Séminaire LJKModèles et Algorithmes Déterministes: EDPMOISEMGMI

Le Jeudi 8 Mars 2018 à 11h00 en Salle 106  Batiment IMAG

Séminaire de Mr Julian PANETTA (Ecole Polytechnique Fédérale de Lausanne (EPFL))

Robust Elastic Metamaterial Design for Additive Fabrication

Résumé:

Modern additive fabrication technologies like 3D printing have made manufacturing specialpurpose, customized objects easy, quick, and affordable. Their widespread availability has spurred the development of computational tools for designing objects to meet specific performance goalsfor instance, designing an object that performs a particular deformation when subject to forces. However, these design goals typically require controlling the fabrication material's elastic properties more precisely than is possible with even the most advanced multimaterial 3D printers. Furthermore, many goals require material behaviors outside the range of properties offered by ordinary fabrication materials (e.g., negative Poisson's ratios).
In this talk, I will present my work on designing elastic metamaterials for 3D printing: periodic microstructures that are tuned to emulate a large space of elastic materials. This approach to tailoring material properties is a perfect fit for additive fabrication, which can produce parts of arbitrary complexity at a cost proportional to only the material consumed. I will also present a design tool to optimally apply these metamaterials to achieve highlevel deformation goals.
Since microstructures typically consist of thin features that concentrate stress, they are prone to plastic deformation and fracture even under mild deformations of the macroscale object. A key goal of my work is to design microstructures that minimize such stress concentrations, improving the metamaterials' robustness in practice. The approach I propose is based on a novel, efficient, exact solution to the worstcase stress analysis problem for periodic microstructures which supports several failure criteria (e.g., maximum principal stress or von Mises stress). I introduce an algorithm for designing microstructures that minimize a worstcase stress measure while achieving a particular target elasticity tensor and satisfying fabrication constraints (minimum thickness, overhang). Structures designed with this algorithm achieve a typical worstcase stress reduction factor of 5x, which translates into practical robustness improvements demonstrated in lab tests of 3D printed samples.
