View
4
Download
0
Category
Preview:
Citation preview
INSTITUTE OF NATURAL AND APPLIED SCIENCES
UNIVERSITY OF CUKUROVA
Bs.C THESIS
Emine GURPINAR
2008 BEAM TEST ANALYSIS OF CASTOR CALORIMETER AND PEDESTALSTABILITY OF HCAL DURING GLOBAL RUNS
DEPARTMENT OF PHYSICS
ADANA, 2009
INSTITUTE OF NATURAL AND APPLIED SCIENCESUNIVERSITY OF CUKUROVA
2008 BEAM TEST ANALYSIS OF CASTOR CALORIMETER AND PEDESTAL
STABILITY OF HCAL DURING GLOBAL RUNS
By Emine GURPINAR
A THESIS OF MASTER OF SCIENCEDEPARTMENT OF PHYSICS
We certify that the thesis titled above was reviewed and approved for the award of degreeof the Master of Science by the board of jury on ...........................
Signature.............................Prof.Dr. Gulsen ONENGUTSUPERVISOR
Signature.............................Prof.Dr. Eda ESKUTMEMBER
Signature.............................Assist.Prof.Dr. Nuri EMRAHOGLUMEMBER
This Ph.D. Thesis is performed in Department of Physics of Institute of Natural andApplied Sciences of Cukurova UniversityRegistration Number:
Prof.Dr. Aziz ERTUNCDirector
The Institute of Natural and Applied Sciences
Not: The usage of the presented specific declarations, tables, figures and photographs either inthis thesis or in any other reference without citation is subject to “The Law of Arts andIntellectual Products” numbered 5846 of Turkish Republic.
CUKUROVA UNIVERSITESIFEN BILIMLERI ENSTITUSU
CASTOR KALORIMETRESININ 2008 HUZME TESTI ANALIZLERI VE
HCAL’IN GENEL VERI ALIMI SIRASINDAKI PEDESTAL KARARLILIGI
Emine GURPINAR
MASTER TEZIFIZIK ANABILIM DALI
Bu tez ......................... tarihinde asagıdaki juri uyeleri tarafından oybirligi/oycoklugu ilekabul edilmistir.
Imza.............................Prof.Dr. Gulsen ONENGUTDANISMAN
Imza.............................Prof.Dr. Eda ESKUTUYE
Imza.............................Yrd.Doc.Dr. Nuri EMRAHOGLUUYE
Bu tez Enstitumuz Fizik Anabilim Dalında hazırlanmıstır.Kod No:
Prof.Dr. Aziz ERTUNCEnstitu MuduruImza ve Muhur
Not: Bu tezde kullanılan ozgun ve baska kaynaktan yapılan bildirislerin, cizelge, sekil vefotografların kaynak gosterilmeden kullanımı, 5846 sayılı Fikir ve Sanat Eserleri Kanunundakihukumlere tabidir.
ABSTRACT
MSc THESIS
2008 BEAM TEST ANALYSIS OF CASTOR CALORIMETER AND
PEDESTAL STABILITY OF HCAL DURING GLOBAL RUNS
Emine GURPINAR
DEPARTMENT OF PHYSICSINSTITUTE OF NATURAL AND APPLIED SCIENCES
UNIVERSITY OF CUKUROVA
Supervisor: Prof.Dr. Gulsen ONENGUTYear: 2009, Pages: 66Jury: Prof.Dr. Gulsen ONENGUT
Prof.Dr. Eda ESKUTAssist.Prof.Dr. Nuri EMRAHOGLU
Centauro and Strange Object Research (CASTOR) which is a tungsten/quartzCerenkov sampling calorimeter, is installed in the very forward region of the CompactMuon Solenoid (CMS) experiment at the Large Hadron Collider (LHC). It will cover thepseudo rapidity range 5.1< η <6.6 and will be placed 14.38 m away from the interactionpoint. In order to test the performance of the CASTOR calorimeter, CASTOR prototypeIV was tested at CERN/SPS H2 beam line in 2008. The energy linearity and resolution,as well as the spatial resolution of the prototype of electromagnetic and hadronic showersare studied with E= 10-200 GeV electrons, E= 20-350 GeV pions, and E= 50-150 GeVmuons in beam tests. In my analysis X-surface scan is studied using E=50 GeV pions andE=100 GeV electrons.
Hadronic Calorimeter (HCAL) which is a subsystem of CMS experiment at the LHC,consists of four subdetectors, Hadronic Barrel (HB), Hadronic Endcap (HE), HadronicOuter (HO) and Hadronic Forward (HF). It will measure hadronic particles’ directionsand energies. In HCAL, pedestal is important to determine the muon energy deposits andfor quality of calibration of HCAL. Also in my analysis, I studied pedestal stability of allsubdetectors of HCAL by using data taken during CRAFT.
Key Words: CASTOR, HCAL, CMS, LHC.
I
OZ
MASTER TEZI
CASTOR KALORIMETRESININ 2008 HUZME TESTI ANALIZLERI VE
HCAL’IN GENEL VERI ALIMI SIRASINDAKI PEDESTAL KARARLILIGI
Emine GURPINAR
CUKUROVA UNIVERSITESIFEN BILIMLERI ENSTITUSU
FIZIK ANABILIM DALI
Danısman: Prof.Dr. Gulsen ONENGUTYıl: 2009, Sayfa: 66Juri: Prof.Dr. Gulsen ONENGUT
Prof.Dr. Eda ESKUTYrd.Doc.Dr. Nuri EMRAHOGLU
Centauro ve Acayip Cisim Arastırmaları (CASTOR), Buyuk Hadron Carpıstırıcısı(LHC)’deki Compact Muon Solenoid (CMS) deneyinin ileri bolgesine yerlestirilecekolan Cerenkov ısıması ilkesine dayanan tungsten-kuartz ornekleme kalorimetresidir.Etkilesme noktasından 14.38 m uzaklıga konulacaktır ve 5.1< |η | <6.6 pseudorapiditearalıgını kaplayacaktır. CASTOR kalorimetresinin performansını test etmek amacıyla2008 yılında CASTOR’un IV. prototipinin CERN/SPS H2 deney alanında huzme testiyapılmıstır. Huzme testinde 50-150 GeV’lik muonlar, 20-300 GeV pionlar ve 10-200 GeV elektronlarla prototipin hem uzaysal cozunurlugu hem de enerji linarite vecozunurlugu arastırılmıstır.
LHC’de CMS deneyinin alt sistemi olan Hadronik Kalorimetre (HCAL) Hadronikfıcı (HB), Hadronik kapak (HE), dıs kısım (HO), ve ileri kalorimetre (HF) gibi 4 altdedektor icermektedir. Bunlar hadronik parcacıkların izlerini ve enerjilerini olceceklerdir.HCAL’de pedestal, muon enerjisini ve kalibrasyonun kalitesini belirledigi icin onemlidir.Analizimde ayrıca Cosmic Run At Four Tesla (CRAFT) sırasında alınan veriler kulla-narak HCAL’in tum altdedektorlerinin pedestal kararlılıgı arastırılmıstır.
Anahtar Kelimeler: CASTOR, HCAL, CMS, LHC.
II
ACKNOWLEDGEMENTS
I first and foremost thank Gulsen Onengut who is my supervisor. I am grateful to her
for guidance throughout the variety of learning experiences and her advice and comments.
Also it is a pleasure for me to work with her. She is not only helpful but moreover
understanding.
I would like to thank Mustafa Numan Bakirci for his help and his support. I am very
grateful to him because he showed me the way during my research at CERN.
Specifically, I would like to thank Pawel De Barbaro and Dmitry Vishnevsky for their
supervision and teaching in pedestal study.
I also would like thank to team members of the High Energy Physics group of
Cukurova University for their helps.
Special thanks are due to Turkish Atomic Energy Authority (TAEK) who sponsored
me during the time I have spent at CERN.
Last, I would like to send many thanks my family and say that I love them very much.
III
CONTENTS PAGE
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I
OZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . III
CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VI
LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII
LIST OF SYMBOLS AND ABBREVATIONS . . . . . . . . . . . . . . . . . . . XII
1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Standard Model of the Particle Physics . . . . . . . . . . . . . . . . . . . 1
1.1.1 Fundamental Forces and Interactions . . . . . . . . . . . . . . . 2
1.2 Shortcomings of SM . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Physics Beyond Standard Model . . . . . . . . . . . . . . . . . . . . . . 5
1.3.1 Higgs Physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3.2 Supersymmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 THE LARGE HADRON COLLIDER AND CMS DETECTOR . . . . . . . . 7
2.1 The Large Hadron Collider (LHC) . . . . . . . . . . . . . . . . . . . . . 7
2.2 Compact Muon Solenoid (CMS) . . . . . . . . . . . . . . . . . . . . . . 8
2.2.1 The Tracker System . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.2 The Electromagnetic Calorimeter (ECAL) . . . . . . . . . . . . . 11
2.2.3 The Hadronic Calorimeter (HCAL) . . . . . . . . . . . . . . . . 12
2.2.4 The Magnet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.2.5 The Muon System . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3 Calorimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.4 Particle Shower . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4.1 Electromagnetic Shower . . . . . . . . . . . . . . . . . . . . . . 16
IV
2.4.2 Hadronic Shower . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3 THE CASTOR CALORIMETER . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2 Detector components . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.1 The Tugsten Plates . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.2 The Quartz Plates . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.3 The Photo Multiplier Tubes (PMTs) and Bases . . . . . . . . . . 25
3.2.4 Air-Core the Light Guides . . . . . . . . . . . . . . . . . . . . . 25
4 ANALYSIS AND RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2 Beam Test of CASTOR Prototype-IV . . . . . . . . . . . . . . . . . . . 27
4.3 The H2 Beam Line and Particle Identification . . . . . . . . . . . . . . . 28
4.4 X-Surface Scan Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4.4.1 X-Surface Scan with electron runs . . . . . . . . . . . . . . . . . 30
4.4.2 X-Surface Scan with pion runs . . . . . . . . . . . . . . . . . . . 33
5 HCAL PEDESTAL STABILITY STUDIES IN CMS . . . . . . . . . . . . . . 41
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.2 Stability of HCAL Pedestals during CRAFT . . . . . . . . . . . . . . . . 41
5.3 Pedestal Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.4 Pedestal Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
5.5 Stability of the HE and HB pedestals during CRAFT . . . . . . . . . . . 43
5.5.1 Stability of the HCAL pedestal average . . . . . . . . . . . . . . 44
5.5.2 Stability of individual channel pedestals . . . . . . . . . . . . . . 45
5.6 Stability of HF pedestal during CRAFT . . . . . . . . . . . . . . . . . . 52
5.7 Stability of HO pedestal during CRAFT . . . . . . . . . . . . . . . . . . 54
6 CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
RESUME . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
V
LIST OF TABLES PAGE
Table 1.1. Three generations of quarks and leptons (Bawa, 2007) . . . . . . . . . 3
Table 1.2. The four forces and their quanta, the gauge bosons (Bawa, 2007) . . . . 3
Table 2.1. Parameters of the CMS superconducting solenoid . . . . . . . . . . . . 14
Table 4.1. List of electron runs used in analysis of X-surface scan during CAS-
TOR test beam 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Table 4.2. List of pion runs used in analysis of X-surface scan during CASTOR
test beam 2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Table 5.1. List of runs used in analysis of pedestal stability during CRAFT (Bar-
baro, 2009). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Table 5.2. The linearized ADC values for the lowest 30 channels are listed in table
(Barbaro, 2009). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Table 5.3. List of 23 individual channels in HB and HE subdetectors with large
run-to-run pedestal RMS. (Barbaro, 2009) . . . . . . . . . . . . . . . . 52
Table 5.4. List of ten individual channels in HF with large run-to-run pedestal
RMS. (Barbaro, 2009) . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Table 5.5. List of seventeen individual channels in HO with large run-to-run
pedestal RMS. (Barbaro, 2009) . . . . . . . . . . . . . . . . . . . . . . 61
VI
LIST OF FIGURES PAGE
Figure 2.1. Overview of CERN’s accelarator layout (Gumus, 2008) . . . . . . . . . 8
Figure 2.2. The CMS detector (The CMS Collaboration 2007) . . . . . . . . . . . 9
Figure 2.3. The CMS tracker (Gumus, 2008) . . . . . . . . . . . . . . . . . . . . . 10
Figure 2.4. Layout of pixel detectors in the CMS tracker (CMS TDR, 2006) . . . . 11
Figure 2.5. Longitudinal view of the silicon strip tracker (one quarter) (Moortgat,
2004). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Figure 2.6. The CMS ECAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Figure 2.7. The CMS Muon system is depicted (Gumus, 2008). . . . . . . . . . . . 15
Figure 2.8. Schematic of electromagnetic shower development in matter (Virdee,
