Opm::NonlinearSolverEbos< TypeTag, PhysicalModel > Class Template Reference

A nonlinear solver class suitable for general fully-implicit models, as well as pressure, transport and sequential models. More...

#include <NonlinearSolverEbos.hpp>

Classes
struct	SolverParameters

Public Types
enum	RelaxType { Dampen , SOR }

Public Member Functions
	NonlinearSolverEbos (const SolverParameters &param, std::unique_ptr< PhysicalModel > model)
	Construct solver for a given model. More...
SimulatorReportSingle	step (const SimulatorTimerInterface &timer)
const SimulatorReportSingle &	failureReport () const
	return the statistics if the step() method failed
int	linearizations () const
	Number of linearizations used in all calls to step().
int	nonlinearIterations () const
	Number of full nonlinear solver iterations used in all calls to step().
int	linearIterations () const
	Number of linear solver iterations used in all calls to step().
int	wellIterations () const
	Number of well iterations used in all calls to step().
int	nonlinearIterationsLastStep () const
	Number of nonlinear solver iterations used in the last call to step().
int	linearIterationsLastStep () const
	Number of linear solver iterations used in the last call to step().
int	wellIterationsLastStep () const
	Number of well iterations used in all calls to step().
std::vector< std::vector< double > >	computeFluidInPlace (const std::vector< int > &fipnum) const
const PhysicalModel &	model () const
	Reference to physical model.
PhysicalModel &	model ()
	Mutable reference to physical model.
void	detectOscillations (const std::vector< std::vector< double > > &residualHistory, const int it, bool &oscillate, bool &stagnate) const
	Detect oscillation or stagnation in a given residual history.
template<class BVector >
void	stabilizeNonlinearUpdate (BVector &dx, BVector &dxOld, const double omega) const
	Apply a stabilization to dx, depending on dxOld and relaxation parameters. More...
double	relaxMax () const
	The greatest relaxation factor (i.e. smallest factor) allowed.
double	relaxIncrement () const
	The step-change size for the relaxation factor.
enum RelaxType	relaxType () const
	The relaxation type (Dampen or SOR).
double	relaxRelTol () const
	The relaxation relative tolerance.
int	maxIter () const
	The maximum number of nonlinear iterations allowed.
int	minIter () const
	The minimum number of nonlinear iterations allowed.
void	setParameters (const SolverParameters &param)
	Set parameters to override those given at construction time.

Detailed Description

template<class TypeTag, class PhysicalModel>
class Opm::NonlinearSolverEbos< TypeTag, PhysicalModel >

A nonlinear solver class suitable for general fully-implicit models, as well as pressure, transport and sequential models.

Constructor & Destructor Documentation

NonlinearSolverEbos()

template<class TypeTag , class PhysicalModel >

Opm::NonlinearSolverEbos< TypeTag, PhysicalModel >::NonlinearSolverEbos ( const SolverParameters &  param, std::unique_ptr< PhysicalModel >  model  )

inline

Construct solver for a given model.

The model is a std::unique_ptr because the object to which model points to is not allowed to be deleted as long as the NonlinearSolver object exists.

Parameters

[in]	param	parameters controlling nonlinear process
[in,out]	model	physical simulation model.

Member Function Documentation

stabilizeNonlinearUpdate()

template<class TypeTag , class PhysicalModel >

template<class BVector >

void Opm::NonlinearSolverEbos< TypeTag, PhysicalModel >::stabilizeNonlinearUpdate ( BVector &  dx, BVector &  dxOld, const double  omega  ) const

inline

Apply a stabilization to dx, depending on dxOld and relaxation parameters.

Implemention for Dune block vectors.

---

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

• opm/simulators/flow/NonlinearSolverEbos.hpp