Zero-Waste Sand Formworks for Lightweight Concrete Structures Daria Kovaleva Institute for Lightweight Structures and Conceptual Design University of Stuttgart Zero-Waste Sand Formworks for Lightweight Concrete Structures A dissertation approved by the Faculty of Architecture and Urban Planning of the University of Stuttgart for the conferral of the title of Doctor of Engineering Sciences (Dr.-Ing.) Submitted by Daria Kovaleva from Moscow Main Supervisor: Co-supervisor: Prof. Dr.-Ing. Dr.-Ing. E.h. Dr. h.c. Werner Sobek Prof. Dr. Benjamin Dillenburger Date of the oral examination: 13.12.2024 Institute for Lightweight Structures and Conceptual Design University of Stuttgart 2025 Zero-Waste Sand Formworks for Lightweight Concrete Structures Von der Architektur und Stadtplanung der Universität Stuttgart zur Erlangung der Würde eines Doktor-Ingenieurs (Dr.-Ing.) genehmigte Abhandlung Vorgelegt von Daria Kovaleva aus Moskau Hauptberichter: Mitberichter: Prof. Dr.-Ing. Dr.-Ing. E.h. Dr. h.c. Werner Sobek Prof. Dr. Benjamin Dillenburger Tag der mündlichen Prüfung: 13.12.2024 Institut für Leichtbau Entwerfen und Konstruieren Universität Stuttgart 2025 Acknowledgments This thesis was written while I was working as a researcher at the Institute for Light- weight Structures and Conceptual Design (ILEK) at the University of Stuttgart. It was an invaluable experience of professional and personal growth for me. The opportunity to participate in unique interdisciplinary projects and to work with many talented, open-minded, and inspiring people from different disciplines far beyond architecture made it an exciting, fulfilling, and unforgettable journey. Thank you all for that. First and foremost, I would like to express my deepest gratitude to Prof. Dr.-Ing. Dr.- Ing. E.h. Dr. h.c. Werner Sobek, who supervised this thesis. You have always been a source of inspiration and motivation for me. Thank you for constantly challenging and setting the bar high to strive for, as well as for your all-encompassing support and trust, which have made this journey rewarding and enriching. I am honored by Prof. Dr. Benjamin Dillenburger being my second reviewer. Your expert knowledge and feedback were very helpful during the final stages of the the- sis. I would also like to thank Prof. Achim Menges for chairmaning the examination committee. I am delighted I had the opportunity to work in such an inspiring, motivating, and creative environment as ILEK, which is driven by teamwork and encouragement. Working with many colleagues has made my time at the institute enlightening and enjoyable. You have become a big family to me. Thank you, Prof. Lucio Blandini, for your trust and support during the research development. Thanks to Oliver Gericke for introducing me to ILEK and the world of structural engineering. You are a colleague one can only dream of. Thanks to David Nigl, Carl Haufe, and Benedict Strahm for energizing coffee breaks and relaxing sips of fresh air in the shade of ILEK trees. Special thanks to Christian Assenbaum for support in the literature search, guidance in translation difficulties, and reviewing. I am grateful to all the researchers and students involved in this work, especially Chris- toph Nething and Maiia Smirnova, for their contributions in the early stages of devel- opment and to Lennon Toeche-Mittler and Justus Schwörer for their tremendous sup- port in realizing the architectural prototype. A long and fruitful interdisciplinary collaboration between ILEK and the Institute for Control Engineering of Machine Tools and Manufacturing Units (ISW) was a key factor in the project's successful development. Thanks to Frederik Wulle for nurturing this collaboration and to Maximilian Nistler for many years of productive cooperation, tire- less help, and constructive feedback while preparing this manuscript. The time spent at the Materials Testing Institute (MPA) of the University of Stuttgart and the knowledge of concrete technology gained there during the fabrication of var- ious prototypes played a decisive role in developing and scaling up the technology. I would like to thank all colleagues at MPA, especially Norbert Schulz, for providing facilities and equipment and for valuable advice, help, and understanding. This research would not have been possible without project funding provided by the German Research Foundation (DFG) in the framework of the Priority Program SPP 2187, "Advanced modularized construction made with flow production methods." Special thanks to Sika Deutschland GmbH for the generous support in concrete sup- ply during the project's duration. Last, I am grateful to my family for their unconditional love and support throughout this long and sometimes challenging journey. I want to thank my parents, Julia and Mikhail, and my grandmother, Tatiana, for their unstinting encouragement in my en- deavors. Finally, thank you, my husband, Ivan Tomovic, for sharing with me all the ups and downs along this way. I dedicate this work to our daughter Xenia, the pure joy and love of this life. Contents vii Contents List of Abbreviations xiii Abstract xv Kurzfassung xvii 1. Introduction 19 1.1. Context 19 1.1.1. Lightweight construction with concrete 20 1.1.2. Problem Statement 24 1.2. Research goals 26 1.3. Research Methodology 28 1.4. Structure of the thesis 28 2. State of the Art 31 2.1. Sustainable formwork systems for complex concrete structures 32 2.1.1. Reconfigurable formworks 32 2.1.2. Stay-in-place formworks 34 2.1.3. Recyclable formworks 38 2.2. Recyclable formworks from soil materials 40 2.2.1. Clay molds 41 2.2.2. Silt molds 42 2.2.3. Sand Molds 43 2.3. Additive manufacturing of sand molds 47 Contents viii 2.3.1. Additive manufacturing of sand molds in the foundry and composites industries 48 2.3.2. Sustainability of 3D printed sand molds and development tendencies 50 2.4. Conclusion 54 3. Zero-waste sand formworks for complex concrete structures 57 3.1. Technology concept 57 3.2. Application range 59 3.3. Technology requirements 60 3.3.1. Formwork requirements 61 3.3.2. Material Requirements 63 3.3.3. Production cycle requirements 63 3.4. Conclusion 66 4. Additive manufacturing of water-soluble sand formworks 67 4.1. Selection of the additive manufacturing process type 67 4.2. 3D Printer development 69 4.2.1. Prototype 69 4.2.2. Full-scale 3D printer 76 4.3. Design-to-production workflow and communication architecture 80 4.3.1. CAD-CAM tool 80 4.3.2. Communication architecture 85 4.4. Conclusion 86 5. Material composition and properties of 3D printed water-soluble sand formworks 88 5.1. Water-soluble sand molds within the classification of sand molds in production industry 88 Contents ix 5.2. Selection of powder and binder materials 90 5.2.1. Aggregate molding materials 90 5.2.2. Water-soluble binder materials 92 5.3. Methods 95 5.3.1. Optical microscopy 96 5.3.2. X-ray microtomography 96 5.3.3. Mechanical properties 97 5.3.4. Hygroscopicity 98 5.3.5. Water solubility 99 5.4. Material properties of water-soluble sand molds 99 5.4.1. Preparation of samples 100 5.4.2. Process-related density, microstructure and geometric accuracy 102 5.4.3. Mechanical properties of water-soluble molds 104 5.4.4. Hygroscopicity of water-soluble sand molds 105 5.4.5. Water solubility of water-soluble sand molds 106 5.5. Selection of preferrable binder type and further experiments 107 5.5.1. Influence of aggregate type on mechanical properties of sand molds 107 5.5.2. Influence of binder content on mechanical properties of sand molds 108 5.5.3. Recycling of sand-binder mixture 109 5.5.4. Mechanical properties of recycled mixtures 110 5.6. Conclusion 111 6. Results 113 6.1. Functionally graded concrete lattice structure 113 6.1.1. CAD-CAM workflow 114 6.1.2. Fabrication of formwork 115 Contents x 6.1.3. Casting and demolding 116 6.1.4. Discussion 116 6.2. Single-span lattice beam 117 6.2.1. Design of the single-span beam 118 6.2.2. Design and production of the formwork 120 6.2.3. Casting and demolding 121 6.2.4. Discussion 123 6.3. Marinaressa Coral Tree 124 6.3.1. Design 125 6.3.2. Fabrication of formwork, reinforcement and connectors 137 6.3.3. Assembly of formwork, reinforcement, and connectors 141 6.3.4. Casting and demolding the modules 141 6.3.5. Assembly of the structure in Venice 143 6.3.6. Time of production 146 6.3.7. Discussion 147 6.4. Conclusion 148 7. Outlook 151 7.1. Contributions 151 7.1.1. Closed-loop production cycle – proof of concept 151 7.1.2. Additive manufacturing of water-soluble sand molds 152 7.1.3. Computational design-to-production workflow 153 7.2. Advantages 153 7.2.1. Reduced resource consumption, emissions, and waste 153 7.2.2. Geometric freedom 155 7.3. Limitations 156 Contents xi 7.4. Future work 157 7.4.1. Automation of the production cycle 157 7.4.2. Life-cycle assessment of the production process 158 7.5. Final remarks 158 Appendices 160 A.1 Mock-up of Marinaressa Coral Tree 160 A.2 Production specification of formworks for Marinaressa Coral Tree 164 A.3 Material properties of water-soluble sand material mixtures 167 A.3.1 Mechanical properties of water-soluble sand molds 167 A.3.2 Hygroscopicity of water-soluble sand molds 169 A.3.3 Water solubility of water-soluble sand molds 169 List of figures 170 List of Tables 177 References 178 Curriculum Vitae 193 xii List of Abbreviations xiii List of Abbreviations 3DP 3-Dimensional Printing AM Additive Manufacturing BJ Binder Jetting BJ3DP Binder-Jet 3D Printing CAD Computer-Aided Design CAM Computer-Aided Manufacturing CIJ Continuous Inkjet CMC Carboxylmethyl Cellulose CNC Computerized Numerical Control CT Computer Tomography CV Convective DoD Drop-on-Demand DoF Degrees of Freedom EPS Expanded Polystyrene FE Finite Element FFF Fused Filament Fabrication FRP Fiber-Reinforced Polymer GRP Glass-Reinforced Plastic HMI Human-Machine Interface IM Indirect Manufacturing IR Infrared ME Material Extrusion MEP Mechanical, Engineering, and Plumping MPA Materialprüfungsanstalt (Material Testing Institute, University of Stuttgart) MWIR Medium-Wave InfRared NURBS Non-Uniform Rational B-Spline PEI Polyethyleneimine PLC Programmable Logic Controller List of Abbreviations xiv PU Polyurethane PUCB Phenolic Urethane Cold-Box PVA Polyvinyl Alcohol PWM Pulse-Width Modulation RTS Real-Time System SBA Selective Binder Activation SDGs Sustainable Development Goals SFF Solid Freeform Fabrication SIP Stay-In-Place SLA Stereolithography SLS Selective Laser Sintering UHPC Ultra-High-Performance Concrete Abstract xv Abstract To address the growing urgent need to reduce resource consumption, embodied en- ergy, and waste in construction, this thesis presents a new method for the zero-waste production of lightweight concrete structures using water-soluble sand formwork. The application of lightweight construction principles allows the creation of efficient and expressive structures with minimal material consumption and, consequently, an eco- logical footprint. Due to its ability to take any conceivable shape, concrete provides architects and engineers with virtually unlimited design freedom and is ideal for putting these principles into practice. However, despite the wide availability of design solu- tions known since the middle of the 20th century, lightweight concrete structures are still not widely used due to the lack of adequate sustainable production methods. This often involves formwork manufacturing, which is still labor-intensive and wasteful and accounts for over two-thirds of the production budget. Digital production methods, such as additive and subtractive manufacturing, enable highly precise creation of geo- metrically complex objects. However, their broader application in formwork production is limited by their narrow specialization in the types of geometry produced, the gener- ation of waste during processing, and the use of toxic and non-recyclable formwork materials. Therefore, the emergence of a flexible and environmentally friendly form- work method suitable for producing geometrically complex structures is