Download presentation
Presentation is loading. Please wait.
Published byDominick Pearson Modified over 8 years ago
1
Reconstructing energy from HERD beam test data Zheng QUAN IHEP 3 rd HERD work shop Xi’an, 20 Jan.2016 0
2
What does ICCD raw data look like Images of 250 WLS fibers on 300×400 pixels CCDs Each pixel has a digitized gray value proportional to the light intensity 1
3
Crystal/WLSF array coordinate One frame of 40GeV electron run C2X crystals: center of crystal array, directly hit by beam particles, most probably deposited high energy ; Higher deposited energy means larger spot, needs more space to weaken or avoid overlapping. 2
4
3 Fitting or merging a. Fitting the shape of 1MIP spot with gaussian function: sigma~1.3 pixels, maybe the bin size is not small enough to make a precise fitting
5
4 b. Change of shape: high energy deposited, over saturated; c. Unsymmetrical shape.
6
Energy reconstruction step 1: Summing gray values(merging pixels) Merging area: approximated circle, radius = 5 pixels, center = COG(gray value weighted) Light intensity distribution on CCD doesn’t change with Light intensity; 5
7
Fitting: Lan×Gaus(left peak) + Lan×Gaus(right peak), MIP = MPV(right) – MPV(left) Energy reconstruction step 2: Calibrating low range ICCD using pion beam Event Selection: ~40% incident particles do not initiate showers 6 1MIP ≈ 0.03GeV
8
Energy reconstruction step 3: Fitting baseline Low range ICCD High range ICCD The pixels on high range ICCD have a larger fluctuation of response, indicating a wider baseline distribution 7
9
The MPV of baseline peak always deviates from 0 after baseline calibration. Fortunately, We can observe the baseline peak in every run. Low range baseline of A02 High range baseline of C24 8 Baseline
10
Energy reconstruction step 4: Image overlap(crosstalk) correction 9
11
Construct a Correction Matrix 10 Relative Crosstalk Value doesn’t change with light intensity. Very rough estimate Beam GeV
12
Energy reconstruction step 5: Calibrating high range ICCD Calibration factors: Great difference among CALO units… Expected factor: 0.025~0.033 11
13
Event Selection 1: Time dependent Good events: Before PS trigger, in a period of 3.3ms, no self trigger event comes 12
14
13
15
Event Selection 2: Shower shape 14
16
100GeV electron beam: contains 13% hadrons, 90% of hadronic events are excluded after selecting. Problem: What’s in this peak? 15
17
16
18
Electron energy Slope<1, due to energy deposition in carbon structure and energy leakage. 17
19
Energy resolution Electron 18
20
Angle reconstruction 19
21
20 Gap = 5mmGap = 2mm
22
Proton Event Selection: Enough valid hits to exclude MIP events; Shower maximum is contained; Shower starts at first few layers; 21
23
22
24
Problems and discussions: Longitudinal Shower Shape 23 Electron 100GeV
25
24 For layer 2, layer 3: Energy deposited in central units are higher than expected(C21,C22); For layer 4, layer 5: Energy deposited in CALO units around the central units are lower than expected.
26
25 C20C21C22
27
26 Problems and discussions: Strange excess on the right of main peak Main peak Right peak
28
27 Self trigger system: no peak on the right Self Trigger system VS reconstructed energy from ICCD Right peak Right peak disappears after excluding E>120GeV events So the right peak is probably from ICCD
29
28
30
Problems and discussions: charge detection C20: the first layer of central CALO Units 29 Quenching?
31
30 Problems and discussions: About energy resolution… Slightly saturated
32
31 With high rangeWithout high range
33
Summary 32
Similar presentations
© 2024 SlidePlayer.com Inc.
All rights reserved.