Base class for FEM element implementations. More...

#include <FiniteElementBase.hpp>

Inheritance diagram for geos::finiteElement::FiniteElementBase:

[legend]

Classes
struct	FunctionSpaceHelper
	An helper struct to determine the function space. More...

struct	MeshData
	Variables used to initialize the class. More...

struct	StackVariables
	Kernel variables allocated on the stack. More...

Public Member Functions
	FiniteElementBase ()=default
	Default Constructor.

GEOS_HOST_DEVICE	FiniteElementBase (FiniteElementBase const &source)
	Copy Constructor. More...

	FiniteElementBase (FiniteElementBase &&)=default
	Default Move constructor.

FiniteElementBase &	operator= (FiniteElementBase const &)=delete
	Deleted copy assignment operator. More...

FiniteElementBase &	operator= (FiniteElementBase &&)=delete
	Deleted move assignment operator. More...

virtual GEOS_HOST_DEVICE	~FiniteElementBase ()
	Destructor.

template<typename LEAF , typename SUBREGION_TYPE >
GEOS_HOST_DEVICE void	setup (localIndex const &cellIndex, typename LEAF::template MeshData< SUBREGION_TYPE > const &meshData, typename LEAF::StackVariables &stack) const
	Abstract setup method, possibly computing cell-dependent properties. More...

virtual GEOS_HOST_DEVICE localIndex	getNumQuadraturePoints () const =0
	Virtual getter for the number of quadrature points per element. More...

virtual GEOS_HOST_DEVICE localIndex	getNumSupportPoints () const =0
	Virtual getter for the number of support points per element. More...

template<typename LEAF >
GEOS_HOST_DEVICE localIndex	numSupportPoints (typename LEAF::StackVariables const &stack) const
	Getter for the number of support points per element. More...

virtual GEOS_HOST_DEVICE localIndex	getMaxSupportPoints () const =0
	Get the maximum number of support points for this element. More...

template<typename LEAF >
GEOS_HOST_DEVICE real64	getGradN (localIndex const k, localIndex const q, real64 const (&X)[LEAF::maxSupportPoints][3], real64(&gradN)[LEAF::maxSupportPoints][3]) const
	Get the shape function gradients. More...

template<typename LEAF >
GEOS_HOST_DEVICE real64	getGradN (localIndex const k, localIndex const q, real64 const (&X)[LEAF::maxSupportPoints][3], typename LEAF::StackVariables const &stack, real64(&gradN)[LEAF::maxSupportPoints][3]) const
	Get the shape function gradients. More...

template<typename LEAF >
GEOS_HOST_DEVICE real64	getGradN (localIndex const k, localIndex const q, int const X, real64(&gradN)[LEAF::maxSupportPoints][3]) const
	Get the shape function gradients. More...

template<typename LEAF >
GEOS_HOST_DEVICE real64	getGradN (localIndex const k, localIndex const q, int const X, typename LEAF::StackVariables const &stack, real64(&gradN)[LEAF::maxSupportPoints][3]) const
	Get the shape function gradients. More...

template<typename LEAF , localIndex NUMDOFSPERTRIALSUPPORTPOINT, bool UPPER = false>
GEOS_HOST_DEVICE void	addGradGradStabilizationMatrix (typename LEAF::StackVariables const &stack, real64(&matrix)[LEAF::maxSupportPoints NUMDOFSPERTRIALSUPPORTPOINT][LEAF::maxSupportPoints NUMDOFSPERTRIALSUPPORTPOINT], real64 const scaleFactor=1.0) const
	Add stabilization of grad-grad bilinear form to input matrix. More...

template<typename LEAF , localIndex NUMDOFSPERTRIALSUPPORTPOINT>
GEOS_HOST_DEVICE GEOS_FORCE_INLINE void	addEvaluatedGradGradStabilizationVector (typename LEAF::StackVariables const &stack, real64 const (&dofs)[LEAF::maxSupportPoints][NUMDOFSPERTRIALSUPPORTPOINT], real64(&targetVector)[LEAF::maxSupportPoints][NUMDOFSPERTRIALSUPPORTPOINT], real64 const scaleFactor=1.0) const
	Add a grad-grad stabilization operator evaluated at a provided vector of dofs to input vector. More...