necessary for the broader application of lightweight construction with concrete. Offering a comprehensive approach to the above-described problem, this thesis pro- poses a novel zero-waste technology to produce lightweight concrete structures using additive manufacturing of a specially developed water-soluble sand and binder mix- ture. The powder-bed-based 3D printing of granular materials gives the greatest free- dom in terms of geometric complexity, while the water-soluble nature of the formwork material mix allows it to be fully recycled after casting and reused in further production cycles. Following the overall goal of promoting lightweight concrete construction, this Abstract xvi technology also has an inverse effect on designing lightweight structures. It makes it possible to realize structural morphologies that would be inefficient or even impossible to produce with conventional formwork methods. The water solubility of the formwork material allows the creation of structures with geometrically complex external shapes and internal configurations. This enables not only improved structural performance but also the integration of other functional elements, such as MEP systems, acoustic and thermal insulation. The work on the thesis includes the conceptualization of closed-loop production cycle, the creation of an automated manufacturing process based on 3D printing of sand molds with a specially developed material mix, and the development of necessary accompanying CAD-CAM tools. The proposed technology is validated in the produc- tion of formworks for lightweight concrete structures of various scales, from small- scale prototypes to architectural demonstrator. Kurzfassung xvii Kurzfassung Als Antwort auf die dringende Notwendigkeit, den Ressourcenverbrauch, die graue Energie und den Abfall im Bauwesen zu reduzieren, wird in dieser Dissertation eine neue Methode zur abfallfreien Herstellung leichter Betonstrukturen mit wasserlösli- chen Sandschalungen vorgestellt. Die Anwendung der Leichtbauprinzipien ermöglicht es, effiziente und ausdrucksstarke Bauwerke mit minimalem Materialeinsatz und so- mit geringem ökologischen Fußabdruck zu schaffen. Aufgrund seiner Fähigkeit, jede denkbare Form anzunehmen, bietet Beton Architekten und Ingenieuren eine nahezu unbegrenzte Gestaltungsfreiheit und eignet sich ideal für die Umsetzung dieser Prin- zipien in die Praxis. Trotz der seit Mitte des 20. Jahrhunderts bekannten gestalteri- schen Möglichkeiten sind Leichtbaukonstruktionen aus Beton jedoch noch nicht weit verbreitet, da es an geeigneten nachhaltigen Produktionsmethoden mangelt. Dies be- trifft häufig die Schalungsherstellung, die nach wie vor arbeitsintensiv und verschwen- derisch ist und mehr als zwei Drittel des Produktionsbudgets ausmacht. Digitale Pro- duktionsverfahren wie additive und subtraktive Fertigung ermöglichen die hochpräzise Herstellung geometrisch komplexer Objekte. Ihre breitere Anwendung in der Scha- lungsproduktion wird jedoch durch ihre enge Spezialisierung auf die Art der hergestell- ten Geometrien, die Entstehung von Abfällen während der Verarbeitung und die Ver- wendung von toxischen und nicht wiederverwertbaren Schalungsmaterialien einge- schränkt. Daher ist die Entwicklung einer flexiblen und umweltfreundlichen Schalungsmethode, die für die Herstellung geometrisch komplexer Strukturen geeig- net ist, für eine breitere Anwendung leichter Betonstrukturen im Bauwesen erforder- lich. Um einen umfassenden Ansatz für die oben beschriebene Problematik zu finden, wird in dieser Dissertation eine neuartige abfallfreie Technologie zur Herstellung leichter Betonstrukturen mittels additiver Fertigung einer speziell entwickelten wasserlösli- chen Sand- und Bindemittelmischung vorgeschlagen. Der pulverbettbasierte 3D- Kurzfassung xviii Druck von granularen Materialien bietet die größte Freiheit hinsichtlich der geometri- schen Komplexität, wobei die Wasserlöslichkeit des Schalungsmaterials ein vollstän- diges Rezyklieren nach dem Gießen und die Wiederverwendung in weiteren Produk- tionszyklen ermöglicht. Im Hinblick auf das übergeordnete Ziel, den Leichtbau mit Beton zu fördern, hat diese Technologie auch einen umgekehrten Effekt auf den Ent- wurf von Leichtbaustrukturen aus Beton. Sie ermöglicht die Realisierung von Struktur- typologien, die mit konventionellen Schalungsmethoden ineffizient oder gar nicht rea- lisierbar wären. Die Wasserlöslichkeit des Schalungsmaterials erlaubt die Herstellung von Strukturen mit geometrisch komplexen Außenformen und Innenstrukturen. Dies ermöglicht nicht nur eine verbesserte strukturelle Wirkung, sondern auch die Integra- tion anderer Funktionssysteme, wie z.B. MEP-Systeme, verbesserte Schall- und Wär- medämmung etc. Die Dissertation umfasst die Konzeption eines geschlossenen Pro- duktionszyklus, die Entwicklung eines automatisierten Herstellungsprozesses basie- rend auf dem 3D-Druck von Sandformen mit einer speziell entwickelten Materialmischung und die Implementierung der notwendigen begleitenden CAD- CAM-Werkzeuge. Die vorgeschlagene Technologie wird bei der Herstellung von Scha- lungen für leichte Betonstrukturen in verschiedenen Größenordnungen, von kleinen Prototypen bis hin zu architektonischen Demonstrationsobjekten, validiert. 1. Introduction 19 1. Introduction 1.1. Context Today, the construction industry is still dominated by an extractive economic model characterized by linear production processes, excessive consumption of natural re- sources, and generation of non-recyclable waste. In concrete production and construc- tion, the situation is aggravated by the cheapness and wide availability of raw materials such as limestone, sand, and gravel, leading to their excessive extraction and dissipa- tion. With ever-increasing demand driven by population growth, the annual global con- sumption of concrete today reaches 30 billion tons, making it the most used building material, responsible for 8% of global CO2 emissions [1]. Consuming half of all materials used and producing a quarter of all emissions in the industry, the cement and concrete sectors play an essential role in meeting the Paris Agreement and decarbonizing the construction industry by 2050 [2]. This goal can only be achieved through concrete and cumulative action by all stakeholders at all levels, from cement to construction, along the entire value chain. To date, the following measures have been identified as necessary to achieve carbon neutrality: kiln improve- ments, alternative fuels, recycling of fines as raw materials, alternative binders, carbon capture and storage, changes in concrete composition, and structural optimization. At the same time, measures to reduce the demand for concrete at the construction level and the introduction of circular economy principles are assessed as the most straight- forward scenario to implement, having the shortest implementation period with the highest effectiveness [3]. Reducing the demand for primary resources in concrete construction can be achieved by optimizing designs, production, and construction processes and increasing the closed-loop use of materials and components [3]. Since concrete is currently used in- efficiently in mass construction due to standardized cross-sections and dimensions of structures caused by the limitations of production technologies, optimization of struc- tural components can already result in more than 50% of material savings [4⁠, 5]. 1.1 Context 20 Moreover, while previously resource-efficient concrete structures were difficult and very expensive to realize on the construction site due to their geometrical complexity, with the advances in digital fabrication, they can be realized off-site, in precast produc- tion plants, where sufficient control of all parameters is possible, and the realization of closed-loop processes can be facilitated. 1.1.1. Lightweight construction with concrete Lightweight construction is a design philosophy that aims to improve resource effi- ciency by designing structures of minimum weight that can carry the required loads. It is an essential approach to designing concrete structures with minimal environmen- tal impact, contributing to the decarbonization of the construction industry, as it pre- sents an integral approach to reducing the consumption of raw materials, energy, and waste. Lightweight design can be implemented on the level of material, structure, or system [8] or the combination thereof. Using high and ultra-high-performance concrete (UHPC) with a favorable ratio of specific weight to strength enables a considerable reduction of component cross-sections, reducing the overall environmental footprint [9]. On the structural level, which presents the center of interest in this thesis, the material is minimized by designing the structure to transfer the required loads towards the supports with the minimum dead weight while satisfying all given constraints. This means solving an optimization (minimization) problem in the design space with a b Figure 1.1: Lightweight concrete structures in the 20th Century: a) Los Manantiales restaurant by Felix Candela, 1958, Mexico city, [6] (Photo: Juan Guzmán/Archivo Fundación Televisa) and b) Deitingen Süd Service Station by Heinz Isler, [7] (Photo: Heinz Isler). 1. Introduction 21 constraints. A lightweight design on a system level can be considered by integrating other functional requirements. At the structural level, optimization can be performed by finding the optimal shape or topology of the structure [10]. In shape optimization, the search for the structure's ideal external shape (surface) is carried out. In the design of concrete structures sub- jected mainly to compressive stresses (favorable for concrete), shape optimization leads to thin-walled shell structural typologies with small cross-sections. Some fa- mous examples of thin shells were created by the pioneers of lightweight concrete construction in the post-war period, such as Felix Candela (Figure 1.1a), Heinz Isler (Figure 1.1b), Eduardo Torroja, and others. However, most structural typologies, espe- cially in high-rise buildings, combine compressive and tensile stresses and contain bending moments due to various constraints. Optimization of such structures can be achieved by optimizing their topology, which includes manipulation of material proper- ties, purposeful placement of cavities, and determining the connectivity of the design domain [11]. One approach to this is the so-called gradation technology, introduced into concrete construction by Werner Sobek. It involves manipulating the mechanical properties of concrete at the material level (microgradation) or introducing hollow bodies (mesogra- dation) into the component's interior according to the structure's stress state under the given loads to homogenize the stress field. In microgradation, mass savings can be as high as 56% due to using concrete mixtures with lightweight aggregates and Figure 1.2: Micrograded concrete beam (ILEK, University of Stuttgart) [12]. 1.1 Context 22 pore-forming agents that achieve a bulk density of 210 kg/m3 [13] (Figure 1.2). The choice of mineral aggregates, such as expanded glass or clay, allows the production of fully recyclable components. At the Institute of Lightweight Structures and Concep- tual Design (ILEK), in collaboration with the Institute of Building Materials (IWB) and the Institute of Systems Dynamics (ISYS), variable density concrete mixtures and an automated method for the production of precast functionally graded concrete ele- ments have been developed [14]. In mesogradation, larger mineral voids ranging in size from 10 mm to 250 mm are introduced into the component’s interior. Similar to hollow core slabs and the BubbleDeck system [15], the potential for optimal placement of mineral hollow spheres with different packing densities has recently been investi- gated at the ILEK [16]. Material savings of ca. 50% were reached in producing the 6m- long mesograded concrete slab (Figure 1.3). Thin-walled concrete hollow spheres, pro- duced in a centrifugal process, allow for the complete recycling of components as they enable the elimination of plastic, which is implemented in commercial bubble deck systems today. A further approach to optimizing concrete structures is based on the distribution of a single material with constant properties within the design domain by manipulating its connectivity. The realization of such designs became possible with the advances in digital production methods, which directly link Computer-Aided Design (CAD) and Manufacturing (CAM) to machines with Computerized Numerical Control (CNC) and robotics to produce the geometries generated by computational optimization tools. For example, the structure of the functionally graded lightweight concrete shell, the Rosenstein Pavilion, created at the ILEK, was optimized by distributing open porosity Figure 1.3: Mesograded concrete slab, (ILEK, University of Stuttgart) [16]. 