DUNE integration

[sethrj]

This is the primary tracking issue for integrating Celeritas with the simulation workflow for DUNE. The primary Celeritas contact points are the "assignees" to the right.

Core capabilities

• [x] Optical photon physics: https://github.com/celeritas-project/celeritas/issues/886
• [x] ORANGE for platform portability: https://github.com/celeritas-project/celeritas/issues/1167
• [ ] Larsoft integration

Key requirements

• [ ] Geometry
• [ ] Scintillation

[sethrj]

Notes from today's meeting (compiled an hour later so hopefully I remember the important bits):

• Cherenkov is neglected in current simulations; it's only a few percent, and the detectors weren't designed for it. (There's currently no directional capabilities, the detector geometry doesn't favor capturing the light, and the photodetectors can't discriminate against UV light.)
• The optical photon mapping includes time dependence, but a finely resolved grid (necessary for higher fidelity) requires too much memory
• Currently the framework decouples energy deposition from optical responses at a framework level: energy deposition from steps are gathered during the Geant4 simulation, and afterward another stage of processing does the optical response calculation
• Future work could move back to a model where optical photons are simulated during the main simulation
• DUNE uses GDML for input (we have a copy)
• They have a limited closed-source standalone simulation we might be able to try out
• Surface physics is important; there will soon be reflective materials to worry about

[sethrj]

@whokion You can forward this issue page to Erica with a link to our (almost current) status? https://indico.cern.ch/event/1466097/contributions/6216512/attachments/2965789/5217806/celeritas-optical-photon.pdf

[sethrj]

To get from DUNE:

• Physics list construction (done through larsoft: larsim, larg4)
• GDML export
• Sensitive detector definitions

Ken to send primer of running larsoft event which should point us to physics list etc.

Note that LarSoft will be mostly replaced by ('migrated to') Phlex

[stognini]

From the discussion with Laura Paulucci:

• DUNE's fast optical simulation is handled in LArSoft via larsim/PhotonPropagation. This is the workflow on a production-like scenario, with a particle depositing energy and generating a light signal in the PDS.
• Particle gun-like primary generation also exists independently for individual use-cases in larsim/EventGenerator.