search

dealii / code-gallery

A collection of codes based on deal.II contributed by deal.II users
Other
44 stars 54 forks source link

Nonlinear heat transfer problem #178

Closed narasimhan-2604 closed 4 months ago

narasimhan-2604 commented 4 months ago

A simple nonlinear heat transfer problem is solved with automatic differentiation. The nonlinear equations are solved using TRILINOS NOX .

narasimhan-2604 commented 4 months ago

Nonlinear heat transfer

This submission is used to demonstrate the use of TRILINOS'x Automatic differentiation and NOX to solve a simple nonlinear heat transfer problem.

narasimhan-2604 commented 4 months ago

This submission is used to demonstrate the use of TRILINOS'x Automatic differentiation and NOX to solve a simple nonlinear heat transfer problem.

blaisb commented 4 months ago

Thank you very much for your contribution. Would it be possible to remove the VTU as well as the LaTeX artifacts from the PR ?

narasimhan-2604 commented 4 months ago

As suggested, I have removed the vtu files and also other files from the Documentation directory.

narasimhan-2604 commented 4 months ago

The following changes were made (1) Left the latex source files in the Documentation folder to possibly generate doxygen documentation (2) Modified CMakeLists.txt to follow norms for variable specs and also removed some unnecessary lines (3) The Latex Documentation now has some output figures as well. (4) Changed the dates to be of the type XX Jun 2024

bangerth commented 4 months ago

Awesome. I've only got two more small issues:

After that, I can take it from there and merge the latex sources into the regular documentation.

Many thanks again for submitting this -- I thought it was fun to talk about the program in Chennai, and it's nice to see it in actual code now!

bangerth commented 4 months ago

Please add a comment here once you've pushed the new version!

narasimhan-2604 commented 4 months ago

Yes. I have changed the version to 9.5.0 in the CMakeLists.txt file. I hope that it is what was needed. Also, I hope I have done the git squash correctly, I have't done that before.

It was indeed very fruitful discussion in Chennai and I really enjoyed every bit of it. All of the steps step-71, step-72 and step-77 were very useful. Umair (My PhD student) also will have a contribution to code_gallery soon on solving a simple phase field model, reproducing the contents of a paper (typically used by all phase field first timers).

Please let me know if anything else is needed.

Awesome. I've only got two more small issues:

  • The tester runs deal.II 9.5.0, but you require 9.5.1 in CMakeLists.txt and so the automatic check fails. I don't think the difference in version is important. Could you set the required version to 9.5?
  • Would you please merge the commits (either via git squash or git fixup) into one commit, and then force-push?

After that, I can take it from there and merge the latex sources into the regular documentation.

Many thanks again for submitting this -- I thought it was fun to talk about the program in Chennai, and it's nice to see it in actual code now!

marcfehling commented 4 months ago

I think you squashed 1 commit too far, but that's not problem.

bangerth commented 4 months ago

This patch has been superseded by #179 which contains all of the important parts and also correctly attributes the contribution to @narasimhan-2604 .

Thank you, @narasimhan-2604 -- I will take it from here to incorporate the documentation.