1. Introduction 23 according to the stress state in the structure. The resulting geometry was fabricated from 69 individual segments cast with concrete into double-sided CNC-milled form- work segments (Figure 1.4a) [17]. Material savings of 40% were achieved compared to the solid shell of the same thickness. At ETH Zürich, extensive research has been conducted on creating topologically optimized concrete structures, particularly floor slabs, using 3D printed formworks (Figure 1.4b) [18]. As reported, up to 70% of mate- rial savings could be achieved compared to the standard concrete slab of the same dimensions. At the University of Ghent, a topologically optimized post-tensioned con- crete beam was cast into robotically extruded concrete formwork segments (Figure 1.4c) [19]. Material savings of ca. 20% were estimated compared to the T-section girder with identical overall dimensions. a b c Figure 1.4: Topologically optimized concrete structures: a) Rosenstein Pavilion (ILEK, University of Stuttgart) [17], b) topologically optimized concrete slab (Chair of Digital Building Technolo- gies, ETH Zurich) [20], c) topologically optimized post-tensioned concrete girder, (Ghent University) [19]. 1.1 Context 24 1.1.2. Problem Statement Although the optimized concrete structures, including those morphologies obtained via topology optimization methods, are favorable in material savings, they are difficult, if not impossible, to realize without an appropriate manufacturing technology or com- plex casting systems [21]. Moreover, if formwork already accounts for up to 50% of the total construction costs for geometrically simple construction elements [22], for components with complex geometries, as lightweight structures often are, costs reach as high as 80% [23]. In the 20th century, to make the production of lightweight concrete structures eco- nomically feasible, engineers tried to rationalize the design towards the producibility of formwork, in particular through the use of ruled surfaces, such as hyperbolic parab- oloids, to produce them with straight timber planks (shells of Felix Candela (Figure 1.5a)), or the use of templates of repeating shapes to reduce the cost of formwork (ferrocement technology of Pier Luigi Nervi (Figure 1.5b)). In fact, the labor cost of formwork led to the slow decline of concrete shells in the 1970s [24], and it became clear that formwork techniques have more influence on the shape of the concrete structures than even the characteristics of the structural material itself [25]. The spread of digital production technologies has enabled the manufacturing of digi- tally created forms, including those generated with structural optimization methods directly from CAD-CAM environment on CNC machines. Easily machinable materials such as wood and expanded polystyrene (EPS) have become state of practice for the formwork production of unique building projects, such as the facade of the New cus- toms yard in Düsseldorf [26], the roof of the Städel Museum [27], the foyer of the Building Academy in Salzburg (Figure 1.6a) [28] or Stuttgart 21 railway station (Figure 1.6b) [29]. Still, subtractive manufacturing is mainly suitable for creating thin-walled and shell-like shapes since the geometrical complexity of the formwork surface is lim- ited by the degrees of freedom (DOFs) of the tools used in CNC milling. Another sig- nificant problem is the generation of waste during machining and the use of non-recy- clable formwork materials. This makes the production of unique structures most Stuttgart#_CTVL00176253408787b4cc8a834d6597a62dd6b 1. Introduction 25 critical and forces designers to rationalize and unify the geometry of concrete ele- ments to minimize the number of unique formwork segments [30]. Additive manufacturing (AM) processes such as Material Extrusion (ME) or Binder Jet- ting (BJ) replace subtractive methods for formwork production due to the more effi- cient use of materials and higher geometrical freedom. While ME is more suitable to produce vertical elements (columns and walls) due to the constraints in the layer over- hang angle, BJ provides the highest geometrical freedom as it can circumvent this limitation due to printing in a powder bed. This technology can be suitable to produce topologically optimized concrete structures. However, in the state of practice (compa- nies such as Voxeljet or ExOne), chemically bound molds made from sand and furan resin binders are mainly used. This leads to irreversible curing and difficulties in dis- mantling and recycling formwork material. Improving the ecological aspects of formwork production methods, without compro- mising the economic ones, is a prerequisite for their competitiveness in the construc- tion industry and has yet to receive sufficient attention [33]. Recently, there has been a growing interest in the development of sustainable formwork production technolo- gies that, for example, use alternative recyclable milling materials, such as wax [34], sand [35], or ice [36]. More recently, these technologies have been expanded into AM a b Figure 1.5: Formwork solutions for 20th-century lightweight structures: a) timber falsework of the hypar umbrella by Felix Candela [31], b) floorplan of the ribbed slab of the Gatti wool fac- tory in Rome by Pier Luigi Nervi and its concrete stay-in-place formwork [32]. 1.2 Research goals 26 of formworks, including 3D printing of wax [37] or water-soluble PVA-based polymers, which can be dissolved in water and are biodegradable [38⁠, 39]. The main challenge here is to find a combination of materials and production techniques that simultane- ously combine technical and environmental advantages and do not limit design. Ad- dressing these issues becomes the primary motivation for the current thesis. 1.2. Research goals In the context of the above-described problems, the thesis aims to conceptualize a fabrication method for zero-waste production of lightweight concrete structures using a recyclable sand formwork system. Based on preliminary research and the analysis of existing sustainable production methods used in construction and manufacturing, a b c d Figure 1.6: Construction projects and formwork solutions for complex concrete structures: a) main foyer in Building academy in Salzburg [28] (Photo: F. Hafele), b) CNC-milled styropor formwork blocks [28] (Photo: soma architecture), c) view of the chalice columns at Stuttgart 21 main railway station (Photo: Ingo Rasp), d) CNC-milled timber formwork for chalice columns [40] (Photo: Achim Birnbaum). 1. Introduction 27 the most promising direction is identified in the application of water-soluble formwork material systems in combination with powder-bed-based AM. To test this hypothesis, the research is conducted in three stages: conceptualization of the closed-loop pro- duction cycle, creation of an automated manufacturing process based on powder bed 3D printing of a specifically developed material mix, and validation of the technology in the production of prototypes on various scales, from components to full-scale struc- tures. Since the printing of the formwork in a powder bed imposes limitations on the maxi- mum dimensions of formwork elements, and the water solubility of the formwork system requires controlled moisture conditions, the field of application for this tech- nology is primarily seen in precast concrete construction. At the same time, the di- mensions of the designed precast concrete elements do not need to be limited to the sizes of the formwork segments, as they can be cast into assemblies from multiple parts. The designed production unit also assumes the modularity of its components, and its dimensions can be adapted to the specific production requirements in future applications. Production technology also has a reverse influence on design decisions. It imposes certain constraints on the geometric characteristics of concrete components, such as requiring the elimination of isolated voids in the component’s interior for the complete removal of formwork material. Therefore, special attention is paid to integrating pro- duction parameters into the design process and ensuring a continuous flow of infor- mation from the design and simulation of the concrete part and its formwork to the export of the machine code for manufacturing. Since production technology requires inputs of resources and energy, it has a particu- lar impact on the environment and is not emission-neutral per se. However, if the energy embodied in the production process is less than the energy saved through lightweight design and waste elimination in production, the technology may prove competitive with alternative formwork production methods. 1.3 Research Methodology 28 1.3. Research Methodology The research is interdisciplinary, situated at the intersection of architecture, structural, and mechanical engineering, and conducted mainly in a physical-empirical manner. It develops iteratively through working with materials, designing computational tools and mechanical components, and running a series of production experiments. The research is organized in three phases: • In the first phase, the closed-loop production cycle is conceptualized. It includes the design of all necessary production steps, including the investigation of suit- able formwork material mixes, the identification of relevant manufacturing pro- cesses, and the detailed development of all process steps involved, including automated formwork production, casting, demolding, and recycling of form- work material mix. • Based on the production requirements defined in the first phase, in the next phase, an automated formwork production unit is designed and built. All essen- tial mechanical components are manufactured and assembled. Production pa- rameters and their dependencies are tested in the series of experiments. Nec- essary CAD and CAM tools are created to ensure continuous design-to-produc- tion workflow. • In the third phase, using the design-to-production workflow created in phase 2, various formwork production series are carried out to validate and improve pro- duction process parameters, and the geometrical quality and mechanical prop- erties of the formwork structures. Further, concrete components of various sizes and levels of complexity are designed and manufactured. 