119
NATIONAL INSTITUTE FOR NUCLEAR PHYSICS AND HIGH-ENERGY PHYSICS
ANNUAL REPORT
2003
Kruislaan 409, 1098 SJ AmsterdamP.O. Box 41882, 1009 DB Amsterdam
Colofon
Publication edited for NIKHEF:Address: Postbus 41882, 1009 DB Amsterdam
Kruislaan 409, 1098 SJ AmsterdamPhone: +31 20 592 2000Fax: +31 20 592 5155E-mail: [email protected]
Editors: Louk Lapikás, Marcel Vreeswijk & Auke-Pieter ColijnLayout & art-work: Kees HuyserCover Photograph: Relativistic fluid mechanics: impression of a fluid current moving on a sphere where scalar fields live.URL: http://www.nikhef.nl
NIKHEF is the National Institute for Nuclear Physics and High-Energy Physics in the Netherlands, in which the Foundation for FundamentalResearch on Matter (FOM), the Universiteit van Amsterdam (UvA), the Vrije Universiteit Amsterdam (VUA), the Katholieke UniversiteitNijmegen (KUN) and the Universiteit Utrecht (UU) collaborate. NIKHEF co-ordinates and supports all activities in experimental subatomic(high-energy) physics in the Netherlands.
NIKHEF participates in the preparation of experiments at the Large Hadron Collider at CERN, notably Atlas, LHCb and Alice. NIKHEFis actively involved in experiments in the USA (D∅ at Fermilab, BaBar at SLAC and STAR at RHIC), in Germany at DESY (Zeus, Hermesand Hera-b) and at CERN (Delphi, L3 and the heavy-ion fixed-target programme). Furthermore astroparticle physics is part of NIKHEF’sscientific programme, in particular through participation in the ANTARES project: a detector to be built in the Mediterranean. DetectorR&D, design and construction of detectors and the data-analysis take place at the laboratory located in Sciencepark Amsterdam as well asat the participating universities. NIKHEF has a theory group with both its own research program and close contacts with the experimentalgroups.
Contents
Preface 1
A Experimental Programmes 3
1 The ATLAS Experiment 3
1.1 ATLAS experiment 3
1.2 D∅ experiment 9
2 The B-Physics Programme 13
2.1 Introduction 13
2.2 Status of HERA-B 13
2.3 Status of the BaBar Experiment 13
2.4 The LHCb Vertex detector 15
2.5 Level-0 Pile-Up Veto System 18
2.6 The Outer Tracker of LHCb 18
2.7 Reconstruction and Physics Studies 22
3 Heavy Ion Physics 25
3.1 Introduction 25
3.2 Results from SPS experiments 25
3.3 Jet production and jet quenching studied in STAR 26
3.4 The Alice experiment at CERN 27
4 ANTARES 29
4.1 Introduction 29
4.2 Prototype Sector Line 29
4.3 On-shore data processing 30
4.4 Analysis methods 30
4.5 Outlook 31
i
5 ZEUS 33
5.1 Introduction 33
5.2 ZEUS beats HERA beam background 33
5.3 Analysis HERA-I from 1991-2000 33
5.4 Performance of the ZEUS vertex detector 34
6 HERMES 37
6.1 Introduction 37
6.2 Data taking 38
6.3 Physics analysis 38
6.4 Instrumentation 41
6.5 Outlook 43
7 DELPHI 45
7.1 DELPHI programme and Detector exhibit 45
7.2 Publications 45
7.3 B physics 45
7.4 QCD 46
7.5 Standard Model Electroweak results 46
8 L3 49
8.1 Introduction 49
8.2 Searches 49
8.3 W and Z physics 49
8.4 QCD 51
8.5 L3+Cosmics 51
B Transition Programme 53
1 New Detector R&D at NIKHEF 53
1.1 Introduction 53
1.2 The readout of drift chambers with the Time-Pix-Grid 53
1.3 CMOS-based pixel detectors 55
1.4 X-ray imaging pixel detectors 55
2 Grid Projects 57
2.1 Introduction 57
2.2 Local Facilities 57
ii
2.3 The LHC Computing Grid Project 57
2.4 The DataGrid Project 57
2.5 AliEn 60
2.6 D∅ Grid 602.7 EGEE 60
2.8 Virtual Laboratory for eScience 60
3 Experiments abroad 61
3.1 Proton-neutron knockout from 3He induced by virtual photons 61
C Theoretical Physics 63
1 Theoretical Physics Group 63
1.1 Introduction 63
1.2 Physics of the standard model 63
1.3 Beyond the standard model 64
1.4 Cosmology, astrophysics and quantum gravity 64
D Technical Departments 67
1 Computer Technology 67
1.1 Computer system management 67
1.2 Project support 69
2 Electronics Technology 73
2.1 Introduction 73
2.2 Department developments 73
2.3 Projects 73
3 Engineering Department 79
3.1 ATLAS 79
3.2 LHCb 80
3.3 Alice Inner Tracking System 81
3.4 ANTARES 81
3.5 Alpha Magnetic Spectrometer 81
3.6 IdePhix 82
4 Mechanical Workshop 83
iii
4.1 Introduction 83
4.2 Projects 83
E Publications, Theses and Talks 87
1 Publications 87
2 PhD Theses 98
3 Invited Talks 99
F Resources and Personnel 105
1 Resources 105
2 Membership of Councils and Committees during 2003 106
3 Personnel as of December 31, 2003 109
iv
Preface
The CERN LHC program will start in 2007. This is animportant date for many physicists worldwide and cer-tainly also for NIKHEF and its scientific and technicalstaff. The construction of the various components ofthe large detectors ATLAS, LHCb and ALICE or insteadsees steady progress. During 2003 we saw completionof 75 out of 96 muon chambers for ATLAS. NIKHEFscientists contributed in an important way to the H8testbeam which validated many aspects of the ATLASmuon System. In LHCb an important achievement wasthe production of the first prototype secondary vacuumbox. The LHCb outer tracker project has passed allnecessary steps to enter the production phase. The AL-ICE experiment is also entering the production phase.NIKHEF contributes to electronics and readout. To-gether with industry, tooling has been developed forthe positioning of the silicon ladders.
These preparations for the LHC program form anenormous challenge for the technical departments ofNIKHEF. Their work is essential in the execution of thescientific program of NIKHEF, and in the preparationfor future activities. At this point, with a deep senseof loss, I recall the decease of Paul Rewiersma of theelectronics technology department. Paul started hiswork at NIKHEF with design work for the ACCMORcollaboration. His last contributions were to theANTARES experiment.
The design and construction of detectors is only one ac-tivity towards the LHC program. In parallel, NIKHEFphysicists participate in the D∅ experiment at Fermi-lab, BaBar at SLAC and the STAR collaboration atBrookhaven. The aim is to be involved in frontline ex-perimental research as well as to prepare for the manyanalysis tasks at similar LHC experiments.
The D∅ experiment has performed a thorough analysisof the properties of the top-quark, with the importantresult that the value of the top-quarks mass is now179.0 ± 3.5 ± 3.8 GeV. This value is higher than theresults from previous analyses. In combination withLEP data it means that the prediction for the Higgsmass is now also higher. A NIKHEF group joined BaBar
at the end of the year 2002. These physicists haveconcentrated on the measurement of the angle γ, one ofthe parameters expressing CP-violation in the B system.The BaBar collaboration found the new charmed mesonDs(2317) and discovered the decay B → π0π0. Itsbranching fraction is an important input in further CP-violation studies. The STAR collaboration has obtainedinteresting results on the behavior of jets produced inheavy-ion collisions. These results are important for theALICE Program.
In order to investigate phenomena like the Higgs mech-anism or CP-violation, sound theoretical predictions arenecessary. The NIKHEF theory group has been work-ing on many aspects of the standard model and beyond.Here it remains clear that the algebra program FORMis an indispensable tool in this theoretical work. Thetheory group entered the field of astroparticle physics,a relatively new area of research.
From the start of LHC operations, the experimentalgroup will have to deal with massive amounts of dataat an unprecedented level. The development of the re-quired grid technology is now well underway. NIKHEFplays a major role in the development of several gridtechnologies and its implementation in several gridprojects such as VL-E and EGEE. The NIKHEF par-ticipation in D∅ and other experiments provides goodopportunities to test the grid for high-energy physicsapplications.
On the 17th of March 2003 the ANTARES collabora-tion successfully put the first data and control lines onthe bottom of the Mediterranean Sea and connectedthe equipment to the junction box so that signals wereavailable on shore on the same day. Although therewas a problem with the reference clock, many aspectsof the design could be validated. The completion ofthis neutrino telescope is foreseen in 2006.
The LEP program has officially come to an end in 2003;the enormous contribution of the LEP program to ourcurrent understanding of the standard model will be inthe textbooks for many years to come. The NIKHEFprogram at the HERA collider has been the subject of
1
an extensive review by a panel chaired by J. Dainton.The committee underlined the high quality of the sci-entific work and made strong recommendations for thefinal stages and completion of the program at HERA.
Jos Engelen
During the year 2003 it became clear that Jos Enge-len would not complete his five year term as directorof NIKHEF. He was appointed by the CERN councilas Chief Scientific Officer and deputy Director Generalof CERN. The NIKHEF community congratulates himwith this great personal and scientific honor.
Karel Gaemers
2
A Experimental Programmes
1 The ATLAS Experiment
1.1 ATLAS experiment
A view of the ATLAS detector is shown in Figure 1.1.NIKHEF has significant detector construction responsi-bilities for the muon spectrometer (the outermost shellsof the ATLAS detector; see Section 1.1) and for thecentral tracker (the part of the ATLAS detector near-est to the beam line; see Section 1.1). The preparationswhich should lead to a prominent presence in ATLASdata analysis can be found in Section 1.1. The progressregarding our participation in the D∅ experiment at Fer-milab is presented in Section 1.2.
