Analysis of BLC Calibration Data at GSFC

1
Analysis of BLC Calibration Data at GSFC

• S. Sturner, C. Shrader, G. Weidenspointer, B.
  Teegarden

2
BLC Data Analysis

• SPIHISTO used to create most spectra of BLC data.
• SPIHIST used to create simulated spectra from
  MGEANT.
• IDL software tools used for data comparison.
• Simple linear channel-to-energy conversion.
• Standard 1 keV/channel spectra created.
• Detector resolution not included in simulations.
• Singles and PSD spectra added.
• For SPIHISTO spectra components of multiples
  stored separately (energies not added).
• SPIHIST modified to produce samespectra as
  SPIHISTO.
• Approximate (3) deadtime correction included.

3
BLC Data Analysis (cont.)S. Sturner, C.
Shrader, G. Weidenspointer, B. Teegarden

• 6 dual-processor Linux boxes procured.
• Runs are computationally intensive
• Runs with isotropic beams typically were 24 hr
  long running 10 processors.

4
BLC - SPI Mass Model for MGEANT
5
Early Simulations

• Early simulations had marked differences with
  the BLC data, including
• Low data/sim photopeak ratios, 0.85
• Very narrow simulated Ge backscatter lines
• Very low simulated continuum(singlesPSD),
  particularly in the passive backscatter region
• ACS veto artifacts in simulated spectra
• gt Changes in GEANT physics, the simulated ACS
  veto, and the SPI site mass models

6
Modeling the SPI ACS

• Modelling has progressed through 4 stages.
• 1. Simple step function threshold with same value
  for all segments.
• 2. Error function threshold with same value for
  all segments.
• 3. Same as 2. Except measured values of ACS
  threshold parameters used.
• 4. Energy losses checked in up to 3 shield
  segments. (Previously only total shield energy
  loss was used.)

7
Effect of Varying the ACS Threshold
Black - Es 100 keV Orange - Es 200
keV Purple - Es 300 keV

• Varying the threshold can have a large effect
  on the continuum spectrum
• gt good calibration of the ACS is essential

8
Modeling the SPI ACS (cont.)

• Findings
• Error function veto adds significant wings to Ge
  backscatter line
• Changing from fixed 100 keV threshold to true
  thresholds for each block had little effect.
• Adding the extra information on energy losses in
  up to 3 blocks has little effect on the final
  result.
• Changing the mean threshold on all blocks from
  100 to 200 or 300 keV has a significant effect on
  the compton continuum.

9
The Mass Model

• 535 different volumes, many repeated several
  times.
• Most inputs came from drawings supplied by CESR,
  CNES, MPE.
• Cryostat model from Euclid CAD files provided by
  Saclay.
• Many refinements over the past several months.
• New information on Ge crystal masses included.
• Total mass in mass model today is 1100 kg.
• Almost all missing mass is outside shield
  (probably structural).
• Mass model continues to be updated.

10
SPI Mass Model
11
Cryostat
12
Detector Capsule
Mave 950.91 g
13
Detector Modelling

• Single fixed model for detector capsule (repeated
  19 times).
• Ge crystal area set to mean area of the 19
  detectors.
• 2 variation in area
• Ge crystal length adjusted to make mass avg.
  detector mass.
• Mass variation 5.
• Detector mass and area variations to be included
  in response function by using table of correction
  factors.

14
Mass Budget
15
The 8 m Collimator Model
16
Effects of the Collimator on the Continuum
Isotropic
Isotropic
16o beam, no collimator

• Isotropic runs have 3e9 triggered photons

17
Collimator Effects

• Earlier simulations did not take collimator
  effects properly into account.
• Leakage and scattering from collimator have
  significant effects on SPI spectra.
• Scattering from collimator support structure is
  important.
• Scattering from walls of room is important.

18
Latest Simulation Results - 137Cs, 8 meters, no
mask
SinglesPSD
Multiples
All
____ BLC Data ____ MGEANT Simulations

• 3 dead time applied to simulations

19
Latest Simulation Results - 60Co, 8 Meters,no mask
SinglesPSD
Multiples
All
____ BLC Data ____ MGEANT Simulations

• 3 dead time applied to simulations

20
Latest Simulation Results - 137Cs, 1.8 Meters,mask
SinglesPSD
Multiples
All
____ BLC Data ____ MGEANT Simulations

• 3 dead time applied to simulations

21
Latest Simulation Results - 54Mn, 8.3 m, ISDC
Processing, Mask
Doubles Triples
____ BLC Data ____ MGEANT Simulation
22
Latest Simulation Results - 137Cs, 8.3 m, ISDC
Processing, Mask
Doubles Triples
____ BLC Data ____ MGEANT Simulation
23
Latest Simulation Results Efficiency and
Effective Area
24
Latest Simulation Results - Photo-Peak
Ratio(data/simulation)
25
Accelerator Target Branching Ratios
26
137Cs - 8m, no mask photopeak (661.6 keV) all
events ? 0, ? 0
27
137Cs - 8m, no mask all multiples ? 0, ? 0
28
137Cs - 125m, mask photopeak (661.6 keV) all
events ? 0, ? 0