Static Public Member Functions
template<typename SUBREGION_TYPE >
static void	fillMeshData (NodeManager const &nodeManager, EdgeManager const &edgeManager, FaceManager const &faceManager, SUBREGION_TYPE const &cellSubRegion, MeshData< SUBREGION_TYPE > &meshData)
	Method to fill a MeshData object. More...

template<typename LEAF , typename SUBREGION_TYPE >
static void	initialize (NodeManager const &nodeManager, EdgeManager const &edgeManager, FaceManager const &faceManager, SUBREGION_TYPE const &cellSubRegion, typename LEAF::template MeshData< SUBREGION_TYPE > &meshData)
	Abstract initialization method. More...

template<typename SUBREGION_TYPE >
GEOS_HOST_DEVICE static GEOS_FORCE_INLINE void	setupStack (localIndex const &cellIndex, MeshData< SUBREGION_TYPE > const &meshData, StackVariables &stack)
	Empty setup method. More...

template<int N>
constexpr static GEOS_HOST_DEVICE PDEUtilities::FunctionSpace	getFunctionSpace ()
	Getter for the function space. More...

template<localIndex NUMDOFSPERTRIALSUPPORTPOINT, localIndex MAXSUPPORTPOINTS, bool UPPER>
GEOS_HOST_DEVICE static GEOS_FORCE_INLINE void	addGradGradStabilization (StackVariables const &stack, real64(&matrix)[MAXSUPPORTPOINTS NUMDOFSPERTRIALSUPPORTPOINT][MAXSUPPORTPOINTS NUMDOFSPERTRIALSUPPORTPOINT], real64 const &scaleFactor)
	Empty method, here for compatibility with methods that require a stabilization of the grad-grad bilinear form. More...

template<localIndex NUMDOFSPERTRIALSUPPORTPOINT, localIndex MAXSUPPORTPOINTS>
GEOS_HOST_DEVICE static GEOS_FORCE_INLINE void	addEvaluatedGradGradStabilization (StackVariables const &stack, real64 const (&dofs)[MAXSUPPORTPOINTS][NUMDOFSPERTRIALSUPPORTPOINT], real64(&targetVector)[MAXSUPPORTPOINTS][NUMDOFSPERTRIALSUPPORTPOINT], real64 const scaleFactor)
	Empty method, here for compatibility with methods that require a stabilization of the grad-grad bilinear form. More...

Value Operator Functions
template<int NUM_SUPPORT_POINTS>
static GEOS_HOST_DEVICE void	value (real64 const (&N)[NUM_SUPPORT_POINTS], real64 const (&var)[NUM_SUPPORT_POINTS], real64 &value)
	Compute the interpolated value of a variable. More...

template<int NUM_SUPPORT_POINTS, int NUM_COMPONENTS>
static GEOS_HOST_DEVICE void	value (real64 const (&N)[NUM_SUPPORT_POINTS], real64 const (&var)[NUM_SUPPORT_POINTS][NUM_COMPONENTS], real64(&value)[NUM_COMPONENTS])
	Compute the interpolated value of a vector variable. More...

Gradient Operator Functions
template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE void	symmetricGradient (GRADIENT_TYPE const &gradN, real64 const (&var)[NUM_SUPPORT_POINTS][3], real64(&gradVar)[6])
	Calculate the symmetric gradient of a vector valued support field at a point using the stored basis function gradients for all support points. More...

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE real64	symmetricGradientTrace (GRADIENT_TYPE const &gradN, real64 const (&var)[NUM_SUPPORT_POINTS][3])
	Calculate the trace of the symmetric gradient of a vector valued support field (i.e. the volumetric strain for the displacement field) at a point using the stored basis function gradients for all support points. More...

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE void	gradient (GRADIENT_TYPE const &gradN, real64 const (&var)[NUM_SUPPORT_POINTS], real64(&gradVar)[3])
	Calculate the gradient of a scalar valued support field at a point using the input basis function gradients. More...

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE void	gradient (GRADIENT_TYPE const &gradN, real64 const (&var)[NUM_SUPPORT_POINTS][3], real64(&gradVar)[3][3])
	Calculate the gradient of a vector valued support field at a point using the input basis function gradients. More...

Multi-Operator Functions
template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE void	valueAndGradient (real64 const (&N)[NUM_SUPPORT_POINTS], GRADIENT_TYPE const &gradN, real64 const (&var)[NUM_SUPPORT_POINTS], real64 &value, real64(&gradVar)[3])
	Calculate the value and gradient of a scalar valued support field at a point using the input basis function gradients. More...

Static Public Attributes
constexpr static int	numSamplingPointsPerDirection = 10
	Number of sampling points.

Scattering Operator Functions
These functions take quadrature data and map it to the support points through some operator.
arrayView4d< real64 const >	m_viewGradN
	View to potentially hold pre-calculated shape function gradients.

arrayView2d< real64 const >	m_viewDetJ

void	setGradNView (arrayView4d< real64 const > const &source)
	Sets m_viewGradN equal to an input view. More...

void	setDetJView (arrayView2d< real64 const > const &source)
	Sets m_viewDetJ equal to an input view. More...

arrayView4d< real64 const >	getGradNView () const
	Getter for m_viewGradN. More...

arrayView2d< real64 const >	getDetJView () const
	Getter for m_viewDetJ. More...

