Mesh export, high order

[Joss G.]

Hi all, 

I am using high order methods ad I am trying to export the mesh points. I am using  grid_out.write_msh(triangulation, out).

Is it any other way to get the  mesh points including the ones corresponding to the high orders?

Thank you

[Marc Fehling]

Hello,

yes, you can use high-order output using the `DataOut` class after setting the corresponding `DataOutFlags` flag write_higher_order_cells.

The tutorials step-11, step-44, step-65, step-67, and step-76 demonstrate that feature.

Best,
Marc

[Joss G.]

Thank you for your answer.

Following the examples you mentioned, I got a vtu file where I can extract the points. As I am using distributed tirangulation, I am getting the (x,y) values of the points of each cell, meaning that the points are not unique, there are repeated.

DataOut<dim> data_out;

DataOutBase::VtkFlags flags;

flags.write_higher_order_cells = true;

data_out.set_flags(flags);

data_out.attach_dof_handler(dof_handler);

data_out.add_data_vector(solution, "sol");

Vector<float> subdomain(triangulation.n_active_cells());

for (unsigned int i = 0; i < subdomain.size(); ++i)

subdomain(i) = triangulation.locally_owned_subdomain();

data_out.add_data_vector(subdomain, "subdomain");

data_out.build_patches();

data_out.write_vtu_with_pvtu_record("./", "solution", 1, mpi_communicator, 2, 8);

I would need to get the unique points corresponding to the domain in the same order as the operators are assembled. Similar to what I got with grid_out_write.msh but including the higher order points. COuld you please suggest me how can I approach that?

Thank you again.

[Wells, David]

Hi,

Unlike a lot of FE software deal.II does not actually store, e.g., 27 points for a HEX27 element. They are always implicitly defined by the FiniteElement, Mapping, and Triangulation put together so you have to compute their values explicitly.

Hence, if you want globally unique nodes, the correct things to compute are the support points for all locally owned degrees of freedom. The function

DoFTools::map_dofs_to_support_points()

would be what you want but that function doesn't work in parallel. Fortunately this function is pretty short and you can write your own version that works on the locally owned dofs pretty easily by starting at

https://github.com/dealii/dealii/blob/master/source/dofs/dof_tools.cc#L2102

and doing something like

1. Set up a quadrature with fe.get_unit_support_points()
2. set up FEValues with your FE, mapping, and that quadrature. Do you have a finite element field that describes a position or displacement? If so take a look at MappingFEField.
3. do fe_values.get_quadrature_points() on each cell, but only save the results if the dof corresponding to the quadrature point (i.e., its support point) is locally owned.

Do you see what to do?

Best,

David Wells

---

From: dea...@googlegroups.com <dea...@googlegroups.com> on behalf of Joss G. <jossg...@gmail.com>
Sent: Wednesday, November 17, 2021 7:24 AM
To: deal.II User Group <dea...@googlegroups.com>
Subject: [deal.II] Re: Mesh export, high order

 

--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en
---
You received this message because you are subscribed to the Google Groups "deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dealii+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/7095bcb6-fccd-4a28-a2ba-5b6803f4f80cn%40googlegroups.com.

[Wolfgang Bangerth]

On 11/17/21 5:24 AM, Joss G. wrote:
> As I am using distributed tirangulation, I am getting the (x,y) values
> of the points of each cell, meaning that the points are not unique,
> there are repeated.

See also here:
https://github.com/dealii/dealii/wiki/Frequently-Asked-Questions#the-graphical-output-files-dont-make-sense-to-me----they-seem-to-have-too-many-degrees-of-freedom

Best
W.

--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/