GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity > Class Template Reference

Physics class for Incompressible Navier-Stokes. More...

#include <low_mach_navier_stokes.h>

Inheritance diagram for GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >:

Collaboration diagram for GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >:

Public Member Functions
	LowMachNavierStokes (const PhysicsName &physics_name, const GetPot &input)
	~LowMachNavierStokes ()
virtual void	auxiliary_init (MultiphysicsSystem &system)
	Any auxillary initialization a Physics class may need. More...
virtual void	register_postprocessing_vars (const GetPot &input, PostProcessedQuantities< libMesh::Real > &postprocessing)
	Register postprocessing variables for LowMachNavierStokes. More...
virtual void	init_context (AssemblyContext &context)
	Initialize context for added physics variables. More...
virtual void	element_time_derivative (bool compute_jacobian, AssemblyContext &context)
	Time dependent part(s) of physics for element interiors. More...
virtual void	element_constraint (bool compute_jacobian, AssemblyContext &context)
	Constraint part(s) of physics for element interiors. More...
virtual void	mass_residual (bool compute_jacobian, AssemblyContext &context)
	Mass matrix part(s) for element interiors. All boundary terms lie within the time_derivative part. More...
virtual void	compute_element_time_derivative_cache (AssemblyContext &context)
virtual void	compute_postprocessed_quantity (unsigned int quantity_index, const AssemblyContext &context, const libMesh::Point &point, libMesh::Real &value)
Public Member Functions inherited from GRINS::LowMachNavierStokesBase< Viscosity, SpecificHeat, ThermalConductivity >
	LowMachNavierStokesBase (const PhysicsName &physics_name, const std::string &core_physics_name, const GetPot &input)
	~LowMachNavierStokesBase ()
virtual void	set_time_evolving_vars (libMesh::FEMSystem *system)
	Sets velocity variables to be time-evolving. More...
libMesh::Real	T (const libMesh::Point &p, const AssemblyContext &c) const
libMesh::Real	rho (libMesh::Real T, libMesh::Real p0) const
libMesh::Real	d_rho_dT (libMesh::Real T, libMesh::Real p0) const
libMesh::Real	get_p0_steady (const AssemblyContext &c, unsigned int qp) const
libMesh::Real	get_p0_steady_side (const AssemblyContext &c, unsigned int qp) const
libMesh::Real	get_p0_steady (const AssemblyContext &c, const libMesh::Point &p) const
libMesh::Real	get_p0_transient (AssemblyContext &c, unsigned int qp) const
virtual void	register_parameter (const std::string &param_name, libMesh::ParameterMultiAccessor< libMesh::Number > &param_pointer) const
	Each subclass will register its copy of an independent. More...
Public Member Functions inherited from GRINS::Physics
	Physics (const GRINS::PhysicsName &physics_name, const GetPot &input)
virtual	~Physics ()
virtual void	init_variables (libMesh::FEMSystem *)
	Initialize variables for this physics. More...
virtual bool	enabled_on_elem (const libMesh::Elem *elem)
	Find if current physics is active on supplied element. More...
void	set_is_steady (bool is_steady)
	Sets whether this physics is to be solved with a steady solver or not. More...
bool	is_steady () const
	Returns whether or not this physics is being solved with a steady solver. More...
virtual void	preassembly (MultiphysicsSystem &)
	Perform any necessary setup before element assembly begins. More...
virtual void	reinit (MultiphysicsSystem &)
	Any reinitialization that needs to be done. More...
virtual void	side_time_derivative (bool, AssemblyContext &)
	Time dependent part(s) of physics for boundaries of elements on the domain boundary. More...
virtual void	nonlocal_time_derivative (bool, AssemblyContext &)
	Time dependent part(s) of physics for scalar variables. More...
virtual void	side_constraint (bool, AssemblyContext &)
	Constraint part(s) of physics for boundaries of elements on the domain boundary. More...
virtual void	nonlocal_constraint (bool, AssemblyContext &)
	Constraint part(s) of physics for scalar variables. More...
virtual void	damping_residual (bool, AssemblyContext &)
	Damping matrix part(s) for element interiors. All boundary terms lie within the time_derivative part. More...
virtual void	nonlocal_mass_residual (bool, AssemblyContext &)
	Mass matrix part(s) for scalar variables. More...
void	init_ics (libMesh::FEMSystem *system, libMesh::CompositeFunction< libMesh::Number > &all_ics)
virtual void	compute_side_time_derivative_cache (AssemblyContext &)
virtual void	compute_nonlocal_time_derivative_cache (AssemblyContext &)
virtual void	compute_element_constraint_cache (AssemblyContext &)
virtual void	compute_side_constraint_cache (AssemblyContext &)
virtual void	compute_nonlocal_constraint_cache (AssemblyContext &)
virtual void	compute_damping_residual_cache (AssemblyContext &)
virtual void	compute_mass_residual_cache (AssemblyContext &)
virtual void	compute_nonlocal_mass_residual_cache (AssemblyContext &)
ICHandlingBase *	get_ic_handler ()
Public Member Functions inherited from GRINS::ParameterUser
	ParameterUser (const std::string &user_name)
virtual	~ParameterUser ()
virtual void	set_parameter (libMesh::Number &param_variable, const GetPot &input, const std::string &param_name, libMesh::Number param_default)
	Each subclass can simultaneously read a parameter value from. More...
virtual void	set_parameter (libMesh::ParsedFunction< libMesh::Number, libMesh::Gradient > &func, const GetPot &input, const std::string &func_param_name, const std::string &param_default)
	Each subclass can simultaneously read a parsed function from. More...
virtual void	set_parameter (libMesh::ParsedFEMFunction< libMesh::Number > &func, const GetPot &input, const std::string &func_param_name, const std::string &param_default)
	Each subclass can simultaneously read a parsed function from. More...
virtual void	move_parameter (const libMesh::Number &old_parameter, libMesh::Number &new_parameter)
	When cloning an object, we need to update parameter pointers. More...
virtual void	move_parameter (const libMesh::ParsedFunction< libMesh::Number, libMesh::Gradient > &old_func, libMesh::ParsedFunction< libMesh::Number, libMesh::Gradient > &new_func)
	When cloning an object, we need to update parameter pointers. More...
virtual void	move_parameter (const libMesh::ParsedFEMFunction< libMesh::Number > &old_func, libMesh::ParsedFEMFunction< libMesh::Number > &new_func)
	When cloning an object, we need to update parameter pointers. More...

Protected Member Functions
void	assemble_mass_time_deriv (bool compute_jacobian, AssemblyContext &context, const CachedValues &cache)
	Helper function. More...
void	assemble_momentum_time_deriv (bool compute_jacobian, AssemblyContext &context, const CachedValues &cache)
	Helper function. More...
void	assemble_energy_time_deriv (bool compute_jacobian, AssemblyContext &context, const CachedValues &cache)
	Helper function. More...
void	assemble_continuity_mass_residual (bool compute_jacobian, AssemblyContext &context)
	Helper function. More...
void	assemble_momentum_mass_residual (bool compute_jacobian, AssemblyContext &context)
	Helper function. More...
void	assemble_energy_mass_residual (bool compute_jacobian, AssemblyContext &context)
	Helper function. More...
void	assemble_thermo_press_elem_time_deriv (bool compute_jacobian, AssemblyContext &context)
void	assemble_thermo_press_side_time_deriv (bool compute_jacobian, AssemblyContext &context)
void	assemble_thermo_press_mass_residual (bool compute_jacobian, AssemblyContext &context)
Protected Member Functions inherited from GRINS::Physics
libMesh::UniquePtr< libMesh::FEGenericBase< libMesh::Real > >	build_new_fe (const libMesh::Elem *elem, const libMesh::FEGenericBase< libMesh::Real > *fe, const libMesh::Point p)
void	parse_enabled_subdomains (const GetPot &input, const std::string &physics_name)
void	check_var_subdomain_consistency (const FEVariablesBase &var) const
	Check that var is enabled on at least the subdomains this Physics is. More...

