Machine-learning for improving efficiency of power generation

Computational Fluid Dynamics (CFD) has become an integral part of the design of any energy generating device, from gas turbines to wind or tidal turbines. In industrial design, CFD relies heavily on models for the complex turbulent flows that feature in nearly all energy applications. However, the current industry-standard models are based on simplified assumptions that can lead to significant errors, and thus lack of accuracy, that reduce the impact CFD can have on technology development. In this project, a unique machine-learning framework for turbulence modelling, developed at the University of Melbourne, is used to improve the accuracy of turbulence models. The models will be developed by using large-scale data sets already available. Importantly, the newly developed models will also be tested in a full a-posteriori context on industrially relevant problems to assess their accuracy.

UoM Supervisors

• Dr. Cat Vreugdenhil, Doreen Thomas Postdoctoral Fellow in the Department of Mechanical Engineering
• Prof. Richard Sandberg, MEI Program Leader, Power Generation & Transport, and Chair of Computational Mechanics, Mechanical Engineering.

How to apply

Applications are closed.