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE void	plusGradNajAij (GRADIENT_TYPE const &gradN, real64 const (&var_detJxW)[6], real64(&R)[NUM_SUPPORT_POINTS][3])
	Inner product of each basis function gradient with a rank-2 symmetric tensor. More...

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE void	plusGradNajAij (GRADIENT_TYPE const &gradN, real64 const (&var_detJxW)[3][3], real64(&R)[NUM_SUPPORT_POINTS][3])
	Inner product of each basis function gradient with a rank-2 symmetric tensor. More...

template<int NUM_SUPPORT_POINTS>
static GEOS_HOST_DEVICE void	plusNaFi (real64 const (&N)[NUM_SUPPORT_POINTS], real64 const (&forcingTerm_detJxW)[3], real64(&R)[NUM_SUPPORT_POINTS][3])
	Product of each shape function with a vector forcing term. More...

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE void	plusGradNajAijPlusNaFi (GRADIENT_TYPE const &gradN, real64 const (&var_detJxW)[3][3], real64 const (&N)[NUM_SUPPORT_POINTS], real64 const (&forcingTerm_detJxW)[3], real64(&R)[NUM_SUPPORT_POINTS][3])
	Inner product of each basis function gradient with a rank-2 symmetric tensor added to the product each shape function with a vector. More...

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >
static GEOS_HOST_DEVICE void	plusGradNajAijPlusNaFi (GRADIENT_TYPE const &gradN, real64 const (&var_detJxW)[6], real64 const (&N)[NUM_SUPPORT_POINTS], real64 const (&forcingTerm_detJxW)[3], real64(&R)[NUM_SUPPORT_POINTS][3])
	Inner product of each basis function gradient with a rank-2 tensor added to the product each shape function with a vector. More...

Detailed Description

Base class for FEM element implementations.

Definition at line 45 of file FiniteElementBase.hpp.

Constructor & Destructor Documentation

◆ FiniteElementBase()

GEOS_HOST_DEVICE geos::finiteElement::FiniteElementBase::FiniteElementBase ( FiniteElementBase const & source )

inline

Copy Constructor.

Parameters

source The object to copy.

Definition at line 61 of file FiniteElementBase.hpp.

Member Function Documentation

◆ addEvaluatedGradGradStabilization()

template<localIndex NUMDOFSPERTRIALSUPPORTPOINT, localIndex MAXSUPPORTPOINTS>

GEOS_HOST_DEVICE static GEOS_FORCE_INLINE void geos::finiteElement::FiniteElementBase::addEvaluatedGradGradStabilization	(	StackVariables const &	stack,
		real64 const (&)	dofs[MAXSUPPORTPOINTS][NUMDOFSPERTRIALSUPPORTPOINT],
		real64(&)	targetVector[MAXSUPPORTPOINTS][NUMDOFSPERTRIALSUPPORTPOINT],
		real64 const	scaleFactor
	)

inlinestatic

Empty method, here for compatibility with methods that require a stabilization of the grad-grad bilinear form.

This method is intended to be used with targetVector being the residual and dofs being the degrees of freedom of the previous solution.

Template Parameters

NUMDOFSPERTRIALSUPPORTPOINT Number of degrees of freedom for each support point.

Parameters

stack	Stack variables as filled by setupStack.
dofs	The degrees of freedom of the function where the stabilization operator has to be evaluated.
targetVector	The input vector to which values have to be added, seen in chunks of length `NUMDOFSPERTRIALSUPPORTPOINT`.
scaleFactor	Scaling of the stabilization matrix.

Definition at line 407 of file FiniteElementBase.hpp.

◆ addEvaluatedGradGradStabilizationVector()

template<typename LEAF , localIndex NUMDOFSPERTRIALSUPPORTPOINT>

GEOS_HOST_DEVICE GEOS_FORCE_INLINE void geos::finiteElement::FiniteElementBase::addEvaluatedGradGradStabilizationVector	(	typename LEAF::StackVariables const &	stack,
		real64 const (&)	dofs[LEAF::maxSupportPoints][NUMDOFSPERTRIALSUPPORTPOINT],
		real64(&)	targetVector[LEAF::maxSupportPoints][NUMDOFSPERTRIALSUPPORTPOINT],
		real64 const	scaleFactor = `1.0`
	)		const

inline

Add a grad-grad stabilization operator evaluated at a provided vector of dofs to input vector.

This method is used to modify a residual consistently when the jacobian includes a stabilization term.

Template Parameters

LEAF	Type of the derived finite element implementation.
NUMDOFSPERTRIALSUPPORTPOINT	Number of degrees of freedom for each support point.

