GmpFemProcess Class Reference

Basic class for the FEM Process plugin object. More...

#include <gmpFemProcess.h>

Inheritance diagram for GmpFemProcess:

Collaboration diagram for GmpFemProcess:

Public Member Functions
	GmpFemProcess (GmSimulationData *simulation, QString id, QString description, const GmLogCategory &logger)
	Constructor. Will be called by the plugin loading code.
virtual	~GmpFemProcess ()
	Destructor.
virtual const char *	pluginName () const
virtual const char *	pluginType () const
virtual void	loadFunctions (LuaTable &table)
virtual QStringList	functionList () const
Public Member Functions inherited from GmProcess
	GmProcess (GmSimulationData *simulation, QString id, QString description, const GmLogCategory &logger)
virtual const char *	pluginCategory () const
virtual bool	loadPrivateData (LuaTable &table)
virtual void	printParameters (const GmLogCategory &logger)
int	createProgressStatsItem (GmRunProgressStats *ps, const char *name)
Public Member Functions inherited from GmPluginObject
	GmPluginObject (GmSimulationData *simulation, QString id, QString description, const GmLogCategory &logger)
QString	id () const
QString	description () const
QString	pluginTypeName () const

Static Public Member Functions
static int	parseSolverParameters (lua_State *L, GmSimulationData *simData, QString fname, GmElementMesh *&emesh, QVector< GmpFemPhysics * > &physics, GmNumSolver *&solver, GmpFemSolverOptions *solverOptions)
	Parses the set of user parameters, detailed in the solve() method documentation, filling the supplied parameters. On errors, this function calls luaL_error and does not return. On success it returns 0.

Private Member Functions
GmProcessMethodRedirector< GmpFemProcess >::MethodRegistry *	processMethodTable () const
	Returns the table storing the process exported function names / methods.
int	solve (lua_State *L)
	The fem.solve() process implementation.
int	initLinearSolver (lua_State *L)
	Creates and initializes a linear fem solver, intended to be used on a loop as described below. Receives the same set of parameters of the solve() call. On success, returns an opaque solver object.
int	linearStep (lua_State *L)
	Runs a new step of the linear solver created by the call to initLinearSolver(). See comments o that function for the intended use case.
int	linearResidual (lua_State *L)
	Calculates the error residual associated with the last linear step calculated by a call to linearStep(). See comments on initLinearSolver() for the intended use case.
int	runSolver (lua_State *L)
int	initSolver (lua_State *L)
int	stepSolver (lua_State *L)
int	geostatic (lua_State *L)
	Runs a Geostatic step to initialize the in-situ conditions.
int	NLInitSolver (lua_State *L)
	Creates and initializes a nonlinear static fem solver, intended to be used on a loop as described below. Receives the same set of parameters of the solve() call plus a parameter defining if the problem is nonlinear or not. On success, returns a opaque solver object.
int	NLStepSolver (lua_State *L)
	Runs a new step of the nonLinearSolver created by the call to init(). See comments on that function for the intended use case.
int	NLGeostatic (lua_State *L)
	Runs a Geostatic step to initialize the in-situ conditions.
int	initTransientSolver (lua_State *L)
	Creates and initializes a transient fem solver, intended to be used on a loop as described below. Receives the same set of parameters of the solve() call plus a parameter defining if the problem is nonlinear or not. On success, returns a opaque solver object.
int	transientStep (lua_State *L)
	Runs a new step of the transient solver created by the call to initTransientSolver(). See comments on that function for the intended use case.
int	transientLinearStep (lua_State *L)
	Runs a new step of the transient solver created by the call to initTransientSolver(). See comments on that function for the intended use case.
int	transientLinearResidual (lua_State *L)
	Calculates the error residual associated with the last linear step calculated by a call to transientLinearStep(). See comments on initTransientSolver() for the intended use case.
int	release (lua_State *L)
	Releases the fem object created by any of the init functions.
int	update (lua_State *L)
	Updates the fem assembler. Expects as parameters the fem object plus either a string with a single update flag or a table with multiple update flags. If the second parameter is empty, a full update will be done. Accepted flags are: 'nodesAdded', 'dofsAdded', 'dofsRemoved', 'elementsAdded', 'elementsRemoved', 'fixedBcsUpdate', 'fixedBcsChanged', 'fixedBcsAll', 'fixedBcsKeepold'.

Additional Inherited Members
Protected Member Functions inherited from GmProcess
void	registerFunctions (LuaTable &table, typename GmProcessMethodRedirector< T >::MethodRegistry *methods, bool noProgressStats=false)
QStringList	buildFunctionList (typename GmProcessMethodRedirector< T >::MethodRegistry *methods) const
Protected Member Functions inherited from GmPluginObject
virtual const GmLogCategory &	logger () const
GmSimulationData *	simulationData () const

Detailed Description

Basic class for the FEM Process plugin object.