Protected Attributes
bool	_pin_pressure
	Enable pressure pinning. More...
PressurePinning	_p_pinning
unsigned int	_rho_index
	Cache index for density post-processing. More...
Protected Attributes inherited from GRINS::LowMachNavierStokesBase< Viscosity, SpecificHeat, ThermalConductivity >
libMesh::Number	_p0_over_R
	Thermodynamic pressure divided by gas constant. More...
libMesh::Number	_p0
libMesh::Number	_R
libMesh::Number	_T0
VelocityVariable &	_flow_vars
PressureFEVariable &	_press_var
PrimitiveTempFEVariables &	_temp_vars
ThermoPressureVariable *	_p0_var
Viscosity	_mu
	Viscosity object. More...
SpecificHeat	_cp
	Specific heat object. More...
ThermalConductivity	_k
	Thermal conductivity object. More...
libMesh::Point	_g
	Gravity vector. More...
bool	_enable_thermo_press_calc
	Flag to enable thermodynamic pressure calculation. More...
Protected Attributes inherited from GRINS::Physics
const PhysicsName	_physics_name
	Name of the physics object. Used for reading physics specific inputs. More...
GRINS::ICHandlingBase *	_ic_handler
std::set< libMesh::subdomain_id_type >	_enabled_subdomains
	Subdomains on which the current Physics class is enabled. More...

Private Member Functions
	LowMachNavierStokes ()

Additional Inherited Members
Static Public Member Functions inherited from GRINS::Physics
static void	set_is_axisymmetric (bool is_axisymmetric)
	Set whether we should treat the problem as axisymmetric. More...
static bool	is_axisymmetric ()
Static Public Attributes inherited from GRINS::ParameterUser
static std::string	zero_vector_function = std::string("{0}")
	A parseable function string with LIBMESH_DIM components, all 0. More...
Static Protected Attributes inherited from GRINS::Physics
static bool	_is_steady = false
	Caches whether or not the solver that's being used is steady or not. More...
static bool	_is_axisymmetric = false
	Caches whether we are solving an axisymmetric problem or not. More...

Detailed Description

template<class Viscosity, class SpecificHeat, class ThermalConductivity>
class GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >

Physics class for Incompressible Navier-Stokes.

This physics class implements the classical Incompressible Navier-Stokes equations.

Definition at line 41 of file low_mach_navier_stokes.h.

Constructor & Destructor Documentation

template<class Mu , class SH , class TC >

GRINS::LowMachNavierStokes< Mu, SH, TC >::LowMachNavierStokes	(	const PhysicsName &	physics_name,
		const GetPot &	input
	)

Definition at line 45 of file low_mach_navier_stokes.C.

References GRINS::LowMachNavierStokesBase< Viscosity, SpecificHeat, ThermalConductivity >::_flow_vars, GRINS::Physics::_ic_handler, GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >::_pin_pressure, GRINS::MultcomponentVectorVariable::dim(), and GRINS::PhysicsNaming::low_mach_navier_stokes().

    : LowMachNavierStokesBase<Mu,SH,TC>(physics_name,PhysicsNaming::low_mach_navier_stokes(),input),
    _p_pinning(input,physics_name),
    _rho_index(0) // Initialize to zero
  {
    this->_ic_handler = new GenericICHandler( physics_name, input );

    this->_pin_pressure = input("Physics/"+PhysicsNaming::low_mach_navier_stokes()+"/pin_pressure", false );

    if( this->_flow_vars.dim() < 2 )
      libmesh_error_msg("ERROR: LowMachNavierStokes only valid for two or three dimensions! Make sure you have at least two components in your Velocity type variable.");
  }

template<class Viscosity , class SpecificHeat , class ThermalConductivity >

GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >::~LowMachNavierStokes ( )

inline

Definition at line 47 of file low_mach_navier_stokes.h.

{};

template<class Viscosity , class SpecificHeat , class ThermalConductivity >

GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >::LowMachNavierStokes ( )

private

Member Function Documentation

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_continuity_mass_residual ( bool  compute_jacobian, AssemblyContext &  context  )

protected

Helper function.

Definition at line 693 of file low_mach_navier_stokes.C.

  {
    // Element Jacobian * quadrature weights for interior integration
    const std::vector<libMesh::Real> &JxW =
      context.get_element_fe(this->_flow_vars.u())->get_JxW();

    // The shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& p_phi =
      context.get_element_fe(this->_press_var.p())->get_phi();

    // The number of local degrees of freedom in each variable
    const unsigned int n_p_dofs = context.get_dof_indices(this->_press_var.p()).size();

    // The subvectors and submatrices we need to fill:
    libMesh::DenseSubVector<libMesh::Real> &F_p = context.get_elem_residual(this->_press_var.p());

    unsigned int n_qpoints = context.get_element_qrule().n_points();

    for (unsigned int qp = 0; qp != n_qpoints; ++qp)
      {
        // For the mass residual, we need to be a little careful.
        // The time integrator is handling the time-discretization
        // for us so we need to supply M(u_fixed)*u' for the residual.
        // u_fixed will be given by the fixed_interior_value function
        // while u' will be given by the interior_rate function.
        libMesh::Real T_dot;
        context.interior_rate(this->_temp_vars.T(), qp, T_dot);

        libMesh::Real T = context.fixed_interior_value(this->_temp_vars.T(), qp);

        for (unsigned int i = 0; i != n_p_dofs; ++i)
          {
            F_p(i) -= T_dot/T*p_phi[i][qp]*JxW[qp];
          } // End DoF loop i

      } // End quadrature loop qp

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_energy_mass_residual ( bool  compute_jacobian, AssemblyContext &  context  )

protected

Helper function.

Definition at line 814 of file low_mach_navier_stokes.C.

  {
    // Element Jacobian * quadrature weights for interior integration
    const std::vector<libMesh::Real> &JxW =
      context.get_element_fe(this->_temp_vars.T())->get_JxW();

    // The shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& T_phi =
      context.get_element_fe(this->_temp_vars.T())->get_phi();

    // The number of local degrees of freedom in each variable
    const unsigned int n_T_dofs = context.get_dof_indices(this->_temp_vars.T()).size();

    // The subvectors and submatrices we need to fill:
    libMesh::DenseSubVector<libMesh::Real> &F_T = context.get_elem_residual(this->_temp_vars.T());

    unsigned int n_qpoints = context.get_element_qrule().n_points();

    for (unsigned int qp = 0; qp != n_qpoints; ++qp)
      {
        // For the mass residual, we need to be a little careful.
        // The time integrator is handling the time-discretization
        // for us so we need to supply M(u_fixed)*u' for the residual.
        // u_fixed will be given by the fixed_interior_value function
        // while u will be given by the interior_rate function.
        libMesh::Real T_dot;
        context.interior_rate(this->_temp_vars.T(), qp, T_dot);

        libMesh::Real T = context.fixed_interior_value(this->_temp_vars.T(), qp);

        libMesh::Real cp = this->_cp(T);

        libMesh::Number rho = this->rho(T, this->get_p0_transient(context, qp));

        for (unsigned int i = 0; i != n_T_dofs; ++i)
          {
            F_T(i) -= rho*cp*T_dot*T_phi[i][qp]*JxW[qp];
          } // End DoF loop i

      } // End quadrature loop qp

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_energy_time_deriv ( bool  compute_jacobian, AssemblyContext &  context, const CachedValues &  cache  )

protected

Helper function.

Definition at line 581 of file low_mach_navier_stokes.C.

References GRINS::CachedValues::get_cached_gradient_values(), GRINS::CachedValues::get_cached_values(), GRINS::Cache::TEMPERATURE, GRINS::Cache::TEMPERATURE_GRAD, GRINS::Cache::THERMO_PRESSURE, GRINS::Cache::X_VELOCITY, GRINS::Cache::Y_VELOCITY, and GRINS::Cache::Z_VELOCITY.

