mech_output_writer_impl.h

Go to the documentation of this file.

/*
 * Copyright (c) 2014-2015:  G-CSC, Goethe University Frankfurt
 * Author: Raphael Prohl
 * 
 * This file is part of UG4.
 * 
 * UG4 is free software: you can redistribute it and/or modify it under the
 * terms of the GNU Lesser General Public License version 3 (as published by the
 * Free Software Foundation) with the following additional attribution
 * requirements (according to LGPL/GPL v3 §7):
 * 
 * (1) The following notice must be displayed in the Appropriate Legal Notices
 * of covered and combined works: "Based on UG4 (www.ug4.org/license)".
 * 
 * (2) The following notice must be displayed at a prominent place in the
 * terminal output of covered works: "Based on UG4 (www.ug4.org/license)".
 * 
 * (3) The following bibliography is recommended for citation and must be
 * preserved in all covered files:
 * "Reiter, S., Vogel, A., Heppner, I., Rupp, M., and Wittum, G. A massively
 *   parallel geometric multigrid solver on hierarchically distributed grids.
 *   Computing and visualization in science 16, 4 (2013), 151-164"
 * "Vogel, A., Reiter, S., Rupp, M., Nägel, A., and Wittum, G. UG4 -- a novel
 *   flexible software system for simulating pde based models on high performance
 *   computers. Computing and visualization in science 16, 4 (2013), 165-179"
 * 
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
 * GNU Lesser General Public License for more details.
 */
 
#ifndef MECH_OUTPUT_WRITER_IMPL_H_
#define MECH_OUTPUT_WRITER_IMPL_H_
 
#include "lib_grid/algorithms/geom_obj_util/geom_obj_util.h"
 
// module intern headers
#include "mech_output_writer.h"
 
namespace ug {
namespace SmallStrainMechanics{
 
template <typename TGridFunction>
void plast_ip(TGridFunction& plastIP,
    MechOutputWriter<typename TGridFunction::domain_type>& mechOut,
    TGridFunction& u)
{
  static const int dim = TGridFunction::dim;
  typedef typename TGridFunction::template dim_traits<dim>::grid_base_object grid_base_object;
  typedef typename TGridFunction::template dim_traits<dim>::const_iterator const_iterator;
 
  //  local indices and local algebra
  LocalIndices ind; LocalVector locU, locPlastIP;
 
  const_iterator iter = u.template begin<grid_base_object>();
  const_iterator end = u.template end<grid_base_object>();
 
  //  loop over all elements
  for(;iter != end; ++iter)
  {
    //  get element
    grid_base_object* elem = *iter;
 
    //  get global indices
    u.indices(elem, ind);
 
    //  adapt local algebra
    locU.resize(ind); locPlastIP.resize(ind);
 
    //  reset contribution of this element
    locPlastIP = 0.0;
 
    //  local vector extract -> locU
    GetLocalVector(locU, u); GetLocalVector(locPlastIP, plastIP);
 
    mechOut.plastIP_elem(locPlastIP, elem, locU,
        u.domain());
 
    //  send local to global plastIP
    AddLocalVector(plastIP, locPlastIP);
  }
}
void plast_ip(TGridFunction& plastIP, {…}
 
template <typename TGridFunction>
void equiv_plast_strain(TGridFunction& eqPlastStrain,
    MechOutputWriter<typename TGridFunction::domain_type>& mechOut,
    TGridFunction& u)
{
  static const int dim = TGridFunction::dim;
  typedef typename TGridFunction::template dim_traits<dim>::grid_base_object grid_base_object;
  typedef typename TGridFunction::template dim_traits<dim>::const_iterator const_iterator;
 
  //  local indices and local algebra
  LocalIndices ind; LocalVector locU, locEqPlastStrain;
 
  const_iterator iter = u.template begin<grid_base_object>();
  const_iterator end = u.template end<grid_base_object>();
 
  //  loop over all elements
  for(;iter != end; ++iter)
  {
    //  get element
    grid_base_object* elem = *iter;
 
    //  get global indices
    u.indices(elem, ind);
 
    //  adapt local algebra
    locU.resize(ind); locEqPlastStrain.resize(ind);
 
