CRT Simulation Updates Richie Diurba, Hans Wenzel, David Rivera - - PowerPoint PPT Presentation

CRT Simulation Updates

Richie Diurba, Hans Wenzel, David Rivera

diurb001@umn.edu February, 2020

1/10

Changes

• Placed CRT reco and matching back into the reco fcl file.
• Fixed AuxDet exit positions (Hans)
• Clustered AuxDet hits, no more splitting of energy depositions. (Hans)
• Added trigger backtracking that connects a CRT simulated trigger to all

MCParticles that are above threshold. (David)

2/10

Change to AuxDet

Hans did an immense amount of work. Thanks to him we have:

• Exit positions fixed.
• Used a clustering of temporary AuxDet hits to make a single AuxDet hit. This

required an intermediate step in larg4.

• This is important because if a hit gets split due to physics in the scintillator we

cannot backtrack the hits as the splitting leads to some hits having a different track ID. To cluster them in detector simulation, we’d have to take all thousands

• f hits and make maps of maps, we agreed it is much easier to do it during the G4

stepping.

3/10

Entry and Exit Positions using new AuxDet

Double checking that Aux Det gives right positions.

cm

4/10

Entry and Exit Positions using new AuxDet

Everything looks good! The gen file used for this had muons going through the front BR CRTs.

cm and ns

5/10

Energy Deposited using new AuxDet

Should give a Landau distribution. It does!

MeV

6/10

Rate Tests

The biggest issue was that matching MCParticle to CRT reco for cosmic events had a 25% reduction in rates. This is due to the reduction in energy deposits from the parent-child bifurcation.

• 2.32 hits per event in legacy for cosmic events
• 1.82 hits per event in old refactored for cosmic events

To test with the new clustering, I fired a single muon one thousand times within a rectangle of around 1 m2 on the BR CRTs. The same gen file was used for all G4 simulations.

7/10

SingleGen Truth-CRT Matching Rate Test

Matched CRT hits to MCParticle using new AuxDet

Displacement between CRT hit and MCParticle position on CRT for BR CRTs using a sample

• f 1k muons. Strips are 5 cm wide overlaid to create an effective width of 2.5 cm.

8/10

Rates using new AuxDet

Out of 1000 single gen muons:

• 856 matched hits in new refactored
• 843 matched hits in legacy
• 658 matched hits in old refactored

The 25% discrepancy seen last month then was due to not clustering the hits. We now match the event rate of legacy.

9/10

Moving Forward

• Hans did extensive work to redo the AuxDet clustering and it looks like it worked!
• Backtracking has been added and entry and exit position issues with the product

revalidated.

• The reco modules give us CRT reco tracks in the new samples. Should look

towards double-checking that those make sense next.

10/10