FELIX A full acceptance detector at the LHC

11 Pages·0.29 MB·English

by  V. Avati

1
FELIX
A full acceptance detector at the LHC
V. Avati, K. Eggert,a and C. Taylor b
(for the FELIX Collaboration)
Presented by K. Eggert
8
9 aCERN, Geneva, Switzerland
9
1 bDept. of Physics, Case Western Reserve University, Cleveland, OH 44106 USA

n
a The FELIX collaboration has proposed the construction of a full acceptance detector for the LHC, to be
J
located at Intersection Region 4, and to be commissioned concurrently with the LHC. The primary mission of
3 FELIXis QCD:toprovidecomprehensiveand definitiveobservations ofa verybroad rangeof strong-interaction
2
processes. Thispaperreviewsthedetectorconceptandperformancecharacteristics, thephysicsmenu,andplans

for integration of FELIX into the collider lattice and physical environment. The current status of the FELIX
1
Letterof Intentis discussed.
v
1
2
0 1. Introduction
1
0 FELIX will be the first full acceptance detec-
h d
8 tor at a hadron collider. It will be optimized for c/
9 studying the structure of individual events over dN
/
x all of phase space (see Figure 1). FELIX will
e observe and measure all charged particles, from
-
p the central region all the way out to diffractive
e protons which have lost only 0.2% of their initial
h energy. Itwillevenseeelasticprotonswhichhave
v: amomentumtransferofatleast10−2GeV2. This
Xi comprehensive,precisiontrackingisaccompanied h d LHC, inelastic collisions
by equally superb electromagnetic and hadronic E/
r d
a calorimetry. FELIX will observe and measure
photons and neutrons down, literally, to zero de-
grees, giving it an unparalleled ability to track
the energy flow. In contrast, the other LHC de-
tectors are sensitive over only a fraction of phase
space and see less than 10% of the typical en-
ergyflow. FELIXis thus uniquely able to pursue
physics complementary to that of the other de- h
tectors planned for the LHC.
TheFELIXdesigninvolvesthecoordinatedar-
rangement of three distinct systems: the mag-
netic architecture responsible for getting the
Figure 1. The pseudorapidity distribution of
beams through the I4 straight section, the track-
charged particles and of the energy-flow at
ing system, and the calorimetry. Each system
√s=14 TeV.
must be complete in its own right, without com-
2
Figure 2. Sketch of the FELIX experiment in central and very forward region.
promising the characteristics of the other sys- the point ofview oftransversespace. The beams
tems. The magneticaperturesmustnotbe limit- are separatedby 42 cm at the location of the RF
ingaperturesofeitherthetrackingorcalorimeter cavities,providingroomforzerodegreecalorime-
systems. There must be sufficient physical space try. Since the existing infrastructure, including
for both tracking and calorimetry. The calorime- the ALEPH solenoid, can be re-used with mini-
tersmustbephysicallylargeenoughtohavegood mal modifications, I4 is clearly a superb location
resolution,andmustnotinterferewith either the for a full acceptance detector. (The central part
tracking or the magnetic systems. of FELIX, which nicely fits into the existing cav-
Allofthisrequiresalotofspace,andthedetec- ern,andtheextensionsupstreamintotheforward
tormustbecarefullyintegratedintothedesignof regions, are shown in Figure 2.)
the machine. Full acceptance cannot be achieved Nevertheless,the task ofintegratinga detector
by“addingon”tocentraldetectorsoptimizedfor with genuinely full acceptance into the available
high p physics. Here FELIX is fortunate. The space at I4 is not trivial. The FELIX Letter of
T
decision to split the RF cavities at I4, moving Intent[1]outlineshowitcanbe done,usingwell-
themto 140mfromthe interactionpoint(IP), understood magnets and compact detectors, for
±
combined with the fact that FELIX’s “low” lu- acomparativelymodestprice: weestimateacost
minosity permits the focusing quadrupoles to be of about 25 MCHF for the machine magnets and
movedmorethan120mfromtheIP,providesthe the infrastructure, and about 50 MCHF for the
necessarylongitudinalspace. I4isalsoidealfrom detector outlined here and presented in more de-
3
tail in the FELIX LoI. Given this richness of phenomena, and given
the importance of QCD to the interpretation of
thenew-physicsdataexpectedtoemergefromthe
2. Physics Overview
LHC, it is clearly very important to improve the
The heart of the FELIX physics agenda is data-base with an LHC detector and experimen-
QCD: FELIX will be the ultimate QCD detector tal group fully dedicated to the observation and
at the LHC. interpretation of as broad a range of QCD phe-
Surprisingly,theneedforsuchadetectorisnot nomena as possible. This is ofcoursethe mission
obvioustomanymembersofthehighenergycom- of the FELIX initiative.
