Feature/contact/normal vector

[thiagordonho]

Describe the PR
Implementation to compute the normal unit vector at every node with more than one materials. In order to do this, a function in Particle class to map the domain gradients (gradient of volumes) is implemented.

  for (unsigned i = 0; i < nodes_.size(); ++i) {
    const auto& gradient = volume_ * dn_dx_.row(i);
    nodes_[i]->update_property(true, "domain_gradients", gradient, material_id_,
                               Tdim);
  }

Then, the normal unit vector is calculated at every node by means of a function implemented in the Node class computing the unit vector as the domain gradient divided by its norm.

    const auto& normal_unit_vector =
        (1 / domain_gradient.norm()) * domain_gradient;

Related Issues/PRs
This PR is related to the RFC #636 .

[codecov]

Codecov Report

Merging #655 into develop will decrease coverage by 0.00%.
The diff coverage is 96.55%.

@@             Coverage Diff             @@
##           develop     #655      +/-   ##
===========================================
- Coverage    96.55%   96.55%   -0.00%     
===========================================
  Files          122      122              
  Lines        25337    25395      +58     
===========================================
+ Hits         24464    24520      +56     
- Misses         873      875       +2     

Impacted Files	Coverage Δ
include/node.h	100.00% <ø> (ø)
include/node_base.h	100.00% <ø> (ø)
include/particles/particle.h	100.00% <ø> (ø)
include/particles/particle_base.h	100.00% <ø> (ø)
include/solvers/mpm_explicit.tcc	61.94% <0.00%> (-0.81%)	⬇️
include/mesh.tcc	84.13% <100.00%> (+0.07%)	⬆️
include/node.tcc	95.81% <100.00%> (+0.14%)	⬆️
include/particles/particle.tcc	94.36% <100.00%> (+0.09%)	⬆️
tests/interface_test.cc	100.00% <100.00%> (ø)

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 9ac85dd...57d5d46. Read the comment docs.

[kks32]

Please complete the requested changes.

[ezrayst]

Thanks @thiagordonho, not sure whether we should use .normalized() or .normalize() - but I think the different is that it returns a copy for the former and modify the current variable for the latter. So I think you are right in using the former!

[kks32]

Suggested change

	const auto& gradient = volume_ * dn_dx_.row(i);
	Eigen::Matrix<double, Tdim, 1> gradient;
	for (unsigned j = 0; j < Tdim; ++j)
	gradient[j] += dn_dx_[i, j] * volume;