RKSolver Class Reference

RKSolver class solves system of differential equations by Runge-Kutta method. More...

#include <RKSolver.hxx>

Inheritance diagram for RKSolver:

Public Member Functions
	RKSolver ()
	Normal constructor.
	RKSolver (const RKSolver &BS)
	Copy constructor.
	~RKSolver () override
	Destructor.
Mains attributes of the RKSolver
void	Solve (double *N0, double **Matrix, int N, double t1, double t2) override
	Solve Bateman equation using Runge-Kutta method.
Runge-Kutta related methods
Algorithms are taken from Numerical Receipes.
void	RungeKutta (double *YStart, double t1, double t2)
	Runge Kutta calling method.
void	BuildEqns (double t, double *N, double *dNdt)
	Builds the Bateman equations for integration.
void	SetTheMatrix (double **f)
Public Member Functions inherited from BatemanSolver
	BatemanSolver ()
	Normal constructor.
	BatemanSolver (const BatemanSolver &BS)
	Copy constructor.
virtual	~BatemanSolver ()
	Destructor.
int	GetNumberOfEquationSize ()
void	SetNumberOfEquations (int N)
void	SetPrecision (double eps=1e-5)
void	SetForbidNegativeValue ()
	Forbid negative value during integration.
void	SetXe135EquilibriumIndex (int Te135, int I135, int Xe135)
	Set index (position) in EvolvingSystem::fEvolvingAtoms vector for Te-135, I-135 and Xe-135.

Private Attributes
Private methods
double **	fTheMatrix
	The evolution Matrix.

Miscellaneous methods
double	fHestimate
	RK Step estimation.
double	fHmin
	RK minimum Step.
double	fMaxHdid
	store the effective RK max step
double	fMinHdid
	store the effective RK min step
void	RK4 (double *y, double *dydx, double x, double h, double *yout)
	Runge Kutta main method.
void	AdaptStepSize (double *y, double *dydx, double *x, double htry, double eps, double *yscal, double *hdid, double *hnext)
	Adaptative Step Size method for RK.

Additional Inherited Members
Protected Attributes inherited from BatemanSolver
int	fNVar
	The size of the composition vector and /or number of ZAIs involved.
double	fPrecision
	Precision of the solution.
double	fTmin
	The lowest value of the decay constant considered.
bool	fIsNegativeValueAllowed
int	fIdxTe135
	index of Te-135 in the fEvolvingAtoms array
int	fIdxI135
	index of I-135 in the fEvolvingAtoms array
int	fIdxXe135
	index of Xe-135 in the fEvolvingAtoms array

Detailed Description

RKSolver class solves system of differential equations by Runge-Kutta method.

A RKSolver is a base class thatsolves system of 1st order of differential equations.

The differential equations are built in RKSolver::BuildEqns. The RK method is a 4 order adaptative step Runge-Kutta method. Algorithms are taken from Numerical Receipes.

Author

PTO.

MBe.

Version

1.1

Constructor & Destructor Documentation

RKSolver() [1/2]

RKSolver::RKSolver

(

)

Normal constructor.

RKSolver() [2/2]

RKSolver::RKSolver

(

const RKSolver &

BS

)

Copy constructor.

~RKSolver()

RKSolver::~RKSolver ( )

overridedefault

Destructor.

Member Function Documentation

AdaptStepSize()

void RKSolver::AdaptStepSize ( double *  y, double *  dydx, double *  x, double  htry, double  eps, double *  yscal, double *  hdid, double *  hnext  )

protected

Adaptative Step Size method for RK.

Call by RK4

Parameters

y	: initial values to integrate
dydx	: ode's equations (variable is x)
x	: new value of the variable after the adaptative step
htry	: try step size for integration
eps	: precision
yscal	: result after hdid step integration
hdid	: did step size for integration
hnext	: next step size for integration

BuildEqns()

void RKSolver::BuildEqns	(	double	t,
		double *	N,
		double *	dNdt
	)

Builds the Bateman equations for integration.

\f[ \frac{d\vec{N}}{dt}=A\vec{N}\f]

Parameters

t	: time at which the equations are built (needed by RK but useless here)
N	: Nucleus Proportions a time t
dNdt	: ode's for each element.

RK4()

void RKSolver::RK4 ( double *  y, double *  dydx, double  x, double  h, double *  yout  )

protected

Runge Kutta main method.

Call by RungeKutta

Parameters

y	: initial values to integrate
dydx	: ode's equations (variable is x)
x	: variable of integration
h	: step size for integration
yout	: result after integration

RungeKutta()

void RKSolver::RungeKutta	(	double *	YStart,
		double	t1,
		double	t2
	)

Runge Kutta calling method.

Parameters

YStart	: input : the initial condition Y(t1) ; output the final value Y(t2)
t1	: initial time
t2	: final time

SetTheMatrix()

void RKSolver::SetTheMatrix ( double **  f )

inline

Solve()

void RKSolver::Solve ( double *  N0, double **  Matrix, int  N, double  t1, double  t2  )

overridevirtual

Solve Bateman equation using Runge-Kutta method.

For a given vector

of initial composition and a given matrix NxN, integrate Bateman equation from t1 to t2.

Parameters

N0	: initial composition vector
Matrix	: matrix of Bateman eqs
N	: initial composition vector
t1	: initial time
t2	: final time

Implements BatemanSolver.

Member Data Documentation

fHestimate

double RKSolver::fHestimate

protected

RK Step estimation.

fHmin

double RKSolver::fHmin

protected

RK minimum Step.

fMaxHdid

double RKSolver::fMaxHdid

protected

store the effective RK max step

fMinHdid

double RKSolver::fMinHdid

protected

store the effective RK min step

fTheMatrix

double** RKSolver::fTheMatrix

private

The evolution Matrix.

---

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

• /gpr/meplan/SMURE/source/include/RKSolver.hxx
• /gpr/meplan/SMURE/source/src/RKSolver.cxx