Expressions of interest are welcomed from potential PhD students. A list of possible research topics for PhD is provided below, organised by the main areas of my research. These are examples of relevant projects either recently started or currently in development. If you have any questions or ideas how to extend or compliment these topics please get in touch!
1) Turbulence Modelling & Simulation
a) Turbulence Simulation for complex geometry environments using Physics-informed Machine Learning
The environmental and economic needs for disruptive improvement in predictive fluid dynamics capability require a new paradigm for the analysis of very large data sets. Complex fluid dynamics applications often push the limits of traditional turbulence models and give rise to non-local flow features which are intrinsically beyond their reach. Particularly so in energy and transport sectors - the focus of this project - where, for instance, engine and aerodynamic efficiency impact directly on the environment. It is computationally unfeasible to apply the more accurate Direct/Large Eddy Simulation across the entire domain and so there is a clear need for methods that offer intelligent compromise. There is growing recognition that future turbulence simulation methods must use machine learning algorithms with the data and results of experiments and high-fidelity simulation methods to improve turbulence models. Physics-informed Machine Learning (PIML) is the process of updating the functional form of Reynolds stresses in turbulence models based on available data and will complement world-leading expertise in turbulence within the School of MACE. The philosophy in the present project is to develop and employ PIML enhanced turbulence models in an embedded eddy simulation approach, which adapts automatically to the geometry according to variations in flow complexity.
b) Dual finite volume - lattice Boltzmann turbulence simulation for interactive aerodynamic analysis
This project develops an existing virtual wind tunnel prototype designed for automotive design and engineering analysis. The existing system allows the import of digital models into a virtual reality (VR) environment as well as real-time simulation of the air flow around the geometry. This project will develop this capability further, extending the approach to allow the simulation of higher Reynolds number flows in real-time. The major proposed innovation versus the prototype is to limit turbulence simulation to a subset of the virtual domain, running on a GPU-accelerated solver, while the rest of the domain is conducted using lower order methods on CPU. In addition, a new, more efficient integration between the solver and the visualisation environment will be developed and more effective means of interaction and visualisation introduced. The approach to achieving this through this project is threefold; 1) The operational range of the flow in the tunnel will be extended by including Hybrid RANS-LES turbulence modelling. 2) Levels of interactivity will be increased to facilitate a wider range of parameter variations. The ability to specify flow direction and upstream turbulent intensity at run-time will be added to the solver interface and exposed to the user through the game engine. 3) Bottlenecks associated with GPU-CPU data transfer and mapping of simulation data will be circumvented through reorganisation of the data storage patterns. The work will be supervised by Drs Adrian Harwood and Alistair Revell based at the School of Mechanical, Aerospace & Civil Engineering.
c) Zonal RANS modelling using Artificial Intelligence
With ever-increasing computational power and improved user understanding of scale resolving methods, novel developments in RANS have reduced significantly. However there remains a clear need for efficient and reliable RANS modelling, which is likely to remain for the foreseeable future. Industrial design in particular would benefit from faster and more accurate schemes to reduce design cycle whilst examining a broader range of parameters. Many other applications need only bulk flow quantities and thus requisite time averaging with scale-resolving methods limits scope. The need for true grid-refined solutions is also difficult to meet with LES and hybrid methods; both of which tend naturally towards DNS. However in spite of these motivations for RANS progress has stalled on account of wide range of accuracy reported for some flows. Part of the problem with RANS is, given their limited degrees of freedom, they tend to be developed to be good predictors of a particular physics at the expense of others. In particular there is a need for simple models capable of predicting turbulence anisotropy; i.e. NLEVM or EARSM. For a specific range of flow types their ‘predictive efficiency’ (i.e. accuracy/convergence time) is significantly higher than full stress transport modelling approaches. The expertise in turbulence modelling then lies in understanding which model works best where. The objective of this project is to develop a framework to identify a general scheme that enables different RANS models to be used at different physical locations in a domain, potentially in a combination of both steady and unsteady mode. A major element of this work would then be the development of a Artificial Intelligence system to guide the sizing of a zone, and the selection of which model to be used within it. Ultimately the framework should become as automated as possible, with minimal user input.