    //  reset contribution of this element
    locEqPlastStrain = 0.0;
 
    //  local vector extract -> locU
    GetLocalVector(locU, u); GetLocalVector(locEqPlastStrain, eqPlastStrain);
 
    mechOut.eqPlastStrain_elem(locEqPlastStrain, elem, locU, u.domain());
 
    //  send local to global eqPlastStrain
    AddLocalVector(eqPlastStrain, locEqPlastStrain);
  }
}
void equiv_plast_strain(TGridFunction& eqPlastStrain, {…}
 
template<typename TGridFunction>
void
normal_stresses_strains(MechOutputWriter<typename TGridFunction::domain_type>& mechOut,
    TGridFunction& sigma, TGridFunction& epsilon,
    TGridFunction& stressFunc, TGridFunction& strainFunc, TGridFunction& u)
{
  static const int dim = TGridFunction::dim;
  typedef typename TGridFunction::template dim_traits<dim>::grid_base_object grid_base_object;
  typedef typename TGridFunction::template dim_traits<dim>::const_iterator const_iterator;
 
  //  local indices and local algebra
  LocalIndices indU, indEps, indSig, indStressFunc, indStrainFunc;
  LocalVector locU, locSig, locEps, locStressFunc, locStrainFunc;
 
  const_iterator iter = u.template begin<grid_base_object>();
  const_iterator end = u.template end<grid_base_object>();
 
  //  loop over all elements
  for(;iter != end; ++iter)
  {
    //  get element
    grid_base_object* elem = *iter;
 
    //  get global indices
    u.indices(elem, indU); epsilon.indices(elem, indEps);
    sigma.indices(elem, indSig); stressFunc.indices(elem, indStressFunc);
    strainFunc.indices(elem, indStrainFunc);
 
    //  adapt local algebra
    locU.resize(indU); locEps.resize(indEps);
    locSig.resize(indSig); locStressFunc.resize(indStressFunc);
    locStrainFunc.resize(indStrainFunc);
 
    //  reset contribution of this element
    locStressFunc = 0.0; locStrainFunc = 0.0; locSig = 0.0; locEps = 0.0;
 
    //  local vector extract -> locU
    GetLocalVector(locU, u);
 
    //  call method of solid mechanics output writer
    mechOut.normal_stress_strain_elem(locSig, locEps, locStressFunc, locStrainFunc, elem,
        locU, u.domain());
 
    //  send local to global
    //  sigma (cauchy-stress tensor), epsilon (linearized strain tensor), devSigma (deviatoric part of Sigma)
    AddLocalVector(stressFunc, locStressFunc);
    AddLocalVector(strainFunc, locStrainFunc);
    AddLocalVector(sigma, locSig);
    AddLocalVector(epsilon, locEps);
  }
}
normal_stresses_strains(MechOutputWriter<typename TGridFunction::domain_type>& mechOut, {…}
 
template <typename TGridFunction>
void invariants_kirchhoff_stress(TGridFunction& invarKirchhoffStress,
    MechOutputWriter<typename TGridFunction::domain_type>& mechOut,
    TGridFunction& u)
{
  static const int dim = TGridFunction::dim;
  typedef typename TGridFunction::template dim_traits<dim>::grid_base_object grid_base_object;
  typedef typename TGridFunction::template dim_traits<dim>::const_iterator const_iterator;
 
  //  local indices and local algebra
  LocalIndices ind, indInvarKirchhoffStress;
  LocalVector locU, locInvarKirchhoffStress;
 
  const_iterator iter = u.template begin<grid_base_object>();
  const_iterator end = u.template end<grid_base_object>();
 
  //  loop over all elements
  for(;iter != end; ++iter)
  {
    //  get element
    grid_base_object* elem = *iter;
 
    //  get global indices
    u.indices(elem, ind); invarKirchhoffStress.indices(elem, indInvarKirchhoffStress);
 
    //  adapt local algebra
    locU.resize(ind); locInvarKirchhoffStress.resize(indInvarKirchhoffStress);
 
    //  local vector extract -> locU
    GetLocalVector(locU, u); GetLocalVector(locInvarKirchhoffStress, invarKirchhoffStress);
 
