The quest for ultralightweight materials: from membranes to architected lattices

Federico Bosi (UCL)

Click here to joinLink opens in a new window

Abstract:

From nature to structural applications, the search for ultralightweight and strong materials drives the progress toward highly efficient systems. In this framework, membranes and cellular solids represent two classes of materials that offer unmatched strength-to-weight ratios. The former enables the realisation of reconfigurable systems that yield fast and large shape changes with minimal energy consumption, while the latter provides the opportunity to tailor the mechanical properties by manipulation of the architecture at micro and nanoscales. However, the homogeneity and high viscosity of membranes prevent accurate modelling and reliable control strategies, while the heterogeneity of architected materials is often characterised by standard unit cells with a suboptimal distribution of solid material.

In this talk, we discuss advances in the modelling, design and characterisation of ultralightweight solids to enhance their mechanical properties and promote their use in several engineering applications, from sustainable buildings to robotic components. In the first part, we present a new imaging technique to precisely characterise the yield strength of membrane materials, thus overcoming the limitations of the available experimental methods. The proposed approach identifies the material yield domain through the abrupt change in strain distribution that develops at the onset of plasticity when soft membranes of different shapes are inflated. The method is based on a finite strain solution for the bulging of circular and elliptical elastoplastic thin films that is analytically obtained, numerically validated and experimentally verified through digital image correlation technique. In the second part of the talk, we demonstrate how the mechanical properties of lightweight bending-dominated lattices can be enhanced by altering their architecture through the redistribution of solid material along the cell walls. Multi-objective optimisation of the strut topology is performed to maximise stiffness, yield and buckling strength. Analytical calculations are employed to optimise the properties when a reduced number of design parameters is considered, while finite element simulations are interfaced with a Bayesian machine learning framework to explore an enlarged design space.