munity. In part, this may be because of the suc- Many of these new opportunities in QCD
cess of the interplay between theory and exper- physics at the LHC are not well known, and the
iment in the case of electron-positron collisions. FELIXcollaborationhasaccordinglyplacedhigh
The cleanliness of the process, together with the priority in in providing a description of them in
low event rate and full-acceptance capability of the FELIX LoI. We briefly summarize a few of
the detectors, has led to an especially fruitful in- the main themes here.
teraction between the QCD aspects of that ex-
perimental programwith the remainder. 2.1. Parton densities can be measured to
The case of hadron-hadron collider physics is extremely small x, below 10−6
quite different. The high-p , low cross sec- The partondensities at small x are themselves
T
tion physics is accessed by highly selective trig- a very important thing to measure. Up to now
gers. The phase-space acceptance of the detec- HERA has provideddata downto x values ofor-
tors is largely limited to the central rapidity re- der10−4forQ2intheperturbativedomainofsev-
gion. Fullacceptancehasnotbeenattainedsince eralGeV2. FELIXwill havethe capability to ex-
the bubble-chamber era of fixed-target physics. tend these measurements to x values below 10−6
Therefore the basic data base is much more lim- via observationofdileptons, low-massdijets, and
ited. low-massjet-photonsystemscarryinglargelongi-
This situation is all the more serious because tudinalmomenta. Inthis regimeoneexpects(es-
of the great variety in event classes for hadron- pecially for proton-ion collisions) the breakdown
hadron collisions. There are soft collisions with of the usual DGLAP/BFKL evolution-equation
large impact parameters; angular momenta of formalism and significant nonlinear effects to be
tens of thousands instead of the unique J = 1 observed.
of the e+e− world. Central collisions produce
much higher multiplicities than are seen in e+e− 2.2. Minijet production in hadron-hadron
annihilation. There are the diffraction classes of collisionsisstronglyenergydependent
events, with and without jet activity, that com- TheneedforavastlyimprovedQCDdata-base
prise several to tens of percent of typical sub- for hadron-hadron collisions is made even more
samples (if seen in full acceptance) and which urgent by the fact that qualitative changes are
present a major challenge to theory. There are expected even in the structure of generic events
poorly understood strong Bose-Einstein-like cor- becauseoftherapidincreasewithenergyofgluon
relations seen at very low p and low relative p parton densities in the primary protons. Thanks
T T
in hadron-hadron collisions which do not occur to the measurements at HERA, this is not only
ine+e− collisions. But atcollider energiesthis is thetheoreticalexpectationbutalsoadata-driven
only based on one sample of low-p data from one. The parton densities at a 5 10 GeV scale
T
−
UA1, because until now no other detector has become so large that minijet production in cen-
had the measurement capability. Finally, there tral collisions may become common place, with
is little if any data in the forward fragmentation minijet p large enough for reasonably clean ob-
T
regions,wherecosmicrayexperimentsinsistently servability. These very high parton densities cre-
claim that anomalies exist. ate, at a perturbative short distance scale, “hot
4
spots” in the spacetime evolution of the collision nomena associated with the conjectured fluctu-
processwithinwhichtheremaybethermalization ation of the initial-state projectile into a trans-
or other nonperturbative phenomena not easy to versely compact configuration, which therefore
anticipate in advance of the data. Particle spec- interacts with an unusually small cross section.