  {
    // The number of local degrees of freedom in each variable.
    const unsigned int n_T_dofs = context.get_dof_indices(this->_temp_vars.T()).size();
    const unsigned int n_u_dofs = context.get_dof_indices(this->_flow_vars.u()).size();

    // Element Jacobian * quadrature weights for interior integration.
    const std::vector<libMesh::Real> &JxW =
      context.get_element_fe(this->_temp_vars.T())->get_JxW();

    // The temperature shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& T_phi =
      context.get_element_fe(this->_temp_vars.T())->get_phi();
    const std::vector<std::vector<libMesh::Real> >& u_phi =
      context.get_element_fe(this->_flow_vars.u())->get_phi();

    // The temperature shape functions gradients at interior quadrature points.
    const std::vector<std::vector<libMesh::RealGradient> >& T_gradphi =
      context.get_element_fe(this->_temp_vars.T())->get_dphi();

    libMesh::DenseSubVector<libMesh::Number> &FT = context.get_elem_residual(this->_temp_vars.T()); // R_{T}

    unsigned int n_qpoints = context.get_element_qrule().n_points();
    for (unsigned int qp=0; qp != n_qpoints; qp++)
      {
        libMesh::Number u, v, T, p0;
        u = cache.get_cached_values(Cache::X_VELOCITY)[qp];
        v = cache.get_cached_values(Cache::Y_VELOCITY)[qp];
        T = cache.get_cached_values(Cache::TEMPERATURE)[qp];
        p0 = cache.get_cached_values(Cache::THERMO_PRESSURE)[qp];

        libMesh::Gradient grad_T = cache.get_cached_gradient_values(Cache::TEMPERATURE_GRAD)[qp];

        libMesh::NumberVectorValue U(u,v);
        if (this->_flow_vars.dim() == 3)
          U(2) = cache.get_cached_values(Cache::Z_VELOCITY)[qp]; // w

        libMesh::DenseSubMatrix<libMesh::Number> &KTu = context.get_elem_jacobian(this->_temp_vars.T(), this->_flow_vars.u()); // R_{u},{u}
        libMesh::DenseSubMatrix<libMesh::Number> &KTv = context.get_elem_jacobian(this->_temp_vars.T(), this->_flow_vars.v()); // R_{u},{u}
        libMesh::DenseSubMatrix<libMesh::Number>* KTw = NULL;

        libMesh::DenseSubMatrix<libMesh::Number> &KTT = context.get_elem_jacobian(this->_temp_vars.T(), this->_temp_vars.T()); // R_{u},{u}

        if( this->_flow_vars.dim() == 3 )
          {
            KTw = &context.get_elem_jacobian(this->_temp_vars.T(), this->_flow_vars.w()); // R_{u},{w}
          }

        libMesh::Number k = this->_k(T);
        libMesh::Number dk_dT = this->_k.deriv(T);
        libMesh::Number cp = this->_cp(T);
        libMesh::Number d_cp = this->_cp.deriv(T);
        libMesh::Number rho = this->rho( T, p0 );
        libMesh::Number d_rho = this->d_rho_dT( T, p0 );

        // Now a loop over the pressure degrees of freedom.  This
        // computes the contributions of the continuity equation.
        for (unsigned int i=0; i != n_T_dofs; i++)
          {
            FT(i) += ( -rho*cp*U*grad_T*T_phi[i][qp] // convection term
                       - k*grad_T*T_gradphi[i][qp]            // diffusion term
                       )*JxW[qp];


            if(compute_jacobian)
              {
                for (unsigned int j=0; j!=n_u_dofs; j++)
                  {
                    //pre-compute repeated term
                    libMesh::Number r0 = rho*cp*T_phi[i][qp]*u_phi[j][qp];

                    KTu(i,j) += JxW[qp]*
                      -r0*grad_T(0);
                    //-rho*cp*u_phi[j][qp]*grad_T(0)*T_phi[i][qp];

                    KTv(i,j) += JxW[qp]*
                      -r0*grad_T(1);
                    //-rho*cp*u_phi[j][qp]*grad_T(1)*T_phi[i][qp];

                    if (this->_flow_vars.dim() == 3)
                      {
                        (*KTw)(i,j) += JxW[qp]*
                          -r0*grad_T(2);
                        //-rho*cp*u_phi[j][qp]*grad_T(2)*T_phi[i][qp];
                      }

                  } // end u_dofs loop (j)


                for (unsigned int j=0; j!=n_T_dofs; j++)
                  {
                    KTT(i,j) += JxW[qp]* (
                                          -rho*(
                                                cp*U*T_phi[i][qp]*T_gradphi[j][qp]
                                                + U*grad_T*T_phi[i][qp]*d_cp*T_phi[j][qp]
                                                )
                                          -cp*U*grad_T*T_phi[i][qp]*d_rho*T_phi[j][qp]
                                          -k*T_gradphi[i][qp]*T_gradphi[j][qp]
                                          -grad_T*T_gradphi[i][qp]*dk_dT*T_phi[j][qp]
                                          );
                  } // end T_dofs loop (j)

              } // end if compute_jacobian
          } // end outer T_dofs loop (i)
      } //end qp loop

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_mass_time_deriv ( bool  compute_jacobian, AssemblyContext &  context, const CachedValues &  cache  )

protected

Helper function.

Definition at line 152 of file low_mach_navier_stokes.C.

References GRINS::CachedValues::get_cached_gradient_values(), GRINS::CachedValues::get_cached_values(), GRINS::Cache::TEMPERATURE, GRINS::Cache::TEMPERATURE_GRAD, GRINS::Cache::X_VELOCITY, GRINS::Cache::X_VELOCITY_GRAD, GRINS::Cache::Y_VELOCITY, GRINS::Cache::Y_VELOCITY_GRAD, GRINS::Cache::Z_VELOCITY, and GRINS::Cache::Z_VELOCITY_GRAD.

  {
    // The number of local degrees of freedom in each variable.
    const unsigned int n_p_dofs = context.get_dof_indices(this->_press_var.p()).size();

    // Element Jacobian * quadrature weights for interior integration.
    const std::vector<libMesh::Real> &JxW =
      context.get_element_fe(this->_flow_vars.u())->get_JxW();

    // The pressure shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& p_phi =
      context.get_element_fe(this->_press_var.p())->get_phi();



    // The velocity shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& u_phi =
      context.get_element_fe(this->_flow_vars.u())->get_phi();

    // The velocity shape function gradients (in global coords.)
    // at interior quadrature points.
    const std::vector<std::vector<libMesh::RealGradient> >& u_gradphi =
      context.get_element_fe(this->_flow_vars.u())->get_dphi();

    // The temperature shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& T_phi =
      context.get_element_fe(this->_temp_vars.T())->get_phi();

    // The temperature shape function gradients (in global coords.)
    // at interior quadrature points.
    const std::vector<std::vector<libMesh::RealGradient> >& T_gradphi =
      context.get_element_fe(this->_temp_vars.T())->get_dphi();



    libMesh::DenseSubVector<libMesh::Number> &Fp = context.get_elem_residual(this->_press_var.p()); // R_{p}

    unsigned int n_qpoints = context.get_element_qrule().n_points();

    // The number of local degrees of freedom in each variable.
    const unsigned int n_t_dofs = context.get_dof_indices(this->_temp_vars.T()).size();

    // The number of local degrees of freedom in each variable.
    const unsigned int n_u_dofs = context.get_dof_indices(this->_flow_vars.u()).size();

    // Check number of dofs is same for _flow_vars.u(), v_var and w_var.
    libmesh_assert (n_u_dofs == context.get_dof_indices(this->_flow_vars.v()).size());

    if (this->_flow_vars.dim() == 3)
      libmesh_assert (n_u_dofs == context.get_dof_indices(this->_flow_vars.w()).size());

    for (unsigned int qp=0; qp != n_qpoints; qp++)
      {
        libMesh::Number u, v, T;
        u = cache.get_cached_values(Cache::X_VELOCITY)[qp];
        v = cache.get_cached_values(Cache::Y_VELOCITY)[qp];

