Compute1DFiniteStrain defines a strain increment for finite strains in 1D problems, handling strains in other two directions. More...

#include <Compute1DFiniteStrain.h>

Inheritance diagram for Compute1DFiniteStrain:

[legend]

Public Member Functions
	Compute1DFiniteStrain (const InputParameters &parameters)

void	computeProperties () override

void	initialSetup () override

Static Public Member Functions
static InputParameters	validParams ()

static MooseEnum	decompositionType ()

Protected Member Functions
virtual Real	computeGradDispYY ()=0
	Computes the current dUy/dY; as a virtual function, this function is overwritten for the specific geometries defined by inheriting classes. More...

virtual Real	computeGradDispYYOld ()=0
	Computes the old dUy/dY; as a virtual function, this function is overwritten for the specific geometries defined by inheriting classes. More...

virtual Real	computeGradDispZZ ()=0
	Computes the current dUz/dz; as a virtual function, this function is overwritten for the specific geometries defined by inheriting classes. More...

virtual Real	computeGradDispZZOld ()=0
	Computes the old dUz/dz; as a virtual function, this function is overwritten for the specific geometries defined by inheriting classes. More...

virtual void	computeQpStrain ()

virtual void	computeQpIncrements (RankTwoTensor &e, RankTwoTensor &r)

virtual void	initQpStatefulProperties () override

void	subtractEigenstrainIncrementFromStrain (RankTwoTensor &strain)

virtual void	displacementIntegrityCheck ()

Protected Attributes
std::vector< RankTwoTensor >	_Fhat

std::vector< const VariableGradient * >	_grad_disp_old

MaterialProperty< RankTwoTensor > &	_strain_rate

MaterialProperty< RankTwoTensor > &	_strain_increment

MaterialProperty< RankTwoTensor > &	_rotation_increment

MaterialProperty< RankTwoTensor > &	_deformation_gradient

const MaterialProperty< RankTwoTensor > &	_mechanical_strain_old

const MaterialProperty< RankTwoTensor > &	_total_strain_old

std::vector< const MaterialProperty< RankTwoTensor > * >	_eigenstrains_old

unsigned int	_ndisp
	Coupled displacement variables. More...

std::vector< const VariableValue * >	_disp

std::vector< const VariableGradient * >	_grad_disp

const std::string	_base_name

MaterialProperty< RankTwoTensor > &	_mechanical_strain

MaterialProperty< RankTwoTensor > &	_total_strain

std::vector< MaterialPropertyName >	_eigenstrain_names

std::vector< const MaterialProperty< RankTwoTensor > * >	_eigenstrains

const MaterialProperty< RankTwoTensor > *	_global_strain

const bool	_volumetric_locking_correction

const Real &	_current_elem_volume

Private Types
enum	DecompMethod { DecompMethod::TaylorExpansion, DecompMethod::EigenSolution }

Private Attributes
const DecompMethod	_decomposition_method

Detailed Description

Compute1DFiniteStrain defines a strain increment for finite strains in 1D problems, handling strains in other two directions.

It contains virtual methods to define the displacement gradients as a general nonzero value.

Definition at line 24 of file Compute1DFiniteStrain.h.

Member Enumeration Documentation

◆ DecompMethod

enum ComputeFiniteStrain::DecompMethod

strongprivateinherited

Enumerator
TaylorExpansion
EigenSolution

Definition at line 40 of file ComputeFiniteStrain.h.

   {
     TaylorExpansion,
     EigenSolution
   };

Constructor & Destructor Documentation

◆ Compute1DFiniteStrain()

Compute1DFiniteStrain::Compute1DFiniteStrain ( const InputParameters & parameters )

Definition at line 25 of file Compute1DFiniteStrain.C.

   : ComputeFiniteStrain(parameters)
 {
 }

Member Function Documentation

◆ computeGradDispYY()

virtual Real Compute1DFiniteStrain::computeGradDispYY ( )

protectedpure virtual

Computes the current dUy/dY; as a virtual function, this function is overwritten for the specific geometries defined by inheriting classes.

Implemented in ComputeAxisymmetric1DFiniteStrain.

Referenced by computeProperties().

◆ computeGradDispYYOld()

virtual Real Compute1DFiniteStrain::computeGradDispYYOld ( )

protectedpure virtual

Computes the old dUy/dY; as a virtual function, this function is overwritten for the specific geometries defined by inheriting classes.

