poroelasticity_elements.cc
Go to the documentation of this file.
1 // LIC// ====================================================================
2 // LIC// This file forms part of oomph-lib, the object-oriented,
3 // LIC// multi-physics finite-element library, available
4 // LIC// at http://www.oomph-lib.org.
5 // LIC//
6 // LIC// Copyright (C) 2006-2024 Matthias Heil and Andrew Hazel
7 // LIC//
8 // LIC// This library is free software; you can redistribute it and/or
9 // LIC// modify it under the terms of the GNU Lesser General Public
10 // LIC// License as published by the Free Software Foundation; either
11 // LIC// version 2.1 of the License, or (at your option) any later version.
12 // LIC//
13 // LIC// This library is distributed in the hope that it will be useful,
14 // LIC// but WITHOUT ANY WARRANTY; without even the implied warranty of
15 // LIC// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
16 // LIC// Lesser General Public License for more details.
17 // LIC//
18 // LIC// You should have received a copy of the GNU Lesser General Public
19 // LIC// License along with this library; if not, write to the Free Software
20 // LIC// Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
21 // LIC// 02110-1301 USA.
22 // LIC//
23 // LIC// The authors may be contacted at oomph-lib@maths.man.ac.uk.
24 // LIC//
25 // LIC//====================================================================
27 
28 namespace oomph
29 {
30  /// Static default value for timescale ratio (1.0 -- for natural scaling)
31  template<unsigned DIM>
33 
34  /// Static default value for the density ratio
35  template<unsigned DIM>
37 
38  /// Static default value for 1/k
39  template<unsigned DIM>
41 
42  /// Static default value for alpha
43  template<unsigned DIM>
45 
46  /// Static default value for the porosity
47  template<unsigned DIM>
49 
50  //======================================================================
51  /// Compute the strain tensor at local coordinate s
52  //======================================================================
53  template<unsigned DIM>
55  const Vector<double>& s, DenseMatrix<double>& strain) const
56  {
57 #ifdef PARANOID
58  if ((strain.ncol() != DIM) || (strain.nrow() != DIM))
59  {
60  std::ostringstream error_message;
61  error_message << "Strain matrix is " << strain.ncol() << " x "
62  << strain.nrow() << ", but dimension of the equations is "
63  << DIM << std::endl;
64  throw OomphLibError(
65  error_message.str(), OOMPH_CURRENT_FUNCTION, OOMPH_EXCEPTION_LOCATION);
66  }
67 
68  /*
69  //Find out how many position types there are
70  unsigned n_position_type = this->nnodal_position_type();
71 
72  if(n_position_type != 1)
73  {
74  throw OomphLibError(
75  "PoroElasticity is not yet implemented for more than one position type",
76  OOMPH_CURRENT_FUNCTION,
77  OOMPH_EXCEPTION_LOCATION);
78  }
79  */
80 #endif
81 
82 
83  // Find out how many nodes there are in the element
84  unsigned n_node = nnode();
85 
86  // Find the indices at which the local velocities are stored
87  unsigned u_nodal_index[DIM];
88  for (unsigned i = 0; i < DIM; i++)
89  {
90  u_nodal_index[i] = u_index(i);
91  }
92 
93  // Set up memory for the shape and derivative functions
94  Shape psi(n_node);
95  DShape dpsidx(n_node, DIM);
96 
97  // Call the derivatives of the shape functions
98  (void)dshape_eulerian(s, psi, dpsidx);
99 
100  // Calculate interpolated values of the derivative of global position
101  DenseMatrix<double> interpolated_dudx(DIM, DIM, 0.0);
102 
103  // Loop over nodes
104  for (unsigned l = 0; l < n_node; l++)
105  {
106  // Loop over velocity components
107  for (unsigned i = 0; i < DIM; i++)
108  {
109  // Get the nodal value
110  const double u_value = this->nodal_value(l, u_nodal_index[i]);
111 
112  // Loop over derivative directions
113  for (unsigned j = 0; j < DIM; j++)
114  {
115  interpolated_dudx(i, j) += u_value * dpsidx(l, j);
116  }
117  }
118  }
119 
120  /// Now fill in the entries of the strain tensor
121  for (unsigned i = 0; i < DIM; i++)
122  {
123  // Do upper half of matrix
124  // Note that j must be signed here for the comparison test to work
125  // Also i must be cast to an int
126  for (int j = (DIM - 1); j >= static_cast<int>(i); j--)
127  {
128  // Off diagonal terms
129  if (static_cast<int>(i) != j)
130  {
131  strain(i, j) =
132  0.5 * (interpolated_dudx(i, j) + interpolated_dudx(j, i));
133  }
134  // Diagonal terms will including growth factor when it comes back in
135  else
136  {
137  strain(i, i) = interpolated_dudx(i, i);
138  }
139  }
140  // Matrix is symmetric so just copy lower half
141  for (int j = (i - 1); j >= 0; j--)
142  {
143  strain(i, j) = strain(j, i);
144  }
145  }
146  }
147 
148 
149  //======================================================================
150  /// Compute the Cauchy stress tensor at local coordinate s for
151  /// displacement formulation.
