‘Short circuiting’ alloy design
Developing TRIP steel, stainless steel and other advanced alloys may become faster using a novel genetic algorithm-based thermokinetic computer model.
Scientists at Delft University of Technology in The Netherlands are using their software to
produce high strength maraging stainless steels of about 1.5-2GPa for use in aircraft landing gear, incorporating composition, as well as austenitisation and ageing temperature and process optimisation parameters.
‘We have tried to mimic nature,’ explains Dr Pedro Rivera at Delft. ‘In nature, you have a series of features in animals that are evolving and these, when interacting with the environment, either become stronger, reproduce or die. We use the same principle. The features we look at are expressed in terms of compositions and heat treatment temperatures. We use the numerical framework of the genetic algorithm to produce structures that are defined by the rules of physical metallurgy.’
He adds, ‘In one run we can consider trillions of samples from different compositions and heat treatments, which may require several iterations with conventional alloy development methods.’
At the nanoscale
The maraging stainless steels created precipitate particles in the range of 5-30nm, incorporating strengthening elements such as copper, nickel, titanium and vanadium. The exact compositions are confidential, but the ‘key engineering focus has been on strengthening potential, sufficient ductility to shape the material and achieving toughness so that the material is resistant to cracking in service’, says Rivera. The toughness and corrosion resistance of the Delft materials are still being tested.
Commercial aircraft landing gear is currently made from cadmium-coated steel for corrosion resistance. High strength stainless steels could edge into the market, as cadmium is to be banned for environmental reasons.
Combining the genetic algorithm with thermodynamic calculations allows the researchers to minimise the production of unwanted precipitates during heat treatment.
Rivera explains, ‘The austenitisation criteria are very important, as, during solidification [of the material], intermetallics and undesirable phases such as niobium carbon nitrides form quickly at between 1,100-1,250ºC. They grow so big that they inhibit and sometimes preclude the subsequent formation of the low temperature, nanometre scale strengthening precipitates.
‘We won’t be able to get rid of the undesirable phases, but we can minimise them so that the properties are not drastically affected.’
Formation of the nanoprecipitates during ageing at 450-575ºC is also optimised. ‘The right time has to be chosen to achieve these fine precipitates without them growing too large to provide adequate strengthening,’ says Rivera.
The researchers are working closely with the Tata/Corus Group, who specified compositional ranges that were suitable for bulk manufacture, and then produced the resulting sample alloys stipulated by the computer models.
Dr Peter Morris of Corus RD&T in Rotherham, UK, sees potential for applying the software to a range of materials. He says, ‘One of the problems with materials development is that you have to explore a series of compositions before you get to the optimum solution. That’s a time-consuming and inefficient process. The researchers at Delft have developed software that potentially allows us to short circuit that process’.
He insists, however, that the physical rules input to the program must be correct and reliable. ‘It does not remove the need for a detailed understanding of metallurgy.’
Scientists at Delft will continue characterising their stainless steels, as well as exploring how their computer modelling technique can redesign other alloys.