1.4. Structure of the thesis This thesis is organized into seven Chapters. Following the Introduction to the thesis, Chapter 2 provides an overview of the state-of-the-art sustainable formwork produc- tion technologies for complex concrete structures and is organized into three parts. In 1. Introduction 29 the first part, various available methods are presented and described. The second part focuses on the recyclable formwork systems from soil granulates, such as clay, silt, and sand. The third part presents the digital manufacturing of sand formworks, with a focus on AM methods. Chapter 3 presents the proposed technology and its individual steps and sets the boundaries for the research, including fields of application and requirements for the production process, formworks, and final concrete structures. Chapter 4 introduces the digital manufacturing process of zero-waste sand molds. The chapter is structured in three parts. The first part describes the choice of the selected AM type. The second part presents the prototypical powder-bed-based 3D printer and the full-scale production unit. In the third part, the necessary design-to-production workflow is introduced, and the computational algorithm for preparing formworks for production is presented. Chapter 5 describes the material composition and mechanical properties of the water- soluble sand mixtures investigated in the thesis and selected for testing on a 3D printer, presented in Chapter 4. The chapter is organized into four parts. In the first part, the positioning of the water-soluble sand molds is given within the classification of the molding systems. In the second part, suitable granulate and binder materials are selected for further investigation. Methods used for the analysis are presented in the third part. Finally, the fourth part provides information on the material properties of the investigated mixtures and gives an overview of the selected mixes for further production in Chapter 6. Chapter 6 presents the results achieved using the proposed technology. It describes the various aspects of designing and fabricating lightweight concrete structures man- ufactured with water-soluble sand formworks. The chapter is organized into three parts, presenting the results achieved on various scales. In the first part, the design and production of the small-scale prototype of a functionally graded concrete structure is presented to validate the main parameters of the developed technology. In the 1.4 Structure of the thesis 30 second part, the manufacturing of the reinforced concrete component – single-span beam is presented to address the issues of scaling and reinforcement integration. The third part describes the design and fabrication process of the architectural demonstra- tor, Marinaressa Coral Tree. Here, the application of the proposed technology in the production of lightweight concrete reinforced structures is demonstrated on an archi- tectural scale. Chapter 7 provides conclusions that can be drawn from the study. It describes the main technological achievements, potential, and limitations, and offers a summary of issues relevant to future work. 2. State of the Art 31 2. State of the Art This chapter discusses the current work on formwork development in concrete con- struction, focusing on sustainable systems. Section 2.1 gives an overview of the types of zero-waste formwork technologies available today suitable for geometrically com- plex concrete structures. Further, in Section 2.2, recyclable soil-based formworks are described in more detail, including those from clay, silt, and sand, as the potential group providing the most design freedom in casting complex concrete structures. Fi- nally, in Section 2.3 the state-of-the-art digital fabrication of sand molds is given, in- cluding the established methods and development trends. Formwork, also known as falsework, shuttering, sheathing, or mold – is a temporary or permanent support for concrete used to shape and maintain fresh concrete when casting or spraying until it reaches adequate strength. The final geometry and surface quality of the completed concrete structure are highly dependent on the formwork system employed in the construction [41]. Several classifications of formwork sys- tems can be found in literature based on the material, functional type [42], or produc- tion method [33]. In this thesis, a classification of formwork systems based on their sustainability criteria is proposed as shown in Figure 2.1 Figure 2.1: Classification of formworks based on the sustainability criteria. 2.1 Sustainable formwork systems for complex concrete structures 32 The distinction between standard and sustainable systems is drawn, where disposa- ble or reusable formworks fall under the standard category. As effective as they are for traditional construction, they do not provide sufficient flexibility for unique concrete structures with complex shapes (reusable formwork) or fail to meet sustainability cri- teria (disposable formwork) such as eliminating production waste. The sustainable methods instead aim to provide geometric freedom and eliminate waste during pro- duction. Their types, including reconfigurable, stay-in-place, and recyclable formworks, will be discussed in more detail further. 2.1. Sustainable formwork systems for complex concrete structures As sustainability is a broad concept that includes economic, social, and environmental aspects, an overview of sustainable formwork systems will be given in terms of the latter, i.e., environmental characteristics, namely the avoidance of waste during pro- duction. This can be achieved through the following three main strategies: 1. to reuse the formwork that can be adjusted to changing configurations of con- crete structures, 2. to keep the formwork permanently, preferably make it functional, 3. to recycle the formwork material. As far as the production of complex shapes is concerned, these methods have mainly emerged with the spread of digital production. However, some examples of sustaina- ble formwork systems have been known since the early 20th century. 2.1.1. Reconfigurable formworks The idea of using flexible, adaptive, or reconfigurable formwork is based on adapting the shape of a single formwork when casting concrete objects with various geome- tries, thus avoiding production waste. Flexible systems can be mechanically or pneu- matically actuated. 2. State of the Art 33 Mechanically actuated formworks A mechanically actuated formwork is based on the principle of deforming a flat surface into a desired curvilinear shape by pressing it against a rubber mat formed by a bed of variable height pins. The earliest developments of pin-controlled adaptive molds date back to the 19th century, with the first patents using automated actuators to create a flexible, adjustable, double-curved mold for concrete casting in 1979 [43]. In the 1960s, Renzo Piano first applied the pin-controlled flexible molding concept to construction applications, designing a pin-controlled machine to produce doubly curved fiber-rein- forced plastic (FRP) elements [44]. Concrete can be poured on a curvilinear surface, e.g., by spraying or in a flat state of the formwork that is later brought to the required geometry after the concrete setting. However, the latter requires careful concrete composition and rheology control to prevent cracking and spalling [43]. Several aca- demic research initiatives pushed the technology forward, including the Denmark- based Adapa company (Figure 2.2a), which transferred it to commercial applications. For the Kuwait International Airport project, they deployed the on-site fabrication plant to produce 37,000 doubly-curved concrete thin-walled panels with 13,000 various ge- ometries using just 85 adaptive molds [45]. Pneumatically actuated formworks Pneumatic formworks consist of a form-maintaining flexible membrane supported by air. As the pressure is constant over the entire membrane surface, the formwork tends to take a spherical shape, which can be influenced to a certain extent by custom-knit- ted or glued membranes or pre-tensioning the membrane with windings or cables [46]. Therefore, This method is the most suitable for producing thin-walled concrete shells, with the range of producible shapes being virtually limitless within the "pneumatically possible" [25]. Since the first patents for pneumatic formwork for concrete shells were granted in 1941, this method was widely used in the second half of the 20th century to build economical domed houses and other similar structures. Several methods of concrete application are known to date, including shotcreting on inflated formwork or on the flat surface of the formwork followed by inflation in a green state of concrete [47], or inflating the formwork with cured concrete (Figure 2.2b) [48]. The theoretical 2.1 Sustainable formwork systems for complex concrete structures 34 analog of pneumatic formworks is vacuumatic formworks, whose structural integrity depends on the negative air pressure difference. The method was pioneered by Frank Huijben from Eindhoven University of Technology [49]. Vacuumatic formworks use granular material enclosed in a flexible membrane and are stabilized by an internal reduced pressure (vacuum), which, acting on the shell, holds the aggregate particles, thus "freezing" their configuration. In the flexible state (no vacuum), the aggregate core can be formed into any shape, while in the rigid state (maximum vacuum), the stabi- lized granular material acts as a temporary load-bearing structure capable of supporting the concrete mortar until it is sufficiently hardened. 2.1.2. Stay-in-place formworks Stay-in-place (SIP), also known as lost formwork, is a formwork type that interacts structurally with concrete and works as a self-supporting formwork during construc- tion processes [52]. It can increase the load-bearing capacity of concrete or function as reinforcement. Moreover, some types of SIP formworks can extend the service life of structures exposed to harsh environmental conditions, such as offshore structures. SIP formworks enable to overcome the constraints that appear when working with more complex geometries when demolding the formwork is impossible without breaking it. Generally described as permanent formworks, these may be of two types: a b Figure 2.2: Reconfigurable formworks: a) adaptive mold D200 by Adapa with the dimensions 3,400 x 2,400 x 900 mm [50], b) pneumatically formed concrete dome (TU Wien / Öhlinger + Partner ZT Ges.m.b.H.) (Photo: Benjamin Kromoser) [51]. 2. State of the Art 35 rigid, such as a metal deck, precast concrete, wood, plastics, and the various types of fiberboards; or flexible, such as fabric, reinforced water-repellent corrugated paper, or wire mesh with waterproof paper backing [53]. Concrete formworks The application of concrete as the functional, stay-in-place formwork for concrete structures was already in use in the 20th century, e.g., by Pier Luigi Nervi, who used the precast ferrocement panels as the SIP formwork in the construction of multiple buildings [54]. The molds were precast in the plant from plaster or polymer templates, and were placed on the scaffold before the concrete was cast [55]. In 1996, Behrokh Khoshnevis developed a so-called Contour Crafting method, where the concrete act- ing as the exterior vertical wall surface was directly extruded using the 3-axis kinematic system with a mounted extruder [56⁠, 57]. This method eliminated the expensive form- work and waste and enabled casting concrete into the core, thus providing the needed strength of the structure as well as the integration of the reinforcement. The imple- mentation of Contour Crafting in construction has increased exponentially during the last decades. Further research has been focused on the improvement of the viability of the process, the integration of functional elements and reinforcement [58], imple- mentation of non-vertical geometries, improvement of mix design, including the use of UHPC [59], and acceleration of the production speed. Most recent applications also focus on using concrete 3D-printed SIP formworks for a wide spectrum of construc- tion elements, including floor slabs [60] and bridges [19] (Figure 2.3a). Foam formworks Various extrudable and fast-curing foam materials with lower density can be used as non-structural SIP formworks for building physics or other functional purposes [61]. The University of Nantes has developed a 3D printing technique, BatiPrint3DTM, that utilizes the 3D printed expandable Polyurethane (PU) foam as a formwork for concrete wall, acting as the thermal insulation for a an internal load-bearing concrete layer [66]. The technology was applied to build a 95m² social dwelling unit built in Nantes 2.1 Sustainable formwork systems for complex concrete structures 36 a b c d e f Figure 2.3: Stay-in-place formwork systems: a) topologically optimized post-tensioned concrete girder produced with concrete 3D printed stay-in-place formwork (Ghent University) [19], b) ribbed slab produced with Foamwork (Chair of Digital Building Technologies, ETH Zurich) (Photo: Patrick Bedarf) [62], c) Stay-in-place 3D printed sand formwork for topologically optimized slab (Chair of Digital Building Technologies, ETH Zurich) [20], d) cut-out of the slab with formwork infill test (Institute for Building Materials, ETH Zurich) [63], e) Mesh Mould – doubly curved concrete wall produced with spatial reinforcement mesh (Gramazio Kohler Research, ETH Zurich) [64], f) KnitCandela concrete shell pro- duced with knitted fabric formwork, (Block Research Group, ETH Zurich) (Photo: Mari- ana Popescu) [65]. 