Computational Design of Active 3D-Printed Multi-State Structures for Shape Morphing

TL;DR

This paper presents a computational framework for designing multi-state 3D-printed active structures capable of shape morphing with a single input, combining topology optimization and multi-material printing to enable versatile, programmable deformation modes.

Contribution

It introduces a novel topology optimization method based on power diagrams and a multi-material dithering scheme for creating multi-state active structures without additional training steps.

Findings

Structures achieved desired deformation modes in simulations and experiments.

The framework is applicable to diverse fields like aerospace, robotics, and computer graphics.

Experimental results show promising shape morphing capabilities despite fabrication limitations.

Abstract

Active structures have the ability to change their shape, properties, and functionality as a response to changing operational conditions, which makes them more versatile than their static counterparts. However, most active structures currently lack the capability to achieve multiple, different target states with a single input actuation or require a tedious material programming step. Furthermore, the systematic design and fabrication of active structures is still a challenge as many structures are designed by hand in a trial and error process and thus are limited by engineers' knowledge and experience. In this work, a computational design and fabrication framework is proposed to generate structures with multiple target states for one input actuation that don't require a separate training step. A material dithering scheme based on multi-material 3D printing is combined with locally applied copper coil heating elements and sequential heating patterns to control the thermo-mechanical properties of the structures and switch between the different deformation modes. A novel topology optimization approach based on power diagrams is used to encode the different target states in the structure while ensuring the fabricability of the structures and the compatibility with the drop-in heating elements. The versatility of the proposed framework is demonstrated for four different example structures from engineering and computer graphics. The numerical and experimental results show that the optimization framework can produce structures that show the desired motion, but experimental accuracy is limited by current fabrication methods. The generality of the proposed method makes it suitable for the development of structures for applications in many different fields from aerospace to robotics to animated fabrication in computer graphics.