One of the highlights in 2003 was the H8 test beamprogram at CERN with NIKHEF physicists taking anactive part in the validation of the ATLAS muon spec-trometer and silicon tracker detector hardware, read-out electronics and data analysis. With the ATLASdetector construction work well underway, the effortscontinue to shift to reconstruction software and physicsperformance studies. This led to major Dutch contri-butions to recent software releases. In December 2003,the Dutch ATLAS group organized a 3-day workshop inLunteren to discuss the preparations leading to physicsdata analysis. This workshop was very successful andwill be repeated in 2004. In 2003, five PhD theses werecompleted including the thesis of our first PhD studenton the D∅ experiment.Muon spectrometer
The ATLAS muon spectrometer consists of a barrel andtwo end-caps. In the barrel region the muon track ismeasured by three concentric shells of muon chambersbased on the Monitored Drift Tube (MDT) principle.In the end-caps the muon track is measured by threedisks covered with MDT chambers. In the barrel re-gion the magnetic field is generated by a huge super-conducting toroid with eight coils. The magnetic fieldin the end-cap regions is generated by smaller super-conducting toroids. NIKHEF is responsible for manycomponents in the ATLAS muon spectrometer: the 96Barrel Outer Large (BOL) MDT chambers; the bar-rel (RASNIK) alignment system; the complete detectormonitoring (temperature and magnetic field) and con-trol (initialization of the front-end electronics) and thehighest level of the data acquisition system (the MuonRead-Out Driver or MROD). In addition Dutch indus-tries construct the two large end-cap toroids.
The NIKHEF muon spectrometer detector constructionresponsibilities are near completion. As a consequence,most physicists are working on the analysis of H8 testbeam data and the development of simulation and re-construction software for the full ATLAS muon spec-trometer. The only muon related hardware project re-maining in the development phase is the precise calibra-tion (order 0.01%) of 1200 magnetic field probes. Thisproject is a collaboration between CERN and NIKHEF.
MDT chamber production
At the end of 2003, 75 of the 96 large BOL MDTchambers were completed at NIKHEF. During 2003, wegained extensive experience in equipping these cham-bers with services like alignment, temperature moni-toring, gas distribution, front-end electronics cards andFaraday cages. The earlier difficulties with malfunc-tioning gas pipes have been solved. At this momentthe only missing component is the so-called ChamberService Module (CSM) which collects all data from thefront-end electronics cards and sends them on a singlefiber to the higher level data acquisition electronics (theMROD, see section 1.1 for more details). The CSMsare provided by our American colleagues and should beavailable in 2005. At NIKHEF we have already six pro-totype CSMs to operate the large cosmic-ray stand totest the 96 BOL MDT chambers under real operationconditions.
In agreement with the schedule, all 96 BOL MDTchambers will be transported to CERN in the fall of
Figure 1.1: Artist impression of the ATLAS detector.
3
Figure 1.2: The assembly of four magnetic field sensorsand at the center one coil calibration card.
2004 to be integrated with the resistive plate triggerchambers (RPCs) prior to their installation into the AT-LAS cavern in 2005. The preparation of the MDT-RPCintegration, including the construction of the commonMDT-RPC mounting fixtures, was started in 2003 andapart from the worrisome RPC production schedule noother problems are foreseen.
RASNIK alignment status
In 2003 the calibration bench for the important so-called projective RASNIK alignment system was devel-oped and demonstrated to be able to reach a precisionof 4 µm, well below the target precision of 15 µm.Also in 2003 the next level in the multiplexed RASNIKreadout scheme, the MasterMux, was designed, testedand submitted for serial production. This leaves, inaccord with the schedule, for 2004 the design of thefinal USA15Mux as the last RASNIK related task to becompleted.
Detector control system
The detector control system (DCS) for the ATLASmuon spectrometer is entirely developed by NIKHEFand comprises: temperature monitoring, magnetic fieldmonitoring, initialization of the MDT front-end elec-tronics and the monitoring of the voltages and temper-atures on the MDT front-end electronics cards. Thedesign of the system was completed in 2003 and, apartfrom the magnetic field sensors, all components are ei-ther available or are in production.
The magnetic field sensors make use of the Hall ef-fect: on each sensor three orthogonally mounted Hallsensors are used to measure the value and the direc-tion of the local magnetic field. The crucial aspect
Figure 1.3: The magnetic field calibrator in the dipolemagnet at CERN.
of these sensors is their calibration to a precision of0.01%. This is realized by means of an ingenuous cal-ibration bench (and procedure) originally developed bya group at CERN. This CERN device demonstrated thefeasibility of the 0.01% precision. However, in 2002 itbecame clear that the CERN bench was not suitable forthe serial calibration of the large number (1200) of AT-LAS magnetic field probes and NIKHEF was asked torealize a robust and easy to operate calibration benchbased on the CERN principle. As a concerted effort ofthe NIKHEF technical departments a state-of-the-artcalibration tool was realized.
Figure 1.2 shows the calibration head with four mag-netic field cards and one so-called coil card. The cali-bration principle requires the head to rotate uniformlyin a very homogeneous (0.001%) magnetic field avail-able at CERN. The magnitude of the homogeneousmagnetic field itself is continuously monitored by anNMR probe. The magnetic induction in the coils isused to determine the instantaneous orientation of thehead which allows to calibrate the three Hall sensors oneach of the four magnetic field cards. Alternatively, thehead can be moved in a stop-and-go mode in which casethe instantaneous head position is taken from decodersmounted onto the two orthogonal rotation axes. Trialruns began late 2003 and revealed some defects whichare under investigation at the moment. So far a pre-cision of 0.1% was reached compared to the expected0.01% precision. The installation of the calibrator inthe gap of the dipole magnet at CERN is shown inFigure 1.3.
H8 test beam
As in the summer of 2002, also in the summer of2003 the H8 muon spectrometer test beam setup was
4
the focal point of many NIKHEF activities. In 2003,a crucial new component in the test beam was theNIKHEF designed and constructed Muon Read-OutDriver (MROD) electronics card. The MRODs are thelast stage in the muon data acquisition chain, eachMROD serves the read-out of six MDT chambers. Thecommissioning of the MRODs in H8 proved less easythan anticipated. Nevertheless, little beam time waslost because during the first month of test beam run-ning the CERN accelerator complex suffered from re-peated breakdowns allowing ample time for MROD de-bugging. During the second half of the test beamperiod the MROD performance met the specificationsthereby validating the full MDT read-out chain to beused in the complete ATLAS muon spectrometer. Asin 2002, also in 2003 the detector control and align-ment components and read-out electronics, providedby NIKHEF, operated as specified.
From the data analysis point of view the test beamprovides a stimulating play-ground for in particular ourPhD students. Significant progress was achieved in thesimulation of the detailed muon induced signals in theindividual muon drift tubes. Similarly the full potentialof the front-end electronics (notably the time-slewingcorrection) was employed to further improve the reso-lution to a level below the original specification in theMuon Technical Design Report.
End-cap toroids
The two cold masses (CMs) are manufactured byBrush-HMA in Ridderkerk. Both CM’s were due tobe delivered and assembled at CERN in fall 2002.This schedule was shifted to a delivery end of 2003for the first CM and end of 2004 for the secondone. This scheme still matched the revised ATLAStime line. Winding of all 16 coils was completedend of 2003 and eight coils were impregnated bythat time (all coils for one CM). Also eight keystoneboxes were available after welding and machining ata subcontractor. Realization of the cooling circuitsfailed due to persistent problems in qualifying forthe (manual) welding of the aluminum cooling pipes.This work was subcontracted. After many trials anddespite expert advise the subcontractor concludedthat these welds could only be realized successfullyby orbital welding. The company had no experiencewith this type of welding and claimed extra costs foracquisition of the equipment, training of the weldersand a partial replacement of the aluminum circuitry bystainless steel. Also the aluminum extrusions were notappropriate for orbital welding because of variations
in the wall thickness. The last few years the CERNworkshop acquired a lot of experience in orbital weldingof aluminum pipes and fortunately it appeared possibleto realize the cooling circuits at CERN. End of 2003an agreement was reached to transfer the welding ofthe cooling circuits to CERN.
Central tracker
The central tracker employs three technologies for themeasurements of charged tracks: nearest to the inter-action point several layers with silicon pixel elements(the PIXEL detector); subsequently several layers withsilicon strip detectors (the SCT detector1) and finally anarrangements of drift tubes (the TRT detector2). Thecomplete central tracking volume sits within a largesolenoid which provides a 2 T magnetic field.
NIKHEF is involved in the silicon strip detector, SCT,project and in particular in the construction of one ofthe two SCT end-caps. NIKHEF’s main responsibilityis the procurement of 2× 9 high-precision carbon-fiberdisks which are used to hold the many silicon detectormodules. In addition NIKHEF constructs 80 silicon de-tector modules and at NIKHEF one complete end-cap isbuilt and tested. The other end-cap is built in England.In 2003 excellent progress was made in this very diffi-cult, multi-disciplinary and international project. Con-struction is in full swing at NIKHEF and the schedule isvery tight for in-time delivery of the complete end-capto CERN for installation in the ATLAS cavern in 2006.
SCT production
All 20 (including two spares) carbon-fiber disks whichhold the silicon strip modules have now been deliveredto NIKHEF. Early in the year, the manufacturer haddifficulty meeting our stringent flatness requirementsof 0.5 mm over the 1.2 m diameter. They refined theirprocessing several times so that finally they could reli-ably achieve all specifications.
Before mounting services and modules on the disk, sev-eral access holes have to be machined, followed by thegluing of hundreds of plastic mounting pads. The tool-ing for this was developed early in the year, with deliveryto England (RAL) of a prototype disk in January.