        T = cache.get_cached_values(Cache::TEMPERATURE)[qp];

        libMesh::Gradient grad_u = cache.get_cached_gradient_values(Cache::X_VELOCITY_GRAD)[qp];
        libMesh::Gradient grad_v = cache.get_cached_gradient_values(Cache::Y_VELOCITY_GRAD)[qp];

        libMesh::Gradient grad_T = cache.get_cached_gradient_values(Cache::TEMPERATURE_GRAD)[qp];

        libMesh::NumberVectorValue U(u,v);
        if (this->_flow_vars.dim() == 3)
          U(2) = cache.get_cached_values(Cache::Z_VELOCITY)[qp]; // w

        libMesh::Number divU = grad_u(0) + grad_v(1);
        if (this->_flow_vars.dim() == 3)
          {
            libMesh::Gradient grad_w = cache.get_cached_gradient_values(Cache::Z_VELOCITY_GRAD)[qp];
            divU += grad_w(2);
          }


        libMesh::DenseSubMatrix<libMesh::Number> &KPu = context.get_elem_jacobian(this->_press_var.p(), this->_flow_vars.u());
        libMesh::DenseSubMatrix<libMesh::Number> &KPv = context.get_elem_jacobian(this->_press_var.p(), this->_flow_vars.v());
        libMesh::DenseSubMatrix<libMesh::Number> &KPT = context.get_elem_jacobian(this->_press_var.p(), this->_temp_vars.T());

        libMesh::DenseSubMatrix<libMesh::Number>* KPw = NULL;

        if( this->_flow_vars.dim() == 3 )
          {
            KPw = &context.get_elem_jacobian(this->_press_var.p(), this->_flow_vars.w());
          }

        // Now a loop over the pressure degrees of freedom.  This
        // computes the contributions of the continuity equation.
        for (unsigned int i=0; i != n_p_dofs; i++)
          {
            Fp(i) += (-U*grad_T/T + divU)*p_phi[i][qp]*JxW[qp];


            if (compute_jacobian)
              {

                for (unsigned int j=0; j!=n_u_dofs; j++)
                  {
                    KPu(i,j) += JxW[qp]*(
                                         +u_gradphi[j][qp](0)*p_phi[i][qp]
                                         -u_phi[j][qp]*p_phi[i][qp]*grad_T(0)/T
                                         );

                    KPv(i,j) += JxW[qp]*(
                                         +u_gradphi[j][qp](1)*p_phi[i][qp]
                                         -u_phi[j][qp]*p_phi[i][qp]*grad_T(1)/T
                                         );

                    if (this->_flow_vars.dim() == 3)
                      {
                        (*KPw)(i,j) += JxW[qp]*(
                                                +u_gradphi[j][qp](2)*p_phi[i][qp]
                                                -u_phi[j][qp]*p_phi[i][qp]*grad_T(2)/T
                                                );
                      }

                  }

                for (unsigned int j=0; j!=n_t_dofs; j++)
                  {
                    KPT(i,j) += JxW[qp]*(
                                         -T_gradphi[j][qp]*U*p_phi[i][qp]/T
                                         +U*p_phi[i][qp]*grad_T*T_phi[j][qp])/(T*T);
                  }
              } // end if compute_jacobian
          } // end p_dofs loop
      } // end qp loop

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_momentum_mass_residual ( bool  compute_jacobian, AssemblyContext &  context  )

protected

Helper function.

Definition at line 735 of file low_mach_navier_stokes.C.

  {
    // Element Jacobian * quadrature weights for interior integration
    const std::vector<libMesh::Real> &JxW =
      context.get_element_fe(this->_flow_vars.u())->get_JxW();

    // The shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& u_phi =
      context.get_element_fe(this->_flow_vars.u())->get_phi();

    // The number of local degrees of freedom in each variable
    const unsigned int n_u_dofs = context.get_dof_indices(this->_flow_vars.u()).size();

    // The subvectors and submatrices we need to fill:
    libMesh::DenseSubVector<libMesh::Real> &F_u = context.get_elem_residual(this->_flow_vars.u());
    libMesh::DenseSubVector<libMesh::Real> &F_v = context.get_elem_residual(this->_flow_vars.v());
    libMesh::DenseSubVector<libMesh::Real>* F_w = NULL;


    if( this->_flow_vars.dim() == 3 )
      F_w  = &context.get_elem_residual(this->_flow_vars.w()); // R_{w}

    unsigned int n_qpoints = context.get_element_qrule().n_points();

    for (unsigned int qp = 0; qp != n_qpoints; ++qp)
      {
        // For the mass residual, we need to be a little careful.
        // The time integrator is handling the time-discretization
        // for us so we need to supply M(u_fixed)*u' for the residual.
        // u_fixed will be given by the fixed_interior_value function
        // while u' will be given by the interior_rate function.
        libMesh::Real u_dot, v_dot, w_dot = 0.0;
        context.interior_rate(this->_flow_vars.u(), qp, u_dot);
        context.interior_rate(this->_flow_vars.v(), qp, v_dot);

        if( this->_flow_vars.dim() == 3 )
          context.interior_rate(this->_flow_vars.w(), qp, w_dot);

        libMesh::Real T = context.fixed_interior_value(this->_temp_vars.T(), qp);

        libMesh::Number rho = this->rho(T, this->get_p0_transient(context, qp));

        for (unsigned int i = 0; i != n_u_dofs; ++i)
          {
            F_u(i) -= rho*u_dot*u_phi[i][qp]*JxW[qp];
            F_v(i) -= rho*v_dot*u_phi[i][qp]*JxW[qp];

            if( this->_flow_vars.dim() == 3 )
              (*F_w)(i) -= rho*w_dot*u_phi[i][qp]*JxW[qp];

            /*
              if( compute_jacobian )
              {
              for (unsigned int j=0; j != n_u_dofs; j++)
              {
              // Assuming rho is constant w.r.t. u, v, w
              // and T (if Boussinesq added).
              libMesh::Real value = JxW[qp]*_rho*u_phi[i][qp]*u_phi[j][qp];

              M_uu(i,j) += value;
              M_vv(i,j) += value;

              if( _dim == 3)
              {
              M_ww(i,j) += value;
              }

              } // End DoF loop j
              } // End Jacobian check
            */

          } // End DoF loop i
      } // End quadrature loop qp

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_momentum_time_deriv ( bool  compute_jacobian, AssemblyContext &  context, const CachedValues &  cache  )

protected

Helper function.

Definition at line 287 of file low_mach_navier_stokes.C.

References GRINS::CachedValues::get_cached_gradient_values(), GRINS::CachedValues::get_cached_values(), GRINS::Cache::PRESSURE, GRINS::Cache::TEMPERATURE, GRINS::Cache::THERMO_PRESSURE, GRINS::Cache::X_VELOCITY, GRINS::Cache::X_VELOCITY_GRAD, GRINS::Cache::Y_VELOCITY, GRINS::Cache::Y_VELOCITY_GRAD, GRINS::Cache::Z_VELOCITY, and GRINS::Cache::Z_VELOCITY_GRAD.