Parameters

stack	Stack variables created by a call to setup.
dofs	The vector of dofs to evaluate the stabilization.
targetVector	The input vector to which values have to be added, seen in chunks of length `NUMDOFSPERTRIALSUPPORTPOINT`.
scaleFactor	Optional scaling of the stabilization matrix. Defaults to 1.0.

Definition at line 435 of file FiniteElementBase.hpp.

◆ addGradGradStabilization()

template<localIndex NUMDOFSPERTRIALSUPPORTPOINT, localIndex MAXSUPPORTPOINTS, bool UPPER>

GEOS_HOST_DEVICE static GEOS_FORCE_INLINE void geos::finiteElement::FiniteElementBase::addGradGradStabilization	(	StackVariables const &	stack,
		real64(&)	matrix[MAXSUPPORTPOINTS NUMDOFSPERTRIALSUPPORTPOINT][MAXSUPPORTPOINTS NUMDOFSPERTRIALSUPPORTPOINT],
		real64 const &	scaleFactor
	)

inlinestatic

Empty method, here for compatibility with methods that require a stabilization of the grad-grad bilinear form.

Template Parameters

NUMDOFSPERTRIALSUPPORTPOINT	Number of degrees of freedom for each support point.
MAXSUPPORTPOINTS	Maximum number of support points allowed for this element.
UPPER	If true only the upper triangular part of `matrix` is modified.

Parameters

stack	Stack variables as filled by setupStack.
matrix	The matrix that needs to be stabilized.
scaleFactor	Scaling of the stabilization matrix.

Definition at line 355 of file FiniteElementBase.hpp.

◆ addGradGradStabilizationMatrix()

template<typename LEAF , localIndex NUMDOFSPERTRIALSUPPORTPOINT, bool UPPER = false>

GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::addGradGradStabilizationMatrix	(	typename LEAF::StackVariables const &	stack,
		real64(&)	matrix[LEAF::maxSupportPoints NUMDOFSPERTRIALSUPPORTPOINT][LEAF::maxSupportPoints NUMDOFSPERTRIALSUPPORTPOINT],
		real64 const	scaleFactor = `1.0`
	)		const

inline

Add stabilization of grad-grad bilinear form to input matrix.

Template Parameters

LEAF	Type of the derived finite element implementation.
NUMDOFSPERTRIALSUPPORTPOINT	Number of degrees of freedom for each support point.
UPPER	If true only the upper triangular part of `matrix` is modified.

Parameters

stack	Stack variables created by a call to setup.
matrix	The input matrix to which values have to be added.
scaleFactor	Optional scaling of the stabilization matrix. Defaults to 1.0.

Definition at line 378 of file FiniteElementBase.hpp.

◆ fillMeshData()

template<typename SUBREGION_TYPE >

static void geos::finiteElement::FiniteElementBase::fillMeshData	(	NodeManager const &	nodeManager,
		EdgeManager const &	edgeManager,
		FaceManager const &	faceManager,
		SUBREGION_TYPE const &	cellSubRegion,
		MeshData< SUBREGION_TYPE > &	meshData
	)

inlinestatic

Method to fill a MeshData object.

Parameters

nodeManager	The node manager.
edgeManager	The edge manager.
faceManager	The face manager.
cellSubRegion	The cell sub-region for which the element has to be initialized.
meshData	MeshData struct to be filled.

Definition at line 136 of file FiniteElementBase.hpp.

◆ getDetJView()

arrayView2d< real64 const > geos::finiteElement::FiniteElementBase::getDetJView ( ) const

inline

Getter for m_viewDetJ.

Returns: A new arrayView copy of m_viewDetJ.

Definition at line 759 of file FiniteElementBase.hpp.

◆ getFunctionSpace()

template<int N>

constexpr static GEOS_HOST_DEVICE PDEUtilities::FunctionSpace geos::finiteElement::FiniteElementBase::getFunctionSpace ( )

staticconstexpr

Getter for the function space.

Template Parameters

The	number of components per support point (i.e., 1 if scalar variable, 3 if vector variable)

Returns: The function space.

◆ getGradN() [1/4]

template<typename LEAF >

GEOS_HOST_DEVICE real64 geos::finiteElement::FiniteElementBase::getGradN	(	localIndex const	k,
		localIndex const	q,
		int const	X,
		real64(&)	gradN[LEAF::maxSupportPoints][3]
	)		const

Get the shape function gradients.

Template Parameters

LEAF	Type of the derived finite element implementation.

Parameters

k	The element index.
q	The quadrature point index.
X	dummy variable.
gradN	Return array of the shape function gradients.

Returns: The determinant of the Jacobian transformation matrix.

This function returns pre-calculated shape function gradients.