1998). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 2.9. Energy loss of electrons and in lead as a function of their energy (Fab-
jan, 1987). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Figure 2.10. Schematic of development of hadronic shower (Virdee,1998) . . . . . . 19
Figure 3.1. The CASTOR Calorimeter . . . . . . . . . . . . . . . . . . . . . . . . 21
Figure 3.2. The details of the CASTOR Calorimeter . . . . . . . . . . . . . . . . . 21
Figure 3.3. Longitudinal cross sectional view of 1/2 of CASTOR calorimeter
(Panagiotou, 2007) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Figure 3.4. Position of CASTOR Calorimeter in CMS (Panagiotou, 2007) . . . . . 22
Figure 3.5. Example of a tungsten plate used for the assembly (Basegmez, 2008) . 23
Figure 3.6. Construction drawing of the EM plate (a), Construction drawing of the
HAD tungsten plate.(CASTOR EDR, 2007) (b). . . . . . . . . . . . . . 23
Figure 3.7. Example of a quartz plate used for the assembly (Basegmez, 2008) . . . 24
Figure 3.8. Construction drawings of the semi-octant quartz plate for the EM sec-
tion (a), Construction drawings of the semi-octant quartz plate for the
HAD section (CASTOR EDR, 2007) (b). . . . . . . . . . . . . . . . . 24
Figure 3.9. Assembly of light guides onto W/Q sampling units (Ayhan, 2008) . . . 25
Figure 3.10. Cross section of the EM light-guide with the PMT and base housing (a),
Cross section of the HAD light-guide with the PMT and base housing
(CASTOR EDR, 2007) (b). . . . . . . . . . . . . . . . . . . . . . . . . 26
VII
Figure 3.11. PMT base and its cable (a), Air-core light guide and reflecting foil cov-
ering inside(b) (CASTOR EDR, 2007). . . . . . . . . . . . . . . . . . 26
Figure 4.1. Photograph of fully instrumented CASTOR octant prototype at the
CERN/SPS H2 line (Ayhan, 2009). . . . . . . . . . . . . . . . . . . . 28
Figure 4.2. Schematic drawing of CASTOR prototype-IV (Aslanoglou et al., 2008). 28
Figure 4.3. Projection of the EM (black lines) and HAD (red lines) sections onto a
45o plane.(Aslanoglou et al., 2008). . . . . . . . . . . . . . . . . . . . 29
Figure 4.4. Schematic figure of the H2 beam line of the SPS at CERN (Aslanoglou
et al., 2008). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Figure 4.5. WC’s profiles for electron run, r47601 . . . . . . . . . . . . . . . . . . 31
Figure 4.6. Beam profile projected onto the front face of the calorimeter using the
hits distribution from the WC-E, before (left) and after (right) circle cut
of 2 mm radius (r47601). . . . . . . . . . . . . . . . . . . . . . . . . . 31
Figure 4.7. ADC distribution of scintillator counters S1, S2 and S4. The data are
fitted by a Gausian and events within 3 sigma are selected as single
particle events. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Figure 4.8. Typical example of the ADC distribution of the muon veto counter, for
r47601 with a electron beam with E=100 GeV. . . . . . . . . . . . . . 32
Figure 4.9. Signal distribution of the sum of the signals in EM1, EM3 and HAD1
channels after applied all cuts (a), Signal distribution of the sum of the
signals in EM1, EM3 and HAD1 channels after applied all cuts (b). . . 33
Figure 4.10. Response of the semi-octant of the EM section (Saleve side) as the
beam scans the front face of the calorimeter (a). The derivative of the
response with respect to x, indicating the width (σ=1.903 mm) of the
EM shower (b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Figure 4.11. Response of the semi-octant of the EM section (Jura side) as the beam
scans the front face of the calorimeter (a). The derivative of the re-
sponse with respect to x, indicating the width (σ=1.601 mm) of the EM
shower (b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Figure 4.12. X-scan along the face of the prototype for 100 GeV electrons. . . . . . 35
Figure 4.13. WC’s profiles for pion run, r48495 . . . . . . . . . . . . . . . . . . . . 36
VIII
Figure 4.14. Beam profile for r48495 without cut (a). Beam profile for r48495
WC(2mm) cut (b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Figure 4.15. For r48495 ADC distribution of scintillator counters S1, S2 and S4.
The data are fitted by a Gausian and events within 3 sigma are selected
as single particle events. . . . . . . . . . . . . . . . . . . . . . . . . . 37
Figure 4.16. Typical example of the ADC distribution of the muon veto counter, for
r48495 with a pion beam with E=80 GeV. . . . . . . . . . . . . . . . . 38
Figure 4.17. Energy spectrum for an pion beam of E=80, before and after applying
all cuts (a), Signal distribution after applied all cuts (b). . . . . . . . . . 38
Figure 4.18. X-scan along the face of prototype for 80 GeV pions (a). The derivative
of the response with respect to X , the width of the HAD shower is given
by (σ=6.081 mm) (b). . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Figure 4.19. X-scan along the face of the prototype for 80 GeV pions. . . . . . . . . 39
Figure 5.1. For a single run (r67647), distribution of mean pedestal values for 2583
HB channels (Barbaro, 2009). . . . . . . . . . . . . . . . . . . . . . . 44
Figure 5.2. The average of all HB pedestal (upper plot) and all HE pedestal (lower
plot) with respect to reference values plotted during CRAFT (Barbaro,
2009). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Figure 5.3. Pedestal stability for two HB channels (eta=1,phi=31,depth=1 on the
left and eta=1, phi=41,depth=1 on the right). Upper plot: pedestal mean
plotted. Lower plot: distribution of pedestal means (Barbaro, 2009). . . 46
Figure 5.4. Distribution of difference between individual HCAL channel pedestal
means (calculated for independently for each run) and reference
pedestal value for individual HCAL channels. Data is shown separately
for HB channels (upper plot) and HE channels (lower plot) (Barbaro,
2009). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Figure 5.5. Distribution of run-to-run RMS of pedestal means for individual HCAL
Barrel (upper plot) and HCAL Endcap (lower plot) channels (Barbaro,
2009). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
IX
Figure 5.6. Distribution of run-to-run RMS of pedestal means for individual HB
HE channels correspond to the mean values calculated using 100k
muon triggers (Barbaro, 2009). . . . . . . . . . . . . . . . . . . . . . 49
Figure 5.7. Pedestal mean versus run number (upper plot) pedestal mean distri-
bution for individual HCAL channel and pedestal mean distributions
(lower plots) for HE channel HE (eta=26, phi=23,depth=1) (Barbaro,
2009). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Figure 5.8. Pedestal mean versus run number (upper plots) and pedestal mean dis-
tributions (lower plots) for two individual HCAL channels, HE (eta=-
20, phi=61, depth=1) and HB (eta=-2, phi=66, depth=1). In case of
HE channel (left plots), a pedestal shift of ∼ 0.060 ADC/Time Slice
took place during CRAFT. Shift of 0.060 ADC/TS is equivalent of shift
of 100 MeV/channel. In case of HB channel shown (right plots), two
shifts took place: one shift upward, the second shift, downward (Bar-
baro, 2009). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
Figure 5.9. Pedestal mean versus run number for individual HCAL channels, HB
(eta=2, phi=23, depth=1) and HE (eta=18, phi=8, depth=1). For these
two channels, pedestal shifts were apparent throughout several runs
(long term drifts, as opposed to shift) (Barbaro, 2009). . . . . . . . . . 50
Figure 5.10. Pedestal mean versus run number for individual HCAL channels, HB
(eta=4, phi=36, depth=1) and HB (eta=6, phi=4, depth=1). For these
two channels, pedestal shifts were apparent throughout several runs
(long term drifts, as opposed to shift) (Barbaro, 2009). . . . . . . . . . 51
Figure 5.11. Pedestal mean versus run number for individual HCAL channels, HE
(eta=20, phi=32, depth=2) and HB (eta=18, phi=35, depth=1). For
these two channels, pedestal shifts were apparent throughout several
runs (long term drifts, as opposed to shift) (Barbaro, 2009). . . . . . . 51
Figure 5.12. Pedestal average for HF versus run number. Average based on global
runs, using 100k events/run (Barbaro, 2009). . . . . . . . . . . . . . . 53
X
Figure 5.13. Run-to-run RMS of HF pedestals.Pedestals are based on global runs,
using 100k events/run. Ten channels (out of a total of 1725) have RMS
larger than 0.005 ADC counts (Barbaro, 2009). . . . . . . . . . . . . . 53
Figure 5.14. Pedestal average for HF (eta=-31, phi=7, depth=1) channel versus run
number (a), Pedestal average for HF (eta=29, phi=71, depth=1) channel
versus run number (b). . . . . . . . . . . . . . . . . . . . . . . . . . . 54
Figure 5.15. Pedestal average for HF (eta=-33, phi=39, depth=2) channel versus run
number (a), Pedestal average for HF (eta=41, phi=47, depth=2) channel
versus run number (b) . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Figure 5.16. Pedestal average for HF (eta=39, phi=43, depth=2) channel versus run
number (a), Pedestal average for HF (eta=30, phi=17, depth=2) channel
versus run number (b). . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Figure 5.17. Pedestal average for HF (eta=29, phi=5, depth=1) channel versus run
number (a). Pedestal average for HF (eta=-36, phi=7, depth=1) channel
versus run number (b). . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Figure 5.18. Pedestal average for HO versus run number. Average based on global
runs, using 100k events/run (Barbaro, 2009). . . . . . . . . . . . . . . 57
Figure 5.19. Run-to-run RMS of HF pedestal. Pedestal are based on global runs,
using 100k events/run. Five channels (out of a total of 2154) have RMS
larger than 0.010 ADC counts (Barbaro, 2009). . . . . . . . . . . . . . 58
Figure 5.20. HO channel (eta=-4, phi=69, depth=4) with unstable pedestal. For
block runs (r67128-r67541) pedestal setting for this channel was incor-
rectly loaded (a). HO channel (eta=15, phi=24, depth=4) with unstable
pedestal. Pedestal value changes by up to 2 ADC counts (b). . . . . . . 59
Figure 5.21. HO channel (eta=-5, phi=34, depth,4) with unstable pedestal (a), HO
channel (eta=-11, phi=21, depth=4) with unstable pedestal (b). . . . . . 59
Figure 5.22. HO channel (eta=-11, phi=36, depth=4) with unstable pedestal (a), HO
channel (eta=-6, phi=8, depth=4) with unstable pedestal (b). . . . . . . 60
Figure 5.23. HO channel (eta=-1, phi=26, depth=4) with unstable pedestal (a), HO
channel (eta=-2, phi=50, depth=4) with unstable pedestal (b). . . . . . . 60
XI
LIST OF SYMBOLS AND ABBREVATIONS
SM Standard Model
SUSY Supersymmetry
LHC Large Hadron Collider
CMS Compact Muon Solenoid
CASTOR Centauro and Strange Object Research Detector
CERN European Nuclear Research Laboratory
GeV Giga electron Volt
MeV Million electron Volt
TeV Billion electron Volt
PS Proton Synchrotron
SPS Super Proton Sychrotron
ECAL Electromagnetic Calorimeter
HCAL Hadronic Calorimeter
HB Hadron Barel
HF Hadron Forward
HE Hadron Endcap
HO Hadron Outer
EM Electromagnetic Channel
HAD Hadronic Channel
SU Sampling Unit
RU Readout Unit
CRAFT Cosmic Run At Four Tesla
QIE Charge Integrator and Encoder
XR Radiation Length
Z Atomic Number
RM Moliere Radius
XII
1. INTRODUCTION Emine GURPINAR
1. INTRODUCTION
High energy physics searches the elementary constituents of matter and the interac-
tions between them. It concentrates on subatomic particles. These contain atomic con-
stituents like electrons, protons, and neutrons. Protons and neutrons are really combined
particles which are made up of quarks. More details about quarks can be found at sec-
tion 1.1. Particles like photons, neutrinos, and muons, as well as a wide range of exotic
particles are produced by radiative and scattering processes.