Member Function Documentation

functionList()

virtual QStringList GmpFemProcess::functionList ( ) const

inlinevirtual

Implements GmProcess.

initLinearSolver()

int GmpFemProcess::initLinearSolver ( lua_State * L )

private

Creates and initializes a linear fem solver, intended to be used on a loop as described below. Receives the same set of parameters of the solve() call. On success, returns an opaque solver object.

Intended use:

local solver = fem.initLinearSolver({'physics'}, 'numeric_solver', solverOptions)
while convergence_condition do
  fem.linearStep(solver)
  local r, max, avg = fem.linearResidual(solver, true)
end

initTransientSolver()

int GmpFemProcess::initTransientSolver ( lua_State * L )

private

Creates and initializes a transient fem solver, intended to be used on a loop as described below. Receives the same set of parameters of the solve() call plus a parameter defining if the problem is nonlinear or not. On success, returns a opaque solver object.

Intended use:

-- Usage for linear problems
local solver = fem.initTransientSolver({'physics'}, 'numeric_solver', solverOptions)
while time < last_time do
  fem.transientStep(solver, dt, [updateDof])
end
 
-- Usage for non linear problems
local solver = fem.initTransientSolver({'physics'}, 'numeric_solver', solverOptions, true)
while time < last_time do
  local iter = 1
  while convergence_condition do
    fem.transientLinearStep(solver, dt, iter, [updateDof])
    ...
    local r, max, avg = fem.transientLinearResidual(solver, true)
    iter = iter + 1
  end
end

linearResidual()

int GmpFemProcess::linearResidual ( lua_State * L )

private

Calculates the error residual associated with the last linear step calculated by a call to linearStep(). See comments on initLinearSolver() for the intended use case.

Expects to receive as first parameter the solver created by initLinearSolver().

This function returns the L2 norm of the residual vector. If a second parameter is present and true, it also returns the maximum and the average values among each of the values in the residual vector. The default is false since this option only makes sense if all degrees of freedom in the result space are comparable.

linearStep()

int GmpFemProcess::linearStep ( lua_State * L )

private

Runs a new step of the linear solver created by the call to initLinearSolver(). See comments o that function for the intended use case.

Expects to receive as parameter the solver created by initLinearSolver().

loadFunctions()

virtual void GmpFemProcess::loadFunctions ( LuaTable & table )

inlinevirtual

Implements GmProcess.

NLInitSolver()

int GmpFemProcess::NLInitSolver ( lua_State * L )

private

Creates and initializes a nonlinear static fem solver, intended to be used on a loop as described below. Receives the same set of parameters of the solve() call plus a parameter defining if the problem is nonlinear or not. On success, returns a opaque solver object.

Intended use:

-- Usage for nonlinear static problems
local solver = fem.init({'physics'}, 'numeric_solver', solverOptions)
while time < last_time do
  fem.step(solver, dt, [updateDof])
end

NLStepSolver()

int GmpFemProcess::NLStepSolver ( lua_State * L )

private

Runs a new step of the nonLinearSolver created by the call to init(). See comments on that function for the intended use case.

Expects to receive as parameter the solver created by init() + the time step For solvers created with nonlinear = false only.

pluginName()

virtual const char * GmpFemProcess::pluginName ( ) const

inlinevirtual

Implements GmPluginObject.

pluginType()

virtual const char * GmpFemProcess::pluginType ( ) const

inlinevirtual

Implements GmPluginObject.

solve()

int GmpFemProcess::solve ( lua_State * L )

private

The fem.solve() process implementation.

This function expects to receive as parameters in the Lua stack:

1) A string with the physical method id or a table with multiple method ids. All methods must share the same mesh 2) A string with the solver id that will be used to solve the linear system 3) An optional table with several different solver options regarding printing intermediate matrices and the mode used to handle dofs

transientLinearResidual()

int GmpFemProcess::transientLinearResidual ( lua_State * L )

private

Calculates the error residual associated with the last linear step calculated by a call to transientLinearStep(). See comments on initTransientSolver() for the intended use case.

Expects to receive as first parameter the solver created by initTransientSolver().

This function returns the L2 norm of the residual vector. If a second parameter is present and true, it also returns the maximum and the average values among each of the values in the residual vector. The default is false since this option only makes sense if all degrees of freedom in the result space are comparable.

transientLinearStep()

int GmpFemProcess::transientLinearStep ( lua_State * L )

private

Runs a new step of the transient solver created by the call to initTransientSolver(). See comments on that function for the intended use case.

Expects to receive as parameter the solver created by initTransientSolver() + the time step + the current nonlinear iteration number. For solvers created with nonlinear = true only.

transientStep()

int GmpFemProcess::transientStep ( lua_State * L )

private

Runs a new step of the transient solver created by the call to initTransientSolver(). See comments on that function for the intended use case.

Expects to receive as parameter the solver created by initTransientSolver() + the time step For solvers created with nonlinear = false only.

---

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

• inc/gmpFemProcess.h
• src/gmpFemProcess.cpp