Incorporate parabolic terms into `main`

[jlchan]

PR for discussion and progress on parabolic terms in Trixi.

[codecov[bot]]

Codecov Report

Merging #1149 (0ac6c59) into main (8dd52e3) will increase coverage by 0.03%. The diff coverage is 97.83%.

:exclamation: Current head 0ac6c59 differs from pull request most recent head 7fb68c8. Consider uploading reports for the commit 7fb68c8 to get more accurate results

@@            Coverage Diff             @@
##             main    #1149      +/-   ##
==========================================
+ Coverage   96.78%   96.81%   +0.03%     
==========================================
  Files         310      326      +16     
  Lines       24300    25066     +766     
==========================================
+ Hits        23517    24266     +749     
- Misses        783      800      +17     

Flag	Coverage Δ
unittests	96.81% <97.83%> (+0.03%)	:arrow_up:

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
src/Trixi.jl	70.00% <ø> (ø)
src/equations/compressible_euler_1d.jl	99.62% <ø> (ø)
src/equations/compressible_euler_2d.jl	99.01% <ø> (ø)
src/equations/compressible_euler_3d.jl	97.63% <ø> (ø)
.../equations/compressible_euler_multicomponent_1d.jl	100.00% <ø> (ø)
.../equations/compressible_euler_multicomponent_2d.jl	100.00% <ø> (ø)
src/solvers/dgmulti/sbp.jl	92.37% <ø> (ø)
src/solvers/dgsem_p4est/dg_2d.jl	100.00% <ø> (ø)
src/solvers/dgsem_p4est/dg_3d.jl	100.00% <ø> (ø)
src/solvers/dgsem_tree/dg.jl	100.00% <ø> (ø)
... and 24 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[ranocha]

The test failure

ERROR: LoadError: UndefVarError: ViscousFormulationBassiRebay1 not defined
Stacktrace:
  [1] top-level scope
    @ ~/work/Trixi.jl/Trixi.jl/src/solvers/dgmulti/dg_parabolic.jl:208

looks real?

[sloede]

The test failure

ERROR: LoadError: UndefVarError: ViscousFormulationBassiRebay1 not defined
Stacktrace:
  [1] top-level scope
    @ ~/work/Trixi.jl/Trixi.jl/src/solvers/dgmulti/dg_parabolic.jl:208

looks real?

Maybe a matter of the include order of files?

[jlchan]

I'll try to debug it now...

[jlchan]

The test failure

ERROR: LoadError: UndefVarError: ViscousFormulationBassiRebay1 not defined
Stacktrace:
  [1] top-level scope
    @ ~/work/Trixi.jl/Trixi.jl/src/solvers/dgmulti/dg_parabolic.jl:208

looks real?

Maybe a matter of the include order of files?

I think that's correct. When I moved a specialization of calc_viscous_penalty! into dg_parabolic.jl, it introduced the parabolic solver type before it was defined. Should be fixed by https://github.com/trixi-framework/Trixi.jl/pull/1149/commits/ffffa34d2a6fc836cff4c111b81e50c46fe6b8b1

[ranocha]

Looking at #1147, I think it's fine to merge this into main once we have

• LDG with TreeMesh2D
• a bit more documentation (e.g., BCs)
• experimental notes everywhere

Then, we should play around with it and implement the other parts, e.g., callbacks for coupled equations.

[sloede]

Looking at #1147, I think it's fine to merge this into main once we have

* LDG with `TreeMesh2D`

* a bit more documentation (e.g., BCs)

* experimental notes everywhere

Then, we should play around with it and implement the other parts, e.g., callbacks for coupled equations.

I agree! We should leave all "extra" stuff like Callbacks for later - as soon as we get this into main and into people's hands, the better for us, I think.

[ranocha]

I will have a look at the PID computation. What are the other actionable items from the GAU ("Gregor as user") test and who is working on them?

[jlchan]

I believe there were two others

• Convergence to Euler as Reynolds's number goes to infinity
• use of viscosity instead of Re

I can look at the former. The latter would require generalizing the formulation from non dimensional to dimensional, so maybe we can discuss it today.