    //  reset contribution of this element
    locInvarKirchhoffStress = 0.0;
 
    mechOut.invariants_kirchhoff_stress_elem(locInvarKirchhoffStress, elem, locU,
        u.domain());
 
    //  send local to global tensor
    AddLocalVector(invarKirchhoffStress, locInvarKirchhoffStress);
  }
}
void invariants_kirchhoff_stress(TGridFunction& invarKirchhoffStress, {…}
 
template<typename TDomain>
void
MechOutputWriter<TDomain>::
plastIP_elem(LocalVector& locPlastIP, TBaseElem* elem,
    const LocalVector& locU, SmartPtr<TDomain> dom)
{
  //  get vertices and extract corner coordinates
  typedef typename TDomain::position_accessor_type position_accessor_type;
  position_accessor_type& aaPos = dom->position_accessor();
  MathVector<dim> coCoord[domain_traits<dim>::MaxNumVerticesOfElem];
  const size_t numVertices = elem->num_vertices();
  for (size_t i = 0; i < numVertices; ++i) {
    coCoord[i] = aaPos[elem->vertex(i)];
  };
 
  //  update the geometry for the Element elem
  DimFEGeometry<dim> geo;
    try{
    geo.update(elem, &(coCoord[0]),
        LFEID(LFEID::LAGRANGE, dim, 1), m_quadOrder);
  }
    UG_CATCH_THROW("SmallStrainMechOutput::plastIP_elem:"
          " Cannot update Finite Element Geometry.");
 
  //  get all neighbor elems which share a vertex with the given element 'elem'
  typename TDomain::grid_type& grid = *(dom->grid());
  typedef typename vector<TBaseElem*>::iterator neighborElemIter;
  vector<TBaseElem*> vNeighborElems;
  CollectNeighbors(vNeighborElems, elem, grid, NHT_VERTEX_NEIGHBORS);
 
  //  pointer to internal variable of current elem
  m_spMatLaw->internal_vars(elem);
 
  /*MathMatrix<dim, dim> GradU;
  number plasticIP = 0.0;
  for (size_t ip = 0; ip < geo.num_ip(); ++ip)
  {
    //  get displacementGradient (GradU)
    m_spMatLaw->template DisplacementGradient<DimFEGeometry<dim> >(GradU, ip, geo, locU);
 
    number gamma = m_spMatLaw->plastic_multiplier(ip, GradU);
    if (gamma > 0.0)
      plasticIP += 1.0;
    if (gamma < 0.0)
      UG_THROW("gamma: " << gamma << "in plastIP_elem \n");
  }*/
 
  MathMatrix<dim, dim> plastStrain;
  number plasticIP = 0.0;
  for (size_t ip = 0; ip < geo.num_ip(); ++ip)
  {
    plastStrain = *(m_spMatLaw->inelastic_strain_tensor(ip));
 
    number plastStrainNorm = MatFrobeniusNorm(plastStrain);
    if (plastStrainNorm > 0.0)
      plasticIP += 1.0;
  }
 
  for (size_t co = 0; co < elem->num_vertices(); ++co)
  {
    //  init 'elemsWithCo' with 1, because co lies in 'elem'
    size_t elemsWithCo = 1;
 
    //  iterate neighbor elems and count how many elems contain the corner 'co'
    for (neighborElemIter it = vNeighborElems.begin();
        it != vNeighborElems.end(); ++it){
      if (ContainsPoint(*it, coCoord[co], aaPos) == true)
        elemsWithCo++;
    }
 
    //  scaling factor for averaging values out of all ips
    //  of the associated elements of corner co
    size_t scaleFac = elemsWithCo * geo.num_ip();
 
    locPlastIP(0,co) = plasticIP / scaleFac;
  }
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
void
MechOutputWriter<TDomain>::
eqPlastStrain_elem(LocalVector& locEqPlastStrain, TBaseElem* elem,
    const LocalVector& locU, SmartPtr<TDomain> dom)
{
  //  get vertices and extract corner coordinates
  typedef typename TDomain::position_accessor_type position_accessor_type;
  position_accessor_type& aaPos = dom->position_accessor();
  MathVector<dim> coCoord[domain_traits<dim>::MaxNumVerticesOfElem];
  const size_t numVertices = elem->num_vertices();
  for (size_t i = 0; i < numVertices; ++i) {
    coCoord[i] = aaPos[elem->vertex(i)];
  };
 
