1 Introduction

In recent years, the development of computational design has led to the advancement of 3D printing technology (3DP) in construction, bringing new solutions to recurrent problems such as labour shortages, long construction times, high costs, and inadequate quality of the resulting product. Most of the research is focused on optimizing construction in medium- and large-scale static architecture. However, despite the apparent need of modern consumers to adapt architectural objects to rapidly changing environments, the same aspects in kinetic architecture (KA) are studied relatively less.

Deployable structures can be considered a subset of KA due to their ability to transform and adapt to changing conditions. In this sense, the field of deployable structures represents an important area within the broader context of KA, which seeks to create dynamic and responsive structures that can interact with their environment in various ways. These structural mechanisms can go from a compact state, suitable for storage or transportation, to an extended and fully deployed configuration in a matter of minutes. We can distinguished four main groups if we classify them according to their structural system (De Temmerman 2007): spatial bar structures consisting of hinged bars, foldable plate structures consisting of hinged plates, tensegrity structures and membrane structures. Focusing on their wide applicability in the field of KA, their high degree of deployability and reliable deployment, two subcategories are established: scissor structures and foldable plate structures. Scissor structures are extendable lattice structures consisting of bars joined by hinges, allowing them to folding into a compact bundle. Foldable plate structures consist of plate elements, which are connected by line joints allowing one rotational degree of freedom (De Temmerman 2007).

Developing a sustainable strategy for building structures is crucial in reducing the enormous amount of waste generated by the construction industry. Deployable structures offer an advantage in terms of reuse as they have the ability to change their shape and function. They can be easily disassembled, transported, and reassembled in a different location or for a different purpose. Additive manufacturing (AM) also offers some advantages, such as faster and cheaper production, geometry optimization, digital workflow and automation, reduced CO2 emissions, use of recycled materials and freedom of design. However, requiring skilled employees and high start-up investments might be considered drawbacks. Moreover, Advanced technology typically involves complex systems and components, which can make it more difficult to diagnose and repair in the event of breakdowns or malfunctions. This complexity can also require specialized expertise and tools, leading to higher costs for repairs and maintenance. Additionally, the use of advanced materials and components in technology can make it more susceptible to damage or wear, further adding to the need for frequent repairs and maintenance. While advanced technology can offer many benefits and advantages, it is important to consider the potential challenges and costs associated with maintaining and repairing these systems.

2 Research goals and methodology

In response to the idea of combining LSAM and KA, this paper introduces the experimental case study of a deployable disaster shelter printed on-site. 3DP can impact on some aspects of disaster management. It can improve response to emerging disasters and provide needed shelter for victims (Gregory 2016).

By printing the necessary number of required parts directly at the disaster site or nearby instead of arranging the transportation logistics, the amount of time and money spent could be reduced significantly during relief efforts. This would help disaster victims find temporary shelter and return to everyday life sooner. There are two options when printing on-site. The first would be to install the equipment directly on the construction plot where the printed structure will be erected. This requires preparing the ground, installing the printing equipment and bringing the material to the plot. The printing is done in one piece (or several pieces of significant size) and if something goes wrong during the process, it is usually difficult to correct or even means starting from scratch. Of course, once the process is finished, it requires dismantling the equipment, moving it and starting the cycle all over again. The second option is to set up a manufacturing facility where the equipment is assembled just once and where we store the material. It may even be the case that this manufacturing facility arrives already pre-assembled, in a container, for example, and all that is required is the delivery and placement of the container to start manufacturing. In this case, the production consists of parts that must be transported to the plot and assembled there. The team opted for the second option as we felt it offered better performance in an emergency situation. In this scenario, it is important that the manufactured parts are not too heavy (if necessary, they should be moved to the construction area without the use of machinery) and that they are not too complex or too different from each other (this facilitates the assembly and manufacturing process). Keeping this in mind, the project focuses on the scissor structure, as we find it more versatile, adaptable, flexible and simple. All these factors make it the ideal object of our research, which aims to explore the computational design of such structures and optimize usability for 3D fabrication.