152  //======================================================================
153  template<unsigned DIM>
155  const Vector<double>& s, DenseMatrix<double>& stress) const
156  {
157 #ifdef PARANOID
158  if ((stress.ncol() != DIM) || (stress.nrow() != DIM))
159  {
160  std::ostringstream error_message;
161  error_message << "Stress matrix is " << stress.ncol() << " x "
162  << stress.nrow() << ", but dimension of the equations is "
163  << DIM << std::endl;
164  throw OomphLibError(
165  error_message.str(), OOMPH_CURRENT_FUNCTION, OOMPH_EXCEPTION_LOCATION);
166  }
167 #endif
168 
169  // Get strain
170  DenseMatrix<double> strain(DIM, DIM);
171  this->get_strain(s, strain);
172 
173  // Now fill in the entries of the stress tensor without exploiting
174  // symmetry -- sorry too lazy. This fct is only used for
175  // postprocessing anyway...
176  for (unsigned i = 0; i < DIM; i++)
177  {
178  for (unsigned j = 0; j < DIM; j++)
179  {
180  stress(i, j) = 0.0;
181  for (unsigned k = 0; k < DIM; k++)
182  {
183  for (unsigned l = 0; k < DIM; k++)
184  {
185  stress(i, j) += this->E(i, j, k, l) * strain(k, l);
186  }
187  }
188  }
189  }
190  }
191 
192 
193  /// Performs a div-conserving transformation of the vector basis
194  /// functions from the reference element to the actual element
195  template<unsigned DIM>
197  const Vector<double>& s,
198  const Shape& q_basis_local,
199  Shape& psi,
200  DShape& dpsi,
201  Shape& q_basis) const
202  {
203  // Call the (geometric) shape functions and their derivatives
204  //(void)this->dshape_eulerian(s,psi,dpsi);
205  this->dshape_local(s, psi, dpsi);
206 
207  // Storage for the (geometric) jacobian and its inverse
208  DenseMatrix<double> jacobian(DIM), inverse_jacobian(DIM);
209 
210  // Get the jacobian of the geometric mapping and its determinant
211  double det = local_to_eulerian_mapping(dpsi, jacobian, inverse_jacobian);
212 
213  // Transform the derivative of the geometric basis so that it's w.r.t.
214  // global coordinates
215  this->transform_derivatives(inverse_jacobian, dpsi);
216 
217  // Get the number of computational basis vectors
218  const unsigned n_q_basis = this->nq_basis();
219 
220  // Loop over the basis vectors
221  for (unsigned l = 0; l < n_q_basis; l++)
222  {
223  // Loop over the spatial components
224  for (unsigned i = 0; i < DIM; i++)
225  {
226  // Zero the basis
227  q_basis(l, i) = 0.0;
228  }
229  }
230 
231  // Loop over the spatial components
232  for (unsigned i = 0; i < DIM; i++)
233  {
234  // And again
235  for (unsigned j = 0; j < DIM; j++)
236  {
237  // Get the element of the jacobian (must transpose it due to different
238  // conventions) and divide by the determinant
239  double jac_trans = jacobian(j, i) / det;
240 
241  // Loop over the computational basis vectors
242  for (unsigned l = 0; l < n_q_basis; l++)
243  {
244  // Apply the appropriate div-conserving mapping
245  q_basis(l, i) += jac_trans * q_basis_local(l, j);
246  }
247  }
248  }
249 
250  // Scale the basis by the ratio of the length of the edge to the length of
251  // the corresponding edge on the reference element
252  scale_basis(q_basis);
253 
254  return det;
255  }
256 
257  /// Output FE representation of soln: x,y,u1,u2,q1,q2,div_q,p at
258  /// Nplot^DIM plot points
259  template<unsigned DIM>
260  void PoroelasticityEquations<DIM>::output(std::ostream& outfile,
261  const unsigned& nplot)
262  {
263  // Vector of local coordinates
264  Vector<double> s(DIM);
265 
266  // Tecplot header info
267  outfile << tecplot_zone_string(nplot);
268 
269  // Loop over plot points
270  unsigned num_plot_points = nplot_points(nplot);
271 
272  for (unsigned iplot = 0; iplot < num_plot_points; iplot++)
273  {
274  // Get local coordinates of plot point
275  get_s_plot(iplot, nplot, s);
276 
277  // Output the components of the position
278  for (unsigned i = 0; i < DIM; i++)
279  {
280  outfile << interpolated_x(s, i) << " ";
281  }
282 
283  // Output the components of the FE representation of u at s
284  for (unsigned i = 0; i < DIM; i++)
285  {
286  outfile << interpolated_u(s, i) << " ";
287  }
288 
289  // Output the components of the FE representation of q at s
290  for (unsigned i = 0; i < DIM; i++)
291  {
292  outfile << interpolated_q(s, i) << " ";
293  }
294 
295  // Output FE representation of div u at s
296  outfile << interpolated_div_q(s) << " ";
297 
298  // Output FE representation of p at s
299  outfile << interpolated_p(s) << " ";
300 
301  const unsigned n_node = this->nnode();
302 
303  Shape psi(n_node);
304  shape(s, psi);
305 
306  Vector<double> interpolated_du_dt(DIM, 0.0);
307 
308  for (unsigned i = 0; i < DIM; i++)
309  {
310  for (unsigned l = 0; l < n_node; l++)