2. State of the Art 37 in 2018 [67]. To address the environmental issues associated with applying synthetic (polystyrene and polyurethane) foams, recent research in this area has focused on the investigation of extrudable inorganic mineral foams for construction applications. Re- searchers at ETH Zurich were recently using the combination of mineral foam 3D print- ing (F3DP) and concrete casting to manufacture lightweight ribbed composite slab prototype [68] (Figure 2.3b). Sand formworks At ETH Zurich, research has been conducted on the use of 3D-printed SIP sand form- works for UHPC prototypes with complex geometries over the past several years. Powder-bed-based 3D printing of molds made from fine silica sand and furan- based organic binder was carried out on an ExOne S-Max commercial 3D printer with a maximum throughput of 86 l/h. The formwork was further infiltrated with phenolic resin to increase its strength, achieving a compressive strength of 12.5 MPa and a flexural strength of 6.5 MPa. The experiments involved the production of topologically optimized slabs with dimensions of 1.8 x 1 x 0.15 m3, with an outer layer of 9 mm thick permanent formwork and short fiber reinforcement [69⁠, 70] (Figure 2.3c, d). Mesh Mold At ETH Zurich, an alternative robotic manufacturing process has been investigated that uses a grid of structural reinforcement as a fixed structural formwork for concrete [71]. This method is based on "Ferrocement" technology (invented by Joseph-Louis Lambot in 1848 and rediscovered by Pier Luigi Nervi), in which thin steel wire mesh is coated with cement mortar. The reinforcement mesh is installed by an industrial robot on a self-propelled platform. The machine automatically bends and welds the steel bar to form a structure ready to be filled with concrete. The precast mesh molds can be assembled on-site, then filled and finished using standard concrete handling and con- struction equipment (Figure 2.3e). 2.1 Sustainable formwork systems for complex concrete structures 38 Fabric formworks Compared to rigid formwork systems, fabric formworks have recommended them- selves for their inherent capacity to deal with non-standard shapes as the material is highly flexible and can deflect under fresh concrete pressure [72]. It is also preferred because of minimal material usage, lightness of the system, and ease of assembly and transport. As the fabric is deformed by the hydrostatic pressure exerted by the wet concrete during casting, the final shape must be known in advance so that the structural elements can be designed accordingly. Furthermore, the geometry must be anticlastic if the formwork must be tensioned to maintain the required stiffness. The research at ETH Zurich has shown that the fabric formwork can be used as a SIP functional surface for thin concrete shells with complex geometry [73]. Combined with computational knitting, it is possible to create non-developable surfaces, reproduce the designed geometry to a great extent, and integrate the design features, such as ribs, channels, and other functional cavities (Figure 2.3f). 2.1.3. Recyclable formworks Although almost any material can be recycled to some extent and with some effort, recyclable formwork includes materials that are relatively easy to mold and whose recycling does not require significant energy and labor costs. Generally, the recyclabil- ity of such formworks is enabled by the reversible stability of their materials, which can be affected by various processes such as heating, cooling or dissolution. Polymer formworks Recyclable plastic formworks utilize the properties of thermoplastics to melt at a rela- tively low temperature (approx. 175 °C), making them relatively easy to recycle. When combined with 3D printing, customized, geometrically complex designs can be cre- ated. 3D printing is performed by heating a filament or pellet of plastic in the nozzle of a print head and converting it from a solid state to a plastic state. When extruded, the thermoplastic material cools down quickly, gaining sufficient strength to withstand the hydrostatic pressure of the concrete. Researchers from the University of Kent exam- ined the recycling possibilities of such formworks, which are mechanically removed 2. State of the Art 39 after the concrete is poured and hardened, the material is cleaned of concrete resi- dues, collected, and crushed by machine for loading into a new production cycle [74]. The biggest challenge is to completely clean the plastic and remove all unwanted par- ticles, as they are known to clog the extrusion nozzle and affect the quality of the recycled plastic. Later, this method was further developed and tested at ETH Zurich in the production of large-scale prototypes of structural, reinforced columns, and slab elements using the same principle of formwork material recycling by shredding it after demolding (Figure 2.4a) [75]. Wax formworks Wax has been used as a molding material in arts and crafts for millennia and as a popular molding technique for metal castings, a so-called lost-wax casting. With the spread of digital fabrication, the interest in wax molds was revived in producing free- form concrete structures. Researchers at ETH Zurich have developed the zero-waste wax mold method TailorCrete, when the wax formworks were formed offsite by an adaptive, pin-actuated mold (described previously) and, after solidifying, were mounted on the standard support structure onsite for casting the concrete (Figure 2.4b) [79]. After concrete hardening, the molds were reused by melting the wax and casting new shapes. Further, the researchers from TU Braunschweig have used CNC a b c Figure 2.4: Recyclable formworks from various materials: a) Eggshell 3D printed PLA formwork (Gramazio Kohler Research, ETH Zurich) [76], b) TailorCrete wax formwork (Gramazio Kohler Research, ETH Zurich) [77], c) ice formwork, (KTH Royal Institute of Technology) (Photo: Vasily Sitnikov) [78]. 2.2 Recyclable formworks from soil materials 40 milling of wax blocks to produce formworks for free-form concrete casting [34]. Re- cent developments have shown that wax formworks can also be reused in the pro- duction of small series of precast elements, such as pollers [80]. Besides CNC milling, the 3D printing of wax molds using robotic arms was proposed to address the problem of latent heat in large wax blocks and to minimize the use of energy and material in production [37]. Ice formworks Castings in frozen molds, also known as the frozen process, were developed in the 1970s in Japan and the Soviet Union [81]. In the foundry industry, sand and water are often used, molded into the desired template, and frozen to form the ice bridges be- tween the grains and gain the required strength. To cast free-form concrete structures, researchers at KTH Royal Institute of Technology froze the water blocks to ice and CNC-milled them to the desired shape to use as formworks (Figure 2.4c) [36]. A special low-temperature hydrating concrete mix was used to cure the concrete properly. Re- cycling such formworks is not a problem and is practically automatic when the form- work melts. Ongoing research is directed towards producing more geometrically com- plex spatial structures. For this, the ice aggregate method has been developed, whereby individual molds are first CNC-milled, ice cells are frozen in them, then as- sembled and poured with concrete [82]. 2.2. Recyclable formworks from soil materials In general, the soil is an unconsolidated mineral or organic material on the immediate surface of the Earth that serves as a natural medium for the growth of land plants. Soil can be of several types, including sand, clay, silt, peat, chalk, or loam, based on the dominating size of the particles within. As these materials are all easily formable and the most present in almost every geographic region, various soil materials have been used since the early civilizations as molds for various applications, including glassmak- ing, foundry, sculpture, or construction [83⁠–85]. A brief description of the applications 2. State of the Art 41 and molding methods of the different types of soil present in the literature will be given next, focusing on their use for concrete casting. 2.2.1. Clay molds Clay has been used as molding material since the Bronze Age primarily for bronze casting, often as the binder to sand molds to improve their cohesiveness [81]. In re- cent years, with the widespread adoption of AM and the strengthening of the environ- mental agenda, robotic extrusion of clay as an inexpensive, accessible, and environ- mentally friendly raw material has become very popular in design, architecture, and construction [88]. One of the early applications of clay as a formwork for geometrically complex concrete structures was at the Paris-Malaquais School of Architecture in col- laboration with the UCL Bartlett School of Architecture, where the production of geo- metrically complex forms for building applications was investigated [86]. During this project, concrete twisted columns up to 2.5 m high and weighing up to 500 kg were fabricated using a robotic arm and manual clay extruder (Figure 2.5a). Further research at the University of Michigan focused on the combined incremental clay 3D printing paralleled with concrete casting to extend the printable height limits. Demolding was a b Figure 2.5: Clay formworks for complex concrete structures: a) 3D printing of clay formwork for 2.5 m tall column and the view of the concrete column (Paris-Malaquais School of Architec- ture in collaboration with the UCL Bartlett School of Architecture) [86], b) 3D printed clay formwork with cast concrete (Taubman College) [87]. 2.2 Recyclable formworks from soil materials 42 carried out after 24-72 h (drying period of the clay formwork), when the clay shrank and cracked upon drying, allowing it to be easily removed and recycled by rehydration and loading into the extruder [89]. 2.2.2. Silt molds Silt is a granular material of the size between sand and clay and is composed mostly of broken grains of quartz. The positive aspects of silt as a molding material for casting were discovered and systematically used by the Italian-American artist and architect Paolo Soleri, who experimented with the earth- particularly silt-casting techniques for more than 25 years starting in the 1960s [90]. He oversaw the large-scale experimental construction site of Cosanti, Arizona, and established the Cosanti Foundation – the experimental place for the earth-casting technique. He saw the earth, silt, and clay materials as informative and alive opportunities to express the architectural and aes- thetic properties of such liquid materials as gypsum and concrete and their structural a b c Figure 2.6: The use of silt molds by Paolo Soleri: a) description of Soleri’s wash away silt technique [90], b) Paolo Soleri works on one of the sculptures, c) Paolo Soleri, Single Cantilever Bridge, early 1960s, plaster, silt, and adhesive; top element: 16 x 15 x 78 in; base: 45.5 x 14.25 x 29 in. Collection of the Cosanti Foundation. Copyright Cosanti Foundation [91]. 