incflo Class Reference

AMR-Wind API: incflo Class Reference

AMR-Wind API v0.1.0

CFD solver for wind plant simulations

Loading...

Searching...

No Matches

#include <incflo.H>

Inheritance diagram for incflo:

[legend]

Collaboration diagram for incflo:

[legend]

Public Member Functions
	incflo ()

	~incflo () override

void	InitData ()

void	Evolve ()

void	ErrorEst (int lev, amrex::TagBoxArray &tags, amrex::Real time, int ngrow) override

void	MakeNewLevelFromScratch (int lev, amrex::Real time, const amrex::BoxArray &new_grids, const amrex::DistributionMapping &new_dmap) override

void	MakeNewLevelFromCoarse (int lev, amrex::Real time, const amrex::BoxArray &ba, const amrex::DistributionMapping &dm) override

void	RemakeLevel (int lev, amrex::Real time, const amrex::BoxArray &ba, const amrex::DistributionMapping &dm) override

void	ClearLevel (int lev) override

void	init_mesh ()

void	init_amr_wind_modules ()

void	prepare_for_time_integration ()

bool	regrid_and_update ()

void	pre_advance_stage1 ()

void	pre_advance_stage2 ()

void	prepare_time_step ()

void	do_advance (const int fixed_point_iteration)

void	advance (const int fixed_point_iteration)

void	prescribe_advance ()

void	post_advance_work ()

amr_wind::CFDSim &	sim ()

const amr_wind::SimTime &	time () const

amr_wind::FieldRepo &	repo ()

const amr_wind::FieldRepo &	repo () const

amr_wind::pde::PDEBase &	icns ()

const amr_wind::pde::PDEBase &	icns () const

amr_wind::pde::PDEMgr::TypeVector &	scalar_eqns ()

const amr_wind::pde::PDEMgr::TypeVector &	scalar_eqns () const

amr_wind::Field &	velocity () const

amr_wind::Field &	density () const

amr_wind::Field &	temperature () const

amr_wind::Field &	pressure () const

amr_wind::Field &	grad_p () const

void	ComputeDt (bool explicit_diffusion)

void	ComputePrescribeDt ()

void	ApplyPredictor (const bool incremental_projection=false, const int fixed_point_iteration=0)

void	ApplyCorrector ()

void	ApplyPrescribeStep ()

void	ApplyProjection (amrex::Vector< amrex::MultiFab const * > density, amrex::Real time, amrex::Real scaling_factor, bool incremental)

void	init_physics_and_pde ()
	Initialize Physics instances as well as PDEs (include turbulence models)

void	ReadCheckpointFile ()

Private Member Functions
amrex::FabFactory< amrex::FArrayBox > const &	Factory (int lev) const noexcept

bool	need_divtau () const

void	CheckAndSetUpDryRun ()

void	ReadParameters ()

void	InitialProjection ()

void	InitialIterations ()

void	PrintMaxValues (const std::string &header)

void	PrintMaxVelLocations (const std::string &header)

void	PrintMaxVel (int lev) const

void	PrintMaxGp (int lev) const

void	CheckForNans (int lev) const

Private Attributes
amr_wind::CFDSim	m_sim

amr_wind::SimTime &	m_time

amr_wind::FieldRepo &	m_repo

amr_wind::OversetOps	m_ovst_ops

std::unique_ptr< amr_wind::RefineCriteriaManager >	m_mesh_refiner

int	m_verbose = 0

bool	m_do_initial_proj = true

int	m_initial_iterations = 3

bool	m_use_godunov = true

bool	m_prescribe_vel = false

bool	m_dry_run = false

int	m_fixed_point_iterations {1}

amrex::Long	m_cell_count {-1}
	number of cells on all levels including covered cells

DiffusionType	m_diff_type = DiffusionType::Crank_Nicolson

bool	m_reconstruct_true_pressure {false}
	reconstruct true pressure

Detailed Description

AMR-Wind driver class

Constructor & Destructor Documentation

◆ incflo()

incflo::incflo ( )

◆ ~incflo()

incflo::~incflo ( )

overridedefault

Member Function Documentation

◆ advance()

void incflo::advance ( const int fixed_point_iteration )

Advance simulation state by one timestep

Performs the following actions at a given timestep

Compute \(\Delta t\)
Advance all computational fields to new timestate in preparation for time integration
Call pre-advance work for all registered physics modules
For Godunov scheme, advance to new time state
For MOL scheme, call predictor corrector steps
Perform any post-advance work

Much of the heavy-lifting is done by incflo::ApplyPredictor and incflo::ApplyCorrector. Please refer to the documentation of those methods for detailed information on the various equations being solved.