311  {
312  interpolated_du_dt[i] += du_dt(l, i) * psi(l);
313  }
314  outfile << interpolated_du_dt[i] << " ";
315  }
316 
317  Vector<double> interpolated_d2u_dt2(DIM, 0.0);
318 
319  for (unsigned i = 0; i < DIM; i++)
320  {
321  for (unsigned l = 0; l < n_node; l++)
322  {
323  interpolated_d2u_dt2[i] += d2u_dt2(l, i) * psi(l);
324  }
325  outfile << interpolated_d2u_dt2[i] << " ";
326  }
327 
328  const unsigned n_q_basis = this->nq_basis();
329  const unsigned n_q_basis_edge = this->nq_basis_edge();
330 
331  Shape q_basis(n_q_basis, DIM), q_basis_local(n_q_basis, DIM);
332  this->get_q_basis_local(s, q_basis_local);
333  (void)this->transform_basis(s, q_basis_local, psi, q_basis);
334 
335  Vector<double> interpolated_dq_dt(DIM, 0.0);
336 
337  for (unsigned i = 0; i < DIM; i++)
338  {
339  for (unsigned l = 0; l < n_q_basis_edge; l++)
340  {
341  interpolated_dq_dt[i] += dq_edge_dt(l) * q_basis(l, i);
342  }
343  for (unsigned l = n_q_basis_edge; l < n_q_basis; l++)
344  {
345  interpolated_dq_dt[i] +=
346  dq_internal_dt(l - n_q_basis_edge) * q_basis(l, i);
347  }
348  outfile << interpolated_dq_dt[i] << " ";
349  }
350 
351  outfile << std::endl;
352  }
353 
354  // Write tecplot footer (e.g. FE connectivity lists)
355  this->write_tecplot_zone_footer(outfile, nplot);
356  }
357 
358  /// Output FE representation of exact soln: x,y,u1,u2,q1,q2,div_q,p at
359  /// Nplot^DIM plot points
360  template<unsigned DIM>
362  std::ostream& outfile,
363  const unsigned& nplot,
365  {
366  // Vector of local coordinates
367  Vector<double> s(DIM);
368 
369  // Vector for coordintes
370  Vector<double> x(DIM);
371 
372  // Tecplot header info
373  outfile << this->tecplot_zone_string(nplot);
374 
375  // Exact solution Vector
376  Vector<double> exact_soln(2 * DIM + 2);
377 
378  // Loop over plot points
379  unsigned num_plot_points = this->nplot_points(nplot);
380 
381  for (unsigned iplot = 0; iplot < num_plot_points; iplot++)
382  {
383  // Get local coordinates of plot point
384  this->get_s_plot(iplot, nplot, s);
385 
386  // Get x position as Vector
387  this->interpolated_x(s, x);
388 
389  // Get exact solution at this point
390  (*exact_soln_pt)(x, exact_soln);
391 
392  // Output x,y,q_exact,p_exact,div_q_exact
393  for (unsigned i = 0; i < DIM; i++)
394  {
395  outfile << x[i] << " ";
396  }
397  for (unsigned i = 0; i < 2 * DIM + 2; i++)
398  {
399  outfile << exact_soln[i] << " ";
400  }
401  outfile << std::endl;
402  }
403 
404  // Write tecplot footer (e.g. FE connectivity lists)
405  this->write_tecplot_zone_footer(outfile, nplot);
406  }
407 
408  /// Output FE representation of exact soln: x,y,u1,u2,div_q,p at
409  /// Nplot^DIM plot points. Unsteady version
410  template<unsigned DIM>
412  std::ostream& outfile,
413  const unsigned& nplot,
414  const double& time,
416  {
417  // Vector of local coordinates
418  Vector<double> s(DIM);
419 
420  // Vector for coordintes
421  Vector<double> x(DIM);
422 
423  // Tecplot header info
424  outfile << this->tecplot_zone_string(nplot);
425 
426  // Exact solution Vector
427  Vector<double> exact_soln(2 * DIM + 2);
428 
429  // Loop over plot points
430  unsigned num_plot_points = this->nplot_points(nplot);
431 
432  for (unsigned iplot = 0; iplot < num_plot_points; iplot++)
433  {
434  // Get local coordinates of plot point
435  this->get_s_plot(iplot, nplot, s);
436 
437  // Get x position as Vector
438  this->interpolated_x(s, x);
439 
440  // Get exact solution at this point
441  (*exact_soln_pt)(time, x, exact_soln);
442 
443  // Output x,y,q_exact,p_exact,div_q_exact
444  for (unsigned i = 0; i < DIM; i++)
445  {
446  outfile << x[i] << " ";
447  }
448  for (unsigned i = 0; i < 2 * DIM + 2; i++)
449  {
450  outfile << exact_soln[i] << " ";
451  }
452  outfile << std::endl;
453  }
454 
455  // Write tecplot footer (e.g. FE connectivity lists)
456  this->write_tecplot_zone_footer(outfile, nplot);
457  }
458 
459  /// Compute the error between the FE solution and the exact solution
460  /// using the H(div) norm for q and L^2 norm for u and p
461  template<unsigned DIM>
463  std::ostream& outfile,
465  Vector<double>& error,
466  Vector<double>& norm)
467  {
468  for (unsigned i = 0; i < 3; i++)
469  {
470  error[i] = 0.0;
471  norm[i] = 0.0;
472  }
473 
474  // Vector of local coordinates
475  Vector<double> s(DIM);
476 
477  // Vector for coordinates
478  Vector<double> x(DIM);
479 
480  // Set the value of n_intpt
481  unsigned n_intpt = this->integral_pt()->nweight();
482 
483  outfile << "ZONE" << std::endl;
484 
485  // Exact solution Vector (u,v,[w])
486  Vector<double> exact_soln(2 * DIM + 2);
487 
488  // Loop over the integration points
489  for (unsigned ipt = 0; ipt < n_intpt; ipt++)