In order to position modules precisely and without riskof clashes, the mounting pads have to be the correctheight. Machining them to a variation of less than0.1 mm over the whole disk proved challenging. Themethod involves attaching the relatively floppy disk to a
1SCT is the acronym for Semi-Conductor Tracker.2TRT is the acronym for Transition Radiation Tracker.
5
Figure 1.4: Photo of a carbon-fiber support disk duringverification of the mechanical accuracy using the 3Dmetrology measurement machine at NIKHEF.
precise and stiff tool-plate, without distorting the disk.The mounting points can then be machined flat withoutthe disk bending or vibrating. Finally the required pre-cision was achieved when we realized the temperaturehad to be controlled to better than 0.5 oC to preventthe tool-plate from distorting. Figure 1.4 shows the fi-nal metrology inspection of a disk. The first productiondisk was delivered to England in October 2003.
NIKHEF will produce 80 of the 4000 modules in theSCT. These consist of two silicon-strip sensors gluedtogether with better than 5 µm precision, with each ofthe 1536 strips bonded to its readout electronics. As-sembly was delayed by the late delivery of version K5of the hybrid containing the readout electronics, butat last in June production began of five “qualification”modules. These were completed and tested, exceedingspecifications. This led to NIKHEF site-qualification inOctober and subsequently NIKHEF started mass pro-duction at the end of the year, reaching a peak rate offour modules per week. NIKHEF module assembly isexpected to be completed late in 2004.
NIKHEF will equip nine disks with services and modules(total 1000), and assemble the disks into a carbon-fiber cylinder to complete one of the two SCT end-caps.The tooling for placing modules was well on its way tocompletion at the end of 2003. This has a rotatingdisk-holder and a module grabber with cameras andmicrometer adjustment stages to align the module tothe mounting pins on the disk. The modules have tobe tested after mounting. The infrastructure for thisincludes cooling plant which was delivered in June, and
a large test-box which was started at the end of theyear.
The first disks should be assembled into the cylinder in2004. The tooling design for this made major advancesin 2003, with the main components ordered toward theend of the year.
Trigger, data-acquisition & detector control
The submission of the ATLAS trigger, data-acquisition(DAQ) and detector control system (DCS) TechnicalDesign Report (TDR) at the end of June constitutedan important milestone for the project. NIKHEF con-tributed in particular results from “paper” and “com-puter” models. Some further development of the exist-ing computer model was required. The modeling resultsclearly show that the design presented in the report inprinciple can satisfy the performance requirements. TheTDR has been formally approved.
The ROBIn prototype, a PCI board able to receive andbuffer two 160 MByte/s data streams and with a Giga-bit Ethernet network interface has become available fortesting in 2003. The board is the result of a commonproject, in which the University of Mannheim, Univer-sity of London (Royal Holloway) and NIKHEF partici-pate. In the course of the year it was decided that theROBIn boards will be located in rack-mounted PCs inthe USA15 underground area. At the end of the year itwas also decided that a new version of the board, ableto receive and buffer three 160 MByte/s data streamsand also with a Gigabit Ethernet network interface, willbe designed. During the year, configuration code for theFPGA of the current prototype has been developed.
Muon Read-Out Driver
In the summer of 2003 the MROD, the Read-Out Driverfor the muon chambers (MDTs), was deployed at theH8 test beam at CERN (see also Section 1.1). TwoMROD modules were used successfully for reading out12 MDTs. These modules use an older version of theADSP-21160 SHARC processor of Analog Devices. Us-ing the same design, for evaluation purposes a newmodule was built with the most recent version of thisprocessor. It has been found that some problems havebeen solved, but reliable communication between theprocessors is still only possible at half of the specifiedspeed (40 MByte/s in stead of 80 MByte/s per com-munication link with a 80 MHz processor clock). The“CSMUX” modules, originally developed as temporaryfacility for sending data from the TDCs on the cham-bers in the cosmic ray test stand to the MROD, havebeen converted into data generators. A CSMUX mod-
6
Figure 1.5: Photo of the MROD test setup. The9U VME crate at the left contains two MROD pro-totype modules. CSMUX modules in the crate at theright send simulated event data, formatted in the sameway as data output by the Chamber Service Modules(CSMs) on the muon chambers, to the MROD undertest.
ule can now emulate the data produced by a chamberand send these to the MROD via an optical fiber. Withthe modules performance tests with realistic data arepossible. A test setup is shown in Figure 1.5.
Operation of the MROD prototype above the maximumevent rate of 100 kHz has been demonstrated for smallevents. For events with realistic sizes it is expectedthat this rate can also be sustained, but more complexsoftware has to be developed to demonstrate this3. Inview of the limited headroom of the current design andon the basis of recommendations by a review panel, in2004 work on an upgrade of the design will be started,with the new Xilinx Virtex-II Pro FPGAs replacing theAltera FPGAs used in the prototypes. In first instanceoperation in the same way as the current prototype isforeseen. However, the new design will have a built-inupgrade possibility, as data transport via the communi-cation links of the SHARC processors can be replacedby transport via the up to 3 Gbit/s serial links of theXilinx FPGAs. This will result in a higher internal band-width of the MROD and reduce the processing load ofthe processors.
Detector Control System
In the detector control system (DCS) the EmbeddedLocal Monitor Board (ELMB) plays an important role
3This is due to the 40 MByte/s in stead of 80 MByte/s band-width available per SHARC communication link, making it nec-essary to transport the data across more links than in the testsdone. Fortunately the MROD prototype offers support for this.
as the main building block of the DCS front-end I/Oof the ATLAS sub detectors. The ELMB is also usedin other LHC experiments. In total nearly 10,000 ofthese boards are being produced. In 2003 the softwarewhich is delivered with the ELMB has been developedfurther by NIKHEF. This software enables the ELMBto be used “off-the-shelf”, or may be customized withlimited effort in case of special requirements. The soft-ware has several features aimed at minimizing the effectof radiation induced bit flips in memories and registers.The tolerance of the hardware for radiation and the ef-fectiveness of these features has been studied in severalradiation tests with satisfactory results.
From simulation to reconstruction to physics anal-ysis
The ATLAS software suite has matured considerablyduring the last year. The core software, called Athena,allows processing of events using detector services, con-ditional data, persistence etc. in a very modular fashionbased on C++. Proton collision events can be gener-ated with numerous event generators and single particleguns. Detector response is available using parameteri-zations (fast Monte Carlo) of the generated final stateparticles as well as by detailed detector simulation basedon Geant4. Enormous effort was put in validating theGeant4 program and work is ongoing to reach correctsimulation and digitization of all aspects of the ATLASdetector. Reconstruction algorithms are subsequentlyrun in the same Athena environment, completing thechain between generation of simulated events and fullreconstruction.
The most important tests of this complete softwaresuite are the ‘computing data challenges and a series oftest beams. In 2003 NIKHEF was actively involved inthe SCT and muon test beams at CERN as well as inthe upcoming data challenge in 2004 (data challenge 2or DC2).
In the ATLAS community the commissioning of the de-tector, both prior to data taking and during the firstrunning periods, receive a lot of attention. A fewmonths before the first proton collisions many sub-detectors will be tested, calibrated and aligned usingcosmic muons, whereas during the first running periodsa number of ‘standard’ physics channels will provide in-dispensable tools for these tasks.
Inner tracker and muon spectrometer software
NIKHEF is playing a major role in developing and de-livering key components of the inner detector softwareand bringing it into a production quality state for DC2
7
Distance to wire [mm]0 2 4 6 8 10 12 14
Re
so
luti
on
[m
m]
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
no slewing corrections
with slewing correction
Figure 1.6: The resolution of a Monitored Drift Tube(MDT) as a function of the distance to the wire withand without the time-slewing correction.
and combined test beam running. NIKHEF is coor-dinating the developments of the detector descriptionsoftware for the inner detector and providing the geom-etry for the SCT and interface to reconstruction for thepixel and SCT detectors. We have reached the goal ofusing a common source for detector description (Geo-Model) for Geant4 simulation, digitization and recon-struction and the full chain is operational. This detectordescription also includes the possibility of introducingalignment corrections and we have demonstrated thisfor the Pixel and SCT detectors. We continue to workon improving the realism of the SCT simulation andworking on improvements to the SCT digitization.
The same software infrastructure used in the full AT-LAS detector is also being used for the combined testbeam. We have successfully demonstrated the use ofcommon software components to simulate and recon-struct tracks in the test beam. During the test beamperiods we actively participated in the data taking ofthe SCT, alignment of the modules, and straight trackreconstruction. In particular we investigated the effectof inserting various targets in the beam-line, to inducea production vertex in the beam-line. Simulation ofthe set-up allowed to cross check the resolution of thespace points and indicated the detector alignment wascorrectly understood. Reconstruction of track segmentsin the SCT system including pattern recognition, usingATLAS standard software tools, is ongoing.
NIKHEF is also actively involved in the Muon subsys-tem software, where we contribute to calibration, sim-ulation, digitization and track reconstruction. We arecoordinating the efforts to set up and implement inthe Athena framework the calibration software for the
Radius muon hit (mm)0 2 4 6 8 10 12 14
Hit
s b
elo
w m
uo
n h
it
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2Contributions
In gas
In wall
Outside tube
Figure 1.7: The fraction of hits in a Monitored DriftTube (MDT) obscured by a δ-electron as a function ofthe distance to the wire.
complete muon spectrometer. We will also provide animplementation of the MDT calibration software, whichis the most complicated calibration of the four detectortechnologies used in the muon system. We participatedin the CERN H8 test-beam this summer, where (amongother things) a new feature in the MDT read-out elec-tronics, the so-called charge measurement, was intro-duced. With this charge measurement the time-slewingof the electronics can be corrected, and the resolutioncan be improved. The NIKHEF calibration software wasextended to include this new measurement.
