In efforts to reduce the cost and time demands of testing and optimising advanced composite materials, researchers at the NYU Tandon School of Engineering have devised a machine learning system which employs artificial neural networks (ANN) capable of extrapolating from data obtained on the basis of a single sample.
This allows the new system to quickly formulate and provide analytics on theoretical composites enhanced with graphene, which could eventually allow manufacturers to arrive at the best material configuration without the usual hassle and extravagant cost.
In a paper to be featured on the inside cover of the journal Advanced Theory and Simulations, lead researchers Nikhil Gupta and PhD student Xianbo Xu at NYU Tandon detail a method of bypassing the need for tensile tests and dynamic mechanical analysis (DMA), both of which are widely used to characterise the viscoelastic properties of materials at different loading rates and temperatures.
First, the ANN-based approach builds a model and then feeds it data from DMA to generate a forecast of how the relevant composite will react to temperature and pressure conditions other than those present under DMA, or, in the words of Gupta himself:
“Testing materials under different conditions during the product development cycle is a major cost for manufacturers who are trying to create composites for numerous applications. This system allows us to conduct one test and then predict the properties under other conditions. It therefore considerably reduces the amount of experimentation needed.”
According to Gupta, the capacity to predict the properties of composite materials using ANNs could eventually lead to system whereby said capacity is deployed to guide the development of materials themselves, as well as the applications thereof.
Commenting on the new system, Head of RD at a New York-based production facility of GrapheneCa – a 2D graphene materials manufacturer involved in the study – Dr Sergey Voskresensky said the team has managed to develop “a new method for predicting the behavior of thermosetting nano-composites over a wide range of temperature and loading rates”.
Furthermore, the same approach might also be applicable to predicting the behavior of thermoplastic materials which, according to Voskresensky, “is a critical step towards advanced composite production”.
Comment this news or article