490  {
491  // Assign values of s
492  for (unsigned i = 0; i < DIM; i++)
493  {
494  s[i] = this->integral_pt()->knot(ipt, i);
495  }
496 
497  // Get the integral weight
498  double w = this->integral_pt()->weight(ipt);
499 
500  // Get jacobian of mapping
501  double J = this->J_eulerian(s);
502 
503  // Premultiply the weights and the Jacobian
504  double W = w * J;
505 
506  // Get x position as Vector
507  this->interpolated_x(s, x);
508 
509  // Get exact solution at this point
510  (*exact_soln_pt)(x, exact_soln);
511 
512  // Displacement error
513  for (unsigned i = 0; i < DIM; i++)
514  {
515  norm[0] += exact_soln[i] * exact_soln[i] * W;
516  // Error due to q_i
517  error[0] += (exact_soln[i] - this->interpolated_u(s, i)) *
518  (exact_soln[i] - this->interpolated_u(s, i)) * W;
519  }
520 
521  // Flux error
522  for (unsigned i = 0; i < DIM; i++)
523  {
524  norm[1] += exact_soln[DIM + i] * exact_soln[DIM + i] * W;
525  // Error due to q_i
526  error[1] += (exact_soln[DIM + i] - this->interpolated_q(s, i)) *
527  (exact_soln[DIM + i] - this->interpolated_q(s, i)) * W;
528  }
529 
530  // Flux divergence error
531  norm[1] += exact_soln[2 * DIM] * exact_soln[2 * DIM] * W;
532  error[1] += (exact_soln[2 * DIM] - interpolated_div_q(s)) *
533  (exact_soln[2 * DIM] - interpolated_div_q(s)) * W;
534 
535  // Pressure error
536  norm[2] += exact_soln[2 * DIM + 1] * exact_soln[2 * DIM + 1] * W;
537  error[2] += (exact_soln[2 * DIM + 1] - this->interpolated_p(s)) *
538  (exact_soln[2 * DIM + 1] - this->interpolated_p(s)) * W;
539 
540  // Output x,y,[z]
541  for (unsigned i = 0; i < DIM; i++)
542  {
543  outfile << x[i] << " ";
544  }
545 
546  // Output u_1_error,u_2_error,...
547  for (unsigned i = 0; i < DIM; i++)
548  {
549  outfile << exact_soln[i] - this->interpolated_u(s, i) << " ";
550  }
551 
552  // Output q_1_error,q_2_error,...
553  for (unsigned i = 0; i < DIM; i++)
554  {
555  outfile << exact_soln[DIM + i] - this->interpolated_q(s, i) << " ";
556  }
557 
558  // Output p_error
559  outfile << exact_soln[2 * DIM + 1] - this->interpolated_p(s) << " ";
560 
561  outfile << std::endl;
562  }
563  }
564 
565  /// Compute the error between the FE solution and the exact solution
566  /// using the H(div) norm for u and L^2 norm for p. Unsteady version
567  template<unsigned DIM>
569  std::ostream& outfile,
571  const double& time,
572  Vector<double>& error,
573  Vector<double>& norm)
574  {
575  for (unsigned i = 0; i < 3; i++)
576  {
577  error[i] = 0.0;
578  norm[i] = 0.0;
579  }
580 
581  // Vector of local coordinates
582  Vector<double> s(DIM);
583 
584  // Vector for coordinates
585  Vector<double> x(DIM);
586 
587  // Set the value of n_intpt
588  unsigned n_intpt = this->integral_pt()->nweight();
589 
590  outfile << "ZONE" << std::endl;
591 
592  // Exact solution Vector (u,v,[w])
593  Vector<double> exact_soln(2 * DIM + 2);
594 
595  // Loop over the integration points
596  for (unsigned ipt = 0; ipt < n_intpt; ipt++)
597  {
598  // Assign values of s
599  for (unsigned i = 0; i < DIM; i++)
600  {
601  s[i] = this->integral_pt()->knot(ipt, i);
602  }
603 
604  // Get the integral weight
605  double w = this->integral_pt()->weight(ipt);
606 
607  // Get jacobian of mapping
608  double J = this->J_eulerian(s);
609 
610  // Premultiply the weights and the Jacobian
611  double W = w * J;
612 
613  // Get x position as Vector
614  this->interpolated_x(s, x);
615 
616  // Get exact solution at this point
617  (*exact_soln_pt)(time, x, exact_soln);
618 
619  // Displacement error
620  for (unsigned i = 0; i < DIM; i++)
621  {
622  norm[0] += exact_soln[i] * exact_soln[i] * W;
623  // Error due to q_i
624  error[0] += (exact_soln[i] - this->interpolated_u(s, i)) *
625  (exact_soln[i] - this->interpolated_u(s, i)) * W;
626  }
627 
628  // Flux error
629  for (unsigned i = 0; i < DIM; i++)
630  {
631  norm[1] += exact_soln[DIM + i] * exact_soln[DIM + i] * W;
632  // Error due to q_i
633  error[1] += (exact_soln[DIM + i] - this->interpolated_q(s, i)) *
634  (exact_soln[DIM + i] - this->interpolated_q(s, i)) * W;
635  }
636 
637  // Flux divergence error
638  norm[1] += exact_soln[2 * DIM] * exact_soln[2 * DIM] * W;
639  error[1] += (exact_soln[2 * DIM] - interpolated_div_q(s)) *
640  (exact_soln[2 * DIM] - interpolated_div_q(s)) * W;
641 
642  // Pressure error
643  norm[2] += exact_soln[2 * DIM + 1] * exact_soln[2 * DIM + 1] * W;
644  error[2] += (exact_soln[2 * DIM + 1] - this->interpolated_p(s)) *
645  (exact_soln[2 * DIM + 1] - this->interpolated_p(s)) * W;
646 
647  // Output x,y,[z]
648  for (unsigned i = 0; i < DIM; i++)
649  {
650  outfile << x[i] << " ";
651  }
652 
653  // Output u_1_error,u_2_error,...
