v0.8.13
Public Member Functions | Public Attributes | List of all members
MixTransport::MixTransportElement::OpEvaluateBcOnFluxes Struct Reference

Evaluate boundary conditions on fluxes. More...

#include <users_modules/basic_finite_elements/mix_transport/src/MixTransportElement.hpp>

Inheritance diagram for MixTransport::MixTransportElement::OpEvaluateBcOnFluxes:
[legend]
Collaboration diagram for MixTransport::MixTransportElement::OpEvaluateBcOnFluxes:
[legend]

Public Member Functions

 OpEvaluateBcOnFluxes (MixTransportElement &ctx, const std::string flux_name)
 
MoFEMErrorCode doWork (int side, EntityType type, DataForcesAndSourcesCore::EntData &data)
 
- Public Member Functions inherited from MoFEM::FaceElementForcesAndSourcesCore::UserDataOperator
 UserDataOperator (const FieldSpace space)
 
 UserDataOperator (const std::string &field_name, const char type)
 
 UserDataOperator (const std::string &row_field_name, const std::string &col_field_name, const char type, const bool symm=true)
 
double getArea ()
 get area of face More...
 
double getMeasure ()
 get measure of element More...
 
VectorDoublegetNormal ()
 get triangle normal More...
 
VectorDoublegetTangent1 ()
 get triangle tangent 1 More...
 
VectorDoublegetTangent2 ()
 get triangle tangent 2 More...
 
auto getFTensor1Normal ()
 get normal as tensor More...
 
DEPRECATED auto getTensor1Normal ()
 
auto getFTensor1Tangent1 ()
 get tangentOne as tensor More...
 
DEPRECATED auto getTensor1Tangent1 ()
 
auto getFTensor2Tangent1 ()
 get tangentTwo as tensor More...
 
DEPRECATED auto getTensor2Tangent1 ()
 
int getNumNodes ()
 get element number of nodes More...
 
const EntityHandlegetConn ()
 get element connectivity More...
 
VectorDoublegetCoords ()
 get triangle coordinates More...
 
auto getFTensor1Coords ()
 get get coords at gauss points More...
 
DEPRECATED auto getTensor1Coords ()
 
MatrixDoublegetGaussPts ()
 get matrix of integration (Gauss) points on Face Element where columns 0,1 are x,y coordinates respectively and column 2 is a value of weight for example getGaussPts()(1,13) returns y coordinate of 13th Gauss point on particular face element More...
 
auto getFTensor0IntegrationWeight ()
 Get integration weights. More...
 
MatrixDoublegetCoordsAtGaussPts ()
 Gauss points and weight, matrix (nb. of points x 3) More...
 
auto getFTensor1CoordsAtGaussPts ()
 get coordinates at Gauss pts. More...
 
DEPRECATED auto getTensor1CoordsAtGaussPts ()
 
MatrixDoublegetHoCoordsAtGaussPts ()
 coordinate at Gauss points (if hierarchical approximation of element geometry) More...
 
auto getFTensor1HoCoordsAtGaussPts ()
 get coordinates at Gauss pts (takes in account ho approx. of geometry) More...
 
DEPRECATED auto getTensor1HoCoordsAtGaussPts ()
 
MatrixDoublegetNormalsAtGaussPts ()
 if higher order geometry return normals at Gauss pts. More...
 
DEPRECATED MatrixDoublegetNormalsAtGaussPt ()
 
ublas::matrix_row< MatrixDoublegetNormalsAtGaussPts (const int gg)
 if higher order geometry return normals at Gauss pts. More...
 
DEPRECATED auto getNormalsAtGaussPt (const int gg)
 
MatrixDoublegetTangent1AtGaussPts ()
 if higher order geometry return tangent vector to triangle at Gauss pts. More...
 
DEPRECATED MatrixDoublegetTangent1AtGaussPt ()
 
MatrixDoublegetTangent2AtGaussPts ()
 if higher order geometry return tangent vector to triangle at Gauss pts. More...
 
DEPRECATED MatrixDoublegetTangent2AtGaussPt ()
 
auto getFTensor1NormalsAtGaussPts ()
 get normal at integration points More...
 
DEPRECATED auto getTensor1NormalsAtGaussPt ()
 
auto getFTensor1Tangent1AtGaussPts ()
 get tangent 1 at integration points More...
 
DEPRECATED auto getTensor1Tangent1AtGaussPt ()
 
auto getFTensor1Tangent2AtGaussPts ()
 get tangent 2 at integration points More...
 
