Radiation transport calculations are key to designing safe and efficient neutron instruments at spallation sources such as ISIS. The radiation transport codes are being developed continually to utilise modern high performance computing facilities to enable us to tackle larger problems within a reasonable timeframe. The two main methods for radiation transport are Monte Carlo methods and deterministic methods. Deterministic methods discretise space, energy and angle to solve the transport equations usually on a regular structured grid which can scale well on 100-100,000 processors. This project will investigate the potential application to neutron instruments of a recent deterministic transport code currently under development at Imperial college. The code aims to use an unstructured mesh to define the geometry and achieve similar levels of scaling on high performance computers. The project will aim to convert a simplified model of a neutron instrument into the new code, and compare with Monte Carlo and structured grid results for the same problem. This will involve meshing the geometry, preparing the nuclear data, using UK leading high performance computing and analysing the data.
Supervisor: Steve Lilley (firstname.lastname@example.org)
You might particularly like this if you are interested in any of these: scientific computing, engineering, physics, applied mathematics.