Here is the call graph for this function:

◆ ApplyCorrector()

void incflo::ApplyCorrector ( )

Corrector step for MOL scheme

Solve transport equation for momentum and scalars

\begin{align} \left[1 - \kappa \frac{\Delta t}{\rho} \nabla \cdot \left( \mu \nabla \right)\right] u^{*} &= u^n - \Delta t C_u + (1 - \kappa) \frac{\Delta t}{\rho} \nabla \cdot \left( \mu \nabla\right) u^{n} + \frac{\Delta t}{\rho} \left( S_u - \nabla(p + p_0)\right) \\ \end{align}

where
\begin{align} \kappa = \begin{cases} 0 & \text{Explicit} \\ 0.5 & \text{Crank-Nicolson} \\ 1 & \text{Implicit} \end{cases} \end{align}
Apply projection

◆ ApplyPredictor()

void incflo::ApplyPredictor	(	const bool	incremental_projection = false,
		const int	fixed_point_iteration = 0 )

Apply predictor step

For Godunov, this completes the timestep. For MOL, this is the first part of the predictor/corrector within a timestep.

Solve transport equation for momentum and scalars

\begin{align} \left[1 - \kappa \frac{\Delta t}{\rho^{n+1/2}} \nabla \cdot \left( \mu \nabla \right)\right] u^{*} &= u^n - \Delta t (u \cdot \nabla) u + (1 - \kappa) \frac{\Delta t}{\rho^n} \nabla \cdot \left( \mu^{n} \nabla\right) u^{n} + \frac{\Delta t}{\rho^{n+1/2}} \left( S_u - \nabla(p + p_0)\right) \\ \end{align}

where
\begin{align} \kappa = \begin{cases} 0 & \text{Explicit} \\ 0.5 & \text{Crank-Nicolson} \\ 1 & \text{Implicit} \end{cases} \end{align}
Apply projection

◆ ApplyPrescribeStep()

void incflo::ApplyPrescribeStep ( )

◆ ApplyProjection()

void incflo::ApplyProjection	(	amrex::Vector< amrex::MultiFab const * >	density,
		amrex::Real	time,
		amrex::Real	scaling_factor,
		bool	incremental )

Perform nodal projection

Computes the following decomposition:

\begin{align} u + \frac{\Delta t}{\rho} \nabla \phi &= u^{*} & \nabla \cdot u = 0 \end{align}

where \(u^{*}\) is a non-div-free velocity field, stored by components in u, v, and w. The resulting div-free velocity field, u, overwrites the value of u* in u, v, and w.

\(\phi\) is an auxiliary function related to the pressure \(p\) by the relation:

\[ p^{\mathrm{new}} = \phi \]

except in the initial projection when

\[ p^{\mathrm{new}} = p^{\mathrm{old}} + \phi \]

Velocity is updated using the following relation

\begin{align} u^{**} &= u^{*} + \frac{\Delta t}{\rho} \nabla p^{\mathrm{old}} \\ \nabla \cdot \left( \frac{\Delta t}{\rho} \nabla \phi \right) &= \nabla \cdot u^{**} \\ u^{\mathrm{new}} &= u^{**} - \frac{\Delta t}{\rho} \nabla p^\mathrm{new} \end{align}

Notes:

scaling_factor equals \(\Delta t\) except when called during initial projection, when it is 1.0
\(\rho\) in the above expressions is either at state n+1 or n+1/2 depending on whether this method was called from incflo::ApplyPredictor or incflo::ApplyCorrector.
If incremental == true, then the pressure term is not added to \(u^{**}\) and the update is in delta-form.

