Programmable Textiles: From Modeling, Structure, and Inverse Design

Abstract

Textiles are an essential and versatile engineering material. They insulate and protect us from our environment, protect astronauts in pressurized spacesuits, support large scale tensile architecture, reduce aerodynamic drag in high-speed sports, and transmit assistive forces in wearable robotic systems. In all of these applications, performance is fundamentally governed by the underlying textile properties. Yet, despite many modern advances in manufacturing technologies such as additive manufacturing and CNC machining, textile development remains constrained by the complex heirarchy and interaction of fibers, yarns and architecture. Inspired by mechanical metamaterials, this dissertation develops a new class of programmable textiles. The goal is to establish an accessible framework that couples structural design with finite element (FE) modeling to enable inverse deign of textile behavior and properties. The first part of this dissertation presents a homogenized FE modeling framework for textiles. This approach captures both in-plane mechanical properties and out-of-plane bending behavior. Material parameters are identified using gradient-free optimization, combining uniaxial and cantilever experiments to calibrate the model. The framework is validated by comparison and between analytical predictions and experimental measurements, demonstrating that the framework can provide accurate prediction of textile response including wrinkling behavior. The second part introduces a fabrication strategy for structured textile metamaterials based on composite layering. A compliant knit is combined with a patterned woven layer to form a composite textile system. By varying the woven architecture, the textile response can be tuned, including the realization of auxetic designs with negative Poisson's ratio. These structures enable controlled deformation modes, including a stretch-induced dimpling mechanism that persists under form-fitting boundary conditions. This functionality allows direct control of surface roughness. As a demonstration we show that the tuning of aerodynamics response of a fabric can be achieved through structural design. The final part of this dissertation integrates modeling and structural design into an inverse design framework for metamaterials. This framework maps prescribed mechanical responses to realizable textile architectures through parameterized unit-cell design and optimization. Using homogenized FE models and gradient-free optimization, the approach enables systematic exploration of the design space and the reproduction of a wide range of textile behaviors. Together, these contributions establish a framework for programmable textiles in which mechanical behavior is specified through structure. This enables new deformation modes, such as auxetic dimpling and shear-controlled creasing, and supports new functional applications including aerodynamic control and wearable actuation.