Algorithmic Design-for-Manufacturing and Programmability of Metamaterials

Abstract

Modern digital design tools offer vast creative freedom, yet a persistent gap remains between conceptual designs and their physical realization, particularly for advanced materials and designs. This challenge is acute for metamaterials, whose novel properties are derived from complex, fine-scale architecture that is often difficult to design for and manufacture. This dissertation argues that design tools should directly translate a user's high-level intent into physical form by managing complexity on two fronts: abstracting complex material behavior into simple, programmable controls, and embedding the physics of the fabrication process directly into the design environment. This approach treats physical realization not as a downstream constraint but as an integral design parameter, enabling more expressive, accessible, and efficient workflows. To substantiate this claim, this thesis presents three interlocking contributions that integrate computational design with advanced manufacturing: First, to address the design complexity of metamaterials, I introduce a framework that abstracts complex mechanical behaviors into simple, programmable primitives. This work uses compliant straight-line mechanisms (SLMs) as reconfigurable building blocks that explicitly encode zero-energy deformation modes. By coordinating these SLMs with planar symmetries, I demonstrate the ability to deterministically program and smoothly interpolate between all 2D extremal classes (nullmode, unimode, bimode, and trimode), enabling in-situ, reversible tuning of emergent properties like Poisson's ratio and chirality without costly re-computation. Second, I present Fabrication-Directed Entanglement (FDE), which combines the previous work on simplifying metamaterial design with topology optimization and viscous thread printing (VTP) to computationally design and fabricate monolithic foam metamaterials with spatially patterned entanglement. By translating intended directions of compliance and rigidity into optimized density fields and controlled VTP coiling patterns, FDE produces single-filament structures exhibiting targeted anisotropy, extreme Poisson's ratios, and significant chirality-driven normal-shear coupling---behaviors previously inaccessible in uniform or multi-material entangled foams. Third, to demonstrate the generality of this fabrication-aware approach beyond mechanics, I extend the VTP-based design methodology to the optical domain. I present a computational pipeline for creating foam-based lithophanes where light transmission is controlled by spatially varying the foam's porosity. This work leverages a calibrated physical simulation of the VTP process and a photorealistic rendering pipeline to automate the translation of digital images into manufacturable, uniform-thickness structures with programmed optical properties. Collectively, these projects demonstrate that abstracting low-level fabrication and material physics into process-aware computational tools expands the space of what can be designed and built. By algorithmically integrating the means of production into the act of creation, this work provides a pathway toward more powerful and intuitive design systems for complex physical objects.