Implemented in ComputeAxisymmetric1DFiniteStrain.

Referenced by computeProperties().

◆ computeGradDispZZ()

virtual Real Compute1DFiniteStrain::computeGradDispZZ ( )

protectedpure virtual

Computes the current dUz/dz; as a virtual function, this function is overwritten for the specific geometries defined by inheriting classes.

Implemented in ComputeAxisymmetric1DFiniteStrain.

Referenced by computeProperties().

◆ computeGradDispZZOld()

virtual Real Compute1DFiniteStrain::computeGradDispZZOld ( )

protectedpure virtual

Computes the old dUz/dz; as a virtual function, this function is overwritten for the specific geometries defined by inheriting classes.

Implemented in ComputeAxisymmetric1DFiniteStrain.

Referenced by computeProperties().

◆ computeProperties()

void Compute1DFiniteStrain::computeProperties ( )

override

Definition at line 31 of file Compute1DFiniteStrain.C.

 {
   for (_qp = 0; _qp < _qrule->n_points(); ++_qp)
   {
     RankTwoTensor A((*_grad_disp[0])[_qp],
                     (*_grad_disp[1])[_qp],
                     (*_grad_disp[2])[_qp]); // Deformation gradient
     RankTwoTensor Fbar((*_grad_disp_old[0])[_qp],
                        (*_grad_disp_old[1])[_qp],
                        (*_grad_disp_old[2])[_qp]); // Old Deformation gradient
  
     // Compute the displacement gradient dUy/dy and dUz/dz value for 1D problems
     A(1, 1) = computeGradDispYY();
     A(2, 2) = computeGradDispZZ();
  
     Fbar(1, 1) = computeGradDispYYOld();
     Fbar(2, 2) = computeGradDispZZOld();
  
     // Gauss point deformation gradient
     _deformation_gradient[_qp] = A;
     _deformation_gradient[_qp].addIa(1.0);
  
     A -= Fbar; // very nearly A = gradU - gradUold, adapted to cylindrical coords
  
     Fbar.addIa(1.0); // Fbar = ( I + gradUold)
  
     // Incremental deformation gradient _Fhat = I + A Fbar^-1
     _Fhat[_qp] = A * Fbar.inverse();
     _Fhat[_qp].addIa(1.0);
   }
  
   for (_qp = 0; _qp < _qrule->n_points(); ++_qp)
     computeQpStrain();
 }

◆ computeQpIncrements()

void ComputeFiniteStrain::computeQpIncrements	(	RankTwoTensor &	e,
		RankTwoTensor &	r
	)

protectedvirtualinherited

Definition at line 137 of file ComputeFiniteStrain.C.

 {
   switch (_decomposition_method)
   {
     case DecompMethod::TaylorExpansion:
     {
       // inverse of _Fhat
       const RankTwoTensor invFhat = _Fhat[_qp].inverse();
  
       // A = I - _Fhat^-1
       RankTwoTensor A(RankTwoTensor::initIdentity);
       A -= invFhat;
  
       // Cinv - I = A A^T - A - A^T;
       RankTwoTensor Cinv_I = A * A.transpose() - A - A.transpose();
  
       // strain rate D from Taylor expansion, Chat = (-1/2(Chat^-1 - I) + 1/4*(Chat^-1 - I)^2 + ...
       total_strain_increment = -Cinv_I * 0.5 + Cinv_I * Cinv_I * 0.25;
  
       const Real a[3] = {invFhat(1, 2) - invFhat(2, 1),
                          invFhat(2, 0) - invFhat(0, 2),
                          invFhat(0, 1) - invFhat(1, 0)};
  
       Real q = (a[0] * a[0] + a[1] * a[1] + a[2] * a[2]) / 4.0;
       Real trFhatinv_1 = invFhat.trace() - 1.0;
       const Real p = trFhatinv_1 * trFhatinv_1 / 4.0;
  
       // cos theta_a
       const Real C1_squared = p +
                               3.0 * Utility::pow<2>(p) * (1.0 - (p + q)) / Utility::pow<2>(p + q) -
                               2.0 * Utility::pow<3>(p) * (1.0 - (p + q)) / Utility::pow<3>(p + q);
       mooseAssert(C1_squared >= 0.0,
                   "Cannot take square root of a negative number. This may happen when elements "
                   "become heavily distorted.");
       const Real C1 = std::sqrt(C1_squared);
  
