No abstract

Positron emission particle tracking (PEPT), a powerful technique for studying fluid and granular flows, has been developed at Birmingham over the last 30 years. In PEPT, a “positron camera” is used to detect the pairs of back-to-back photons emitted from positron annihilation. Accurate high-speed tracking of small tracer particles requires a positron camera with high sensitivity and data rate. In this paper, we compare the sensitivity and data rates obtained from the three principal cameras currently used at Birmingham. The recently constructed SuperPEPT and MicroPEPT systems have much higher sensitivity than the longstanding ADAC Forte and can generate data at much higher rates, greatly extending the potential for PEPT studies.

Positron emission particle tracking (PEPT) is a non-invasive technique used to study fluid, granular, and multi-phase systems of interest to academia and industry which employs position-sensitive radiation detectors to record gamma rays in coincidence and track the movement of discrete sources. A modular detector array, the Large Modular Array (LaMA), has been constructed at the University of Birmingham's Positron Imaging Centre (PIC) to enable custom detector geometries. In order to estimate the LaMA's performance characteristics prior to experimentation, assist in developing optimised camera geometries, and determine ideal PEPT tracer characteristics, a Monte Carlo model of LaMA is created and subsequently validated with experimental measurements. Validation is achieved through comparisons of the spatial resolution and count-rate response following the National Electrical Manufacturers Association (NEMA) industry standard protocol. Notably, the model's pulse-processing chain is autonomously calibrated to match experimental measurements using a recently developed technique which applies an evolutionary algorithm. Ultimately, the simulated spatial resolution matches the experiment to within 5%. In addition, the total, true, and corrupted count-rates are reproduced to a mean error of 3.41% over the range of source activities tested. This calibrated detector model strengthens the PIC's modelling capabilities. To facilitate future research, this model has been made publicly available through the PIC's GitHub repository.