All the particles and their interactions observed until now can almost be described en-
tirely by a quantum field theory called Standard Model (SM). The Standard Model is the
common theory of quarks and leptons and their electromagnetic, weak and strong interac-
tions. But it is not a complete theory because it has many important unanswered questions.
Because of this, beyond the Standard model physics research is needed. Beyond the SM
physics will be studied of the experiments A Torodial LHC Apparatus (ATLAS), Compact
Muon Solenoid (CMS), A Large Ion Collider Experiment (ALICE) and A Large Hadron
Collider Beauty (LHC-B) on the Large Hadron Collider (LHC) ring at European Nuclear
Research Laboratory (CERN).
1.1 Standard Model of the Particle Physics
SM is a theory of three of the four known fundamental interactions and of the fun-
damental particles that take part in these interactions :strong, weak and electromagnetic
forces. SM defines these fundamental forces using mediating gauge bosons which are
photons for the electromagnetic, gluons for the strong, W± and Z bosons for the weak
force. The theory does not include the effects of gravitational interactions. They are neg-
ligible in high-energy physics. The model also contains fundamental particles which are
the constituents of matter (Bawa, 2007).
In 1963, the first step towards Standard Model was Sheldon Glashow’s discovery com-
bine the electromagnetic and weak interactions. Steven Weinberg and Abdus Salam added
the Higgs mechanism into Glashow’s electroweak theory, in 1967.
In 1979, Glashow, Salam, and Weinberg won Nobel Prize in Physics because of dis-
covering the electroweak theory. The W and Z bosons were discovered experimentally in
1
1. INTRODUCTION Emine GURPINAR
1981, and their masses were found to be as the Standard Model predicted.
In 1964, Gell-Mann and Zweig independently proposed that the three quarks and anti-
quarks combined in many different ways in accordance with the rules of symmetry. They
were named up-quark, down-quark and strange-quark. In 1974, a new particle was dis-
covered at Stanford Linear Accelerator Center. It was J/Psi which was interpieted as the
bound state of a new quark and its antiquark. The new fourth quark was named charm.
Kobayashi, Cabbibo and Maskawa (CKM) added bottom and top quarks to previously
discovered the four quarks. All the mesons and baryons till now are made up of vari-
ous combinations of these quarks (Bawa, 2007). In the SM, the elementary particles are
fermions. They have spin-1/2 and obey the to the Pauli Exclusion Principle. There are
12 fermions (6 quarks, 6 leptons) and their corresponding antiparticles. They are catego-
rized according to how they interact. Table 1.1. gives the properties of quarks and leptons.
The quarks carry color charge and hence, interact via the strong force. The leptons are
light particles and they don’t carry color charge. The three neutrinos do not carry electric
charge either, so they interact only via the weak nuclear force. Particles that have zero or
integer spin are called bosons. Hadrons contain either a quark and an antiquark (mesons)
or three quarks (baryons). The familiar proton and the neutron are the two baryons having
the smallest mass. Quarks also carry electric charge and weak isospin. Hence they interact
with other fermions both electromagnetically and through the weak nuclear interaction.
In conclusion, SM was the victory of particle physics of the 1970’s. Today, SM is a
well established theory applicable over a wide range of conditions.
1.1.1 Fundamental Forces and Interactions
In nature there are four fundamental forces. They are Gravity, the Weak Force, Elec-
tromagnetic and the Strong Force (see Table 1.2.). Gravitation is the weakest of the four
interactions. But it is important for macroscopic objects. Gravity and electromagnetism
control our world. The strong force adheres quarks together and holds nucleons in nuclei.
The weak is one of the four fundamental forces, it is responsible for the radioactive decay
of unstable nuclei and for interactions of neutrinos and other leptons with matter. These
forces arise from the exchange of specific particles called gauge bosons, the quanta of the
2
1. INTRODUCTION Emine GURPINAR
Table 1.1. Three generations of quarks and leptons (Bawa, 2007)
Quarks (u) up-quark (c) charm-quark (t) top-quark Charge=+2/3
Mass 0.0015-0.004 GeV 1.3 GeV 171.4 GeV
Quarks (d) down-quark (s) stranges-quark (b) bottom-quark Charge=-1/3
Mass 0.004-0.008 GeV 0.08-0.13 GeV 4 GeV
Leptons (νe) elec-neutrino (νµ ) muon-neutrino (ντ ) tau-neutrino Charge=0
Mass 0 0 0
Leptons (e) electron (µ) muon (τ) tau Charge=-1
Mass 0.0015-0.004 GeV 1.3 GeV 171.4 GeV
Table 1.2. The four forces and their quanta, the gauge bosons (Bawa, 2007)
Force Relative Strength Gauge boson Mass Charges Spin
Strong 1 Gluon (g) 0 0 1
Electromagnetic 1/137 Photon (γ) 0 0 1
Weak 10−9 W±, Z 86,97 ± 1, 0 1
Gravity 10−38 Graviton (G) 0 0 2
force field.
1.2 Shortcomings of SM
Despite the success of the SM which has answers to many of the questions related to
the structure and stability of matter with its six types of quarks, six types of leptons, and
four forces most physicists are assured that it is only a low energy effective theory and
has shortcomings which are listed below;
It does not include the fourth force in nature: gravity. This is believed to be mediated
by the exchange of gravitons, no one has ever been able to incorporate gravity into the
SM. So the theory is surely incomplete.
3
1. INTRODUCTION Emine GURPINAR
Another problem of this model is that one has to assume the existence of different
forces and their carriers. Electricity and magnetism were used to be thought of as two
forces, however now we know they are different aspects of the same (electro-magnetic)
force.
Many astrophysical observations point towards the existence of non-relativistic, neu-
tral, non-baryonic dark matter. Since neutrinos are not massive enough to explain the
observations, a new candidate for this cold dark matter is needed.
In the SM, neutrinos are massless. But recent experiments have provided evidence for
very small but non-zero neutrino masses. A number of possibilities for accommodating
neutrino masses in the SM have been discussed in the literature.
In particle physics, the hierarchy problem is the question why the weak force is 1032
times stronger than gravity. Both of these forces include constants of nature, Fermi’s con-
stant for the weak force and Newton’s constant for gravity. In addition if the SM is used
to calculate the quantum corrections to Fermi’s constant, it appears that Fermi’s constant
is unnaturally large and should be closer to Newton’s constant unless there is a delicate
cancelation between the bare value of Fermi’s constant and the quantum corrections to it.
The most popular theory to solve the hierarchy problem is supersymmetry. This explains
how a tiny Higgs mass can be produced from quantum corrections. Supersymmetry re-
moves the power-law divergences of the radiative corrections to the Higgs mass; however,
there is no understanding of why the Higgs mass is so small in the first place.
Another question is why the Higgs boson is so much lighter than the Planck mass
which is the unit of mass in the system of natural units: one would expect that the large
quantum contributions to the square of the Higgs boson mass would inevitably make
the mass huge, comparable to the scale at which new physics appears, unless there is an
incredible fine-tuning cancelation between the quadratic radiative corrections and the bare
mass.
The SM will be extended by beyond the SM physics studies that will be probed in
LHC experiments.
4
1. INTRODUCTION Emine GURPINAR
1.3 Physics Beyond Standard Model
1.3.1 Higgs Physics
The masses of fermions and gauge bosons fall under the SM through the Higgs mech-
anism, which is satisfactory technically but is not understood physically. We do not know
what nature really does to give mass to particles, nor what experimental clues will lead us
to nature’s solution. Understanding Higgs physics is necessary in order to complete the
Standard Model, and to learn how to extend it and improve its foundations.
In the SM, the elementary particles acquire their mass through the Higgs mechanism.
This mechanism foresees the existence of a Higgs boson. Higgs mass is the only yet
unknown parameter of the SM.
One of the basic physics programs of LHC will be to confirm or disprove the existence
of Higgs bosons. This will require the study of several final states, depending on the mass
of the Higgs boson. Assuming that at least one type of Higgs boson is discovered, the next
step will be, the detailed study of its properties such as mass, decay width, production
rates, branching ratios, spin-parity, couplings etc.
1.3.2 Supersymmetry
Supersymmetry (SUSY) is a symmetry between bosons (one unit of spin) and
fermions (half a unit of spin). In a supersymmetric theories, every fundamental fermion
has a bosonic superpartner and vice versa. If SUSY exists at the electroweak scale (around
246 GeV), its discovery should be direct at the LHC. If SUSY exists close to the TeV en-
ergy scale, it allows the solution of two problems in particle physics. One is the hierarchy
problem. On the other is the unification of the weak interactions, the strong interactions
and electromagnetism. Gluino and squark production which are strongly produced with
large cross-sections, control the SUSY cross-section. Then gluinos and squarks decay
though a series of steps into the Lightest SUSY Particles (LSP). These decay chains lead
to a variety of signatures involving multiple jets, leptons, photons, W and Z bosons, and
missing energy. The combination of a large production cross-section and distinctive sig-
natures makes it easy to separate SUSY from the Standard Model background. So, the
5
1. INTRODUCTION Emine GURPINAR
main dispute is not only to discover SUSY, but to separate the many SUSY processes that
occur and to measure the masses and other properties of the SUSY particles. The back-
grounds from other SUSY events are greater than the reducible SM backgrounds in most
cases (Bawa, 2007).
6
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR
2.1 The Large Hadron Collider (LHC)
The Large Hadron Collider (LHC) which is the world’s highest-energy particle accel-
erator, was built by the European Organization for Nuclear Research (CERN). LHC is a
large proton-proton collider located at CERN, near Geneva, Switzerland and it is included
in a circular tunnel 27 kilometres in circumference at a depth ranging from 50 to 150 me-
ters. LHC aimed to collide opposing particle beams, protons at a center of mass energy
of 14 TeV. Experiments on the LHC are believed strongly to help scientist to answer the
existence of mysterious questions like what gives mass to a particle?, what is the nature
of dark matter?, do extra dimensions exist? etc.
Protons started circulating in the beam line in September 2008 but the operation was
halted because of a faulty connection between two superconducting bending magnets. It
was not possible to collide the protons for physics studies. The LHC will be working
again in fall 2009.
Proton bunches will collide at 40 MHz frequency, or every 25 ns. The accelerator
luminosity can be obtained by Eqn (1.1). The maximum design luminosity of the LHC
is 1034 cm−2 s−1. The luminosity of an accelerator which collides bunches containing n1
and n2 particles at a frequency f is given by
L =f n1n2
4πσ1σ2(2.1)
where σ1 and σ2 characterize the Gaussian transverse beam profiles. Protons are injected
to different accelerators to raise their energy to 7 TeV. Firstly they will be injected into
Linear Accelerator (Linac) generating 50 MeV protons. There the protons will be acceler-
ated to 1.4 GeV by a booster and injected into the Proton Synchrotron (PS) increasing their
energy to 25 GeV and then protons will be accelerated by the Super Proton Sychrotron
(SPS) up to 450 GeV. Finally they will be injected into the main ring to accelerate up
to 7 TeV. At every 25 ns proton bunches will collide at the four intersection points. The
schematic view of LHC machine can be seen in Figure 2.1.
LHC has four experiments. They are the Compact Muon Solenoid (CMS), A Large
Torodial LHC Apparatus (ATLAS), Large Hadron Collider b-quark experiment (LHC-b)
7
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
Figure 2.1. Overview of CERN’s accelarator layout (Gumus, 2008)
and A Large Ion Collider Experiment (ALICE). The CMS and ATLAS are multipurpose
experiments. They have the same scientific aims but the technical solution and design
of detector magnet system are different. The LHC-b is a specialized experiment which
will be investigating the differences between matter and antimatter by studying a type of
particle called the ’beauty quark’. The ALICE will study the quark-gluon plasma in heavy
ion collisions.
2.2 Compact Muon Solenoid (CMS)
The CMS experiment is a general-purpose detector. Approximately three thousand
scientists and engineers from scientific institutes all over the world formed a collaboration
to construct this detector. CMS experiment will investigate new physics at TeV scale,
8
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
discover the Higgs boson and look for evidence of physics beyond the SM, SUSY or
extra dimensions. Having extraordinary electronic eyes, the CMS can be regarded as a
particle hunter because it has;
1. a large selonoid magnet,
2. a good muon system,
3. a high quality central tracking,
4. the best electromagnetic calorimeter with good energy resolution.