◆ getGradN() [2/4]

template<typename LEAF >

GEOS_HOST_DEVICE real64 geos::finiteElement::FiniteElementBase::getGradN	(	localIndex const	k,
		localIndex const	q,
		int const	X,
		typename LEAF::StackVariables const &	stack,
		real64(&)	gradN[LEAF::maxSupportPoints][3]
	)		const

Get the shape function gradients.

Template Parameters

LEAF	Type of the derived finite element implementation.

Parameters

k	The element index.
q	The quadrature point index.
X	dummy variable.
stack	Stack variables relative to the element `k` created by a call to setup.
gradN	Return array of the shape function gradients.

Returns: The determinant of the Jacobian transformation matrix.

This function returns pre-calculated shape function gradients.

◆ getGradN() [3/4]

template<typename LEAF >

GEOS_HOST_DEVICE real64 geos::finiteElement::FiniteElementBase::getGradN	(	localIndex const	k,
		localIndex const	q,
		real64 const (&)	X[LEAF::maxSupportPoints][3],
		real64(&)	gradN[LEAF::maxSupportPoints][3]
	)		const

Get the shape function gradients.

Template Parameters

LEAF	Type of the derived finite element implementation.

Parameters

k	The element index.
q	The quadrature point index.
X	Array of coordinates as the reference for the gradients.
gradN	Return array of the shape function gradients.

Returns: The determinant of the Jacobian transformation matrix.

This function calls the function to calculate shape function gradients.

◆ getGradN() [4/4]

template<typename LEAF >

GEOS_HOST_DEVICE real64 geos::finiteElement::FiniteElementBase::getGradN	(	localIndex const	k,
		localIndex const	q,
		real64 const (&)	X[LEAF::maxSupportPoints][3],
		typename LEAF::StackVariables const &	stack,
		real64(&)	gradN[LEAF::maxSupportPoints][3]
	)		const

Get the shape function gradients.

Template Parameters

LEAF	Type of the derived finite element implementation.

Parameters

k	The element index.
q	The quadrature point index.
X	Array of coordinates as the reference for the gradients.
stack	Stack variables relative to the element `k` created by a call to setup.
gradN	Return array of the shape function gradients.

Returns: The determinant of the Jacobian transformation matrix.

This function calls the function to calculate shape function gradients.

◆ getGradNView()

arrayView4d< real64 const > geos::finiteElement::FiniteElementBase::getGradNView ( ) const

inline

Getter for m_viewGradN.

Returns: A new arrayView copy of m_viewGradN.

Definition at line 750 of file FiniteElementBase.hpp.

◆ getMaxSupportPoints()

virtual GEOS_HOST_DEVICE localIndex geos::finiteElement::FiniteElementBase::getMaxSupportPoints ( ) const

pure virtual

Get the maximum number of support points for this element.

This should be used to know the size of pre-allocated objects whose size depend on the number of support points.

Returns: The number of maximum support points for this element.

◆ getNumQuadraturePoints()

virtual GEOS_HOST_DEVICE localIndex geos::finiteElement::FiniteElementBase::getNumQuadraturePoints ( ) const

pure virtual

Virtual getter for the number of quadrature points per element.

Returns: The number of quadrature points per element.

◆ getNumSupportPoints()

virtual GEOS_HOST_DEVICE localIndex geos::finiteElement::FiniteElementBase::getNumSupportPoints ( ) const

pure virtual

Virtual getter for the number of support points per element.

Returns: The number of support points per element.

◆ gradient() [1/2]

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::gradient	(	GRADIENT_TYPE const &	gradN,
		real64 const (&)	var[NUM_SUPPORT_POINTS],
		real64(&)	gradVar[3]
	)

static

Calculate the gradient of a scalar valued support field at a point using the input basis function gradients.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape function gradients.

Parameters

gradN	The basis function gradients at a point in the element.
var	The vector valued support field that the gradient operator will be applied to.
gradVar	The gradient.

More precisely, the operator is defined as:

$grad_{j} = \sum_a^{nSupport} \left ( \frac{\partial N_a}{\partial X_j} var_{a}\right ),$

◆ gradient() [2/2]

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::gradient	(	GRADIENT_TYPE const &	gradN,
		real64 const (&)	var[NUM_SUPPORT_POINTS][3],
		real64(&)	gradVar[3][3]
	)

static

Calculate the gradient of a vector valued support field at a point using the input basis function gradients.

Calculate the gradient of a scalar valued support field at a point using the input basis function gradients.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape function gradients.

Parameters

gradN	The basis function gradients at a point in the element.
var	The vector valued support field that the gradient operator will be applied to.
gradVar	The gradient.

