CMS - The Compact Muon Solenoid

CMS is a general purpose proton-proton detector designed to run at the highest luminosity at the LHC. It is also well adapted for studies at the initially lower luminosities. The CMS Collaboration consists of over 1800 scientists and engineers from 151 institutes in 31 countries. The main design goals of CMS are: \begin{enumerate} \item a highly performant muon system, \item the best possible electromagnetic calorimeter \item high quality central tracking \item hermetic calorimetry \item a detector costing less than 475 MCHF. \end{enumerate} All detector sub-systems have started construction. Engineering Design Reviews of parts of these sub-systems have been successfully carried-out. These are held prior to granting authorization for purchase. The schedule for the LHC machine and the experiments has been revised and CMS will be ready for first collisions now expected in April 2006. \\\\ ~~~~$\bullet$ Magnet \\ The detector (see Figure) will be built around a long (13~m) and large bore ($\phi$=5.9~m) high-field (4T) superconducting solenoid leading to a compact design for the muon spectrometer. The magnetic flux is returned through $\approx$1.5~m of saturated iron yoke (1.8~T) instrumented with muon chambers. The construction of the magnet is well advanced. It will be tested on the surface in July 2004. Three of the five barrel yokes (YB) have been assembled in the surface building at Point 5. The assembly of the endcap yokes (YB) will start in April 2001. Four good lengths of Rutherford cable and three lengths of the insert (Rutherford cable co-extruded with pure aluminium), out of 21, have been produced. Each one has a length of 2.65km. The contracts for the winding machine have been placed. \\\\ ~~~~$\bullet$ Inner Tracking \\ All high $p_t$ muons, isolated electrons and charged hadrons, produced in the central rapidity region, are reconstructed with a momentum precision of $\Delta p_t / p_t \approx$~0.005~+~0.15$p_t$ ($p_t$ in TeV). The high momentum precision is a direct consequence of the high magnetic field. The tracking volume is given by a cylinder of length 6~m and a diameter of 2.6~m. In order to deal with high track multiplicities tracking detectors with small cell sizes are used. Silicon microstrip detectors provide the required granularity and precision in the bulk of the tracking volume. Stereo information is provided by back-to-back microstrip detectors with strips at a small angle. Pixel detectors placed close to the interaction region improve the measurement of the track impact parameter and secondary vertices. High track finding efficiencies are achieved for isolated high $p_t$ tracks. It is also fairly high for tracks in jets. \\ In June 2000 the LHCC approved the `All-Silicon Tracker' to be built in a single stage. The layout has been optimized with the removal of the central support tube. A pre-production comprising 200 detectors has been launched to exercise the automated production procedure. \\ The short bunch crossing time at the LHC (25ns) places challenging requirements on the readout electronics. Furthermore, the detectors and the read-out electronics have to withstand high levels of irradiation. A test in an LHC-like bunched beam was successfully carried out to test the functionality of the full electronics chain. Beam tests using electronics designed in the 0.25$\mu$m technology have confirmed the expected performance. \\ Good progress is also being made on the electronics and mechanics of the pixel detectors. \\\\ ~~~~$\bullet$ Muon System \\ Centrally produced muons are measured three times, in the inner tracker, after the coil and in the return flux. They are then identified and measured in four identical muon stations (MB) inserted in the return yoke. Special care has been taken to avoid pointing cracks and to maximize the geometric acceptance. Each muon station consists of twelve planes of aluminium drift tubes designed to give a muon vector in space, with 100~$\mu$m precision in position and better than 1~mrad in direction. The four muon stations include RPC triggering planes that also identify the bunch crossing and enable a cut on the muon transverse momentum at the first trigger level. The endcap muon system also consists of four muon stations (ME). Each station consists of six planes of Cathode Strip Chambers. The chambers are arranged such that all muon tracks traverse four stations at all rapidities, including