An Exploration of Methods to increase the speed of Array-based Ultrasound Imaging

Abstract

Ultrasound imaging is a widely used technique for non-destructive testing of large structures and
surfaces and various other range of inspection needs. The use of array based imaging techniques
has grown in recent years, with use of techniques, such as the Full Matrix Capture (FMC) and
Total Focusing Method (TFM), becoming industry standards. In an ever growing industry,
the methods of non-destructive testing of in-service becoming more and more necessary for
safety critical analysis. Limitations of the current state of the art techniques due to their data
acquisition rate means long inspection times for large surface area scans for ultrasound testing.
This thesis aids TWI ltd, the sponsoring company for this project, to increase data acquisition
rates by reducing the number of transmission signals transmitted into a sample.
This thesis explores several techniques to increase the rate of data acquisition in array
based imaging. Allowing for quicker scan speeds in specific industrial scenarios where faster
scans would be a desirable outcome. Current fast based algorithms which are being developed
for industrial deployment are investigated with the use of a hybrid linear model, in the forms
of Plane Wave Imaging (PWI), Virtual Source Aperture (VSA) and Radial Virtual Source
Aperture (R-VSA). These algorithms are then scrutinised and compared with FMC and TFM
with the use of a performance index defined as the Array Performance Index (API). After a
thorough evaluation, the results define a basis for optimised characteristics for each algorithm.
In each case, it is determined PWI and VSA perform similarly and are affected in different
ways depending on the location of an artefact in the x-z direction, while R-VSA was found to
result in worse performance relative to PWI due to its wavefront not being circular but more
plane-wave like.
Coded Excitation signals have a wide range of use, and have been implemented for ultrasonics
as a means of increasing signal-to-noise ratio (SNR). Following on from the above investigation,
this thesis uses these signals in the form of Linear Frequency Modulated (LFM) signals and
Pseudo-Random Bipolar Signals (PRBS) to increase the rate of data acquisition in (PWI) by
using a characteristic of these signals defined as partial orthogonality. The PWI algorithm is
adapted into the form of Rapid Plane-Wave Imaging (R-PWI) and Ultra-Rapid Plane-Wave
Imaging (UR-PWI) respectively. From analysis, the results show that R-PWI is able to increase
the rate of acquisition by approximately double, while UR-PWI is capable of a speed increase
more than 12.5 times. However, both techniques have disadvantages and advantages. It was
determined that R-PWI is best performing at fast frame-rates (1300-2500 theoretical frames
per second {fps}), while UR-PWI performs the best at ultra fast frame rates (> 2500 fps),
albeit at a lower SNR than if PWI was used at its relatively low frame-rates.
This thesis has provided a basis of optimising and standardising new techniques, PWI and
VSA, and introduced new techniques with the use of coded-excitation signals to further reduce
the number of transmission cycles required to perform an inspection.

Date of Award	19 Mar 2024
Original language	English
Awarding Institution	University of Bristol
Supervisor	Bruce W Drinkwater (Supervisor), Paul D Wilcox (Supervisor) & Miles Weston (Supervisor)