The US and CMS

CMS detector
Lowering of one slice of the CMS detector into its cavern. Image © CERN
The US CMS collaboration, with 49 institutions, 800 physicists, nearly 200 graduate students, and over 300 engineers, technicians and computer scientists is the largest national group in the CMS collaboration. US groups have made significant contributions to nearly every aspect of the detector throughout all phases including construction, installation and preparation for data-taking. The US collaboration also makes major contributions to the construction and operation of the computing facilities that are needed to analyze the unprecedented amount of data generated by CMS, and to the software that allows physicists to operate the CMS detector, reconstruct the data, analyze it and extract from it physics discoveries.

The CMS detector is designed to detect the basic objects that are identified by physicists as being truly fundamental: electrons, muons, tau leptons, photons, quark jets and missing energy due to very weakly interacting particles such as neutrinos. Massive particles such as the Higgs boson decay into these fundamental objects, the properties of which are measured in the CMS detector’s many subsystems.

Hadron Calorimetry

US scientists led the international team that designed, constructed and tested the hadron calorimeter. This device measures the energy and direction of jets, which are sprays of fundamental particles that are the manifestation of quark and gluon production in high energy interactions.

Electromagnetic Calorimetry

The electromagnetic calorimeter is used to identify electrons and photons and measure their energy and position. In CMS, these measurements are accomplished with a system of lead-tungstate crystals that record the energy and position of electromagnetic showers. The crystal material was developed by CMS and provides excellent energy resolution while surviving the high radiation doses that will exist in the detector. The full calorimeter contains about 80,000 crystals. The US contributed to the barrel electromagnetic calorimeter that forms a cylinder surrounding the interaction point. US scientists developed calibration systems for the calorimeter, and worked on front-end sensors that detect the electromagnetic showers.

Muon Detector

The US led the project to provide muon identification in one subset of the large muon detector subsystem. Muon identification uses a type of detector called Cathode Strip Chambers, a gas detector that simultaneously reads out information about muon positions in two dimensions. In total, 468 planar detectors were constructed, organized into eight huge wheels, each ten meters in diameter.

Silicon Strip Tracking system

CMS has constructed a system made entirely of silicon detectors that tracks charged particles emitted from the collision point. This system, with more than 200 square meters of silicon, is the largest silicon strip detector ever constructed. The full detector has 11 million silicon strips of varying sizes and positions organized into 15,000 modules. The US built all 5,200 modules for one part of the system, the Tracker Outer Barrel, and assembled them into rods. It also constructed over 2,000 modules for another part, the Tracker Endcap. This amounts to about two-thirds of the surface area of the detector.

Forward Silicon Pixel Tracking System

Close to the proton beams, the density of particles emitted from the collision is very high. Tracking of particles in this region must have high precision to determine whether particles come from the main interaction or from decays that happen a very small distance away. This requires a highly segmented silicon pixel detector that can operate from a few centimeters to about 20 centimeters from the beam. US groups have led the project to build the endcap, or forward, section of this device. It consists of 18 million silicon pixels of dimension 100x150 mm 2.

Trigger System and Data Acquisition

At full luminosity, the LHC beams produce one billion interactions in the center of the CMS detector. The trigger system selects about 100 of these events as promising indicators of new physics beyond the Standard Model. US groups played a leadership role in the trigger project, especially in developing hardware for the lowest level of the trigger selection project, which must examine every collision and therefore must operate at the highest speed. After this lowest-level trigger, only the surviving events must be examined by subsequent phases, which therefore can spend more time on each event. Once the trigger decides to accept an event, it must be recorded in an accessible way for subsequent analysis. The US developed the hardware and software to coordinate the flow of events in the data acquisition system and to bring the event data from the underground enclosure housing the experiment to the surface where the events are recorded.

Computing

The discovery of new physics at CMS depends upon the successful processing and storage of an unprecedented amount of data. Each year, the detector produces more than one petabyte—one million gigabytes—of raw data. Even larger amounts of simulated data are produced, necessary for understanding the performance of the detector and what signals from new physics will look like in the collected data. All of this data must be processed, stored, transmitted to physicists spread around the world and analyzed.

To cope with these demands, CMS and the other LHC experiments use a distributed computing solution. In this tiered system, the CERN computing center is responsible for doing a first-pass processing of the collision events, writing the data to tape for permanent storage, and then sending copies of the data to seven Tier-1 sites. One of the Tier-1 facilities is located at Fermi National Accelerator Laboratory in Illinois. Fermilab will further reduce the new data, re-process older data with improved software tools, and archive the data.

Generating simulated data and making data available to physicists around the world is the job of the Tier-2 facilities, which in aggregate have as much computing power as the Tier-0 facility. There are about thirty Tier-2 sites spread around the world, seven of which are in US universities: University of California, San Diego; Caltech; University of Florida; MIT; University of Nebraska-Lincoln; Purdue University; and University of Wisconsin-Madison. Each has sufficient computing resources for about forty physicists to study the data received from Tier-1 sites and make the measurements which will lead to publications and discoveries. Tier-2 computers are also used for detector simulations that are archived at Tier-1 sites. By 2008, each US Tier-2 site achieved CPU power equivalent to about 300 of today's desktop PC's, 0.2 PB of disk, and a 10 gigabit/sec network link to Tier-1 sites.

List of US institutions participating in the CMS experiment