Title: New g2 Calorimeters
1New g-2 Calorimeters
2Calorimeter Segmentation
4cm
3Structure of BlocksDecisions
- Depth of block
- W/SciFi ratio
- Number of layers
- Close packing
- Harder to construct
- Better resolution
4Energy Resolution and Statistical Error
- Simulated data
- Inserted energy resolution
- 2 GeV cut
- Fit 5 parameter function
- 10 in resolution means 20 more run time
Error on vs Energy Resolution
5GEANT 4 Calorimeter
- Recently completed
- Layers of Tungsten and SciFi (and glue)
- Input electrons from old g2GEANT
- Parameters
- Number of Layers
- Fiber Diameter
- Depth of Block
- 10000 events/hr
6Output Data
- Initial Position
- Initial Momentum/Energy
- Total energy deposited in calorimeter
- Energy deposited in the fibers of each block
- Output as text file, converted to ROOT ntuple
7Determining Energy Resolution
- Calibrate to convert fiber energy to total
- Plot Energy (MeasuredActual)/Actual
- Use hits in detector center
- Mean of is offset in calibration
- RMS is fractional energy resolution
8Resolution vs Energy
20 Layers 0.8mm Fibers 1.2mm Tungsten
Expected
GeV
9More Layers, Same W/SciFi
40 Layers 0.4mm Fibers 0.6mm Tungsten Same W/SciFi
Better resolution but still wrong energy
dependence
GeV
10Larger Proportion of Fiber
20 Layers 1.2mm Fibers 0.8mm Tungsten
Again Better resolution, wrong energy dependence
11Whats the Problem?
- Depth of Calorimeter 15cm (Artificially high)
- Fiber geometry
- GEANT shower parameters
Variation in resolution vs angle for 2GeV Beam
12Number of Blocks Hit
of Hits gt 200 MeV
of Hits gt 20 MeV
Hit
13Number Hit vs Electron Energy
of Hits gt 200 MeV
of Hits gt 20 MeV
14To Do
- Get GEANT4 running on more machines
- Make parameters adjustable, not hard coded.
- Questions
- Why does the resolution have the wrong energy
dependence? - Is the close packing scheme significantly better
than fiber strips? - What block depth and W/SciFi ratio should we use?
- At what level can we separate pileup events?
- What is the best layout of blocks?