Since the main goal is to gain insight into the structural behaviour of LSAM regarding scissor structure, we experimented with different designs (bar and joints) and printing capabilities as part of the concept phase. During fabrication, however, we were constrained by a narrow time frame, so we selected only a representative portion of the model for the manufacturing phase. The workflow is presented in the diagram Fig. 1.

Fig. 1
figure 1

The diagram provides a visual representation of the structured and systematic flow of steps within the project, highlighting the interdependencies between tasks

Full size image

3 Theory review

3.1 Large scale additive manufacturing (LSAM)

LSAM is a natural extension of 3DP using AM techniques, which proposes the idea of constructing building-sized objects rapidly with less manual intervention than using traditional construction techniques (Barnett and Gosselin 2015).

AM technology is currently being developed across various industries, including electronics, biomedicine, defence, and aerospace (Pasco et al. 2022). For example, in the electronics industry, AM is being utilized to fabricate intricate structures for printed circuit boards and other electronic parts. Additionally, in the biomedicine sector, AM is being employed to produce patient-specific implants, like customized prosthetics and orthopaedics. The defence industry is also leverage AM to design and manufacture lightweight and robust components for military equipment and vehicles. Furthermore, in the aerospace industry, AM is being employed to create complex, lightweight structures for aircraft and spacecraft, such as engine parts and airframe components. This layer-by-layer technique allows optimizing designs by accelerating up production, speeding up delivery, reducing material waste and creating new geometries unattainable with traditional manufacturing methods. Unlike traditional fabrication processes such as injection moulding or machining, 3DP does not start with an existing mould or material, but it uses a computer model and a printing strategy for production (Gregory 2016).

LSAM builds on advances in AM technology to provide a solution for producing large objects with the same precision and efficiency as smaller components. This technology has several advantages over traditional manufacturing techniques, including the ability to reduce waste and increase production speed. Regarding our project about the applicability of LSAM to the fabrication of kinetic structural elements, our research explores whether LSAM can be used to produce scissor structures. By using LSAM to fabricate these structures, we believe that the same benefits of reduced cost and increased efficiency can be realized.

On the market, different machines exist to carry out this large-scale printing task. However, industrial robots provide greater flexibility and enhance fabrication possibilities. Not only do they bring unprecedented design freedom and resource management to manufacturing, they offer a superior solution thanks to their versatility and the possibility of equipping them with a variety of tools. In addition, LSAM expands both the possibilities of off-site construction and directly on-site.

In recent years, there have been buildings and architectural objects illustrating the capabilities of LSAM technology. In 2021, a stainless steel 2.5 m wide and 10 m long bridge was printed by MX3D to cross a canal in Amsterdam. The project demonstrates the potential of multi-axis printing technology to produce robust, complex and graceful metal structures (MX3D 2021). Another project developed in the USA is a prototype for an initiative called Additive Manufacturing Integrated Energy (AMIE), which aims to shape new perspectives on producing, storing, and using electricity (Brake 2016). In addition, an important feature of this method is the possibility of waste-free construction (Pasco et al. 2022). The ETFE Plastic Roof Canopy of the 6 Bevis Marks Building in England shows the potential for stability in construction. Printed metal parts are complex connections between a building’s columns and canopy brackets (Beyhan et al. 2018). Finally one example using concrete. On 30 April 2021, the first tenant of the first Dutch house made of 3D printed concrete received the key to his home. The house in Eindhoven, the first of five houses in the Milestone Project, meets all the strict Dutch building requirements (Eindhoven University of Technology 2021).

