(91b) Evolutionary Optimisation of a Pharmaceutical Feeder System

The transition from conventional batch secondary manufacturing to continuous direct
compression in the pharmaceutical industry requires significant efforts to optimise the
processes. This project focuses on the feeding stage, a challenge due to the absence of
intermediate operations to improve the flowability of the matter.

The initial aim of this project is to develop a well-validated in-silico discrete element method
model of a twin-screw loss in weight feeder, using a simple, free-flowing powder. Positron
emission particle tracking (PEPT) will be used to image a radioactive tracer within the system in
order to gain an insight into its internal dynamics. Various powder characterisation tools along
with their digital twins will be used to calibrate the measurable powder properties with the
corresponding in-silico parameters. The most suitable characterisation tool for this system can
then be established by comparing the model with the post-processed PEPT data and
determining which characterisation tool leads to the most accurate calibration.

Once a suitable characterisation method has been established, powders with more complex
flow behaviours can be placed into the model. The use of an evolutionary algorithm can then be
used to optimise the geometries of equipment, with the purpose of minimising the influence of
material attributes on the overall process robustness.