PhD Studentship – Processing and Analysis Methods for Large Historical Ultrasonic Structural Integrity Monitoring Data
Institution: National Structural Integrity Research Centre (NSIRC)
PhD Supervisor/s: Professor Keith Worden, Dr Kai Yang
Application Deadline: Ongoing
Funding Availability: Funded PhD Project (Students Worldwide)

NSIRC is a state-of-the-art postgraduate engineering facility established and managed by structural integrity specialist TWI, working closely with lead academic partner Brunel University, the universities of Cambridge, Manchester, Loughborough, Birmingham, Leicester and a number of leading industrial partners. NSIRC aims to deliver cutting edge research and highly qualified personnel to its key industrial partners.
Ultrasonic NDE (Non-Destructive Evaluation) is used in man-made structures to find flaws and to assure their structural integrity. Sometimes this involves the use of relatively low frequency ultrasound (typically 20 kHz to 100 kHz) that is transmitted over tens of meters to rapidly screen for flaws in large volumes (volumes which might otherwise be inaccessible to other techniques). Data gathered automatically by permanently-installed instrumentation, allows for trends to be detected and flaw growth to be detected at an early stage.
Conventionally, these structures are subjected to time-driven preventive maintenance. However, if no repairs are required, the costs of shutting down, emptying and cleaning the tank are largely wasted. Alternatively, an ultrasonic guided wave-based structural health monitoring technique has been developed to evaluate the condition of tank floors and to increase the time between the services. Another common example is the use of long range ultrasonics to monitor tens of meters of pipeline (which is potentially buried or otherwise inaccessible) from instrumentation placed at a single axial location. Such pipeline monitoring systems are often used to monitor pipeline sections where they pass under roads.
Project Outline:
This study will design and integrate machine learning methods to build a model based on historical ultrasonic monitoring data. The collection of large data sets is often expensive because it puts a greater demand on the design and production of electronics and other systems, which means there is a desire to develop monitoring methods that require less data. The trade-off between quantity of data and defect detection capability is as yet uncharacterised, so it is currently not possible to design a system to meet an informed compromise. In addition, various corrosion levels may require different confidence levels. For example, for 75% metal loss, a fairly high confidence level (e.g. 99%) is necessary; for 25% metal loss, a less confidence level (e.g. 75%) is requested. There is a need to characterise how the amount of data and the way it is collected affects both the system’s detection capability and the confidence in that capability.
In order to estimate the corrosion severity and predict the safety of structures, a supervised machine learning algorithm (such as an artificial neural network), will be developed and used. Moreover, the environment temperature or operating condition is highly relevant to the guided wave transmission. This includes the temperature variation from stored liquids and temperature effects on the sensor coupling materials. Data normalisation issues are vital for developing a robust Structural Health Monitoring system.
About University/Department:
The University of Sheffield is a public research university in Sheffield, South Yorkshire, England. Sheffield is a multi-campus university organised into five academic faculties composed of multiple departments. It had 19,555 undergraduate and 8,370 postgraduate students in 2015/16. Sheffield was placed 84th worldwide by The QS World University Rankings 2016.It is a member of the Russell Group of research-intensive universities, the Worldwide Universities Network, the N8 Group of the eight most research intensive universities in Northern England and the White Rose University Consortium. There are seven Nobel Prize laureates amongst Sheffield academics, six of which are its alumni or former staff
Entry Requirements:
Candidates should have a relevant degree at 2.1 minimum, or an equivalent overseas degree in Computer Science, Electrical Engineering or another scientific field that includes digital signal processing. Candidates with suitable work experience and strong capacity in machine learning, programming or signal processing are particularly welcome to apply. Overseas applicants should also submit IELTS results (minimum 6.5) if applicable.

Administrative Contacts & How To Apply

For more details and how to apply please see here
For any more information or general enquiries please email:
Data Protection Statement
By registering your application you agree that your details may be held, processed and disclosed within TWI for the sole purpose of recruitment (and subsequent employment if applicable) and for compliance with applicable laws and regulations to which TWI is subject to. We will never provide, share or sell your personal information to a third party. TWI follow strict procedures in the storage of information and will retain the information you submit for recruitment for up to six months in case any questions arise in relation to the recruitment process. Thereafter, the information will be destroyed in a secure manner.

Funding Notes

This project is funded by TWI and academic partners. For Home students a £24k per annum scholarship is available which will cover tuition fees and provide a competitive enhanced STIPEND of £16-20kper annum for the duration of three years. Partial scholarships are provided for international students with funding up to £24k per annum for three years.

Funding Information

Add to list