5. a hadron calorimetry with enough energy resolution and good hermiticity
The CMS detector consists of subdetectors which are a silicon tracker, an electromag-
netic calorimeter and a hadron calorimeter, surrounded by a solenoid which generates a
strong magnetic field of 4 T, in order to measure the tracks, energy and momentum of
photons, electrons, muons and the other particles over a large energy range and at high
luminosity. An overall picture of the CMS can be seen in Figure 2.2.
Figure 2.2. The CMS detector (The CMS Collaboration 2007)
9
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
Figure 2.3. The CMS tracker (Gumus, 2008)
2.2.1 The Tracker System
The charged particles tracks are bent by the magnetic field and their charge and mo-
mentum can be measured from the curvature of particle tracks by segmented silicon sen-
sors (pixel and strips). The central tracking system is placed close to the interaction re-
gion, covering the pseudorapidity region |η |<2.5. The CMS Tracker consists of a silicon
pixel detector and a silicon microstrip detector. Three pixel layers and ten silicon strip
layers will be installed in the barrel, and in each endcap, two pixel layers, three inner and
nine outer forward disks of silicon detectors will be placed. The outer radius of the CMS
Tracker extends up to 107-110 cm. The total length is approximately 540 cm. Figure 2.3.
shows the tracking system.
The Pixel Detector consists of 3 barrel layers with 2 endcap disks on each side of
them. The 3 barrel layers are located at mean radii of 4.4 cm, 7.3 cm and 10.2 cm, and
have a length of 53 cm. The pixel detector consists of 4.4 million pixels in a square
measuring 150 µm per side. It provides a spatial resolution of 15 µm. Figure 2.4. shows
the CMS pixel detector
The Silicon Strip Detector consists of four inner layers (Tracker Inner Barrel, TIB)
and six outer barrel layers (Tracker Outer Barrel, TOB). On each side of the TIB, there
are three mini disks (Tracker Inner Disks, TID). Each endcap region consists of nine disks
(Tracker End-Cap, TEC). The silicon strip detector consists of 25000 strips. Figure 2.5.
10
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
Figure 2.4. Layout of pixel detectors in the CMS tracker (CMS TDR, 2006)
shows longitudinal view of the silicon strip tracker (one quarter).
Figure 2.5. Longitudinal view of the silicon strip tracker (one quarter) (Moortgat, 2004).
2.2.2 The Electromagnetic Calorimeter (ECAL)
The ECAL is designed to measure the energies of electrons and photons. It is a ho-
mogenous calorimeter, made of lead tungstate (PbWO4) crystals. This material was cho-
sen because of its high density (8.2 g/cm3) leading to a short radiation length (X0=0.89
cm) and a small Moliere radius (RM=2.19 cm).
The ECAL consists of two parts: The electromagnetic barrel (EB) and electromag-
netic endcap (EE). Inner radius of the barrel section (EB) is 129 cm, covers the region
|η | <1.48. The EB has 36 supermodules and each supermodule contains four modules.
11
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
Front face area of each lead tungstate crystal is 22×22 mm2. Its length is 230 mm (25.8
X0). The endcaps (EE) is at a distance of 314 cm from the vertex, covering the pseudo-
rapidity range of 1.5< |η | <3.0. The Endcaps has 7300 crystals which have a front face
cross section of 28.6 ×28.6 mm2 and a length of 220 mm, about 24.7 X0.
Preshower detector (ES) covers 1.6 < |η | < 2.6, is in front of EE. The aim of
Preshower detector (ES) to reject π0s which decay into two closely separated photons.
Vacuum Photo Triodes (VPT) in the barrel and Avalanche Photo Diodes (APD) in the
endcaps are used as photodetectors to read out the signals at the back of the crystals. The
layout of the CMS ECAL is shown in Figure 2.6.
Figure 2.6. The CMS ECAL
2.2.3 The Hadronic Calorimeter (HCAL)
HCAL which will measure quark, gluon and neutrino directions and energies by
means of measuring the energy and direction of particle jets and of the missing trans-
verse energy flow, is subsystem of the CMS detector. The measurement of the missing
transverse energy will light the way for new particles and phenomena.
The HCAL consist of four subdetectors which are Hadronic Barrel (HB), Hadronic
Endcap (HE), Hadronic Outer (HO) and Hadronic Forward (HF). HB covers the η
range -1.4< |η | < 1.4 and the HCAL endcaps (HE) cover the pseudorapidity range
1.3< |η | <3.0. They are the sampling calorimeters which consist of plastic scintillators
12
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
as active material inserted between copper absorber plates, which are placed between the
ECAL and the magnet. Light collected from the scintillators are read out by the Hybrid
Photo Diodes (HPD).
The HB is not deep enough to contain a hadronic shower fully. Thus, the HO comes
in to play to catch the tails of a hadronic shower. The HO contains scintillators with a
thickness of 10 mm, is physically located inside the barrel muon system. It covers the
region -1.26< |η | <1.26. It is divided into 5 sections along η , called rings -2, -1, 0, 1,
and 2.
The HF calorimeters, the last subdetector of HCAL, are placed∼ 11 m away from the
interaction point. The HF calorimeter improves the measurement of the missing trans-
verse energy and allows very forward jets to be identified and reconstructed. The HF
calorimeter is located at 3.0< |η | <5.0. It uses the quartz fibers as the active medium.
There are two HFs on both sides of CMS. Each HF has an active radius of 1.4 m and
consists of iron absorbers, fibers embedded into the absorbers, and phototubes. Fibers
have two different lengths to differentiate between shower processes. Longer fibers will
provide light from electromagnetic and hadronic showers in the absorber. Shorter fibers
will only see the hadronic showers. The long and the short fibers are read out by separate
PMTs.
2.2.4 The Magnet
CMS has a large superconducting solenoid which will produce a 4T magnetic field.
The charge/mass ratio of particles are determined from the curved track that they follow
in the magnetic field. The solenoid has a radius of 5.9 m. To achieve the reconstruction of
1 TeV muons with a 15% PT (transverse momentum) resolution is the goal of the magnet
(Gumus, 2008). The connection between PT and the magnetic field is given by
PT = 0.3BR (2.2)
where R is the radius of the curvature of the particle in m and B is the magnetic field in
tesla. Parameters of the CMS’s superconducting selonoid are listed inTable 2.1.
13
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
Table 2.1. Parameters of the CMS superconducting solenoid
Field 4T
Inner Bore 5.9m
Length 12.9m
Number of Turns 2168
Current 19.5kA
Stored energy 2.7GJ
Hoop stress 64atm
2.2.5 The Muon System
The purpose of the muon system of the CMS detector are muon identification,
muon trigger, and muon momentum measurement. Muons transverse through the whole
detector, leaving minimal ionization behind. So, they are easy to identify in the detector.
They also play an important role in interesting physics. The muon system uses three
types of detector which are Drift Tubes (DT) in the barrel region to provide a precise
track measurement in the bending plane, Cathode Strip Chambers (CSC) in the endcap
region to provide high precision measurements in the presence of a large and varying
magnetic field, and Resistive Plate Chambers (RPC) in both the barrel and endcap to
identify muons and measure their momentum. A layout of the muon system is given in
Figure 2.7.
2.3 Calorimetry
Calorimetry is an experimental method to carry out energy measurements. A
calorimeter is a block of matter with useful instrumentation. It measures the energy of
incident particles. Its typical feature is that the resolution of the energy measurement
improves with energy. In addition to energy measurement, calorimeters can also supply
position measurement and particle identification.
14
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
Figure 2.7. The CMS Muon system is depicted (Gumus, 2008).
In a calorimeter the signal generation mechanism is:
1. The incident particle interacts inside the calorimeter volume and starts a shower of
secondary particles,
2. The particle transfers all its energy to this shower. The shower develops and its
products generate signals by passing through the sensitive material of the calorime-
ter.
Structure and dimensions of a shower depend on the type of incoming particle and
the detector material. The operation principle of calorimeters depends on the dE/dx
technique. The signal which is generated by scintillation or ionization, can be light, for
calorimeters with scintillators, or ionization charge, for those with gaseous or semicon-
ductive sensitive material. Calorimeters with a different principle of operation are those
composed of lead glass or quartz fibers. For these, the signal which is generated, is based
on the Cherenkov effect. There are two basic types of showers. Electromagnetic showers
are produced by a particle that interacts primarily via the electromagnetic force, usually
a photon or electron. They serve to measure the energy of incident electrons, positrons
and photons. Hadronic showers are produced by hadrons i.e. nucleons and other particles
made of quarks, and proceed mostly via the strong nuclear force. Depending on their
construction, they are also classified as homogeneous and sampling calorimeters.
In a homogeneous calorimeter, all detector volume is used as an absorber. Shower
production and development and signal generation and collection happens at the same
material. It has the best possible energy resolution, but is expensive and is used only as
an electromagnetic calorimeter.
15
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
In sampling calorimeters, there are two different materials; passive and active (sensi-
tive) material.
• Passive material; the absorber, where the shower develops. Copper, steel, lead or
the other dense metals (Fe, Pb, U) are used as absorber
• Active (Sensitive) material; where the signal is produced and collected.Plastic scin-
tillators, silicon detectors, lead glass, liquid, gas are used as active material.
There are a lot of disadvantages of a sampling calorimeter. Such as, it has worse
energy resolution because of statistical fluctuations. Their signals are significantly large
compared to homogeneous ones and only part of the particle energy is deposited in detec-
tor layers.
On the other hand there are a lot of advantages too. For example it is very economical.
The materials can be optimized for their purpose and calorimeter can be constructed in a
compact design.
2.4 Particle Shower
In particle physics, a shower is a cascade of secondary particles produced on account
of a high-energy particle interacting with dense matter. Firstly the incoming particle in-
teracts then multiple new particles are produced with lesser energy; each of these then
interacts in the same way, a process carries on until many thousands, millions, or even
billions of low-energy particles are produced. Then these are stopped in the matter and
absorbed.
There are two basic types of showers: electromagnetic showers, hadronic showers
2.4.1 Electromagnetic Shower
Electrons and positrons lose energy via bremsstrahlung, and photons through pair
production for energies above 1 GeV. The secondary particles produced interact through
the same processes and a shower develops inside the calorimeter (Mavromanolakis, 2003).
Figure 2.8. shows the schematic electromagnetic shower development.
16
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
Figure 2.8. Schematic of electromagnetic shower development in matter (Virdee, 1998).
At low energies, electrons lose their energy mainly through collisions with the atoms
and molecules of the material thus giving rise to ionization and thermal excitation. The
main process for the energy loss of photons is Compton scattering and the photoelectric
effect (Fabjan, 2003).
Figure 2.9. Energy loss of electrons and in lead as a function of their energy (Fabjan, 1987).
The total energy loss dE/dX of electrons can be written as
17
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
dEdX
=−α(E)− EX0
(2.3)
where α(E) is the ionization energy loss given by the Bethe-Bloch formula. It depends
logarithmically on energy. X0 is the radiation length. Particle multiplication and ioniza-
tion energy loss are competing processes in showers. The energy at which an electron’s
rates of losing energy from ionization and from bremsstrahlung become equal is called
critical energy. The critical energy is an important parameter which characterizes the ma-
terial of a calorimeter. Ec is approximately 800/Z+1.2 and is expressed Ec = 610/Z +1.24
MeV for liquids or solids and as Ec = 710/Z +0.92 MeV for gases, where Z is the atomic
number of material.
X0 is also a characteristic quantity of electromagnetic showers. It is defined as the
mean distance that an electron traverses to lose (1-1/e) of its energy by bremmsstrahlung.
It is roughly X0 ≈ 180A/Z2(in gr/cm2), where A is atomic mass in g/mole and Z is the
atomic number of the material.
X0 =716.4.A
Z(Z +1).ln(287/√
Z)(2.4)
The moliere radius (RM) is another characteristic parameter of the material giving the
scale of the transverse dimension of the fully contained electromagnetic showers initiated
by an incident high energy electron or photon. It is calculated as;
RM = 0.0265X0(Z +1.2) (2.5)
where Z is the atomic number. Moliere radius is useful in experimental particle physics for
the design of calorimeters. If the moliere radius is small we obtain good shower position
resolution, and better shower separation due to a smaller degree of shower overlaps.
2.4.2 Hadronic Shower
Hadronic showers are produced by hadrons such as nucleons and other particles made
of quarks, and proceed mostly through the strong nuclear force. The hadronic showering
process is dominated by a succession of inelastic hadronic interactions. At high energy,
it is qualified by multiparticle production and particle emission activating from nuclear
18
2. THE LARGE HADRON COLLIDER AND CMS DETECTOR Emine GURPINAR
decay of excited nuclei because of relatively ordinary generation of π0’s, there is also an
electromagnetic component present in a hadronic shower. Figure 2.10. shows the hadronic
shower.
