Speaker
Description
The future Electron Ion Collider (EIC) in BNL, USA, will be the ultimate facility to study Quantum Chromodynamics (QCD) with an unprecedented accuracy. The ePIC detector is a general purpose detector, capable of providing the entire physics programme documented in the EIC Yellow report. The requirements for the EIC accelerator and the ePIC detector are challenging. In the EIC highly polarized electrons (~70%) will be colliding with highly polarized nucleons and light nuclei (~70%); unpolarized nuclei can be as heavy as Uranium nuclei. The centre-of-mass energy for such collisions will vary from 20 GeV to 140 GeV with peak luminosity reaching 10$^{33}$-10$^{34}$ cm$^{-2}$ s$^{-1}$.
The forward (hadron going) direction is characterized as the high x-Q$^2$ region, therefore in this region the majority of the high momentum hadrons will be produced. An efficient identification of those high momentum charged hadrons is essential to several key physics channels. A dual Radiator Ring Imaging Cherenkov (dRICH) counter will perform the identification of the high momentum charged hadrons. Furthermore, it will also be used to reject low momentum pions from electrons to complement the Calorimeter in reconstructing the Deep Inelastic Scattering (DIS) events. By providing charged pion–kaon separation from approximately 3 GeV/c up to 50 GeV/c (and even higher for kaon–proton separation) and electron–pion separation from a few hundred MeV/c up to 15 GeV/c, with at least 3 σ separation over a wide pseudorapidity range (1.5 < η < 3.5), the dRICH will serve as a cornerstone for the successful Semi-Inclusive DIS (SIDIS) programme of the ePIC collaboration.
The design of the dRICH faces several challenges coming from the global detector: shorter radiator length, presence of the solenoidal magnetic field of the ePIC detector and high radiation in the sensor region. A set of thorough and systematic simulation studies are essential to overcome these challenges and to deliver an optimal design that provides the ambitious physics requirements.
In this contribution we describe the realistic geometric model, optical parameters for the sensors and the two radiators used in the simulation studies as well as impact of noise in the detector performance. The simulation has been performed using the official simulation framework of the ePIC collaboration. We also present the identification and confusion matrix obtained from these studies. The simulation framework is capable of providing not only the detector performance used in physics studies, but also several microscopic design features that allow us to optimize the detector parameters and mitigate the design risk. Incorporation of a Bayesian optimization serves as an important tool to perform such detector optimization. We also introduce an AI/ML based algorithm involving Convoluted Neural Network (CNN) for future application for the dRICH PID.
| Speaker confirmation | Yes |
|---|