  //  prepare geometry for type and order
  DimFEGeometry<dim> geo;
    try{
    geo.update(elem, &(coCoord[0]),
        LFEID(LFEID::LAGRANGE, dim, 1), m_quadOrder);
  }
    UG_CATCH_THROW("SmallStrainMechOutput::eqPlastStrain_elem:"
          " Cannot update Finite Element Geometry.");
 
  //  get all neighbor elems which share a vertex with the given element 'elem'
  typename TDomain::grid_type& grid = *(dom->grid());
  typedef typename vector<TBaseElem*>::iterator neighborElemIter;
  vector<TBaseElem*> vNeighborElems;
  CollectNeighbors(vNeighborElems, elem, grid, NHT_VERTEX_NEIGHBORS);
 
  //  pointer to internal variable of current elem
  m_spMatLaw->internal_vars(elem);
 
  MathMatrix<dim, dim> GradU;
  number SumEqPlastStrainIP = 0.0;
  for (size_t ip = 0; ip < geo.num_ip(); ++ip)
  {
    //  get displacementGradient (GradU)
    m_spMatLaw->template DisplacementGradient<DimFEGeometry<dim> >(GradU, ip, geo, locU);
 
    number gamma = m_spMatLaw->plastic_multiplier(ip, GradU);
 
    //  sum over all ip`s
    SumEqPlastStrainIP += gamma * sqrt(2.0/3.0);
  }
 
  for (size_t co = 0; co < elem->num_vertices(); ++co)
  {
    //  init 'elemsWithCo' with 1, because co lies in 'elem'
    size_t elemsWithCo = 1;
 
    //  iterate neighbor elems and count how many elems contain the corner 'co'
    for (neighborElemIter it = vNeighborElems.begin();
        it != vNeighborElems.end(); ++it){
      if (ContainsPoint(*it, coCoord[co], aaPos) == true)
        elemsWithCo++;
    }
 
    //  scaling factor for averaging values out of all ips
    //  of the associated elements of corner co
    size_t scaleFac = elemsWithCo * geo.num_ip();
 
    locEqPlastStrain(0, co) = SumEqPlastStrainIP / scaleFac;
  }
}
MechOutputWriter<TDomain>:: {…}
template<typename TDomain>
void
MechOutputWriter<TDomain>::
normal_stress_strain_elem(LocalVector& locSigma, LocalVector& locEps, LocalVector& locStressFunc,
    LocalVector& locStrainFunc, TBaseElem* elem, const LocalVector& locU,
    SmartPtr<TDomain> dom)
{
//  get all neighbor elems which share a vertex with the given element 'elem'
  typename TDomain::grid_type& grid = *(dom->grid());
  vector<TBaseElem*> vNeighborElems;
  CollectNeighbors(vNeighborElems, elem, grid, NHT_VERTEX_NEIGHBORS);
  typedef typename vector<TBaseElem*>::iterator neighborElemIter;
 
  typedef typename TDomain::position_accessor_type position_accessor_type;
  position_accessor_type& aaPos = dom->position_accessor();
 
  //  coord and vertex array
  MathVector<dim> coCoord[domain_traits<dim>::MaxNumVerticesOfElem];
 
  //  get vertices and extract corner coordinates
  const size_t numVertices = elem->num_vertices();
  for (size_t i = 0; i < numVertices; ++i) {
    coCoord[i] = aaPos[elem->vertex(i)];
  };
 
  //  prepare geometry for type and order
  DimFEGeometry<dim> geo;
    try{
    geo.update(elem, &(coCoord[0]),
        LFEID(LFEID::LAGRANGE, dim, 1), m_quadOrder);
  }
    UG_CATCH_THROW("SmallStrainMechOutput::normal_stress_strain_elem:"
          " Cannot update Finite Element Geometry.");
 
  //  pointer to internal variable of current elem
  m_spMatLaw->internal_vars(elem);
 
