Add heat flux postprocessor #953

mivaia · 2023-12-06T23:15:21Z

Description of the problem

In the context of conjugated heat transfer simulations, there was no way to get the local heat fluxes of the domain.

Description of the solution

A new DataPostprocessorVector child class was made, namely, HeatFluxPostprocessor. This class postprocesses the local heat fluxes of the fluids and solids of the domain. The values for each fluid and solid is outputed individually.

How Has This Been Tested?

CI

Comments

An example using this HeatFluxPostprocessor values will be made soon.

oguevremont

Very minor comments, good work! After my suggestions it is approved

source/solvers/heat_transfer.cc

include/solvers/postprocessors.h

blaisb

Minor minor changes, but please add a bit more comments

blaisb · 2023-12-07T14:59:32Z

include/solvers/heat_transfer.h

+                                          thermal_conductivity_models;
+  std::vector<HeatFluxPostprocessor<dim>> heat_flux_postprocessors;


Do we need to recreate those or we could just get them from the physical property manager?

Since its a private member, we have to make a copy... However, I can simply initialize it with values in the constructor instead of in attach_solution_to_output. I'll do this change!

blaisb · 2023-12-07T15:02:10Z

include/solvers/postprocessors.h

+  HeatFluxPostprocessor(
+    std::shared_ptr<ThermalConductivityModel> p_thermal_conductivity_model,
+    const std::string                         p_material_string = "f",
+    const unsigned int                        p_material_id     = 0,
+    const unsigned int                        p_id              = 0)
+    : DataPostprocessorVector<dim>("heat_flux_" + p_material_string +
+                                     Utilities::to_string(p_id, 2),
+                                   update_values | update_gradients)
+    , thermal_conductivity_model(p_thermal_conductivity_model)
+    , material_id(p_material_id)
+  {}


Can you explain a bit more why you need to pass the p_id and the p_material id. It's not clear at this stage

p_id is the number of either the fluid or the solid. So when, for example, we have a 1 fluid and 1 solid, they both have a p_id of 0 (fluid_id = 0 and solid_id = 0).

The material_id is the cumulative id of fluids and solids (a more global id that's give). From what I've seen fluids are counted first followed by solids. Therefore, in our examples, the fluid would have a material_id of 0 and the solid of 1.

The material_id is used to which cells belong in which material. And p_id is used for the name of the outputted vector. I could've done it with the a if and the p_material_string, but I wanted to avoid the if.

I'll try to make it more clear. :)
Also, I just realized that the arguments are not entered by reference, I'll change that!

blaisb · 2023-12-07T15:03:45Z

include/solvers/postprocessors.h

+    Vector<double>              null_vector(dim);
+    std::vector<Vector<double>> null_computed_quantities(n_quadrature_points,


What are those null? Can you comment out why they are required and what's their point?

It's for when the cell doesn't lie in the right material, to avoid doing a for loop to insert zeros. I'll move it down, where the comment is already there and to avoid making this null-vector for nothing.
I'll also add more comments in general. :)

hepap

Nice addition! I agree with Bruno that a bit more comments would help :) Also, may be you can add a test to verify your implementation? I don't know if it is relevant at this point, but a pseudo 1D case with different layers of different materials could be interesting (there are analytical solutions for that in the BSL I think!)

hepap · 2023-12-07T19:20:04Z

include/solvers/postprocessors.h

+    , material_id(p_material_id)
+  {}
+
+  virtual void


I noticed that every overriden methods in postprocessors.h are virtual. Are we expecting to further override them? It is more for my own knowledge, because I'm not sure I get why we do this :)

AmishgaAlphonius · 2023-12-07T20:24:35Z

Nice addition! I agree with Bruno that a bit more comments would help :) Also, may be you can add a test to verify your implementation? I don't know if it is relevant at this point, but a pseudo 1D case with different layers of different materials could be interesting (there are analytical solutions for that in the BSL I think!)

Since it the values can only be read from an output file, a simple test cannot be made without postprocessing the data in the exported files. However, an example is on the way!