  {
    // The number of local degrees of freedom in each variable.
    const unsigned int n_u_dofs = context.get_dof_indices(this->_flow_vars.u()).size();
    const unsigned int n_p_dofs = context.get_dof_indices(this->_press_var.p()).size();
    const unsigned int n_T_dofs = context.get_dof_indices(this->_temp_vars.T()).size();

    // Check number of dofs is same for _flow_vars.u(), v_var and w_var.
    libmesh_assert (n_u_dofs == context.get_dof_indices(this->_flow_vars.v()).size());

    if (this->_flow_vars.dim() == 3)
      libmesh_assert (n_u_dofs == context.get_dof_indices(this->_flow_vars.w()).size());

    // Element Jacobian * quadrature weights for interior integration.
    const std::vector<libMesh::Real> &JxW =
      context.get_element_fe(this->_flow_vars.u())->get_JxW();

    // The pressure shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& u_phi =
      context.get_element_fe(this->_flow_vars.u())->get_phi();
    const std::vector<std::vector<libMesh::Real> >& p_phi =
      context.get_element_fe(this->_press_var.p())->get_phi();
    const std::vector<std::vector<libMesh::Real> >& T_phi =
      context.get_element_fe(this->_temp_vars.T())->get_phi();

    // The velocity shape function gradients at interior quadrature points.
    const std::vector<std::vector<libMesh::RealGradient> >& u_gradphi =
      context.get_element_fe(this->_flow_vars.u())->get_dphi();

    libMesh::DenseSubVector<libMesh::Number> &Fu = context.get_elem_residual(this->_flow_vars.u()); // R_{u}
    libMesh::DenseSubVector<libMesh::Number> &Fv = context.get_elem_residual(this->_flow_vars.v()); // R_{v}
    libMesh::DenseSubVector<libMesh::Real>* Fw = NULL;

    if( this->_flow_vars.dim() == 3 )
      {
        Fw  = &context.get_elem_residual(this->_flow_vars.w()); // R_{w}
      }

    unsigned int n_qpoints = context.get_element_qrule().n_points();
    for (unsigned int qp=0; qp != n_qpoints; qp++)
      {
        libMesh::Number u, v, p, p0, T;
        u = cache.get_cached_values(Cache::X_VELOCITY)[qp];
        v = cache.get_cached_values(Cache::Y_VELOCITY)[qp];

        T = cache.get_cached_values(Cache::TEMPERATURE)[qp];
        p = cache.get_cached_values(Cache::PRESSURE)[qp];
        p0 = cache.get_cached_values(Cache::THERMO_PRESSURE)[qp];

        libMesh::Gradient grad_u = cache.get_cached_gradient_values(Cache::X_VELOCITY_GRAD)[qp];
        libMesh::Gradient grad_v = cache.get_cached_gradient_values(Cache::Y_VELOCITY_GRAD)[qp];

        libMesh::Gradient grad_w;
        if (this->_flow_vars.dim() == 3)
          grad_w = cache.get_cached_gradient_values(Cache::Z_VELOCITY_GRAD)[qp];

        libMesh::NumberVectorValue grad_uT( grad_u(0), grad_v(0) );
        libMesh::NumberVectorValue grad_vT( grad_u(1), grad_v(1) );
        libMesh::NumberVectorValue grad_wT;
        if( this->_flow_vars.dim() == 3 )
          {
            grad_uT(2) = grad_w(0);
            grad_vT(2) = grad_w(1);
            grad_wT = libMesh::NumberVectorValue( grad_u(2), grad_v(2), grad_w(2) );
          }

        libMesh::NumberVectorValue U(u,v);
        if (this->_flow_vars.dim() == 3)
          U(2) = cache.get_cached_values(Cache::Z_VELOCITY)[qp]; // w

        libMesh::Number divU = grad_u(0) + grad_v(1);
        if (this->_flow_vars.dim() == 3)
          divU += grad_w(2);

        libMesh::Number rho = this->rho( T, p0 );
        libMesh::Number d_rho = this->d_rho_dT( T, p0 );
        libMesh::Number d_mu = this->_mu.deriv(T);

        libMesh::DenseSubMatrix<libMesh::Number> &Kuu = context.get_elem_jacobian(this->_flow_vars.u(), this->_flow_vars.u()); // R_{u},{u}
        libMesh::DenseSubMatrix<libMesh::Number> &Kuv = context.get_elem_jacobian(this->_flow_vars.u(), this->_flow_vars.v()); // R_{u},{v}
        libMesh::DenseSubMatrix<libMesh::Number>* Kuw = NULL;

        libMesh::DenseSubMatrix<libMesh::Number> &Kvu = context.get_elem_jacobian(this->_flow_vars.v(), this->_flow_vars.u()); // R_{v},{u}
        libMesh::DenseSubMatrix<libMesh::Number> &Kvv = context.get_elem_jacobian(this->_flow_vars.v(), this->_flow_vars.v()); // R_{v},{v}
        libMesh::DenseSubMatrix<libMesh::Number>* Kvw = NULL;

        libMesh::DenseSubMatrix<libMesh::Number>* Kwu = NULL;
        libMesh::DenseSubMatrix<libMesh::Number>* Kwv = NULL;
        libMesh::DenseSubMatrix<libMesh::Number>* Kww = NULL;

        libMesh::DenseSubMatrix<libMesh::Number> &Kup = context.get_elem_jacobian(this->_flow_vars.u(), this->_press_var.p()); // R_{u},{p}
        libMesh::DenseSubMatrix<libMesh::Number> &Kvp = context.get_elem_jacobian(this->_flow_vars.v(), this->_press_var.p()); // R_{v},{p}
        libMesh::DenseSubMatrix<libMesh::Number>* Kwp = NULL;

        libMesh::DenseSubMatrix<libMesh::Number> &KuT = context.get_elem_jacobian(this->_flow_vars.u(), this->_temp_vars.T()); // R_{u},{p}
        libMesh::DenseSubMatrix<libMesh::Number> &KvT = context.get_elem_jacobian(this->_flow_vars.v(), this->_temp_vars.T()); // R_{v},{p}
        libMesh::DenseSubMatrix<libMesh::Number>* KwT = NULL;

        if( this->_flow_vars.dim() == 3 )
          {
            Kuw = &context.get_elem_jacobian(this->_flow_vars.u(), this->_flow_vars.w()); // R_{u},{w}
            Kvw = &context.get_elem_jacobian(this->_flow_vars.v(), this->_flow_vars.w()); // R_{v},{w}
            Kwu = &context.get_elem_jacobian(this->_flow_vars.w(), this->_flow_vars.u()); // R_{w},{u};
            Kwv = &context.get_elem_jacobian(this->_flow_vars.w(), this->_flow_vars.v()); // R_{w},{v};
            Kww = &context.get_elem_jacobian(this->_flow_vars.w(), this->_flow_vars.w()); // R_{w},{w}
            Kwp = &context.get_elem_jacobian(this->_flow_vars.w(), this->_press_var.p()); // R_{w},{p}
            KwT = &context.get_elem_jacobian(this->_flow_vars.w(), this->_temp_vars.T()); // R_{w},{T}
          }

        // Now a loop over the pressure degrees of freedom.  This
        // computes the contributions of the continuity equation.
        for (unsigned int i=0; i != n_u_dofs; i++)
          {
            Fu(i) += ( -rho*U*grad_u*u_phi[i][qp]                 // convection term
                       + p*u_gradphi[i][qp](0)                           // pressure term
                       - this->_mu(T)*(u_gradphi[i][qp]*grad_u + u_gradphi[i][qp]*grad_uT
                                       - 2.0/3.0*divU*u_gradphi[i][qp](0) )    // diffusion term
                       + rho*this->_g(0)*u_phi[i][qp]                 // hydrostatic term
                       )*JxW[qp];

            Fv(i) += ( -rho*U*grad_v*u_phi[i][qp]                 // convection term
                       + p*u_gradphi[i][qp](1)                           // pressure term
                       - this->_mu(T)*(u_gradphi[i][qp]*grad_v + u_gradphi[i][qp]*grad_vT
                                       - 2.0/3.0*divU*u_gradphi[i][qp](1) )    // diffusion term
                       + rho*this->_g(1)*u_phi[i][qp]                 // hydrostatic term
                       )*JxW[qp];
            if (this->_flow_vars.dim() == 3)
              {
                (*Fw)(i) += ( -rho*U*grad_w*u_phi[i][qp]                 // convection term
                              + p*u_gradphi[i][qp](2)                           // pressure term
                              - this->_mu(T)*(u_gradphi[i][qp]*grad_w + u_gradphi[i][qp]*grad_wT
                                              - 2.0/3.0*divU*u_gradphi[i][qp](2) )    // diffusion term
                              + rho*this->_g(2)*u_phi[i][qp]                 // hydrostatic term
                              )*JxW[qp];
              }

            if (compute_jacobian && context.get_elem_solution_derivative())
              {
                libmesh_assert (context.get_elem_solution_derivative() == 1.0);

                for (unsigned int j=0; j != n_u_dofs; j++)
                  {
                    //precompute repeated terms
                    libMesh::Number r0 = rho*U*u_phi[i][qp]*u_gradphi[j][qp];
                    libMesh::Number r1 = u_gradphi[i][qp]*u_gradphi[j][qp];
                    libMesh::Number r2 = rho*u_phi[i][qp]*u_phi[j][qp];


