Clean up code for single-group radiation

[chongchonghe]

Description

This PR updates the code structure for single-group radiation so that functions like ComputePlanckOpacity return a double. This change enforces the use of the correct opacity models (i.e., piecewise_constant_opacity or PPL_opacity_fixed_slope_spectrum, or PPL_opacity_full_spectrum) when dealing with multiple radiation groups. The previous code provides an option to naively extend the single-group formulation in the IMEX paper to the multigroup case, i.e. replacing $\kappa_P$ and $\kappa_F$ with $\kappa_g$. However, that method does not correctly describe multigroup radiation transport and should be removed, because it misses the important delta term in the flux equation. See Equation 16 of the multigroup paper and the discussion thereafter.

In summary, here's how you should define gas opacity in a problem:

• Single-group opacity definition:
  • Define ComputePlanckOpacity and optionally ComputeFluxMeanOpacity and ComputeEnergyMeanOpacity. By default, the latter two are set equal to the first.
• Multigroup opacity definition:
  • Choose an opacity model by setting opacity_model in RadSystem_Traits to one of the following: piecewise_constant_opacity, PPL_opacity_fixed_slope_spectrum, or PPL_opacity_full_spectrum
  • Define DefineOpacityExponentsAndLowerValues. See the example in RadhydroShockMultigroup.

Bug fix:

• fixed bug related to multigroup opacity in ComputeCellOpticalDepth. This function is only used when you turn on wave speed correction by hand and is not turned on by default. We can't have a regression test for this because it is controlled by use_wavespeed_correction inside radiation_system.hpp and we don't have a way to turn it on for a specific problem. I tested it on my computer and it worked as expected.

Checklist

Before this pull request can be reviewed, all of these tasks should be completed. Denote completed tasks with an x inside the square brackets [ ] in the Markdown source below:

• I have added a description (see above).
• I have added a link to any related issues see (see above).
• I have read the Contributing Guide.
• I have added tests for any new physics that this PR adds to the code.
• I have tested this PR on my local computer and all tests pass.
• I have manually triggered the GPU tests with the magic comment /azp run.
• I have requested a reviewer for this PR.

[chongchonghe]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).

[BenWibking]

We could createa new version of an existing test that uses the wavespeed correction in the Riemann solver to test this. I think that would help avoid this problem.

[chongchonghe]

We could createa new version of an existing test that uses the wavespeed correction in the Riemann solver to test this. I think that would help avoid this problem.

The issue is, you can't turn it on for any specific problem. It's controlled by use_wavespeed_correction which is not in any traits. You have to change radiation_system.hpp to turn it on. I can, of course, add another parameter in the Radiation Trait, but I think it adds another fuss to the user and it's not worth it since we will almost never use it.

[BenWibking]

We could createa new version of an existing test that uses the wavespeed correction in the Riemann solver to test this. I think that would help avoid this problem.

The issue is, you can't turn it on for any specific problem. It's controlled by use_wavespeed_correction which is not in any traits. You have to change radiation_system.hpp to turn it on. I can, of course, add another parameter in the Radiation Trait, but I think it adds another fuss to the user and it's not worth it since we will almost never use it.

We could do this: AMReX-Codes/amrex#2954

[chongchonghe]

We could createa new version of an existing test that uses the wavespeed correction in the Riemann solver to test this. I think that would help avoid this problem.

The issue is, you can't turn it on for any specific problem. It's controlled by use_wavespeed_correction which is not in any traits. You have to change radiation_system.hpp to turn it on. I can, of course, add another parameter in the Radiation Trait, but I think it adds another fuss to the user and it's not worth it since we will almost never use it.

We could do this: AMReX-Codes/amrex#2954

That feature is interesting. Have you tried it? Is there any example in Quokka? I'm not sure how to set runtime_option in problem.cpp.

[BenWibking]

That feature is interesting. Have you tried it? Is there any example in Quokka? I'm not sure how to set runtime_option in problem.cpp.

No, it's not currently used in Quokka.

In the code below, you can set runtime_option just like any other variable in the code. You could initialize its value by querying the ParmParse database, for instance, so that it's set from a command-line parameter.

    int runtime_option = ...; // set this based on the value of any variable, read it from an input file, read it from the command-line, do whatever you want
    enum All_options : int { A0, A1, A2, A3};
    // Four ParallelFors will be generated.
    ParallelFor(TypeList<CompileTimeOptions<A0,A1,A2,A3>>{}, {runtime_option},
                box, [=] AMREX_GPU_DEVICE (int i, int j, int k, auto control)
    {
        ...
        if constexpr (control.value == A0) {
            ...
        } else if constexpr (control.value == A1) {
            ...
        } else if constexpr (control.value == A2) {
            ...
        else {
            ...
        }
        ...
    });

[BenWibking]

Here's an example in WarpX:
https://github.com/ECP-WarpX/WarpX/blob/675369d9797aa2bcfda8b64c8e500fff6a2b9f04/Source/Particles/RigidInjectedParticleContainer.cpp#L424

[chongchonghe]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).

[chongchonghe]

@BenWibking Can you review this PR?

[chongchonghe]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).

[BenWibking]

I am confused. Can you explain more explicitly what opacity functions should be used by gray problems and what opacity functions should be used (and how) by multi-group problems?

[chongchonghe]

I am confused. Can you explain more explicitly what opacity functions should be used by gray problems and what opacity functions should be used (and how) by multi-group problems?

I updated the PR comments to better explain this.

[chongchonghe]

/azp run

[azure-pipelines]

Azure Pipelines successfully started running 2 pipeline(s).