PhD student, 2016 intake

University: University of Strathclyde

Sponsor: University of Strathclyde/NDEvR

Background: Undergraduate MSci Mathematics, University of Glasgow June 2015.

Ultrasonic phased arrays generate a rich data set – the Full Matrix Capture(FMC) – that contains knowledge about the material in which the ultrasonic waves has travelled through as well as any flaws it encountered. In heterogeneous materials the material microstructure has a significant effect on the recorded waveforms suggesting that it is possible to reconstruct the material map (the so called inverse problem). Some progress has made in this direction in time-of-flight data from the FMC. However this discards a rich data set contained in the full waveform.

The aim of the project will be to develop a method to solve the inverse problem of recovering the material map in heterogeneous materials and, in so doing, improve the imaging of flaws. This will be done by:

– using Voronoi diagrams to mathematically parameterise the material geometry

– using a ray-based forward model to chart the path of the wave through the material.

– devising a suitable metric to quantify the misfit between the modelled waveform and the experimental counterpart.

– developing Bayesian and deterministic protocols for solving the inverse problem.

– using synthetic FMC datasets from finite element simulations complemented a series of experimental investigations on industrially relevant test samples.