Piezoelectric truss metamaterials: data-driven design and additive manufacturing

TL;DR

This paper introduces a data-driven approach combining machine learning and 3D printing to design and fabricate customizable piezoelectric truss metamaterials with enhanced and unconventional electromechanical properties.

Contribution

It presents an integrated framework for designing and manufacturing piezoelectric metamaterials with tailored anisotropic responses using generative ML and in-gel 3D printing techniques.

Findings

Achieved over 48% improvement in specific hydrostatic piezoelectric coefficient.

Discovered designs with auxetic, unidirectional, and omnidirectional piezoelectric behaviors.

Fabricated metamaterials with higher transverse than longitudinal piezoelectric coupling coefficients.

Abstract

In the development of active animate materials, electromechanical coupling is highly attractive to realize mechanoresponsive functionality. Piezoelectricity is the most utilized electromechanical phenomenon due to the wide availability of materials that display precise and reliable coupling. However, the inherent directionality of these materials is constrained by the symmetry of their crystal structure, which limits the choice of available properties in natural piezoelectric materials. A solution to alleviate this limitation could be to leverage geometry or architecture at the mesoscale. Here, we present an integrated framework to design and 3D-print piezoelectric truss metamaterials with customizable anisotropic responses. To explore the vast design space of truss metamaterials, we employ generative machine learning to optimize the topology and geometry of truss lattices and achieve target piezoelectricity. Then, we develop an in-gel-3D printing method to fabricate polymer-ceramic piezoelectric truss metamaterial structures using a composite slurry of photo-curable resin and lead-free piezoelectric particles. The ML framework discovers designs exhibiting unconventional behaviors, including auxetic, unidirectional, and omnidirectional piezoelectricity, while the additive manufacturing technique ensures shaping freedom and precision in fabricating these metamaterials at small scales. Our results show an improvement of over 48% in the specific hydrostatic piezoelectric coefficient in optimized metamaterials over bulk lead zirconate titanate (PZT). We successfully achieved metamaterials with higher transverse piezoelectric coupling coefficient than its longitudinal coefficient, which is a phenomenon that is rare in bulk materials. Our approach enables customizable piezoelectric responses and paves the way towards the development of a new generation of electro-active animate materials.