This project is to develop and exploit an ecosystem of digital twins of the main components of the FASTBLADE Facility.
The School of Engineering at the University of Edinburgh is currently building a first-in-class EPSRC funded structural composites research facility (FASTBLADE) for fatigue testing of tidal turbine blades. This facility is mainly used to a) determine the static loading performance of the blade (stiffness-deflection curve plus full strain mapping of the surface in strategic sections of the blade, b) perform a cyclic loading test to 10 million cycles in cantilever mode. www.fastblade.eng.ed.ac.uk
FASTBLADE can be divided into five systems that complex interact together in a single environment. These systems are:
1. FASTBLADE Reaction Frame FEA .
2. Hydraulic System (Pumps, pipe network and actuators).
3. Control System and Data Acquisitions System.
4. Cooler Network and Oil Conditioning System.
5. Building Information Management System.
These systems, and the fact that the facility is located in an industrial-academic setting, provide a unique opportunity to develop robust digital ecosystems of Digital Twins that can improve asset management, structural health monitoring, and industrial processes to deliver environmental, economic benefit. Create and combine the different digital twins into a digital ecosystem will be undertaken in this PhD research.
This research consists of five main parts;
A) Improvement of current digital twin systems (systems 1-2),
B) Design and creation of missing digital twins systems (systems 3-5),
C) Design of protocols to communicate and exchange data between digital systems,
D) Validate the digital twins with collect data,
E) Identify better ways of operating the facility to keeping cost and energy usage to a minimum.
During the first part of the project, the PhD candidate will review and optimize the existing digital twins. The candidate will develop the rest of the digital twins in the second stage and transfer lessons from stage one. In the third stage, the candidate will create the ecosystems where the five different digital twins interact and exchange information. At this stage, the candidate will develop protocols to exchange data from different software and programming languages. The fourth stage will require integrating measured data and anomaly detection algorithms to perform the system validation, involving any required tuning. The final stage implies the application of machine learning to understand the holistic performance of the facility is a step change from simply understanding the local response of an asset. Machine learning outputs will be used to correlate the main features of interest, i.e., energy consumption, test cycle velocity and specimen properties such as stiffness. This will reduce the uncertainty in the initial configuration (parameter selection) and reduce the computational times.
Minimum entry qualification - an Honours degree at 2:1 or above (or International equivalent) in Mathematics, Engineering or equivalent, and demonstrable knowledge in programming languages (Python and Matlab), possibly supported by an MSc Degree.
This position is only open to Home and EU students.
Funding (tuition fees and stipend) is available at the Home tuition fee rate.