Karlsruhe 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
T: Fachverband Teilchenphysik
T 109: Neutrino physics 12
T 109.5: Talk
Friday, March 8, 2024, 10:00–10:15, Geb. 30.21: Gerthsen-HS
Simulating LiquidO detectors for prototype research and development — •Ben Cattermole for the CLOUD collaboration — University of Sussex, Brighton, United Kingdom
LiquidO is a novel detector technology that makes use of the stochastic confinement of scintillator light around its origin in an opaque medium. To collect this light a lattice of wavelength-shifting fibers runs through the medium, with each fiber end leading to a SiPM. By analysing event topology LiquidO style detectors have strong particle identification down to the MeV scale. Subsequent background rejection capabilities of the LiquidO technology make it ideal for neutrino detection. LiquidO will be used in the Chooz LiquidO Ultra near Detector, CLOUD, planned to be a 5 to 10 ton above ground detector for reactor anti neutrinos.
I will report on my work which involves simulations of LiquidO based detectors for research and development purposes. These simulations are built in Geant4 and include the geometries of prototypes, the scintillator material itself, the reflectivity of the vessel, fiber position and simulations of light in the fibers themselves. Alongside prototyping, these simulations are used to generate machine learning datasets. The main machine-learning technique being considered is a convolutional neural network due to the lattice of fibers used by LiquidO detectors being easily mapped to a pixel grid image format.
Keywords: opaque scintilator; monte carlo; particle identification; nuclear reactor; machine learning