  MathMatrix<dim, dim> GradU, eps, sigma, devSigma, SumEpsIP, SumSigmaIP, inelasticStrainTens, eps_elast;
  number normDevSig, normSig, normEps, SumNormDevSigIP, SumNormSigIP, SumNormEpsIP,
  SumVolSigmaIP, normElastStrain, SumNormElastStrainIP, normInelastStrain, SumNormInelastStrainIP;
 
  //  init summation-values
  SumEpsIP = 0.0; SumSigmaIP = 0.0; SumNormDevSigIP = 0.0; SumNormSigIP = 0.0;
  SumNormEpsIP = 0.0; SumVolSigmaIP = 0.0; SumNormElastStrainIP = 0.0; SumNormInelastStrainIP = 0.0;
 
  for(size_t ip = 0; ip < geo.num_ip(); ++ip)
  {
    //  get displacementGradient (GradU)
    m_spMatLaw->template DisplacementGradient<DimFEGeometry<dim> >(GradU, ip, geo, locU);
    //  get Cauchy stress-tensor (sigma) at ip
    m_spMatLaw->stressTensor(sigma, ip, GradU);
 
    //  get inelastic strain tensor at ip
    inelasticStrainTens = *(m_spMatLaw->inelastic_strain_tensor(ip));
 
    //  get linearized strain tensor (eps) at ip
    for (size_t i = 0; i < (size_t) dim; ++i)
      for (size_t J = 0; J < (size_t) dim; ++J)
      {
        eps[i][J] = 0.5 * (GradU[i][J] + GradU[J][i]);
        eps_elast[i][J] = eps[i][J] - inelasticStrainTens[i][J];
      }
 
    //  add values to local vector
    for (size_t i = 0; i < (size_t) dim; ++i)
    {
      SumEpsIP[i][i] += eps[i][i];
      SumSigmaIP[i][i] += sigma[i][i];
    }
 
    //  stress functions
    number trSigma = MatDeviatorTrace(sigma, devSigma);
    normDevSig = MatFrobeniusNorm(devSigma);
    normSig = MatFrobeniusNorm(sigma);
    SumVolSigmaIP += 1.0/3.0 * trSigma;
    SumNormDevSigIP += normDevSig;
    SumNormSigIP += normSig;
 
    //  strain functions
    normEps = MatFrobeniusNorm(eps);
    SumNormEpsIP += normEps;
 
    normElastStrain = MatFrobeniusNorm(eps_elast);
    SumNormElastStrainIP += normElastStrain;
 
    normInelastStrain = MatFrobeniusNorm(inelasticStrainTens);
    SumNormInelastStrainIP += normInelastStrain;
 
  } //end (ip)
 
  // loop corner
  for (size_t co = 0; co < numVertices; ++co)
  {
    //  TODO: is this geometry update necessary here?
 
    //  update geometry
    try{
      geo.update(elem, &(coCoord[0]),
          LFEID(LFEID::LAGRANGE, dim, 1), m_quadOrder);
    }
    UG_CATCH_THROW("SmallStrainMechOutput::normal_stress_strain_loc:"
            " Cannot update Finite Element Geometry.");
 
    //  init 'elemsWithCo' with 1, because co lies in 'elem'
    size_t elemsWithCo = 1;
 
    //  iterate neighbor elems and count how many elems contain the corner 'co'
    for (neighborElemIter it = vNeighborElems.begin();
        it != vNeighborElems.end(); ++it){
      if (ContainsPoint(*it, coCoord[co], aaPos) == true)
        elemsWithCo++;
    }
 
    //  scaling factor for averaging values out of all ips
    //  of the associated elements of corner co
    size_t scaleFac = elemsWithCo * geo.num_ip();
 
    //  add values to local vector
    for (size_t i = 0; i < (size_t) dim; ++i)
    {
      locEps(i, co) += SumEpsIP[i][i] / scaleFac;
      locSigma(i, co) += SumSigmaIP[i][i] / scaleFac;
    }
 
    locStressFunc(0, co) += SumNormSigIP / scaleFac;
    locStressFunc(1, co) += SumNormDevSigIP / scaleFac;
    locStressFunc(2, co) += SumVolSigmaIP / scaleFac;
 
    locStrainFunc(0, co) += SumNormEpsIP / scaleFac;
    locStrainFunc(1, co) += SumNormElastStrainIP / scaleFac;
    locStrainFunc(2, co) += SumNormInelastStrainIP / scaleFac;
  } //end(co)
 
