Final Year Dissertations

Design of invisibility cloakes

Meta-materials are materials that are made from human engineered molecules and that are designed to exhibit very specific material properties that often times do not appear in nature. Meta-materials can be designed that are non-uniform and non-isotropic and that can be used to built invisibility cloaks. These cloaks are areas that, when illuminated by an electromagnetic wave, do not result in reflection and transmission. This effectively renders these structures, and any object placed inside of them, invisible.

In this project you will investigate how to model scattering of an electromagnetic wave from such a cloaking cavity. This modelling strategy will then allow you to further opyimise the design of the bunker.

Organic design with subdivision surfaces

In scientific computing, computer aided design, and gaming, it is important to be able to store geometries in a format that allows both accurate description of the real-life system of interest and efficient manipulation and/or computation.

In all these fields, a simple enumeration of triangles approximately covering the structure is very popular. In CAD especially other models are prevalent: boundary representations, desccriptions using B-splines, constructive solid geometry, etc. For organic and other highly smooth surfaces, subdivision geometries provide an appealing and space efficicient description. In this project you will directly use such a description to start the modelling of physical scattering and transmission phenomena.

Accelerating boundary element methods on the GPU

GPUs are the ideal hardware to parallelise simple computations that have to be repeated over and over. This type of computations appears commonly in graphics and gaming, but also in scientific computing GPU hardware can be exploited to arrive at a solution in only a fraction of the time a traditional CPU would take to complete it.

In this project, you will write generic code that can execute on both traditional CPUs with limited parallel capability and GPUs that have many processing units that can simultaneously deal with indepent parts of the overal simulation. You will use these techniques to speed up, by a factor of ten or more, the modelling of highly complex optical and photonic systems.

Modelling of faster-than-light physics

What happens when light bounces of a mirror moving faster than light (in the encompassing medium)? Intuitively it is clear that light is not fast enough to escape the mirror, which will be chasing it through the medium. Clearly there must be a resolution to this paradox.

Using both analytic and numerical techniques, you will inspect this situation and come up with a modelling paradigm for the scattering and transmission of waves in such a situation.

This will open up avenues, not only in the context of this highly exotic and maybe purely hypothetical problem, but also in situations where a transmissive medium moves at much more moderate speeds. This for example of a beating heart that is inspected by illuminating it with a microwave signal, for diagnostic purposes.