                    Kuu(i,j) += JxW[qp]*(
                                         -r0
                                         //-rho*U*u_gradphi[j][qp]*u_phi[i][qp]
                                         -r2*grad_u(0)
                                         //-rho*u_phi[i][qp]*grad_u(0)*u_phi[j][qp]
                                         -this->_mu(T)*(
                                                        r1
                                                        //u_gradphi[i][qp]*u_gradphi[j][qp]
                                                        + u_gradphi[i][qp](0)*u_gradphi[j][qp](0) // transpose
                                                        - 2.0/3.0*u_gradphi[i][qp](0)*u_gradphi[j][qp](0)
                                                        ));
                    Kvv(i,j) += JxW[qp]*(
                                         -r0
                                         //-rho*U*u_gradphi[j][qp]*u_phi[i][qp]
                                         -r2*grad_v(1)
                                         //-rho*u_phi[i][qp]*grad_v(1)*u_phi[j][qp]
                                         -this->_mu(T)*(
                                                        r1
                                                        //u_gradphi[i][qp]*u_gradphi[j][qp]
                                                        + u_gradphi[i][qp](1)*u_gradphi[j][qp](1) // transpose
                                                        - 2.0/3.0*u_gradphi[i][qp](1)*u_gradphi[j][qp](1)
                                                        ));

                    Kuv(i,j) += JxW[qp]*(
                                         +2.0/3.0*this->_mu(T)*u_gradphi[i][qp](0)*u_gradphi[j][qp](1)
                                         -this->_mu(T)*u_gradphi[i][qp](1)*u_gradphi[j][qp](0)
                                         -r2*grad_u(1)
                                         //-rho*u_phi[i][qp]*u_phi[j][qp]*grad_u(1));
                                         );

                    Kvu(i,j) += JxW[qp]*(
                                         +2.0/3.0*this->_mu(T)*u_gradphi[i][qp](1)*u_gradphi[j][qp](0)
                                         -this->_mu(T)*u_gradphi[i][qp](0)*u_gradphi[j][qp](1)
                                         -r2*grad_v(0)
                                         //-rho*u_phi[i][qp]*u_phi[j][qp]*grad_v(0));
                                         );



                    if (this->_flow_vars.dim() == 3)
                      {
                        (*Kuw)(i,j) += JxW[qp]*(
                                                +2.0/3.0*this->_mu(T)*u_gradphi[i][qp](0)*u_gradphi[j][qp](2)
                                                -this->_mu(T)*u_gradphi[i][qp](2)*u_gradphi[j][qp](0)
                                                -r2*grad_u(2)
                                                //-rho*u_phi[i][qp]*u_phi[j][qp]*grad_u(2)
                                                );

                        (*Kvw)(i,j) += JxW[qp]*(
                                                +2.0/3.0*this->_mu(T)*u_gradphi[i][qp](1)*u_gradphi[j][qp](2)
                                                -this->_mu(T)*u_gradphi[i][qp](2)*u_gradphi[j][qp](1)
                                                -r2*grad_v(2)
                                                //-rho*u_phi[i][qp]*u_phi[j][qp]*grad_v(2)
                                                );

                        (*Kwu)(i,j) += JxW[qp]*(
                                                +2.0/3.0*this->_mu(T)*u_gradphi[i][qp](2)*u_gradphi[j][qp](0)
                                                -this->_mu(T)*u_gradphi[i][qp](0)*u_gradphi[j][qp](2)
                                                -r2*grad_w(0)
                                                //-rho*u_phi[i][qp]*u_phi[j][qp]*grad_w(0)
                                                );

                        (*Kwv)(i,j) += JxW[qp]*(
                                                +2.0/3.0*this->_mu(T)*u_gradphi[i][qp](2)*u_gradphi[j][qp](1)
                                                -this->_mu(T)*u_gradphi[i][qp](1)*u_gradphi[j][qp](2)
                                                -r2*grad_w(1)
                                                //-rho*u_phi[i][qp]*u_phi[j][qp]*grad_w(1)
                                                );

                        (*Kww)(i,j) += JxW[qp]*(
                                                -r0
                                                //-rho*U*u_gradphi[j][qp]*u_phi[i][qp]
                                                -r2*grad_w(2)
                                                //-rho*u_phi[i][qp]*grad_w(2)*u_phi[j][qp]
                                                -this->_mu(T)*(
                                                               r1
                                                               //u_gradphi[i][qp]*u_gradphi[j][qp]
                                                               + u_gradphi[i][qp](2)*u_gradphi[j][qp](2) // transpose
                                                               - 2.0/3.0*u_gradphi[i][qp](2)*u_gradphi[j][qp](2)
                                                               ));

                      } // end if _dim==3
                  } // end of the inner dof (j) loop

                for (unsigned int j=0; j!=n_T_dofs; j++)
                  {

                    //precompute repeated term
                    libMesh:: Number r3 = d_rho*u_phi[i][qp]*T_phi[j][qp];

                    // Analytical Jacobains
                    KuT(i,j) += JxW[qp]*(
                                         -r3*U*grad_u
                                         //-d_rho*T_phi[j][qp]*U*grad_u*u_phi[i][qp]
                                         -d_mu*T_phi[j][qp]*grad_u*u_gradphi[i][qp]
                                         -d_mu*T_phi[j][qp]*grad_u(0)*u_gradphi[i][qp](0) // transpose
                                         +2.0/3.0*d_mu*T_phi[j][qp]*divU*u_gradphi[i][qp](0)
                                         +r3*this->_g(0)
                                         //+d_rho*T_phi[j][qp]*this->_g(0)*u_phi[i][qp]
                                         );

                    KvT(i,j) += JxW[qp]*(
                                         -r3*U*grad_v
                                         //-d_rho*T_phi[j][qp]*U*grad_v*u_phi[i][qp]
                                         -d_mu*T_phi[j][qp]*grad_v*u_gradphi[i][qp]
                                         -d_mu*T_phi[j][qp]*grad_v(1)*u_gradphi[i][qp](1) // transpose
                                         +2.0/3.0*d_mu*T_phi[j][qp]*divU*u_gradphi[i][qp](1)
                                         +r3*this->_g(1)
                                         //+d_rho*T_phi[j][qp]*this->_g(1)*u_phi[i][qp]
                                         );

                    if (this->_flow_vars.dim() == 3)
                      {
                        (*KwT)(i,j) += JxW[qp]*(
                                                -r3*U*grad_w
                                                //-d_rho*T_phi[j][qp]*U*grad_v*u_phi[i][qp]
                                                -d_mu*T_phi[j][qp]*grad_w*u_gradphi[i][qp]
                                                -d_mu*T_phi[j][qp]*grad_w(2)*u_gradphi[i][qp](2) // transpose
                                                +2.0/3.0*d_mu*T_phi[j][qp]*divU*u_gradphi[i][qp](2)
                                                +r3*this->_g(2)
                                                //+d_rho*T_phi[j][qp]*this->_g(2)*u_phi[i][qp]
                                                );

                      } // end if _dim==3
                  } // end T_dofs loop

                // Matrix contributions for the up, vp and wp couplings
                for (unsigned int j=0; j != n_p_dofs; j++)
                  {
                    Kup(i,j) += JxW[qp]*p_phi[j][qp]*u_gradphi[i][qp](0);
                    Kvp(i,j) += JxW[qp]*p_phi[j][qp]*u_gradphi[i][qp](1);
                    if (this->_flow_vars.dim() == 3)
                      (*Kwp)(i,j) += JxW[qp]*p_phi[j][qp]*u_gradphi[i][qp](2);
                  } // end of the inner dof (j) loop

              } // end - if (compute_jacobian && context.get_elem_solution_derivative())

          } // end of the outer dof (i) loop
      } // end of the quadrature point (qp) loop

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_thermo_press_elem_time_deriv ( bool  compute_jacobian, AssemblyContext &  context  )

protected

Definition at line 860 of file low_mach_navier_stokes.C.