 
  //  temporarily output for linear-elastic-benchmark-output:
  for (size_t co = 0; co < numVertices; ++co)
  {
    if ((coCoord[co][0] == 90.0) && (coCoord[co][1] == 0.0))
    {
      UG_LOG("At 2 (90.0,0.0): \n");
      for (size_t i = 0; i < (size_t) dim; ++i)
        UG_LOG("u("<< i << "," << co << "): " << locU(i,co) << "\n");
 
      UG_LOG("\n");
 
      vector<size_t> vNextIP;
      next_ips_to_point(vNextIP, coCoord[co], geo);
 
      MathMatrix<dim, dim> sigma;
      for (vector<size_t>::iterator it = vNextIP.begin();
              it != vNextIP.end(); ++it)
      {
        //  get displacementGradient (GradU) at ip
        m_spMatLaw->template DisplacementGradient<DimFEGeometry<dim> >(GradU, *it, geo, locU);
        //  get the Cauchy Stress Tensor (sigma) at ip
        m_spMatLaw->stressTensor(sigma, *it, GradU);
 
        UG_LOG("At " << geo.global_ip(*it) << ": \n");
        UG_LOG("sigma_yy: " << sigma[1][1] << "\n");
        UG_LOG("sigma: " << sigma << "\n");
      }
    }
 
    if ((coCoord[co][0] == 100.0) && (coCoord[co][1] == 100.0))
    {
      UG_LOG("At 4 (100.0,100.0): \n");
      for (size_t i = 0; i < (size_t) dim; ++i)
        UG_LOG("u("<< i << "," << co << "): " << locU(i,co) << "\n");
 
      UG_LOG("\n");
    }
 
    if ((coCoord[co][0] == 0.0) && (coCoord[co][1] == 100.0))
    {
      UG_LOG("At 5 (0.0,100.0): \n");
      for (size_t i = 0; i < (size_t) dim; ++i)
        UG_LOG("u("<< i << "," << co << "): " << locU(i,co) << "\n");
 
      UG_LOG("\n");
    }
 
  };
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
void
MechOutputWriter<TDomain>::
invariants_kirchhoff_stress_elem(LocalVector& locInvarKirchhoffStress,
    TBaseElem* elem, const LocalVector& u, SmartPtr<TDomain> dom)
{
  //  get vertices and extract corner coordinates
  typedef typename TDomain::position_accessor_type position_accessor_type;
  position_accessor_type& aaPos = dom->position_accessor();
  MathVector<dim> coCoord[domain_traits<dim>::MaxNumVerticesOfElem];
  const size_t numVertices = elem->num_vertices();
  for (size_t i = 0; i < numVertices; ++i) {
    coCoord[i] = aaPos[elem->vertex(i)];
  };
 
  //  prepare geometry for type and order
  DimFEGeometry<dim> geo;
  try{
    geo.update(elem, &(coCoord[0]),
        LFEID(LFEID::LAGRANGE, dim, 1), m_quadOrder);
  }
  UG_CATCH_THROW("MechOutputWriterFinite::invariants_kirchhoff_stress_elem:"
          " Cannot update Finite Element Geometry.");
 
  //  get all neighbor elems which share a vertex with the given element 'elem'
  typename TDomain::grid_type& grid = *(dom->grid());
  typedef typename vector<TBaseElem*>::iterator neighborElemIter;
  vector<TBaseElem*> vNeighborElems;
  CollectNeighbors(vNeighborElems, elem, grid, NHT_VERTEX_NEIGHBORS);
 
  //  pointer to internal variable of current elem
  m_spMatLaw->internal_vars(elem);
 
  MathMatrix<dim, dim> tauTens, tauTens2, GradU, sigma, F, Ident;
  MatIdentity(Ident);
  for (size_t co = 0; co < elem->num_vertices(); ++co)
  {
    number invar1_ip = 0.0;
    number invar2_ip = 0.0;
    number invar3_ip = 0.0;
 