2. State of the Art 43 characteristics. During the experimental workshops, he established the so-called wash-away silt technique, which was used for casting complex objects and formworks with spatial structure (Figure 2.6a). His technique used silt to create a positive model of the future final geometry cast in plaster or concrete. It was covered with a layer of plaster and, if necessary, reinforced. Then, the silt was washed away from the cast plaster model's interior, and the final plaster or concrete was cast. Soleri has used the wash-away silt technique to create objects of various scales, from small sculptures and models (Figure 2.6b, c) to large-scale structures, particularly building domes at Arcosanti. 2.2.3. Sand Molds Sand has more than 5000 years of history of use as the molding material for various casting applications, mainly for metal casting. Besides some evidence of its use in the Bronze Age, its official introduction into foundry as molding material happened in the 16th century in France, and the first application of bound sand molds can be traced back to the end of the 19th century [92]. Further, a brief overview of the early develop- ment of concrete sand casting will be given, followed by the use of sand in the con- struction of free-form shells in the post-war period and its revival in construction today as a molding material due to its favorable forming and environmental characteristics. Early developments The early application of sand molds in the concrete industry is known from the manu- facturing of artificial stone or cast stone – precast concrete elements, molded and finished to resemble cut natural stone (Figure 2.7a) [93]. This technique became pop- ular in late 19th and early 20th-century architecture in the United States and Canada (Figure 2.7b). The production method involved applying a pattern to wet sand in a fresh state to give it the desired profile, which was a negative relief mold for pouring liquid concrete and was quickly patented [94]. The interest in sand molds for casting was due to their cheapness and wide availability, as well as the increased strength of the cast concrete due to the lack of mold release restrictions. The positive aspect of the sand molds for concrete strength was also considered to be the absorption of excess 2.2 Recyclable formworks from soil materials 44 moisture by the mold, which improved the properties of the concrete mix [93]. The Onondaga Litholite Company became the pioneer in manufacturing sand-cast stone with a production capacity of up to 17,000 m3 a year (Figure 2.7c) [95]. The introduction of the sand molding process in the first decades of the 20th century led to a dramatic increase in cast stone production, improvements in surface finish and appearance, and eventually, its use in the construction of many prominent buildings [96]. Postwar period After WWII, due to the scarcity of building materials, sand and soil were often used as molding materials for large-scale free-form concrete shells. In 1955, Heinz Isler used a mound of earth to cast a concrete bomb shelter, withstanding the load of 60 t/m2 (Figure 2.8a) [98]. One early example of a full-scale structure using the same method was the Albuquerque Civic Auditorium, an indoor arena designed by Fergu- son, Stevens, and Associates and opened in 1957 [99]. The 19m high, 66m span con- crete dome was cast on a mound of sand, gravel, and organic binder mix (Figure 2.8b). When choosing a fabrication method, this was the cheapest option that saved the contractor $50,000. The shell was reinforced in longitudinal and transverse directions and had a variable thickness from 12 cm at the apex to 60 cm at the base. After the a b c Figure 2.7: Early developments in Cast Stone technology: a) elaborate capital being tooled by hand after casting in sand, Buffalo Litholite Company [97], b) the cover of the book “Concrete from Sand Molds” – the earliest monograph on sand casting concrete method, 1910 [93], c) View of the molding room at the Onondaga Litholite Co. [95]. 2. State of the Art 45 concrete was cured, the earth was removed from under the dome to create an arena space. In 1957, sand mounds were used to produce the concrete envelope of the Philips Pavilion designed by Le Corbusier and Yannis Xenakis for the Brussels Expo 58 [102]. The pavilion’s envelope consisted of nine concrete hyperbolic paraboloids prefabri- cated from individual cassettes and assembled on-site. Following the progressive and futuristic character of the pavilion’s architecture, it was decided to produce it with 5cm thick prestressed concrete shells. Each hypar shell was, in turn, prefabricated off-site in sections on the open sand-bed molds, split into individual cassettes up to 1 m2 in size to facilitate transport (Figure 2.8c). One sand pad was formed for each section consisting of several dozen cassettes. As the hypar is a ruled surface, the shape of each cassette was formed by marking it out on a ruler using straight boards. Some loam was added to the mix to ensure the stability of the sand. On-site, the cassettes were assembled, and the joints were grouted with mortar, followed by prestressing of the whole structure. a b c Figure 2.8: The use of sand as molding material in the 20th Century post-war concrete construction: a) freely shaped earth hill method by Heinz Isler, 1955 (Photo: Heinz Isler) [98], b) the pouring of concrete of the Albuquerque Civic Auditorium, 1956 [100], c) casting con- crete panels on the sand bed for the Philips Pavilion at Brussels Expo 58 [101]. 2.2 Recyclable formworks from soil materials 46 Contemporary developments Recently, there has been a renewed interest in using sand molds in architecture and construction due to the need for sustainable production methods for individualized concrete elements, the manufacturing of which can now be facilitated by digital pro- duction. The research has been mainly conducted on small-scale academic projects and student workshops [103⁠–105]. In 2011, during student seminars at ETH Zurich, the production of custom concrete components using new formwork strategies was investigated. In particular, the potential for the robotic forming of loose sand followed by the concrete pouring of 1 x 2 m wall panels was investigated (Figure 2.9a) [106]. Following the idea of the free-form earth hill described above, researchers at Cornell University used a mechanically formed mound of loose sand as the mold for robotically extruded doubly curved lattice concrete structure [107]. The 1-5 mm gravel was cho- sen as a supporting aggregate for its jagged geometry, offering a relatively high repose angle. The loose pile of gravel was mechanically formed to the needed shape using the printer’s gantry end-effector (Figure 2.9b). The previously mentioned frozen method was further developed at the University of Stuttgart with CNC-milling frozen sand and water blocks to produce free-form thin- walled concrete façade and shell elements (Figure 2.9c) [35]. The two-part formworks a b c Figure 2.9: Digital production of sand molds for complex concrete structures: a) robotic forming of loose sand, Procedural Landscapes seminar (Gramazio Kohler Research, ETH Zurich) [106], b) concrete extrusion on top of the gravel mound (Cornell University) [107], c) CNC-milling of the frozen sand mold (ILEK, University of Stuttgart) [35]. 2. State of the Art 47 were CNC-milled, then assembled and cast with self-compacting concrete. To im- prove the hydration of concrete, the formworks equipped with integrated copper pipes were warmed up to maintain the minimum temperature needed for curing. Other pro- jects that have explored CNC milling of sand molds have used green sand: in this case, sand was often mixed with oil, clay, or bentonite to improve the adhesion of the molds during machining. Before milling, the sand-oil mass was compacted to increase the density and avoid unwanted cavities [108]. 2.3. Additive manufacturing of sand molds As seen so far, digital manufacturing technologies have given a new impetus to pro- ducing complexly shaped individual objects in concrete construction, enabling accu- rate reproduction of digitally designed shapes, particularly those obtained through structural optimization. From the available methods shown in Figure 2.10, AM has been gaining more attention in recent years, as, when compared with subtractive and formative manufacturing, AM has three main advantages: 1. removes traditional manufacturing constraints and provides more design free- dom, 2. can shorten the supply chain and increase the profit for manufacturers, 3. has the potential to reduce the environmental impact [109]. Figure 2.10: Digital fabrication methods suitable for formwork production. 2.3 Additive manufacturing of sand molds 48 AM originated as a method for rapid prototyping of design concepts, allowing design- ers and engineers from various manufacturing and construction areas to test experi- mental models without significant investment in downstream manufacturing pro- cesses. As the processes became more accurate and versatile, the industry's focus shifted from rapid prototyping to rapid manufacturing (at that time Solid Freeform Fab- rication (SFF)) that focused on the rapid production of three-dimensional objects from a variety of molding materials: in particular polymers, metals, and ceramics [110]. As technologies evolved, an appropriate prototyping method has been identified and de- veloped for each material group, which has been categorized as Selective Laser Sin- tering (SLS) for metals and ceramics, Stereolithography (SLA) for liquid resins, Fused Filament Fabrication (FFF) for plastics and 3D-Printing (3DP) (later more often called BJ) for sand molds and cores of carious applications [111]. In the following, the most critical milestones in the development of the technology of AM of sand molds in pro- duction and the adaptation of this method to the manufacturing of formworks for con- crete structures will be briefly described. Finally, main directions for increasing the sustainability of production processes and current open areas of research will be out- lined. 2.3.1. Additive manufacturing of sand molds in foundry and composites industries As molds serve as an intermediate step towards the final product – concrete compo- nent – their fabrication using AM is usually referred to as Indirect Manufacturing (IM) [112]. The AM of sand molds gained popularity in the industry, as it enabled to remove the additional labor- and time-consuming steps from the classical mold preparation process, such as the manufacturing of the pattern and the core box), and reduced the overall production time form several weeks to 24-48 hours [113⁠, 114]. Currently, two powder-bed-based AM processes, SLS and 3DP, are considered suita- ble for digitally fabricating sand molds. The SLS method, patented in 1989 at the Uni- versity of Texas in Austin, is based on selective sintering of the binder-coated sand particles using a laser beam to induce the thermopolymerisation process of phenolic resin (Figure 2.11a) [117]. To start the reaction, the laser heats the binder to at least 2. State of the Art 49 180 °C. After the part is sintered, additional curing of the green mold in the muffle furnace is required to get the needed strength [118]. The SLS method was further developed by a US-based DTM company (now 3D Systems) and by a German EOS company in 1996. Typically, silica sand with an average grain size from 0.1 to 0.2 mm is used and coated with a phenolic resin. It is essential to properly select the sand and binder type to get appropriate thermal, physical, and chemical properties. Therefore, the equipment and precoated sand powders are usually acquainted from the produc- ers [119]. The critical process parameters are the scan spacing and speed, laser beam power, and spot diameter. These last two determine the thermal radiant power den- sity, which depends directly on the laser power and inversely on the spot diameter. The scan speed directly impacts the machine's productivity [120]. Usual scan speeds typically vary from 10 to 100 mm/s [121]. Due to the complexity of the technology, low speed, and high cost of production, the companies mentioned above have recently shifted their focus to manufacturing metal and plastic products. The AM of sand molds is more dominated today by the 3DP technologies. 