trathemselvesmayevolvetosomethingquitedis- Evidence for this is seen in vector-meson pho-
tinct from what has been so far observed, with toproduction at HERA, especially J/ψ produc-
strangeness,heavyflavors,and/orbaryonandan- tion, which exhibits the expected rapid increase
tibaryonproductionenhanced. Especiallyincen- of cross section with energy. Also at Fermilab,
tral proton-ion collisions, where the total gluon- diffractiondissociationofa highenergypioninto
gluon luminosity per collision is maximized, and dijets, with all the initial pion energy going into
wheretheevolutionofasingleprotonfragmentis the dijet system, is being studied by experiment
followed, one can expect this class of phenomena E791. Exactlythesameprocessisavailableatthe
to be most prominent and surprises most proba- LHC with FELIX, as well as a similar process
ble. where one beam proton dissociates diffractively
into threejets, oneforeachquark. TheAdepen-
2.3. Diffractive final states are endemic, dence of these processes is remarkable, roughly
many are important, and some are A4/3, because this diffractive process should oc-
spectacular cur even in central collisions,thanks to the small
Diffractive final states will comprise almost size of the initial configuration.
50% of all final states at the LHC. The soft
diffractionatverylargeimpactparameter,which 2.4. Particle production from deep within
perhaps sheds light on pion-cloud or glueball the light cone may exist and deserves
physics, is at one extreme, and hard diffraction, careful searches
where rapidity gaps coexist with jets, is at the The existence of events with a very high final-
other. There are a large variety of hard diffrac- state multiplicity ofminijets andtheir associated
tionprocesses,includingsomewithtwoandthree hadrons has other implications. The products of
rapiditygaps,whichareofbasicinteresttostudy. such interactions for the most part can be ex-
In this class there are expected to be, for exam- pected to explode from the initially compact col-
ple, an extraordinary class of events where the lision volume in all directions at the speed of
completeeventconsistsofacoplanardijetaccom- light. Because of the high multiplicity density,
panied by the two unfragmented beam protons the time of hadronization of all these degrees of
detected in Roman pots, and absolutely nothing freedom will be lengthened from the usual low-
else in the detector. Certainly ATLAS and CMS energy value of 1-2 fm to several fm. Up to this
can also detect such events, provided they sac- time of hadronization, the expanding “fireball”
rifice a luminosity factor of about 30 relative to containingmostofthepartoniccollisionproducts
their hard-earned peak luminosity. However, to is arguably a rather thin sphericalshell, of thick-
really understand this event class, one will need, ness of order a fm. So even before hadroniza-
atthe veryleast,to examine the t-distributionof tion there is a large interior volume of hundreds
the Roman-pot protons, as well as to study the of fm3, isolated from the exterior vacuum, which
generalizations of this process to the cases where may evolve toward a chirally disordered vacuum.
one orbothofthe protonsundergoessoftdiffrac- Consequently in such events there might be a
tiondissociationtoalowmassresonanceorahigh large pulse of semiclassical,coherentpions of rel-
mass continuum, orto a high-p systemcontain- atively low p emitted when this false vacuum
T T
ing a tagging jet. Only FELIX would have such eventually decays: disoriented chiral condensate.
a capability. This is at present only a speculative possibility,
Inadditionto this classofharddiffractionand although experimental searches,especially in the
very soft diffraction processes, there is another context of ion-ion collisions, are underway.
very interesting class ofsemiharddiffractive phe- Moregenerally,onemayask: ifdisorientedvac-
5
uum is not what is in the interior of this quasi- gree measurements of nuclear fragments and by
macroscopicfireball,whatis? Iftheinterior“vac- theamountoftransverseenergyproduced. Atthe
uum” is broken into domains of various chiral LHC, the FELIX instrumentation in the forward
orientations, then topological obstructions might direction allows a data-driven approach for at-
lead to production of (Skyrmionic) baryons and tacking the problem by the former method. The
antibaryons of unusually low p . And if there large yield of minijets, strongly dependent upon
T
is activity deep inside the light cone, no matter impact parameter, may allow the latter method,
what it is, then this activity has eventually to based upon transverse energy production, to be
be turned into emission of particles; hence a new used more effectively at the LHC (by all detec-
particleproductionmechanismwhichdeservesto tor groups)because of the strongercorrelationof
be studied. It would seem that the only alterna- multiplicity withimpactparameterthanatlower
tive available for the absence of new phenomena energy. A combination of both methods, unique
emergent from the deep interior of the light cone to FELIX, is likely to be the best of all.
