simulation.cpp.

◆ get_filter_width()

template<int dim, int nstate, typename real >
double PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >::get_filter_width ( const dealii::types::global_dof_index  cell_index) const

Compute the nondimensionalized filter width used by the SGS model given a cell index.

Reference: Marta de la Llave Plata, et al. "On the performance of a high-order multiscale DG approach to LES at increasing Reynolds number." Computers and Fluids 194 (2019), Page 4, Eq.(14).

Definition at line 183 of file large_eddy_simulation.cpp.

◆ max_convective_eigenvalue()

template<int dim, int nstate, typename real >
real PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >::max_convective_eigenvalue ( const std::array< real, nstate > &  soln) const
virtual

Maximum convective eigenvalue of the additional models' PDEs.

For LES model, this value is assigned to be zero

Implements PHiLiP::Physics::ModelBase< dim, nstate, real >.

Definition at line 148 of file large_eddy_simulation.cpp.

◆ max_convective_normal_eigenvalue()

template<int dim, int nstate, typename real >
real PHiLiP::Physics::LargeEddySimulationBase< dim, nstate, real >::max_convective_normal_eigenvalue ( const std::array< real, nstate > &  soln,
const dealii::Tensor< 1, dim, real > &  normal 
) const
virtual

Maximum convective normal eigenvalue (used in Lax-Friedrichs) of the additional models' PDEs.

For LES model, this value is assigned to be zero

Implements PHiLiP::Physics::ModelBase< dim, nstate, real >.

Definition at line 156 of file large_eddy_simulation.cpp.


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