Figure 1.6 shows the MDT resolution as a function ofthe distance to the wire with and without this new time-slewing correction. The average resolution improvesfrom 90 micron to 70 micron by using this new fea-ture. For the simulation of the MDTs we improvedthe handling of δ-electrons, which are the main causeof inefficiency. Figure 1.7 shows the fraction of MDThits that are hidden by δ-electrons, as a function ofthe distance to the wire. The various contributions toδ-electrons as generated in the Monte Carlo, after ourimprovements, sum up to the values as measured in ourcosmic ray test set-up.
Effort is still ongoing in the MDT digitization, where amore detailed description of the MDT response is mod-eled. In particular, the new charge measurement andits correlation with the timing measurement is mod-eled by generating the full signal shape. The current
8
Figure 1.8: Simulation of the top-quark signal in tt̄production at the LHC with one top quark decayingsemi-leptonically.
muon reconstruction software is not suited to recon-struct muons from cosmic origin, because it assumesthat muons originate from the interaction point andthis assumption is not valid for these cosmic muons.NIKHEF has started to develop muon reconstructionsoftware that can cope with muons that do not comefrom the interaction point, in order to fully benefit fromthe detector commissioning period.
Physics studies
One of our primary interests in ATLAS is the discoveryof the Higgs particle, and we have performed variousstudies on the Higgs particle in ATLAS. For example,after discovery of the Higgs particle we showed that itsCP state can be determined from its decay to τ -leptonpairs. We also estimated Higgs production uncertain-ties originating from uncertainties in the underlying par-ton densities.
In addition our group focuses on top physics for a num-ber of reasons. We have experience in top physics fromthe D∅ experiment, top quark properties provide inter-esting checks on the Standard Model, the top quarkproduction cross section is huge and will be used forcalibration studies and provides feed-back on the de-tector during the commissioning period, it provides anexcellent starting point for more involved Higgs studies,and NIKHEF plays a key role in theoretical aspects of(single) top quark physics.
NIKHEF is coordinating the validation of many genera-tors for top production and background events, and weare responsible for event production in DC2. We willobserve the top signal within a few days of data tak-ing and are testing the measurement robustness againstvarious detector commissioning stages.
[GeV/c]Tp100 200 300 400 500 600
[p
b /
(G
eV
/c)]
〉 T
/ d
pσ
d〈
10-3
10-2
10-1
1
10
102
103
|y| < 0.5
1.5 < |y| < 2.0
2.0 < |y| < 2.4
NLO (JETRAD) CTEQ6MmaxT = 0.5 pRµ = Fµ=1.3, sep R
Run II preliminaryOD
-1 = 143 pbintL
=1.96TeVs
data , Cone R=0.7OD
Figure 1.9: Cross section for jet production (cone algo-rithm with R = 0.7), as a function of jet pT , in variousranges of jet rapidity y , and a comparison to a NLOcalculation.
For example, in Figure 1.8 we show the single-leptonictop signal using MC@NLO with background comingfrom W plus 4 additional jets using the AlpGen gen-erator, after a couple of days of data taking. In thisfigure no b-tagging is assumed to be present, as a pes-simistic scenario for the startup period. We are involvedin the mass determination of the top quark, and theobservation of single top events in order to determinethe Kobayashi-Maskawa matrix element Vtb. Further-more we are in the process of developing strategies forthe unexpected at ATLAS. We are studying the excitingpossibility that with additional extra dimensions gravitybecomes strong at the TeV scale, observable in gravitonexcitations in the TeV range.
1.2 D∅ experimentRun II of the Tevatron is aimed at collecting an inte-grated luminosity of at least 2 fb−1 before first LHCcollisions. During 2003, the performance of the Teva-tron has further improved upon the 2002 performance,and D∅ collected data corresponding to an integratedluminosity of 150 pb−1, which combined with the 2002data gives more than 200 pb−1 for physics analysis.Further accelerator work is going on to improve per-formance, in particular the inclusion of the recycler forincreased anti-proton storage capacity.
Detector hardware
The detector is complete and largely commissioned, andtaking data with 85% efficiency. In 2003, a Level 1
9
Figure 1.10: Measured cross section for the productionof b-quark jets as a function of jet ET , and comparisonto a NLO calculation (MNR). The uncertainty on theNLO calculation is shown as dashed lines.
track trigger, CTT, became operational, and commis-sioning was started for a Level 2 displaced track trigger,STT. This trigger should be of significant help in trig-gering on particles containing bottom quarks. The per-formance of the tracker system (silicon strips and scin-tillating fibers) is good, and progress has been made intrack reconstruction software. The calorimeter is oper-ational, but has been suffering in 2003 from noise in-duced in the precision readout, leading to a non-optimalenergy resolution and occasional fake jets. Extensivestudies are going on to cure this in hardware and soft-ware.
The NIKHEF hardware responsibilities, a radiationmonitor for the silicon tracker, a beam loss monitorsystem, and a Hall probe magnetic field monitoringsystem, have operated stably. The forward protondetectors, for which NIKHEF has designed and madepositioning components, are now operational duringdata taking.
D∅ computingThe NIKHEF group plays a role in pursuing full-scaledata analysis outside Fermilab. With a much-improvedevent reconstruction program, all D∅ Run II data takenup to September 2003 have been reprocessed in the lastmonths of 2003, partly on NIKHEF computers. In orderto achieve this, the D∅ software was adapted to work onthe European DataGrid, and the D∅ data reprocessinghas been a first real data test of that Grid system. In2003, some seven million fully simulated Monte Carloevents were requested and produced. Data transfer toand from the D∅ data management system SAM is au-tomatic.
The D∅ agenda server that has been setup in Nijmegen
mass [GeV]4.5 5 5.5 6 6.5
eve
nts
/50 M
eV
0
10
20
30
40
50
60
70
-0.1 0 0.1 0.2 0.3
even
ts
10-1
1
10
102
-1
2
All events
Signal
Prompt J/psi
Missing dec.prods.
Unrelated K0
Proper Decay Length [cm]
Figure 1.11: Distribution of mass (left) and proper de-cay length (right) for reconstructed J/ψ K0S combina-tions, and the result of a combined mass-lifetime fit.The Bd → J/ψ K0S component resulting from the fit isshown as the shaded histogram.
to allow for a homogeneous interface to all of D∅ś meet-ings and to facilitate posting of relevant documentationis widely used.
D∅ software and physics analysisInteresting physics results from Run I data are still ap-pearing. Noteworthy is a new measurement of the topquark mass in the lepton plus jets decay channel, whichcombined with the other decay channels leads to thenew D∅ result mt = 179.0± 3.5± 3.8 GeV. This is themost precise individual measurement of the top quarkmass. The new top quark mass result is somewhathigher than the previous result, leading to an upwardshift in the allowed Higgs boson mass range in StandardModel fits.
Physics analysis of Run II data has progressed well in2003, boosted by the data reprocessing effort. Manyanalyzes now surpass the precision of the Run I ana-lyzes, and some are new for D∅. Results have beensubmitted to conferences on searches for new particlesand phenomena, W and Z boson production and decay,jet production (see Figure 1.9), B-meson productionand lifetimes, and top quark production.
The NIKHEF group is interested in pursuing top physicsand Higgs-boson searches, and look for new physics.In order to do so, students and staff are involved inthe development of software for b-quark-tagging, jetidentification, and lepton reconstruction. These toolshave been used in 2003 for electroweak and top andbottom quark physics.
An important milestone in 2003 was the successful de-fense of the first dutch D∅ PhD-thesis. The productionof b-quarks can be recognized by secondary vertex re-construction, track impact parameter tagging, and softlepton tagging (b → cℓν). A NIKHEF PhD-student
10
0
2.5
5
7.5
10
12.5
15
17.5
20
22.5
1750 1800 1850 1900 1950 20000
2.5
5
7.5
10
12.5
15
17.5
20
22.5Kidonakis NNLO-NNNLL (hep-ph/0303186)
Cacciari et al. (hep-ph/0303085)
mt = 174.3 ± 5.1 GeV/c2
D∅ Run I/II l+jets/topo (prel.)
D∅ Run I/II l+jets/µ (LP03, prel.)
D∅ Run I/II ll/topo
D∅ Run I/II alljets
√s (GeV)
σ (
pb
)
Figure 1.12: Measured cross section for tt̄ productionin various final states, in Run I (
√s = 1.8 TeV) and in
Run II (√
s = 1.96 TeV).
has measured the cross section for the production ofb-quark jets in the first Run II data, using soft muontagging. The measured b-jet production cross sectionis shown in Figure 1.10 as a function of jet ET , andis on the high side, but not incompatible with, a NLOcalculation.
Algorithms have been developed for track impact pa-rameter tagging, where signed impact parameters areused to calculate probabilities for tracks to originatefrom the primary vertex. These algorithms are beingused to measure the dynamics of b-quark production,with the aim to disentangle the contributing processesof flavor creation, flavor excitation and gluon splitting.Also, the algorithms are used to search for Z → bb̄, anobvious prelude to Higgs-boson searches.
A further analysis focuses on an exclusive b-decaymode, namely Bd → J/ψK 0S , which is one of a numberof decays that are interesting for studies of CP violationin the B system. In Amsterdam, a PhD-student hasreconstructed events in this decay mode and optimizedthe selection. The Bd lifetime was measured in theseevents from reconstructed B decay vertexes, as shownin Figure 1.11, as a first step toward CP violationmeasurements in D∅.A b-tagging algorithm based on secondary vertex prob-abilities has been used for the study of top-quark-pair(tt̄) production, and their hadronic decay into (at least)six jets. A first measurement of the tt̄ production crosssection in this all-hadronic decay mode has been made.