Despite the outstanding results, the technology of LSAM still faces multiple challenges. Although the complexity and size of the planned structure become more approachable with more delicate detailing during the architectural design process, the final fabrication of large-scale geometric shapes often fails due to fabrication difficulties (Teizer et al. 2016). A common LSAM failure scenario occurs when the project team has designed the building using this technology, but during the fabrication process discovers that the equipment used is not capable of handling the size and weight of the structural components. The fabrication process then faces several problems such as overheating, material distortion and dimensional inaccuracies, leading to the decision to stop the project and redesign the building using traditional construction methods. This can lead to significant delays in project completion and additional costs, highlighting the challenges that still need to be overcome in the implementation of LSAM technology in large-scale construction projects.

3.2 Scissor structures

Deployable scissor structures can be fast and reversibly transformed from the stored state, in which they form a compact and easily transportable bundle, to the deployed state, in which they perform their function. Scissor grids usually form mechanisms with one degree of freedom, which, when fully unfolded, are fixed and become load-bearing structures. Deployable scissor grids have many different shapes, describing linear, single-curved, or double-curved geometries. The simplest and most widespread in the application are the single-curved structures—for instance, a semi-cylindrical shell.

Scissor structures can be classified as translational, polar, and angulated, respectfully to their unit type. A scissor unit, or scissor-like element (SLE), is the basic structural unit of any deployable scissor grid (Roovers et al. 2014). A scissor unit consists of two bars interconnected by a hinge, allowing the bars to move in the unit plane. An SLE unit can only move on one axis plane. Translational and polar units consist of two straight bars, and the angular unit consists of two angled bars (see Fig. 2).

Fig. 2
figure 2

Scissor structures classification and their corresponding SLE. Translational units with equal bar size can only form planar shapes (a) while translational units with various bars’ sizes enable forming curvature (b)

Full size image

The translational units exist between two parallel lines (the unit lines) that will never intersect during deployment. The unit lines connect the upper and lower ends of the bars. Translational units with equal bar size can only form planar shapes ((a) in Fig. 2). Translational units with various bars’ sizes enable forming curvature ((b) in Fig. 2). Polar units also consist of same-size bars, but the hinge connection is shifted from the middle; therefore, it is possible to create curvature. The angulated unit consists of two angled bars so that the elements are divided into two semi-bars at the hinge.

The deployment angle between units changes depending on the deployment stage. The final deployment configuration can be classified based on the overall curvature (plane, single, or double), the grid directions (two-, three-, or four-way grid (see Fig. 3) and the type(s) of units used (translational, polar or both).

Fig. 3
figure 3

Two-way grid with directions A and B and Three-way grid with directions C, D and E

Full size image

This research focuses only on single curvature two-way grid composed of translational and polar units (see Fig. 4).

Fig. 4
figure 4

Polar and translational units on a two-way grid

Full size image

4 Computational workflow: from design to assembly

4.1 Design decisions

This research considers two different approaches to the design of emergency shelters using the scissor structure, regardless of whether, in the end, we produced only the one that offered the best performance and was best suited to our manufacturing conditions (see Fig. 5).

Fig. 5
figure 5

The approach (a) performs a single curvature and forms a semi-cylindrical shape while the approach (b) performs a double curvature and forms a doughnut shape

Full size image

The first approach (a) performs a single curvature and forms a semi-cylindrical shape. It consists of polar and translational units flowing in opposite directions. The first ones are responsible for the curvature formation, and the second for the linear deployment. All bars are the same size. Translational units are connected using the intermediate hinge at the center of the bar; the hinges in polar units are shifted to achieve forming the arc-like shape.

The second approach (b) performs a double curvature and forms a doughnut shape. It consists of polar units that form double curvature in both directions. All bars in each horizontal layer have different sizes that differ from layer to layer. This gives it the ability to form a second curvature.

Finally, the case study developed the first approach because the bars are all the same, simplifying the manufacturing and assembly process, which can save time, reduce reliance on unskilled labour and make it easier to replace broken or lost elements, especially in emergency situations.