More precisely, the operator is defined as:

$grad_{j} = \sum_a^{nSupport} \left ( \frac{\partial N_a}{\partial X_j} var_{a}\right ),$

More precisely, the operator is defined as:

$grad_{ij} = \sum_a^{nSupport} \left ( \frac{\partial N_a}{\partial X_j} var_{ai}\right ),$

◆ initialize()

template<typename LEAF , typename SUBREGION_TYPE >

static void geos::finiteElement::FiniteElementBase::initialize	(	NodeManager const &	nodeManager,
		EdgeManager const &	edgeManager,
		FaceManager const &	faceManager,
		SUBREGION_TYPE const &	cellSubRegion,
		typename LEAF::template MeshData< SUBREGION_TYPE > &	meshData
	)

inlinestatic

Abstract initialization method.

It calls the fillMeshData method of the specific element implementation.

Template Parameters

LEAF	Type of the derived finite element implementation.

Parameters

nodeManager	The node manager.
edgeManager	The edge manager.
faceManager	The face manager.
cellSubRegion	The cell sub-region for which the element has to be initialized.
meshData	The struct to be filled according to the `LEAF` class needs.

Definition at line 160 of file FiniteElementBase.hpp.

◆ numSupportPoints()

template<typename LEAF >

GEOS_HOST_DEVICE localIndex geos::finiteElement::FiniteElementBase::numSupportPoints ( typename LEAF::StackVariables const & stack ) const

inline

Getter for the number of support points per element.

Template Parameters

LEAF	Type of the derived finite element implementation.

Parameters

stack Stack variables created by a call to setup.

Returns: The number of support points per element.

Definition at line 252 of file FiniteElementBase.hpp.

◆ operator=() [1/2]

FiniteElementBase& geos::finiteElement::FiniteElementBase::operator= ( FiniteElementBase && )

delete

Deleted move assignment operator.

Returns: deleted

◆ operator=() [2/2]

FiniteElementBase& geos::finiteElement::FiniteElementBase::operator= ( FiniteElementBase const & )

delete

Deleted copy assignment operator.

Returns: deleted

◆ plusGradNajAij() [1/2]

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::plusGradNajAij	(	GRADIENT_TYPE const &	gradN,
		real64 const (&)	var_detJxW[3][3],
		real64(&)	R[NUM_SUPPORT_POINTS][3]
	)

static

Inner product of each basis function gradient with a rank-2 symmetric tensor.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape function gradients.

Parameters

gradN	The basis function gradients at a point in the element.
var_detJxW	The rank-2 tensor at `q` scaled by J*W.
R	The vector at each support point which will hold the result from the tensor contraction.

More precisely, the operator is defined as:

$R_i = \sum_a^{nSupport} \left( \frac{\partial N_a}{\partial X_j} var_{ij}\right),$

where $\frac{\partial N_a}{\partial X_j}$ is the basis function gradient, $var_{ij}$ is the rank-2 symmetric tensor.

Inner product of each basis function gradient with a rank-2 tensor.

◆ plusGradNajAij() [2/2]

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::plusGradNajAij	(	GRADIENT_TYPE const &	gradN,
		real64 const (&)	var_detJxW[6],
		real64(&)	R[NUM_SUPPORT_POINTS][3]
	)

static

Inner product of each basis function gradient with a rank-2 symmetric tensor.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape function gradients.

Parameters

gradN	The basis function gradients at a point in the element.
var_detJxW	The rank-2 tensor at `q` scaled by J*W.
R	The vector at each support point which will hold the result from the tensor contraction.

More precisely, the operator is defined as:

$R_i = \sum_a^{nSupport} \left( \frac{\partial N_a}{\partial X_j} var_{ij}\right),$

where $\frac{\partial N_a}{\partial X_j}$ is the basis function gradient, $var_{ij}$ is the rank-2 symmetric tensor.

◆ plusGradNajAijPlusNaFi() [1/2]

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::plusGradNajAijPlusNaFi	(	GRADIENT_TYPE const &	gradN,
		real64 const (&)	var_detJxW[3][3],
		real64 const (&)	N[NUM_SUPPORT_POINTS],
		real64 const (&)	forcingTerm_detJxW[3],
		real64(&)	R[NUM_SUPPORT_POINTS][3]
	)

static

Inner product of each basis function gradient with a rank-2 symmetric tensor added to the product each shape function with a vector.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape function gradients.

Parameters

gradN	The basis function gradients at a point in the element.
var_detJxW	The rank-2 symmetric tensor at `q` scaled by J*W.
N	The shape function value at a predetermined coordinate in the element.
forcingTerm_detJxW	A vector scaled by detJxW
R	The vector at each support point which will hold the result from the tensor contraction.

