Hybrid Discrete Vault Structure
An Integrated Multi-Modal and Multi-Scalar Design, Production, and Assembly Workflow
OVERVIEW | OBJECTIVES
Hybrid Discrete Vault investigates how computational design, non-planar additive manufacturing, and subtractive fabrication can be coordinated to produce a discrete, multi-material architectural assembly. Developed through Mark Segovia’s Master of Science in Architecture thesis at Texas Tech University, the project brings together large-format polymer printing, ceramic 3D printing, CNC-milled timber, and mechanical connections within a shared design-to-production workflow.
The research connects the overall geometry of a vault with the fabrication and assembly requirements of its individual components. Form-finding, subdivision, print orientation, toolpath development, material behavior, and connection detailing inform one another throughout the process. The work develops across three related investigations: non-planar clay vaults, a compression-informed tripartite LFAM vault, and a hybrid assembly integrating printed ceramics with a polymer substrate and timber registration system.
The resulting full-scale demonstrator examines how components produced through different manufacturing processes can be brought into alignment. Its contribution lies in coordinating geometric relationships, material tolerances, and assembly sequences across the complete system.
Initial experiments investigate how curved toolpaths and slope-responsive deposition can produce vaulted clay components. Variations in printing speed and extrusion conditions are studied in relation to local slope, material accumulation, and print continuity. Wind-informed geometric studies connect these fabrication experiments to broader questions of environmental response.
Non-Planar Clay Printing
RESEARCH | DEVELOPMENT
The research extends to a tripartite vault derived through thrust-network form-finding. Its geometry is divided into twelve primary shell components, with additional base and interface elements, for non-planar large-format additive manufacturing. Component boundaries, production orientation, and toolpaths are developed together to translate the overall form into printable and assemblable parts. The fabricated LFAM vault demonstrates assembly feasibility and self-support under its own weight.
Large-Format Additive Manufacturing
Hybrid Material Assembly
The final system combines glass-fiber-reinforced PETG components, printed Cinco Rojo ceramics, CNC-milled timber, and mechanical hardware. Timber members register and align adjacent components, while individual ceramic attachments accommodate installation within the larger assembly. Connections are designed to support disassembly and component replacement.
Physical evaluation examines print fidelity, dimensional tolerance, and assembly fit. Ceramic studies additionally investigate water absorption and preliminary evaporative-cooling behavior, extending the fabrication workflow toward questions of environmental performance.
The project develops through iterative computational modeling, material prototyping, and physical assembly. Decisions are coordinated across three scales: the overall vault geometry, the organization and interfaces of individual components, and the local conditions of material deposition.
Design, Production, and Assembly
METHODS | WORKFLOW
Thrust-network form-finding establishes the compression-informed tripartite geometry. Parametric modeling translates this surface into discrete components, coordinating subdivision boundaries with fabrication limits, production orientation, and assembly interfaces.
Form-Finding and Discretization
Custom non-planar toolpaths connect surface geometry with deposition behavior. Clay and large-format polymer printing are developed through material-specific trials, examining how printing speed, extrusion settings, and local slope affect bead formation and geometric fidelity. Physical prototypes inform revisions to component geometry and production parameters.
Toolpath Development and Material Prototyping
Printed polymer components, ceramic elements, CNC-milled timber, and mechanical hardware are developed as a coordinated kit of parts. Timber members provide registration and alignment, while connection details address fabrication tolerances and installation access. The assembly sequence informs component interfaces, with removable connections designed to allow disassembly and individual component replacement.
Component Coordination and Assembly
Fabricated components and assemblies are assessed for print fidelity, dimensional fit, and assembly feasibility. Earlier clay-vault studies incorporate computational fluid dynamics and wind-chamber testing, while ceramic material studies examine water absorption and preliminary evaporative-cooling behavior. These observations inform subsequent design and fabrication iterations.
Physical Evaluation and Feedback
PUBLICATION | RECOGNITION
Publications:
Publication 01
Non-Planar 3D Printing of Vaulted Structures: Using Variable Feed and Extrusion Rates for Wind-Informed Climate Resilience: Segovia, M. A., Rodriguez, D., Hunt, E., & Mostafavi, S. (2025). Non-planar 3D printing of vaulted structures: Using variable feed and extrusion rates for wind-informed climate resilience. In Computing for Resilience: Proceedings of the 45th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA 2025) (Vol. 2, pp. 86–101). ACADIA. https://doi.org/10.52842/conf.acadia.2025.2.086
Integrated Design-to-Production of Tripartite Hybrid Vault Structures Using Non-Planar Large-Format Additive Manufacturing: Forthcoming paper, IASS–IWSS 2026.
Non-Planar Large-Format Additive Manufacturing of Hybrid Vault Systems: Robotic Fabrication and Ceramic 3D Printing: Forthcoming special-session contribution, essay and poster, ROB|ARCH 2026.
Hybrid Vault: A Multi-Scale Fabrication- and Assembly-Aware Workflow for Discrete Construction Integrating Subtractive Methods, Ceramic 3D Printing, and Large-Format Additive Manufacturing: Forthcoming peer-reviewed paper, ACADIA 2026.
AWARD | RECOGNITION
Coming Soon…
INFORMATION | CREDITS
Principal Investigators:
Mark Segovia
Texas Tech University Hi-DARS Team:
TTU HCoA Additional Support:
External Community Partners:
