This website is no longer maintained. Its content may be obsolete. Please visit http://home.cern/ for current CERN information.
The basic philosophy of our detector is that it will be a survey instrument: sensitive to almost everything, and optimized for almost nothing, i.e. a detector with good capabilities to see the unexpected as well as ``engineered" discoveries. We note that some of the items on the physics agenda place stringent requirements on the detector. For example, the analysis of events with rapidity gaps requires coverage without holes over the maximum range of pseudorapidity attainable, while studies of jets-within-jets requires fine segmentation. Studies of disoriented chiral condensates require the ability to simultaneously see both charged particles and photons. The challenge is to design a detector which can meet all these requirements in the difficult environment of forward physics at a hadron collider.
The strength of a full acceptance detector will be in the perception of complicated patterns in individual events. This leads essentially to the following design goals:
should be observed and their momenta
well-measured over the entire phase space (both arms)
Figure 2: Sketch of the FELIX experiment
at I4 (central and very forward region).
Fig. 2 is a sketch of the central
and the very forward part of FELIX in the
intersection region I4, which can be equipped with an insertion well suited
for a forward experiment. The basic magnetic architecture is given by the
central ALEPH solenoid (1.6 Tesla), already in place, complemented on both
sides by dipole magnets with a horizontal field of 0.7 - 1 Tesla. These
magnets, excited by superconducting coils, can be easily constructed out of
the modular UA1 magnet. Since the experiment has to extend well into
the forward beam lines it is important that the magnetic architecture of the
experiment is fully integrated into the design of the machine lattice. The two
horizontal dipoles UA1 and D0 provide the vertical crossing-angle of about 0.6
mrad needed to avoid long range tune shift effects at large 
-values. The
vertical dipoles D1 (split into three magnets) and D2 (two magnets) separate
the beams horizontally to the required distance of 42 cm. All forward magnets
(D0,D1 and D2) with a free bore of 10 cm radius and a field strength of 4
Tesla are identical in their design (taken from RHIC) and differ only in their
length. Thus the forward magnets are used to guide the beams and at the same
time they serve as powerful spectrometer magnets.
The measurements of very forward particles are especially difficult since they
have to be performed before the particles shower in the surrounding material.
The vacuum-chamber architecture, the calorimetry close to the beams, and the
implementation of ``Roman pots" have to be optimized and thus present a major
challenge. Forward particles are bent out of the beam profile according to
their momenta in the different magnets and their momenta are measured with
detectors placed as close as 
away from the beam. Thus each forward
spectrometer covers a certain range in momentum. The
spectrometers are designed so that they have overlapping acceptances. Thus
a charged particle track is acquired in one magnet
and the measurement of it's momentum is improved
in the following one. As a consequence of the
larger bending power of the very forward magnets the momentum resolution stays
well below 0.5 % up to the beam momentum (assuming a spatial resolution
of 
per detector plane).
For the detectors close to the beam we envisage a combination of pixel
detectors with 
pixel size (for pattern recognition) and
silicon strip detectors (for resolution improvement).
The design of the beam pipe plays an important role in the optimization of the very forward detector. It has to be as transparent as possible to avoid particle showering, but it should also permit the placement of detectors close to the beam. Hence large diameter vacuum pipes can only be considered if a radio-frequency screen around the beam minimizes electrical losses and disturbences to the beams.
The charged particle measurements will be complemented by complete electromagnetic and hadronic calorimetric coverage extending down to zero degrees. Such a hermetic detector will, for the first time, provide a measurement of total missing energy, a powerful tool in the search for new physics. The interplay between calorimetry and charged particle tracking is delicate, however. In principle, charged particle momenta should be measured by tracking before they enter the calorimeters. This is complicated, however, because each stage of calorimetry also defines the aperture restriction for subsequent stages of tracking. Thus care is needed to avoid splash and leakage from the calorimeters interfering with subsequent stages of tracking. Detailed GEANT simulations are in progress to help optimize the forward detectors, and to determine the occupancy of the forward trackers and the granularity of the calorimeters.
Finally, an almost complete muon coverage will extend down close to the beams.
We have devoted comparatively little effort to detailed work on the detector architecture in the central region. Unlike the forward direction, the technical challenges in the central region are reasonably well understood, and we will be able to profit from the experience of the many existing or planned central detectors.
FELIX will run at a reduced luminosity of 
while the other experiments are at full luminosity. In order to reduce
the luminosity locally in I4 by up to three orders of magnitude relative
to the other experiments, and to minimize the beam divergence for the
elastic scattering measurement, -values at the intersection point
(IP) of up to 
or more have to be achieved. Fortunately the
group of T. Taylor (CERN-LHC) has designed a new insertion for
I4 (Fig. 3) which
takes into account the FELIX experimental magnets together with the
constraint that the beams be separated by 
at the location of the RF
cavities [8].
Figure 3: The proposed insertion at I4, also indicating the positions
of the ``Roman pots'' upstream.
The basic feature of the design is that the inner
quadrupole triplet is located behind the beam separating magnets, leaving
sufficient space between the magnets D1 and D2 for the very forward
spectrometer and the zero degree calorimeter. Further, the proposed
insertion can be tuned so that the -value in
IP4 is either in the range of 
or 
. While precluding reaching the very highest luminosities of the
major detectors ( 
),
this is nicely matched to the FELIX physics goals. After completion of
the low-luminosity running necessary to achieve the physics goals outlined
here, it should be possible to modify the insertion in order to achieve
lower values of 
.
Figure 4: The 10 
beam profile as a function of the distance from
the IP and the excursion of (a)
a diffractive proton ( 
); (b) an elastically scattered proton
( 
).
FELIX has to extend several hundred meters upstream in order to measure
diffractive and elastically scattered protons. The detectors are placed
in ``Roman" pots so that they can be moved
as close to the beam as possible (probably 
) during stable
beam conditions, and be retracted during
beam injection and acceleration.
The optimized locations of these tracking stations are indicated in
Fig. 3.
Fig. 4 plots the beam profile for the two
directions perpendicular to the beam (x lies in the beam-bend plane)
for a 
. For 
, the
beam width is around 
.
An elastically scattered
proton with 
can be measured well
due to it's excursion of about 
(Fig. 4(b)).
The excursion of a diffractive proton with a relative momentum
loss 
is largest at 
, where the dispersion of the machine
is about 
(see Fig. 4(a)). It should be possible
to measure diffractive protons with 
.
