Description

Loads acting on multiple ChLoadable items.

One must inherit from this and implement ComputeQ() directly. The ComputeQ() must write the generalized forces Q into the "load_Q" vector of this object. Given that multiple ChLoadable objects are referenced here, their sub-forces Q are assumed appended in sequence in the "load_Q" vector, in the same order that has been used in the std::vector "mloadables" for ChLoadCustomMultiple creation. The same applies for the order of the sub-matrices of in the Jacobian matrices K, R, M.

#include <ChLoad.h>

Inheritance diagram for chrono::ChLoadCustomMultiple:

[legend]

Collaboration diagram for chrono::ChLoadCustomMultiple:

[legend]

Public Member Functions
	ChLoadCustomMultiple (std::vector< std::shared_ptr< ChLoadable >> &loadable_objects)

	ChLoadCustomMultiple (std::shared_ptr< ChLoadable > loadableA, std::shared_ptr< ChLoadable > loadableB)

	ChLoadCustomMultiple (std::shared_ptr< ChLoadable > loadableA, std::shared_ptr< ChLoadable > loadableB, std::shared_ptr< ChLoadable > loadableC)

virtual int	LoadGetNumCoordsPosLevel () override
	Gets the number of DOFs affected by this load (position part).

virtual int	LoadGetNumCoordsVelLevel () override
	Gets the number of DOFs affected by this load (speed part).

virtual void	LoadGetStateBlock_x (ChState &mD) override
	Gets all the current DOFs packed in a single vector (position part).

virtual void	LoadGetStateBlock_w (ChStateDelta &mD) override
	Gets all the current DOFs packed in a single vector (speed part).

virtual void	LoadStateIncrement (const ChState &x, const ChStateDelta &dw, ChState &x_new) override
	Increment a packed state (e.g., as obtained by LoadGetStateBlock_x()) by a given packed state-delta. More...

virtual int	LoadGetNumFieldCoords () override
	Number of coordinates in the interpolated field. More...

virtual void	ComputeJacobian (ChState state_x, ChStateDelta state_w) override
	Compute Jacobian matrices K=-dQ/dx, R=-dQ/dv, and M=-dQ/da. More...

virtual void	LoadIntLoadResidual_F (ChVectorDynamic<> &R, double c) override
	Add the internal loads Q (pasted at global offsets) into a global vector R, multiplied by a scaling factor c. More...

virtual void	LoadIntLoadResidual_Mv (ChVectorDynamic<> &R, const ChVectorDynamic<> &w, double c) override
	Increment a vector R with the matrix-vector product M*w, scaled by the factor c. More...

virtual void	LoadIntLoadLumpedMass_Md (ChVectorDynamic<> &Md, double &err, const double c) override
	Add the lumped mass to an Md vector, representing a mass diagonal matrix. More...

virtual void	CreateJacobianMatrices () override
	Create the Jacobian loads if needed and set the ChVariables referenced by the sparse KRM block.

virtual ChVectorDynamic &	GetQ ()
	Access the generalized load vector Q.

Public Member Functions inherited from chrono::ChLoadBase
virtual void	ComputeQ (ChState state_x, ChStateDelta state_w)=0
	Compute the generalized load(s). More...

ChLoadJacobians *	GetJacobians ()
	Access the Jacobians (if any, i.e. if this is a stiff load).

virtual void	Update (double time)
	Update, called at least at each time step. More...

virtual bool	IsStiff ()=0
	Report if this is load is stiff. More...

virtual void	InjectKRMMatrices (ChSystemDescriptor &descriptor)
	Register with the given system descriptor any ChKRMBlock objects associated with this item.

virtual void	LoadKRMMatrices (double Kfactor, double Rfactor, double Mfactor)
	Compute and load current stiffnes (K), damping (R), and mass (M) matrices in encapsulated ChKRMBlock objects. More...

Public Member Functions inherited from chrono::ChObj
	ChObj (const ChObj &other)

virtual ChObj *	Clone () const =0
	"Virtual" copy constructor.

int	GetIdentifier () const
	Get the unique integer identifier of this object. More...

void	SetTag (int tag)
	Set an object integer tag (default: -1). More...

int	GetTag () const
	Get the tag of this object.

void	SetName (const std::string &myname)
	Set the name of this object.

const std::string &	GetName () const
	Get the name of this object.

double	GetChTime () const
	Gets the simulation time of this object.

void	SetChTime (double m_time)
	Sets the simulation time of this object.

virtual void	ArchiveOut (ChArchiveOut &archive_out)
	Method to allow serialization of transient data to archives.

virtual void	ArchiveIn (ChArchiveIn &archive_in)
	Method to allow de-serialization of transient data from archives.

virtual std::string &	ArchiveContainerName ()

