Environmental Technology: Clean engines on the front burner

Researchers have joined forces in an attempt to work out how jet engines can be operated using more air and less fuel

winner/flowtool/rolls-royce/cambridge university

Jet engines are a simple concept but a complex reality. Sometimes described as ‘a bucket that you burn fuel in’, their precisely machined components and painstakingly chosen materials are designed to eke as much of the energy embodied in the fuel into the force that propels aircraft through the air.

Making these engines as environmentally friendly as possible is also a complicated matter. As well as carbon dioxide, jet engines can produce poisonous oxides of nitrogen (NOx) under certain conditions. Understanding the complex interplays of gas flows, combustion and temperature that give rise to these conditions – and then finding out how to avoid them – is a daunting prospect, which Matthew Juniper of Cambridge University took on with the assistance of Rolls-Royce.

‘If you want to operate cleaner, with low NOx production, you need to burn with lots of air and less fuel,’ he said. ‘Having lots of air keeps the temperature down in the combustion chamber and it’s high temperatures that give rise to the NOx; essentially, you start burning the nitrogen in the air. So you want to burn colder, but of course the colder it is, the harder it is to keep a flame going.’

To make sure their engine is burning lean and not producing NOx, engine designers must ensure that the fuel and air are mixed as well as possible. Otherwise, the regions that are richer in fuel will burn hotter and the resulting high temperature will lead to NOx generation. However, the small ratio of fuel to air causes other problems. The conditions put the flame on the very edge of instability. ‘Everyone who cooks with gas burners has seen this effect,’ said Juniper. ‘If your pan boils over and blocks up half the holes, the other holes start to burn much faster and then, as the blocked holes clear, you get an oscillating effect across the burner as the flames grow and shrink until they settle back down.

‘In a jet engine combustion chamber, you get pressure waves that can interact with the flame in the same sort of way,’ he added. ‘The flame comes from the fuel injector into the combustion chamber and a pressure wave hits it and perturbs it. A little later, the flame will give out a bit more heat, which will increase NOx generation. If you’re unlucky, the heat release comes at just the wrong moment and adds to the oscillation, and that can lock into the resonances of the combustion chamber. That’s much more likely to happen in lean flames than rich ones.’

What is needed is some way of understanding how the design of fuel injectors and the geometry of the combustion chamber interact to form this sort of instability, but the complexities of the fluid flows make this extremely difficult. However, Juniper’s team has used some recent advances in applied mathematics to develop some software that could solve this problem.

The system, called FlowTool, generates simulations that work in a similar way to MRI. ‘We take a slice through the injector and the system tells you which bits of the injector are responsible for good mixing and instabilities,’ he said. ‘Then, crucially, it tells you what frequency and shape the motion will be – will the gas jet flap backwards and forwards or side to side? With that information on one injector, you can look at how it will fit into the acoustic modes – the pressure waves – in the combustion chamber.’

This, said Juniper, is a tough problem. The research team is using applied mathematics first published 10-15 years ago, which is an unusually short time for maths to be applied to practical engineering. ‘Applied mathematicians tend to lose interest once the basic problem has been solved,’ he said. ‘The processes behind this tool had been applied to a flow behind a cylinder, which is a standard problem in applied maths, but it hadn’t been applied to anything more complex than that. I don’t think anyone quite believed we would get this to work.’

Cambridge University’s engineering department has close links with Rolls-Royce and the company began to sponsor the project in 2005, two years after Juniper started working on FlowTool. ‘We had a PhD student working on the maths, which was the side we really had to nail. Then I worked with another PhD student with Rolls-Royce this year, where we sat with the company’s design team and worked out what it wanted to feed into this thing and what it would like to get out.’

The current version of FlowTool accepts the outputs from Rolls-Royce’s computational fluid dynamics (CFD) system. The output, which is generated in a few hours, is graphical – a slice through the injector highlighting the position and shape of instabilities. ‘It’s almost at the stage where you press a button and just let it run,’ said Juniper.

The FlowTool code was written in a modular form using Matlab software. Juniper’s team is currently trying to speed it up by removing modules, reprogramming them in the more streamlined C++ language and plugging them back in. ‘The speed is so important to the design process,’ he said. ‘Who wants to wait a week?’

“We went for a very ambitious project. I’m surprised at how well it works, but perhaps not as surprised as many”

The team is also developing plug-in modules, such as FlowTweak, which shows the effects of changing the flow profile. ‘You’ll have a flow profile from CFD and that gives rise to a certain instability,’ explained Juniper. ‘If you want to make that flow a bit quicker through the centre of the injector, you can drag a line with a couple of mouse clicks and it will recalculate to see what that does to the instability.’

The next phase of the project is to validate the data produced by FlowTool against systems that have already been fully examined. The flow profiles in fuel injectors for jets are very similar to those encountered in papermaking equipment, and Juniper has been testing FlowTool against the analysis from these systems. ‘A Swedish papermaking firm had done full-blown CFD using Sweden’s biggest cluster of PCs, which took several weeks,’ he said. ‘We did the same analysis on a laptop in an hour. It agreed to within five per cent.’

Juniper believes that FlowTool could offer Rolls-Royce a significant advantage in the gas turbine industry. ‘We went for a very ambitious project, applying this to something as big and complicated as a jet engine fuel injector,’ he said. ‘I’m surprised at how well it works, but perhaps not as surprised as many.’

Runners up

Environmental technology
The other shortlisted candidates in this category were:

Warwick University, Sorption Energy
This project aims to commercialise research into low-energy cooling systems developed by Prof Robert Critoph of Warwick University. These use activated carbon beds that adsorb and desorb ammonia, replacing the electrically driven compressors that usually power refrigeration and air-conditioning systems. The partners believe that there is a significant market for such systems in India and China

UPS Systems, Glamorgan University
The Renewable Hydrogen Research and Development Centre at Glamorgan University is the first facility in the UK to use renewable energy to generate hydrogen — a key technique for overcoming the intermittency of wind turbines and solar photovoltaics. UPS provided the fuel cells for the centre, which has now formed a model for other facilities in the UK.