Programmable Mud: 3D Printing earth to achieve low-carbon, low-cost construction automation

Abstract

Large-scale additive manufacturing (LSAM) with locally sourced materials, such as earth, presents a promising approach to addressing the urgent challenges of rapid urbanization and construction-related carbon emissions. This dissertation establishes a comprehensive framework for integrating low-carbon materials, particularly minimally processed earth, with computational design methodologies and robotic fabrication processes for architectural-scale applications. Through systematic material characterization, novel testing protocols, and case studies across multiple building systems, the research demonstrates that minimally processed earthen materials can be transformed into high-performance building elements uniquely suited to local environmental conditions and design considerations. The developed computational framework employs multi-objective optimization and material-aware toolpath generation to balance structural performance, thermal comfort, embodied carbon, and construction time. Four case studies validate this approach: (1) toolpath optimization for shell structures, (2) a hybrid floor system combining shape-optimized concrete beams with 3D-printed ceramic blocks, (3) zero-waste earthen formwork for reinforced concrete, and (4) thermally optimized wall systems for passive climate control. Life cycle assessment reveals that 3D-printed earth structures have approximately one-fifth the embodied carbon of conventional concrete and one-fiftieth that of industry-standard 3D-printed mortar. This research bridges the gap between additive computational design and material circularity, offering scalable approaches to sustainable construction that can be implemented across diverse environmental and economic contexts.