Kill photons as soon they enter into a OpticalTracker

[atolosadelgado]

BEGINRELEASENOTES

• Geant4OpticalTracker: Kill optical photons (instead of all other particles) as soon they enter into a Optical Tracker sensitive detector

ENDRELEASENOTES

Hi,

Please, correct me, but I think the condition to kill the photons when they enter an Optical Tracker is the other way around.

I ran a simulation of optical photon transport with and without changing that condition. The default DD4hep do not kill photons when they enter the Optical Tracker sensitive detector, but that behavior happens when the condition is reversed (as in the suggested change in this PR).

Test: I did some tests with version 1.23 of DD4hep. The geometry is simple: a box (laboratory) made of "optical air", showed as a semitransparent, contains a sensitive detector made of "optical silicon", showed as red. The particle gun fires photons from the right to the left. The refractive index in the laboratory and the sensor is the same. The photons are stopped when they try to jump from the "laboratory" to the "world" (because World has no optical properties)

1. This plot shows a track of a photon that crosses the sensor, the laboratory and it is only stopped when trying to cross to the world.
   
2. This plot shows a track of a photon after changing the condition of killing particle in OpticalTracker SD Action
   
3. This plot shows a track of a proton, same direction, with default DD4hep
   
4. This plot shows a track of a proton, same direction, after changing the condition in OpticalTracker SD Action
   

The following code reproduce the plots:

git clone -b optical_tracker_test https://github.com/atolosadelgado/dd4hep_cube.git
cd dd4hep_cube
cmake -B build -S . -D CMAKE_INSTALL_PREFIX=install
cmake --build build -j 6 -- install
export LD_LIBRARY_PATH=$PWD/install/lib:$LD_LIBRARY_PATH
# Gun = optical photon
python3 test.py --compactFile compact/simple_detector.xml --runType qt --macroFile vis.mac --gun.energy 4*eV --gun.particle opticalphoton
# Gun = proton
python3 test.py --compactFile compact/simple_detector.xml --runType qt --macroFile vis.mac --gun.energy 4*GeV --gun.particle proton

[veprbl]

cc @wdconinc

[github-actions]

Test Results

       6 files         6 suites   7h 4m 21s ⏱️
   356 tests    356 ✔️ 0 💤 0 ❌
1 058 runs  1 058 ✔️ 0 💤 0 ❌

Results for commit 6725f8d.

♻️ This comment has been updated with latest results.

[andresailer]

@wdconinc : As the original code came from you in #967

Could you review this please?

[wdconinc]

I'll try to take a look today, at the latest by the end of the week. (cc @veprbl @c-dilks who may have comments too)

[c-dilks]

The RICH tests still pass, so I assume this change is okay for our ePIC dRICH.

[atolosadelgado]

the bug is more evident if you remove these lines from the steering file

    # Some detectors are only sensitive to optical photons
    SIM.filter.filters["opticalphotons"] = dict(
        name="ParticleSelectFilter/OpticalPhotonSelector",
        parameter={"particle": "opticalphoton"},
        )
    SIM.filter.mapDetFilter["ACUBE"] = "opticalphotons"

[andresailer]

@wdconinc Any objection from your side?

[wdconinc]

I haven't had a chance to test this (swamped with a review), but I looked at the code and it seems like the right fix so don't let me hold this up. I'm not completely sure why it seemed to have worked for us in the past, but likely because we had thin detector volumes and the same thing happened as described here: photons are killed when leaving the detector, or are even absorbed in the detector (silicon PM).