Figure 2.10. Schematic of development of hadronic shower (Virdee,1998)
The transverse development of a hadronic shower is determined by the mean trans-
verse momentum of the produced particles, which is roughly 〈pt〉 ' 350 MeV/c so
hadronic showers tend to be more laterally spreadout than electromagnetic ones. Hadronic
showers also take much longer to develop and are much deeper than the electromagnetic
showers.
Mainly pions and nucleons are produced as secondary particles, multiplicity which
increases logarithmically with energy per collision and approximately 1/3 of the produced
pions are π0’s that immediately after decay into photons and generate electromagnetic
showers. A fraction of the energy of hadrons is lost via interactions of with the nuclei.
λI , Interaction length is the quantity of hadronic showers, which governs the longitu-
dinal development of hadronic showers. It is essentially energy-independent. It is defined
as the mean distance that a hadron transverses to lose (1−1/e) of its energy.
The transverse shower development does not scale with interaction length however
one interaction length contains almost 95% of the shower energy in a cylinder of radius
RM (Mavromanolakis, 2003).
19
3. THE CASTOR CALORIMETER Emine GURPINAR
3. THE CASTOR CALORIMETER
3.1 Introduction
The Centauro and Strange Object Research (CASTOR) calorimeter which will search
the Centauro-type events in heavy-ion collisions, is one of the forward detectors of CMS.
The CASTOR calorimeter (see Figure 3.1.) has been a part of the CMS detector since
June 2009. It will search the electromagnetic and hadronic contents of the interactions by
measuring the energies of the particles. The main goal of the CASTOR is to study some
anomalous events like centauros and stranglets. It will also make possible to study the
diffractive and low-x physics.
CASTOR which is a tungsten/quartz Cerenkov electromagnetic and hadronic sam-
pling calorimeter, an octagonal cylinder in shape. Its outer diameter is 36 cm and its inner
diameter is 3.8 cm (to contain the beam pipe). Castor will cover the region 5.2≤ |η | ≤6.4. The η-range covered will be 5.3 < |η | < 6.46 for the electromagnetic section with
99% containment and 5.25 < |η | < 6.31 for the hadronic section with 95% containment
(CASTOR EDR, 2007). It is divided into 16 sectors in azimuth. Also it is divided longi-
tudinally into 14 sections, 2 sections for the EM part and 12 sections for the HAD parts
in depth. The electromagnetic section consists of 2x16 channels and its depth is 2x10 =
20 radiation lengths, Xo. For hadrons its depth corresponds to 0.77 interaction lengths,
λI . The hadronic section has 12x16 channels and its depth is 12x0.77 = 9.24 λI . The
calorimeter has total depth 10 λI . CASTOR calorimeter consists of successive layers
of tungsten plates (W) as absorber and fused silica quartz (Q) plates as active medium.
Thicknesses of W plates and Q-plates are 5mm and 2mm respectively for hadronic sec-
tion the W and Q plates have thicknesses of 10mm and 4mm larger, than the W plates
and Q plates of EM, tilted at 450 with respect to the direction of the impinging particles
due to capture maximum of Cerenkov light in the quartz. Cerenkov light is produced by
the passage of particles through the medium and is collected in sections of 5 W/Q then
focused by air-core light guides onto the PMTs.
The CASTOR Calorimeter has 224 (16×14) subdivisions in total. The Cerenkov
light produced in each one is collected and focused by air-core light guides onto the
corresponding PMTs. There are 5 tungsten/quartz layers called Sampling Units (SU) in
20
3. THE CASTOR CALORIMETER Emine GURPINAR
Figure 3.1. The CASTOR Calorimeter
Figure 3.2. The details of the CASTOR Calorimeter
both the EM and HAD sections, each read by a Readout Unit (RU) (CASTOR EDR,
2007). This calorimeter design and components are shown in Figure 3.2. The front view
of the calorimeter design is shown in Figure 3.3., which includes the four octants forming
half of the calorimeter. The calorimeter will be placed at 14.385 m from the interaction
point and it is constructed in two semi-circular sections of 4-octants each, in order to
be positioned around the fixed beam pipe. The location of the calorimeter in the CMS
forward region is shown Figure 3.4..
21
3. THE CASTOR CALORIMETER Emine GURPINAR
Figure 3.3. Longitudinal cross sectional view of 1/2 of CASTOR calorimeter (Panagiotou, 2007)
Figure 3.4. Position of CASTOR Calorimeter in CMS (Panagiotou, 2007)
3.2 Detector Components
3.2.1 The Tugsten Plates
The tungsten plates contain 97% W, 1.3% Fe and 1.7% Ni. The density is 18.5±0.2
g/cm3. For the EM and HAD sections the thicknesses are5 mm, 10 mm respectively,
but the effective thickness increases to 7.07 mm for the EM section, 14.14 for the HAD
section at 45o inclination. Example of the tungsten plates can be seen in Figure 3.5.. The
22
3. THE CASTOR CALORIMETER Emine GURPINAR
Figure 3.5. Example of a tungsten plate used for the assembly (Basegmez, 2008)
(a) (b)
Figure 3.6. Construction drawing of the EM tungsten plate (a), Construction drawing of the HADtungsten plate (CASTOR EDR, 2007) (b).
CERN Metrology Section measured the geometry, dimensions, consistency and density
of samples of the W-plates (CASTOR EDR, 2007). Figure 3.6. (a-b) shows the octant
trapezoidal shape and dimensions of the tungsten plates.
3.2.2 The Quartz Plates
The quartz plates which were designed in semi-octant geometry, define the 16-fold
azimuthal segmentation of the calorimeter. Figure 3.7. shows one of the quartz plates.
The two semi-octants Right, Left (R, L) are positioned side by side along their vertical
23
3. THE CASTOR CALORIMETER Emine GURPINAR
Figure 3.7. Example of a quartz plate used for the assembly (Basegmez, 2008)
(a) (b)
Figure 3.8. Construction drawings of the semi-octant quartz plate for the EM section (a), Con-struction drawings of the semi-octant quartz plate for the HAD section (CASTOR EDR, 2007)(b).
side. The semi-octant construction designs of the EM and HAD Q-plates are shown in
figure Figure 3.8. (a-b). They are optically separated by thin Al foils. The light guides are
placed on top of the 5 tungsten/quartz layers, RU, in both the EM and HAD sections.
Tyvek paper is used as a diffuser and it protects the surface of polished quartz plates
from contact with the tungsten ones.Assembly of light guides onto W/Q sampling units
are shown in figure Figure 3.9..
24
3. THE CASTOR CALORIMETER Emine GURPINAR
Figure 3.9. Assembly of light guides onto W/Q sampling units (Ayhan, 2008)
3.2.3 The Photo Multiplier Tubes (PMTs) and Bases
The Cherenkov light produced in the quartz plates is collected and focused by air-core
light guides onto the corresponding PMTs. We use two different types of PMTs for the
light collection. These are (i) a Hamamatsu R5380Q, (ii) a radiation-hard multi-mesh
FEU-187. The air core light guides transmit the Cherenkov light to the light-reading
devices.
3.2.4 Air-Core The Light Guides
In the quartz plates, the Cherenkov light is produced and collected then it is transmit-
ted to the PMTs through air-core light guides. The efficiency of light transmission and
its dependence on the light-source position are crucial parameters, which characterize the
light guide and significantly affect the performance of the calorimeter. The design and
25
3. THE CASTOR CALORIMETER Emine GURPINAR
(a) (b)
Figure 3.10. Cross section of the EM light-guide with the PMT and base housing (a), Crosssection of the HAD light-guide with the PMT and base housing (CASTOR EDR, 2007) (b).
(a) (b)
Figure 3.11. PMT base and its cable (CASTOR EDR, 2007)(a), Air-core light guide and reflect-ing foil covering inside(b) (CASTOR EDR, 2007).
dimensions of the air-core light guides for the electromagnetic and hadronic sections are
obtained from Monte Carlo simulations. Figure 3.10. (a,b) show the construction draw-
ings of the EM and HAD light guides, respectively.
For the transmission of the Cherenkov light, the inside walls of the light guides are
covered with reflecting foil. An air-core light guide with reflecting foil covering inside
can be seen Figure 3.11.
26
4. ANALYSIS AND RESULTS Emine GURPINAR
4. ANALYSIS AND RESULTS
4.1 INTRODUCTION
In this chapter, I present the analysis results of the CASTOR calorimeter test beam of
prototype IV data collected at CERN in the summer of 2008. My analysis consists of two
parts:I studied the X-surface scan by using E=50 GeV pion and E=100 GeV electrons.
Before the beam test, the performance studies of the three prototypes of CASTOR
calorimeter were conducted with the data collected at the H2 CERN/SPS beam line, Beam
Test 2003 of CASTOR Prototype I, Beam Test 2004 of CASTOR Prototype II, Beam Test
2007 of CASTOR Prototype III. References [1,2,3] present detailed beam-test studies of
the energy and position performances of those three prototypes of the CASTOR detector.
4.2 Beam Test of CASTOR Prototype-IV
The beam test of prototype IV was performed in the H2 line at CERN Super Proton
Synchrotron (SPS). More details about H2 beam line can be found at section 4.3.
The energy linearity, resolution and uniformity, as well as the surface scan were stud-
ied for electrons, pions and muons of various energies. On the other hand channel inter-
calibration, gain variation and pedestal were studied by using muons.
The prototype IV was a full-length octant which consisted of EM and HAD sections
with a total of 28 readout-units (RUs). W plates, as absorber, and Q plates as active
medium were installed in one octant of Castor prototype-IV. Figure 4.1. shows a photo-
graph of the one octant of prototype IV. Light is produced by the passage of relativistic
particles via Q medium and collected by 5 W/Q layers. Then it is focused by air-core light
guides onto the PMTs.
Schematic drawing of the beam test with 28 RUs indicated are shown in Figure 4.2.
The beam comes from the left impinging on the EM sections. Figure 4.3. shows the
two semi-octants of the EM (black) and of the HAD (red) sections onto a plane at 450
with respect to the beam axis. Due to the different sizes of the W/Q-plates there isn’t a
complete overlap of the two sections. The beam profile for each point was subdivided
into a number of smaller parts, each of diameter less than 1 cm, so more impact points
27
4. ANALYSIS AND RESULTS Emine GURPINAR
Figure 4.1. Photograph of fully instrumented CASTOR octant prototype at the CERN/SPS H2
line (Ayhan, 2009).
could finally be used (see Figure 4.3.).
Figure 4.2. Schematic drawing of Castor prototype-IV (Aslanoglou et al., 2008).
4.3 The H2 Beam Line and Particle Identification
The beam test of prototype IV was performed in the H2 line. Figure 4.4. shows a
schematic figure of the H2 beam line of the SPS at CERN. The beam line provides 400
GeV/c protons from the Super Proton Synchrotron (SPS). A production target (T2) was
located at 590.9 m to produce different particles such as pions and electrons. Protons
28
4. ANALYSIS AND RESULTS Emine GURPINAR
Figure 4.3. Projection of the EM (black lines) and HAD (red lines) sections onto a 45o
plane.(Aslanoglou et al., 2008).
produce various particles of momentum between 10 and 350 GeV/c. Useful maximum
momentum for pions is 350 GeV/c and for electrons 100 GeV/c. To produce very low
energy (VLE) particles with momentum range between 1 and 9 GeV/c, another target
T22 was placed 97 m upstream of the calorimeters. For momentum selection and particle
identification, the produced particles were forced to follow a dog-leg path (see in Figure
4.4.).
Figure 4.4. Schematic figure of the H2 beam line of the SPS at CERN (Aslanoglou et al., 2008).
In the beam line, there were eight wire chambers (WC1 through WC3 and WCA
through WCE) but WCA, WCB, WCC, WCD, WCE were used for CASTOR calorime-
29
4. ANALYSIS AND RESULTS Emine GURPINAR
ter. They provided the tracking information of particles.Four scintillation counters (S1
through S4) were used for triggering. The beam trigger typically consisted of the co-
incidence of S1,S2,S4 counters which defined a 4×4 cm2 area on the front face of the
calorimeter.Muon Veto Front (MVF), Muon Veto Back (MVB), and Muon Veto Wall
(MVW) counters were used. The MVF and MVB were (80×80 cm2) scintillation coun-
ters. They were placed well behind the calorimeters. MVB was located behind the
calorimeter and was used in order to tag muons which penetrated the whole length of
the apparatus.