Please consult AMReX Linear Solvers documentation for more information on the nodal projection operator.

Parameters

density	Vector of multifabs containing the density field at all levels
time	Current time
scaling_factor	Scaling factor \(\Delta t\)
incremental	Flag indicating if it is an incremental projection.

◆ CheckAndSetUpDryRun()

void incflo::CheckAndSetUpDryRun ( )

private

◆ CheckForNans()

void incflo::CheckForNans ( int lev ) const

private

◆ ClearLevel()

void incflo::ClearLevel ( int lev )

override

◆ ComputeDt()

void incflo::ComputeDt ( bool explicit_diffusion )

Estimate the new timestep for adaptive timestepping algorithm

Parameters

explicit_diffusion Flag indicating whether user has selected explicit treatment of diffusion term.

Compute new \(\Delta t\) by using the formula derived in "A Boundary Condition Capturing Method for Multiphase Incompressible Flow" by Kang et al. (JCP).

\begin{align} \text{CFL} &= \frac{\Delta t}{2} \left[\left(C+V \right) + \sqrt{\left(C+V\right)^2 + 4 \left(\frac{|F_x|}{\Delta x} + \frac{|F_y|}{\Delta y} + \frac{|F_z|}{\Delta z} \right)} \right] && \\ C &= \text{max} \left(\frac{|U_x|}{\Delta x} ,\ \frac{|U_y|}{\Delta y} ,\ \frac{|U_z|}{\Delta z}\right) && \text{Convection} \\ V &= 2 \left(\frac{\mu}{\rho}\right)_\mathrm{max} \left[\frac{1}{\Delta x^2} + \frac{1}{\Delta y^2} + \frac{1}{\Delta z^2} \right] && \text{Diffusion} \end{align}

\(F_x, F_y, F_z\) are acceleration due to forcing terms. By default, \(C\) is always used for computing CFL. The term \(V\) is only used when the user has chosen implicit or Crank-Nicolson scheme for diffusion, and contributions from forcing term when time.use_force_cfl is true (default is true).

◆ ComputePrescribeDt()

void incflo::ComputePrescribeDt ( )

◆ density()

amr_wind::Field & incflo::density ( ) const

inline

◆ do_advance()

void incflo::do_advance ( const int fixed_point_iteration )

◆ ErrorEst()

void incflo::ErrorEst	(	int	lev,
		amrex::TagBoxArray &	tags,
		amrex::Real	time,
		int	ngrow )

override

◆ Evolve()

void incflo::Evolve ( )

Perform time-integration for user-defined time or timesteps.

Here is the call graph for this function:

◆ Factory()

amrex::FabFactory< amrex::FArrayBox > const & incflo::Factory ( int lev ) const

inlineprivatenoexcept

◆ grad_p()

amr_wind::Field & incflo::grad_p ( ) const

inline

◆ icns() [1/2]

amr_wind::pde::PDEBase & incflo::icns ( )

inline

◆ icns() [2/2]

const amr_wind::pde::PDEBase & incflo::icns ( ) const

inline

◆ init_amr_wind_modules()

void incflo::init_amr_wind_modules ( )

Initialize AMR-Wind data structures after mesh has been created.

Modules initialized:

Registered Physics models classes
Registered PDE systems
Registered post-processing classes

◆ init_mesh()

void incflo::init_mesh ( )

Initialize AMR mesh data structure.

Calls the AmrMesh/AmrCore functions to initialize the mesh. For restarted simulations, it loads the checkpoint file.

◆ init_physics_and_pde()

void incflo::init_physics_and_pde ( )

Initialize Physics instances as well as PDEs (include turbulence models)

◆ InitData()

void incflo::InitData ( )

Perform all initialization actions for AMR-Wind.

This is a wrapper method that calls the following methods to perform the actual work.