One PhD-student has finalized a NLO calculation ofthe single top quark production cross section, and hasmeasured the tt̄ production cross section in the elec-
Center of Mass Energy (TeV)1
)b
p(
rB
×Z,
Wσ
210
310
RunII
RunI
(e)∅D
CDF(e)
(e)∅D
CDF(e)
UA1
)µ(∅D
)µCDF(
)µ(∅D
)µCDF(
UA2
)πτ(∅D
)ρτ(∅D
+Xν l → W+X →pp
ll +X→ Z+X →pp
RunII Preliminary∅CDF and D
Figure 1.13: Cross section for W and Z production anddecay into electrons, muons, and taus final states.
tron plus jets decay mode. In this decay mode, and inthe muon plus jets decay mode, likelihood discriminantsare constructed to separate signal and background. Thett̄ cross section measurements are summarized in Fig-ure 1.12.
Further focus is placed on the identification of τ lep-tons using tracking and calorimetry. In Nijmegen, aPhD-student has measured the pp̄ → Z → τ+τ− crosssection, using a final state in which one τ decays toa muon and neutrino’s, and the other τ hadronicallyinto π±ν or π±π0ν. The result, together with othermeasurements of Z and W production and decay intoelectrons and muons, is shown in Figure 1.13. Further-more, students are looking to improve the trigger forτ ’s and jets.
For 2004, the Tevatron foresees an additional 300 pb−1
of data, and a total of 5-8 fb−1 of data is expected tohave been delivered by 2009. In 2003 it was decidedthat the D∅ trigger system will be upgraded to copewith the higher luminosities foreseen, but that D∅ willremain operating the current silicon micro-vertex de-tector until the end of Run II. A new inner layer of sil-icon strip detectors (Layer 0) will be installed in 2005,to improve the impact parameter resolution and coun-terbalance the radiation damage expected to occur inLayer 1.
11
250mrad
100mrad
M1
M3M2
M4 M5
RICH2
HCALECAL
SPD/PSMagnet
T1T2
T3
z5m
y
5m
− 5m
10m 15m 20m
TTVertexLocator
RICH1
The LHCb detector layout.
12
2 The B-Physics Programme
2.1 Introduction
The main focus of the B physics group of NIKHEFis the participation in the LHCb experiment at CERN.The group is also active in the Hera-B experiment atDESY and the BaBar experiment at the Stanford LinearAccelerator Center (SLAC) in Palo Alto, U.S.A. Giventhe limited performance of both the experiment andthe HERA accelerator, Hera-B will not contribute in asignificant manner to our understanding of CP violationin the B system. Consequently, the NIKHEF group hasdecided to end its contributions toward Hera-B. In orderto participate in a leading experiment on CP violation inthe B system, NIKHEF joined the BaBar experiment atSLAC. Various accurate measurements to pinpoint theorigin of CP violation can be made by BaBar, in thisway preparing the group for measurements in LHCb.The LHCb experiment at the LHC collider of CERN,is planned to come into operation in April 2007. Thedevelopment of the detector components for LHCb isgenerally proceeding on schedule. The same is true forsoftware development and the study of data analysismethods.
2.2 Status of HERA-B
After a decision of the directorate of DESY, Hera-B hasbeen shutdown by March 3rd, 2003. Until this moment,much less beam time has been available than was orig-inally allotted to Hera-B. In fact, only about between10 and 20 % of the envisaged data sample could berecorded in the period between November 2002, whenHERA operation was stable enough for regular data tak-ing with Hera-B, and beginning of March 2003, whenHERA was shut down for the luminosity upgrade.
Nevertheless, this amounts to about 300,000 J/ψevents and about 220 million minimum bias events.The data are being analysed under many differentaspects, but none of these analyses has yet beenfinalized. We give here an (incomplete) list of thesubjects which are at present under work:
1. J/ψ polarization and differential distributions inthe muon channel;
2. the b − b̄ production cross section through de-tached J/ψ → µµ events;
3. analysis of double semi-leptonic B-decays.
4. hyperon production, including Σ(1385), Ξ, Λ1520,Ξ∗1530, and others ;
5. double-φ production;
6. D-meson production;
7. A-dependence and xF -dependence of J/ψ produc-tion;
8. ratio of J/ψ and ψ′ production;
9. Υ production;
10. Bose-Einstein correlations;
11. upper limit on D0 → µµ. Our result appears to be2 to 3 times more sensitive than the present upperlimit.
12. Pentaquark analysis, including searches for Θ+ andΞ−−. Until now, we can only place upper limits onproduction cross sections, which are competitivein the light of the cross sections reported by otherexperiments.
13. χc - production ratio of χc → J/ψ+γ with respectto. direct J/ψ production.
It is worthwhile to mention that the detector has beenwell under control after the commissioning period pre-ceding the data taking period. This concerns the effi-ciencies and reliability of the different subsystems, andespecially the understanding and operation of the FirstLevel and Second Level triggers, which worked betterthan expected. At present, a major effort is investedin a better understanding of the efficiency map of theFLT, which is vital for the extraction of absolute crosssections.
2.3 Status of the BaBar Experiment
Introduction
At the end of 2002, NIKHEF joined the BaBar experi-ment (shown in Figure 2.1) at the Stanford Linear Ac-celerator Center in Palo Alto, U.S.A.
During 2003, the integrated luminosity recorded by theexperiment increased by 57 fb−1 to a total of 152 fb−1.Part of the large increase is due to the introduction inthe second half of the year of the almost continuous (or‘trickle’) injection into the low-energy ring of the PEP-II accelerator. Plans for next year include the same‘trickle’ injection into the high-energy ring of the accel-erator. Not only does this procedure bring the average
13
Figure 2.1: Front view of the BaBar detector duringmaintenance.
luminosity closer to the peak luminosity, but it also in-sures more stable running conditions of the accelerator.This in turn helps to increase the overall running effi-ciency, and thus the collected integrated luminosity. Asa result, current extrapolations predict that by 2007 asample of 500 fb−1 can be collected.
The increased luminosity collected during 2003 makes itpossible to consider various new analysis opportunities.This is also clear from the record number of 25 paperspublished in refereed journals during 2003, bringing thetotal number of such papers to 49.
Physics Results
Even though the emphasis of the BaBar experiment ison the measurements of CP asymmetries in B decays,one of the surprises of 2003 was the discovery of a newcharmed meson, the Ds(2317). The mass of this ex-cited state is lower than models had predicted. As aresult, it has dropped below the threshold for the decaymode expected to dominate, leading in turn to a greatlyreduced width. This discovery spawned renewed (theo-retical) interest in the modeling of the D meson system.In addition to this discovery, BaBar also published newlimits on D mixing.
Another first observation was that of the decay of B →π0π0. The importance of this mode lies in the fact thatthe magnitude of the branching fraction of this modedetermines how well one can relate the measured CPasymmetry in the decay B→ π±π∓ to the CKM angleα. Unfortunately, although small in absolute terms, therelatively large value of this branching ratio, (2.1±0.6±0.3) · 10−6, only allows one to constrain the difference
π−
Κ−π+
π+ ∝+
∝−
π−π+
B0→ ψ(2S)K0sB0→D*+π −
ψ(2S) → ∝+ ∝−
K0s→ π+ π−
D*+→D0π+
D0→Κ−π+
Figure 2.2: A fully reconstructed event as seen by theBaBar detector. One of the B mesons has decayedthrough the decay chain B0 → J/ψK 0S , followed byJ/ψ→ µ+µ− and K 0S → π+π− , while the other B hasbeen reconstructed as the decay B̄0 → D∗+π− , withD∗+→ D0π+ with the subsequent decay D0→ K−π+.
between the observed ’effective’ angle and the CKMangle to be less than 47o at 90% CL.
Fortunately, the same construction also holds for themodes B → ρ0ρ0 and B → ρ+ρ− . In this casethere is still only an upper limit on the branching ra-tio of B(B→ ρ0ρ0) of 2.1 · 10−6, which implies themuch stronger constraint that the angle α deviates lessthan 17o at 90%CL from the observed effective an-gle, making the determination of the CP asymmetry inB→ ρ+ρ− a priority for next year.The NIKHEF group has focused its interest on mea-surements which constrain the CP angle γ. The firststep in this program is the measurement of CP asym-metry in the mode B → D(∗)∓π± , where the decayis fully reconstructed. An example of such a recon-structed decay is shown in Figure 2.2. The CP asym-metry in this mode is proportional to sin(2β + γ), and,as sin 2β is well known from previous BaBar measure-ments, this measurement will thus constrain γ. Thesensitivity to sin(2β + γ) originates from the interfer-
14
ence between the two amplitudes which contribute tothis decay. The difficulty of this analysis lies in thefact that the CP asymmetry is not only proportional tosin(2β + γ), but also to r , the ratio of the two inter-fering amplitudes. Unfortunately r is too small to bemeasured directly in the foreseeable future, and mustbe obtained from other measurements and theoreticalarguments. An additional complication is due to thecoherent production of B and B̄ at BaBar: the pos-sibility of CP violation in the decay of the ’other’ Bmeson produced simultaneously with the signal decaymust be taken into account.
Notwithstanding these challenges, the first direct con-straints on the angle γ could be set (see Figure 2.3),and a paper was submitted to Phys.Rev.Lett.
The next step in this program will be to use the decayB → D∗∓ρ± . Due to the fact that the final stateconsists of two vector mesons, there are several ampli-tudes that contribute to this decay, depending on thehelicity of the final state. The additional observablesthus generated make the analysis more complicated,but do allow the determination the ratio of the ampli-tudes required to extract sin(2β + γ) without relyingon theoretical assumptions which would otherwise limitthe attainable accuracy.