654  for (unsigned i = 0; i < DIM; i++)
655  {
656  outfile << exact_soln[i] - this->interpolated_u(s, i) << " ";
657  }
658 
659  // Output q_1_error,q_2_error,...
660  for (unsigned i = 0; i < DIM; i++)
661  {
662  outfile << exact_soln[DIM + i] - this->interpolated_q(s, i) << " ";
663  }
664 
665  // Output p_error
666  outfile << exact_soln[2 * DIM + 1] - this->interpolated_p(s) << " ";
667 
668  outfile << std::endl;
669  }
670  }
671 
672  /// Fill in residuals and, if flag==true, jacobian
673  template<unsigned DIM>
675  Vector<double>& residuals, DenseMatrix<double>& jacobian, bool flag)
676  {
677  // Get the number of geometric nodes, total number of basis functions,
678  // and number of edges basis functions
679  const unsigned n_node = nnode();
680  const unsigned n_q_basis = nq_basis();
681  const unsigned n_q_basis_edge = nq_basis_edge();
682  const unsigned n_p_basis = np_basis();
683 
684  // Storage for the geometric and computational bases and the test functions
685  Shape psi(n_node), u_basis(n_node), u_test(n_node), q_basis(n_q_basis, DIM),
686  q_test(n_q_basis, DIM), p_basis(n_p_basis), p_test(n_p_basis),
687  div_q_basis_ds(n_q_basis), div_q_test_ds(n_q_basis);
688 
689  DShape dpsidx(n_node, DIM), du_basis_dx(n_node, DIM),
690  du_test_dx(n_node, DIM);
691 
692  // Get the number of integration points
693  unsigned n_intpt = integral_pt()->nweight();
694 
695  // Storage for the local coordinates
696  Vector<double> s(DIM);
697 
698  // Storage for the elasticity source function
699  Vector<double> f_solid(DIM);
700 
701  // Storage for the source function
702  Vector<double> f_fluid(DIM);
703 
704  // Storage for the mass source function
705  double mass_source_local;
706 
707  // Storage for Lambda_sq
708  double lambda_sq = this->lambda_sq();
709 
710  // Get the value of 1/k
711  double k_inv = this->k_inv();
712 
713  // Get the value of alpha
714  double alpha = this->alpha();
715 
716  // Get the value of the porosity
717  double porosity = this->porosity();
718 
719  // Get the density ratio
720  double density_ratio = this->density_ratio();
721 
722  // Precompute the ratio of fluid density to combined density
723  double rho_f_over_rho =
724  density_ratio / (1 + porosity * (density_ratio - 1));
725 
726  // Get continuous time from timestepper of first node
727  double time = node_pt(0)->time_stepper_pt()->time_pt()->time();
728 
729  // Local equation and unknown numbers
730  int local_eqn = 0, local_unknown = 0;
731 
732  // Loop over the integration points
733  for (unsigned ipt = 0; ipt < n_intpt; ipt++)
734  {
735  // Find the local coordinates at the integration point
736  for (unsigned i = 0; i < DIM; i++)
737  {
738  s[i] = integral_pt()->knot(ipt, i);
739  }
740 
741  // Get the weight of the intetgration point
742  double w = integral_pt()->weight(ipt);
743 
744  // Call the basis functions and test functions and get the
745  // (geometric) jacobian of the current element
746  double J = shape_basis_test_local_at_knot(ipt,
747  psi,
748  dpsidx,
749  u_basis,
750  u_test,
751  du_basis_dx,
752  du_test_dx,
753  q_basis,
754  q_test,
755  p_basis,
756  p_test,
757  div_q_basis_ds,
758  div_q_test_ds);
759 
760  // Storage for interpolated values
761  Vector<double> interpolated_x(DIM, 0.0);
762  Vector<double> interpolated_u(DIM, 0.0);
763  DenseMatrix<double> interpolated_du_dx(DIM, DIM, 0.0);
764  double interpolated_div_du_dt_dx = 0.0;
765  Vector<double> interpolated_d2u_dt2(DIM, 0.0);
766  Vector<double> interpolated_q(DIM, 0.0);
767  double interpolated_div_q_ds = 0.0;
768  Vector<double> interpolated_dq_dt(DIM, 0.0);
769  double interpolated_p = 0.0;
770 
771  // loop over the vector basis functions to find interpolated x
772  for (unsigned l = 0; l < n_node; l++)
773  {
774  // Loop over the geometric basis functions
775  for (unsigned i = 0; i < DIM; i++)
776  {
777  interpolated_x[i] += nodal_position(l, i) * psi(l);
778 
779  interpolated_d2u_dt2[i] += this->d2u_dt2(l, i) * u_basis(l);
780 
781  // Get the nodal displacements