under these circumstances is that that region re- A second important tag available to FELIX is
laxes back to the true vacuum, despite its being the choice of beam. By tagging on a leading
isolated from the true vacuum by a fireball shell neutron or ∆++ at very low t, one can reason-
and despite there not being enough elapsed time ablycleanlyisolatetheone-pion-exchangecontri-
forchiralorientationto be distinguishedenerget- bution, and thereby replace the LHC pp collider
ically from chiral disorientation. with a somewhat lower energy, lower luminosity
πp collider. In a similar spirit, and including Λ
2.5. Collisions with very high impact- tags, one can study collisions of any combina-
parameter may probe the chiral vac- tion of π, K, or p with each other. The beam-
uum structure dependence of phenomena has historically been
Ingeneral,thechiralvacuumcondensateisdis- of considerable importance, and it may find im-
torted in the neighborhood of impurities such as portant applicability, especially with respect to
an isolated proton. This is just the long-range questionsofvalence-partonstructure,attheLHC
pion cloud surrounding it. The pion-cloud struc- energy scale.
ture can be probed especially well in high energy A special case of these tags is that of a photon
ppcollisionsatverylargeimpactparameters,say tag in ion-ion collisions, via forward detection of
2 to 3 fm. These interactions are, because of theundissociatedions. Theluminosityforγγcol-
the larger radii of interaction at the LHC, a big- lisions is very high, and the capability of FELIX
ger component of the cross section, and can lead to exploit this luminosity is also very high.
to larger final-state multiplicities than found at Another class of tags which has been under-
lower energies. Perhaps here too there may be utilized is the diffractive tag, where leading pro-
coherence in the structure of the pion emission, tons are detected via Roman-pots. As discussed
and this class of events may turn out to be of above, this leads to a very rich stratum of up-
special interest. Again a detection capability at to-now poorly-measured, poorly understood, but
very low p , 100 MeV and less, as possessed by potentially important physics.
T
FELIX, is important for such studies. Finally, there may be pattern tags. The event
structure in final states containing jets is depen-
2.6. New opportunities exist for tagging dent upon the color flow. Typically, neighboring
event classes jets in phase space are connected by a partonic
Together with these many novel phenomena, color line (antenna). For quarks, one antenna
therewillbenewmethodsforexperimentallytag- line emerges from the jet, for gluons two. Along
ging different kinds of events. The impact pa- these antennalinesinphasespace,hadronization
rameter of the collision is obviously of impor- and minijet production is enhanced. Recently
tance to be determined event-by-event. This is the Tevatron collider experiments have observed
done routinely in ion-ion collisions via zero de- these effects. In principle this technique might
6
0.2
0.1
0
-0.1
-0.2
0 20 40 60 80 100 120
Figure 3. The top view of the FELIX detector. The different magnets, calorimeters (hatched areas),
tracking stations (vertical lines) and the beam trajectories in the horizontal plane are indicated.
∗
allow one in the future to identify in an indi- insertion which can be tuned from β = 23 m
∗
vidual multijet event quarks versus gluons, and to β = 900 m without changing the magnetic
even fully classify the event structure according elements.
to the color flow. Clearly such a pattern-analysis There are two significant features of this in-
technique is very difficult, and needs to be data- sertion. First, the final-focus quadrupoles can
driven. FELIX, with full acceptance, will be op- be placed more than 120 m from the IP, provid-
timal for making the attempt. ingthespaceneededtoaccommodatetheFELIX
dipoles. Second, it is economical. The necessary
quadrupoles are already in the LHC baseline de-
3. The FELIX detector
sign.
We now introduce the major features of the The ability to tune the insertion also has sev-
FELIX design. More details can be found in the eral nice features. At β∗ = 900 m, FELIX is
FELIX LOI. [1]. optimized for the study of low-t elastic scatter-
∗
ing. At β = 110 m, where FELIX’s luminosity
3.1. A tunable insertion at I4 is about 4 x 1031 cm−2 s−1 when the LHC is at
A full acceptance detector must be able to an- design luminosity, the beam size in the heart of
alyze the global structure event-by-event. This FELIXdetector ( 120m) is minimized, permit-
means that it should run at a luminosity no ting the Roman p±ot detectors in these locations
greater than 1032 cm−2 s−1; that is, with less to come as close as 3 mm to the beam. Finally,
L∼
than about one interactionper crossing. This lu- β∗ = 23 m permits FELIX to reach luminosities
minosity can be achieved at I4 by means of an
7
as high as 2 x 1032cm−2s−1. the typical momenta of the particles, resulting
inmomentumresolutionwhichisreasonablyuni-
3.2. Well-understood magnets form over all of phase space.
FELIX will implement a “kissing scheme” in Finally, we note that all magnets can be ac-
which the two beams are brought together at 0o commodated in the existing Aleph collision hall
in the horizontal plane and then returned to the andadjacenttunnelswithoutanysignificantcivil
same inner or outer arc (See Figure 3). To ac- construction.