The dimensions of the shelter have been optimized to meet both functional requirements and practical considerations. The width of 2.50 m and height of 2.35 m provide ample space for its intended use, including accommodation for personnel and equipment. The thickness of the scissor structure, at 0.25 m, ensures adequate strength and stability, while keeping the overall set as compact as possible. The length of the bar unit, 0.60 m, has been determined to optimize structural integrity, while facilitating efficient transportation and storage of the unit (see Fig. 6).

Fig. 6
figure 6

The image displays both the elevation and plan views of the shelter, with dimensions measured in millimetres

Full size image

The project also includes anchors for more stability and two layers of cover. The inner skin is a waterproof tensile membrane. The outer skin is a water harvesting membrane, e.g. a fishing net might be used. Both covers can be attached using the joins after the shelter deployment (see Fig. 7).

Fig. 7
figure 7

The image portrays the constituent parts of the shelter

Full size image

4.2 Computational design

The shelter was created with Grasshopper 3D for Rhinoceros 3D, an algorithmic editor that offers flexibility in the design process. Computational design involves engineering an algorithm, allowing for analysis and simulation from start to finish. This is highly advantageous in emergency situations where adaptability is crucial. Using the presented algorithm, a variety of shelters can be easily fabricated by adjusting the parameters to suit different scenarios.

  1. 1.

    The initial shape of the structure conforms to a barrel vault, so the starting inputs for the algorithm are the parameters needed to create a primary arc that generates such a geometry from its translation. Specifically, three points in space are set to build this initial arch, determining the height and width of the structure in its simplified version. The simplified version is the one that considers the bars of the structure only as lines, without any volume.

  2. 2.

    Next, we set the values of the following parameters, number of polar units and offset distance for the arc (thickness of polar units deployed), thus completing a polar unit system that traverses the arc shape in its unfolded state.

  3. 3.

    In the next stage, we decide how many of these systems our structure requires (total length of the structure), copy them at the appropriate distance according to the dimension of the bars and tie them together by means of translational units. This generates several translational unit systems running perpendicular to the polar ones.

  4. 4.

    It is time to combine all the units in a simplified model, enabling analysis and simulations of the complete structure. The selected software tool to simulate the folding and unfolding of the structure was Kangaroo 2 (Piker 2013), a plug-in for Grasshopper that simulates physical behaviours.

  5. 5.

    At this point, we are ready to provide volume to our simplified model. We start with the design of the joints. We opted for a simple solution consisting of a central tube-shaped core with four attached flanges with their respective holes to receive and dock the unit bars, both polar and translational. When designing those mentioned above joints, it was also taken into consideration that they could be used to anchor the structure to the terrain. Of course, the joint is also expressed in algorithmic mode, meaning that just like the simplified structure, we control the design by setting parameters.

  6. 6.

    The final step is to introduce our joints into the simplified model and shorten the lines representing the bars, making room to convert them into solids that align and fit perfectly with the joints. The lines will first be converted into 2D sections and then extruded in the corresponding section’s normal direction to generate the 3D solid. Again the width and thickness of the bars are parameters supplied to the algorithm as inputs that the user can modify.

The full workflow involves creating a 1:10 scale model using an Ultimaker 2+ desktop printer. The model was printed in parts, then assembled to check its structural functionality before it was produced at full size (see Fig. 8).

Fig. 8
figure 8

The image showcases a 1:10 scale printed model of the shelter in various stages of deployment, ranging from a fully compact configuration to a fully deployed state

Full size image

Our Grasshopper algorithm, shown as pseudo-code in the diagram below (see Fig. 9), includes suggestions for adding interior and exterior skins to the structure to provide protection from the elements, shade, privacy, and even water collection capabilities.

