HydroFemPhysics
The GeMA Hydraulic FEM Physics Plugin
Loading...
Searching...
No Matches
GmpHydraulicNeuralNetwork Class Reference

Basic class for the hydraulic physics plugin object. More...

#include <gmpHydraulicNeuralNetwork.h>

Inheritance diagram for GmpHydraulicNeuralNetwork:
Collaboration diagram for GmpHydraulicNeuralNetwork:

Public Member Functions

 GmpHydraulicNeuralNetwork (const char *pluginType, GmSimulationData *simulation, QString id, QString description, const GmpFemPhysicsCommonMaterialFactory *matFactory, const GmLogCategory &logger)
 Constructor. Will be called by the plugin loading code.
 
virtual ~GmpHydraulicNeuralNetwork ()
 Destructor.
 
virtual const QVariantMap * physicsMetaDataMap ()
 Returns a reference for the single Hydraulics attribute map, built when the function is called for the first time.
 
virtual FemResultType fillElementData (const GmElement *e, GmpFemMatrixSet &elemMatrices, GmpFemVectorSet &elemVectors)
 Fills the element matrices and vectors for a generic element.
 
double fast_acos (double x)
 Returns the inverse cosine of x in the range [0, pi] for -1 <= x <= 1 Fast implementation for the inverse cosine calculation Implementation based on "Handbook of Mathematical Functions" by M. Abramowitz and I.A. Stegun, Ed. and "https://developer.download.nvidia.com/cg/acos.html" Mean absolute error: 2.8453e-05 rad (0.0016°) Max absolute error: 6.7527e-05 rad (0.0039°)
 
double fast_atan (double x)
 Returns the inverse tangent of x in the range [0, pi/4] for -1 <= x <= 1 Fast implementation for the inverse tangent calculation Implementation based on "Efficient approximations for the arctangent function" by S. Rajan et al. and "https://yal.cc/fast-atan2/" Mean absolute error: 9.7010e-04 rad (0.0556°) Max absolute error: 1.5089e-03 rad (0.0865°)
 
double fast_atan2 (double y, double x)
 Returns the angle in the range [-pi, pi] between the positive x-axis and the ray to the point (x, y) Fast implementation for the 2-argument inverse tangent calculation Implementation based on "Efficient approximations for the arctangent function" by S. Rajan et al. and "https://yal.cc/fast-atan2/".
 
- Public Member Functions inherited from GmpHydraulic
 GmpHydraulic (const char *pluginType, GmSimulationData *simulation, QString id, QString description, const GmpFemPhysicsCommonMaterialFactory *matFactory, const GmLogCategory &logger)
 Constructor. Will be called by the plugin loading code.
 
virtual ~GmpHydraulic ()
 Destructor.
 
virtual const char * pluginName () const
 
virtual const char * pluginType () const
 
virtual bool supportsStateDumping ()
 
virtual bool fixedNodalForcesBc (QVector< int > &nodes, QVector< int > &dof, QVector< double > &values) const
 See comments on base class. Fills vectors with prescribed nodal forces - flow.
 
virtual bool fixedNodalDofsBc (QVector< int > &nodes, QVector< int > &dof, QVector< double > &values, bool *constantValues) const
 See comments on base class. Fills vectors with prescribed node pore pressures.
 
virtual FemResultType fillElementDataForBc (const GmElement *e, const GmBoundaryCondition *bc, int bcIndex, int bcListIndex, int border, GmpFemMatrixSet &elemMatrices, GmpFemVectorSet &elemVectors)
 
virtual FemResultType fillElementDataBc (const GmElement *e, const GmBoundaryCondition *bc, int bcIndex, int bcListIndex, int border, GmMatrix &elemMat, GmVector &elemVec)
 
virtual bool calcDerivedResults (bool nonLinearSolver)
 Calc velocity and flow on nodes and/or Gauss points.
 
virtual void fillStorageCompMatrix (const GmElement *e, const GmVector *coord, int ip, GmMatrix &Sc, const GmVector &N, double c)
 Calculates the stiffness matrix of "storage / compressibility" for element 'e' at the given integration point (coord, ip) storing the result at the Sc matrix. N is the shape function vector and c a constant equal to w * h * detJ.
 
virtual void fillFlowMatrix (const GmElement *e, const GmVector *coord, int ip, GmMatrix &Kc, const GmMatrix &Bp, double c) const
 Calculates the flow stiffness matrix due to permeability contribution or permeability matrix for element 'e' at the given integration point (coord, ip) storing the result at the Kc matrix. Bp is the strain pore pressure matrix and c a constant equal to w * h * detJ.
 
virtual void fillFluidVelocity (const GmElement *e, const GmVector *coord, int ip, GmVector &Vip, const GmVector &Pe, const GmMatrix &Bp) const
 Evaluates velocity at integration points (it should be improved to consider gravity effect)
 
virtual void fillElementPorePressure (const GmElement *e, GmVector &pe)
 Given an element, fills the vector pe with nodal pore pressure. The vector should have size equal to the number of calculated nodes.
 
