Difference between revisions of "HDGeant4 Meeting, June 1, 2021"

From GlueXWiki
Jump to: navigation, search
(first draft)
(Charged Hadrons and BCAL Timing: add Tegan's figures)
 
Line 26: Line 26:
 
Tegan showed several plots, which he previewed at the last meeting, comparing timing distributions for charged hadrons between data and Monte Carlo. On the simulation side he plots both Geant4 and GEANT/HADR4. He used exclusive ρ events for both data and MC. He plots the difference in the time predicted by charged particle tracking to that measured in the BCAL.
 
Tegan showed several plots, which he previewed at the last meeting, comparing timing distributions for charged hadrons between data and Monte Carlo. On the simulation side he plots both Geant4 and GEANT/HADR4. He used exclusive ρ events for both data and MC. He plots the difference in the time predicted by charged particle tracking to that measured in the BCAL.
  
* For protons the agreement in the core Gaussian part of the distributions is good for all three species; see ???Figure 1. The data does show non-Gaussian tails on the log plot shown that could be due to event selection. We suggested that he try a tighter cut on the &chi;<sup>2</sup> of the kinematic fit to see if that reduces the tails. Tegan reported that on a linear plot, these tails are not visible to the eye.
+
* For protons the agreement in the core Gaussian part of the distributions is good for all three species; see [[Media:Proton tegan.pdf|this plot]]. The data does show non-Gaussian tails on the log plot shown that could be due to event selection. We suggested that he try a tighter cut on the &chi;<sup>2</sup> of the kinematic fit to see if that reduces the tails. Tegan reported that on a linear plot, these tails are not visible to the eye.
* For pions the agreement among all three is very good, with a high side tail visible in all three species; see ???Figure 2.
+
* For pions the agreement among all three is very good, with a high side tail visible in all three species; see [[Media:PiPlus tegan.pdf|this plot]].
  
 
Mark suggested that if Tegan does not uncover further problems, we should go back to the original studies of event efficiency that launched us on this path. Those studies showed overall event reconstruction efficiency showing large differences between the various simulation engines. We have since concluded that GEANT/HADR1 does not give reliable timing distributions for charged hadrons in the calorimeters, but the comparison between Geant4 and GEANT/HADR4 is still relevant. There have been many changes in the code since those studies were performed and we should see where we stand with the latest versions. Zisis mentioned that Ahmed Foda's &omega;&pi; analysis shows good agreement in cross sections between Spring 17, Spring 18, and Fall 18 lately, a new development.
 
Mark suggested that if Tegan does not uncover further problems, we should go back to the original studies of event efficiency that launched us on this path. Those studies showed overall event reconstruction efficiency showing large differences between the various simulation engines. We have since concluded that GEANT/HADR1 does not give reliable timing distributions for charged hadrons in the calorimeters, but the comparison between Geant4 and GEANT/HADR4 is still relevant. There have been many changes in the code since those studies were performed and we should see where we stand with the latest versions. Zisis mentioned that Ahmed Foda's &omega;&pi; analysis shows good agreement in cross sections between Spring 17, Spring 18, and Fall 18 lately, a new development.

Latest revision as of 18:07, 1 June 2021

HDGeant4 Meeting
Tuesday, June 1, 2021
3:00 pm EDT
BlueJeans: 968 592 007

Agenda

  1. Announcements
  2. Review of minutes from the last meeting (all)
  3. Issues on GitHub
  4. Pull Requests on GitHub
  5. Action Item Review

Minutes

Present: Alex Austregesilo, Tegan Beattie, Sean Dobbs, Mark Ito (chair), Igal Jaegle, Richard Jones, Zisis Papandreou, Simon Taylor, Beni Zihlmann

There is a recording of this meeting on the BlueJeans site. Log into the BlueJeans site first to gain access (use your JLab credentials).

Review of minutes from the last meeting

We went over the minutes from the meeting on May 18 without a lot of discussion.

Charged Hadrons and BCAL Timing

Tegan showed several plots, which he previewed at the last meeting, comparing timing distributions for charged hadrons between data and Monte Carlo. On the simulation side he plots both Geant4 and GEANT/HADR4. He used exclusive ρ events for both data and MC. He plots the difference in the time predicted by charged particle tracking to that measured in the BCAL.

  • For protons the agreement in the core Gaussian part of the distributions is good for all three species; see this plot. The data does show non-Gaussian tails on the log plot shown that could be due to event selection. We suggested that he try a tighter cut on the χ2 of the kinematic fit to see if that reduces the tails. Tegan reported that on a linear plot, these tails are not visible to the eye.
  • For pions the agreement among all three is very good, with a high side tail visible in all three species; see this plot.

Mark suggested that if Tegan does not uncover further problems, we should go back to the original studies of event efficiency that launched us on this path. Those studies showed overall event reconstruction efficiency showing large differences between the various simulation engines. We have since concluded that GEANT/HADR1 does not give reliable timing distributions for charged hadrons in the calorimeters, but the comparison between Geant4 and GEANT/HADR4 is still relevant. There have been many changes in the code since those studies were performed and we should see where we stand with the latest versions. Zisis mentioned that Ahmed Foda's ωπ analysis shows good agreement in cross sections between Spring 17, Spring 18, and Fall 18 lately, a new development.

The Schick Effect

We discuss the problem Andrew Schick discovered in his Bethe-Heitler analysis where the total energy of the B-H pairs in the FCAL were lower than the beam photon beam energy by 4 to 5%. Even in simulation some of this deficit is seen, though only by 1 to 2%. Beni mentioned that there is also a difference seen between electrons and positrons. Alex also reminded us that Drew Smith (Compton scattering) and Mark Dalton (ω production) both showed effects in the cross sections that pointed to inefficiencies in the inner blocks of the FCAL.

At this meeting, the question was whether there might be some insight to be gained by studying these effects in simulation in the future.

Simulation GlueX Doc

Richard floated the idea of the group writing a GlueX Doc summarizing the knowledge gain by this group vis-a-vis the fidelity of our simulation so that the ideas don't get lost. This would be along the lines of the document being produced by the Beamline Working Group on beamline-related systematics.

Exploiting New Features of Geant4

Richard reported on work he has been doing to incorporate circular polarization into the simulation. One aspect of his work is to use recently introduced polarization dependent cross sections from Geant4. He finds that it is sufficient to go to version 10.6.p1 to pick these up, but that requires GCC 5.3.1 (again sufficient if not necessary) which is beyond that provided by CentOS 7. He is using Devtoolset 4[?] to get the compiler. He stressed the need for GlueX to press forward with using more modern versions of Geant4 and other packages.

Bethe-Heitler Event Generator

Richard has been working on a built-in Bethe-Heitler generator for HDGeant4. This is to support Andrew's work, mentioned above, with a more accurate physics model that the one Andrew has been using thus far.

Action Items

  1. Re-do proton-in-BCAL timing study with tighter cuts on the χ2 from the kinematic fit. (Tegan)
  2. Re-do comparisons of overall event reconstruction efficiencies.
  3. Think about how to study the Schick-Smith-Dalton effect(s). (all)
  4. Think about a GlueX Doc on simulation. (all)
  5. Think about how to get to modern versions of external software packages. (all)