Reconstruction of Non-prompt Tracks with the ATLAS Detector

Introduction

A variety of new physics scenarios give rise to long-lived particles that will tend to decay to charged particles far from the primary interaction point inside the ATLAS experiment. In some models, such as R-parity violating SUSY, these particles may be isolated while in others, such as "Hidden Valley" sectors, they are produced as part of a jet. SLAC is actively involved in preparing to search for these signatures with early LHC data. The key problem in this effort is the fact that reconstruction of tracks without a constraint to the interaction point is very challenging in the dense LHC event environment. The critical task of triggering on these events so that they are recorded is even more difficult given the limited algorithmic complexity allowed by the brief time for trigger decisions.

With a large number of measurement points in the tracking system and a calorimeter having fine spatial resolution, the ATLAS experiment may have a unique capability to trigger and reconstruct non-prompt tracks at the LHC. While some algorithms already exist that may reconstruct such tracks, these algorithms were not optimized with non-prompt tracks in mind and it is not yet known how well they will perform when triggering on and reconstructing these signatures. Meanwhile, other approaches that may be critical to success have not yet been tried. A prime example is calorimeter-seeded tracking, which uses measurements in the calorimeter to "seed" a trajectory from outside the tracking volume. This approach was developed for the CDF experiment and has been successful at providing pure and efficient tracking well beyond the acceptance originally envisioned for the tracking system. Due to the precise pointing resolution of the ATLAS calorimeter, this approach should allow the reconstruction of isolated non-prompt tracks and possibly tracks in jets as well. Implementation of this algorithm in the trigger will be difficult, but may also be possible.

Possible Projects

• The study and optimization of existing tracking algorithms for non-prompt tracks for R-parity violating SUSY and/or Hidden Valley searches. This would include creation of simulated data for these physics models, running standard tracking code to study the performance of the various default algorithms and optimizing those tracking algorithms for non-prompt tracks in the context of these physics scenarios. In the most ambitious outcome, this optimization could evolve into an entirely new tracking algorithm for non-prompt tracks.
• Study the reconstruction of particles in the calorimeter to devise an algorithm for creating calorimeter-based track stubs that can be used to seed trajectories in the tracking detectors. This would include running and understanding the calorimeter reconstruction, producing a set of initial track parameters from reconstructed calorimeter objects and devising a technique to use these seeds to constrain the standard tracking algorithms in adding hits from the Transition Radiation Tracker (TRT) and the silicon tracking detectors. Work would begin with reconstruction of prompt tracks in Standard Model physics samples for benchmarking against existing algorithms and proceed to the study of performance for non-prompt tracks in the context of R-parity violating SUSY or Hidden Valleys.

For additional information, contact Tim Nelson.