• Trigger
• Tracking
• Fisica del sapore
• LHCb

Over the course of the last 30 years, the Standard Model (SM) of elementary particles and their interactions has been extensively tested by experiments, which confirmed all its predictions up to the TeV energy scale, establishing it as an extremely predictive theory of microscopic physics. Nevertheless, open questions remain, which suggest the need for a more fundamental theory involving new particles and interactions at higher energies than probed so far. Identification of such theory is the main target of current high-energy physics.

Precision measurements of lower-energy observables that are sensitive to the potential exchange of virtual non-SM particles and are accurately predicted by the SM allow exploring energies far higher than those directly accessible to particles accelerators.
Weak interactions of quarks offer a rich phenomenology, in which possible discrepancies among many independent measurements may provide early indications of contributions from non-SM phenomena. The dynamics of the charm quark is particularly promising. Charmed hadrons are directly sensitive to non-SM particles that couple to up-type quarks (electric charge +2/3), whose phenomenology is less effectively constrained using kaon or bottom hadrons. However, the most interesting charm observables are so suppressed that only recently experiments started to access event samples copious enough to provide discrimination between SM and non-SM phenomena. This led already to a few intriguing anomalies observed in charm decays into charged hadrons. While a conclusive interpretation of such deviations is lacking, consensus is that more accurate experimental measurements are strongly desirable to fully exploit the charm potential.

The LHCb experiment at the LHC proton-proton (pp) collider is leading the exploration of charm transitions. A single-arm forward spectrometer dedicated to the study of heavy-flavor decays, LHCb collected a sample corresponding to nearly 3 fb-1 of integrated luminosity at 7-8 TeV center-of-mass energy in 2010-2012, and will resume its operations in early 2015.
The large charm-quark production cross-section offers access to charm dynamics with unprecedented sensitivity. However, even larger background processes and technical limitations of the data acquisition system make the efficient exploitation of the full physics potential challenging. Moreover, the increase of LHC collision energy and rate expected in 2015 will put an additional strain on the experiment's capabilities to collect charm decays, which currently consume a large fraction of the online and offline computing resources.

In this work, I explore simpler, more effective selection strategies of hadronic heavy-flavor decays than currently used in LHCb.
Typically, hadronic bottom and charm decays yield final states with multiple charged particles. Such particles are likely to originate in a space-point displaced from the pp collision point (primary vertex), due to 0.5-1.5 ps typical values for heavy-flavor lifetimes. Current selection of hadronic heavy-flavor decays in LHCb is based on variables sensitive to the displacement of one charged-particle trajectory (track) from the primary vertex. The need to reconstruct the primary vertex position in complex events with O(100) tracks and up to 5 primary vertices makes this strategy computationally complex and overly sensitive to the details of the environment. My work demonstrates that simpler selection methods exist, which offer a comparable discriminating performance in terms of efficiency on charm signal, background rejection, and event accept-rate.

The transverse impact parameter is the two-dimensional distance of closest approach, in the plane transverse to the beam, of the track to the beam itself. By using only information from the plane transverse to the beam, it does not require primary-vertex reconstruction. Additionally, for two or more final-state particles, the relative sign between the transverse impact parameters of any pair of them provides additional discriminating power.
I study the discriminating performance of the transverse impact parameter using fully reconstructed D0->Kpi decays as benchmark. The D0->Kpi decays are among the most generic representatives of typical features of hadronic heavy-flavor decays. In addition, they are copious enough to allow a comprehensive study based on experimental data only, without relying on simulation. Due to this abundance, I am able to reconstruct a special sample of such decays in LHCb data, that is free from biases from the usual online event-selections (trigger).

With this sample, I compare the discriminating performance of transverse impact parameter and LHCb standard tracking variables. I explore selections imposed on either one or two tracks. In the one-track scenario, the kaon or the pion are required to have transverse impact parameter in excess of a given threshold. In the two-track scenario, both particles are required to meet the criterion, with or without an additional requirement on the relative sign. Analogous requirements are applied on standard tracking variables and the resulting performances are then compared.
In spite of its simplicity, I find that the transverse impact parameter has comparable performance to standard tracking variables, especially if high background rejection is required.
The decrease in signal efficiency at given background rejection never exceeds 10%. I also find that at given signal efficiency, requirements on two tracks improve background rejection of nearly 15% with respect to one-track selections.

In view of a possible future application of a transverse-impact-parameter-based selection in the LHCb trigger, I also study the event accept-rates associated to various thresholds in a sample of pp collisions unbiased by any online or offline selection. Online track parameters are used, as accessible in the first stage of the current trigger, to impose displacement criteria. I find that transverse impact parameter and standard tracking variables have comparable performances also in terms of event accept-rates.

This work demonstrates that the transverse impact parameter is a promising alternative to standard tracking variables for selecting hadronic heavy-flavor decays. It shows comparable discriminating performance against backgrounds, but is easier and faster to calculate, more robust against environmental conditions (such as high primary-vertex multiplicity), and therefore a better choice whenever computing complexity is an issue. It also shows that selections based on more than one track provide superior discriminating performance with respect to one-track selections, with no additional penalty in complexity. As a result of this work, the transverse impact parameter is now part of the standard LHCb variables available for offline data analysis, and is being considered for implementation in the software-based high-level trigger in the next LHCb run, in 2015.
A longer-term prospect is the incorporation of similar selection methodologies in a fast hardware trigger device, capable of performing a first selection of hadronic decays at 40 MHz, the full interaction frequency of the upgraded LHC, allowing an even greater improvement of the LHCb capability of collecting such decays in 2018.

This work has been partially developed during my permanence at CERN as a technical student, and large portions of it are documented as an LHCb internal note which is being finalized in these days.