complishthis,weneedsome67T-mtofirstbring
the beams together (D2 magnets), and then an-
3.3. Compact, precise tracking
other 67 T-m (D1 magnets) to make them par-
Some50trackingstations,locatedasfaras430
allel. This has to be accomplished within the
m from the IP, are needed to ensure full accep-
120 m available. Both sets of magnets must be
tance and uniform resolution. The positions of
superconducting machine dipoles. The D1 mag-
most of the stations (vertical lines) are indicated
nets must also have large bores, to accomodate
in Figure 3. FELIX will instrument radially out-
both beams and to provide acceptable tracking
ward,emphasizingcompact,near-beamtracking.
and calorimetry apertures.
How close we will approach the beams depends
FELIX is fortunate that Brookhaven National
on the location. In general, we will use Roman
Laboratory (BNL) has designed large aperture
potdetectorstoaggressivelyapproachthebeams
superconductingdipolemagnetsforuseatRHIC.
wherever the location is accessible and the pot
With a coil aperture of 18 cm and a design field
mechanicalstructuredoesnotinterferewithother
of 4.28 T (FELIX will use them at 3.62 T),
trackingorcalorimetry. Elsewhere,weproposeto
thesemagnetsaresuitableforuseasD1magnets.
use fixed-radius tracking, approaching to within
BNL is committed to producing these magnets
2.5 cmofthe beams. The acceptancefor charged
for RHIC and thus will be able to supply well-
particles as a function of pseudorapidity (a) and
understood magnets on the FELIX time scale.
their momenta(b)(seeFigure 4)isalmost100%
The constraints on the D2 magnets are some-
over the entire phase space.
what less severe, and several options are avail-
An important consideration is the occupancy
able. Of these, FELIX proposes existing super-
within the tracking detectors. High particle den-
conducting dipoles constructed as prototypes for
sities close to the beam pose a significant pat-
UNK. While these are single aperture magnets,
tern recognition problem. Each tracking sta-
the42cmbeamseparationpermitstwoUNKcold
tionshouldthushavesufficientresolutionandre-
massesto be assembledina commoncryostatfor
dundancy to be able to locally reconstruct track
use as D2 magnets.
segments. Track segments are then matched,
In order to avoid parasitic beam-beam inter-
station-to-station, resulting in a very powerful
actions and long-range tune shift effects, the
spectrometer.
beams will collide with a vertical crossing angle
These considerations lead to a common con-
of 0.5 mrad. To do this while optimizing the
± ceptual designfor mostFELIXtrackingstations,
match of the magnetic architecture to tracking
based on two technologies: Si pixel detectors out
andcalorimetry,weproposetore-usetheexisting
toradiiofabout8cm,supplementedbyGasElec-
UA1 magnet, split longitudinally into two halves
tron Multiplier (GEM) chambers at larger radii.
and equipped with new coils. We will also build
We are also exploring the possibility of using
two5meterlong,2Twarmdipole(D0)magnets.
GEMasthebasisforverycompactmicro-TPC’s.
The magnetic architecture is completed by the
Aconceptualdesignfora“standard”fixed-radius
re-use of the existing ALEPH solenoid, which is
tracking station is shown in Figure 5. The same
well-matched with the use of the UA1 magnet.
technologieswillbe usedforacompactmicrover-
An important feature of this overall design is
tex detector.
that the strengths of the magnetic fields increase
in the forward direction, always well-matched to
8
3.4. Forward calorimetry
FELIXproposesfour calorimetersoneachside
ce 1 of the IP to provide complete electromagnetic
n
epta 0.75 a) and hadronic calorimetry for angles θ < 0.2 ra-
c dian, that is, for η > 2.3. The coverage of the
Ac | |
0.5 calorimeters is illustrated in Figure 6. The in-
terplaywith the magnetsandtrackingsystemsis
0.25
illustrated in Figures 2.
