Suggested Tests

Introduction

Suggested Simulations:

• Simulate a single ACD tile
  • 50,000 normally incident 300 MeV muons
  • 50,000 normally incident 1 GeV muons
  • N 300 MeV gammas - test Compton Scattering - does the cross-section make sense?
• Simulate a single ACD ribbon
  • Need to update the flight.xml file to reflect the current geometry.
• Simulate full GLAST
  • Simulate 50,000 300 MeV muons, normally incident.  Perform simulations of muons normally incident on each face with ACD tiles - to test all tiles.
  • Same test, 1 GeV muons
  • Simulate off-axis muons
  • Simulate backgndmix - let's say 100,000 events
• Backsplash simulations
  • High energy mono-energetic gamma - shot into CAL module of a side tower.
  • High energy mono-energetic gamma - shot into the side of the CAL.

Run all simulations with and without noise, poisson fluctuations, and gain factors.

Low level tests
The packages that contain acd digitization and reconstruction should provide some very basic tests:

• Check for memory leaks.
• Check conversion from MIPs to photo-electrons (pe)
• Poisson fluctuations - does the resulting pe distribution look correct?
• Gaussian noise - does the resulting MIP distribution look correct?
• Are the ACD discriminators set when they should be?
• Do the ACD_DOCAs correspond to the actual ACD tile that was hit?
• Test the actual DOCA algorithm to be sure it accurately calculates the distance from the edge of the appropriate tile.

Plots:

• Plot deposited energy for each tile and ribbon.
  • Does the Monte Carlo output support the reported energies?
• Plot PHA for each PMT.
  • Perform pedestal subtraction first - if necessary.
• Plot efficiency versus threshold for each PMT
  • Determine nominal threshold - 0.3 of peak PHA
  • What fraction of events are over threshold?  Where the threshold is taken from 0.3 peak PHA through 1.0 corresponding to the peak PHA (1 MIP).
  • Compare efficiencies to our target 0.9997
• Plot combined efficiency of the 2 PMTs per tile and ribbon.
  • Take the OR of the 2 PMTs
  • Is the combined efficiency of the 2 PMTs for a tile higher than the individual efficiency of one PMT?
• ACD and CAL response
  • Plot Total Cal sum (ADC bin) vs ACD PHA
• Distribution of ACD tile count
  • Does the distribution correspond to the number of tiles that should have been hit?
  • Does the Monte Carlo output support the number of reported hit tiles?