$R_i = \sum_a^{nSupport} \left ( \frac{\partial N_a}{\partial X_j} var_{ij} + N_a f_i \right ),$

◆ plusGradNajAijPlusNaFi() [2/2]

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::plusGradNajAijPlusNaFi	(	GRADIENT_TYPE const &	gradN,
		real64 const (&)	var_detJxW[6],
		real64 const (&)	N[NUM_SUPPORT_POINTS],
		real64 const (&)	forcingTerm_detJxW[3],
		real64(&)	R[NUM_SUPPORT_POINTS][3]
	)

static

Inner product of each basis function gradient with a rank-2 tensor added to the product each shape function with a vector.

Inner product of each basis function gradient with a rank-2 symmetric tensor added to the product each shape function with a vector.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape function gradients.

Parameters

gradN	The basis function gradients at a point in the element.
var_detJxW	The rank-2 symmetric tensor at `q` scaled by J*W.
N	The shape function value at a predetermined coordinate in the element.
forcingTerm_detJxW	A vector scaled by detJxW
R	The vector at each support point which will hold the result from the tensor contraction.

$R_i = \sum_a^{nSupport} \left ( \frac{\partial N_a}{\partial X_j} var_{ij} + N_a f_i \right ),$

◆ plusNaFi()

template<int NUM_SUPPORT_POINTS>

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::plusNaFi	(	real64 const (&)	N[NUM_SUPPORT_POINTS],
		real64 const (&)	forcingTerm_detJxW[3],
		real64(&)	R[NUM_SUPPORT_POINTS][3]
	)

static

Product of each shape function with a vector forcing term.

Template Parameters

NUM_SUPPORT_POINTS The number of support points for the element.

Parameters

N	The shape function value at a predetermined coordinate in the element.
forcingTerm_detJxW	A vector scaled by detJxW
R	The vector at each support point which will hold the result from the tensor contraction.

◆ setDetJView()

void geos::finiteElement::FiniteElementBase::setDetJView ( arrayView2d< real64 const > const & source )

inline

Sets m_viewDetJ equal to an input view.

Parameters

source The view to assign to m_viewDetJ.

Definition at line 738 of file FiniteElementBase.hpp.

◆ setGradNView()

void geos::finiteElement::FiniteElementBase::setGradNView ( arrayView4d< real64 const > const & source )

inline

Sets m_viewGradN equal to an input view.

Parameters

source The view to assign to m_viewGradN.

Definition at line 719 of file FiniteElementBase.hpp.

◆ setup()

template<typename LEAF , typename SUBREGION_TYPE >

GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::setup	(	localIndex const &	cellIndex,
		typename LEAF::template MeshData< SUBREGION_TYPE > const &	meshData,
		typename LEAF::StackVariables &	stack
	)		const

inline

Abstract setup method, possibly computing cell-dependent properties.

Template Parameters

LEAF	Type of the derived finite element implementation.

Parameters

cellIndex	The index of the cell with respect to the cell sub region to which the element has been initialized previously (see initialize).
meshData	A MeshData object previously filled.
stack	Object that holds stack variables.

Definition at line 204 of file FiniteElementBase.hpp.

◆ setupStack()

template<typename SUBREGION_TYPE >

GEOS_HOST_DEVICE static GEOS_FORCE_INLINE void geos::finiteElement::FiniteElementBase::setupStack	(	localIndex const &	cellIndex,
		MeshData< SUBREGION_TYPE > const &	meshData,
		StackVariables &	stack
	)

inlinestatic

Empty setup method.

Parameters

cellIndex	The index of the cell with respect to the cell sub region.
meshData	MeshData struct filled by fillMeshData.
stack	Object that holds stack variables.

Definition at line 184 of file FiniteElementBase.hpp.

◆ symmetricGradient()

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::symmetricGradient	(	GRADIENT_TYPE const &	gradN,
		real64 const (&)	var[NUM_SUPPORT_POINTS][3],
		real64(&)	gradVar[6]
	)

static

Calculate the symmetric gradient of a vector valued support field at a point using the stored basis function gradients for all support points.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape

Parameters

gradN	The basis function gradients at a point in the element.
var	The vector valued support field that the gradient operator will be applied to.
gradVar	The symmetric gradient in Voigt notation.

More precisely, the operator is defined as:

$grad^s_{ij} = \frac{1}{2} \sum_a^{nSupport} \left ( \frac{\partial N_a}{\partial X_j} var_{ai} + \frac{\partial N_a}{\partial X_i} var_{aj}\right ),$

◆ symmetricGradientTrace()

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE real64 geos::finiteElement::FiniteElementBase::symmetricGradientTrace	(	GRADIENT_TYPE const &	gradN,
		real64 const (&)	var[NUM_SUPPORT_POINTS][3]
	)

static

Calculate the trace of the symmetric gradient of a vector valued support field (i.e. the volumetric strain for the displacement field) at a point using the stored basis function gradients for all support points.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape

Parameters

gradN	The basis function gradients at a point in the element.
var	The vector valued support field that the gradient operator will be applied to.