  {
    // Element Jacobian * quadrature weights for interior integration
    const std::vector<libMesh::Real> &JxW =
      context.get_element_fe(this->_temp_vars.T())->get_JxW();

    // The number of local degrees of freedom in each variable
    const unsigned int n_p0_dofs = context.get_dof_indices(this->_p0_var->p0()).size();

    // The subvectors and submatrices we need to fill:
    libMesh::DenseSubVector<libMesh::Real> &F_p0 = context.get_elem_residual(this->_p0_var->p0());

    unsigned int n_qpoints = context.get_element_qrule().n_points();

    for (unsigned int qp = 0; qp != n_qpoints; ++qp)
      {
        libMesh::Number T;
        T = context.interior_value(this->_temp_vars.T(), qp);

        libMesh::Gradient grad_u, grad_v, grad_w;
        grad_u = context.interior_gradient(this->_flow_vars.u(), qp);
        grad_v = context.interior_gradient(this->_flow_vars.v(), qp);
        if (this->_flow_vars.dim() == 3)
          grad_w = context.interior_gradient(this->_flow_vars.w(), qp);

        libMesh::Number divU = grad_u(0) + grad_v(1);
        if(this->_flow_vars.dim()==3)
          divU += grad_w(2);

        //libMesh::Number cp = this->_cp(T);
        //libMesh::Number cv = cp + this->_R;
        //libMesh::Number gamma = cp/cv;
        //libMesh::Number gamma_ratio = gamma/(gamma-1.0);

        libMesh::Number p0 = context.interior_value( this->_p0_var->p0(), qp );

        for (unsigned int i = 0; i != n_p0_dofs; ++i)
          {
            F_p0(i) += (p0/T - this->_p0/this->_T0)*JxW[qp];
            //F_p0(i) -= p0*gamma_ratio*divU*JxW[qp];
          } // End DoF loop i
      }

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_thermo_press_mass_residual ( bool  compute_jacobian, AssemblyContext &  context  )

protected

Definition at line 958 of file low_mach_navier_stokes.C.

  {
    // The number of local degrees of freedom in each variable.
    const unsigned int n_p0_dofs = context.get_dof_indices(this->_p0_var->p0()).size();
    const unsigned int n_T_dofs = context.get_dof_indices(this->_temp_vars.T()).size();
    const unsigned int n_p_dofs = context.get_dof_indices(this->_press_var.p()).size();

    // Element Jacobian * quadrature weights for interior integration
    const std::vector<libMesh::Real> &JxW =
      context.get_element_fe(this->_temp_vars.T())->get_JxW();

    // The temperature shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& T_phi =
      context.get_element_fe(this->_temp_vars.T())->get_phi();

    // The temperature shape functions at interior quadrature points.
    const std::vector<std::vector<libMesh::Real> >& p_phi =
      context.get_element_fe(this->_press_var.p())->get_phi();

    // The subvectors and submatrices we need to fill:
    libMesh::DenseSubVector<libMesh::Real> &F_p0 = context.get_elem_residual(this->_p0_var->p0());
    libMesh::DenseSubVector<libMesh::Real> &F_T = context.get_elem_residual(this->_temp_vars.T());
    libMesh::DenseSubVector<libMesh::Real> &F_p = context.get_elem_residual(this->_press_var.p());

    unsigned int n_qpoints = context.get_element_qrule().n_points();

    for (unsigned int qp = 0; qp != n_qpoints; ++qp)
      {
        libMesh::Number T;
        T = context.fixed_interior_value(this->_temp_vars.T(), qp);

        libMesh::Number cp = this->_cp(T);
        libMesh::Number cv = cp + this->_R;
        libMesh::Number gamma = cp/cv;
        libMesh::Number one_over_gamma = 1.0/(gamma-1.0);

        libMesh::Number p0_dot;
        context.interior_rate(this->_p0_var->p0(), qp, p0_dot);

        libMesh::Number p0 = context.fixed_interior_value(this->_p0_var->p0(), qp );

        for (unsigned int i=0; i != n_p0_dofs; i++)
          {
            F_p0(i) -= p0_dot*one_over_gamma*JxW[qp];
          }

        for (unsigned int i=0; i != n_T_dofs; i++)
          {
            F_T(i) += p0_dot*T_phi[i][qp]*JxW[qp];
          }

        for (unsigned int i=0; i != n_p_dofs; i++)
          {
            F_p(i) += p0_dot/p0*p_phi[i][qp]*JxW[qp];
          }

      }
    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::assemble_thermo_press_side_time_deriv ( bool  compute_jacobian, AssemblyContext &  context  )

protected

Definition at line 908 of file low_mach_navier_stokes.C.

  {
    // The number of local degrees of freedom in each variable.
    const unsigned int n_p0_dofs = context.get_dof_indices(this->_p0_var->p0()).size();

    // Element Jacobian * quadrature weight for side integration.
    //const std::vector<libMesh::Real> &JxW_side = context.get_side_fe(this->_temp_vars.T())->get_JxW();

    //const std::vector<Point> &normals = context.get_side_fe(this->_temp_vars.T())->get_normals();

    //libMesh::DenseSubVector<libMesh::Number> &F_p0 = context.get_elem_residual(this->_p0_var->p0()); // residual

    // Physical location of the quadrature points
    //const std::vector<libMesh::Point>& qpoint = context.get_side_fe(this->_temp_vars.T())->get_xyz();

    unsigned int n_qpoints = context.get_side_qrule().n_points();
    for (unsigned int qp=0; qp != n_qpoints; qp++)
      {
        /*
          libMesh::Number T = context.side_value( this->_temp_vars.T(), qp );
          libMesh::Gradient U = ( context.side_value( this->_flow_vars.u(), qp ),
          context.side_value( this->_flow_vars.v(), qp ) );
          libMesh::Gradient grad_T = context.side_gradient( this->_temp_vars.T(), qp );


          libMesh::Number p0 = context.side_value( this->_p0_var->p0(), qp );

          libMesh::Number k = this->_k(T);
          libMesh::Number cp = this->_cp(T);

          libMesh::Number cv = cp + this->_R;
          libMesh::Number gamma = cp/cv;
          libMesh::Number gamma_ratio = gamma/(gamma-1.0);
        */

        //std::cout << "U = " << U << std::endl;

        //std::cout << "x = " << qpoint[qp] << ", grad_T = " << grad_T << std::endl;

        for (unsigned int i=0; i != n_p0_dofs; i++)
          {
            //F_p0(i) += (k*grad_T*normals[qp] - p0*gamma_ratio*U*normals[qp]  )*JxW_side[qp];
          }
      }

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::auxiliary_init ( MultiphysicsSystem &  system )

virtual

Any auxillary initialization a Physics class may need.

This is called after all variables are added, so this method can safely query the MultiphysicsSystem about variable information.

Reimplemented from GRINS::Physics.

Definition at line 59 of file low_mach_navier_stokes.C.

  {
    if( _pin_pressure )
      _p_pinning.check_pin_location(system.get_mesh());
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::compute_element_time_derivative_cache ( AssemblyContext &  context )

virtual

Reimplemented from GRINS::Physics.

Definition at line 1020 of file low_mach_navier_stokes.C.

References GRINS::AssemblyContext::get_cached_values(), GRINS::Cache::PRESSURE, GRINS::CachedValues::set_gradient_values(), GRINS::CachedValues::set_values(), GRINS::Cache::TEMPERATURE, GRINS::Cache::TEMPERATURE_GRAD, GRINS::Cache::THERMO_PRESSURE, GRINS::Cache::X_VELOCITY, GRINS::Cache::X_VELOCITY_GRAD, GRINS::Cache::Y_VELOCITY, GRINS::Cache::Y_VELOCITY_GRAD, GRINS::Cache::Z_VELOCITY, and GRINS::Cache::Z_VELOCITY_GRAD.

  {
    CachedValues & cache = context.get_cached_values();

    const unsigned int n_qpoints = context.get_element_qrule().n_points();

    std::vector<libMesh::Real> u, v, w, T, p, p0;
    u.resize(n_qpoints);
    v.resize(n_qpoints);
    if( this->_flow_vars.dim() > 2 )
      w.resize(n_qpoints);

    T.resize(n_qpoints);
    p.resize(n_qpoints);
    p0.resize(n_qpoints);

    std::vector<libMesh::Gradient> grad_u, grad_v, grad_w, grad_T;
    grad_u.resize(n_qpoints);
    grad_v.resize(n_qpoints);
    if( this->_flow_vars.dim() > 2 )
      grad_w.resize(n_qpoints);

    grad_T.resize(n_qpoints);

    for (unsigned int qp = 0; qp != n_qpoints; ++qp)
      {
        u[qp] = context.interior_value(this->_flow_vars.u(), qp);
        v[qp] = context.interior_value(this->_flow_vars.v(), qp);

        grad_u[qp] = context.interior_gradient(this->_flow_vars.u(), qp);
        grad_v[qp] = context.interior_gradient(this->_flow_vars.v(), qp);
        if( this->_flow_vars.dim() > 2 )
          {
            w[qp] = context.interior_value(this->_flow_vars.w(), qp);
            grad_w[qp] = context.interior_gradient(this->_flow_vars.w(), qp);
          }
        T[qp] = context.interior_value(this->_temp_vars.T(), qp);
        grad_T[qp] = context.interior_gradient(this->_temp_vars.T(), qp);

        p[qp] = context.interior_value(this->_press_var.p(), qp);
        p0[qp] = this->get_p0_steady(context, qp);
      }

    cache.set_values(Cache::X_VELOCITY, u);
    cache.set_values(Cache::Y_VELOCITY, v);

    cache.set_gradient_values(Cache::X_VELOCITY_GRAD, grad_u);
    cache.set_gradient_values(Cache::Y_VELOCITY_GRAD, grad_v);

    if(this->_flow_vars.dim() > 2)
      {
        cache.set_values(Cache::Z_VELOCITY, w);
        cache.set_gradient_values(Cache::Z_VELOCITY_GRAD, grad_w);
      }

    cache.set_values(Cache::TEMPERATURE, T);
    cache.set_gradient_values(Cache::TEMPERATURE_GRAD, grad_T);

    cache.set_values(Cache::PRESSURE, p);
    cache.set_values(Cache::THERMO_PRESSURE, p0);
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::compute_postprocessed_quantity ( unsigned int  quantity_index, const AssemblyContext &  context, const libMesh::Point &  point, libMesh::Real &  value  )

virtual

Reimplemented from GRINS::Physics.

Definition at line 1083 of file low_mach_navier_stokes.C.

  {
    if( quantity_index == this->_rho_index )
      {
        libMesh::Real T = this->T(point,context);
        libMesh::Real p0 = this->get_p0_steady(context,point);

        value = this->rho( T, p0 );
      }

    return;
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::element_constraint ( bool  , AssemblyContext &    )

virtual

Constraint part(s) of physics for element interiors.

Reimplemented from GRINS::Physics.

Definition at line 129 of file low_mach_navier_stokes.C.

  {
    // Pin p = p_value at p_point
    if( this->_pin_pressure )
      this->_p_pinning.pin_value( context, compute_jacobian, this->_press_var.p());
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::element_time_derivative ( bool  , AssemblyContext &    )

virtual

Time dependent part(s) of physics for element interiors.

Reimplemented from GRINS::Physics.

Definition at line 115 of file low_mach_navier_stokes.C.

References GRINS::AssemblyContext::get_cached_values().

  {
    const CachedValues & cache = context.get_cached_values();

    this->assemble_mass_time_deriv( compute_jacobian, context, cache );
    this->assemble_momentum_time_deriv( compute_jacobian, context, cache );
    this->assemble_energy_time_deriv( compute_jacobian, context, cache );

    if( this->_enable_thermo_press_calc )
      this->assemble_thermo_press_elem_time_deriv( compute_jacobian, context );
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::init_context ( AssemblyContext &  context )

virtual

Initialize context for added physics variables.

Reimplemented from GRINS::LowMachNavierStokesBase< Viscosity, SpecificHeat, ThermalConductivity >.

Definition at line 95 of file low_mach_navier_stokes.C.

References GRINS::LowMachNavierStokesBase< Viscosity, SpecificHeat, ThermalConductivity >::init_context().

  {
    // First call base class
    LowMachNavierStokesBase<Mu,SH,TC>::init_context(context);

    // We also need the side shape functions, etc.
    context.get_side_fe(this->_flow_vars.u())->get_JxW();
    context.get_side_fe(this->_flow_vars.u())->get_phi();
    context.get_side_fe(this->_flow_vars.u())->get_dphi();
    context.get_side_fe(this->_flow_vars.u())->get_xyz();

    context.get_side_fe(this->_temp_vars.T())->get_JxW();
    context.get_side_fe(this->_temp_vars.T())->get_phi();
    context.get_side_fe(this->_temp_vars.T())->get_dphi();
    context.get_side_fe(this->_temp_vars.T())->get_xyz();
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::mass_residual ( bool  , AssemblyContext &    )

virtual

Mass matrix part(s) for element interiors. All boundary terms lie within the time_derivative part.

Reimplemented from GRINS::Physics.

Definition at line 139 of file low_mach_navier_stokes.C.

  {
    this->assemble_continuity_mass_residual( compute_jacobian, context );

    this->assemble_momentum_mass_residual( compute_jacobian, context );

    this->assemble_energy_mass_residual( compute_jacobian, context );

    if( this->_enable_thermo_press_calc )
      this->assemble_thermo_press_mass_residual( compute_jacobian, context );
  }

template<class Mu , class SH , class TC >

void GRINS::LowMachNavierStokes< Mu, SH, TC >::register_postprocessing_vars ( const GetPot &  input, PostProcessedQuantities< libMesh::Real > &  postprocessing  )

virtual

Register postprocessing variables for LowMachNavierStokes.

Reimplemented from GRINS::Physics.

Definition at line 66 of file low_mach_navier_stokes.C.

References GRINS::PhysicsNaming::low_mach_navier_stokes(), and GRINS::PostProcessedQuantities< NumericType >::register_quantity().

  {
    std::string section = "Physics/"+PhysicsNaming::low_mach_navier_stokes()+"/output_vars";

    if( input.have_variable(section) )
      {
        unsigned int n_vars = input.vector_variable_size(section);

        for( unsigned int v = 0; v < n_vars; v++ )
          {
            std::string name = input(section,"DIE!",v);

            if( name == std::string("rho") )
              {
                this->_rho_index = postprocessing.register_quantity( name );
              }
            else
              {
                std::cerr << "Error: Invalue output_vars value for "+PhysicsNaming::low_mach_navier_stokes() << std::endl
                          << "       Found " << name << std::endl
                          << "       Acceptable values are: rho" << std::endl;
                libmesh_error();
              }
          }
      }
  }

Member Data Documentation

template<class Viscosity , class SpecificHeat , class ThermalConductivity >

PressurePinning GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >::_p_pinning

protected

Definition at line 81 of file low_mach_navier_stokes.h.

template<class Viscosity , class SpecificHeat , class ThermalConductivity >

bool GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >::_pin_pressure

protected

Enable pressure pinning.

Definition at line 79 of file low_mach_navier_stokes.h.

Referenced by GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >::LowMachNavierStokes().

template<class Viscosity , class SpecificHeat , class ThermalConductivity >

unsigned int GRINS::LowMachNavierStokes< Viscosity, SpecificHeat, ThermalConductivity >::_rho_index

protected

Cache index for density post-processing.

Definition at line 84 of file low_mach_navier_stokes.h.

---

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

• src/physics/include/grins/low_mach_navier_stokes.h
• src/physics/src/low_mach_navier_stokes.C