Here is the call graph for this function:

◆ InitialIterations()

void incflo::InitialIterations ( )

private

Perform initial pressure iterations

Performs a user-defined number of iterations to compute the pressure based on the initial conditions. This method is only invoked for new simulations and skipped for restarted simulations from a checkpoint file.

◆ InitialProjection()

void incflo::InitialProjection ( )

private

Ensure initial velocity field is divergence-free.

Performs a projection step to ensure that the user-provided initial velocity field is divergence free. This method is only invoked for new simulations. For restarted simulations using a checkpoint file, this is not necessary.

◆ MakeNewLevelFromCoarse()

void incflo::MakeNewLevelFromCoarse	(	int	lev,
		amrex::Real	time,
		const amrex::BoxArray &	ba,
		const amrex::DistributionMapping &	dm )

override

◆ MakeNewLevelFromScratch()

void incflo::MakeNewLevelFromScratch	(	int	lev,
		amrex::Real	time,
		const amrex::BoxArray &	new_grids,
		const amrex::DistributionMapping &	new_dmap )

override

◆ need_divtau()

bool incflo::need_divtau ( ) const

inlineprivate

◆ post_advance_work()

void incflo::post_advance_work ( )

Perform actions after a timestep

Performs any necessary I/O before advancing to next timestep

◆ pre_advance_stage1()

void incflo::pre_advance_stage1 ( )

◆ pre_advance_stage2()

void incflo::pre_advance_stage2 ( )

◆ prepare_for_time_integration()

void incflo::prepare_for_time_integration ( )

Initialize flow-field before performing time-integration.

This method calls the incflo::InitialProjection step to ensure that the velocity field is divergence free. Then it performs a user-defined number of initial iterations to compute \(p^{n - 1/2}\) necessary for advancing into the first timestep. These actions are only performed when starting a new simulation, but not for a restarted simulation.

◆ prepare_time_step()

void incflo::prepare_time_step ( )

◆ prescribe_advance()

void incflo::prescribe_advance ( )

◆ pressure()

amr_wind::Field & incflo::pressure ( ) const

inline

◆ PrintMaxGp()

void incflo::PrintMaxGp ( int lev ) const

private

◆ PrintMaxValues()

void incflo::PrintMaxValues ( const std::string & header )

private

◆ PrintMaxVel()

void incflo::PrintMaxVel ( int lev ) const

private

◆ PrintMaxVelLocations()

void incflo::PrintMaxVelLocations ( const std::string & header )

private

◆ ReadCheckpointFile()

void incflo::ReadCheckpointFile ( )

◆ ReadParameters()

void incflo::ReadParameters ( )

private

Parse the input file and populate parameters

◆ regrid_and_update()

bool incflo::regrid_and_update ( )

Perform regrid actions at a given timestep.

Returns: Flag indicating if regrid was performed

◆ RemakeLevel()

void incflo::RemakeLevel	(	int	lev,
		amrex::Real	time,
		const amrex::BoxArray &	ba,
		const amrex::DistributionMapping &	dm )

override

◆ repo() [1/2]

amr_wind::FieldRepo & incflo::repo ( )

inline

◆ repo() [2/2]

const amr_wind::FieldRepo & incflo::repo ( ) const

inline

◆ scalar_eqns() [1/2]

amr_wind::pde::PDEMgr::TypeVector & incflo::scalar_eqns ( )

inline

◆ scalar_eqns() [2/2]

const amr_wind::pde::PDEMgr::TypeVector & incflo::scalar_eqns ( ) const

inline

◆ sim()

amr_wind::CFDSim & incflo::sim ( )

inline

◆ temperature()

amr_wind::Field & incflo::temperature ( ) const

inline

◆ time()

const amr_wind::SimTime & incflo::time ( ) const

inline

◆ velocity()