DEPRECATED auto getTensor1Tangent2AtGaussPt ()
 
DEPRECATED const FaceElementForcesAndSourcesCoregetFaceElementForcesAndSourcesCore ()
 
DEPRECATED const FaceElementForcesAndSourcesCoregetTriFE ()
 
const FaceElementForcesAndSourcesCoregetFaceFE ()
 return pointer to Generic Triangle Finite Element object More...
 
MoFEMErrorCode loopSideVolumes (const string &fe_name, VolumeElementForcesAndSourcesCoreOnSide &method)
 
- Public Member Functions inherited from MoFEM::ForcesAndSourcesCore::UserDataOperator
 UserDataOperator (const FieldSpace space, const char type=OPLAST, const bool symm=true)
 
 UserDataOperator (const std::string &field_name, const char type, const bool symm=true)
 
 UserDataOperator (const std::string &row_field_name, const std::string &col_field_name, const char type, const bool symm=true)
 
virtual ~UserDataOperator ()
 
boost::shared_ptr< const NumeredEntFiniteElementgetNumeredEntFiniteElementPtr () const
 Return raw pointer to NumeredEntFiniteElement. More...
 
EntityHandle getFEEntityHandle () const
 Return finite element entity handle. More...
 
MoFEMErrorCode getProblemRowIndices (const std::string filed_name, const EntityType type, const int side, VectorInt &indices) const
 Get row indices. More...
 
DEPRECATED MoFEMErrorCode getPorblemRowIndices (const std::string filed_name, const EntityType type, const int side, VectorInt &indices) const
 
MoFEMErrorCode getProblemColIndices (const std::string filed_name, const EntityType type, const int side, VectorInt &indices) const
 Get col indices. More...
 
virtual MoFEMErrorCode setPtrFE (ForcesAndSourcesCore *ptr)
 
DEPRECATED MoFEMErrorCode getPorblemColIndices (const std::string filed_name, const EntityType type, const int side, VectorInt &indices) const
 
const FEMethodgetFEMethod () const
 Return raw pointer to Finite Element Method object. More...
 
int getOpType () const
 Get operator types. More...
 
void setOpType (const OpType type)
 Set operator type. More...
 
void addOpType (const OpType type)
 Add operator type. More...
 
int getNinTheLoop () const
 get number of finite element in the loop More...
 
int getLoopSize () const
 get size of elements in the loop More...
 
const std::string & getFEName () const
 Get name of the element. More...
 
Vec getSnesF () const
 
Vec getSnesX () const
 
Mat getSnesA () const
 
Mat getSnesB () const
 
Vec getTSu () const
 
Vec getTSu_t () const
 
Vec getTSf () const
 
Mat getTSA () const
 
Mat getTSB () const
 
int getTSstep () const
 
double getTStime () const
 
double getTSa () const
 
- Public Member Functions inherited from MoFEM::DataOperator
 DataOperator (const bool symm=true, const bool do_vertices=true, const bool do_edges=true, const bool do_quads=true, const bool do_tris=true, const bool do_tets=true, const bool do_prisms=true)
 
virtual ~DataOperator ()
 
virtual MoFEMErrorCode doWork (int row_side, int col_side, EntityType row_type, EntityType col_type, DataForcesAndSourcesCore::EntData &row_data, DataForcesAndSourcesCore::EntData &col_data)
 Operator for bi-linear form, usually to calculate values on right hand side. More...
 
virtual MoFEMErrorCode opLhs (DataForcesAndSourcesCore &row_data, DataForcesAndSourcesCore &col_data, bool symm=true)
 
virtual MoFEMErrorCode opLhs (DataForcesAndSourcesCore &row_data, DataForcesAndSourcesCore &col_data)
 
virtual MoFEMErrorCode doWork (int side, EntityType type, DataForcesAndSourcesCore::EntData &data)
 Operator for linear form, usually to calculate values on left hand side. More...
 
virtual MoFEMErrorCode opRhs (DataForcesAndSourcesCore &data, const bool do_vertices, const bool do_edges, const bool do_quads, const bool do_tris, const bool do_tets, const bool do_prisms, const bool error_if_no_base=true)
 
virtual MoFEMErrorCode opRhs (DataForcesAndSourcesCore &data, const bool error_if_no_base=true)
 
bool getSymm () const
 Get if operator uses symmetry of DOFs or not. More...
 
void setSymm ()
 set if operator is executed taking in account symmetry More...
 
void unSetSymm ()
 unset if operator is executed for non symmetric problem More...
 

