In high energy physics bigger is usually better. Now a team at CERN in Geneva, Switzerland, has decided to look at different instead of big. The Compact LInear Collider (CLIC) team are investigating the potential of a new kind of particle accelerator, and to help them they are simulating their designs using grid resources, such the UK computing grid for particle physics, GridPP.
Accelerators can be split into two broad categories, linear and circular. The circular ones are known as discovery machines, the experiments where new physics/particles are seen, while the linear machines are about accuracy and really nailing down specific properties and information.
The 27 kilometer-long Large Hadron Collider at CERN is probably the best known of the discovery machines and CLIC is being designed to complement it (and others) by allowing researchers to flesh out the discoveries made and providing the detail needed for future discoveries/experiments.
The general design of CLIC is well understood; it will use electrons and positrons (identical to the electron in all respects but with a positive charge) and operate over a range of energies. However, using current methods to accelerate particles, this would require a machine hundreds of kilometres long, but CLIC is designed to be a 'mere' 50km.
One way the researchers will achieve this reduction in length is by using two beam pipes to move the particles from low to high energies. The two beam pipes are not the only innovation; CLIC plans to have two detectors that will alternate between taking data. They will move in and out of the beam as needed, so when one detector is in the beam, the other one is in a parked position.
"The research behind these detectors intends to be as generic as possible, having these requirements in mind really helps to push the envelope of what is possible today and provides ample opportunity for further detector research and development," said Jan Strube, one of the team members working on the CLIC detectors. "However we needed computer simulations to test them. We investigated how well they carried out a set of six physics measurements that would look at different aspects of their designs. This was not a trivial task and we found that the necessary resources of CPU time and storage were only available on the grid."
It fell to Stephane Poss, at CERN, to sort out how they could get up and running on the grid. "Running our simulations was not just computationally intensive but we were also going to be shifting around a large number of files so it was the obvious solution for us. I had used the grid before and was familiar with the DIRAC system, so that was the best starting point. The developers were very helpful and its modular nature made it relatively easy to customise for us. Early on I decided that any work I did should benefit the entire Linear Collider community so after six months I had built ILCDirac, a grid environment for both CLIC and similar experiments."
By December last year the team were ready to start using the grid on a large scale. Having been accepted into the ILC virtual organisation they had access to a wide range of resources across the grid. They started relatively slowly, but by mid-July they were able to ramp up their usage and in the last 10 months they have used five million hours of CPU time, with the UK GridPP being the largest single national contributor, providing 34%.
The grid-enabled studies were used extensively in CLIC's second design report, which was finished in September. This was presented to the community in October and the team are looking for signatures of support from them. For now though they are considering some additional physics analyses to be used in a third report but they will not be as in-depth as the work already done. The plan is to submit this third report in mid-2012, and it will contain the final technical solutions and recommendations for building a CLIC experiment.
"CLIC is not the only linear collider project on the table. The International Linear Collider is also looking in a similar area but it is important for the community to have options," said Strube.
"There is still no clear consensus in the community about which of the two machines is the one we should build, but studies such as the ones that we have done using the grid can help define the benefits and differences of each."
The 3rd CLIC report will also be used in an update to the European Strategy for Particle Physics, which will be updated in spring 2013.
This is an edited version of an article that first appeared on the GridPP website.