0 The calorimetersmusthavesuperbenergyand
0 2 4 6 8 10 12
h spatialresolution,andmustprovidethe informa-
tion needed to identify neutrons, electrons and
e
c 1
n gammas. Thismustbedoneinlimitedspace,and
a
ept 0.75 b) in a high-radiationenvironment. These consider-
c
Ac ations determine the structure of the calorime-
0.5 ters, the choice of sampling materials and the
kinds of photodetectors and front end electron-
0.25
ics which can be used for the readout.
00 1000 2000 3000 4000 5000 6000 7000 The UA1 endwall calorimeter, which is ex-
P (GeV) pected to have a radiation dose of less than 5
Mradfor10yearsrunning,isasamplingcalorime-
ter based on plastic scintillators and wavelength
shifting fibers. The very forward (D0, D1 and
Figure 4. The acceptance in FELIX for charged
Zero Degree calorimeters) see much higher ra-
particle momentum measurements as a function
diation levels, and will thus be “spaghetti”-type
of(a)thepseudorapidity;and(b)themomentum
calorimeters,basedoneitherthincapillariesfilled
of the particles.
with liquid scintillator or on quartz fibers. All
threeveryforwardcalorimetersaresimilarincon-
struction, differing only in their overall dimen-
sions. Each consists of a preshower detector, an
EMcalorimeter,and twohadroncalorimetersec-
tions.
3.5. Trigger
The basic structure of the trigger is a multi-
level triggering scheme which must reduce the
triggerratefrom40MHz to arateacceptable for
data recording. A schematic view of the trigger
scheme together with the rates and the latencies
of the three trigger levels is shown in Figure 7.
The initial Level 1 triggering must be done
deadtimeless for each bunch crossing. It must
establish clean evidence of the presence of an
event from the triggering constituents and pro-
Figure 5. A schematic view of a tracking sta- duce a constituent-basedtriggerdecision. This is
tionbasedonSipixeldetectorsandamicro-TPC. achieved with simple algorithms based upon the
Note thatseverallarge-areaGEMchambershave topology of the elements of the detectors which
been removed to improve visibility of the micro- fired, or upon special trigger counters and en-
TPC. ergy deposited therein. These local trigger de-
9
UA1 (cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1) D0 (cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)
(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1) (cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)
(cid:0)(cid:1)(cid:0)(cid:1) (cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1) (cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)
00..21mm 2h=020.3mrad (cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1) (cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)h=(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)26(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1).6.6mh=1r44am.d9rad (cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1) h=7.1 1.6mrad D1 (cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1) ZD (cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:0)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)(cid:1)
(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)
IP4 (cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)
0 m 10 m 20 m 30 m 100 m
(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1) (cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)(cid:0)(cid:1)
(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)EM-calorimeter (cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1)(cid:0)(cid:0)(cid:1)(cid:1) Hadron calorimeter Preshower Coil
Figure 6. Schematic view of FELIX forward calorimetry.
cisions are produced near the detector front-end 3.5.1. Run scenarios
electronics by special circuitry. During the Level The first major data set containing inelastic
1 latency, the sampled data of all channels of all events used for serious physics analyses is as-
detectors are stored in data pipelines. After a sumedtobeoforder108 ppcollisions,takenwith
positive Level 1 trigger decision, the output data essentially a minimum-bias trigger. The only se-
are associated with specific bunch crossings, for- lectivity provided at trigger level would address
matted into event fragments, and stored for sub- the rejection of bunch crossings with no interac-
sequent filtering by the Level 2 trigger. The out- tions (or more than one interaction), and rejec-
put rate from Level I is estimated to be around tion of candidate events seriously contaminated
100 kHz. with beam-gas background.
