01 Fakultät Architektur und Stadtplanung

Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/2

Browse

Filters

2020 - 20223

Settings

Institute: search.filters.UBSInstitut.Institut für Computerbasiertes Entwerfen und BaufertigungSubject: search.filters.subject.620

Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    ItemOpen Access
    Spatial winding : cooperative heterogeneous multi-robot system for fibrous structures
    (2020) Duque Estrada, Rebeca; Kannenberg, Fabian; Wagner, Hans Jakob; Yablonina, Maria; Menges, Achim
    This research presents a cooperative heterogeneous multi-robot fabrication system for the spatial winding of filament materials. The system is based on the cooperation of a six-axis robotic arm and a customized 2 + 2 axis CNC gantry system. Heterogeneous multi-robot cooperation allows to deploy the strategy of Spatial Winding: a new method of sequential spatial fiber arrangement, based on directly interlocking filament-filament connections, achieved through wrapping one filament around another. This strategy allows to create lightweight non-regular fibrous space frame structures. The new material system was explored through physical models and digital simulations prior to deployment with the proposed robotic fabrication process. An adaptable frame setup was developed which allows the fabrication of a variety of geometries within the same frame. By introducing a multi-step curing process that integrates with the adaptable frame, the iterative production of continuous large-scale spatial frame structures is possible. This makes the structure’s scale agnostic of robotic reach and reduces the necessary formwork to the bare minimum. Through leveraging the capacities of two cooperating machines, the system allows to counteract some of their limitations. A flexible, dynamic and collaborative fabrication system is presented as a strategy to tailor the fiber in space and expand the design possibilities of lightweight fiber structures. The artifact of the proposed fabrication process is a direct expression of the material tectonics and the robotic fabrication system.
  • Thumbnail Image
    ItemOpen Access
    Granular architectures : granular materials as "designer matter" in architecture
    (Stuttgart : Institute for Computational Design and Construction, University of Stuttgart, 2020) Dierichs, Karola; Menges, Achim (Prof.)
    The thesis investigates designed granular materials in architecture. Granular materials are defined as high numbers of particles larger than a micrometre, between which mainly short-range repulsive contact forces are acting. In a designed granular material the geometry and material of the individual particle are determined by a designer. Consequently, the overall granular material can have characteristics which are novel in comparison to non-designed granular materials. In architecture, designed granular materials are understood to have new characteristics which fulfil specific architectural performance criteria. The relevance of designed granular materials in architecture is threefold. All granular materials are both fully recyclable and reconfigurable due to the fact that the individual particles are in no way bound to each other. These first two aspects alone make any granular material, whether it is designed or not, a highly pertinent strand of architectural design research. However, designed granular materials, in addition to being recyclable and reconfigurable, bear the potential for the development of entirely novel material behaviours. In the context of architecture, designed granular materials can be considered as a form of "material systems", and more specifically as a sub-group of "aggregate systems". In the wider transdisciplinary context, designed granular materials for architecture can be considered a form of "designer matter (DM)". "Designer matter (DM)" is understood as matter which is designed in its structural characteristics at its mesoscale rather than its macro- or its microscale. The current state of research into designed granular materials is presented for both architecture and granular physics, on a conceptual as well as on a project-based level. In this context the thesis aims to establish and validate a first version of a comprehensive design system for exploring designed granular materials in architecture and for interfacing with granular physics. The research development of this thesis is presented and evaluated with respect to the practical, methodological and conceptual foundations which have been laid during this phase. The methods are introduced in terms of methodological frameworks, tools and techniques and the applied research methodology. The core part of the thesis comprises a design system with a related design system catalogue as well as two case studies. The design system is established for particle systems and for related construction systems. It formulates the basic system categories and corresponding parameters. The design system catalogue is presented in the appendix and summarizes tests which investigate individual aspects of the overall design system for particle and construction systems. Each of the two case studies explores the integration of a different set of design system categories. They were conducted both through full-scale prototypes and a related set of tests with statistical repetition. Case study 1 investigates vertical structures made from a designed granular material consisting of highly non-convex particles. Case study 2 combines two designed granular materials, one consisting of convex particles and the other of highly non-convex particles, in order to form spatial enclosures. The case studies are evaluated with respect to their practical, methodological and conceptual contributions to architectural design research. The thesis is summarized and its contributions are assessed in conclusion both with respect to the field of architecture and for the field of granular physics. Further research in the field of designed granular materials in architecture can be conducted on the practical, methodological and conceptual levels of design. On the practical level, in the area of particle systems the investigation of graded granular materials, of different mechanical properties of the particles' material or of designed granular materials consisting of particles with variable geometry is highly promising. In the area of construction systems, the development of behavioural models of robotic construction is very relevant. Another key direction is for the construction systems to become increasingly simple, while the particles are progressively designed to perform parts of the construction process by themselves. On the methodological level, the integration of "inverse" design methods is the logical next step. This can be done on the basis of the proposed design system. On the conceptual level, the framework of "designer matter (DM)" needs to be further established both as a transdisciplinary model and within the field of architecture. Only then can designed granular materials be fully discussed as one form of "designer matter (DM)" in architecture. Key to any further development of the overall research field is the integration of the two fields of architecture and granular physics.
  • Thumbnail Image
    ItemOpen Access
    Autonomous robotic additive manufacturing through distributed model‐free deep reinforcement learning in computational design environments
    (2022) Felbrich, Benjamin; Schork, Tim; Menges, Achim
    The objective of autonomous robotic additive manufacturing for construction in the architectural scale is currently being investigated in parts both within the research communities of computational design and robotic fabrication (CDRF) and deep reinforcement learning (DRL) in robotics. The presented study summarizes the relevant state of the art in both research areas and lays out how their respective accomplishments can be combined to achieve higher degrees of autonomy in robotic construction within the Architecture, Engineering and Construction (AEC) industry. A distributed control and communication infrastructure for agent training and task execution is presented, that leverages the potentials of combining tools, standards and algorithms of both fields. It is geared towards industrial CDRF applications. Using this framework, a robotic agent is trained to autonomously plan and build structures using two model-free DRL algorithms (TD3, SAC) in two case studies: robotic block stacking and sensor-adaptive 3D printing. The first case study serves to demonstrate the general applicability of computational design environments for DRL training and the comparative learning success of the utilized algorithms. Case study two highlights the benefit of our setup in terms of tool path planning, geometric state reconstruction, the incorporation of fabrication constraints and action evaluation as part of the training and execution process through parametric modeling routines. The study benefits from highly efficient geometry compression based on convolutional autoencoders (CAE) and signed distance fields (SDF), real-time physics simulation in CAD, industry-grade hardware control and distinct action complementation through geometric scripting. Most of the developed code is provided open source.