A kaon trigger for FOPI [Elektronische Ressource] : development and evaluation of a trigger system for strange particles / presented by Oliver Brosch

Dissertationsubmitted t o theCombined Faculties for the N atural S ciences and for Mathematicsof the Ruperto- Carola University of Heidelberg, Germanyfor the degree ofDoctor of Natural Sciencespresented byDiplom-Ph ysiker: Oliver Broschborn in: Jugenheim an der BergstraßeOral examination: May 12, 2004A Kaon Trigger for FO PIDevelopment and Evaluation of aTrigger System for Strange ParticlesReferees: Prof. Dr. Reinhard MännerProf. Dr. Norbert HerrmannEin K aon-Trigger für FOPI –Entwicklung un d Bewertung eines Triggersystems für seltsame TeilchenIn Stößen schwe rer Ionen bei hohen Ene rgien am SIS/GSI kann Kernma terie großer Di cht eund Temperatur erzeugt werden. Im FO PI-E xperiment werden die Spur en der bei denKolli sionen entstehenden geladenen Teilchen nach gewiesen. D urch deren Analyse, insb eson-dere im Hinbli ck auf die seltsame n K-Mes onen, können unsere Ke nntnisse üb er die Strukturder Materie und die Vorgänge beim Entstehen von Ne utronensternen und Schw arzen Löchernvertieft werden. Kaon en werden jedoch selten beob acht et, und so sind die erzielb arenphysikalischen Er gebniss e meist mit großen Fehlern behaf tet. Um die Ausbe ute an Kaonendeutlich zu erhöhen, wurde im Rahmen dieser Arb eit ein Trigger entwick elt, der das Zeit rau-bende Au slesen der kompletten D etektordaten für nicht interessant e Ereignisse verhindernkann. D afür wurde ein auf der H ough-Transforma tion basierender Al gorithmus gesc haffen.
Publié le : jeudi 1 janvier 2004
Lecture(s) : 12
Tags :
Source : ARCHIV.UB.UNI-HEIDELBERG.DE/VOLLTEXTSERVER/VOLLTEXTE/2004/4651/PDF/DISS_BROSCH.PDF
Nombre de pages : 167
Voir plus Voir moins

Dissertation
submitted t o the
Combined Faculties for the N atural S ciences and for Mathematics
of the Ruperto- Carola University of Heidelberg, Germany
for the degree of
Doctor of Natural Sciences
presented by
Diplom-Ph ysiker: Oliver Brosch
born in: Jugenheim an der Bergstraße
Oral examination: May 12, 2004A Kaon Trigger for FO PI
Development and Evaluation of a
Trigger System for Strange Particles
Referees: Prof. Dr. Reinhard Männer
Prof. Dr. Norbert HerrmannEin K aon-Trigger für FOPI –
Entwicklung un d Bewertung eines Triggersystems für seltsame Teilchen
In Stößen schwe rer Ionen bei hohen Ene rgien am SIS/GSI kann Kernma terie großer Di cht e
und Temperatur erzeugt werden. Im FO PI-E xperiment werden die Spur en der bei den
Kolli sionen entstehenden geladenen Teilchen nach gewiesen. D urch deren Analyse, insb eson-
dere im Hinbli ck auf die seltsame n K-Mes onen, können unsere Ke nntnisse üb er die Struktur
der Materie und die Vorgänge beim Entstehen von Ne utronensternen und Schw arzen Löchern
vertieft werden. Kaon en werden jedoch selten beob acht et, und so sind die erzielb aren
physikalischen Er gebniss e meist mit großen Fehlern behaf tet. Um die Ausbe ute an Kaonen
deutlich zu erhöhen, wurde im Rahmen dieser Arb eit ein Trigger entwick elt, der das Zeit rau-
bende Au slesen der kompletten D etektordaten für nicht interessant e Ereignisse verhindern
kann. D afür wurde ein auf der H ough-Transforma tion basierender Al gorithmus gesc haffen.
Di eser rekonstruiert Teilc henspuren aus einem kleinen Teil der D aten der D riftkammer CDC.
Di e geometrisc he Verknüpfung mit den Information en des neuen, hoch auf lösenden Flugzeit-
detektors G RPC erlaub t die Bestimmung der gefundenen Teilche narten. U m den An -
forderungen durch die Datenb andbr eite und die Re chenint ensität zu genügen, sind für die Im-
plementierung spezielle Pr ozessoren nötig. Etwa 5 bis 6 der an der Univ ersität M annheim
entwickelten, FPG A-basierten MP RA CE-Karten können diese Leistung kostengünstig
erbringen. D ie konsequente Parallelisierung der einzelnen Pro grammschr itte ermöglicht das
Aus nutzen der vollen Leistungsfähigkeit von MP RA CE und damit eine Re chenze it von
weniger als 100 s pro Er eignis. Ausführlich e Simulation en des Triggersystems zeigen, dass

in E xperimenten mit leicht en Kernen wie Nick el bei Str ahl energien von 1.93 GeV /u die K -

