Integrate picsar_qed in ablastr : move in the Schwinger process #6427
Conversation
lucafedeli88
added
cleaning
for more information, see https://pre-commit.ci
lucafedeli88
changed the title
lucafedeli88
changed the title
lucafedeli88
changed the title
![github-advanced-security[bot]](https://avatars.githubusercontent.com/in/57789?s=60&v=4){kind=small}
| doQEDEvents(); | ||
| mypc->doQEDSchwinger(); | ||
| #endif | ||
| mypc->doQEDSchwinger(); |
There was a problem hiding this comment.
Why this?
I think it makes sense that the Schwinger effect can only be used when compiling with QED ON, no?
Even if it's not technically essential...
There was a problem hiding this comment.
We have to discuss this.
In principle, if we integrate picsar QED into ablastr, we could enable QED effects at no cost (no additional libraries). We could have a compilation flag only for lookup tables generation (this feature requires the boost library).
Since I see this PR as a step towards the full integration of picsar QED in ablastr it seemed reasonable to me to start putting stuff that can be used without picsar out of the #ifdef regions. However, this change can be certainly delayed till the end of the integration.
There was a problem hiding this comment.
Ahh I see! We shall discuss at the developers meeting then.
I'll add it to the agenda.
There was a problem hiding this comment.
Same as above, I don't see why we would prefer this.
|
|
||
|
|
||
| void | ||
| MultiParticleContainer::doQEDSchwinger () |
There was a problem hiding this comment.
Why doesn't this code belong to ablastr?
There was a problem hiding this comment.
I think that it is a bit too WarpX - specific to be in ablastr. Would you suggest moving the whole function to ablastr? Adding a doxygen comment is certainly a good idea.
There was a problem hiding this comment.
Actually, I've noticed that the doxygen statement is in the MultiParticleContainer.H header file:
/** If Schwinger process is activated, this function is called at every
* timestep in Evolve and is used to create Schwinger electron-positron pairs.
* Within this function we loop over all cells to calculate the number of
* created physical pairs. If this number is higher than 0, we create a single
* particle per species in this cell, with a weight corresponding to the number of physical
* particles.
*/
void doQEDSchwinger ();
Looking again at the code, I still think that it is a bit too warpx-specific to be in ablastr...
Source/ablastr/CMakeLists.txt
Outdated
| add_subdirectory(parallelization) | ||
| #add_subdirectory(particles) | ||
| #add_subdirectory(profiler) | ||
| add_subdirectory(qed) |
There was a problem hiding this comment.
yes, that's a good idea. I'll do that.
| convert_from_to<unit_system::heaviside_lorentz, UnitSystem, quantity::rate>(1.0_rt); | ||
|
|
||
| constexpr auto pi = ablastr::constant::math::pi; | ||
| if(epsi != 0.0_rt && eta != 0.0_rt){ |
Check notice
Code scanning / CodeQL
Equality test on floating-point values Note
There was a problem hiding this comment.
I think it's OK here. It protects against a corner case replacing an expression that would give a NaN with a valid expression obtained by computing the limits eta -> 0 or epsi -> 0.
| convert_from_to<unit_system::heaviside_lorentz, UnitSystem, quantity::rate>(1.0_rt); | ||
|
|
||
| constexpr auto pi = ablastr::constant::math::pi; | ||
| if(epsi != 0.0_rt && eta != 0.0_rt){ |
Check notice
Code scanning / CodeQL
Equality test on floating-point values Note
| if(epsi != 0.0_rt && eta != 0.0_rt){ | ||
| return coeff*rate_conv*epsi*eta*std::exp(-pi/epsi)/std::tanh(pi*eta/epsi); | ||
| } | ||
| else if(epsi == 0.0_rt){ |
Check notice
Code scanning / CodeQL
Equality test on floating-point values Note
3 participants
Summary
This PR is the second step of a long-overdue process: the integration of picsar_qed into ablastr. Specifically, this PR moves the Schwinger pair production process into
ablastr/sfqedand makes the necessary changes in the rest of the code . In addition, some details on the Schwinger process are added to the documentation.The code is adapted from PICSAR QED, with minor changes.
Main changes:
Calculations for the Schwinger process are performed using
amrex::Realandamrex::RealVectND<3>types.Source/ablastr/sfqed/SchwingerProcess.H: this new source file contains the methods that compute the pair production rateSource/ablastr/sfqed/UnitConversion.H: QED-related calculations are performed internally using the Heaviside-Lorentz units. This new file contains helper functions to convert quantities between these units and SI units.Source/Particles/ElementaryProcess/QEDInternals/SchwingerProcessWrapper.H: this file is no-longer needed. The only one method that was implemented there is now inablastr/sfqed/SchwingerProcess.H.Source/Particles/ElementaryProcess/QEDSchwingerProcess.H: in this file a simple name change to call the new method inablastrinstead of the one in the wrapper is performed.Source/Particles/MultiParticleContainer.H/cpp: the Schwinger process no longer requires the PICSAR QED library. Therefore, the relevant code is moved out of the#ifdef WARPX_QED ... #endifsections.Source/Evolve\WarpXEvolve.cpp: same as aboveDocs/source/usage/parameters.rst: the documentation is updated to reflect that the Schwinger process no longer requires the PICSAR QED library.Docs/source/theory/multiphysics/qed.rst: some details on the physics of the Schwinger process are added to the documentation.Docs/source/refs.biba reference to a paper cited inDocs/source/theory/multiphysics/qed.rstis added.Regression/Checksum/benchmarks_json/test_3d_qed_schwinger_4.json: I had to reset the checksum, but the change is really tiny (and the python script passes) .