    //  loop ip`s and determine the next ip`s to corner co!
    vector<size_t> vNextIP;
    next_ips_to_point(vNextIP, coCoord[co], geo);
 
    for (vector<size_t>::iterator it = vNextIP.begin();
                it != vNextIP.end(); ++it)
    {
      //  get deformation gradient at ip
      m_spMatLaw->template DisplacementGradient<DimFEGeometry<dim> >(GradU, *it, geo, u);
 
      //  get Cauchy stress-tensor (sigma) at ip
      m_spMatLaw->stressTensor(sigma, *it, GradU);
 
      //  compute the kirchhoff stress-tensor at ip
      MatAdd(F, GradU, Ident);
      number detF = Determinant(F);
      MatScale(tauTens, 1.0/detF, sigma);
 
 
      MatMultiply(tauTens2, tauTens, tauTens);
      number traceTauTens_ip = Trace(tauTens);
      number traceTauTens2_ip = Trace(tauTens2);
 
      invar1_ip += traceTauTens_ip;
      invar2_ip += 0.5 * (traceTauTens_ip * traceTauTens_ip - traceTauTens2_ip);
      invar3_ip += Determinant(tauTens);
    }
 
    //  init 'elemsWithCo' with 1, because co lies in 'elem'
    size_t elemsWithCo = 1;
 
    //  iterate neighbor elems and count how many elems contain the corner 'co'
    for (neighborElemIter it = vNeighborElems.begin();
        it != vNeighborElems.end(); ++it){
      if (ContainsPoint(*it, coCoord[co], aaPos) == true)
        elemsWithCo++;
    }
 
    //  scaling factor for averaging values out of all ips
    //  of the associated elements of corner co
    const size_t scaleFac = elemsWithCo * vNextIP.size();
 
    locInvarKirchhoffStress(0, co) = invar1_ip/scaleFac;
    locInvarKirchhoffStress(1, co) = invar2_ip/scaleFac;
    locInvarKirchhoffStress(2, co) = invar3_ip/scaleFac;
  }
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
void
MechOutputWriter<TDomain>::
preprocess()
{
  /*if (m_stressEV)
  {
    #ifdef UG_PARALLEL
      if (pcl::ProcRank() == 0){
        m_fileStressEV = fopen("sig_eigen.dat", "w");
      }
    #else
      m_fileStressEV = fopen("sig_eigen.dat", "w");
    #endif
  }*/
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
template<typename TFEGeom>
void
MechOutputWriter<TDomain>::
post_timestep_elem(const number time, SmartPtr<TDomain> dom, TFEGeom& geo,
    TBaseElem* elem, const LocalVector& u)
{
  typedef typename TDomain::position_accessor_type position_accessor_type;
  position_accessor_type& aaPos = dom->position_accessor();
 
  if (m_stressEV && (!m_bIP_values_written))
    if (ContainsPoint(elem, m_evalPointEV, aaPos))
      stress_eigenvalues_near_point(time, geo, u);
 
  //  COMPUTE NORMAL STRESSES OF THE STRESSTENSOR SIGMA AT IP
  if (m_normalStress)
    if (ContainsPoint(elem, m_evalPointNormStress, aaPos))
      normal_stress_near_point(time, geo, u);
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
void
MechOutputWriter<TDomain>::
post_timestep(const number time)
{
  //  write some material law data to console
  m_spMatLaw->write_data_to_console(time);
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
void
MechOutputWriter<TDomain>::
postprocess()
{
  /*if (m_stressEV)
  {
    #ifdef UG_PARALLEL
      if (pcl::ProcRank() == 0){
        fclose(m_fileStressEV);
      }
    #else
      fclose(m_fileStressEV);
    #endif
  }*/
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
template<typename TFEGeom>
void
MechOutputWriter<TDomain>::
stress_eigenvalues_near_point(const number time, TFEGeom& geo,
    const LocalVector& u)
{
  vector<size_t> vNextIP;
  next_ips_to_point(vNextIP, m_evalPointEV, geo);
 