To identify particles in the VLE mode, there were two Cherenkov counters (CK2 and
CK3), two time-of-flight counters (TOF1 and TOF2) in the beam line but they were not
used for the CASTOR calorimeter.
4.4 X-Surface Scan Analysis
4.4.1 X-Surface Scan with electron runs
Electron beams at 100 GeV energy with various X-positions were used to study X-
surface scan for electron runs. I have analyzed 16 electron runs. Table 4.1. shows the list
of electron runs used in this study.
For r47601, scintillator-wire-chamber information are shown in Figure 4.5. WCA is
not used in the test beam and there are some missing events on the WCD. The distribution
of the particles in the scintillator-wire-chamber E was used, which was the closest working
wire chamber to the prototype.
A beam cut was applied to the beam profile for all runs, the beam for each point
was subdivided into a number of smaller parts, each of diameter 2 cm, so more impact
points could finally be used. Figure 4.6. shows the distribution of the wire chamber’s hits
position which was exploited in order to select a well focused beam of a small radius
(2 mm), compared to the very wide beam spot initially obtained without any cut, for
an incoming electron beam of E = 100 GeV, hitting the right semi-octant (Saleve side),
before and after the applied cut (r47601).
Also some spatial cuts were applied for all runs. These are;
Scintillator cuts (SC1, SC2, SC4) were applied to tag single particle events. The
30
4. ANALYSIS AND RESULTS Emine GURPINAR
Figure 4.5. WC’s profiles for electron run, r47601
(a) (b)
Figure 4.6. Beam profile projected onto the front face of the calorimeter using the hits distributionfrom the WC-E, before (left) and after (right) circle cut of 2 mm radius (r47601).
trigger counters are polystyrene scintillators of 1 cm thickness, hence electrons and pions
behave as Minimum Ionizing Particles (MIP). The peak position and width of an MIP was
fitted by a Gaussian distribution and only events with ADC counts within 3 sigma were
selected as single particle events (see Figure 4.7.).
Gausian Distribution simplified means that if you measure the same thing many times
and make a vertical bar for each measurement value you can get, the bar graph will have
a bell shape centered around the actual measurement. The width compared to the height
of the bell shaped graph can be described with a statistical measurement called standard
deviation or sigma. Basically, this is a measurement of how much the measurements vary
around the actual value.
31
4. ANALYSIS AND RESULTS Emine GURPINAR
(a) (b) (c)
Figure 4.7. ADC distribution of scintillator counters S1, S2 and S4. The data are fitted by aGausian and events within 3 sigma are selected as single particle events.
Three sigma is 3* the standard deviation, which statistically mean that 99.73% of the
time a measurement is made it will be within 3*the standard deviation of the actual value.
It is thus a way to compare how good the measurement method is.
Muon cut muons were rejected using the muon veto counter (MVB) placed behind
the CASTOR prototype. Figure 4.8. shows an example of the ADC distribution of the
muon counter, as well as the Gaussian fit which was used to obtain the mean and width
of the spectrum. Events with ADC counts within 3 sigma were treated as muons and they
were rejected.
Figure 4.8. Typical example of the ADC distribution of the muon veto counter, for r47601 witha electron beam with E=100 GeV.
32
4. ANALYSIS AND RESULTS Emine GURPINAR
(a) (b)
Figure 4.9. Signal distribution of the sum of the signals in EM1, EM3 and HAD1 channels afterapplied all cuts (a), Signal distribution of the sum of the signals in EM1, EM3 and HAD1 channelsafter applied all cuts (b).
Electromagnetic Fraction (FEM)=Mean value of total electromagnetic channels
(EM)/mean value of total channels (EM+HAD)> 0.95 used for rejection of pions. Figure
4.9. exhibit the signal distribution for the electron beam of E = 100 GeV before and after
all the cuts.
The results of the X-scan analysis are shown in Figure 4.10. Figure 4.11. and Figure
4.12. They show the response of the EM two semi-octants (Saleve and Jura side) as the
beam impact point moves across the front face of the calorimeter. The sigmoid nature of
the response curve is evident. The X-derivative of the response is calculated, giving the
width of the electromagnetic shower. We observe that one standard deviation amounts to
σEM=1.903 mm for the (Saleve side) EM shower and σEM=1.601 for the (Jura side) EM
shower.
4.4.2 X-Surface Scan with pion runs
Pion beams at 80 GeV energy were used to study X-surface scan with pion runs.
The main goals of X-surface scan with pion runs are to determine the width of the HAD
33
4. ANALYSIS AND RESULTS Emine GURPINAR
(a) (b)
Figure 4.10. Response of the semi-octant of the EM section (Saleve side) as the beam scans the
front face of the calorimeter (a). The derivative of the response with respect to x, indicating the
width (σ= 1.903 mm) of the EM shower (b).
(a) (b)
Figure 4.11. Response of the semi-octant of the EM section (Jura side) as the beam scans the
front face of the calorimeter (a). The derivative of the response with respect to x, indicating the
width (σ=1.601 mm) of the EM shower (b).
34
4. ANALYSIS AND RESULTS Emine GURPINAR
Figure 4.12. x-scan along the face of the prototype for 100 GeV electrons.
Table 4.1. List of electron runs used in analysis of X-surface scan during CASTOR test beam2008
Run x-position y-position47599 16.82 65.2847600 22.04 65.3247601 26.97 65.3247602 31.99 65.3647605 12.1 65.7847614 11.84 65.6547626 7.3 65.747636 2.5 65.5847650 -3.14 65.6247658 -8.02 65.6447663 -12.65 65.6647684 -23.27 65.5947700 -27.89 65.3747711 -32.99 65.2447722 -38.31 65.3347728 -42.89 65.0547746 -47.93 64.94
35
4. ANALYSIS AND RESULTS Emine GURPINAR
shower profile. To study the X-surface scan of the hadronic section of the calorimeter,
a central point in the Saleve side sector was exposed to beams of various position at 80
GeV. I have analyzed 16 pion runs. List of pion runs used in this study is shown Table
4.2.
Figure 4.13. shows scintillator-wire-chamber information for r48495 Beam Profiles
of 80 GeV pions impinging on center of the Saleve side. For all runs, a beam cut was
applied to the beam profile. The beam for each point was subdivided into a number of
smaller parts, each of diameter is 2 cm (see Figure 4.14.).
Figure 4.13. WC’s profiles for pion run, r48495
In X-surface scan with pion runs analysis, some cuts were used to select the pion
events such as Scintillator cut, muon cut, FEM cut. Figure 4.15. shows distribution of
scintillator counters S1, S2 and S4. The data are fitted by a Gausian and events within
mean + 3sigma are selected as single particle events. Muon cut is applied for the the
muons to be rejected using the MVB. Figure 4.16. shows an example of the ADC dis-
tribution of the muon counter, as well as the Gaussian fit which was used to obtain the
mean and sigma of the spectrum and last FEM<0.95 was used for rejection of electron.
As can be seen in Figure 4.17., the quality of the spactra was significantly improved after
applying all the cuts, although a significant fraction of the events was finally filtered out
reducing the available statistics.
The surface response of the prototype calorimeter to pions was obtained from hadronic
semi-octant sectors, by moving the beam along the X-direction. Figure 4.18. (a) (b) show
the X-scan for pions of 80 GeV energy (right) and the derivative of this response with
36
4. ANALYSIS AND RESULTS Emine GURPINAR
(a) (b)
Figure 4.14. Beam profile for r48495 without cut (a). Beam profile for r48495 WC(2mm) cut(b).
(a) (b) (c)
Figure 4.15. For r48495 ADC distribution of scintillator counters S1, S2 and S4. The data arefitted by a Gausian and events within 3 sigma are selected as single particle events.
37
4. ANALYSIS AND RESULTS Emine GURPINAR
Figure 4.16. Typical example of the ADC distribution of the muon veto counter, for r48495 witha pion beam with E=80 GeV.
(a) (b)
Figure 4.17. Energy spectrum for an pion beam of E=80, before and after applying all cuts (a),Signal distribution after applied all cuts (b).
38
4. ANALYSIS AND RESULTS Emine GURPINAR
respect to x (left).
Figure 4.18. X-scan along the face of prototype for 80 GeV pions (a). The derivative of the
response with respect to X , the width of the HAD shower is given by (σ=6.081 mm) (b).
Figure 4.19. X-scan along the face of the prototype for 80 GeV pions.
39
4. ANALYSIS AND RESULTS Emine GURPINAR
Table 4.2. List of pion runs used in analysis of X-surface scan during CASTOR test beam 2008
Run x-position y-position48481 18.28 65.2848482 12.95 65.5648483 7.497 65.7248484 2.892 65.7548487 -1.269 65.6448488 -12.07 65.6448489 -16.89 65.7148490 -21.91 65.6148491 -27.22 65.2248492 -31.91 65.5348493 28.07 65.6748495 33.38 65.5748496 38.01 65.6248497 44.02 65.43
40
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
5. HCAL PEDESTAL STABILITY STUDIES IN CMS
5.1 Introduction
In HCAL, pedestal subtraction is important to determine the muon energy deposits
and for calorimetry based muon isolation. Precision of pedestal determination has also a
direct impact on the quality of calibration of HCAL. In this method, pedestal is defined
on an event-by-event basis.
During September and November 2008 (Cosmic Run At Four Tesla (CRAFT) period)
we have studied the stability of HCAL pedestals. We have analyzed over 50 CRAFT
global runs using cosmic ray events (Drift Tubes (DT), Resistive Plate Chambers (RPC)
and Cathode Strip Chambers (CSC) triggers) to determine the average pedestal for each
HCAL readout channel for each run. The list of runs in this study used are shown in Table
5.1. Runs were needed to have at least 1 M muon triggers and B field=3.8T. Calorimeter
triggers for instance ECAL, HCAL based triggers which are HPD noise triggers or jet
triggers were removed from this analysis (Barbaro 2009).
In general, during CRAFT individual HCAL channels were stable. Run to run varia-
tion (RMS) of pedestals for most of the channels was in the range of 0.001 to 0.002 ADC
per Time Slice (TS). Assuming 8 time slices were used to reconstruct the HCAL energy,
this variation is equivalent to RMS of 2 to 4 MeV per readout channel. For few individual
channels, pedestals were not stable and exhibited not just a shift, but short or long term
drifts. Pedestal shifts of individual channels were not correlated (Barbaro, 2009).
5.2 Stability of HCAL Pedestals during CRAFT
5.3 Pedestal Definition
Cosmic ray muons permeate only few HCAL towers (out of a total of over ten thou-
sand HCAL readout channels). So reliable source of data to monitor the stability of HCAL
pedestals is provided by events triggered by cosmic ray muons (Barbaro, 2009). HCAL
uses Charge Integrator and Encoder (QIE) cards which are 7 bit ADC. Table 5.2. shows
41
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Table 5.1. List of runs used in analysis of pedestal stability during CRAFT (Barbaro, 2009).
Run
66615 66722 67139 67573 68129 69276
66637 66733 67141 67645 68141
66644 66739 67173 67647 68273
66676 66740 67219 67810 68276
66692* 66746 67225 67818 68279
66703* 66748 67534 67838 68288
66706* 66752 67538 67977 68500
66709* 66757 67539 68000 68926
66711* 66783 67541 68021 68949
66714* 66904 67544 68094 68958
66716 67126** 67548 68100 68253
66720 67128 67557 68124 68269
linearized ADC values for the lowest 30 QIE channels. The charge starts from -0.5 fC up
to 13.5 fC and it is incremented in 1.0 fC steps in the first 15 ADC channels which is the
lowest QIE range. Actually, ADC→ fC conversion factors change from QIE to QIE, with
mean a conversion factor of 0.91 ADC/fC and RMS of 3% (Barbaro, 2009).
HCAL ADCs rotate via four independent capacitors (CapIDs) for each readout chan-
nel to measure charge. Pedestal using values averaged over four CapIDs are more stable
that values calculated for individual CaIDs (Barbaro, 2009). We can calculate average
pedestal value of four CapIDs:
ped(channel,run) =1
8×NTRIG∑ADCTS (5.1)
where NT RIG is number of trigger and ADCT S is ADC per time slice.
