Start Counter Simulations

From GlueXWiki
Jump to: navigation, search

Geant4 Simulations

Geometry and Setup

  • Scintillator: Imported AutoCAD drawing into GEANT4 using CADMESH utility
  • SiPM: 12x12x10 mm volume
  • Surrounded by Air
  • Spacing between scintillator and SiPM: 0.1 mm
  • Generated 10K optical photons at every 2.5 cm spacing inside the scintillator (-300, 300) mm or (0, 60) cm

Material Properties

  • Scintillator: ELJAN 212 material composition properties
    • Density: 1.023 g/cm3
    • Scintillation yield: 10000 photons/MeV
    • Optical photon energy range: 0.5 eV - 3.76 eV
    • Index of refraction: 1.58
    • Absorption length: 400cm (not correct!)
  • SiPM: Silicon
    • Efficiency: 100%
    • Used just to track optical photon hits

Optical Surface

  • The actual scintillators under microscope: [4] [5]
  • An optical surface between air and scintillator is set
  • Properties of optical surface are defined by
    • Material of surfaces: dielectric_dielectric
    • Physics model: unified Levin and Moisan, 1996
    • Transmission efficiency and reflection
    • Finish of the surface: polished or ground
      • polished: Fresnel reflection, Total Internal Reflection, Fresnel Refraction
      • ground: Spike reflection, lobe reflection, backscatter, Lambertian reflection, Fresnel refraction
Spike reflection: the photon is reflected like a perfect mirror or about the average surface normal
Lambertian reflection: the photon is reflected with a Lambertian distribution
Back scatter: the photon is reflected in the direction of incidence
Lobe reflection: based on the orientation of the micro-facet on the surface. 
    The sigma_alpha parameter defines the standard deviation of the distribution of the micro-facets orientation
  • Measurement of Scintillator 1 (at FIU): [6]
  • Variation of transmission efficiency with polished finish: [7]
  • Variation of sigma_alpha with ground finish: [8]
  • Variation of lobe reflectivity with ground finish: [9]

Optical Tests

We want to implement different optical surface properties to different edges of the scintillator because the scintillators we have obtained from McNeal indicate edges have different finishing under the microscope. The flat face have better polishing but the thin edges have groove-like polishing. We want to implement this behavior in Geant4 and study it.

  • Test with a straight 3x15x600mm scintillator (Blue) surrounded by a large Hall filled with Air (Mother volume)
  • Place a 0.15mm thin bar (magenta), filled with Air to match the properties of Hall, running along one of the edges of the scintillator
  • Define a G4LogicalBorderSurface between the scintillator and the thin bar that have an optical surface property that is much different than that between scintillator and Hall
  • Pics [10] [11] [12]
  • The number of hits as a function of distance of SiPM from the source of optical photons is plotted for the above three configurations [13]

  • Place another 0.15mm thin bar (magenta), filled with Air, at the bottom edge of the scintillator and define another G4LogicalBorderSurface between these two volumes
  • Pic [14]
  • Hits vs distance [15]

Side Test: Thickness Variation

  • Simulate the impact of varying the thickness (x=first dimension) of the scintillator with two rough edges and the flat surfaces perfectly polished
  • The original scintillator was of the dimensions: 3x15x600 mm
  • Hits vs distance [16]

November 15, 2012:

Now, the photons produced are spread over a line in the x-axis (the dimension that is being varied) at (y = 0) and the size of the detector is (2*x_dim, 2*y_dim, 3mm) of the scintillator.

  • For a typical run (x = 5mm) [17]
  • Zoomed-in: [18]
  • If the number of photons generated inside the scintillator are scaled proportionally to the size of the x-dimension of the scintillator (x*5000) [19]
  • If the number of photons generated inside the scintillator is not scaled (num photons = 10000 always) [20]

November 29, 2012:

  • Plot and fit of attenuation length vs thickness for num photons = 10K: [21]
  • Varying \sigma _{\alpha } and the thickness [22]