782  const double u_value = this->raw_nodal_value(l, u_index(i));
783  interpolated_u[i] += u_value * u_basis(l);
784 
785  // Loop over derivative directions
786  for (unsigned j = 0; j < DIM; j++)
787  {
788  interpolated_du_dx(i, j) += u_value * du_basis_dx(l, j);
789  }
790 
791  // divergence of the time derivative of the solid displacement
792  interpolated_div_du_dt_dx += this->du_dt(l, i) * du_basis_dx(l, i);
793  }
794  }
795 
796  // loop over the nodes and use the geometric basis functions to find the
797  // interpolated flux and its time derivative
798  for (unsigned i = 0; i < DIM; i++)
799  {
800  // Loop over the edge basis vectors
801  for (unsigned l = 0; l < n_q_basis_edge; l++)
802  {
803  interpolated_q[i] += q_edge(l) * q_basis(l, i);
804  interpolated_dq_dt[i] += dq_edge_dt(l) * q_basis(l, i);
805  }
806  // Loop over the internal basis vectors
807  for (unsigned l = n_q_basis_edge; l < n_q_basis; l++)
808  {
809  interpolated_q[i] += q_internal(l - n_q_basis_edge) * q_basis(l, i);
810  interpolated_dq_dt[i] +=
811  dq_internal_dt(l - n_q_basis_edge) * q_basis(l, i);
812  }
813  }
814 
815  // loop over the pressure basis and find the interpolated pressure
816  for (unsigned l = 0; l < n_p_basis; l++)
817  {
818  interpolated_p += p_value(l) * p_basis(l);
819  }
820 
821  // loop over the u edge divergence basis and the u internal divergence
822  // basis to find interpolated div u
823  for (unsigned l = 0; l < n_q_basis_edge; l++)
824  {
825  interpolated_div_q_ds += q_edge(l) * div_q_basis_ds(l);
826  }
827  for (unsigned l = n_q_basis_edge; l < n_q_basis; l++)
828  {
829  interpolated_div_q_ds +=
830  q_internal(l - n_q_basis_edge) * div_q_basis_ds(l);
831  }
832 
833  // Get the solid forcing
834  this->force_solid(time, interpolated_x, f_solid);
835 
836  // Get the fluid forcing
837  this->force_fluid(time, interpolated_x, f_fluid);
838 
839  // Get the mass source function
840  this->mass_source(time, interpolated_x, mass_source_local);
841 
842  // Premultiply the weights and the Jacobian
843  double W = w * J;
844 
845  // Linear elasticity:
846 
847  for (unsigned l = 0; l < n_node; l++)
848  {
849  for (unsigned a = 0; a < DIM; a++)
850  {
851  local_eqn = this->nodal_local_eqn(l, u_index(a));
852 
853  if (local_eqn >= 0)
854  {
855  residuals[local_eqn] +=
856  (lambda_sq * (interpolated_d2u_dt2[a] +
857  rho_f_over_rho * interpolated_dq_dt[a]) -
858  f_solid[a]) *
859  u_test(l) * W;
860 
861  // Stress term
862  for (unsigned b = 0; b < DIM; b++)
863  {
864  if (a == b)
865  {
866  residuals[local_eqn] -=
867  alpha * interpolated_p * du_test_dx(l, b) * W;
868  }
869 
870  for (unsigned c = 0; c < DIM; c++)
871  {
872  for (unsigned d = 0; d < DIM; d++)
873  {
874  // Add the stress terms to the residuals
875  residuals[local_eqn] += this->E(a, b, c, d) *
876  interpolated_du_dx(c, d) *
877  du_test_dx(l, b) * W;
878  }
879  }
880  }
881  // Jacobian entries
882  if (flag)
883  {
884  // d(u_eqn_l,a)/d(U_l2,c)
885  for (unsigned l2 = 0; l2 < n_node; l2++)
886  {
887  for (unsigned c = 0; c < DIM; c++)
888  {
889  local_unknown = this->nodal_local_eqn(l2, u_index(c));
890  if (local_unknown >= 0)
891  {
892  if (a == c)
893  {
894  jacobian(local_eqn, local_unknown) +=
895  lambda_sq *
896  this->node_pt(l2)->time_stepper_pt()->weight(2, 0) *
897  u_basis(l2) * u_test(l) * W;
898  }
899 
900  for (unsigned b = 0; b < DIM; b++)
901  {
902  for (unsigned d = 0; d < DIM; d++)
903  {
904  jacobian(local_eqn, local_unknown) +=
905  this->E(a, b, c, d) * du_basis_dx(l2, d) *
906  du_test_dx(l, b) * W;
907  }
908  }
909  }
910  }
911  }
912 
913  // d(u_eqn_l,a)/d(Q_l2)
914  for (unsigned l2 = 0; l2 < n_q_basis; l2++)
915  {
916  TimeStepper* timestepper_pt = 0;
917 
918  if (l2 < n_q_basis_edge)
919  {
920  local_unknown = q_edge_local_eqn(l2);
921  timestepper_pt =
922  this->node_pt(q_edge_node_number(l2))->time_stepper_pt();
923  }
924  else // n_q_basis_edge <= l < n_basis
925  {
926  local_unknown = q_internal_local_eqn(l2 - n_q_basis_edge);