Aus beut e etwa um den Faktor 6 und die K -Ausb eute um den Faktor 11 gesteigert werden
kann.
A Kaon Trigger for FOPI –
Development and Eval uation of a Trigger System for Strange P articles
Du ring col lisions of heavy ions at high energies at SIS/GSI, nucl ear matt er can be exposed to
high densities and temperatu res. The FOP I experiment detects the traces of the charged parti-
cles p roduced i n th e nuc lear r eaction. Th eir analysis, in particular wit h r espect t o th e st range K
mesons, can extend our knowledge about the structure of nucl ear matt er and the proces ses
during the evolution of neutron stars and bl ack holes. Ho wever, kaons are rarely obse rved,
hence the derived physics results suffer from large unce rtainti es. In order to signific ant ly en-
hance the kaon yield, a trigger, which is abl e to prevent the time consuming readout of the
complete detector data for uninteresting events, was developed within the scope of this work.
For that purpose, a Hou gh trans form based algorithm was created. It recon structs particl e
track s from a small fract ion of the data of the drift ch amb er CDC. A geometric al ma tching to
the informa tion from the new high-prec ision time-of-fli ght detect or GR PC allows th e determi-
nation of the found particl es' speci es. In order to fulfill the requirements from the data band -
width and the computing intensity, speci al-purpose proc essors are required. About 5 to 6 of
the FPGA based MP RA CE bo ards developed at Mannh eim Univ ersity can provide this perfor-
manc e cost-e ffectively. The co nsistent parallelization of the individual program steps makes
100 sthe exploitation of the full power of MP RA CE and thus a processing time of less than
per event possible . D etailed simulations of the trigger systems show, that in experiments with

light nuc lei like ni ckel at beam energies of 1.93 GeV /u the K yield can be enhanced by a fac-