Public Attributes

MixTransportElementcTx
 
MatrixDouble NN
 
VectorDouble Nf
 
FTensor::Index< 'i', 3 > i
 
- Public Attributes inherited from MoFEM::ForcesAndSourcesCore::UserDataOperator
char opType
 
std::string rowFieldName
 
std::string colFieldName
 
FieldSpace sPace
 
- Public Attributes inherited from MoFEM::DataOperator
bool sYmm
 If true assume that matrix is symmetric structure. More...
 
bool doVertices
 If false skip vertices. More...
 
bool doEdges
 If false skip edges. More...
 
bool doQuads
 
bool doTris
 
bool doTets
 
bool doPrisms
 

Additional Inherited Members

- Public Types inherited from MoFEM::ForcesAndSourcesCore::UserDataOperator
enum  OpType { OPROW = 1 << 0, OPCOL = 1 << 1, OPROWCOL = 1 << 2, OPLAST = 1 << 3 }
 Controls loop over entities on element. More...
 
- Protected Attributes inherited from MoFEM::ForcesAndSourcesCore::UserDataOperator
ForcesAndSourcesCoreptrFE
 

Detailed Description

Evaluate boundary conditions on fluxes.

Note that Neumann boundary conditions here are essential. So it is opposite what you find in displacement finite element method.

Here we have to solve for degrees of freedom on boundary such base functions approximate flux.

Definition at line 1174 of file MixTransportElement.hpp.

Constructor & Destructor Documentation

◆ OpEvaluateBcOnFluxes()

MixTransport::MixTransportElement::OpEvaluateBcOnFluxes::OpEvaluateBcOnFluxes ( MixTransportElement ctx,
const std::string  flux_name 
)

Definition at line 1177 of file MixTransportElement.hpp.

1179  flux_name, UserDataOperator::OPROW),
1180  cTx(ctx) {}
ForcesAndSourcesCore::UserDataOperator UserDataOperator

Member Function Documentation

◆ doWork()

MoFEMErrorCode MixTransport::MixTransportElement::OpEvaluateBcOnFluxes::doWork ( int  side,
EntityType  type,
DataForcesAndSourcesCore::EntData &  data 
)

Definition at line 1186 of file MixTransportElement.hpp.