Fig. 9
figure 9

The diagram displays a pseudo-code representation of our Grasshopper algorithm

Full size image

4.3 Fabrication

Once all our parameters have been adjusted and the model has been consolidated, it is time to move the design to the construction phase. The elements to be manufactured are of three types: the bars, the joints and the anchors for the foundation. Eventually, only a portion of the structure was fabricated. The selected section is composed just of bars and joints, so no anchor was printed and tested. The LSAM of the parts was carried out using a 6-axis industrial robot (Kuka KR 16-2) equipped with a polymer extruder with interchangeable nozzles (see Fig. 10).

Fig. 10
figure 10

Our 6-axis industrial robot (Kuka KR 16-2) equipped with a polymer extruder including a 6 mm diameter nozzle, engaged in the process of printing one of the bars

Full size image

The nozzle diameter mounted on the extruder was 6 mm, and as for printing material, a cellulose-filled plastic bio-composite (UPM Formi 3D 20/19—UPM). This material is a bio-based plastic containing cellulose fibres from wood. These fibres are key to achieving a more stable printing, helping to balance the internal stresses of the material as it solidifies. The following Table 1 summarizes the technical specifications of the UPM Formi 20/19.

Table 1 Physical/mechanical properties UPM Formi 3D
Full size table

After several printing trials with the configuration described above, we found that the most suitable layer height was 3 mm to achieve a homogeneous wall thickness of 8 mm. The 8 mm thickness was chosen after thorough analysis and testing to ensure an optimum balance between structural integrity and material efficiency. It provides sufficient resistance to bending and buckling while minimising material usage. We also realised that print quality is drastically affected by the following factors: robot speed, extrusion speed and cooling process. It is therefore necessary to find the right balance between these parameters, which may not be constant throughout the part. Thus, a process of trial and error is necessary to find the ideal setting for the parts.

The remaining step was to translate our geometries into points in space that had to be orderly navigated by the robot, depositing material along the entire path and thus generating the printed part. In our case, we also had to consider an additional challenge: our extruder does not allow intermittent running. It does allow adjusting the extrusion speed of the material, but once the printing starts, the material continues to flow until the end of the process.

This inconvenience forced us to be more careful when designing our printing strategies, paying special attention to the robot’s paths, maximizing its effectiveness and minimizing material waste.

The robot programs were generated using KukaPRC (Geddes et al. 1992), a plug-in for Grasshopper 3D, thus keeping the whole computational workflow within the same algorithmic editor.

Bars:

The manufactured bars have a length of 600 mm and a width of 60 mm, dimensions resulting from the parametric modelling and optimization of the structure. As for the bar height, several tests were carried out by applying different numbers of printing layers. Through extensive testing and analysis, it was established that a height of four layers for the bar provides the optimal trade-off between structural rigidity and weight optimization. This height ensures that the bar has the sufficient mechanical strength to withstand the applied stresses while minimizing the overall weight of the structure. The fabrication time for each bar was 35 to 37 min and the final height ranged from 15 to 16 mm.

Considering that the bars should be light, we realized from the beginning that one of the advantages of this fabrication method was that it provided us with the option of manufacturing lattice bars, i.e. that they were not just a solid block. However, we opted to solidify those sections completely where the connections to the joints and the hinges would be placed. Therefore the structure of the bar can be expressed as the sum of five parts: solid joint-lattice body-solid hinge-lattice body-solid joint.

As for the layers, two printing strategies were developed and used alternately until the four layers of each bar were completed. Both strategies used different directions for material deposition, thus achieving a more coherent and homogeneous final piece (see Fig. 11).

Fig. 11
figure 11

The image displays a representation of the four constituent layers of a bar, arranged in a bottom-to-top sequence

Full size image

Joints:

We designed the joints to a maximum size of 260 mm \(\times\) 260 mm and a height of 70 mm, using the results of our printing tests which showed that 18 layers of material are required to achieve the desired height. We arranged four flanges in a cross pattern around a hollow central core to compose the whole joint. We then applied a continuous spiral path printing strategy, which resulted in the wall thickness of the flanges being made up of two adjacent lines of material, giving a total thickness of between 17 and 18 mm for each of the flanges. The production time for each joint varies between 24 and 26 min (see Fig. 12).