tor of 6, and th e K yield by a facto r of 11.Table o f Contents
Table of Contents 7
Cha pter 1 : Introduction 11
1.1 Kaons and St ran ge Pa rticl es........................................................................................... 11
1.1.1 Kaon Pr oduction ................................................................................................... 11
1.1.2 Strangeness........................................................................................................... 12
1.2 Compressed Nu cl ear Matter........................................................................................... 13
1.2.1 Nu clear Phas e Di agram ........................................................................................ 13
1.2.2 In-M edium Effects................................................................................................ 13
1.2.3 Th eoretical App roa ch es........................................................................................ 14
1.3 Obs erva bles.................................................................................................................... 15
1.3.1 Produc tion an d R atios.......................................................................................... 16
1.3.2 Colle ctive E ffec ts................................................................................................. 17
1.4 About this Work.............................................................................................................. 17
Cha pter 2: The F OPI Experiment 19
2.1 GS I Acce lerato r Facility................................................................................................. 19
2.2 FOP I Detecto r Setup ....................................................................................................... 20
2.2.1 Centr al Dr ift Chamber (CDC).............................................................................. 20
2.2.2 G RPC Time-of -Flight De tector ............................................................................ 23
2.2.3 Oth er De tecto rs....................................................................................................25
2.3 D ata Acquisiti on............................................................................................................. 26
2.3.1 Components.......................................................................................................... 26
2.3.2 Pe rformanc e......................................................................................................... 28
2.3.3 Interfaces.............................................................................................................. 28
2.4 Ev ent Re cons truction ..................................................................................................... 29
Cha pter 3: A Kaon Trigger fo r FOPI 33
3.1 Hi gh-Level Trigger Systems........................................................................................... 33
3.1.1 Trigger Idea.......................................................................................................... 33
3.1.2 Trackin g Appr oaches............................................................................................ 34
3.1.3 H ardware Platfo rms.............................................................................................. 35
3.1.4 E xisting Hi gh-Level Triggers............................................................................... 36
3.2 System Spe cifi cation...................................................................................................... 38
3.2.1 Timing Constr aint s............................................................................................... 38
3.2.2 Pre requisites......................................................................................................... 40
3.2.3 Trigger Conc ept.................................................................................................... 41
3.3 Tracki ng Al gorithm........................................................................................................ 42
3.3.1 Hou gh Transform................................................................................................. 43
3.3.2 Hou gh M ap........................................................................................................... 44
7Table of Contents
3.3.3 Lookup Tabl e........................................................................................................ 45
3.3.4 Al gorithm Discu ssion........................................................................................... 46
3.4 Trigger Hardwa re............................................................................................................ 47
3.4.1 Re quirements........................................................................................................ 47
3.4.2 FPGA Proc essors.................................................................................................. 48
3.4.3 MP RACE............................................................................................................. 48
3.4.4 H ardware Dis cussion............................................................................................ 50
Cha pter 4: Algo rithm Implementation 51
4.1 Lookup Tables................................................................................................................ 51
4.1.1 Track Symmetries................................................................................................. 51
4.1.2 Peak Shap e Opti mization..................................................................................... 52
4.1.3 Inverse Transverse M omentum App roa ch............................................................ 54
4.1.4 TO F Matching App roach...................................................................................... 55
4.1.5 CD C C overage Appr oach..................................................................................... 56
4.1.6 Track Width.......................................................................................................... 57
4.1.7 LU T Structure....................................................................................................... 58
4.1.8 Re verse LU T........................................................................................................ 59
4.2 Pr ocessing M odules........................................................................................................ 59
4.2.1 Hit Recognition.................................................................................................... 59
4.2.2 Hit Filter............................................................................................................... 59
4.2.3 Histo gram mer....................................................................................................... 60
4.2.4 Histo gram Fil ter................................................................................................... 60
4.2.5 M axi mum Finder.................................................................................................. 61
4.2.6 M axi mum Filter...................................................................................................62
4.2.7 G ap Filt er.............................................................................................................. 62
4.2.8 Track Fit............................................................................................................... 63
4.2.9 TO F Matching...................................................................................................... 64
4.2.10 Particl e Identific ation......................................................................................... 65
Cha pter 5: Trigger Simulation 67
5.1 Simulat ion F ram ework................................................................................................... 67
5.1.1 Ext ernal Events.................................................................................................... 67
5.1.2 Event C reation ...................................................................................................... 68
5.1.3 Batch Pr oces sing.................................................................................................. 69
5.1.4 Effici ency C alcul ation.......................................................................................... 69
5.2 Ev ent Viewer.................................................................................................................. 70
5.2.1 D etector Data........................................................................................................ 70
5.2.2 Track His togram............................................ 72
5.2.3 Track R econs truction............................................................................................ 72
5.2.4 Particl e Histograms.............................................................................................. 72
5.2.5 Effici ency Di agram s............................................................................................. 73
Cha pter 6: Hardware Implementation 75
6.1 Parallelization I ssues...................................................................................................... 75
6.1.1 LU T Acc ess.......................................................................................................... 76
6.1.2 Histo gram Acces s................................................................................................. 77
6.2 Implementation L ayout................................................................................................... 77
6.2.1 LU T Addressing................................................................................................... 77
6.2.2 Histo gram Perfo rmance........................................................................................ 78
8Table of Contents
6.2.3 Filter Pipel ine....................................................................................................... 79
6.2.4 E xecution Spe ed................................................................................................... 80
6.3 Resou rce Con sumption..................................................................................................82
6.3.1 Parameter De pendence......................................................................................... 82
6.3.2 E xample Conf igurations....................................................................................... 82
Cha pter 7: Performance R esults 85
7.1 Simulat ion Condit ions.................................................................................................... 85
7.1.1 E mployed Eve nt Data................ 85
7.1.2 Stand ard P arameters............................................................................................. 87
7.2 Trigger E ffici ency Results.............................................................................................. 88
7.2.1 Trackin g Effici ency.............................................................................................. 88
7.2.2 Particl e Re co gnition............................................................................................. 90
7.2.3 Trigger Pe rforma nc e............................................................................................. 93
7.2.4 Algorithm Paramete rs........................................................................................... 95
7.2.5 Pe rformanc e Variations ........................................................................................97
7.3 Resou rce Optim ization ................................................................................................. 100
7.3.1 Pe rformanc e St ability......................................................................................... 100
7.3.2 System Di mensions............................................................................................ 100
Cha pter 8: Conclusion 103
8.1 System Performan ce..................................................................................................... 103
8.1.1 Be nch mar k Resul ts............................................................................................. 103
8.1.2 Compar ison an d Dis cus sion............................................................................... 104
8.2 Trigger Application....................................................................................................... 106
8.2.1 FOP I Experiments.............................................................................................. 106
8.2.2 Implementation Co sts......................................................................................... 106
8.2.3 Integration Ef fort................................................................................................ 107
8.3 Outlook ......................................................................................................................... 108
8.3.1 Algorithm Enh ance ment..................................................................................... 108
8.3.2 Implementation Improvements........ 109
8.3.3 Oth er Fields of Appli cation ................................................................................ 110
Cha pter 9: Summary 111
References 115
Ap pendix A: Application Use r Docu mentation 123
Ap pendix B: Ap plication Development 165
Acknowledgments 167
9

Soyez le premier à déposer un commentaire !

17/1000 caractères maximum.