EngD Studentship in Improving Ultrasonic Imaging using Machine Learning


University of Strathclyde

Supervisors: Prof Charles Macleod & Dr Katy Tant

Industrial Partner: Rolls Royce

This EngD will work collaboratively between the Centre for Ultrasonic Engineering (CUE) the Department of Mathematics and Statistics (Dr. K. Tant) and Rolls Royce.

The ultimate aim of the project is to investigate enhanced ultrasonic inspection directly at the point of manufacture to deliver high-quality components right, first time. It is vital that wave propagation and the effect of refraction and scattering is well understood within these complex samples as we seek to utilise advanced imaging and machine learning approaches to compensate for these undesirable effects. Additionally, characterisation of the spatially varying material properties within the sample is sought to further enhance defect detection and characterisation capabilities.

This EngD project will look at thermal and micro-structural compensation methods to consider inspections of complex build geometries. Two research questions will be addressed: 

i) Can complex and dynamic component geometries be accurately mapped out in near real-time using in-process ultrasonic time of flight data, where extreme thermal gradients cause distortion of the expected wave paths? 

ii) Can this knowledge of the component geometry, coupled with models of the thermal gradient (as a function of process parameters) be used to better constrain and drive the microstructure mapping problem? Fully automated in loop capability will be demonstrated on test structures manufactured for aerospace and energy applications (with other industrial members applications supported as appropriate). 

The student will work in collaboration and spend time on site working with the lead industry partner to gain a greater appreciation of the specific industrial challenges and opportunity for automated inspection during fusion welding.

The student will receive an enhanced EPSRC stipend, while also having access to substantial international travel and project funds.

This project promises to be an exciting, fun and industrially relevant project, working alongside skilled engineers and scientists with state-of-the-art robotic equipment to delivery meaningful industrial change. 

Get in touch at FIND-CDT@Bristol.ac.uk