the transition region between the barrel and the endcaps. The last muon stations are after a total of $\geq$~20$\lambda$ of absorber so that only muons can reach them. The four muon stations lead to a redundant and robust muon system. \\ Facilities for mass production have been set up in the institutes participating in the construction of the muon chambers. One site, CIEMAT (Madrid), has built two pre-production drift tube prototype chambers using the final assembly tools and procedures. The commissioning of two more sites (Aachen and Legnaro, Padova) is nearly complete. Around thirty CSC chambers have been manufactured at Fermilab. Parts and tooling have been procured for the sites at PNPI, St. Petersburg and IHEP, Beijing. The procurement of parts for the barrel RPCs also commenced in 2000. Mass production of DTs and CSCs at various sites is expected to reach the final rates in 2001. \\ The combined (using the inner tracker as well as the muon chambers) muon momentum resolution is better than 5\% at 0.3 TeV in the central rapidity region $|\eta| <$ 2, and $\approx$ 10\% for $p_t$ = 2 TeV. Low-momentum ($p_t <$ 100 GeV) muons are measured before the absorber with a precision of about 1.5\% up to a pseudorapidity of 2. \\\\ ~~~~$\bullet$ Calorimetry \\ The coil radius is large enough to install essentially all the calorimetry inside and hence avoid the coil-electromagnetic calorimeter interference. A high precision electromagnetic calorimeter (ECAL) using lead tungstate (PbWO$_4$) crystals has been chosen. Lead tungstate is a dense and relatively easy crystal to grow. \\ The scintillation light is detected by silicon avalanche photodiodes in the barrel region (EB, $|\eta|<$1.48) and vacuum phototriodes in the endcap region (EE, 1.48$<|\eta|<$3.0). The expected energy resolution is better than 0.6\% for electrons and photons with energies greater than 75 GeV. A preshower system (SE) is installed in front of the endcap calorimeter (1.65 $\leq|\eta|\leq$ 2.6). \\ The ECAL is followed by a copper/scintillator sampling hadronic calorimeter (HB, HE). The light is channelled by clear fibres fused to wave-length shifting fibres embedded in scintillator plates. The light is detected by photodetectors that can provide gain and operate in high axial magnetic fields (proximity focussed hybrid photodiodes). Coverage up to rapidities of 5.0 is provided by a steel/quartz fibre calorimeter (HF). The Cerenkov light emitted in the quartz fibres is detected by photomultipliers. \\ The pre-production (6000) of crystals from Russia has been completed. A contract for a further 30000 crystals has been placed in Russia. A breakthrough in crystal growing in Russia means that ingots can be grown of a diameter large enough for two crystals to be cut out. This will considerably increase the yield of crystals per oven. The crystal producers in China, using 28-fold pulling furnaces, have delivered 100 preliminary pre-production crystals that are being evaluated. The contract for the remaining 40000 crystals should be placed in 2001. The infrastructure at the centres where the crystals will be assembled into modules for installation has been set up. The photo-detectors, meeting the specifications, have been developed in collaboration with industry. The front-end chain consists of a preamplifier/range selector (FPPA), an ADC and a serializer/optical link. A 0.25$\mu$m technology version of the serializer has been chosen. \\ Photon-pizero separation in the forward region requires a preshower detector (SE) in front of the crystals. Silicon sensors for the pre-shower detector, of the required quality, have now been produced in Russia, Taiwan and India. A large dynamic range preamplifier in a radiation-hard technology has been fabricated and successfully tested. \\ The absorber for the first HCAL half-barrel, HB-1 (18 wedges), was trial-assembled at Felguera, Spain. It was dismantled and the wedges have been delivered to CERN. The optics, scintillator plus embedded fibres, for more than half the barrel wedges have also been manufactured and delivered to CERN. The trial assembly of one endcap is nearing completion at the manufacturer, MZOR (Byelorussia). Optics manufacture for HE has started. The HF design was changed from bricks to 18 wedges per side. The fibre spacing was changed from 2.5mm to 5mm but the packing fraction is preserved. \\\\ ~~~~$\bullet$ Trigger and Data Acquisition \\ The trigger