3DP (BJ or BJ3DP) is a layer-by-layer process in which a liquid binder is selectively applied with an ink-jet technique in accordance with the CAD model to a layer of powdered material (usually silica or zircon sand) (Figure 2.11b) [124]. The first patent was granted to Sachs at. al. from the Massachusetts Institute of Technology in 1993 a b Figure 2.11: Powder-bed-based 3D printing methods of sand molds: a) Selective Laser Sintering [115] and b) Binder Jetting [116]. 2.3 Additive manufacturing of sand molds 50 [125], and since then, the working principle remained unchanged. Today, several major companies position themselves in the market of additively manufactured sand molds and produce printing equipment, including Voxeljet, ExOne, Viridis3D, and Z Corp. Voxeljet and ExOne use cartesian displacement machines while Viridis3D uses an in- dustrial robotic arm as the displacement mechanism (Figure 2.12). The common ma- terials used in this process are silica sand pre-coated with sulphonic acid as a catalyst and furan resin as a primary binding agent [126]. The process presents a no-bake tech- nology where the mold is consolidated via a chemical reaction between powder and binder material via a catalyst. The powder and binder materials are typically used in proportions of 99% to 1%. Depending on the equipment's size, the production speed varies from 30 to 140 l/h with a layer thickness of ca. 300 µm. After printing, the mold can harden with uncured and cured conditions. In the first case, the part consolidates in 12 hours under ambient conditions. To improve its strength, the mold can also be post-processed by heating (curing) in an air-ventilated oven at 110 °C for 2 hours. 2.3.2. Sustainability of 3D printed sand molds and development tendencies One of the industry's main developmental foci is the transition to "green production," which implies reduced emissions/waste and a healthier and safer working environ- ment. Stringent environmental regulations will severely restrict future molding and core production technologies. As a response to this, there is an increasing focus on creating more sustainable binder systems and processes for 3D printing sand molds. a b Figure 2.12: Binder-Jet 3D printers with a) cartesian displacement mechanism (Voxeljet VX 4000) [122], and b) robotic arm displacement mechanism (Viridis3D) [123]. 2. State of the Art 51 Research in this area focuses on transitioning from synthetic resins to inorganic bind- ers and developing new organic binder systems based on biodegradable polymers [127]. Sand molds with inorganic binders Binders based on sodium silicate (also known as waterglass or water glass) are the current main alternative to organic binders with synthetic resins. Moreover, starting in the 1990s, instead of the CO2-based hardening process used in molding production in the 20th century, the drying-based hardening of waterglass molds is gaining more at- tention [81]. Recently, several research groups have investigated the combination of a thermally hardened sodium silicate binder with an AM as a sustainable mold-making method [128]. The dry sodium silicate powder in the amount of 10 wt.% was premixed with the sand prior to printing. Then, it was dissolved by the thickened water and hardened by physical dehydration. After printing, the mold was infiltrated by an alcohol-water solution to dissolve the remaining sodium silicate powders further and form more stouter bonding bridges. After that, the sand mold was buried and baked in glass beads of zircon sand at 250 °C to eliminate water and form a glass phase (SiO2), through which the strength and gas volume of the sand mold was further improved. The resulting flexural strength of 4.5 – 5.65 MPa could be reached [129⁠, 130]. Water-soluble sand cores With the development of digital manufacturing, the design of molded parts has ad- vanced significantly toward functional integration and complex geometries, including cavities and undercuts. Traditionally, channels and cavities were formed using sand cores from the same materials as the outer molds (chemically bonded sand and resin mixtures). However, their removal from inside the casting by hot stripping, chemical corrosion, or mechanical stripping, which destroyed the casting or required a special device, made the casting more expensive and reduced production efficiency and sus- tainability. Therefore, the development of easily removable sand cores, especially 2.3 Additive manufacturing of sand molds 52 water-soluble ones, has attracted increasing interest in recent decades, mainly in the foundry and fiberglass composite industries. Water-soluble sand molds are composed of a mixture of sand and one or a combina- tion of binding agents with water-soluble properties. The cores are hardened via the physical process of dehydration as opposed to the chemical reaction of polymerization described previously. In recent years, microwave heating has been successfully used to overcome the high energy consumption of convective heating. The physical bond enables the cores to be decomposed by solvents, such as water or other liquids. Cur- rently, inorganic, and organic binder systems can be used as water-soluble binders, including geo-polymers, salts, proteins, polysaccharides, and vinyls. Over the decades, multiple patents were granted for the development of various material systems with water-soluble binders using classical core preparation methods, including the inorganic binders, such as borax-phosphate [133], or magnesium sulfate [134] and organic bind- ers, such as polyvinyl alcohol [135], gelatine [136], or dextrin [137]. With the advances of AM, and primarily of BJ in core production, several attempts have been made to combine the idea of producing sand molds, hardened with water- soluble binders, with 3D printing. Minimal information is available on this topic, and no open-source information on the recyclability possibilities of the soluble sand cores and their binding materials can be found yet. In 2014, Voxeljet AG and Fluidsolids AG pa- tented the method of producing water-soluble sand molds using the BJ technique, described very generally [138]. By their invention, the water-soluble mold should com- prise at least one water-soluble material, be produced with the BJ technique, and then sealed on the surface with insoluble material. The mold could be cast with any hard- ening flowable material. After hardening, it should be washed out, preferably with an aqueous solution of a warm temperature (preferably up to 80 °C), typically higher than the melting temperature of the bonding material used. No specified binder material was described; the possible materials for use were listed, including polysaccharides, proteins, salts, silicates, tannins, polyvinyl acetates, polyvinyl alcohol, polyvinylpyrroli- done or a mixture thereof. 2. State of the Art 53 One public mention has been found of Voxeljet's use of 3D-printed water-soluble sand cores for the production of the FRP motorcycle rocker arm in 2015 [131]. A water- soluble sand core with an inorganic binder was fabricated using the BJ method and then cured in the oven to gain the necessary strength. The core surface was sealed with a slurry and dried at 60-80 °C in a convection oven. Further, it was covered by a water-based agent, which was brushed on the surface and dried repeatedly in the oven at 60-80 °C. Finally, the core was laminated with a polyester material reinforced by glass fiber tissue and washed out after hardening (Figure 2.13a). No further infor- mation on recyclability/reusability was given. The binder specifications were not pro- vided. In 2016, researchers at ETH Zurich experimented with the AM of a geometrically com- plex robotic leg structure. The component was produced on the ExOne S-Max printed using the quartz sand and inorganic water-soluble binder, then sprayed with a sealing layer and cured in the oven [139]. Another public mention was found of ExOne using washout tooling to remove sacrificial cores in the production of FRP composites (Fig- ure 2.13b) [132]. This is probably related to ExOne researching the BJ of water-soluble sand cores using a polyethyleneimine (PEI) binder in 2021 [140]. The sand cores were printed on ExOne X1-lab printer with a resolution of 200 µm. The PEI binder was used a b Figure 2.13: Use of water-soluble sand cores for laminated composite products: a) Process steps of manufacturing a motorbike swing arm made of GRP (from left to right) produced by Voxeljet: 3D printed core; sized core; fine-coated core; laminated, already washed-out core; painted motor-bike swing arm with metal inserts [131], b) washout tooling used by ExOne company in production of laminated FRP composites [132]. 2.4 Conclusion 54 as the solution supplied through the printhead. After printing, the part was cured in the oven at 180 °C for 2 hours to gain strength. The maximum green strength of 6.28 MPa was reached. The tested specimens were used to laminate fiber composites, and the core material could be washed out after hardening. 2.4. Conclusion As can be seen from the state-of-the-art presented in this chapter, sustainable form- work methods have a long history and diversity. Each group has evolved to fulfill a specific purpose. With the advent of digital manufacturing, these methods have been further developed by adapting available digital production techniques. Common to all presented technologies is their narrow specialization in a particular type of producible geometry, which is directly related to the forming method and materials used. Thanks to their reusability, reconfigurable formwork systems present zero- waste solutions. They are the most suitable for producing thin-walled or shell-like serial objects or unique structures within possible configurations. Adaptive formworks are suitable for façade and wall panels and pneumatic systems – for thin-walled shells of a specific range of shapes. Stay-in-place formworks provide more freedom in the choice of form as they do not need to be dismounted. However, they must be func- tional to be effective, so materials should be selected that can fulfill a specific function, e.g., load bearing, insulating, reinforcing, etc. Recyclable formworks are the most flexible sub-type of sustainable formwork sys- tems when combining geometric flexibility with waste minimization. They are the most flexible when dealing with complex spatial structures with undercuts, bottle- necks, inner channels, etc. However, the challenge is finding the best possible com- bination of formwork properties, such as formability, sufficient temporary strength, demoldability, and recyclability of the formwork material. The variety of recyclable formwork systems presented in the state-of-the-art indicates a great interest in this subject in academia and practice and a desire to find a balance between geometry, material, and production process requirements. However, the processes of 2. State of the Art 55 freezing/heating (as in the case of wax and frozen processes) lead to additional energy expenses that need to be further investigated. The author’s experience with frozen sand formwork [35] has shown the limitations in terms of comfort in working in a climatized environment, particularly the formwork handling (cleaning) at low tempera- tures, making it difficult to use them in conventional production halls. The demolding and recyclability procedures pose another challenge – the multistep production pro- cess, which requires additional time, energy, and labor expenses. In other manufacturing areas, mold production processes are increasingly focused on the requirements of green production, i.e., with reduced energy and waste. The pop- ular production of chemically bonded sand molds with furan resins will soon cease, giving way to alternative, more environmentally friendly technologies. Current trends in research and practice in AM of casting molds are focused on finding sustainable alternatives, biomaterials, and non-toxic binders. Considering and adapting industrial methods for use in the construction industry is of interest for more sustainable pro- duction of concrete structures. In this context, using water-soluble formwork systems is the most gentle and prom- ising method, as it allows easy demolding of complex-shaped castings at room tem- perature with virtually no restrictions, which also has no adverse effect on the con- crete. The open questions, in this case, are: • the investigation of material system, which would combine the necessary strength and geometric stability during formwork production, pouring and cur- ing of concrete with sufficient water solubility at room temperature in the short- est possible time, and • the conceptualization of a sustainable automated manufacturing process to pro- duce geometrically complex molds using the developed material mix. Thus, this thesis aims to find a combined solution to these two problems and proposes technology for the automated production of geometrically complex concrete struc- tures using 3D-printed molds from a mixture of sand and a specially selected binder. 