ProfessorBen Thornber

Project Opportunities

Title: Computational Automotive Aeroacoustics

Summary of opportunity:

Current automotive aerodynamic design is driven by a myriad of requirements including minimum drag, minimum noise, favourable handling, thermal management and vehicle soiling. Our group have developed new approaches coupling steady Reynolds-Averaged Navier-Stokes methods with unsteady turbulence resolving methods such as Large-Eddy-Simulation. These are implemented in our very-high order accurate compressible in-house solver, permitting computations at flow speeds of interest to industry. This PhD will focus on the further development and application of this hybrid method to understand the fundamentals of automotive aerodynamics (particularly base-flows) and aero-acoustic predictions.

Opportunity synopsis:

The FLuD group has ongoing collaborations with industry and international Universities to push forward the state of the art in modelling of fluid flow around generic cars. Two recent problems have focussed on the flow physics of large scale separations at the rear of a car which substantially impact the overall drag. This involves the application of very high order accurate numerical methods and new unsteady turbulence models to resolve the time dependent evolution of the base flow. To date, these studies have been conducted in collaboration with Jaguar Land Rover, UK.  This PhD will take the existing solver and further improve the hybrid RANS-LES approach targeting practical Reynolds numbers, with aeroacoustics as a particular focus. Initial computations will be undertaken with the existing solver in comparison to experimental results from Hyundai detailing noise measurements around a simplified automotive body. Your role would be to derive and implement novel hybrid RANS-LES blending functions which are flow adaptive, requiring minimal a priori knowledge in particular about the boundary layers around the vehicle. The PhD will combine large scale computations on HPC systems with advanced numerical methods, in a collaboration with industry, to provide a fundamental understanding of noise production. In the later years, noise mitigation will be explored.

--------------------------------------------------

Project Opportunities

Title: Efficient Large Eddy Simulations of Helicopter Aerodynamics and Aeroacoustics

Summary of opportunity:

This project aims to push forward the state of the art in unsteady turbulent simulations of helicopters and UAVs flying in atmospheric turbulence generated over ships and through urban canyons. The developments aim to inform safe decision making for helicopter operations (potentially multiple concurrent platforms) around large structures.

Opportunity synopsis:

Within the Fluid Dynamics Research group we have developed unique algorithms to tackle this challenge, which enable us to run unsteady computations efficiently for platforms incorporating one or many rotor blades, such as helicopters or quadrotors. These numerical methods use cutting edge high-order accurate, massively-parallel CFD methods, yet with an unsteady Actuator Surface Model for the rotating components, which reduce the computational effort by factors of 100+ compared to fully resolving the rotor blades.  The aim of this project would be to further advance this extremely promising research thrust, targeting (i) aeroacoustic prediction of rotor blade noise using the newly developed platform and (ii) extension of the existing unsteady rotor blade model to co-axial blade systems. The project will be in close collaboration with industry giving you the opportunity to gain industrial experience as part of the PhD, and also for Australian candidates a generous top-up stipend is available.

--------------------------------------------------

Project Opportunities

Title: Understanding the growth of hydrodynamic instabilities in Inertial Confinement Fusion

Summary of opportunity:

The National Ignition Facility (NIF) in the USA has managed to achieve a fusion fuel gain of 1 in landmark experiments. However, the targeted gain is substantially higher than this value. Here, we partner with the NIF to develop algorithms and run computations to advance our understanding of the hydrodynamics instability and turbulence during the implosion process. Should this be understood and controlled, then much higher fuel gains are possible. This project builds on more than a decade’s work developing algorithms to simulate compressible turbulent mixing due to shock waves, and is currently supported by an Australian Research Council Discovery Project.

Opportunity synopsis:

Meeting the global energy demands for the second half of the 21st century is the key challenge of our civilisation. A future energy source which could address this with ease is nuclear fusion. The energy released from fusing one gram of Deuterium is 100,000kWh which is equivalent to burning 15 tonnes of coal. Given that there is 33g of Deuterium in every ton of sea water, the oceans could supply the world's future energy needs for over a thousand years even with conversion efficiencies of a fraction of a percent. With this in mind, fusion is the ultimate power source. Recently a landmark result has been obtained: for the first time on a laboratory scale a fusion reaction has generated a fuel gain of greater than 1. This outstanding achievement has been recorded at the US National Ignition Facility (NIF) at Lawrence Livermore National Laboratory (LLNL). This is a $4bn program aiming to demonstrate nuclear fusion of millimetre sized capsules containing Deuterium-Tritium (DT) fuel, in a technique known as Inertial Confinement Fusion (ICF). In ICF, fusion is attained through rapid compression of fusion fuel driven by 192 lasers delivering up to 500 Tera-Watt pulses over a period of nanoseconds. Despite this achievement, a practical power source would require a gain many times greater than the total laser power employed, which at present is 1.5MJ for a yield of about 15kJ. If fusion ignition can be achieved, the energy released will be approximately 1000 times greater than the current experiments. This significant gap must be closed before a practical power station can be realised. So why are the capsules under-performing? As the capsule implodes the initially solid material is transformed into a dense plasma by the high temperatures and pressures. The shell (also called `ablator') of the capsule and the layers of DT fuel within the capsule are not perfectly smooth, and this unavoidable surface roughness seeds fluid instabilities at the interfaces between the different layers. The amplitude of the surface roughness initially grows exponentially in time, and at peak compression can be 1000 times larger than the initial roughness height. This growth leads to penetration and mixing of the ablator into the nuclear fuel. This PhD will focus on the development of state-of-the-art massively parallel fluid dynamics software within the School of AMME to further our understanding of the development of these instabilities in NIF-like conditions. This project is in collaboration with scientists at Lawrence Livermore National Laboratory. The ideal student will have an affinity for programming (or keen to learn), at least an Honours 1 WAM and a solid mathematical/engineering/physics background.

--------------------------------------------------

Project Opportunities

Title: Compressible Turbulent Mixing

Summary of opportunity:

A fundamental understanding of compressible turbulent mixing is critical to advancing technologies in aeronautical, automotive, clean energy and acoustics. This project aims to develop state-of-the-art methods and further push forward our understanding of the role of turbulence.

Opportunity synopsis:

Compressible turbulent mixing is responsible for the growth of boundary layers around wings thus dictating drag, for the mixing of fuel and air in a combustor thus dictating reaction rates, and in producing noise in aircraft and helicopters. These flows are incredibly complex, consisting of a wide range of vortex sizes, where often the largest vortices are more than a million times larger than the smallest vortices. In addition, where two or more species are present, the mixing of these two species occurs due to diffusion, which acts at a very small length scale, yet can impact the evolution of the whole mixing layer (particularly in reacting flows). When turbulence interacts with solid surfaces, or with other fluids (when jets meet for example) then the deformation of structures within the turbulent flow field can cause noise. All of the above mechanisms are made more complex by the underlying compressibility of the fluid. So why not just undertake computations of these flows to give the needed insight? The reason is that these computations are extremely challenging, needing to resolve the aforementioned large range of vortex sizes, plus multiple species, and in some cases shock waves. Due to these reasons, it is not yet possible to undertake the `perfect computation’, and will not be for some time.  This project aims to shed light into compressible turbulent mixing problems with a particular focus on the development of algorithms capable of significantly improving on the current state-of-the-art, then applying these algorithms to understand outstanding fundamental problems in turbulence including self-similarity, influence of initial conditions, mixing in reacting flows and noise production.