927  timestepper_pt = this->internal_data_pt(q_internal_index())
928  ->time_stepper_pt();
929  }
930 
931  if (local_unknown >= 0)
932  {
933  jacobian(local_eqn, local_unknown) +=
934  lambda_sq * rho_f_over_rho * timestepper_pt->weight(1, 0) *
935  q_basis(l2, a) * u_test(l) * W;
936  }
937  }
938 
939  // d(u_eqn_l,a)/d(P_l2)
940  for (unsigned l2 = 0; l2 < n_p_basis; l2++)
941  {
942  local_unknown = p_local_eqn(l2);
943  if (local_unknown >= 0)
944  {
945  jacobian(local_eqn, local_unknown) -=
946  alpha * p_basis(l2) * du_test_dx(l, a) * W;
947  }
948  }
949  } // End of Jacobian entries
950  } // End of if not boundary condition
951  } // End of loop over dimensions
952  } // End of loop over u test functions
953 
954 
955  // Darcy:
956  // Loop over the test functions
957  for (unsigned l = 0; l < n_q_basis; l++)
958  {
959  if (l < n_q_basis_edge)
960  {
961  local_eqn = q_edge_local_eqn(l);
962  }
963  else // n_q_basis_edge <= l < n_basis
964  {
965  local_eqn = q_internal_local_eqn(l - n_q_basis_edge);
966  }
967 
968  // If it's not a boundary condition
969  if (local_eqn >= 0)
970  {
971  for (unsigned i = 0; i < DIM; i++)
972  {
973  residuals[local_eqn] +=
974  (k_inv * interpolated_q[i] - rho_f_over_rho * f_fluid[i] +
975  rho_f_over_rho * lambda_sq *
976  (interpolated_d2u_dt2[i] +
977  (1.0 / porosity) * interpolated_dq_dt[i])) *
978  q_test(l, i) * w * J;
979  }
980 
981  // deliberately no jacobian factor in this integral
982  residuals[local_eqn] -= (interpolated_p * div_q_test_ds(l)) * w;
983 
984  // Jacobian entries
985  if (flag)
986  {
987  // d(q_eqn_l)/d(U_l2,c)
988  for (unsigned l2 = 0; l2 < n_node; l2++)
989  {
990  for (unsigned c = 0; c < DIM; c++)
991  {
992  local_unknown = this->nodal_local_eqn(l2, u_index(c));
993  if (local_unknown >= 0)
994  {
995  jacobian(local_eqn, local_unknown) +=
996  rho_f_over_rho * lambda_sq *
997  this->node_pt(l2)->time_stepper_pt()->weight(2, 0) *
998  u_basis(l2) * q_test(l, c) * W;
999  }
1000  }
1001  }
1002 
1003  // d(q_eqn_l)/d(Q_l2)
1004  for (unsigned l2 = 0; l2 < n_q_basis; l2++)
1005  {
1006  TimeStepper* timestepper_pt = 0;
1007 
1008  if (l2 < n_q_basis_edge)
1009  {
1010  local_unknown = q_edge_local_eqn(l2);
1011  timestepper_pt =
1012  this->node_pt(q_edge_node_number(l2))->time_stepper_pt();
1013  }
1014  else // n_q_basis_edge <= l < n_basis
1015  {
1016  local_unknown = q_internal_local_eqn(l2 - n_q_basis_edge);
1017  timestepper_pt =
1018  this->internal_data_pt(q_internal_index())->time_stepper_pt();
1019  }
1020 
1021  if (local_unknown >= 0)
1022  {
1023  for (unsigned c = 0; c < DIM; c++)
1024  {
1025  jacobian(local_eqn, local_unknown) +=
1026  q_basis(l2, c) * q_test(l, c) *
1027  (k_inv + rho_f_over_rho * lambda_sq *
1028  timestepper_pt->weight(1, 0) / porosity) *
1029  W;
1030  }
1031  }
1032  }
1033 
1034  // d(q_eqn_l)/d(P_l2)
1035  for (unsigned l2 = 0; l2 < n_p_basis; l2++)
1036  {
1037  local_unknown = p_local_eqn(l2);
1038 
1039  // deliberately no jacobian factor in this term
1040  jacobian(local_eqn, local_unknown) -=
1041  p_basis(l2) * div_q_test_ds(l) * w;
1042  }
1043  } // End of Jacobian entries
1044  } // End of if not boundary condition
1045  } // End of loop over q test functions
1046 
1047  // loop over pressure test functions
1048  for (unsigned l = 0; l < n_p_basis; l++)
1049  {
1050  // get the local equation number
1051  local_eqn = p_local_eqn(l);
1052 
1053  // If it's not a boundary condition
1054  if (local_eqn >= 0)
1055  {
1056  // solid divergence term
1057  residuals[local_eqn] +=
1058  alpha * interpolated_div_du_dt_dx * p_test(l) * w * J;
1059  // fluid divergence term - deliberately no jacobian factor in this
1060  // term
1061  residuals[local_eqn] += interpolated_div_q_ds * p_test(l) * w;
1062  // mass source term
1063  residuals[local_eqn] -= mass_source_local * p_test(l) * w * J;
1064 
1065  // Jacobian entries
1066  if (flag)
1067  {
1068  // d(p_eqn_l)/d(U_l2,c)
1069  for (unsigned l2 = 0; l2 < n_node; l2++)
