Atomistically-informed continuum interface models for functional composites

点击翻译按钮获取中文摘要

Overall energy density of nanoparticle-based functional composites can be enhanced by tailoring their constituent dielectric and mechanical properties. This project will develop predictive approaches for the enhancement of energy density in functional composites by combining a new data-driven computing paradigm based on Gaussian processes (GPs). The GPs will be employed to determine surrogate continuum material laws from coarse-grained molecular models, describing the behaviour of composite constituents. Those molecularly-informed continuum laws will be used in the microscopic continuum modelling of the composite with the RVE concept, and then propagated onto the macroscale via computational homogenisation to predict the overall response of the composite with error bars.