2.4 The LHCb Vertex detector
This year significant progress was made on the pro-duction of the mechanical components for the vertexdetector. An important achievement was the produc-tion of the first prototype secondary vacuum box for thedetectors shown in Fig. 2.4. The top, side and end foilswere produced with the hot-metal-gas forming method,in which an 0.3 mm thick AlMg3 foil (an aluminum al-loy with 3% magnesium) is heated in a special moldin 4 hours to 350◦C, after which the foil is pressed byheated gas into the mold. A start has been made withthe verification of the required shape on a 3D measur-ing machine.Also the first prototype of the rectangular bellow hasbeen produced. The bellow consists of two sets of20 stainless steel shells. The fully welded assembly isshown in Fig. 2.5.
A large amount of production drawings for the me-chanical components of the Vertex detector havebeen submitted to the EDMS system at CERN. Thedrawings for the pipe frame stand, the Y-translationframe, the center frame and the detector support havebeen approved. Several other ones are under approval.
Figure 2.3: Constraints on the plane spanned by the ρand η parameters which appear in the Wolfenstein pa-rameterization of the CKM matrix. The definitions ofthe three CP angles α, β and γ are indicated. Theconstraints from the analysis of fully reconstructedB → D(∗)∓π± , combined with the results from par-tially reconstructed B → D∗∓π± are shown in colouraccording to their confidence level. In addition, the con-straints on the angle β from the measurements of CPasymmetries in the ’golden’ CP modes like B→ J/ψK 0Sare indicated. The region allowed by these two mea-surements of CP asymmetries is consistent with thearea obtained from the Standard CKM fit.
One of the principal components is the vacuum vessel.The results of the Finite Element Analysis have beenapproved by TIS (the Technical Inspection and SafetyDepartment at CERN).An issue that is still under discussion is if a VitonO-ring can be applied between the detector vacuumand the beam vacuum. In the present design such aring with small compression forces is required to avoiddeformation of the flanges, and consequently of thedetector housing. If the system needs to be ventedfor detector maintenance the permeation through suchan O-ring might saturate the NEG material in thebeam pipe. Both CERN and NIKHEF have thereforeperformed permeation measurements. From theNIKHEF results we concluded that the amount of gasthat permeates through the O-rings can considerablybe reduced if both the beam vacuum and detector
15
Figure 2.4: The first completed secondary vacuum box.
Figure 2.5: The prototype rectangular bellow.
vacuum will be vented with ultra-pure neon, and ifthe access to the detector system can be limited to 50hours. Under those conditions the detectors can beaccessed about 20 times before the NEG material has
Figure 2.6: The pipe frame stand and the Y-translationframe have been placed in the South hall of the formerPIMU building. Here, the complete system will be as-sembled and tested.
to be reactivated.From the approved items, the machine shops atNIKHEF and the VU have produced the pipe framestand and the Y-translation frame. The center frameundergoes its final machining. The detector supportis under construction. The whole system will beassembled in the South hall of the former PIMU-hall.The pipe frame stand and the Y-translation framehave been installed already, as shown in Fig. 2.6.
The VU has also produced several cylindrical vacuumvessels, shown in Fig. 2.7. They will provide realisticvolumes (1900 l for the primary vacuum system, 500l for the secondary one) to perform evacuation andventing simulation tests in order to optimize theevacuation and venting procedures. The vessels willalso be used for integral leak tests of the secondaryvacuum boxes and the bellows. Furthermore, two ofthem will be used as container for coating the outsideof the secondary vacuum boxes with a layer of NEGmaterial. This process will be performed at CERN.For controlling the venting and evacuation processes,membrane switches will be used. The reliability ofthese devices has been checked: their behavior re-mained constant within 0.1 mbar over 20,000 switchingcycles. This roughly corresponds with 800 evacuations.The switches have been modified at NIKHEF in orderto equip them with all-metal seals at the beam vacuumside. The complete set of switches that will be used inLHCb is shown in Fig. 2.8.
16
Figure 2.7: Test vessels for the evacuation and ventingsimulations. The complete volume is comparable tothe volumes in the actual set-up. With this system theevacuation and venting tests can be performed withrealistic volumes.
Figure 2.8: Membrane switch assembly to be used tocontrol the evacuation and venting procedures at LHCb.
A service agreement between the LHC vacuum group(AT-VAC) and NIKHEF is being worked on. CERNwill take the operational responsibility, NIKHEF willperform maintenance tasks. A functional specificationdocument from the AT-VAC is under discussion.
The kapton cables, that transport the signals from thesilicon sensors to the trigger electronics, have under-gone mechanical stress tests. The set-up is shown inFig. 2.9. The end parts of the cable were translated
Figure 2.9: Set-up for endurance tests of the flat cablesthat will be used in the vacuum system. The 30 mmdisplacement corresponds with the expected movementduring injection of the LHC beams.
Figure 2.10: The optimized wakefield suppressor whichproduces an almost perfect rf match between the beampipe and the detector vacuum boxes, both in open andclosed position.
over 30 mm. In part of the tests the end parts werealso moved by 5 mm in the vertical direction. Thiscorresponds to the expected displacement duringinjection of the LHC beams. After 40,000 cycles withonly the horizontal displacement and another seriesof 30,000 cycles with the combined one no cracks orother damage was observed.The optimized wake field suppressor as shown inFig. 2.10 eliminates almost all resonances, both in
17
open and closed position of the detector halves. Alsothe endurance tests for this wake field suppressor havebeen performed. Opening the wake field suppressor by30 mm (as required during injection) and a perpendic-ular displacement over 5 mm resulted in no damageafter 30,000 cycles.Extensive measurements have been done on the out-gassing from all components of the detector system.Therefore, samples have been tested from the kaptoncables and the components of the detector module(hybrid, TPG material, paddle base and heat connectorinterface). The dominant contribution comes fromthe paddle base and the flat cable connectors. Thetotal pressure in the detector vacuum is estimated tobe 3 × 10−5 mbar. This is quite acceptable for thesecondary detector vacuum.A start has been made with the construction of aCO2 cooling set-up. The water cooling and the freoncooling system have been installed. With this systemthe properties and cooling performance will be studied.Also extensive heat transfer model calculations havebeen performed. A first indication is that the TPGmaterial is sufficient to cool the detector, but thatthe heat connection interface needs a considerableimprovement. The contribution from the rf heatingof the secondary vacuum box is quite small. This isespecially due to the fact that the torlon coating at theinside of the secondary vacuum box strongly improvesthe emissivity of the box.
Development of a radiation hard front-end chipfor the vertex detector of LHCb
The development of a radiation hard front-end chipin 0.25 µm CMOS technology for the vertex detec-tor of LHCb is approaching its final stage. In a col-laboration between the ASIC-lab in Heidelberg andNIKHEF new submissions have been prepared and pro-duced. Presently, the 128 channel 40 MHz full-sizechip Beetle1.3, is under investigation in laboratory testsat Heidelberg, Zürich, CERN, Lausanne and NIKHEF.Beam tests and measurements with a radioactive sourceare performed on silicon sensor and hybrid prototypes,equipped with Beetle1.2 and Beetle1.3 chips. It hasbeen decided that one more iteration will be made, i.e.,a Beetle1.4 is going to be produced, while the produc-tion schedule remains within the overall VELO plan-ning.
2.5 Level-0 Pile-Up Veto System
The Level-0 Pile-Up Veto System is designed to rejectcrossings with multiple interactions at the first triggerlevel of LHCb. An LHCb Trigger Trigger Design Report
(TDR) with a description of the Pile-Up system hasbeen submitted in September, the approval of the TDRby the LHCC followed later.
Crucial elements of the system for which a close collab-oration with other institutes within LHCb now exists,are
• The Beetle chip for the readout of 2 planes of sili-con strip detectors. The 1.3 version Beetle chiphas been tested extensively with respect to thecomparator part for which NIKHEF is responsible
• The optical digital transfer of signals from thedetector to the processor system. A commonproject has started concerning all optical linkswithin LHCb.
At NIKHEF tests of the prototype of the processorboard have been concluded: it fulfills the requirementsof being able to run at 40 MHz. The work is now fo-cused on the second submit of the Pile-Up hybrid (sincethe first submit failed) and on the optical data transfersystem.
2.6 The Outer Tracker of LHCb
The Outer Tracker project has gone through all phasesleading to the launch of the mass production of thedetector modules:
• the engineering design of the detector has beenfinalised, presented to the LHCb collaboration andreviewed by an external panel in the EngineeringDesign Review (EDR) in May 2003;
• production tools have been designed and realizedand shipped to all production sites;
• contracts for the production of all detector ma-terial and parts have been signed and quality as-surance plans drafted. The flow of most materi-als toward the production centers passes throughNIKHEF;
• many detector prototypes have been constructedand tested. On the basis of those results, the pro-duction has been launched. Work instructions andQuality Assurance (QA) plans for the productionhave been written up and will be reviewed in theProduction Readiness Review (PRR) of May 2004.
Considerable progress in the Front-End Electronics, en-tirely NIKHEF responsibility, was made: two full pro-
18
Figure 2.11: Straw Preparation Tool.
totypes of the FE Electronics boxes have been con-structed and tested at NIKHEF. Moreover, the engi-neering design of the detector infrastructure (supportframes, detector services, etc.) has been started andforesees the assembly of 1/4 station (two layers of mod-ules, FE Electronics, frames and services) at NIKHEFbefore the end of 2004.
OT Module Production
Once the engineering design of the detector moduleswas reviewed and approved, of crucial importancebecame the completion of production and QA tools.NIKHEF played a central role, being responsible forthe design, production and commissioning of two ofthe main production tools, the Straw Preparation Tool(SPT) and the Straw Template Tool (STT), as wellas of the Wire Tension Meter, HV testing, and sourcescanning QA tools.
