| Abstract Scope |
Key sensors for advanced and everyday technology, such as smartphones and drones, rely on inertial micro-electro-mechanical systems (MEMS). In some extreme applications they cannot yet be used, in others they cannot be substituted upon failure. Contrarily, in consumer applications they are not exploited to the end of life, in a wasteful planned obsolescence cycle. We propose a multiscale modelling framework which, once developed, will be rooted in atomistic simulations, which are cleverly extrapolated to the microscale and interpreted using Machine Learning, all to predict material degradation under a wide range of operating conditions. The result will be a new class of smart constitutive laws that will seamlessly integrate in continuum models to quantify the degradation of MEMS during prolonged use or extreme loads. A key advantage is that the effect of defects, microstructure and loads becomes transparent and mechanistic predictions become possible. |