Fig. 12
figure 12

A subset of the outputted joints with dimensions of 260 \(\times\) 260 \(\times\) 70 mm

Full size image

4.4 Assembly

The mounting operation of the printed parts can be divided into two phases: post-printing treatment and assembly.

We refer to post-printing as the process of preparing the parts prior to assembly. During this stage, the residues of the printing operation were removed, and the parts were drilled to insert the mounting hardware. This task was performed manually in the workshop using conventional tools such as a hand press machine, a bench vise and a power drill.

All the printed bars were evenly divided into two groups. With the help of cardboard templates, one for the translational units and one for the polar units, we marked and drilled the holes to host the joints. In parallel, nine joints were processed in the same way, also using their own template. In this case, the four flanges were marked and drilled.

As for the assembly phase, we used M10 bolts of 60 mm length for the joints (bar–joint connection) and 80 mm length for the hinges (bar–bar connection). Everything was tightened together with their corresponding nuts and washers (see Fig. 13).

Fig. 13
figure 13

Different steps involved in the assembly process, including: a drilling of the bars to accommodate hardware, b mounting of the translational units, c installation of the polar units, d close-up of the joint assembly, e detail of the hinge connections between bars, and f a final representation of the fully assembled structure in its compact configuration

Full size image

5 Conclusion

5.1 Final model description

The team successfully developed a prototype incorporating three polar and three translational systems, comprising a total of 12 bars for the polar units and 12 bars for the translational units. The systems were integrated using nine articulating joints and secured via the use of fasteners. The resulting prototype demonstrates good deployability characteristics, as the design enables smooth and efficient folding and unfolding operations.

One bar weighs between 499 to 515 g; one joint weighs between 705 to 719 g, and the whole prototype weighs 27.90 kg. At least two people are recommended to handle the prototype safely and steadily. This means that the installation of a complete structure made up of 3x9 modules may require the assistance of a crew of 2 to 4 workers, along with some necessary equipment such as a crane or handcart to transport the folded structure to the designated site. The aforementioned structure would consist of 132 bars and 40 joints and, based on extrapolation, is estimated to weigh approximately 120 kg. (see Fig. 14).

Fig. 14
figure 14

Final prototype consisting of three polar and three translational systems, comprising a total of 12 bars for the polar units and an additional 12 bars for the translational units

Full size image

5.2 Critical review and proposed improvements

Improvements for future developments were identified, particularly in material selection and geometry. A comprehensive design approach that considers both factors is necessary. The design-space of DfMA (Design for Manufacturing and Assembly) framework is useful for exploring constraints and opportunities, aiding in selecting efficient materials, manufacturing processes, and assembly techniques. More research is needed to advance our understanding of the complex relationship between materials, geometry, and manufacturing in the context of DfMA. The discussion can be broken down into three sections: material, printing strategies, and technical issues.

Materials:

The primary material selected for the project is PLA plastic, which was reinforced with wood fibres to enhance its structural stability for 3D printing. However, this approach also introduced some drawbacks, such as decreased cohesion between layers, which was particularly evident in the printed bars. During testing, the rigidity of the material and connectors resulted in damage to the structure after nine consecutive deployments. Delamination of the layers was observed, particularly in areas where the bar featured openings rather than being fully solidified. This was likely due to the increased stress concentration caused by the lack of structural continuity in these regions. Analysis indicates that the properties of the chosen material for manufacturing may have contributed to this behaviour. To improve performance, it would be valuable to experiment with different polymer-based materials, such as those with greater flexibility, as alternatives to PLA plastic. Other materials that could potentially outperform PLA plastic are listed in the next Table 2.