2) Bioinspiration for engineering
d) Elucidating the role of wing flexibility in insect flight mechanics
The understanding of flapping flight of insects and bio-inspired micro air vehicles requires detailed insight into aerodynamics and structural mechanics. This is an interesting multi-disciplinary problem involving not only the generation of aerodynamic forces for weight support against gravity, but also the significance of structural properties of the wing frame itself and insight into the feedback control mechanism governing flapping kinematics. Most of the flapping flight research to date assumes rigid wings. Clearly, insects exploit their wing flexibility in different ways; however, understanding in this area remains incomplete. This project offers an exceptional PhD candidate the opportunity to work with a world-leading team of engineers to advance this field. The project will investigate the flight mechanics of flexible insect-inspired wings using a combination of numerical, theoretical and experimental techniques. In the first instance we will employ the lattice Boltzmann method, coupled to a finite element solver via the immersed boundary. We will then provide detailed validation of our in-house computational software against a series of lab-based tests designed to investigate the role that wing flexibility plays on flight mechanics. Following subsequent model refinement we aim to envisage a wide parametric investigation in terms of wing construction, shape and flapping motion. A detailed investigation of this kind into the role of wing flexibility has hitherto remained beyond reach of engineering research and as such this project will enable significantly improved understanding of flight mechanics relevant to the advancement of the field of micro-air vehicles and will inform evolutionary biologists on the role of wing morphology within life style of insects.
e) Piezo-electric/photo-voltaic material for passive energy-harvesting at scale
In this project a concept for simultaneously harvesting both solar and wind energy via an array of flexible flaps combining photovoltaic cells with piezoelectric material will be developed. The primary application is to provide continuous power supply to remote sensors or platforms where a power supply is required over a long duration. Inspiration is taken from the motion of flexible structures in the natural world. Whilst yield from a single such device would be low, the concerted motion of an array of many is envisaged to provide a low cost and reliable source of power. The project would contribute to existing work in this area at the School of MACE and will provide a framework for learning aspects of both numerical and experimental work. Specific tasks include: 1) Survey of state-of-the-art for energy harvesting technology; 2) Testing of available of-the-shelf micro solar panels and identification of potential candidates for dual piezo-solar energy harvesters. 3) Development of semi-empirical model for projected energy generation. 4) Conduct basic wind tunnel tests of the inverted-flag piezo energy harvester. 5) Comparison of experiment with numerical simulations to provide validation of the numerical tool. Conduct simulations to extend parametric testing of the inverted flag concept.
f) Nature inspired flow control for aircraft aerodynamics
The field of biomimetics allows one to mimic designs and ideas gleaned from the observation of the natural world. In the context of flow control, aerodynamicists have long taken inspiration from birds and fish in search of viable techniques for drag reduction and lift enhancement. Inspired by the pop up of birds feathers in certain flight modes, the amelioration of aerodynamic performance via a Porous and ELastic (PEL) coating has been the focus of recent research. The PEL coating works by reconfiguring/adapting to the separated flow, thereby directly changing the near-wall flow and the subsequent vortex shedding; which can lead to reduced form drag by decreasing the intensity and the size of the recirculation region. In contrast to classical Vortex Generators, which are fixed, and act upstream of the separation point, the flexible PEL coating will be activated only when the boundary layer is separated, and will interact with the flow recirculation zone to reduce drag via interaction with the near wall flow. The objective of this project is to investigate the performance benefit these technologies can deliver for flow at higher Reynolds number, relevant for the next generation of aircrafts. This achievement would contribute to reduced CO2/NOx emissions and improved flight performance in terms of delayed stall and reduced vibrations induced by e.g. gust loading; especially during take-off and landing where boundary layer separation plays a crucial role.
3) Biomedical Engineering
g) Biomechanical Investigation of Repaired Tetralogy of Fallot and Coarctation of Aorta
Congenital Heart Disease (CHD) is a heart condition resulting from an abnormality in heart structure or function that is present at birth. In the UK alone, there are about 4,600 babies born with congenital heart disease each year. Computational Fluid Dynamics (CFD) has had a profound impact on cardiovascular medicine in the past decade and will be used in this PhD project to assess its potential as an assistive tool in diagnosis and treatment of two CHD conditions, namely Tetralogy of Fallot (ToF) and Coarctation of Aorta (CoA). This interdisciplinary project will also identify haemodynamic features that correlate with need for re- operation in the case of ToF and as a tool to predict hypertension in CoA. The overall aim of this interdisciplinary PhD project is to explore the potential of novel Computational Fluid Dynamic (CFD) approaches in modelling two CHD lesions including Tetralogy of Fallot (ToF) and Coarctation of Aorta (CoA) and to accelerate the development of innovative nonsurgical (catheter/transcutaneous) solutions. This is achieved through the following main objectives: 1) Data collection from 8-10 child patients; 2) Conduct CFD simulations to evaluate key hemodynamic conditions for both ToF and CoA; 3) Perform morphological characterisation of the image sets and compare with CFD to identify post-repair evolution in the patient-specific models; 4) Incorporate additional data from the registry such as genetic characterisation and modifiable/non-modifiable risk factors into a data-mining framework.