2017 CMS-Caltech-CERN Summer Students

Instrumentation

Student 1

AI Methods

Vitória Barim Pacela, CERN, LCD particle ID and regression with Amir, Maurizio and JR

Silviu-Marian Udrescu, CERN, LCD particle ID and regression with Vitoria et al.

Yikai Huo, Caltech, Charge Particle Tracking (with Dustin and Stephan)

Sarang Mittal, Caltech,

Nishad Maskara, Caltech,

Avikar Periwal, Caltech, Based Multi-class Jet Tagging with Deep Learning (idea from Jesse Thaler)

Physics

Jamie Bamber, Caltech, Dark Matter (low mass)

Gillian Kopp, CERN?, Dark Matter + displaced photons / Timing

Daniel Gawerc, CERN?, Dark Matter + displaced jets / Timing

Projects

Detailed simulation of planar semicondutor sensors for precision timing

Planar semiconductor sensors are used for precision timing detectors in high energy physics. In calorimetric applications the sensors are exposed to a shower of ionizing radiation induced by a high energetic primary particle. The goal of the project is to perform simulation studies of the shower interactic with the sensor to get a precise understanding of the energy deposited in the sensor material. Different sensor materials such as Silicon, Silicon-Germanium and Cadmium-Tellurid will be studied as well as different sensor geometries exposed to different parts of the particle shower. The simulations are preformed with the GEANT simulation package either in an enbedded application derived from CMS sofware or in a standalone application.

Simulation studies of the CMS ECAL photon timing using crystal calorimeter and timing layer information.

The Phase II upgrade of the CMS detector will enhance the timing performance of the crystal calorimter to about 20 ps and possibly add a precision timing layer in from of the calorimeter. The calorimeter is primarily destined to measure the time of arrival for high energy electromagentic clusters and the timing layer will target charged particles. However high energy electrons will feature a precise time measurement in the timing layer and the calorimeter. Further, the electron impacting on the calorimeter will result in a certain amount of front leakage. The goal of the project is to study the timing performance for electrons using both detectors, the impact on front leakage on timing measurements in the timing layer and to study the viability to use front leakage for calibration of the timing layer as well as timing measurements of photons.

Detailed study of regression based photon energy reconstruction

Regression algorithms are used to reconstruct the energy of photons in CMS. The algorithms use the topology of the hit pattern in the CMS ECAL and additional information from other subdetectors if available. The algorithms are trained on large MC samples. The validation of the trained algorithms is performed on MC samples as well as data. Choosing the input variable to the regression and evaluating the performance of the regression requires detailed knowledge of the respective detectors, the physics underlying the detection process and the capabilities of the simulation. The goal of the study is to develop a evaluation strategy for the regression results which is based on numerical evaluation of the results. As a second step a feedback mechanism shall be implemented which systematically tests the choice of input variables and their impact on the regression results.

Evaluation of the timing performance of light based detectors

Scintillator based detectors with SiPM or MCP readout are used for precision timing measurements of ionizing particles. The goal of this project is to use the exisiting picosecond laser test setup at CIT to test the performance of SiPMs, Scintillator/SiPM as well as Scintillator/WLS/SiPM combinations to evaluate their timing performance.