42
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Table 5.2. The linearized ADC values for the lowest 30 channels are listed in table (Barbaro,2009)
qie 0 1 2 3 4 5 6 7 8 9convert2fC -0.50 0.50 1.50 2.50 3.50 4.50 5.50 6.50 7.50 8.50qie 10 11 12 13 14 15 16 17 18 19convert2fC 9.50 10.50 11.50 12.50 13.50 15.00 17.00 19.50 21.00 23.00qie 20 21 22 23 24 25 26 27 28 29convert2fC 25.00 27.00 29.50 32.50 35.50 38.50 42.00 46.00 50.00 54.50
5.4 Pedestal Calculation
Pedestal mean is given by,
4(ped) =RMS(ped)√
8×NTRIG(5.2)
RMS(ped) is RMS of pedestal for a single channel in a single run, is defined as:
RMS(ped) =√
1NTS
∑(ADC(channel,TS)−ped(channel,run))2 (5.3)
where ADC(channel,TS) is ADC of a channel per time slice and ped(channel,run) is
pedestal of a channel per run.
5.5 Stability of the HE and HB pedestals during CRAFT
We have analyzed stability of the pedestals of HE and HB channels. In this analysis,
six runs were removed, as HCAL data was taken with readout in Zero Suppression mode.
Runs marked with (*) were taken with HCAL readout in Zero Suppression mode and
they were removed from analysis. In addition, single run r67126 was also removed, as in
this run most of the channels in HBplus, RBX08 had incorrectly loaded pedestal settings
(Barbaro, 2009).
During CRAFT the average pedestal of all HB-HE channels is very stable. Run to run
RMS of the average pedestal is equal to 0.9 MeV for HB and 0.5 MeV for HE. 23 out of
a total of 5184 channels in HB and HE, less 0.5 percent of the total showed run-to-run
shifts of pedestals above the level of 0.010 ADC per TS. Such uncorrected pedestal shifts
would imply shift in HCAL reconstructed energy (RecHit) at the level of 20 MeV per
43
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.1. For a single run (r67647), distribution of mean pedestal values for 2583 HB channels(Barbaro, 2009).
readout channel. For few individual channels, pedestals were not stable. There were not
just a shift, but short or long term drifts. Pedestal shifts of individual channels were not
correlated. We propose to update pedestal constant of an individual HCAL channel only
if there is a shift for a particular channel is above some significant (5 sigma) threshold.
The possible threshold to change pedestal offset could be 0.004 ADC per TS (1 MeV per
TS) (Barbaro, 2009).
5.5.1 Stability of HCAL pedestal average
One can define the pedestal average for the HCAL subsystem as:
PED(run) =1
Nchn∑ped(channel,run) (5.4)
Distribution of mean pedestal values for 2583 HB channels for a single run (r67647) is
shown in Figure 5.1. Note that there were nine missing channels from the readout (HB
has a total of 2592 channels). The average pedestal is equal to 2.97 ADC counts. RMS of
pedestal distribution is 0.16 ADC counts.
Figure 5.2. shows the plot of the average of HB and HE pedestals with respect
to pedestal values for a reference run (r67647) during CRAFT. Average is equal to
44
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.2. The average of all HB pedestal (upper plot) and all HE pedestal (lower plot) withrespect to reference values plotted during CRAFT (Barbaro, 2009).
PED(run)-PED(r67467)
Run-to-run RMS of the pedestal average is equal to 0.00044 ADC/TS for HB and
0.00025 ADC/TS for HE. Assuming that one uses 8 time slices to define signal and con-
version factor of 250 MeV/ADC, this variation is equivalent to RMS of 0.9 MeV/channel
for HB and 0.5 MeV/channel for HE:
∆(PED)HB = 0.00044ADC/TS×8TS×250MeV/ADC = 0.9MeV (5.5)
∆(PED)HE = 0.00025ADC/TS×8TS×250MeV/ADC = 0.5MeV (5.6)
This result implies that on average HCAL pedestals were indeed very stable throughout
entire CRAFT period (October 16 - November 11, 2008) (Barbaro, 2009).
5.5.2 Stability of individual channel pedestals
We search long term stability of pedestals of individual HCAL channels. For exam-
ple, Figure 5.3. shows pedestal stability of two individual HB channels (eta=1, phi=31,
depth=1 on the left and eta=1, phi=41, depth=1 on the right). In the upper plots pedestal
mean of individual channel, ped (run,channel) is plotted as a function of the run number.
45
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.3. Pedestal stability for two HB channels (eta=1,phi=31,depth=1 on the left and eta=1,phi=41,depth=1 on the right). Upper plot: pedestal mean plotted. Lower plot: distribution ofpedestal means (Barbaro, 2009).
In the lower plots, distribution of the pedestal means for each run are shown. Each data
point corresponds to the average of the pedestal for each run calculated using 100k muon
triggers. RMS of pedestal means for these two particular channels are 0.0013 and 0.0022
ADC counts/TS. Note that as shown in Eqn. 2, statistical uncertainty of determination
of pedestal mean using 100k triggers is about 0.0007 - 0.0009 ADC counts/TS (Barbaro,
2009).
In order to check how stable throughout the CRAFT are individual pedestals of all
HCAL channels, we have looked at the distribution of difference between individual
HCAL channel pedestal means (calculated independently for each run) and reference
pedestal value for individual HCAL channels: PED(run, channel)-PED(r67467, channel)
Data is shown in Figure 5.4. separately for HB channels (upper plot) and HE channels
(lower plot). Here each entry in this plot corresponds to the pedestal mean for a particular
channel and a particular run with respect to the reference run. RMS of the distributions
is 0.0023 ADC counts/ TS. However, there are some data points (channels, runs) with
deviation of pedestal larger than 0.010 ADC counts.
We investigated these points by checking distribution of run-to-run RMS of pedestal
means for individual HCAL channels:
46
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.4. Distribution of difference between individual HCAL channel pedestal means (calcu-lated for independently for each run) and reference pedestal value for individual HCAL channels.Data is shown separately for HB channels (upper plot) and HE channels (lower plot) (Barbaro,2009).
RMS(channels) =1
Nruns(√
∑ (ped(run,channel)-ped(channel))2) (5.7)
where ped (channel) corresponds to the average (overall all runs) of pedestals for particu-
lar channel:
PED(channel) =1
Nruns∑ped(run,channel) (5.8)
Figure 5.5. shows distributions of pedestal RMS for HB (upper plot) and HE (lower plot)
channels. Average RMS is equal to 0.0017 ADC/Time Slice. However, there are 9 chan-
nels in HB and 14 channels in HE with RMS larger than 0.0100 ADC/Time Slice. Sta-
tistical error of pedestal calculation using 100k muon triggers is 0.0007 to 0.0009 ADC
counts/TS (Barbaro, 2009).
Figure 5.6. shows comparison of distribution of run-to-run RMS of pedestal means for
individual HB, HE channels. Two histograms correspond to the mean values calculated
47
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.5. Distribution of run-to-run RMS of pedestal means for individual HCAL Barrel (upperplot) and HCAL Endcap (lower plot) channels (Barbaro, 2009).
using 100k muon triggers. The average RMS is equal to 0.0017 ADC/TS (Barbaro, 2009).
Figure 5.7. shows pedestal mean vs run number (upper plot) pedestal mean distribu-
tion for individual HCAL channel with the largest run-to-run pedestal RMS. HE channel
(eta=26, phi=23, depth=1) has RMS equal to 0.035 ADC/Time Slice. Before run r68273,
pedestal mean was 3.07 ADC counts/TS. After run 68273, pedestal mean shifted upward
by 0.100 ADC counts, to value of 3.17 ADC counts/TS. For most of the channels, run-to-
run RMS is below 0.003 ADC/TS (Barbaro, 2009).
Figure 5.8. through Figure 5.11. show plots of pedestal mean vs run number for indi-
vidual HCAL channels with large RMS. For some of these channels, pedestal shifts took
place during single run. However, for few channels, there were apparent pedestal shifts
throughout several runs (long term drifts, as opposed to shift).
Table 5.3. shows list of 23 particular channels with large long-term instability.
48
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.6. Distribution of run-to-run RMS of pedestal means for individual HB HE channelscorrespond to the mean values calculated using 100k muon triggers (Barbaro, 2009).
Figure 5.7. Pedestal mean versus run number (upper plot) pedestal mean distribution for indi-vidual HCAL channel and pedestal mean distributions (lower plots) for HE channel HE (eta=26,phi=23,depth=1) (Barbaro, 2009).
49
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.8. Pedestal mean versus run number (upper plots) and pedestal mean distributions (lowerplots) for two individual HCAL channels, HE (eta=-20, phi=61, depth=1) and HB (eta=-2, phi=66,depth=1). In case of HE channel (left plots), a pedestal shift of∼ 0.060 ADC/TS took place duringCRAFT. Shift of 0.060 ADC/TS is equivalent of shift of 100 MeV/channel. In case of HB channelshown (right plots), two shifts took place: one shift upward, the second shift, downward (Barbaro,2009).
Figure 5.9. Pedestal mean versus run number for individual HCAL channels, HB (eta=2, phi=23,depth=1) and HE (eta=18, phi=8, depth=1). For these two channels, pedestal shifts were apparentthroughout several runs (long term drifts, as opposed to shift) (Barbaro, 2009).
50
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.10. Pedestal mean versus run number for individual HCAL channels, HB (eta=4,phi=36, depth=1) and HB (eta=6, phi=4, depth=1). For these two channels, pedestal shifts wereapparent throughout several runs (long term drifts, as opposed to shift (Barbaro, 2009)
Figure 5.11. Pedestal mean versus run number for individual HCAL channels, HE (eta=20,phi=32, depth=2) and HB (eta=18, phi=35, depth=1). For these two channels, pedestal shifts wereapparent throughout several runs (long term drifts, as opposed to shift (Barbaro, 2009)
51
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Table 5.3. List of 23 individual channels in HB and HE subdetectors with large run-to-runpedestal RMS. (Barbaro, 2009)
Subdetector eta phi depth RMS(ADC/TS)HE -27 35 1 0.013HE -23 19 2 0.028HE -20 20 2 0.014HE -20 55 1 0.015HE -20 61 1 0.023HE -19 19 2 0.013HE -18 36 2 0.015HE -18 44 2 0.011HE -16 18 3 0.019HE 18 8 1 0.017HE 18 35 1 0.011HE 20 32 2 0.012HE 26 23 1 0.036HB -5 15 1 0.026HB -2 23 1 0.020HB -2 66 1 0.023HB 3 5 1 0.011HB 4 36 1 0.021HB 6 4 1 0.010HB 7 17 1 0.013HB 15 3 2 0.015HB 15 16 1 0.018
5.6 Stability of HF pedestal during CRAFT
We have analyzed stability of pedestals of HF channels.In this analysis, six runs were
removed, as HCAL data was data taken with readout in Zero Suppression mode. Runs
are marked with *. Figure 5.12. shows average of HF pedestal, with respect to average
HF pedestal for r67647, as a function of run number. Run-to-run RMS of average HF
pedestal is 0.00029 ADC counts. A systematic shift in average pedestal value of 0.0007
ADC counts took place between runs 68000 and 68500 (Barbaro, 2009).
Figure 5.13. shows run-to-run RMS of individual HF channels. In general most of HF
channels are stable, with average RMS is 0.0016 ADC counts. However, few channels
exhibit instabilities. There are ten channels (out of a total of 1728) with run-to-run RMS
above 0.0050 ADC counts. Three channels in HF were missing during CRAFT. Individual
52
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.12. Pedestal average for HF versus run number. Average based on global runs, using100k events/run (Barbaro, 2009).
Figure 5.13. Run-to-run RMS of HF pedestals. Pedestals are based on global runs, using 100kevents/run. Ten channels (out of a total of 1725) have RMS larger than 0.005 ADC counts (Bar-baro, 2009).
channels with large RMS are listed in Table 5.4.
Figure 5.14. through Figure 5.17. show values of pedestal for ten HF channels with
large run-to-run RMS. In some cases, large RMS is the result of a single shift, in other
cases, pedestal is unstable throughout the entire CRAFT. In particle, a single with largest
RMS (eta=41, phi=47, depth=2) has a shift of 0.5 ADC counts.