and data acquisition consists of four parts: the front-end detector electronics, the calorimeter and muon first level trigger processors, the readout network and an on-line event filter system. The first two parts are synchronous and pipelined with a pipeline depth corresponding to $\approx$3~$\mu$s. The latter two are asynchronous and based on industry standard data communication components and commercial PCs. The resources that would have been required for a hardware second level trigger processors are invested in a high bandwidth ($\approx$~500~Gbit/s) readout network and in the event filter processing power (10$^{6}$-10$^{7}$ MIPs), both of which are more suitable for upgrading as commercially available technology develops. \\ The CMS Level-1 trigger decision is based upon the presence of physics objects such as muons, photons, electrons, and jets, as well as global sums of E$_t$ and missing E$_t$ (to find neutrinos). The Level-1 Trigger Technical Design Report was submitted at the end of 2000. The DAQ system has to assemble the data from the triggered event, contained in about 500 front-end buffers (readout units), into a single processor in a ``farm'' for executing physics algorithms so that the input rate of 100 kHz is reduced to the 100 Hz of sustainable physics. A new Event Builder setup has been installed that consists of 64 Intel-PCs interconnected by two networks based on the most advanced technologies: a 64 port Gbit Ethernet (Foundry) and a 128-port Myrinet switch (Myricom). The setup will be used to evaluate all the software and hardware design options that will be considered for the TDR, destined for end-2002. \\\\ ~~~~$\bullet$ Computing and Core Software \\ For complex systems, such as the CMS detector, an `object oriented' approach, implemented in C++, is now the choice of software developers. The move to this mainstream software technology will help to manage the process of change over the long lifetime of the experiment. C++ releases have been made of the functional prototypes of the software comprising the framework (CARF), the reconstruction program (ORCA), a basic GEANT4-based simulation program (OSCAR), and an interactive graphics toolkit (IGUANA). The OO technology has been used in the production of Level-1 and High Level Trigger simulation data. ORCA has been used for detector, trigger and physics studies. \\ The data storage, networking and processing power needed to analyse CMS data is well in excess of those of today's facilities. Technological advances will help to make the data analysis possible in a distributed environment, where physicists are scattered all over the world. The optimum mix of storage, networking and processing will change as technology develops. A multi-Tier model, similar to that developed by the MONARC project, underpinned by Grid Technology to provide efficient resource utilization and rapid turnaround time will be prototyped. \\\\ ~~~~$\bullet$ Physics Reconstruction and Selection \\ With the construction phase starting in earnest, physics simulation work has begun to focus on the development of the eventual reconstruction code. As mentioned above this development is taking place using C++ and object-oriented methods. CMS has decided that the first priority is a full understanding and verification of the Higher Level Triggers (HLT). Since CMS does not employ distinct physical intelligences for the would-be Level-2 and Level-3 triggers, but only a single processor farm, the task of selecting events is intimately linked with that of reconstructing the associated detector information online. With this in mind, four ``Physics Reconstruction and Selection'' (PRS) groups were started (electron/photon, muon, jet/missing E$_t$, and b/$\tau$ vertexing) in April 1999. The aim of the groups is to develop the reconstruction and selection procedures (algorithms and software) starting from the output of the Level-1 trigger, and aiming ultimately at the full off-line reconstruction. During 2000, the four groups delivered the first algorithms that correspond to a reduction of the event rate after Level-1 by about a factor 10 using information from single CMS sub-detectors. The activity now continues as a new CMS project, the PRS project, which has close ties with the Computing and Trigger/DAQ projects. The PRS groups are now working on reconstructing physics objects using information from multiple CMS sub-detectors. \\\\ ~~~~$\bullet$ Physics Performance \\ Although high luminosity is essential to cover the entire range of mechanisms of electroweak symmetry-breaking, the LHC machine will start at significantly lower luminosities (L~$\leq$10$^{33}$ cm$^{-2} $s$^{-1}$). The pixel detectors and the PbWO$_4$ crystal electromagnetic calorimeter considerably enhance the discovery potential of CMS at low luminosities. \\ A Standard Model (SM) Higgs boson with mass between 95 and 150~GeV would be discovered via its two photon decay after an integrated luminosity of about 3$\times$10$^4$ pb$^{-1}$. The same integrated luminosity gives a discovery range covering masses from 135 to 525~GeV in the four lepton (e or $\mu$) channel, with only a small gap in the coverage around 2 m$_W$. An integrated luminosity of 10$^5$ pb$^{-1}$ (taken at 10$^{34}$ cm$ ^{-2}$s$^{-1} $) would allow discovery via these channels over the full range between 85 and 700 GeV. Tagging the events produced by WW- and ZZ-fusion by detecting characteristic forward jets, and using decay modes with larger branching ratios (H $\rightarrow$ WW $\rightarrow$ l$\nu$jj, and H~$\rightarrow$ ZZ $\rightarrow$ lljj), should allow the discovery range for a SM Higgs boson to be extended up to 1~TeV for the same integrated luminosity. \\ The two photon and four lepton channels are also crucial for the discovery of a Higgs boson in the Minimal Supersymmetric Standard Model (MSSM). Various channels involving the tau lepton (h$^0$, H $^0$, A $^0$ $\rightarrow$ $\tau$ $\tau$, H$^\pm $ $\rightarrow$ $\tau$ $\nu$) help to cover much of the remaining allowed (m$ _{A}$, tan $\beta$) parameter space. Precise impact parameter measurements using the pixel detector play an important role here. \\ Many physics studies have been carried out in the context of supergravity models (SUGRA). A many-point scan of the gaugino / scalar mass parameter space has been conducted. Squarks and gluinos weighing up to 2~TeV can be detected using, as signature, events with one or more charged leptons, missing transverse energy and two or more jets. Sleptons weighing as much as 400~GeV can be found by looking for events without hadronic jets, but with lepton pairs and missing transverse energy with distinctive kinematic characteristics. Three-lepton states are particularly promising for the detection of charginos and neutralinos. In many cascade decays a heavier neutralino is produced that subsequently decays into the lightest one with the emission of a pair of charged leptons. For low to moderate values of tan$\beta$ the spectrum of the di-lepton invariant masses shows a strikingly sharp end-point determined by the difference in neutralino masses. This feature can be used to select and almost fully reconstruct some events yielding e.g. the mass of the bottom squark. \\ The above studies of specific SUSY models indicate that it is possible to detect and measure a large fraction of the expected SUSY spectrum in CMS. Within the SUGRA models it should be possible to determine the fundamental parameters at the GUT scale. \\ The copious production of B mesons at LHC opens the way for significant measurements of CP violation effects in the B system. Using the B$^0_s$ and B$^0_d$ channels two of the angles in the unitarity triangle can be measured. Furthermore, by observing the time development of B$^0_s$ oscillations, the mixing parameter x$_s$ can be measured. \\ In addition to running as a proton-proton collider, LHC will be used to collide heavy ions at a centre of mass energy of 5.5~TeV per nucleon pair. A new form of deconfined hadronic matter, the quark-gluon plasma (QGP), should be formed at the resulting high energy densities (4-8 GeV/fm$^3$). The formation of quark-gluon plasma in the heavy ion collisions is predicted to be signalled by a strong suppression of $\Upsilon'$ and $\Upsilon''$ production relative to $\Upsilon$ production when compared to pp collisions. The CMS detector is well suited to detect low momentum muons and reconstruct the $\Upsilon$, $\Upsilon'$ and $\Upsilon''$ mesons produced. The good mass resolution ($\sigma$=37 MeV at $\Upsilon$ mass), afforded by the 4T field, enables the measurement of the suppression. Work has been carried out to obtain detailed understanding of the capabilities of CMS for heavy ion physics especially for signatures involving dimuon production, jet quenching and Z production. A detailed document has been prepared outlining the capabilities of CMS