Using modelling and simulation tools is now standard practice in the composite industry. For example, modelling is used to predict aircraft performance. But when it comes to modelling new composite materials, such as nanocomposites containing flakes of graphene, there is a gap in the market. Research has shown that graphene can enhance many properties of composites but tools to model these effects are sadly lacking.
This gap urgently needs plugging, according to Gerry Boyce, managing director of Haydale Composite Solutions (HCS). HCS uses functionalised graphene to improve the electrical, thermal and structural properties of composites and currently does all its development work in the lab. “We use modelling to design composite structures, but what we really need is tools to help us reduce the huge number of experiments we carry out on a daily basis,” said Boyce. “We need modelling tools that can predict the properties of materials depending on the amount and type of functionalised graphene we use.”
This type of simulation technology requires simultaneous modelling on two different scales – the molecular modelling of the graphene platelets and how they interact with their surroundings; and the modelling of how this then affects the bulk properties of the polymer matrix in which the platelets are dispersed.
This type of multiscale modelling is complex and requires large amount of computing power but Boyce believes tools such as this would dramatically speed up the commercialisation of graphene. “We have to work with a frightening number of variables such as flake size, functionalisation, aspect ratio, dispersion, polymer matrix and more,” said Boyce. “Whenever we develop a new nanocomposite we need to test its electrical, thermal, mechanical and optical properties and we quickly can end up with hundreds of millions of tests that we need to perform. We urgently need a modelling tool that can help us design nanocomposites and reduce the number of these experiments.”
James Baker, graphene business director at the National Graphene Institute, The University of Manchester, agrees. “There are gaps in all areas of graphene modelling, but the most urgent one is in multiscale modelling tools,” said Baker. “It is these tools that will help industry de-risk the development of graphene-based materials and this will speed up the adoption of graphene into various applications.”
Academia is both a developer and a user of modelling tools for this application and there is a growing amount of activity in this field in the UK. At The University of Manchester's School of Chemical Engineering and Analytical Science there is a multiscale modelling group looking at a variety of materials including graphene nanocomposites. As well as Manchester, there are projects underway at several UK universities including Warwick University, Imperial College London, and University College London.
However, academia and industry work on different timescales, with industry often frustrated at the slow pace of academia. But if industry wants academia to speed up development of modelling technology, there is a simple solution. “There is a clear correlation between how much money you throw at a problem and how quickly you get results,” says Baker “The challenge with graphene is to get industry to engage. If we can show that we can model complex materials in different applications, reducing development costs for industry, then industry would be more likely to engage.”
At the University of Warwick's International Institute for Nanocomposites Manufacturing (IINM), Dr Lukasz Figiel is supervising a new PhD studentship on multiscale modelling for optimum processing and end-use properties of polymer-graphene nanocomposites. The ultimate objective of this project is to develop an advanced 3D non-linear, multiscale (molecular-continuum), computational model for the prediction and optimisation of the melt and quasi-solid state processing of polymer nanocomposites. This can also be further adapted to predict end-use functional properties such as mechanical, thermal/electrical, conductivity of the polymer nanocomposites.
“Interaction with industry has not been as responsive as I had expected, in terms of a better appreciation of the importance of multiscale modelling for predicting optimum formulations of the polymer nanocomposites,” says Figiel. “Perhaps industry is not sufficiently convinced about polymer-graphene nanocomposites, and companies prefer to wait for signs of the so-called killer application before putting their hands in their pockets and investing in the development of advanced modelling technology.”
He also points out that the very nature of academia means that each research group is taking its own approach tothe problem, each looking to publish novel research. “It seems that there is no common strategy in the UK to underpin the multiscale modelling of polymer nanocomposites,” says Figiel. “There is a strong need for a call in predictive multiscale modelling of polymer nanocomposites, which can bring together best people in that field in the UK to address both fundamental and industry-oriented aspects and needs.”
Innovate UK is attempting to do just that with several initiatives. It has set up a Graphene Special Interest Group (SIG); the SimBest project which collates and disseminates best practices and state-of-the-art in simulation and modelling; and its Uncertainty Quantification SIG uses a capability map to help pull the community together.
Dr Lien Ngo is a technologist at Innovate UK and supports the Graphene SIG. “I think the academics have a very good base and they’re doing excellent work, but what they might be missing is the problems that industry would like to solve,” she said. “If industry doesn't want to invest in looking more closely at those problems until there’s a killer application, then what we can do is to encourage companies to develop that killer application. It's a classic chicken-and-egg scenario: it would be quicker to develop that application if we had better modelling software, but it’s hard to convince people to invest in software until we have the killer application!”
Back in the 1970s, when carbon fibre composites were waiting for that killer application, we did not have the computer power that we have today. With high-performance computing now available to academia and industry, it is hoped that modelling and simulation tools can speed up the development of graphene nanocomposites.
David Standingford is lead technologist at the Centre for Modelling & Simulation (CFMS) in Bristol, an independent non-for-profit organisation that works with industry and academia to reduce risk in the design phase, product development costs and time-to-market. He said: “The issues in modelling carbon fibre composites including graphene are representative of the general issues in modelling materials at a range of scales. Previously, it was felt computer resources would never be able to progress the task. Some of those scales are now becoming computationally tractable."
This is good news for companies such as HCS. “We are at the cusp of the same revolution that reinforced composites started 35 years ago,” says HCS's Gerry Boyce. “It has taken 35 years for modelling tools to become standard practice in the composites industry. I realise nanocomposites are more complicated than standard reinforced composites but with today's advances in computing and the availability of enormous computing power, I sincerely hope this time it wont take 35 years for modelling to become standard practice in this emerging industry.”