Transcript ppt
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Imaging crystals with TKR
Benoît Lott (CENBG)
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Imaging the crystals
The information from the TKR can help investigate the CAL response in different
ways by:
- mapping out the response as a function of the longitudinal and transverse positions
of the particle trajectory within the crystal: “images”.
cross-check of the light tapering (1D instead of 2D); non uniformity in response (see next
talk); possible local flaws.
- enabling the determination of the trigger efficiency for the CAL;
- checking the trajectories as determined from the CAL;
- determining that the particle will leave significant “direct” energy in
the photodiodes;
- determining valid hits for the calibration: on-orbit calibration with heavy ions.
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
CRYSTAL IMAGING:
Plot the average measured energy as a function of trajectory
position within the crystal.
The trajectory determined by the TKR must be extrapolated into
the CAL and only “valid” hits are kept.
Trajectories must intersect the top and bottom faces of a crystal
(no crossing of vertical sides).
At most 1 valid hit per layer.
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
The real question is how accurate is the extrapolation, as two adverse
effects come into play:
- the finite resolution of the tracker, the effect being amplified by the lever arm
between the Tkr layers and CAL;
- multiple scattering within the Tkr and CAL.
Tracker recon must provide the best estimate of the actual trajectory taking
two important facts into account:
1) the incident particle is a muon, not a gamma-ray;
The energy deposited in the CAL (~100 MeV) does not reflect the kinetic
energy.
jobOptions.txt:
TkrInitSvc.SetMinEnergy= 2000 MeV;
TkrIter.Members={};
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
2) the final trajectory (i.e leaving the tracker) is of interest here, not the
initial one.
The track reconstruction algorithm was designed to determine the initial
direction of photons. It makes use of the information available as close as
possible to the conversion point, to avoid the adverse effect of multiple
scattering.
For the present purpose, it is more sensible to use the information provided
by the bottom trays:
end-of-track parameters (CalValTools or xxx_recon.root)
Thanks to Bill and Leon for their help.
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Trajectory-extrapolation algorithm
Simple root macro, using modified (thanks, Anders) svac
ntuple:
VtxX0,…,VtxXDir,… (start of track)
Tkr1EndPos[0],…, Tkr1EndDir[0],…(end of track)
CsILength=326 mm
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
« lTkr – ltrue » distributions
position evaluated at mid-height of first CsI layer
blue: start of track red: end of track
ltrue= true (MC) position in firts cal layer
Benoît Lott
lTkr – ltrue (mm)
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Multiple scattering at play (1)
position evaluated at mid-height of first CsI layer for a pencil beam
l_Tkr_top
l
4 GeV
l_Tkr_bottom
ltrue
ltrue= actual (MC) position
Benoît Lott
position with respect to launch position (mm)
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Multiple scattering at play (2)
l_asym= actual (MC) position at first
calorimeter layer
Benoît Lott
The effect of multiple scattering in the
tracker can be partly corrected for
by using the bottom parameters.
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Position resolution
start of track
4 GeV
end of track
muon: sl=10.5 mm
start of track
end of track
Benoît Lott
500 MeV
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Muon energy distribution
The tracker can help efficiently discard low-energy muons, associated with
large multiple scattering.
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Position resolution
start of track
end of track
Lattest Svac tuple
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Deposited-energy distributions
4M evts: 219041 triggering evts, 303191 valid hits
valid hits
total
<E>=13.60 MeV
0.2% of valid hits have
no energy.
« doubles »
<E>=21.87 MeV
2.9% (8791/303191) of hits have more than 2 MeV in a neighboring log. For
these hits, the deposited energy in the « selected » log is much higher than
average: emission of d electrons !
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
width (mm)
Images
width (mm)
length (mm)
low
Benoît Lott
light tapering
high
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
“Valid events” for calibration
These values correspond to a period of 4000 s (55 Hz at the
trigger level).
top
tower 8
Benoît Lott
bottom
tower 9
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
How is it going to look like in real data?
asym
lTkr – lasym (mm)
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Gsi data: protons at 1.7 GeV
X layers
Y layers
Simulation results
The width of the distribution of the residues of a linear fit is proportional to the resolution s:
0.55 s for the outer layers, 0.83 s for fhe inner layers (simulations).
X layers: s = 10 mm
Y layers: s = 7 mm
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
To be continued…
A note summarizing these results will be written up with David Smith
(SLAC).
This work will be extended with David, to apply it to real data.
The GAM( Montpellier) group will join us on some particular studies
(comparison between trajectories determined by Tkr and CAL) .
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
MIPs in Photodiodes
Same pre-amps, amps, adc, daq
as for
Ganil, GSI, and CERN
(testbeam CsI stack at left…)
Smaller (top) scintillator:
2cm x 2cm
We also used:
i)
A 2cm long “CDE”
ii) A “naked” photodiode
(thanks to G. Bogaert for providing the latter)
Benoît Lott
GLAST LAT Project
Instrument Analysis Workshop, SLAC, September 27, 2004
Muons in a naked photodiode -data
In the CDE we said
1 MIP ~ 12 MeV ~ 1300 dc ~ 1.3 volts
For the photodiode without CsI,
we see ~250 dc
Zoooooom….
( 350-100=250 )
Benoît Lott