  MathMatrix<dim, dim> SYMMSig, Sig, GradU;
  for (vector<size_t>::iterator it = vNextIP.begin();
          it != vNextIP.end(); ++it)
  {
    m_spMatLaw->template DisplacementGradient<TFEGeom>(GradU, *it, geo, u);
    m_spMatLaw->stressTensor(Sig, *it, GradU);
 
    if (dim == 3)
    {
      MatMultiplyMTM(SYMMSig, Sig);
 
      MathMatrix<3, 3, number> SigTens;
      for (size_t a = 0; a < 3; ++a)
        for (size_t b = 0; b < 3; ++b){
          SigTens[a][b] = SYMMSig[a][b];
        }
 
      MathVector<3, number> evMin, evMed, evMax;
      number SiglambdaMin, SiglambdaMed, SiglambdaMax;
 
      CalculateEigenvalues(SigTens, SiglambdaMin,
          SiglambdaMed, SiglambdaMax, evMin, evMed,
          evMax);
 
      number sqrSigLambdaMin = sqrt(SiglambdaMin);
      number sqrSigLambdaMed = sqrt(SiglambdaMed);
      number sqrSigLambdaMax = sqrt(SiglambdaMax);
 
      UG_LOG("minimal EigenValueSigma: " << sqrSigLambdaMin << "\n");
      UG_LOG("medium EigenValueSigma: " << sqrSigLambdaMed << "\n");
      UG_LOG("maximal EigenValueSigma: " << sqrSigLambdaMax << "\n");
 
    /*  number absSiglambdamax = abs(sqrSigLambdaMax);
 
    #ifdef UG_PARALLEL
      if (pcl::ProcRank() == 0){
        fprintf(m_fileStressEV, "%lg %lg \n ", time, absSiglambdamax);
      }
    #else
      fprintf(m_fileStressEV, "%lg %lg \n ", time, absSiglambdamax);
    #endif*/
 
    }
 
    //  to be sure that data is only written once:
    m_bIP_values_written = true;
  } // end vNextIP-iteration
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
template<typename TFEGeom>
void
MechOutputWriter<TDomain>::
normal_stress_near_point(const number time, TFEGeom& geo, const LocalVector& u)
{
  vector<size_t> vNextIP;
  next_ips_to_point(vNextIP, m_evalPointNormStress, geo);
 
  MathMatrix<dim, dim> Sig, GradU;
  for (vector<size_t>::iterator it = vNextIP.begin();
          it != vNextIP.end(); ++it)
  {
    m_spMatLaw->template DisplacementGradient<TFEGeom>(GradU, *it, geo, u);
    m_spMatLaw->stressTensor(Sig, *it, GradU);
 
    UG_LOG("At " << geo.global_ip(*it) << ": \n");
    UG_LOG("sigma_xx: " << Sig[0][0] << "\n");
    UG_LOG("sigma_yy: " << Sig[1][1] << "\n");
    UG_LOG("sigma: " << Sig << "\n");
    UG_LOG("######### \n");
 
    //  to be sure that data is only written once:
    //m_bIP_values_written = true;
  } // end vNextIP-iteration
}
MechOutputWriter<TDomain>:: {…}
 
template<typename TDomain>
template<typename TFEGeom>
void
MechOutputWriter<TDomain>::
next_ips_to_point(vector<size_t>& vNextIP, const MathVector<dim>& point,
    const TFEGeom& geo)
{
  number dist = std::numeric_limits<number>::max();
 
  //  determine shortest distance from an ip to a point
  for (size_t ip = 0; ip < geo.num_ip(); ++ip)
  {
    number dist_ip = VecDistance(geo.global_ip(ip), point);
    if (dist_ip < dist)
      dist = dist_ip;
  }
 
  //  determine all ip`s with shortest distance to a point
  for (size_t ip = 0; ip < geo.num_ip(); ++ip)
  {
    number dist_ip = VecDistance(geo.global_ip(ip), point);
    if (dist_ip == dist)
      vNextIP.push_back(ip);
  }
}
MechOutputWriter<TDomain>:: {…}
 
 
} // namespace SmallStrainMechanics
} // namespace ug
 
#endif /* MECH_OUTPUT_WRITER_IMPL_H_ */