Epic-backward-hcal-l mailing list

[Leszek Kosarzewski <leszek.kosarzewski AT gmail.com>]

Dear Colleagues

Thank you to all who made this very productive mini-overview happen. We discussed a few important points:

1. Well defined physics requiring nHCal, crucial to the core EIC physics program (low-x, low-Q^2), not well known structure of p,A
2. We may keep in mind the vendors unofficially, but official selection of vendors is done by the Project via competition
3. nHCal installation to happen after magnet acceptance tests: https://indico.bnl.gov/event/27365/attachments/60823/104491/Installation%20Schedule%20for%20the%20EIC.12.2024.docx
4. Beam tests with high energy hadrons may not be necessary, not a focus
5. Instead do tests with cosmic muons and low energy neutrons (ORNL spallation source O(100) MeV neutrons?)
6. Test performance of different tile configurations: SiPM on tile, WLS fibers/plates (see STAR FCS HCal)

1. https://drupal.star.bnl.gov/STAR/files/ForwardUpgrade.v20.pdf
2. https://iopscience.iop.org/article/10.1088/1742-6596/587/1/012053
3. 10 cm x 10 cm tiles may have efficiency/uniformity issues, smaller tiles offer smaller light yield, use thicker ones if needed

e/h~1 not important, because EEEMCal breaks compensationNeutron and muon detection efficiency most important for geometry determination

1. Most likely need readout from separate layers to identify MIP signal
2. Include EEMCal with a full test
3. For neutrons, need good signal detection in the first few layers

Use single particles to study performance, check extreme casesNext, include background with a full simulation and confirm realistic performance vs. backgroundMake adjustments to simple geometry, implement realistic module design once fixed

To answer the question about the long integration time, it's illustrated by the picture in slide 64:

https://www.desy.de/~garutti/LECTURES/ParticleDetectorSS12/L10_Calorimetry.pdf

There is a large delayed component coming from gammas from slow neutron capture. However, I looked into the details and this is not what's affecting the detection efficiency.

We did it with pure simulation in npsim only, just to integrate energy deposits over time to study the neutron detection efficiency. This is a simple calculation and we are aware of this. The cut of 100 ns seemed like the best choice. We used it to provide parameters for the current, simple digitization model, which indeed applies the same cut with respect to the start time of the simulation (t=0). This is the limitation of the current digitization model. It will be improved with the new model, discussed last week. We did a check, where we used the time of the first hit as a reference/start time, as a realistic digitization would do. This is shown in slide 28:

https://indico.bnl.gov/event/20473/contributions/84203/

This points to no difference, which is good. We can go lower. However, we still used the start time t=0, since that's what the current digitization really does. Slide 25 also illustrates the delay of the first signal, due to neutron flight time.

Best regards, Leszek

pon., 24 mar 2025 o 14:10 Leszek Kosarzewski <leszek.kosarzewski AT gmail.com> napisał(a):

Dear Colleagues

This is a reminder, we are going to have a backward HCal DSC meeting tomorrow at 1400h:

https://indico.bnl.gov/event/27365/

This time the plan is to have a planned mini-review with experts.

Best regards, Leszek