2.4 Conclusion 56 The general concept of the technology representing a closed-loop production cycle is presented in Chapter 3. The conceptualization and production of an automated 3D printing unit are described in Chapter 4 and the investigation of the suitable material mix is presented in Chapter 5. The developed method is tested in the production of various prototypes, described in Chapter 6. 3. Zero-waste sand formworks for complex concrete structures 57 3. Zero-waste sand formworks for complex concrete structures This chapter describes the general concept of the technology to produce geometrically complex concrete structures using a recyclable sand formwork system. The technol- ogy is conceptualized as a closed-loop production cycle, where formwork material cir- culates as long as possible, eliminating waste production. Following the general de- scription of the approach in Section 3.1, Section 3.2 describes the expected application range of the technology, and Section 3.3 describes the requirements imposed on the technology, including the production cycle in general, the involved production process steps, the formwork, and the concrete component as the final product. 3.1. Technology concept Figure 3.1 presents the concept of technological process for zero-waste production of lightweight concrete structures using recyclable sand formworks. The closed produc- tion cycle consists of 3D printing of formwork, pouring of concrete, demolding of con- crete structure, and recycling of formwork material for use in subsequent production cycles. The formwork 3D geometry is generated in a CAD program. For large-scale concrete elements, the formworks are divided into individual segments depending on the di- mensions of the operational volume of the 3D printer. Further, the 3D volumes are sliced into layers, the printing routines are prepared for every layer and compiled into the machine code for production. The overview of the computational workflow will be described in Subsection 4.3.1. The formworks are produced on a powder-bed-based 3D printer, whose design is explained in Section 4.2. The printing process combines all necessary steps for printing and curing the formwork, so no post-processing steps for further hardening are needed and the formwork is ready for casting right after the printing is done. The material system for formworks consists of a sand and water- soluble binder that enables 3D printing of geometrically stable yet water-soluble form- works that can be easily recycled. Various material mixes are investigated in Chapter 4 and the most preferable one is selected to produce prototypes, described in Chapter 6. 3.1 Technology concept 58 Figure 3.1: Closed-loop production cycle of water-soluble sand molds. 3. Zero-waste sand formworks for complex concrete structures 59 When 3D printing is accomplished, the formwork is cleaned up from the unbound sand that is ready for reuse and loaded into the 3D printer again for further production. The formwork is then prepared for casting. For large-scale components, several formwork segments are assembled with each other. The reinforcement is placed, and the con- necting elements are installed. The classical casting process then proceeded with. After the concrete has hardened, the component can be demolded by submerging it in the water bath or splashing it with room-temperature water from the hose. During the research, both options were tested, with the first one prevailing for small-scale specimens and the second one for bigger structures. In both cases, the container for collecting sand and binder mix is required to collect all formwork material for further recycling. The recycling part of the production cycle starts with the separation of sand and binder mix from the excess water. A filter system can be used for this. The wet mix is loaded into the industrial, commercially available drying oven at 70 – 105 °C for 24 hours. After the formwork material is dried, it is taken out and ground on the commercially available grinding machine. After the grinding, the mixture is loaded into the 3D printer for fur- ther production cycles. The comparison of the freshly prepared and recycled material mix will be given in the Subsection 5.5.4. 3.2. Application range Due to the properties of the formwork material and the multistep production process that requires controlled environmental conditions, the main application area of the pro- posed technology is seen in the off-site production in the precast production plant. All manufacturing steps of the closed-loop production cycle can be well adapted to flow production line. With this technology, concrete structures and their parts, components, and elements can be produced. Moreover, consulting with precast production companies has indi- cated the need for more sustainable formwork solutions for structures with custom 3.3 Technology requirements 60 opening systems, often requested by clients. Until now, such inlays have been exe- cuted using EPS or timber formworks that also involve manual labor in mold prepara- tion and demolding by plant staff (which is laborious due to concrete shrinkage and often destroys the formwork, making it single-use). Thus, the broader application of this technology is expected in the manufacturing of precast concrete structures that need some degree of customization, including inlays, undercuts, and profiles that are economically or ecologically unreasonable to produce otherwise. Types of structures, producible with this technology: 1. structurally optimized precast concrete components, 2. components with internal cavity system for functional integration: o concrete core activation, o thermal insulation, o acoustic insulation, o heating/cooling, another ductwork, 3. façade elements with functional integration: o thermal insulation, o daylight control, o support biodiversity, 4. non-construction components: o furniture and landscape objects, o thermal coolers, o engineering equipment made from mineral, self-hardening materials, 5. underwater structures. 3.3. Technology requirements The requirements for developing technology include those for the production prod- ucts, including concrete structures and formwork, and the production processes. Each category should meet functional and non-functional requirements, i.e., environmental. In the light of the Sustainable Development Goals (SDGs), technology should be 3. Zero-waste sand formworks for complex concrete structures 61 aligned with the Responsible Consumption and Production Goal (SDG12) [141], which aims at the efficient utilization of natural resources and reduction of industrial waste through prevention, reduction, recycling, and reuse. 3.3.1. Formwork requirements As a product of the developing technology, the formwork needs to meet the require- ments applied to formworks for concrete construction to fulfill the three main criteria, namely quality, safety, and economy [142], with a detailed description below: quality: • the accuracy of formwork geometry (shape) during production, casting and hardening, • the quality of formwork surface; safety: • sufficient strength to sustain own weight during production, han- dling, assembly, and demolding, • sufficient strength to sustain weight and hydrostatic pressure of concrete during casting and hardening; econ- omy: • reasonable production time, • reasonable production cost. Quality The formwork must be geometrically precise and stable at several stages of produc- tion, from 3D printing, handling, and transportation to concrete casting and hardening. The geometric precision set in CAD when designing concrete structures must be maintained during the 3D printing. The dimensional tolerances must be within the ad- missible range of those set in DIN 1045-4 [143] for precast concrete structures. The formwork should be geometrically stable, i.e., should not deform under the loads ap- plied when handling, transporting, and, most importantly, casting. The combination of 3.3 Technology requirements 62 strength and cross-sectional thickness should be designed appropriately to minimize deformations. The surface quality of concrete components depends on the mutual selection of two parameters of formwork: the grain size of the formwork material mix and the height of the print layer. If fine sand mixtures are used (0.1-0.3 mm), and the submillimeter layer heights are applied, the surface quality (very rough to rough) comparable with traditional timber formworks can be expected. Strength The formwork must safely carry the designed loads along all necessary process steps. The formwork should carry its weight at all stages of the production process, including 3D printing, cleaning, transportation for assembly, assembly, and casting. After con- crete is hardened and no deviations are caused during this process, further reduction of formwork strength (i.e., due to moisture absorption) is no longer crucial. Formwork should sustain the hydrostatic pressure of the fresh concrete, which increases linearly with the height of the component in the vertical direction. Anticipating the relatively low strength of recyclable formwork mix, compared to steel or timber analogs, the use of the formworks for precast concrete structures is seen as beneficial, as it ena- bles the orientation of components for casting to minimize their height, thus reducing the hydrostatic pressure of concrete to a minimum. Economy The economic aspects of the final product consist of the total production time and cost. Being a multi-stage production process, to be competitive with modern digital production methods for complex concrete structures, the total production time of sim- ilar quality products should not exceed the production time of similar processes. Cost is expected to be affected by the duration of production as it consumes energy and resources, human labor if applicable, and the cost of resources used in the process. Thus, it is essential to keep the energy consumption of the production process low. Also, the automation of as many production steps as possible will reduce the costs. 3. Zero-waste sand formworks for complex concrete structures 63 Finally, the circulation of materials within the production process for as long as possi- ble will reduce material consumption costs. 3.3.2. Material Requirements Requirements that are set for materials used in production technology can be divided into two groups: • product-oriented – provide needed quality and strength of the manufactured formwork and concrete component. • process-oriented – quality and costs of production and processing of the mate- rial within the production cycle, the possibility of its recycling. The raw