Bridging knowledge across architectural heritage and digital fabrication technologies

Abstract

This thesis investigates the integration of computational making approaches within architectural heritage studies by exploring workflows for customized digital fabrication toolpath generation, with a focus on case studies involving carved stone ornaments from selected monumental buildings in medieval Anatolia. Medieval Anatolian architectural ornaments reflect a rich design culture that merges craftsmanship with formal experimentation. On-site traditional construction, performed by skilled artisans, involves more than just executing predefined designs; it transforms abstract design ideas into a cohesive making process. The geometric decorative patterns have been adapted across various materials, such as wood, stone, brick, tile mosaic, and stucco, over an extensive geographical range. Despite the material variety, the diverse craft tools and methods used to create these intricate patterns remain under-examined, with most existing literature emphasizing their abstract geometric shapes rather than construction techniques. The literature on computational design is increasingly highlighting the importance of integrating materiality and craft knowledge into computational models, considering the act of making as a fundamental component of design ideation. While grammar-based techniques have been introduced and diversified methodologies, their application within heritage contexts is still limited and not fully developed. In the past two decades, digital fabrication has emerged as a new medium in cultural heritage, prompting new discussions on materiality. The focus of research is shifting from acquiring high-precision morphological data to emphasizing the accuracy and authenticity of materials and construction techniques, thus opening new avenues for exploration. Existing studies have predominantly utilized standardized prototyping methods designed for industrial mass production, which may not adequately capture the nuances of traditional craftsmanship. Consequently, further research is necessary to develop tailored fabrication technologies that offer deeper insights into historical construction techniques. While much experimental work on customized robotic stone-working has concentrated on technological challenges, integrating these advancements with the historical contexts and stylistic variations of stoneworking cultures holds significant potential. Bridging the gap between technology-driven and knowledge-driven research is crucial for effectively incorporating digital fabrication technologies into architectural heritage studies. Although accelerating processes and improving accuracy through automation are the primary aims of digitizing workflows, digital fabrication technologies also introduce systematization in architectural construction. Innovations in digital fabrication tools and software have enabled integrated data flows in architectural construction by allowing parametric fabrication toolpaths associated with geometric models and facilitating customized automation techniques. This thesis frames the emerging role of digital fabrication in architectural heritage research as formalizing construction knowledge and thus generating and conveying new layers of information about material forms and construction processes. The medieval stone-carved ornamentation styles of Anatolia are unique to the region yet remain inadequately documented. In the 13th century, Anatolia witnessed the emergence of a new aesthetic in stone ornamentation, distinguished by its plasticity and curvilinear forms. This distinctive style emerged from a fusion of the region's abundant stone resources and skilled masons with geometric pattern design traditions. It serves as an exemplary illustration of how materials and craftsmanship can drive the evolution of a specific visual style. Computational making approaches offer promising avenues to enhance our understanding of these understudied generative processes in crafts. In the two workflows developed in this thesis, formalizing the relations between part-whole relationships and subtractive actions enabled the identification of variations in stone carvings. The first case study employs a grammar-based approach with a focus on rule formalizations. The workflow involves transferring information from the making rules to parametric digital models. The outputs include digital representations of the carved shapes, newly generated boundary shapes, G-code for CNC milling, and simulations of the cutting tool movements. Three parametric models corresponding to the three types of rules were developed as custom user objects in the Rhinoceros-Grasshopper modeling environment. Consequently, making rule formalizations in the developed workflow are not merely representational but also actively inform the digital fabrication process. This case study proposes a novel application of making grammars in architectural heritage research, offering several key insights. First, making grammars provide a framework to analyze formal relationships among abstract shapes, material forms, and construction parameters. Formalizing making rules enables the integration of visual and spatial computation into the study of implicit formal relations between geometric compositions, tools, and crafting techniques in historical artisan traditions. In the context of specific medieval Anatolian stone carvings, rule formalizations made it possible to examine the interrelations among geometric patterns, cutting tool profiles, and cut depths. Additionally, combining making rules with shape rules introduces a broader range of outcomes within craft processes. Breaking down the construction sequence into computational steps also helps to distinguish historical stylistic variations, which partly stem from similar carving techniques. Ultimately, making grammars enable the formalization of processes through a multimodal language, recognizing the various layers of knowledge embedded in making. In the second case study, a holistic workflow is explored to facilitate a continuous data flow from photogrammetric survey data of historical stone ornaments to digital and physical parametric reconstructions. This approach involves developing two parametric models for robotic milling toolpath generation. The engraved column sample from the Hunat Hatun Complex serves as an example for modeling carved forms, drawing from the two-dimensional geometric patterns typical of the Anatolian Seljuk period. The spiral-fluted columns from the Karatay Madrasa, Sahabiye Madrasa, and Sultan Han illustrate the variations of three-dimensional forms that can emerge from similar design layouts when using different cutting orientations. Using a custom algorithm to calculate robotic milling toolpaths within a parametric modeling environment enables precise and integrated data flows to modeling historical structures. This approach makes it possible to uncover historical construction parameters embedded in ornamental variations from a specific era, something standard CAM methods cannot achieve. Adding a rotary mechanism as an external axis provided new insights into the formation of various spiral-fluted columns in the case study. Experiments adjusting cutting orientations by changing the angles between the rotary axis and the cutting tool on the cylindrical surface demonstrated that form-making evolves throughout the construction phase of medieval stone ornamentation rather than adhering to a pre-set design. The developed workflow uses KUKA|prc parametric robot control to generate and simulate robot toolpaths, enabling the exploration of multiple parameters, such as cutting orientation, geometric ratios, and tool attributes that affect material outcomes. Developing customized robotic fabrication not only achieves rapid, precise, and efficient fabrication but also transcends visual resemblance, conveying aspects of materiality, tactile qualities, and material behavior in historical structures. By programming robotic movements within widely adopted parametric software, the workflow integrates established optimization and accuracy-testing algorithms, enabling the validation of geometry and fabrication parameters based on actual material results. This system exemplifies how parametric robotic fabrication can systematize the modeling of historical stone ornamentation, providing measurable and comparable data on parameters and material outcomes.