amr_wind::Field & incflo::velocity ( ) const

inline

Member Data Documentation

◆ m_cell_count

amrex::Long incflo::m_cell_count {-1}

private

number of cells on all levels including covered cells

◆ m_diff_type

DiffusionType incflo::m_diff_type = DiffusionType::Crank_Nicolson

private

◆ m_do_initial_proj

bool incflo::m_do_initial_proj = true

private

◆ m_dry_run

bool incflo::m_dry_run = false

private

◆ m_fixed_point_iterations

int incflo::m_fixed_point_iterations {1}

private

◆ m_initial_iterations

int incflo::m_initial_iterations = 3

private

◆ m_mesh_refiner

std::unique_ptr<amr_wind::RefineCriteriaManager> incflo::m_mesh_refiner

private

◆ m_ovst_ops

amr_wind::OversetOps incflo::m_ovst_ops

private

◆ m_prescribe_vel

bool incflo::m_prescribe_vel = false

private

◆ m_reconstruct_true_pressure

bool incflo::m_reconstruct_true_pressure {false}

private

reconstruct true pressure

◆ m_repo

amr_wind::FieldRepo& incflo::m_repo

private

◆ m_sim

amr_wind::CFDSim incflo::m_sim

private

◆ m_time

amr_wind::SimTime& incflo::m_time

private

◆ m_use_godunov

bool incflo::m_use_godunov = true

private

◆ m_verbose

int incflo::m_verbose = 0

private

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

/home/runner/work/amr-wind/amr-wind/amr-wind/incflo.H
/home/runner/work/amr-wind/amr-wind/amr-wind/incflo.cpp
/home/runner/work/amr-wind/amr-wind/amr-wind/incflo_advance.cpp
/home/runner/work/amr-wind/amr-wind/amr-wind/incflo_compute_dt.cpp
/home/runner/work/amr-wind/amr-wind/amr-wind/incflo_regrid.cpp
/home/runner/work/amr-wind/amr-wind/amr-wind/projection/incflo_apply_nodal_projection.cpp
/home/runner/work/amr-wind/amr-wind/amr-wind/setup/init.cpp
/home/runner/work/amr-wind/amr-wind/amr-wind/utilities/diagnostics.cpp
/home/runner/work/amr-wind/amr-wind/amr-wind/utilities/io.cpp

incflo Class Reference

Public Member Functions

Private Member Functions

Private Attributes

Detailed Description

Constructor & Destructor Documentation

◆ incflo()

◆ ~incflo()

Member Function Documentation

◆ advance()

◆ ApplyCorrector()

◆ ApplyPredictor()

◆ ApplyPrescribeStep()

◆ ApplyProjection()

◆ CheckAndSetUpDryRun()

◆ CheckForNans()

◆ ClearLevel()

◆ ComputeDt()

◆ ComputePrescribeDt()

◆ density()

◆ do_advance()

◆ ErrorEst()

◆ Evolve()

◆ Factory()

◆ grad_p()

◆ icns() [1/2]

◆ icns() [2/2]

◆ init_amr_wind_modules()

◆ init_mesh()

◆ init_physics_and_pde()

◆ InitData()

◆ InitialIterations()

◆ InitialProjection()

◆ MakeNewLevelFromCoarse()

◆ MakeNewLevelFromScratch()

◆ need_divtau()

◆ post_advance_work()

◆ pre_advance_stage1()

◆ pre_advance_stage2()

◆ prepare_for_time_integration()

◆ prepare_time_step()

◆ prescribe_advance()

◆ pressure()

◆ PrintMaxGp()

◆ PrintMaxValues()

◆ PrintMaxVel()

◆ PrintMaxVelLocations()

◆ ReadCheckpointFile()

◆ ReadParameters()

◆ regrid_and_update()

◆ RemakeLevel()

◆ repo() [1/2]

◆ repo() [2/2]

◆ scalar_eqns() [1/2]

◆ scalar_eqns() [2/2]

◆ sim()

◆ temperature()

◆ time()

◆ velocity()

Member Data Documentation

◆ m_cell_count

◆ m_diff_type

◆ m_do_initial_proj

◆ m_dry_run

◆ m_fixed_point_iterations

◆ m_initial_iterations

◆ m_mesh_refiner

◆ m_ovst_ops

◆ m_prescribe_vel

◆ m_reconstruct_true_pressure

◆ m_repo

◆ m_sim

◆ m_time

◆ m_use_godunov

◆ m_verbose