Putting turbulence to work

John Cater, Department of Engineering Science

Home

•

Project

•

Putting turbulence to work

Optimising the new generation of wind turbines

As the energy sector seeks to provide more of our power consumption needs from renewable resources, there is a constant pressure to extract more generation from existing infrastructure, such as existing wind farms, or to produce more from a particular new site. Research in the Department of Engineering Science led by Dr. John Cater, working along with the Department of Mechanical Engineering, is focussed on optimising the power generated from the new generation of wind turbines currently being installed around the world. High performance computing is used to solve the non-linear equations of motion for a wide number of wind flows and weather scenarios, and the output data is then used to change the way turbine controllers behave in extreme wind events to decrease turbine damage and to increase power output. The figure below shows a gust, a region of high speed wind flow (shown in red), encountering an off-shore wind farm.

In this simulation, each turbine operates individually and responds to the increased speed by changing its blade rotation speed and pitch angle. Although the gust starts out a large coherent structure (at the left-hand side of the picture), it is broken up by the presence of the front row of turbines and turbulent wakes are generated, which produce a very ‘messy’ flow at subsequent turbines. These messy wakes are responsible for high fluctuating loads on the turbine shafts. This work is world-leading, and has led to a number of recent high-profile publications. Other projects using the computing facilities of the NeSI Pan cluster include simulating the flow through the human upper airway, where turbulence is important in the mixing of gases, such as CO2 during the breathing cycle, and in the generation of some parts of speech (called fricatives).

Simulating turbulent flows

The Pan cluster at the University of Auckland is used for a wide range of turbulent flow problems to generate statistically significant data sets to better model momentum and energy transport. These simulations produce enormous file sizes that are too large to be processed on desktop computers (more than 30GB per timestep, and hundreds of timesteps are needed). The large memory and storage capacity of Pan made these simulations possible. The turbulent flow data generated to date has been used to improve the efficiency of offshore wind farms, to create more sophisticated breathing support devices and to improve our understanding of sound production. The computational fluid dynamics (CFD) software called ANSYS with CFX is used for the simulations.

The next projects that will use Pan will be focussed on using turbulent flows to improve the productivity of the primary industries sector in NZ. Dr Cater’s group will use the parallel computing facilities on the Pan cluster to model the complex mechanics of the bovine rumen and to study the discharge of effluent in marine aquaculture such as that from mussel farms and fish cages.