Table 2 Alternatives to UPM Formi 3D
Full size table

Printing strategies:

PLA plastic can be brittle when it cools quickly, so it may sometimes be difficult to merge layers evenly. For example, considering the constraints of our extruder (it does not allow material flow interruptions during the printing process), we opted for a zigzag layer printing strategy. That is, the top layer starts printing at the same location where the bottom layer finished being printed. This implies that in the area described previously, the end of the first layer and the beginning of the second layer, the material cools down for only a few seconds before receiving the top layer. This facilitates the adhesion of layers (the material remains in a more fluid state) but hinders the support of the top layer (the print tends to collapse). On the other hand, if we look at the opposite end, the beginning of the bottom layer and the end of the top one, the material waited minutes to receive the top layer. Therefore it has cooled and solidified to a greater extent than the other end. The consequences are reversed here, the material is stiff enough to support the top layer, but the adhesion between layers is compromised. This problem of insufficient adhesion between layers is something that was not noticed immediately after the printing of the parts. It was identified and aggravated in the following days after the manufacturing process.

After our experience, we believe that it would be more successful and we would achieve better results, better bars, if we change the zigzag strategy to a loop strategy. That is, all layers start printing at the same end. Meaning when the bottom layer is finished printing, the extruder must relocate to the opposite end before starting the printing of the top layer. Considering our limitations once again, it would mean that the extruder must describe a path far enough away from the geometry of the part to be manufactured since the material extruded during this path will be discarded.

Technical Issues:

To minimize frailties after the printing and simplify the post-printing treatment process, designated holes for bolts and nuts should be incorporated in the design stage. PLA plastic joint and bar advised attaching a protective tube to the holes to avoid brittleness. Furthermore, incorporating table markings parameters in the Grasshopper design stage before the bars and joint fabrication process will allow the industrial robot performance to be more precise.

5.3 Applications

From the very beginning, even before starting our research, we imagined that a potential application could be the erection of temporary shelters in emergencies after natural disasters such as earthquakes, hurricanes or floods. In the aftermath of catastrophes, materials are highly restricted. At least the new materials coming from factories, because usually these catastrophes also release tons of what normally is considered garbage. Our proposal is to recover and reuse part of this waste, in particular plastic waste, thus improving the response time to these disasters (see Fig. 15).

Fig. 15
figure 15

The image showcases a visual depiction of our idea of using the technology to build temporary shelters in the aftermath of natural disasters

Full size image

The entire printing process could be accomplished on-site. Adaptable shelters are not only versatile in terms of assembly but also expand the user experience. Users can rearrange the layout as needed depending on the activity or function.

The concept of a deployable and adaptable shelter is not limited to catastrophes. These types of designs can also be exploited in other fields, such as space exploration. Returning to the Moon to establish a permanent station is currently gaining momentum. It also seems that LSAM could play a major role in achieving this dream. Undoubtedly, it is much cheaper and more efficient to send tools to manufacture on the lunar surface than to send already made pieces from Earth. Of course, there are still many challenges to overcome, the most important of which is how to transform regolith into a printable material.

5.4 Future

The study builds the foundation for merging the scissor structure theory with LSAM using an industrial robot. The experiment results envisioned a further journey in experimenting with this type of structure, considering different technical factors such as the selection of a lighter and stronger material or the optimization of the printing strategy.

Furthermore, developing a project with sustainability as the key objective, researching LSAM to seek a deployable structure to be fabricated with less time, lower cost, and better quality are definitely advantageous not just in the field of architecture but also in other aspects

On the other hand, more could be further researched to explore the design of the scissor structure to enhance the practicality of being seen as a shelter to house people in the future. Along the research development, the functionality and spatial quality could be pushed to a better level to have a double degree of freedom instead of a single degree of freedom. A double degree of freedom (DOF) scissor structure design allows the structure’s shape to stretch in two axes, forming either a parallel structure or a trapezoidal structure. Spatially, it can stand alone or combine with other units creating public, semi-public, semi-private and private spaces for different spatial usage.