Co-authored-by: Amishga Alphonius <[email protected]>

Co-authored-by: Olivier Guévremont <[email protected]>

Co-authored-by: Bruno Blais <[email protected]>

Description of the problem In the context of conjugated heat transfer simulations, there was no way to get the local heat fluxes of the domain. Description of the solution A new DataPostprocessorVector child class was made, namely, HeatFluxPostprocessor. This class postprocesses the local heat fluxes of the fluids and solids of the domain. The values for each fluid and solid is outputed individually. How Has This Been Tested? CI Comments An example using this HeatFluxPostprocessor values will be made soon. Co-authored-by: Amishga Alphonius <[email protected]> Co-authored-by: Olivier Guévremont <[email protected]> Co-authored-by: Amishga Alphonius <[email protected]> Co-authored-by: Bruno Blais <[email protected]> Former-commit-id: d9b9d74

mivaia assigned mivaia and AmishgaAlphonius Dec 6, 2023

mivaia added Enhancement New feature or request Lethe Hackathon - Fall 2023 Post-processing Modifications to the post-processing module and removed Post-processing Modifications to the post-processing module labels Dec 6, 2023

mivaia requested a review from blaisb December 6, 2023 23:16

AmishgaAlphonius requested a review from oguevremont December 7, 2023 00:45

oguevremont approved these changes Dec 7, 2023

View reviewed changes

source/solvers/heat_transfer.cc Outdated Show resolved Hide resolved

include/solvers/postprocessors.h Outdated Show resolved Hide resolved

blaisb requested changes Dec 7, 2023

View reviewed changes

hepap reviewed Dec 7, 2023

View reviewed changes

AmishgaAlphonius force-pushed the add_heat_flux_postprocessor branch from 6db8a49 to 920e2a4 Compare December 7, 2023 20:02

blaisb force-pushed the add_heat_flux_postprocessor branch from 920e2a4 to 58143f4 Compare December 7, 2023 21:05

mivaia and others added 4 commits December 7, 2023 22:35

Add heat flux postprocessor

8d398e1

Co-authored-by: Amishga Alphonius <[email protected]>

Apply suggestions from code review

3359301

Co-authored-by: Olivier Guévremont <[email protected]>

Indent

851199b

Apply code review comments

6abe21d

Co-authored-by: Bruno Blais <[email protected]>

blaisb force-pushed the add_heat_flux_postprocessor branch from 58143f4 to 6abe21d Compare December 8, 2023 03:35

AmishgaAlphonius approved these changes Dec 8, 2023

View reviewed changes

blaisb self-requested a review December 8, 2023 19:16

blaisb approved these changes Dec 8, 2023

View reviewed changes

blaisb merged commit d9b9d74 into master Dec 8, 2023
15 checks passed

blaisb deleted the add_heat_flux_postprocessor branch December 8, 2023 19:16

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add heat flux postprocessor #953

Add heat flux postprocessor #953

mivaia commented Dec 6, 2023 •

edited by AmishgaAlphonius

Loading

oguevremont left a comment

blaisb left a comment

blaisb Dec 7, 2023

AmishgaAlphonius Dec 7, 2023

blaisb Dec 7, 2023

AmishgaAlphonius Dec 7, 2023

blaisb Dec 7, 2023

AmishgaAlphonius Dec 7, 2023

hepap left a comment

hepap Dec 7, 2023

AmishgaAlphonius commented Dec 7, 2023

		thermal_conductivity_models;
		std::vector<HeatFluxPostprocessor<dim>> heat_flux_postprocessors;

		Vector<double> null_vector(dim);
		std::vector<Vector<double>> null_computed_quantities(n_quadrature_points,

Add heat flux postprocessor #953

Add heat flux postprocessor #953

Conversation

mivaia commented Dec 6, 2023 • edited by AmishgaAlphonius Loading

Description of the problem

Description of the solution

How Has This Been Tested?

Comments

oguevremont left a comment

Choose a reason for hiding this comment

blaisb left a comment

Choose a reason for hiding this comment

blaisb Dec 7, 2023

Choose a reason for hiding this comment

AmishgaAlphonius Dec 7, 2023

Choose a reason for hiding this comment

blaisb Dec 7, 2023

Choose a reason for hiding this comment

AmishgaAlphonius Dec 7, 2023

Choose a reason for hiding this comment

blaisb Dec 7, 2023

Choose a reason for hiding this comment

AmishgaAlphonius Dec 7, 2023

Choose a reason for hiding this comment

hepap left a comment

Choose a reason for hiding this comment

hepap Dec 7, 2023

Choose a reason for hiding this comment

AmishgaAlphonius commented Dec 7, 2023

mivaia commented Dec 6, 2023 •

edited by AmishgaAlphonius

Loading