Creep prediction to extend jet engine life

Materials World magazine
,
1 May 2009
Jet engine – image courtesy of Rolls-Royce

UK scientists have developed a computer model that could help improve the performance of jet engines by making more accurate predictions of creep within turbine blades.

High temperature and huge loads on the blades can pull them out of shape. The program from Imperial College London could assist design engineers to extend the life of jet engines. Materials scientists could also use the software to develop better alloys that run at higher temperatures for improved fuel efficiency.

Nickel superalloys have good creep resistance at up to 80% of their melting point, making them the material of choice for turbine blades. Comparable materials begin to creep at around 40% of their melting point.

Dr David Dye, who led the project at Imperial, describes current methods of creep prediction as ‘curve fitting’, which work by extrapolating from test results. He says that his team’s modelling methods promise to be more accurate because they can account for variations in microstructure
in the alloys, in particular the evolution of dislocation distribution. The models are based on knowledge from over 40 years of creep research, including many microscopic studies.

‘We’ve worked out the physics of what’s going on, written that down mathematically and implemented it in code,’ he says.

These physical rules are then embedded into finite element analysis (FEA) software used by gas turbine manufacturers.

This creates a more complex model of creep development. It can account for the two main ways in which creep occurs – at high temperature and low stress (in which case the main nickel matrix deforms), or at high stress and low temperature (when the secondary aluminium phase is sheared).

Accurate creep prediction will give engineers more confidence in ‘lifing’ an engine, as jet engines must be taken out of service well before they fail, so a suitable safety margin is built in. ‘It would allow you to run closer to the true operating limits,’ says Dye.

Jet engine manufacturer Rolls-Royce partners the project, enabling the researchers at Imperial to validate their models against actual behaviour observed in service.

Robbie Hobbs, who manages university research at the company, explains that the high temperature and load on the blades are ‘the equivalent of hanging a double decker bus off the blade. The blades need to resist the urge to stretch, otherwise they can impact on the casing’.

Because the model gives a better understanding of microstructure, materials scientists could also use it to develop alloys with higher creep resistance. Aerospace companies spend vast amounts of money to enable a 20ºC improvement per year, so this predictive software could help them to accelerate the process.

‘A 20ºC increase in turbine entry temperature is about a 0.5% improvement in fuel efficiency,’ explains Dye. ‘That might save £500,000 over a 15-year lifetime.’

Duncan McLachlan of Rolls-Royce’s Lifing Methods Group adds, ‘Anything that gives more accurate estimates of material behaviour is always of interest. This model, [however] probably has a way to go before it can be commercialised. It would need more validation’.