Public Attributes
std::vector< std::shared_ptr< ChLoadable > >	loadables

ChVectorDynamic	load_Q

Additional Inherited Members
Protected Member Functions inherited from chrono::ChObj
int	GenerateUniqueIdentifier ()

Protected Attributes inherited from chrono::ChLoadBase
ChLoadJacobians *	m_jacobians

Protected Attributes inherited from chrono::ChObj
double	ChTime
	object simulation time

std::string	m_name
	object name

int	m_identifier
	object unique identifier

int	m_tag
	user-supplied tag

Member Function Documentation

◆ ComputeJacobian()

void chrono::ChLoadCustomMultiple::ComputeJacobian	(	ChState *	state_x,
		ChStateDelta *	state_w
	)

overridevirtual

Compute Jacobian matrices K=-dQ/dx, R=-dQ/dv, and M=-dQ/da.

This default implementation uses finite differences for computing the K, R, M matrices if the load is stiff. If possible, a derived class should provide analytical Jacobians. Note: Given that multiple ChLoadable objects are referenced here, sub-matrices of K and R are pasted in (i,j) block positions that reflect the order in which loadable objects were specified.

Parameters

state_x	state position to evaluate Jacobians
state_w	state speed to evaluate Jacobians

Implements chrono::ChLoadBase.

◆ LoadGetNumFieldCoords()

int chrono::ChLoadCustomMultiple::LoadGetNumFieldCoords ( )

overridevirtual

Number of coordinates in the interpolated field.

For example, 3 for a tetrahedron finite element or a cable, 1 for a thermal problem, etc.

Implements chrono::ChLoadBase.

◆ LoadIntLoadLumpedMass_Md()

void chrono::ChLoadCustomMultiple::LoadIntLoadLumpedMass_Md	(	ChVectorDynamic<> &	Md,
		double &	err,
		const double	c
	)

overridevirtual

Add the lumped mass to an Md vector, representing a mass diagonal matrix.

In other words, perform the operation: Md += c*diag(M). Used by lumped explicit integrators. If mass lumping is impossible or approximate, adds scalar error to "err" parameter. If no mass matrix M is present (i.e., no inertial effects), implement as no-op. This default implementation uses the computed Jacobians (if available) to extract M = -dQ/da and then calculates Md += c*diag(M). A derived class should override this method if an analytical expressions for c*M*w is available.

Parameters

Md	result: Md vector, diagonal of the lumped mass matrix
err	result: not touched if lumping does not introduce errors
c	a scaling factor

Implements chrono::ChLoadBase.

◆ LoadIntLoadResidual_F()

void chrono::ChLoadCustomMultiple::LoadIntLoadResidual_F	(	ChVectorDynamic<> &	R,
		double	c
	)

overridevirtual

Add the internal loads Q (pasted at global offsets) into a global vector R, multiplied by a scaling factor c.

In other words, perform the operation: R += forces * c

Implements chrono::ChLoadBase.

◆ LoadIntLoadResidual_Mv()

void chrono::ChLoadCustomMultiple::LoadIntLoadResidual_Mv	(	ChVectorDynamic<> &	R,
		const ChVectorDynamic<> &	w,
		double	c
	)

overridevirtual

Increment a vector R with the matrix-vector product M*w, scaled by the factor c.

In other words, perform the operation: R += c*M*w (i.e., R += c*(-dQ/da)*w). If no mass matrix M is present (i.e., no inertial effects), implement as no-op. This default implementation uses the computed Jacobians (if available) to extract M = -dQ/da and then calculates R += c*M*w. A derived class should override this method if an analytical expressions for c*M*w is available.

Parameters

R	result: the R residual, R += cMw
w	the w vector
c	scaling factor

Implements chrono::ChLoadBase.

◆ LoadStateIncrement()

void chrono::ChLoadCustomMultiple::LoadStateIncrement	(	const ChState &	x,
		const ChStateDelta &	dw,
		ChState &	x_new
	)

overridevirtual

Increment a packed state (e.g., as obtained by LoadGetStateBlock_x()) by a given packed state-delta.

Compute: x_new = x + dw. This method is used in calculating Jacobians with finite difference approximations (default implementations of ComputeJacobian).

Implements chrono::ChLoadBase.

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

/builds/uwsbel/chrono/src/chrono/physics/ChLoad.h
/builds/uwsbel/chrono/src/chrono/physics/ChLoad.cpp

Description

Public Member Functions

Public Attributes

Additional Inherited Members

Member Function Documentation

◆ ComputeJacobian()

◆ LoadGetNumFieldCoords()

◆ LoadIntLoadLumpedMass_Md()

◆ LoadIntLoadResidual_F()

◆ LoadIntLoadResidual_Mv()

◆ LoadStateIncrement()