       Real C2;
       if (q > 0.01)
         // (1-cos theta_a)/4q
         C2 = (1.0 - C1) / (4.0 * q);
       else
         // alternate form for small q
         C2 = 0.125 + q * 0.03125 * (Utility::pow<2>(p) - 12.0 * (p - 1.0)) / Utility::pow<2>(p) +
              Utility::pow<2>(q) * (p - 2.0) * (Utility::pow<2>(p) - 10.0 * p + 32.0) /
                  Utility::pow<3>(p) +
              Utility::pow<3>(q) *
                  (1104.0 - 992.0 * p + 376.0 * Utility::pow<2>(p) - 72.0 * Utility::pow<3>(p) +
                   5.0 * Utility::pow<4>(p)) /
                  (512.0 * Utility::pow<4>(p));
  
       const Real C3_test =
           (p * q * (3.0 - q) + Utility::pow<3>(p) + Utility::pow<2>(q)) / Utility::pow<3>(p + q);
       mooseAssert(C3_test >= 0.0,
                   "Cannot take square root of a negative number. This may happen when elements "
                   "become heavily distorted.");
       const Real C3 = 0.5 * std::sqrt(C3_test); // sin theta_a/(2 sqrt(q))
  
       // Calculate incremental rotation. Note that this value is the transpose of that from Rashid,
       // 93, so we transpose it before storing
       RankTwoTensor R_incr;
       R_incr.addIa(C1);
       for (unsigned int i = 0; i < 3; ++i)
         for (unsigned int j = 0; j < 3; ++j)
           R_incr(i, j) += C2 * a[i] * a[j];
  
       R_incr(0, 1) += C3 * a[2];
       R_incr(0, 2) -= C3 * a[1];
       R_incr(1, 0) -= C3 * a[2];
       R_incr(1, 2) += C3 * a[0];
       R_incr(2, 0) += C3 * a[1];
       R_incr(2, 1) -= C3 * a[0];
  
       rotation_increment = R_incr.transpose();
       break;
     }
  
     case DecompMethod::EigenSolution:
     {
       std::vector<Real> e_value(3);
       RankTwoTensor e_vector, N1, N2, N3;
  
       RankTwoTensor Chat = _Fhat[_qp].transpose() * _Fhat[_qp];
       Chat.symmetricEigenvaluesEigenvectors(e_value, e_vector);
  
       const Real lambda1 = std::sqrt(e_value[0]);
       const Real lambda2 = std::sqrt(e_value[1]);
       const Real lambda3 = std::sqrt(e_value[2]);
  
       N1.vectorOuterProduct(e_vector.column(0), e_vector.column(0));
       N2.vectorOuterProduct(e_vector.column(1), e_vector.column(1));
       N3.vectorOuterProduct(e_vector.column(2), e_vector.column(2));
  
       const RankTwoTensor Uhat = N1 * lambda1 + N2 * lambda2 + N3 * lambda3;
       const RankTwoTensor invUhat(Uhat.inverse());
  
       rotation_increment = _Fhat[_qp] * invUhat;
  
       total_strain_increment =
           N1 * std::log(lambda1) + N2 * std::log(lambda2) + N3 * std::log(lambda3);
       break;
     }
  
     default:
       mooseError("ComputeFiniteStrain Error: Pass valid decomposition type: TaylorExpansion or "
                  "EigenSolution.");
   }
 }

Referenced by ComputeFiniteStrain::computeQpStrain().

◆ computeQpStrain()

void ComputeFiniteStrain::computeQpStrain ( )

protectedvirtualinherited

Definition at line 105 of file ComputeFiniteStrain.C.