The SPT is used to “prepare” straws by cutting themto length, inserting wire locators, preparing a piece ofstraw at one end (“tongue”) for GND connection andsoldering this tongue ultrasonically. Three SPTs weredesigned and produced by the VU group, and deliveredand commissioned to all production centers.
The STT is essentially an alignment jig, where all 64straws forming a mono-layer are accurately aligned nextto each other, to be later glued to their supporting
Figure 2.12: Straw Template Tool aligned with laserinterferometry.
panel. The high intrinsic resolution of the drift-cells(about 200µm) requires very tight mechanical toler-ances (of the order of ±50µm) in the straw-alignmentpattern. The STT also plays a crucial role in defin-ing the detector planarity; this results in the requestof a jig planarity tolerance of ±100µm, by no meansa trivial requirement for objects about 40 cm wide andup to 5 m long. The High Tech Aerospace unit of thePhilips Enabling Technologies Group produced 5 STTs(3 of 5m and 2 of 2.5m) based on the design of theEngineering Department of NIKHEF. These jigs werealigned to the required accuracies with the help of laserinterferometry, as shown in Figs. 2.12 and 2.13.
Once production tools were ready, a long period ofprototype production and QA tests followed, at theend of which the first full-size (5m) modules, shownin Fig. 2.14, were produced at NIKHEF.
At that stage of the project, we concentrated on con-trolling the quality of the production in a stable and reli-able way. The Wire Tension Meter (WTM) tool, shownin Fig. 2.15, was designed and produced at NIKHEF:the tensions of the strung wires is measured by induc-ing a mechanical oscillation in each wire by means ofa short electrical pulse, and measuring the main res-onance frequency of the wire in a magnetic field. Wecommissioned three of these devices and delivered themto all production centers; the details can be found in the
19
Figure 2.13: Alignment precision of the STT along thewire coordinate (top) and flatness (bottom) achievedthrough laser interferometry.
LHCb Note LHCb-2004-034. Prior to closing a modulegas box, the quality of all strung wires in a mono-layer iscontrolled by checking that the tension measured withthe WTM falls in within a few grams from the nominalvalue (70 gr); wires outside this acceptance window arethen replaced.
Once the tension of all strung wires is checked, eachanode wire in a mono-layer is brought to high voltage(about 1600 V) in air and the leakage current (of theorder of 1 nA) of each wire is measured as a function ofincreasing HV with a precise computer-controlled cur-rent meter designed and realized at NIKHEF. A distri-bution of measured currents for a mono-layer is shownin Fig. 2.16; wires showing currents above few tens ofnA are then replaced.
Prior to gluing two straw mono-layers into a mod-ule, the wire pitch is checked at four positions forall 64 wires in a mono-layer with the aid of a specialcomputer-controlled tool developed at NIKHEF. Aftersome initial problems encountered with the first fewmodule prototypes, a precision in the pitch distribution
Figure 2.14: Gluing of the two mono-layers of an OTmodule (top); finishing of an OT module (bottom).
of the order of σPITCH ≃ 30µm was achieved. In thepitch distribution in Fig. 2.17 practically all wire pitchesfall in a window of ±50 µm around the nominal (5.25mm) value.
Two straw mono-layers are then glued into a moduleand the gas box is sealed. After a curing time of about19 hours at 20 oC, the module is filled with gas andthe gas tightness of the module is checked; at the endof the procedure, volume losses below 10−6 l/s are ob-tained (significantly better than the specified 5% lossper volume exchange).
After modules are flushed with the counting gas for afew days, they are systematically tested by scanningtheir entire surface with a 90Sr source and measuringthe resulting leakage current. A current distributionfrom the 90Sr scan over the surface of a half module is
20
Figure 2.15: Wire Tension Meter measuring the wiretensions of a straw mono-layer.
0
2
4
6
8
10
12
14
16
18
20
10-2
10-1
1 10 102
103
Current (nA)
Entr
ies
V = 1600V
Panel: FP016 B U
Gas : air
Figure 2.16: Distribution of measured currents of amono-layer in air, at HV = 1600 V.
shown in in Fig. 2.18: the typical patterns due to thepresence of the wire locators are easily recognized.
In addition, dedicated investigation of the pulse heightdistribution from an 55Fe source are carried out forgiven sub-portions of the module surface. The gain ofeach straw channel can thus be determined, as shownin Fig. 2.19.
Wire
Xw
ire -
X(w
ire-1
)
4.9
5
5.1
5.2
5.3
5.4
5.5
5.6
0 10 20 30 40 50 60
Figure 2.17: Wire pitch distribution of a straw mono-layer: the red band is ±50 µm, and the green one ±100µm.
20
40
60
25 50 75 100 125 150 175 200 225 25020
30
40
50
60
70
80
90
10
P iti ( )
Wir
e
V = 1650V
Panel: 5012 B L (nA
Figure 2.18: Leakage current over the surface of a halfmodule from a 90Sr scan at HV = 1650 V.
Front-End Electronics
The extremely compact design of the FE Electronicsforesees the complete digitisation of the drift-time datato take place on board the FE Electronics, where dataare also serialized and optically transmitted on opticalfibres to the readout electronics in the counting room.NIKHEF is responsible for the design, production andcommissioning of the Outer Tracker FE Electronics, aswell as for the design and realization of all related ser-vices: Timing and Fast Control (TFC), Electronic Con-trol System (ECS), Low and High Voltages supply (LVand HV), analog monitoring etc.
We completed the design of the FE Electronics, which isto be globally reviewed by an independent panel in April2004. We also constructed two complete prototypesof the whole FE Electronics box, shown in Fig. 2.20,including among the others:
• HV boards, containing the HV-decoupling capaci-tors buried in the PCB layers;
• the ASDBLR service boards, where the small straw
21
0
50
100
150
200
250
-0.2 -0.15 -0.1 -0.05 0Peak Minimum (V)
55F
e puls
es
Wire: 14
0
50
100
150
200
250
-0.2 -0.15 -0.1 -0.05 0Peak Minimum (V)
55F
e puls
es
Wire: 15
0
1000
2000
3000
4000
5000
6000
7000
0 10 20 30 40 50 60Wire
Gain
Figure 2.19: Pulse height distribution from two anodewires measured with an 55Fe source (top); channel gainsof a straw mono-layer (bottom).
Figure 2.20: Prototype FE Electronics box mounted onits support and cooling chassis.
pulses get amplified and discriminated by the AS-DBLR ASICs;
• the OTIS service boards, where the time delay be-tween the digital straw signals passing over a cer-tain threshold (typically of few thousands electroncharges) and the LHC experimental clock gets digi-tised by the OTIS ASICs;
• the mechanical support and cooling frame and theshielding aluminum box.
Vthreshold (mV)600 650 700 750 800 850 900
rela
tiv
e E
ffic
ien
cy
(%
)
0
10
20
30
40
50
60
70
80
90
100
thresholdASDBLR Efficiency vs. V
Chip 5442 Channel 8
ASDBLR Noise behaviour
= 3.4 fCinput
Q
Vthreshold (mV)600 650 700 750 800 850 900
Inp
ut
Ch
arg
e (
fC)
1
2
3
4
5
6
7
8
at 50% eff.threshold
ASDBLR Input charge vs. V
Chip 5442 Channel 8
ASDBLR Charge calibration
Figure 2.21: Threshold (top) and gain curve (bottom)of a readout channel.
All boards were tested and qualified. E.g. in Fig. 2.21,the measured threshold and the gain curve of a strawchannel is shown, from which the electronics EquivalentNoise Charge (ENC) can be determined.
2.7 Reconstruction and Physics Studies
Detailed Monte Carlo simulation studies are carried outwith a twofold purpose:
• to develop reconstruction algorithms for future realdata,
• to validate the final modifications in the proposedLHCb detector set-up.
The final layout of the experiment is shown in the figureon page 12.
Event Generator and Detector Simulation
Minimum bias proton-proton interactions at a center-of-mass energy of
√s = 14TeV are generated with the
PYTHIA program. A sample of bb̄-events is obtainedby selecting events with at least one b or b̄ hadron in thefinal state. The decay of all unstable particles is per-formed with the QQ program developed by the CLEOcollaboration. In this procedure both the PYTHIA and
22
QQ parameters have been tuned using available pub-lished data in order to reproduce the expected environ-ment of the LHCb machine in the LHCb interactionpoint.
Generated particles are subsequently traced through theLHCb detector using the GEANT package, simulatingall interactions between these final state particles andthe detector materials. In the simulation program theentrance and exit points of each particle traversing asensitive detector layer are registered, together with theenergy loss in that layer. This information is then usedto mimic the detector “raw data”, taking into accountthe details of the sensitivity and response of each de-tector, including detection inefficiencies and electronicnoise.
Reconstruction
In the track reconstruction program the registered hitsof the VELO, the TT, the IT and the OT detectors arecombined to form particle trajectories across the detec-tor. For efficient track finding two complementary ap-proaches have been used. In the first approach a trackcandidate is searched in the VELO and extrapolatedto the T stations where hits matching the track areadded. In the second approach an independent trackseed is searched in the T stations which is matchedto unused track in the VELO detector. The overallprocedure leads to an efficiency of 95% for B decaytracks with a corresponding ghost-track rate of 9%.The quality of the reconstructed tracks is indicated bytheir momentum resolution and their impact param-eter mismatch at the track vertex. Distributions forthese parameters are shown in Fig. 2.22. For B decaytracks the average resolutions are < σIP >= 40 µmand < δp/p >= 0.37%.
(prec − ptrue ) / ptrue
0
250
500
750
1000
1250
1500
1750
2000 a) σ=0.37%
−0.04 −0.02 0 0.02 0.04IP mismatch [mm]
0
100
200
300
400
500 b) = 40 ∝m
0 0.05 0.1 0.15 0.2
Figure 2.22: (a) Momentum resolution with a Gaussianfit, and (b) impact parameter precision, for B-decaytracks.