TheLevel2systemperformsafastdecisionon After such a minimum-bias run, an
the eventquality, basedupon data from separate intermediate-luminosity run could be naturally
detectors or portions thereof, using more sophis- implemented using only the Level2 triggerstrat-
ticated algorithms. These may be realized by a egy and a choice of interaction rate compatible
set of dedicated hardware or software processors. with the lackofrejectionthatLevel1wouldpro-
The output rate from Level 2 is estimated to be vide. Such a strategy would seem to maximize
roughly 10 KHz. the flexibility of choice of detailed triggers, and
The Level 2 filtered data are combined into leave to a minimum the amount which is hard-
complete events by the event builder, and are wired into the front end of the DAQ system. We
filtered by the Level 3 computer farm. Level 3 assume here that this intermediate-luminosity
produces the complete event-based decision, and run might yield of order 108 recorded events for
its task is to reduce the trigger rate from 10 1011 pp interactions in the FELIX detector, i.e.
KHztooneacceptablefordatarecording(around have a rejection of order 103.
100 MBytes/sec). We estimate the mean FELIX Finally, the design run, fully implementing the
event size to be around 0.5 MBytes. Therefore Level 1 and Level 2 triggers, should be expected
the Level 3 rate might be around 200 Hz. to yield 108 recordedevents per 1014 interactions
in the detector.
The detailed choice of the trigger algorithms
10
Event data physics, whilst bearing in mind the LHC physics
priorities already established”
When the possibility for a new interaction re-
40 MHz ECal HCal Preshwr Rpot Muons
gion in I4 became reality (summer 1995) several
ne
128 ev peli workshops took place to discuss the layout of a
Pi full acceptance detector.
Level 1
100 kHz Decision I Fixed latency In May 1996 the LHCC defined new rules for
(~3 ms)
coming activities : “ The LHCC urges that any
ECal HCal Tracker Rpot Muons new experimental initiative should be consistent
100 - 1000 ev uffer uffer uffer uffer with the restricted resources likely to be available,
b b b b
and combined as far as possible with one of the
Level 2
10 kHz Decision II Variable latency foreseen experiment.”
(1 - 10 ms)
InanOct. 1996memorandum[2]totheLHCC,
Level 3 FELIX responded to these new guidelines by de-
Event Builder Variable latency
(~1.0 s) scribing, in detail, the FELIX set-up, strategy
and financial assumptions. The group received
200 Hz CPU CPU CPU CPU Decision III
general encouragement from the CERN manage-
ment to go ahead with the Letter of Intent.
100 MByte/s Data recording Duringthespringandsummerof1997,theFE-
LIX collaboration mobilized for the preparation
of the LoI, which was submitted to the LHCC in
August 1997.
Figure 7. The FELIX triggering and DAQ
InNovember1997,the LHCC chosetoaddress
scheme.
the FELIX Letter of Intent, finding
... that the FELIX LoI is not responsive to
these guidelines. While the physics topics ad-
should probably be deferred for as long as possi- dressed by the programme proposed in the LoI
ble,foratleasttworeasons. Oneistechnological; are of interest (particularly the complete recon-
the rate of change is so largethat one should opt struction of diffractive events), the likely costs of
for as modern hardware (and software) as possi- constructing the proposed dedicated detector and
ble. Theotheristhephysicsitself,whichisevolv- of the modifications to the LHC collider are very
ingandwillcontinuetoevolveratherrapidlywith high in comparison with the probable physics out-
time. It is not easy to anticipate in detail what put. Finally, the composition and strength of the
will be the highest-priority QCD physics to ad- collaboration seem inadequate for carrying out a
dress at the time of LHC commissioning. strongprogrammeaddressingthesephysicstopics.
[3]
The CERNResearchBoardhassinceendorsed
4. Recent History
the decision of the LHCC.
After the presentations of J.D. Bjorken and The FELIX collaboration believes that these
K. Eggert about possible forward physics in pp decisions were reached in a precipitate manner,
and p-A collisions at the LHCC ”Workshop on withgrossviolationofdueprocess. Inparticular,
Further Physics Topics” (Nov. 1994) the LHCC therehasbeennothoroughscientificreviewofthe
Committee recommended this kind of physics by FELIX proposal. Indeed, a primary grievance is
noting: “The LHCC noted the interest in diffrac- that the LHCC referees never contacted the pro-
tion, and expects that such studies may also form ponentsbeforearrivingatitsnegativeconclusion,
part of the LHC experimental programme. The nor were the proponents permitted to directly
committee encourages interested parties to work presentthe initiative inpersonto the committee.
together on an integrated approach towards this Important issues, including possible staging sce-