virtual void gravitySupplyFluid (const GmElement *e, const GmVector *coord, int ip, GmVector &qext_gsf, GmMatrix &Bp, double c)
 
virtual void fillSourceTermVector (const GmElement *e, const GmVector *coord, int ip, GmVector &St, const GmVector &N, double c)
 Calculates the source term due the porosity rate of porous eskeleton for element 'e' at the given integration point (coord, ip) storing the result at external vector. N is the shape function vector and c a constant equal to w * h * detJ.
 
virtual double fillBpMatrix (const GmShape *shape, const GmVector &ncoord, const GmMatrix &X, const GmVector &N, const GmMatrix &J, GmMatrix &Bp)=0
 Given a point 'ncoord', the element coordinate matrix 'X', the point shape functions 'N' and the transformation Jacobian 'J', fills the matrix with cartesian partial derivatives Bp using the provided shape function and returns the scaled jacobian determinant.
 
virtual void gravityVector (const GmElement *e, const GmVector *coord, int ip, GmVector &grav) const
 Returns the gravity accelaration vector.
 
virtual bool isAxisymmetric ()
 Returns TRUE only for axisymmetric models.
 
virtual double axisymmetricFactor (const GmElement *e, const GmMatrix &X, const GmVector &N) const
 Returns the factor that should be applied when using axis symetric elements. 1.0 for other etypes.
 
HMaterialPointAccessorshydraulicMaterialPoint ()
 Returns the hydraulic material point accessor.
 

Protected Member Functions

virtual bool checkAndLoadPrivateData (LuaTable &table)
 
- Protected Member Functions inherited from GmpHydraulic
virtual bool checkAndLoadAttributeAccessors (LuaTable &nodeTable, LuaTable &cellTable, LuaTable &gaussTable)
 Checks the loaded data.
 
virtual void calcElementDerivedResultAtPoints (const GmElement *e, int resultId, const GmMatrix &evalPoints, bool evalAtIp, GmMatrix &result)
 The virtual function used by calcDerivedResults() to effectivelly calculate the velocity (resultId = 0) or the flow (resultId = 1) at the given evaluation points.
 
void calcElementVelocityAtPoints (const GmElement *e, const GmMatrix &evalPoints, GmMatrix &v)
 The function responsible for effectivelly calculating the velocity at the given evaluation points, filling v.
 
void calcElementFlowAtPoints (const GmElement *e, const GmMatrix &evalPoints, bool evalAtIp, GmMatrix &q)
 The function responsible for effectivelly calculating the flow at the given evaluation points, filling q.
 

Private Attributes

const char * _pluginType
 The plugin type name.
 
GmpFemNeuralNetwork * _netK
 Neural Network data to evaluate K matrix.
 
GmpFemNeuralNetwork * _netC
 Neural Network data to evaluate C matrix.
 

Additional Inherited Members

- Public Types inherited from GmpHydraulic
enum  HMaterialPointStates { saturated , unsaturated }
 
enum  PhysicsAttributeIds {
  ISOPARAMETRIC_ID , POROSITY_UPDATE_ID , PERMEABILITY_UPDATE_ID , USE_CONSISTENT_MATRIX_ID ,
  VELOCITY_MODE_ID , FLOW_MODE_ID , DIRECT_ID , CGRAV_ID ,
  VF_MODE_ID , NUM_ATTRI_IDS
}
 IDs for physics attributes. More...
 
enum  StateVarIds { P_ID , NUM_STATEVAR_IDS }
 IDs fo physics state vars. More...
 
enum  NodeAttributeIds { V_NA_ID , Q_NA_ID , VF_NA_ID , NUM_NA_IDS }
 IDs for physics node attributes. More...
 
enum  cellAttributeIds { NUM_CA_IDS }
 IDs for physics Cell attributes. More...
 
enum  GaussAttributeIds {
  V_GA_ID , Q_GA_ID , FT_GA_ID , FTOLD_GA_ID ,
  NUM_GA_IDS
}
 IDs for physics Gauss attributes. More...
 
enum  ElementPropertyIds { THICKNESS_ID , GRAV_ID , SOURTERM_ID , NUM_PROPER_IDS }
 IDs for physics element properties. More...
 
enum  BoundaryConditionIds { FIXED_P_BC_ID , FIXED_Q_BC_ID , FIXED_SPF_BC_ID , NUM_BC_IDS }
 IDs for accepted boundary condition types. More...
 
enum  BoundaryConditionValueIds { BC_P_ID , BC_QW_ID , BC_IF_ID , NUM_BCV_IDS }
 IDs for property values from accepted boundary condition types. More...
 
- Protected Attributes inherited from GmpHydraulic
HMaterialPointAccessors _hmaterialPointAccessor
 Access to hydraulic material point.
 
bool _setInitialCond
 Sets the initial conditions required by the physics.
 

Detailed Description

Basic class for the hydraulic physics plugin object.

Member Function Documentation

◆ fillElementData()

GmpFemPhysics::FemResultType GmpHydraulicNeuralNetwork::fillElementData ( const GmElement * e,
GmpFemMatrixSet & elemMatrices,
GmpFemVectorSet & elemVectors )
virtual

Fills the element matrices and vectors for a generic element.

Reimplemented from GmpHydraulic.

◆ physicsMetaDataMap()

const QVariantMap * GmpHydraulicNeuralNetwork::physicsMetaDataMap ( )
virtual

Returns a reference for the single Hydraulics attribute map, built when the function is called for the first time.

Reimplemented from GmpHydraulic.


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