Hi developers. Thank you for a nice software which is also being used (forked) for muon collider simulation.
I have a question about DDPlanarDigiProcessor. I'm simulating a background-only data sample, where "background" is a diffuse, low-energy beam-induced background. The background is a mix of mostly photons, neutrons, and electrons. And the background is diffuse because it's mostly produced when beam muons decay to electrons which strike shielding outside the detector itself. None of the background is produced at a collision point at the origin.
We have tracking detectors, and I noticed approximately 40% of the tracker sim hits are flagged as failing surf->insideBounds, so the sim hits aren't digitized by DDPlanarDigiProcessor. 40% seems large to me, since we already require tracker sim hits to deposit at least 1 keV of energy with SIM.filter.tracker = "edep1kev" in ddsim. Btw: "Inside bounds" is defined as "distance of the hit from the surface of the sensor is less than epsilon", where distance from the sensor surface is ( point - origin() ) * normal().
Digitzation: https://github.com/iLCSoft/MarlinTrkProcessors/blob/master/source/Digitisers/src/DDPlanarDigiProcessor.cc#L254
Inside bounds: https://github.com/AIDASoft/DD4hep/blob/master/DDRec/src/Surface.cpp#L207
Distance: https://github.com/AIDASoft/DD4hep/blob/master/DDRec/src/Surface.cpp#L271
I made a few plots of tracker sim hits which are inside-bounds vs outside-bounds, and it seems like most of the outside-bounds hits are passing through the sensor at a large angle (cos_theta ~= 0), and they usually have a path-length less than the depth of the sensor, which is 100um. Plots are below.
Do you have a physical intuition for what is happening to these hits: is their path-length less than the depth of the sensor because they are from MC particles stopping/starting in the sensor? Or is there another effect which can cause this?
And from a simulation perspective, do you have an intuition about whether we should be removing these hits at digitization? Naively, these seem like reasonable hits to me. I'm not sure why they should be removed if their distance from the central plane of the sensor is non-zero. But I'm a beginner at simulating silicon detectors!
Thanks very much for any help.
Alex
Notes about muon collider and its backgrounds, in case its helpful:
https://indico.uchicago.edu/event/479/contributions/2022/attachments/815/1295/spagan-detector-usmcc%20(4).pdf
cc @TC01 @madbaron @tmadlener
Hi developers. Thank you for a nice software which is also being used (forked) for muon collider simulation.
I have a question about DDPlanarDigiProcessor. I'm simulating a background-only data sample, where "background" is a diffuse, low-energy beam-induced background. The background is a mix of mostly photons, neutrons, and electrons. And the background is diffuse because it's mostly produced when beam muons decay to electrons which strike shielding outside the detector itself. None of the background is produced at a collision point at the origin.
We have tracking detectors, and I noticed approximately 40% of the tracker sim hits are flagged as failing
surf->insideBounds, so the sim hits aren't digitized by DDPlanarDigiProcessor. 40% seems large to me, since we already require tracker sim hits to deposit at least 1 keV of energy withSIM.filter.tracker = "edep1kev"in ddsim. Btw: "Inside bounds" is defined as "distance of the hit from the surface of the sensor is less than epsilon", where distance from the sensor surface is( point - origin() ) * normal().Digitzation: https://github.com/iLCSoft/MarlinTrkProcessors/blob/master/source/Digitisers/src/DDPlanarDigiProcessor.cc#L254
Inside bounds: https://github.com/AIDASoft/DD4hep/blob/master/DDRec/src/Surface.cpp#L207
Distance: https://github.com/AIDASoft/DD4hep/blob/master/DDRec/src/Surface.cpp#L271
I made a few plots of tracker sim hits which are inside-bounds vs outside-bounds, and it seems like most of the outside-bounds hits are passing through the sensor at a large angle (cos_theta ~= 0), and they usually have a path-length less than the depth of the sensor, which is 100um. Plots are below.
Do you have a physical intuition for what is happening to these hits: is their path-length less than the depth of the sensor because they are from MC particles stopping/starting in the sensor? Or is there another effect which can cause this?
And from a simulation perspective, do you have an intuition about whether we should be removing these hits at digitization? Naively, these seem like reasonable hits to me. I'm not sure why they should be removed if their distance from the central plane of the sensor is non-zero. But I'm a beginner at simulating silicon detectors!
Thanks very much for any help.
Alex
Notes about muon collider and its backgrounds, in case its helpful:
https://indico.uchicago.edu/event/479/contributions/2022/attachments/815/1295/spagan-detector-usmcc%20(4).pdf
cc @TC01 @madbaron @tmadlener