 {
   RankTwoTensor total_strain_increment;
  
   // two ways to calculate these increments: TaylorExpansion(default) or EigenSolution
   computeQpIncrements(total_strain_increment, _rotation_increment[_qp]);
  
   _strain_increment[_qp] = total_strain_increment;
  
   // Remove the eigenstrain increment
   subtractEigenstrainIncrementFromStrain(_strain_increment[_qp]);
  
   if (_dt > 0)
     _strain_rate[_qp] = _strain_increment[_qp] / _dt;
   else
     _strain_rate[_qp].zero();
  
   // Update strain in intermediate configuration
   _mechanical_strain[_qp] = _mechanical_strain_old[_qp] + _strain_increment[_qp];
   _total_strain[_qp] = _total_strain_old[_qp] + total_strain_increment;
  
   // Rotate strain to current configuration
   _mechanical_strain[_qp] =
       _rotation_increment[_qp] * _mechanical_strain[_qp] * _rotation_increment[_qp].transpose();
   _total_strain[_qp] =
       _rotation_increment[_qp] * _total_strain[_qp] * _rotation_increment[_qp].transpose();
  
   if (_global_strain)
     _total_strain[_qp] += (*_global_strain)[_qp];
 }

Referenced by ComputeFiniteStrain::computeProperties(), computeProperties(), Compute2DFiniteStrain::computeProperties(), and ComputeRSphericalFiniteStrain::computeProperties().

◆ decompositionType()

MooseEnum ComputeFiniteStrain::decompositionType ( )

staticinherited

Definition at line 17 of file ComputeFiniteStrain.C.

 {
   return MooseEnum("TaylorExpansion EigenSolution", "TaylorExpansion");
 }

Referenced by TensorMechanicsActionBase::validParams(), and ComputeFiniteStrain::validParams().

◆ displacementIntegrityCheck()

void ComputeStrainBase::displacementIntegrityCheck ( )

protectedvirtualinherited

Reimplemented in Compute2DFiniteStrain, Compute2DIncrementalStrain, and Compute2DSmallStrain.

Definition at line 94 of file ComputeStrainBase.C.

 {
   // Checking for consistency between mesh size and length of the provided displacements vector
   if (_ndisp != _mesh.dimension())
     paramError(
         "displacements",
         "The number of variables supplied in 'displacements' must match the mesh dimension.");
 }

Referenced by ComputeStrainBase::initialSetup().

◆ initialSetup()

void ComputeIncrementalStrainBase::initialSetup ( )

overrideinherited

Definition at line 38 of file ComputeIncrementalStrainBase.C.

 {
   ComputeStrainBase::initialSetup();
   for (unsigned int i = 0; i < 3; ++i)
   {
     if (_fe_problem.isTransient() && i < _ndisp)
       _grad_disp_old[i] = &coupledGradientOld("displacements", i);
     else
       _grad_disp_old[i] = &_grad_zero;
   }
 }

Referenced by ComputeAxisymmetric1DFiniteStrain::initialSetup(), ComputeAxisymmetricRZIncrementalStrain::initialSetup(), ComputeRSphericalFiniteStrain::initialSetup(), ComputeRSphericalIncrementalStrain::initialSetup(), and ComputeAxisymmetric1DIncrementalStrain::initialSetup().

◆ initQpStatefulProperties()

void ComputeIncrementalStrainBase::initQpStatefulProperties ( )

overrideprotectedvirtualinherited

Reimplemented from ComputeStrainBase.

Reimplemented in ComputeCosseratIncrementalSmallStrain.

Definition at line 51 of file ComputeIncrementalStrainBase.C.

 {
   _mechanical_strain[_qp].zero();
   _total_strain[_qp].zero();
   _deformation_gradient[_qp].setToIdentity();
 }

Referenced by ComputeCosseratIncrementalSmallStrain::initQpStatefulProperties().

◆ subtractEigenstrainIncrementFromStrain()

void ComputeIncrementalStrainBase::subtractEigenstrainIncrementFromStrain ( RankTwoTensor & strain )

protectedinherited

Definition at line 59 of file ComputeIncrementalStrainBase.C.

 {
   for (unsigned int i = 0; i < _eigenstrains.size(); ++i)
   {
     strain -= (*_eigenstrains[i])[_qp];
     strain += (*_eigenstrains_old[i])[_qp];
   }
 }

Referenced by ComputeIncrementalSmallStrain::computeProperties(), ComputeCosseratIncrementalSmallStrain::computeQpProperties(), and ComputeFiniteStrain::computeQpStrain().

◆ validParams()

InputParameters Compute1DFiniteStrain::validParams ( )

static

Definition at line 17 of file Compute1DFiniteStrain.C.

 {
   InputParameters params = ComputeFiniteStrain::validParams();
   params.addClassDescription("Compute strain increment for finite strain in 1D problem");
  
   return params;
 }