Figure 2.23: (a) Kaon identification efficiency (solidpoints) and pion misidentification rate (open points)as a function of momentum.
Particle identification within LHCb is provided by thetwo RICH detectors (pion, kaon and proton identifi-cation), the Calorimeter system (electron and photonidentification) and the Muon Detector (muon identifi-cation). A full pattern reconstruction procedure is alsoput in place for these detectors. For all reconstructedparticles likelihoods are defined for each particle hy-pothesis. The performance of the kaon identification,the most difficult task, is plotted in Fig. 2.23.
Event Selection
The main challenge in the offline selection of B decayfinal states is to maintain a high efficiency for the signalB decay events, while providing a very large rejectionfactor for the combinatorial background. The offlineselections of a large number of B0 and Bs decay chan-nels are studied in detail using the simulated events. Ingeneral the selection criteria make use of:
• a precise invariant mass requirement of recon-structed B mesons,
• a high resolution measurement of the B decay ver-tex that is inconsistent with the primary event ver-tex,
• a particle identification consistent with the re-quired decay particles.
23
Channel (c.c included) efficiency yieldB0 → π+π− 0.69% 26kB0s → D−s π+ 0.34% 80kB0s → D∓s K± 0.27% 5.4kB0s → J/ψ(µ+µ−)φ 1.67% 100k
Table 2.1: Estimates for the total efficiencies and theannual yields of benchmark b-hadron decays for the re-optimized LHCb detector.
Table 2.1 lists the total efficiency and the expectedyearly event yield for four benchmark B decay chan-nels.
In order to demonstrate the robustness of the simula-tion and reconstruction, the whole procedure has beenrepeated by using conservative settings of all resolu-tion and efficiency parameters simultaneously in theprogram. The expected loss of events as compared tothe standard procedure is ∼ 30%, which is consideredacceptable for this unlikely hypothetical situation.
CP Sensitivity
The expected sensitivities to CP observables have beenassessed with “fast Monte Carlo” programs, using theefficiency and resolution parameters obtained by the fullsimulation procedure.
The NIKHEF group focused on three benchmark decaysof the Bs meson:
• the decay channel Bs → D−s π+; a flavor specificdecay that can be used to measure the mass dif-ference (∆ms) between the CP-even and CP-oddeigenstates of the Bs meson. This decay mode isalso used as a calibration channel to determine theexperimental mis-tag fraction ωtag directly fromthe data. The expected decay time distributionsfor one year data-taking are shown in Fig. 2.24.The highest value of ∆ms that can be observedwith 5σ statistical significance is 68 ps−1, far be-yond the expected value in the Standard Model.
• the decay channel Bs → D∓s K±; a decay in whicha large CP violation observation is expected. Ameasurement of the time dependent asymmetries,as shown for simulated events in Fig. 2.25, leadsto a determination of the unitary parameter γ′ =γ − 2χ.
• the decay channel Bs → J/ψφ; a decay in whichthe Standard Model predicts no CP violation. Anon-zero observation of CP violation, labeled by
Even
ts
100
200
300
−1 = 15 pssm∀
t [ps]0 1 2 3 4 5 6
Even
ts
100
200
300
−1 = 25 pssm∀
Figure 2.24: Proper-time distribution of simulatedDs → D−s π+ candidates, for two different values of∆ms . The data points represent one year of data, whilethe curves correspond to the maximized likelihood.
the parameter 2χ, would indicate the presence ofnew flavor changing interactions.
A measurement of the time dependent asymmetries ofthese three decay modes results in a measurement ofthe CKM parameter γ with an expected error of 14 de-grees after one year of data-taking, without theoreticaluncertainty.
−0.5
−0.25
0
0.25
0.5
Asy
m (
Ds −
K+)
t [ps]
Asy
m (
Ds +
K−)
−0.5
−0.25
0
0.25
0.5
0 0.5 1 1.5 2 2.5 3 3.5 4
Figure 2.25: Time-dependent Bs -Bs asymmetry of sim-ulated D−s K
+ (top) and D+s K− (bottom) candidates
for a value of ∆ms = 20ps−1 and γ′ = 65 degrees.
24
3 Heavy Ion Physics
3.1 Introduction
The strong interaction is one of the four fundamentalforces in Nature. Quantum Chromo Dynamics (QCD),part of the standard model of particle physics, is a verysuccessful microscopic theory of the strong interaction.It successfully describes the “zoo” of observed hadronsand, at large momentum transfer, the interactions oftheir constituent quarks and gluons. However, despiteintense experimental efforts no free quarks or gluonshave ever been observed. This so called confinement ofquarks and gluons inside a hadron is one of the mostinteresting and puzzling properties of the strong inter-action and is one of the features of QCD that from firstprinciples is still poorly understood. At an energy den-sity larger than in a proton, QCD predicts the existenceof a deconfined form of matter, called a Quark GluonPlasma (QGP), where the quark and gluon degrees offreedom are liberated instead of being confined withinhadrons. The transition between confinement at lowtemperature and deconfinement at high temperaturemay provide better understanding of the fundamentalproperties of the strong interaction. Colliding heavy-ions at the highest energies available is expected to bethe best possibility to create and study such a largehigh temperature system in the laboratory.
NIKHEF has been actively involved in fixed targetexperiments at the CERN SPS (WA98, NA49, andNA57), which still bear new results. The major focusof the experimental activity in the field has, however,shifted to the Relativistic Heavy Ion collider (RHIC) atBrookhaven National Laboratory. The NIKHEF heavyion group is actively participating in data taking andanalysis in the STAR experiment. Moreover, the groupcontinues its effort in the preparation of the ALICEdetector at the future LHC collider. In this reporthighlights of all these activities will be summarized.
3.2 Results from SPS experiments
The WA98 experiment has been able to extract thefirst direct photon measurement in heavy ion collisions,which created a lot of theoretical interest, because di-rect photon spectra contain unique information aboutthe early dense phase of the reaction. The system-atic uncertainties allowed a reliable extraction only atrelatively high transverse momentum. Now interfero-metric methods have been used to estimate the yieldof direct photons at low transverse momenta [1]. Themeasured source radii are similar to those obtained withcharged pions, which is consistent with an emission of
10-7
10-6
10-5
10-4
10-3
10-2
10-1
1
10
102
0 1 2 3 4
lowest value
most probable value
Correlation method
Subtraction method
Upper limit,
subtraction method
Predictions
(Gale & Rapp)
Hadr. Gas
QGP
pQCD
Sum
pT (GeV/c)
E d
N/d
3p
(G
eV
-2)
WA98
Pb + Pb central
Figure 3.1: Direct photon yields extracted from inten-sity interferometry measurements in central Pb-Pb col-lisions at 158AGeV studied in the WA98 experiment.
these photons from the hadronic phase. In Fig. 3.1 theestimated yields are compared to those obtained fromthe earlier measurement (subtraction method). Alsoshown are theoretical calculations including both QCDprompt photons and thermal photons emitted from ahadron gas or a QGP. While these calculations can rel-atively well describe the data at intermediate and hightransverse momenta pt , they are considerably below thenew low-pt results.
To search for possible signals of the onset of decon-finement, the NA49 experiment has taken central Pb-Pb data at five different energies in the years 1996–2002. Preliminary results from the 2002 runs at 20 and30 AGeV supplement those from 40, 80 and 158 AGeVpublished earlier [2]. The new data confirm the sharpmaximum at about 30 AGeV in the ratio of K+/π+
yields as shown in Fig. 3.2. The curves in this figureshow predictions from a statistical hadron-gas model [3]and from two microscopic transport models [4, 5]. It isclear that, at least at present, hadronic models fail todescribe the data. More results from this energy scanhave been obtained, e.g. it is observed that the inverseslope of kaon spectra is constant in the energy rangeof the SPS. The ratio of Λ/π, also measured by NA49
25
(GeV)NN
s
10 102
〉+
π〈/〉+
K〈
0
0.1
0.2
RQMD
URQMD
Hadron Gas
Figure 3.2: Energy dependence of the ratio K+/π+ incentral Pb-Pb collisions from AGS (triangles), NA49(squares) and RHIC (Au-Au collisions). The curvesshow predictions from three different hadronic models.
[6], shows a similar energy dependence as the K+/π+
ratio.
Figure 3.3 shows the ratio Es of total strangeness (Kaonplus Lambda) to pion yields versus the Fermi energymeasure F ≈ s1/4. Also plotted is the prediction fromthe SMES model [7] which assumes a phase transitionto the QGP at SPS energies. The data are in fair agree-ment with this model.
If the anomalies observed by NA49 do indeed indicatea phase transition at SPS energies or that they canjust be explained in a hadronic scenario is still an openquestion.
3.3 Jet production and jet quenching stud-ied in STAR
High-pt partons are produced in the very early stage ofa collision. When traversing a dense system they areexposed to the color charges in the medium. Similarto QED bremsstrahlung the partons will lose energy, aphenomenon called jet-quenching. This makes them anideal probe to study the initial condition of heavy-ioncollisions. Because the jets resulting from these high-ptpartons are difficult to isolate from the soft backgroundin an heavy-ion collision, it was proposed to measurethe leading particles which typically take about half theenergy of the jet. In fact, one of the earliest observa-
)1/2
F (GeV
6420s
E0
0.1
0.2
0.3
NA49
AGS
p+p
Figure 3.3: Energy dependence (F ≈ s1/4) of the ratioEs of strangeness to pion yields in Pb-Pb collisions. Thecurve is the prediction from the SMES model. Protondata (open symbols) are shown for comparison.
tions at RHIC
