PhD Studentship: Developing a machine learning based approach to 2D and 3D hydride characterisation in zirconium alloys

This 4 year PhD project is funded by EPSRC (funding from Materials 4.0 CDT and Rolls-Royce Submarines). The funding is for UK students only; tuition fees will be covered and you will received a stipend set at the UKRI rate (£19,237 for 2024/25). The start date is 1st October 2024.

Zirconium (Zr) alloys are used to make the fuel claddings in nuclear reactors. These rods spend their life sitting in hot, high-pressure water. Hydrogen from the water can enter the zirconium and form regions of brittle zirconium hydride. These hydrides degrade the mechanical properties of the zirconium, which could cause premature failure, particularly during the transportation and handling of the claddings at the end of their lifetime. Over the past few decades, research into Zr-hydrides has been extensive but one challenge hindering further progression is the lack of a high-throughput, automated, unbiased approach to analysing and characterising hydrides in 2D and 3D. The use of deep learning (DL) based algorithms to tackle materials science challenges is gaining increasing popularity. DL methods have been effectively applied to solve a broad range of microstructural classification problems and yield state-of-the-art results as they can provide a scalable, unbiased, and highly reliable approach to feature recognition and classification. This project aims to apply DL methods to detect and extract hydride features from both 2D and 3D datasets. A series of functions will then be developed to extract metrics such as hydride length, orientation, and connectivity from the data. From this, we will be able to for the first time quantitatively characterise hydride micrographs reliably to gain an improved understanding of their precipitation behaviour.

Applicants should have, or expect to achieve, at least a 2.1 honours degree or a master’s (or international equivalent) in a relevant science or engineering related discipline.

This project is supported by the Materials 4.0 CDT (https://www.royce.ac.uk/cdt/) and Rolls-Royce Submarines Ltd

Before you apply, please contac tDr Mia Maric:  mia.maric@manchester.ac.uk

Advert information