Overview
The International Linear Collider (ILC) will be the next major project of the international high-energy physics community to follow the LHC, which is currently in the final phase of its construction. Different from the proton-proton collider LHC, which mainly aims for measurements at the highest reachable energies (up to 14 TeV centre-of-mass energy), the electron-positron collider ILC (with a centre-of-mass energy of 500 GeV and the possibility for a later upgrade to 1 TeV) will first and foremost be an instrument for measurements with maximum precision, thus following the exceedingly successful heritage of LEP. In this way the interplay of LHC and ILC will not only allow to study the Higgs particle which is predicted by many theories, but also to discover and to understand physics beyond the standard model, e. g. supersymmetric particles as possible constituents of dark matter, extra dimensions, or even the completely unexpected.
To carry out the required precision measurements, the ILC needs a detector of unprecedented measuring accuracy. Therefore different detector concepts are currently being studied and optimised in terms of the achievable track, momentum, and energy resolution. The so-called Large Detector Concept (LDC), which is amongst others being developed by the DESY FLC Group, features a Time Projection Chamber (TPC) as the main tracker and a highly granular calorimeter. In contrast to the conventional readout with wires, the TPC should have microstructures for its gas amplification – in this context both Gas Electron Multipliers (GEMs) and Micromegas are being investigated with the help of prototypes and simulations.
The experimental environment at the ILC, colliding point-like particles with an initial state of well-defined energy and polarisation, is generally considered very clean in comparison with the challenging LHC conditions. However there exist background processes as well which should not be neglected: It has to be ensured that this background does not degrade the detector performance.
A significant source of background are electron-positron pairs which are created from the so-called “beamstrahlung” in the interaction of the two strongly-focused beams. Those particles hit the forward calorimeters in large numbers and induce showers which can have a disturbing influence on the inner tracking detectors (vertex detector and main tracker) through the backscattering of charged particles. Furthermore photons and neutrons are created which can cover large distances inside the detector and initiate secondary processes in other places, e. g. in the TPC gas.
The simulation program Mokka is used for the investigation of these processes. It is based on the Geant4 framework and contains various models of the ILC detector. Besides that, Guinea-Pig is used as a particle generator to simulate the beam-beam interaction for different beam parameter sets.
Recent Work
During the last two years various modifications and extensions have been made to Mokka in order to enable it to simulate the effects of machine-induced backgrounds on the ILC detector: Apart from an interface to Guinea-Pig, much time has been spent to provide a model the forward region of the detector as realistic as possible, since even small changes of the geometry may have very large influence on the detector backgrounds. In that context the first and so far only detector models with beam crossing angles (2 mrad, 14 mrad, and 20 mrad) as well as different configurations of the magnetic field (DID, i. e. Detector-Integrated Dipole, and anti-DID) have been implemented.
The most suitable modelling of the underlying physical processes – called “physics list” in Geant4-speak – has been evaluated in direct collaboration with the Geant4 developers at CERN. The reason for this is that, contrary to the showers in the forward calorimeters which are mostly of electromagnetic nature and therefore usually well-reproduced by the simulations, the handling of neutrons in computer simulations is tricky. In the initial phase of the work it was anticipated that the creation of neutrons might pose a significant problem for the TPC and that they might have a strong influence on the choice of the chamber gas (with or without hydrogen content). However, according to the current simulation results, this worry was needless.
In the course of the past months, a couple of large-scale simulations with different detector geometries and beam parameters have been run. Doing so, the usage of grid resources was the only feasible way to process large amounts of data and to react to arising questions about machine and detector design. As examples, the “medium” crossing angle of 14 mrad and the “low-power option”, two topics which appeared in mid-2006, can be stated. Whereas the crossing angle is no problem for the detector as long as an appropriate configuration of the magnetic field is chosen, a change of the beam parameters towards a smaller number of more strongly-focused bunches would raise the detector backgrounds drastically up to a level which would probably be inacceptable for the vertex detector.
Upcoming Work
After the customisation of Mokka and the modelling of detector geometries has now been more or less finished, realistic simulations of the machine-induced backgrounds can be made. Still several tasks remain to be done and a couple of questions have to answered – possibly during the next year 2007:
Can the forward region be optimised further in order to reduce background signals? Especially small modifications to the BeamCal and the accompanying graphite absorber might further suppress the backscattering of low-energy particles into the inner detector.
With which amount of background signals will the TPC have to deal during standard operation? It has to be taken into account that the TPC is read out rather slowly and that the signals of several hundred bunch crossings will pile up.
Which impact will the choice of the chamber gas have? Can one take a usual gas with a content of five to ten percent of methane acting as a quencher, or should hydrogen be avoided? In the latter case, one would have to fall back to tetrachloromethane, which is known not to be completely uncritical.
How large will the occupancy of the TPC be, i. e. the relative number of voxels in which a signal can be found? If the occupancy is too large it will deteriorate pattern recognition, particle reconstruction, and finally the detector resolution.
What exactly happens to the neutrons which reach the inner detector? Are they captured by heavy nuclei, can they escape as thermal neutrons, or will they decay after a mean life of fifteen minutes?
How can a “background library”, containing a large number of readymade simulated bunch crossings, be set up and shared with others? The events in such a library could then easily be superimposed on other, interesting events in order to study the impact of detector backgrounds on reconstruction and analysis. In order to get sufficient statistics for that purpose, it is necessary to simulate a very large number of bunch crossings for a few selected, interesting points in the parameter space.
However, it will probably not be possible to give final answers – e. g. concerning the detector performance with and without background signals – since the required ILC Software (digitisation, pattern recognition, tracking, and “particle flow”) will not be developed far enough in the near to medium future.
The results of these studies are described in DESY-THESIS-2008-036: "Beam-Induced Backgrounds in Detectors at the ILC" by Adrian Vogel.
