Timed Track Seeding for the Future Circular Collider

Particle accelerators and detectors have been the principal tools to investigate nature at the smallest scale for more than 50 years. The Large Hadron Collider (LHC) at the European Organization for Nuclear Research (CERN) in Geneva is scheduled to end its operation after the high-luminosity upgrade program HL-LHC concludes around 2040. An effort to create experiments that can succeed the LHC in pushing the high-energy frontier resulted in the launch of the Future Circular Collider design study in 2014. This design study explores the opportunities a 100 km circumference tunnel in the Geneva area would offer. Such a tunnel could be used for another hadron collider, called FCC-hh, reaching a maximum center-of-mass collision energy of $\sqrt s = 100\,$TeV. This increase, which offers great discovery potential for new phyics, is possible due to the increased tunnel circumference and a research and development program that will result in $16\,$T magnets. Complimentary uses for the tunnel include an electron-positron collider (FCC-ee) and an electron-proton collider (FCC-eh), that can probe new physics via precision measurements. FCC-hh reaches the statistics required for many of the searches for new physics at the cost of an increase in simultaneous collisions per bunch crossing (pile-up), which will average at about $1000$, a five-fold increase over HL-LHC. This means an overall increased demand for computing resources and efficent reconstruction techniques, but poses particular problems for the processes of event reconstruction that scale unfavorably with pile-up. Track seeding, i.e. the problem of identifying track candidates from a subset of the track detector measurements, is a combinatorial problem that exhibits the worst scaling. This thesis studies the performance of current approaches to track seeding as applied to simulated FCC-hh data. It is shown that while track seeding as done at LHC experiments is not feasible for FCC-hh, the inclusion of timing information can reduce the combinatorics to a level that is comparable to HL-LHC.