Model behaviour of ceramics

Materials World magazine
,
1 Jan 2009
Computer modelling atom by atom

Multi-scale mathematical and computer modelling of the sintering of ceramics could help ensure the integrity and quality of components during manufacture, saving money and reducing waste.

A team led by researchers at the University of Leicester, UK, claim to have discovered a way to eliminate the existing ‘trial and error’ approach to sintering ceramic parts. ‘The reason why industry has not been using such technology before was because the data needed are expensive and difficult to get,’ says Lead Investigator, Professor Jingzhe Pan of the Department of Engineering at Leicester. ‘People thought you needed the full constitutive law of a material. But we have demonstrated that this is not the case. For pressureless sintering, which covers a wide range of ceramic products, all you need are
densification data.’

A mathematical framework has been developed using measurements of the densification of different ceramics during firing. This is fed into specially developed computer simulation software to produce models that predict shape change (shrinkage) and density evolution before manufacturing even begins.

Pan explains, ‘Shape and dimensions need to be as precise as possible for an engineering component as post-processing can be difficult. Another issue is after firing you might get defects as the part does not necessarily shrink uniformly. The mechanical strength depends on the microstructure’.

The research has also given unanticipated scientific insight, says Pan. He claims that American sintering pioneer David Kingery’s textbook theory on pore stability has been proved ‘wrong’. Kingery predicted that a critical coordination number exists, above which a pore does not contract but expands during sintering. However, computer simulations of a large pore embedded in a dense polycrystalline solid have revealed the pore does continue shrinking.

‘We have done a lot of experimental studies and found it difficult to reconcile pore stability condition with our observations. So it came as a great relief to see somebody challenging the criterion,’ says Professor Ian Nettleship of the University of Pittsburgh, USA.

Mike Thomas, Technical Director at manufacturer Morgan Technical Ceramics’ UK division, believes there could be benefits to materials modelling if it can fulfil its alleged potential. He says, it ‘would allow modelling to be used as an accurate predictive [instrument] to enable tooling to
be right first time and to eliminate errors and scrap during the early stages of production ramps.

‘Ceramic manufacturing could benefit, rather than relying on historical data and know-how of designing products’.

He adds, ‘I’d like to see how well the model can work on a range of ceramics, as we know empirically the differences going from one material type to another’.

Pan believes the technique could be applied across a range of ceramics, noting that experiments on both high purity alumina and low purity clays have proved the accuracy of the models.

But both Thomas and Pan agree that challenges remain in converting the technology into a user-friendly system for industry uptake. There is now a ‘gap to take this forward’, says Pan. ‘Investment from industry can [help].’