In HF (eta=-31, phi=7, depth=1) channel there was a systematic shift of about 0.030
ADC counts/TS after run 68100.HF (eta=29, phi=71, depth=1) channel was unstable,
53
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
(a) (b)
Figure 5.14. Pedestal average for HF (eta=-31, phi=7, depth=1) channel versus run number (a),Pedestal average for HF (eta=29, phi=71, depth=1) channel versus run number (b).
with long term drift (down and up) 0.100 ADC counts/TS took place during CRAFT.
In HF(eta=-33, phi=39, depth=2) channel, a systematic shift up by 0.060 ADC
counts/TS took place after run 68100. In addition, this channel had pedestal instabil-
ity in the early part of the CRAFT. In HF(eta=41, phi=47, depth=2) channel there was a
systematic shift up by 0.500 ADC counts/TS after run 68300.
In HF(eta=39, phi=43, depth=2) channel, a long term shift up by 0.150 ADC
counts/TS took place during CRAFT. In HF(eta=30, phi=17, depth=2) channel, there is a
long term shift about 0.200 ADC counts/TS.
In HF(eta=29, phi=5, depth=1) channel a long term shift up by 0.150 ADC counts/TS
took place during CRAFT. In HF (eta=-36, phi=7, depth=1) channel, a long term shift up
by 0.150 ADC counts/TS took place during CRAFT.
5.7 Stability of HO pedestal during CRAFT
We have checked the stability of pedestal in HO channels during CRAFT.Figure 5.18.
shows average of HO pedestals, with respect to average HO pedestal for r67647. Run-
to-run RMS average HO pedestal is 0.00044 ADC counts. A systematic shift in average
54
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
(a) (b)
Figure 5.15. Pedestal average for HF (eta=-33, depth=39, depth=2) channel versus run number(a), Pedestal average for HF (eta=41, phi=47, depth=2) channel versus run number (b).
(a) (b)
Figure 5.16. Pedestal average for HF (eta=39, phi=43, depth=2) channel versus run number (a),Pedestal average for HF (eta=30, phi=17, depth=2) channel versus run number (b).
55
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
(a) (b)
Figure 5.17. Pedestal average for HF (eta=29, phi=5, depth=1) channel versus run number (a).Pedestal average for HF (eta=-36, phi=7, depth=1) channel versus run number (b).
Table 5.4. List of ten individual channels in HF with large run-to-run pedestal RMS. (Barbaro,2009)
channel eta phi depth RMS (ADC/TS)1 -36 7 1 0.0072 -31 7 1 0.0113 -30 51 2 0.0054 29 5 1 0.0055 29 71 1 0.0326 30 17 2 0.0057 33 39 2 0.0288 33 53 1 0.0399 39 43 2 0.00510 41 47 2 0.189
56
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.18. Pedestal average for Ho versus run number. Average based on global runs, using100k events/run (Barbaro, 2009).
pedestal value of 0.0007 ADC counts took place between runs 68000 and 68500. Three
runs: 66740, 66746 and 66748 were removed from this analysis. During these runs,
pedestals setting for many HO channels were set incorrectly and also six runs were re-
moved, as HCAL data was data taken with readout in Zero Suppression mode. Runs
marked with (*) were taken with HCAL readout in Zero Suppression mode.
Run-to-run RMS of individual HO channels is shown Figure 5.19. In general, 99.9%
of HO channels are stable. Average RMS is 0.0012 ADC counts.But few channels (0.1%)
exhibit instabilities. There are seventeen (out of total of 2160 channels) with run-to-run
RMS above 0.0050 ADC counts. Also there were six individual channels missing in HO
during CRAFT. Channels with RMS are listed in Table 5.5.
Figure 5.20. through Figure 5.23. show values of pedestals for ten HO channels with large
run-to-run RMS. In some cases, large RMS is result of a single shift, in other cases,
pedestal is unstable throughout entire CRAFT. In particular, there is a single channel with
largest RMS eta=-4, phi=69, depth=4 Figure 5.20. (a). For a group of runs, pedestal set-
ting was incorrectly loaded for this channel. Another channel (eta=15,phi=24, depth=4)
Figure 5.20. (b) is very unstable, with run-to-run variations exceeding 2 ADC counts.
There are several channels with a single shift (step) in pedestal value (Figure 5.21.-Figure
5.22.). There are also several channels with systematic downward or upward pedestal
57
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Figure 5.19. Run-to-run RMS of HF pedestal. Pedestal are based on global runs, using 100kevents/run. Five channels (out of a total of 2154) have RMS larger than 0.010 ADC counts (Bar-baro, 2009).
drift throughout entire CRAFT period.(see Figure 5.23.-Figure 5.23.)
58
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
(a) (b)
Figure 5.20. HO channel (eta=-4, phi=69, depth=4) with unstable pedestal. For block runs(r67128-r67541) pedestal setting for this channel was incorrectly loaded (a). HO channel (eta=15,phi=24, phi=4) with unstable pedestal. Pedestal value changes by up to 2 ADC counts (b).
(a) (b)
Figure 5.21. HO channel (eta=-5, phi=34, depth=4) with unstable pedestal (a), HO channel(eta=-11, phi=21, depth=4) with unstable pedestal (b).
59
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
(a) (b)
Figure 5.22. HO channel (eta=-11, phi=36, depth=4) with unstable pedestal (a), HO channel(eta=-6, phi=8, depth=4) with unstable pedestal (b).
(a) (b)
Figure 5.23. HO channel (eta=-1, phi=26, depth=4) with unstable pedestal (a), HO channel(eta=-2, phi=50, depth=4) with unstable pedestal (b).
60
5. HCAL PEDESTAL STABILITY STUDIES IN CMS Emine GURPINAR
Table 5.5. List of ten individual channels in HO with large run-to-run pedestal RMS. (Barbaro,2009)
channel eta phi RMS (ADC/TS)1 -14 16 0.0112 -11 21 0.0053 -11 36 0.0064 -6 8 0.0055 -5 34 0.0076 -4 69 1.1017 -3 70 0.0208 -2 50 0.0059 -1 26 0.01210 1 25 0.00911 1 26 0.00912 1 30 0.00713 1 40 0.00814 1 29 0.00615 3 42 0.00716 10 9 0.00717 15 24 0.293
61
6. CONCLUSION Emine GURPINAR
6. CONCLUSION
My thesis contains two major studies. The first one is the beam test of the final pro-
totype IV of CASTOR calorimeter. In the second study, I present my analysis of stability
of HCAL pedestal by using data taken during CRAFT.
The beam test analysis of the final prototype IV of CASTOR calorimeter, quartz-
tungsten calorimeter of the CMS experiment as performed with the data which were col-
lected during the test beam of 2008 summer at the H2 CERN/SPS beam line.
In the first part of the thesis, surface scan were studied with an electron beam at 100
GeV, pion beam at 80 GeV. The purposes of the area scanning are to check the uniformity
of the EM calorimeter response to electrons hitting at different points on the sector area,
to estimate the width of the EM shower profile and to assess the amount of the effects and
lateral leakage from the calorimeter which could lead to cross-talk between neighboring
sectors. Spacial cuts were applied to select only clean events for instance without con-
tamination from unwanted particles, beam cut, scintilator cut, muon veto cut and FEM
cut. I used analysis tools such as Root, CMSSW. The detector was scanned with the elec-
trons at 100 GeV and the pions at 80 GeV horizontally and detector response versus beam
impact points were studied. The derivative of the response is calculated and the electro-
magnetic shower width is found to be 1.903 mm for Saleve side and 1.601 mm for Jura
side, hadronic shower width is found to be 6.081 mm for Saleve side,
As expected, the pion shower was larger than the corresponding electromagnetic
shower. It should be noted, however, that in Cerenkov calorimeters the electromagnetic
and hadronic shower widths are approximately a factor of 10 narrower than those in a
dE/dx scintillator calorimeter.
In the second part of the thesis HCAL pedestal stability have been studied. I checked
HCAL pedestal stability using CRAFT global runs and looked at over 50 runs, 100k muon
triggers per each run, HPD noise triggers were excluded. Calorimetry triggers, such as
HPD noise or jet triggers were not included in this analysis.
Most (99.9%) of HCAL readout channels have very stable pedestal. So 0.1% of HCAL
channels are unstable. Typical RMS is 0.001 to 0.002 ADC counts in HB and HE RMS
is 0.001-0.002 ADC/TS. This value is equivalent to RMS of 2-3 MeV per channel. Some
channels (20 out of 5k, < 0.5% of total) showed pedestal variation above the level of 0.010
62
6. CONCLUSION Emine GURPINAR
ADC/TS (equivalent to 15 MeV/channel). In few cases, channels are not stable (not just
a shift, but short or long term drifts). Typical (average) deviation of a single channels
is 0.4 MeV/100 channels. Most of HF channels are stable, with RMS is 0.0016 ADC
counts. There are ten with run-to-run RMS above 0.005 ADC counts. These channels are
unstable. Individual channels with RMS are listed in Table 5.4.
HF(41,47,2) channel which has a shift RMS (0.5 ADC counts), in HO, RMS is 0.0012
ADC counts. There are seventeen with run to run RMS above 0.05 ADC counts (see
Table 5.5.). Single channel in HO(15,24) very unstable pedestal drifts by more than 2
ADC counts. Few channels show slow-term drifts.
63
REFERENCES
ASLANOGLOU et al., 2006, First performance studies of a prototype for the CASTOR
forward calorimeter at the CMS experiment, CMS-NOTE2006-142.
ASLONLOGLOU et al., 2007, Performance Studies of Prototype II for the CASTOR
forward Calorimeter at the CMS Experiment, arXiv:0706.2641v2 [physics.ins-det].
ASLANOGLOU et al., 2008, Performance studies of the final prototype for the CASTOR
forward calorimeter at the CMS experiment CMS NOTE 2008/-(in preparation)
AKSAY, N. 2006. CMS(Compact Muon Solenoid) deneyindeki Castor detektorunun Test
veri Analizleri, Adana, 86s.
BASEGMEZ, S., 2008. CMS-CASTOR forward detector and 2007 Test Beam Data Anal-
ysis, Adana, 91s.
CASTOR Engineering Design Report, 2007, CERN,
CMS Collaboration, The Tracker Project Technical Design Report, CERN/LHCC 1998-
006 (1998).
CMS Collaboration, ECAL Technical Design Report, CERN/LHCC 1997-033 (1997).
CMS Collaboration, The Hadron Calorimeter Technical Design Report, CERN/LHCC
1997-031 (1997)
CMS Collaboration, The Muon Project, Technical Design Report, CERN/LHCC 1997-10
(1997)
CMS Physics Technical Design Report, 2007, Vol II, CERN
DE BARBARO et al., 2009, Stability of HCAL hardware pedestals during CRAFT and
Proposal for definition of pedestal tags and Intervals of Validity (IOVs) for HCAL
energy reconstruction, CMS IN-2009/005
FABJAN, C.W. and GIANOTTI, F., 2003, Calorimetry for Particle Physics,
CERNEP/2003-075, CERN
GUMUS, K., 2008. Search for new Physics in the Compact Muon Solenoid (CMS) exper-
iment and the response of the CMS Calorimeters to particles and jets, Texas Tech
University, 113s.
LEO, W.R., 1994.Techniques for Nuclear and Particle Physics Experiments, NewYork
MAVROMANOLAKIS, G., 2003. Quartz fiber calorimetry and calorimeters, Cambridge,
UK
MOORTGAT, F., 2004 Discovery potential of MSSM Higgs bosons using supersymmet-
ric decay with the CMS detector, Antwerpen
64
6. CONCLUSION Emine GURPINAR
SINGH BAWA, H., 2007 Some Aspects of High ET jets in p-p collisions at LHC
The CMS Collaboration 2007 J. Phys. G: Nucl. Part. Phys. 34 995-1579
VIRDEE, T.S., 1998, Calorimetry. CMS Collaboration,CERN, Geneva, CMS Conference
Report,CMS-CR, 1998/026
WIGMANS, R., 2000. Calorimetry Energy Measurement in Particle Physics, Oxford
University Press, UK, 17,726.
https://en.wikipedia.org/wiki/StandardModel
https://twiki.cern.ch/twiki/bin/view/CMS/CastorEDR
65
RESUME
I was born in Adana. I graduated from primary and secondary schools in Gaziantep.
After that, I enrolled in the high school of Bingol Lisesi and graduated in 2001. I enrolled
in the Physics Department of Cukurova University and I graduated in 2007. I continued
to study for my Masters degree in High Energy Physics, at the Institute of Natural and
Applied Sciences in Cukurova University.
66
Recommended