1187  {
1189  if (data.getFieldData().size() == 0)
1191  EntityHandle fe_ent = getNumeredEntFiniteElementPtr()->getEnt();
1192  int nb_dofs = data.getFieldData().size();
1193  int nb_gauss_pts = data.getHdivN().size1();
1194  if (3 * nb_dofs != static_cast<int>(data.getHdivN().size2())) {
1195  SETERRQ(PETSC_COMM_WORLD, MOFEM_DATA_INCONSISTENCY,
1196  "wrong number of dofs");
1197  }
1198  NN.resize(nb_dofs, nb_dofs);
1199  Nf.resize(nb_dofs);
1200  NN.clear();
1201  Nf.clear();
1202 
1203  // Get normal vector. Note that when higher order geometry is set, then
1204  // face element could be curved, i.e. normal can be different at each
1205  // integration point.
1206  double *normal_ptr;
1207  if (getNormalsAtGaussPts().size1() == (unsigned int)nb_gauss_pts) {
1208  // HO geometry
1209  normal_ptr = &getNormalsAtGaussPts(0)[0];
1210  } else {
1211  // Linear geometry, i.e. constant normal on face
1212  normal_ptr = &getNormal()[0];
1213  }
1214  // set tensor from pointer
1215  FTensor::Tensor1<const double *, 3> t_normal(normal_ptr, &normal_ptr[1],
1216  &normal_ptr[2], 3);
1217  // get base functions
1218  auto t_n_hdiv_row = data.getFTensor1HdivN<3>();
1219 
1220  double nrm2 = 0;
1221 
1222  // loop over integration points
1223  for (int gg = 0; gg < nb_gauss_pts; gg++) {
1224 
1225  // get integration point coordinates
1226  double x, y, z;
1227  if (getNormalsAtGaussPts().size1() == (unsigned int)nb_gauss_pts) {
1228  x = getHoCoordsAtGaussPts()(gg, 0);
1229  y = getHoCoordsAtGaussPts()(gg, 1);
1230  z = getHoCoordsAtGaussPts()(gg, 2);
1231  } else {
1232  x = getCoordsAtGaussPts()(gg, 0);
1233  y = getCoordsAtGaussPts()(gg, 1);
1234  z = getCoordsAtGaussPts()(gg, 2);
1235  }
1236 
1237  // get flux on fece for given element handle and coordinates
1238  double flux;
1239  CHKERR cTx.getBcOnFluxes(fe_ent, x, y, z, flux);
1240  ;
1241  // get weight for integration rule
1242  double w = getGaussPts()(2, gg);
1243  if (gg == 0) {
1244  nrm2 = sqrt(t_normal(i) * t_normal(i));
1245  }
1246 
1247  // set tensor of rank 0 to matrix NN elements
1248  // loop over base functions on rows and columns
1249  for (int ll = 0; ll != nb_dofs; ll++) {
1250  // get column on shape functions
1252  &data.getHdivN(gg)(0, HDIV0), &data.getHdivN(gg)(0, HDIV1),
1253  &data.getHdivN(gg)(0, HDIV2), 3);
1254  for (int kk = 0; kk <= ll; kk++) {
1255  NN(ll, kk) += w * t_n_hdiv_row(i) * t_n_hdiv_col(i);
1256  ++t_n_hdiv_col;
1257  }
1258  // right hand side
1259  Nf[ll] += w * t_n_hdiv_row(i) * t_normal(i) * flux / nrm2;
1260  ++t_n_hdiv_row;
1261  }
1262 
1263  // If HO geometry increment t_normal to next integration point
1264  if (getNormalsAtGaussPts().size1() == (unsigned int)nb_gauss_pts) {
1265  ++t_normal;
1266  nrm2 = sqrt(t_normal(i) * t_normal(i));
1267  }
1268  }
1269  // get global dofs indices on element
1270  cTx.bcIndices.insert(data.getIndices().begin(), data.getIndices().end());
1271  // factor matrix
1273  // solve local problem
1274  cholesky_solve(NN, Nf, ublas::lower());
1275 
1276  // cerr << Nf << endl;
1277  // cerr << data.getIndices() << endl;
1278 
1279  // set solution to vector
1280  CHKERR VecSetValues(cTx.D0, nb_dofs, &*data.getIndices().begin(),
1281  &*Nf.begin(), INSERT_VALUES);
1282  ;
1283 
1285  }
MatrixDouble & getNormalsAtGaussPts()
if higher order geometry return normals at Gauss pts.
MatrixDouble & getHoCoordsAtGaussPts()
coordinate at Gauss points (if hierarchical approximation of element geometry)
boost::shared_ptr< const NumeredEntFiniteElement > getNumeredEntFiniteElementPtr() const
Return raw pointer to NumeredEntFiniteElement.
#define MoFEMFunctionReturn(a)
Last executable line of each PETSc function used for error handling. Replaces return() ...
Definition: definitions.h:495
#define MoFEMFunctionReturnHot(a)
Last executable line of each PETSc function used for error handling. Replaces return() ...
Definition: definitions.h:526
MatrixDouble & getCoordsAtGaussPts()
Gauss points and weight, matrix (nb. of points x 3)
size_t cholesky_decompose(const MATRIX &A, TRIA &L)
decompose the symmetric positive definit matrix A into product L L^T.
Definition: cholesky.hpp:52
MatrixDouble & getGaussPts()
get matrix of integration (Gauss) points on Face Element where columns 0,1 are x,y coordinates respec...
void cholesky_solve(const TRIA &L, VEC &x, ublas::lower)
solve system L L^T x = b inplace
Definition: cholesky.hpp:221
#define CHKERR
Inline error check.
Definition: definitions.h:614
virtual MoFEMErrorCode getBcOnFluxes(const EntityHandle ent, const double x, const double y, const double z, double &flux)
essential (Neumann) boundary condition (set fluxes)
#define MoFEMFunctionBegin
First executable line of each MoFEM function, used for error handling. Final line of MoFEM functions ...
Definition: definitions.h:439

Member Data Documentation

◆ cTx

MixTransportElement& MixTransport::MixTransportElement::OpEvaluateBcOnFluxes::cTx

Definition at line 1176 of file MixTransportElement.hpp.

◆ i

FTensor::Index<'i', 3> MixTransport::MixTransportElement::OpEvaluateBcOnFluxes::i

Definition at line 1184 of file MixTransportElement.hpp.

◆ Nf

VectorDouble MixTransport::MixTransportElement::OpEvaluateBcOnFluxes::Nf

Definition at line 1183 of file MixTransportElement.hpp.

◆ NN

MatrixDouble MixTransport::MixTransportElement::OpEvaluateBcOnFluxes::NN

Definition at line 1182 of file MixTransportElement.hpp.


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