HeatSolver Class Reference

#include <solver.hpp>

Collaboration diagram for HeatSolver:

Public Member Functions
	HeatSolver (const PhysParams &phys, const NumParams &num, SchemeKind scheme)
	Construct the solver with physical and numerical parameters and a scheme kind.
void	run (const std::function< void(double, const Grid &, const std::vector< double > &)> &onDump)
	Run the simulation from t=0 to t=tEnd.

Private Member Functions
void	apply_dirichlet (std::vector< double > &T) const
	Apply Dirichlet boundary conditions (Tsur) at both ends of the vector.
void	init_field ()
	Initialize the temperature field with Tin and enforce boundary conditions.
void	bootstrap_if_needed ()
	Perform a warm-up step when the method needs two previous time levels.

Private Attributes
PhysParams	phys_
	Physical parameters (copied for convenience).
NumParams	num_
	Numerical parameters (copied for convenience).
Grid	grid_
	Spatial grid built from phys/num input.
std::unique_ptr< Method >	method_
	Selected time integration method.
bool	uses_prev_step_ = false
	True when the method requires T^{n-1}.
std::vector< double >	T_
	Current solution (typically holds T^n).
std::vector< double >	Tprev_
	Previous solution (stores T^{n-1} when required).
std::vector< double >	next_
	Buffer used to store the next time layer T^{n+1}.

Detailed Description

Definition at line 14 of file solver.hpp.

Constructor & Destructor Documentation

HeatSolver()

HeatSolver::HeatSolver	(	const PhysParams &	phys,
		const NumParams &	num,
		SchemeKind	scheme )

Construct the solver with physical and numerical parameters and a scheme kind.

Parameters

phys	Physical parameters (D [cm^2/h], L [cm], Tin, Tsur).
num	Numerical parameters (dx [cm], dt [h], tEnd [h], outEvery [h]).
scheme	Selected time-stepping scheme (e.g. DuFort-Frankel).

Definition at line 10 of file solver.cpp.

Here is the call graph for this function:

Member Function Documentation

apply_dirichlet()

void HeatSolver::apply_dirichlet ( std::vector< double > & T ) const

private

Apply Dirichlet boundary conditions (Tsur) at both ends of the vector.

Parameters

T	Temperature field updated in-place

Definition at line 34 of file solver.cpp.

Here is the caller graph for this function:

bootstrap_if_needed()

void HeatSolver::bootstrap_if_needed ( )

private

Perform a warm-up step when the method needs two previous time levels.

We call the scheme once with T^{n-1} = T^{n} = T^0 so that T_ holds T^1 while Tprev_ keeps a copy of T^0 (needed by three-level methods).

Definition at line 47 of file solver.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

init_field()

void HeatSolver::init_field ( )

private

Initialize the temperature field with Tin and enforce boundary conditions.

Definition at line 41 of file solver.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

run()

void HeatSolver::run

(

const std::function< void(double, const Grid &, const std::vector< double > &)> &

onDump

)

Run the simulation from t=0 to t=tEnd.

Parameters

onDump

Callback executed at t=0 and then every outEvery hours: onDump(time, grid, T). Use it to write CSV, compute errors, etc.

Definition at line 83 of file solver.cpp.

Here is the call graph for this function:

Here is the caller graph for this function:

Member Data Documentation

grid_

Grid HeatSolver::grid_

private

Spatial grid built from phys/num input.

Definition at line 35 of file solver.hpp.

method_

std::unique_ptr<Method> HeatSolver::method_

private

Selected time integration method.

Definition at line 36 of file solver.hpp.

next_

std::vector<double> HeatSolver::next_

private

Buffer used to store the next time layer T^{n+1}.

Definition at line 41 of file solver.hpp.

num_

NumParams HeatSolver::num_

private

Numerical parameters (copied for convenience).

Definition at line 34 of file solver.hpp.

phys_

PhysParams HeatSolver::phys_

private

Physical parameters (copied for convenience).

Definition at line 33 of file solver.hpp.

T_

std::vector<double> HeatSolver::T_

private

Current solution (typically holds T^n).

Definition at line 39 of file solver.hpp.

Tprev_

std::vector<double> HeatSolver::Tprev_

private

Previous solution (stores T^{n-1} when required).

Definition at line 40 of file solver.hpp.

uses_prev_step_

bool HeatSolver::uses_prev_step_ = false

private

True when the method requires T^{n-1}.

Definition at line 37 of file solver.hpp.

---

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

• include/solver.hpp
• src/solver.cpp