Search

erik.laurini@dia.units.itThe integration of modelling and simulation techniques to support material selection processes (MSPs) is one of the most compelling needs in advancing material industry and manufactory, due to the necessity of effective/efficient design and production of sophisticated materials, components and systems with advanced/extreme performance on a competitive time scale. In this arena, and specifically for complex structural materials such as polymer-based nanocomposites (PNCs) there is a strong industrial demand for chemistry/physics-based models and modelling workflows able to predict relevant materials properties (aka Key Performance Indicators or KPIs) in an accurate and reliable way and prior to any experimental set-up. With the aim of filling the gap between business processes and materials science/engineering workflows, this work reports the application – within the framework of the EU H2020 project COMPOSELECTOR – of a multi-disciplinary, multi-model approach for the accurate, reliable, efficient and cost-effective industry-driven KPIs determination for PNC materials. Specifically, three examples of questions and answers pertaining to aerospace and automotive applications are presented and discussed.

The research progress in the field of information technology shows a trend towards the gradual miniaturisation of devices, aiming at optimizing the exploitation of materials, the dimensions, energy consumption as well as user portability. The resulting systematic decrease of dimensions down to the nanometric scale implies the need to consider materials’ behaviour pertaining to quantomechanics theory, generally considered in the development of quantum computers but, in the herein considered case, introducing the necessity to consider the fundamental operating principles of factual technological applications. MEMS and NEMS (Micro/Nano Electro-Mechanical System) devices belong to this set of systems which, with their dimensions in different applications progressively scaled down, brought to light a significant contribution of friction force to the energetic performance drop. Indeed, at the nanoscale the dynamic position of each atom affects the slipping behaviour due to the formation and breakage of weak bonds between the surfaces in relative motion, hence inducing a complex elasto-plastic physical behaviour. The friction force is, therefore, not only related to normal load and other influential parameters of the tribological pair, but also to the crystalline structure of the materials in contact and their surface chemistry and roughness. In this frame, the present study is mainly focused on the in silico investigation of the friction force through molecular dynamics (MD) simulations. In particular, a system comprising a hemispheric silicon tip and an aluminium flat surface is studied, aiming at modelling the interactions between the materials from an atomistic perspective. The dimension of the simulated system (2.7M atoms, tip diameter: 20 nm) required the adoption of HPC capacities available at the “Bura” facility of the University of Rijeka, Croatia. The modelling procedure is thus set via four subsequent steps: at first the tip/surface system is created and a suitable force field is chosen to unveil the interactions of the involved atoms. In the second step, system equilibration to a 300 K temperature is performed by employing a Nose-Hoover thermostat. A normal load is then applied on the upper part of the tip to induce its interaction with the aluminium surface. Lastly, lateral velocity is applied to the tip to model, in combination with the normal load and while avoiding any influence from the Nose- Hoover thermostat, its sliding on the aluminium surface. Several case studies are hence considered, exploring the influence of different normal loads as well as different sliding velocities. The model is currently being validated by using experimental test data attained on an Atomic Force Microscope employing a silicon tip probe mounted on a Si3N4 microcantilever. The measurements are performed in a Lateral Force Microscopy mode on an aluminium thin film sample synthesized by using Pulsed Laser Deposition. The measurements allowed attaining data on the applied contact pressure that is used in tuning the used MD models. The iterative enhancement of the MD models based on the experimental datasets will provide means for gaining a deeper insight into the molecular interactions occurring at single asperity contacts, thus providing better predictive tools for obtaining novel nanoscale friction models.