Returns: The trace of the symetric gradient tensor.

◆ value() [1/2]

template<int NUM_SUPPORT_POINTS>

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::value	(	real64 const (&)	N[NUM_SUPPORT_POINTS],
		real64 const (&)	var[NUM_SUPPORT_POINTS],
		real64 &	value
	)

static

Compute the interpolated value of a variable.

Template Parameters

NUM_SUPPORT_POINTS The number of support points for the element.

Parameters

N	Array (for each support point) of shape function values at the coordinate the variable is to be interpolated.
var	Array of variable values for each support point.
value	The interpolated value of `var`.

This is the standard finite element interpolation operator of a discrete variable defined at the support points. The operator is expressed as:

$value = \sum_a^{numSupport} \left ( N_a var_a \right ),$

Note: The shape function values N must be evaluated prior to calling this function.

◆ value() [2/2]

template<int NUM_SUPPORT_POINTS, int NUM_COMPONENTS>

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::value	(	real64 const (&)	N[NUM_SUPPORT_POINTS],
		real64 const (&)	var[NUM_SUPPORT_POINTS][NUM_COMPONENTS],
		real64(&)	value[NUM_COMPONENTS]
	)

static

Compute the interpolated value of a vector variable.

Template Parameters

NUM_COMPONENTS Number of components for the vector variable.

Compute the interpolated value of a variable.

Template Parameters

NUM_SUPPORT_POINTS The number of support points for the element.

Parameters

N	Array (for each support point) of shape function values at the coordinate the variable is to be interpolated.
var	Array of variable values for each support point.
value	The interpolated value of `var`.

This is the standard finite element interpolation operator of a discrete variable defined at the support points. The operator is expressed as:

$value = \sum_a^{numSupport} \left ( N_a var_a \right ),$

Note: The shape function values N must be evaluated prior to calling this function.

◆ valueAndGradient()

template<int NUM_SUPPORT_POINTS, typename GRADIENT_TYPE >

static GEOS_HOST_DEVICE void geos::finiteElement::FiniteElementBase::valueAndGradient	(	real64 const (&)	N[NUM_SUPPORT_POINTS],
		GRADIENT_TYPE const &	gradN,
		real64 const (&)	var[NUM_SUPPORT_POINTS],
		real64 &	value,
		real64(&)	gradVar[3]
	)

static

Calculate the value and gradient of a scalar valued support field at a point using the input basis function gradients.

Template Parameters

NUM_SUPPORT_POINTS	The number of support points for the element.
GRADIENT_TYPE	The type of the array object holding the shape

Parameters

N	Array (for each support point) of shape function values at the coordinate the variable is to be interpolated.
gradN	The basis function gradients at a point in the element.
var	The vector valued support field that the gradient operator will be applied to.
value	The value at the point for which N was specified.
gradVar	The gradient at the point for which gradN was specified.

Member Data Documentation

◆ m_viewDetJ

arrayView2d< real64 const > geos::finiteElement::FiniteElementBase::m_viewDetJ

protected

View to potentially hold pre-calculated weighted jacobian transformation determinants.

Definition at line 771 of file FiniteElementBase.hpp.

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

coreComponents/finiteElement/elementFormulations/FiniteElementBase.hpp

Classes

Public Member Functions

Static Public Member Functions

Static Public Attributes

Scattering Operator Functions

Detailed Description

Constructor & Destructor Documentation

◆ FiniteElementBase()

Member Function Documentation

◆ addEvaluatedGradGradStabilization()

◆ addEvaluatedGradGradStabilizationVector()

◆ addGradGradStabilization()

◆ addGradGradStabilizationMatrix()

◆ fillMeshData()

◆ getDetJView()

◆ getFunctionSpace()

◆ getGradN() [1/4]

◆ getGradN() [2/4]

◆ getGradN() [3/4]

◆ getGradN() [4/4]

◆ getGradNView()

◆ getMaxSupportPoints()

◆ getNumQuadraturePoints()

◆ getNumSupportPoints()

◆ gradient() [1/2]

◆ gradient() [2/2]

◆ initialize()

◆ numSupportPoints()

◆ operator=() [1/2]

◆ operator=() [2/2]

◆ plusGradNajAij() [1/2]

◆ plusGradNajAij() [2/2]

◆ plusGradNajAijPlusNaFi() [1/2]

◆ plusGradNajAijPlusNaFi() [2/2]

◆ plusNaFi()

◆ setDetJView()

◆ setGradNView()

◆ setup()

◆ setupStack()

◆ symmetricGradient()

◆ symmetricGradientTrace()

◆ value() [1/2]

◆ value() [2/2]

◆ valueAndGradient()

Member Data Documentation

◆ m_viewDetJ