1070  {
1071  for (unsigned c = 0; c < DIM; c++)
1072  {
1073  local_unknown = this->nodal_local_eqn(l2, u_index(c));
1074 
1075  if (local_unknown >= 0)
1076  {
1077  jacobian(local_eqn, local_unknown) +=
1078  alpha * this->node_pt(l2)->time_stepper_pt()->weight(1, 0) *
1079  du_basis_dx(l2, c) * p_test(l) * W;
1080  }
1081  }
1082  }
1083 
1084  // d(p_eqn_l)/d(Q_l2)
1085  for (unsigned l2 = 0; l2 < n_q_basis; l2++)
1086  {
1087  if (l2 < n_q_basis_edge)
1088  {
1089  local_unknown = q_edge_local_eqn(l2);
1090  }
1091  else // n_q_basis_edge <= l < n_basis
1092  {
1093  local_unknown = q_internal_local_eqn(l2 - n_q_basis_edge);
1094  }
1095 
1096  if (local_unknown >= 0)
1097  {
1098  jacobian(local_eqn, local_unknown) +=
1099  p_test(l) * div_q_basis_ds(l2) * w;
1100  }
1101  }
1102  } // End of Jacobian entries
1103  } // End of if not boundary condition
1104  } // End of loop over p test functions
1105  } // End of loop over integration points
1106  }
1107 
1108  // Force building of templates
1109  template class PoroelasticityEquations<2>;
1110 
1111 } // namespace oomph
static char t char * s
Definition: cfortran.h:568
cstr elem_len * i
Definition: cfortran.h:603
A Class for the derivatives of shape functions The class design is essentially the same as Shape,...
Definition: shape.h:278
unsigned long nrow() const
Return the number of rows of the matrix.
Definition: matrices.h:485
unsigned long ncol() const
Return the number of columns of the matrix.
Definition: matrices.h:491
void(* SteadyExactSolutionFctPt)(const Vector< double > &, Vector< double > &)
Function pointer for function that computes vector-valued steady "exact solution" as .
Definition: elements.h:1763
void(* UnsteadyExactSolutionFctPt)(const double &, const Vector< double > &, Vector< double > &)
Function pointer for function that computes Vector-valued time-dependent function as .
Definition: elements.h:1769
An OomphLibError object which should be thrown when an run-time error is encountered....
void output_fct(std::ostream &outfile, const unsigned &nplot, FiniteElement::SteadyExactSolutionFctPt exact_soln_pt)
Output FE representation of exact soln: x,y,u1,u2,div_q,p at Nplot^DIM plot points.
static double Default_porosity_value
Static default value for the porosity.
double transform_basis(const Vector< double > &s, const Shape &q_basis_local, Shape &psi, DShape &dpsi, Shape &q_basis) const
Performs a div-conserving transformation of the vector basis functions from the reference element to ...
static double Default_density_ratio_value
Static default value for the density ratio.
void get_stress(const Vector< double > &s, DenseMatrix< double > &sigma) const
Return the Cauchy stress tensor, as calculated from the elasticity tensor at specified local coordina...
virtual void fill_in_generic_residual_contribution(Vector< double > &residuals, DenseMatrix< double > &jacobian, bool flag)
Fill in residuals and, if flag==true, jacobian.
static double Default_lambda_sq_value
Static default value for timescale ratio (1.0 – for natural scaling)
static double Default_alpha_value
Static default value for alpha.
void compute_error(std::ostream &outfile, FiniteElement::SteadyExactSolutionFctPt exact_soln_pt, Vector< double > &error, Vector< double > &norm)
Compute the error between the FE solution and the exact solution using the H(div) norm for q and L^2 ...
static double Default_k_inv_value
Static default value for 1/k.
void output(std::ostream &outfile)
Output with default number of plot points.
void get_strain(const Vector< double > &s, DenseMatrix< double > &strain) const
Return the strain tensor.
A Class for shape functions. In simple cases, the shape functions have only one index that can be tho...
Definition: shape.h:76
void shape(const double &s, double *Psi)
Definition for 1D Lagrange shape functions. The value of all the shape functions at the local coordin...
Definition: shape.h:564
//////////////////////////////////////////////////////////////////// ////////////////////////////////...