HDGeant4 Meeting, August 10, 2021

From GlueXWiki
Jump to: navigation, search

HDGeant4 Meeting
Tuesday, August 10, 2021
3:00 pm EDT
BlueJeans: 968 592 007


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


Present: Alex Austregesilo, Tegan Beattie, Mark Ito (chair), Igal Jaegle, Richard Jones, 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 reviewed the minutes from the meeting on July 13.

Mark has succeeded building the GlueX software stack with ROOT 6.24.0 and Geant4 10.06.p01 using GCC 8 using three different methods:

  1. On a CentOS 7 singularity container with Developers Toolset 8. This was announced on the offline list.
  2. Using GCC 8.3.0 supplied by a module on the ifarm.
  3. Using a CentOS 8 singularity container using its native GCC, 8.3.1

We marked several issues as ready-to-be-closed. We need to make sure that this is done.

Efficiency Tables

Igal asked if there is a way to check if the correct efficiency tables are being applied to simulated data. He received guidance from Richard (use the MC variation of the CCDB) and Alex (there are dead wire maps for the FDC).

Issue #181: G3/G4 Difference in FDC wire efficiency at the cell boundary

There has been a lot of work on this issue. It has morphed from a G3 vs. G4 comparison question to an effort to do a faithful simulation of hit efficiency as a function of track position in the cell.

The process of changing the absolute level of the efficiency curve to get agreement with data has converged. It was complicated by the fact that changing the level affected both the numerator and denominator in the "efficiency" measure used in the comparison. This is probably due to the measure requiring 4 of 6 hits in any package being considered; a loss of a single hit could significantly reduce the number of "tracks" being considered.

Richard also repeated the comparison between data and Monte Carlo with a recent version of tracking using bggen MC to compare with the ρ data used in all other studies. Here the "efficiency" measure was higher that that obtained by Alex with a much older version of reconstruction, but the agreement between data and MC was just as good.

Alex will do a check that he gets a similar result with bggen simulated, including random triggers in the simulation, a feature not present in Richard's study.

In all we appear to be very close to closing this issue.

Issue #192: Vertex generation used with BHgen

Richard described work getting realistic generation of Bethe-Heitler pairs (both electrons and muons) from a lead target, as will be used for CPP, in response to this issue. The details of his effort are recorded in his notes, in the section New pair converter targets. He showed a model that incorporates a nuclear form factor for elastic scattering and for inelastic processes quasi-elastic scattering from protons and neutrons, where the Fermi motion of the nucleons is taken into account. He had compared his model with one that comes from Geant4. The Geant4 model does not have a nuclear form factor and has no model for inelastic scattering. Richard sees differences in the cross section as a function of Q2 which he attributes to these lacks in the Geant4 model.

As a consistency check, he re-did the comparison, restricting Q2 to values less than 10-4 (GeV/c)2. In the this region the agreement between his model and Geant4 is all but exact.

Future Projects

We (aka Richard (mainly)) discussed ideas for other developments in simulation for GlueX.

  • General QED process event generator a la the BH generator. User would be able to specify the process. Like the BH generator, it would take into account the spatial distribution of the photon energy and polarization as a function of position of emergence from the collimator.
  • Radiative corrections for cross section measurements.
  • Neutrons and KLs.
  • CDC efficiency.
  • Variance and co-variance agreement between data and simulation.