Introduction

Both natural and man-made disasters continue to cause mass displacement, disrupt livelihoods, and strain humanitarian systems globally (UNHCR, 2024). The 2023 Turkey-Syria earthquake sequence is a stark reminder of these impacts. On February 6, two powerful earthquakes (M7.8 and M7.6) struck southeastern Turkey and northern Syria, causing over 50,000 deaths, widespread infrastructure damage, and the displacement of around 3.2 million people (AFAD, 2023). The response was widely criticized for delays and poor coordination among key actors, including AFAD, the Red Crescent, the military, and municipal authorities (IHD, 2023). A major shortage of emergency shelters worsened the situation, leaving vulnerable populations at risk (Hatay Planning Center, 2024).

These systemic weaknesses are especially concerning for Istanbul, where experts have predicted a 35–70% chance of an earthquake of magnitude 7.0 or higher by 2033 (Ergintav et al., 2014). Istanbul’s dense population of over 15 million (TÜİK, 2023), its 1.1 million buildings (Veri Kaynağı, n.d.), and complex infrastructure make the city highly susceptible to additional risks such as fires, liquefaction, and tsunamis (Alpar et al., 2003; JICA & IMM, 2002; Bozkurt & Duran, 2018). Factors like aging, informally built structures, lack of open spaces, and growing urban inequalities further limit the city's disaster resilience (TMMOB, 2025).

The 2023 earthquakes revealed critical gaps in Turkey’s disaster management, shelter provision, and preparedness (AFAD, 2023; TMMOB, 2023; IHD, 2023). In Istanbul’s densely built and highly vulnerable urban environment, the absence of effective shelter solutions could lead to consequences that exceed current projections in the event of a major earthquake, with temporary sheltering needed for upwards of 3 million people. (Ergintav et al., 2014; Alpar et al., 2003; Bozkurt & Duran, 2018). These overlapping risks demand shelter responses that are not only structurally responsive but also socially attuned to the diverse needs of displaced populations.

Aim of the Study

The aim of this research is to develop a dynamic sheltering system that is attuned to the evolving needs of displaced populations in the aftermath of a potential large scale earthquake in Istanbul. Rather than focusing on the shelter unit alone, the research attempts to design the entire sheltering process — starting with the rapid deployment of emergency shelters and transitioning into flexible, adaptable structures capable of supporting prolonged residence as the recovery progress unfolds. Recognizing the limitations of existing standardized shelter models, the research seeks to develop a structurally resilient, contextually responsive, and sustainable solution with the ability to respond dynamically to the different phases of post-disaster recovery.

Research Questions

The central research question guiding this research is:

How can a deployable, modular, and adaptable sheltering system be designed to address the evolving needs of displaced populations following a large-scale earthquake in Istanbul?

To explore this central question, the below sub-questions are proposed:

• What are the design and structural requirements for a shelter system that can be rapidly deployed after an earthquake and remain functional throughout extended recovery phases?
• How can shelter design incorporate flexibility and adaptability to meet the changing needs of displaced populations during different recovery phases?
• What cultural, climatic, and social factors specific to Istanbul must be considered when designing a shelter system to ensure its appropriateness and usability across different communities?
• How can contemporary technologies be integrated into the shelter design process and throughout its life cycle?

Disaster, Response and Management

Disasters are major disturbances triggered by natural or human-made hazards, often leading to physical destruction, loss of life, and long-term socio-economic consequences (Sawalha, 2018; Takagi et al., 2021). They typically involve sudden onset, wide societal impact, and complex resource management challenges. Disasters are commonly categorized as natural, man-made, or hybrid events that combine both (Shaluf, 2007).

The type and scale of a disaster influence its impact. For example, earthquakes often lead to building collapses and fatalities, especially in dense urban areas (Takagi et al., 2021). Vulnerability varies across populations, with groups such as the elderly and economically marginalized facing greater risks (Han & Suh, 2023). Effective disaster management consists of four key phases: mitigation, which focuses on reducing risks through hazard assessments and public awareness (Alrehaili et al., 2022); preparedness, which ensures quick, coordinated response through emergency planning and early warning systems; response, which prioritizes life-saving actions and emergency aid immediately after a disaster (Neal, 1997); and recovery, which involves rebuilding infrastructure, restoring services, and providing psychological support (Alrehaili et al., 2022).

Turkey’s inefficiencies in disaster management became visible during the 2025 Marmara Sea earthquake. Despite its moderate magnitude (6.2), the earthquake triggered widespread panic in Istanbul, with 127 aftershocks and 151 people jumping from windows in fear (AA, 2025; Governorship of Istanbul, 2025). Many assembly areas were inaccessible due to uncontrolled urbanization, and communication failures worsened the panic (Gazete Oksijen, 2025).

Shelter / What defines it?

Shelter, along with food, water, and sanitation, is a fundamental pillar of post-disaster recovery and humanitarian response. It is not merely a physical structure but a critical element that preserves life, health, security, and human dignity (Maffei, 2012). The right to adequate housing is internationally recognized, including in the Universal Declaration of Human Rights and the International Covenant on Economic, Social and Cultural Rights, underscoring its importance in disaster contexts.

Shelter provides immediate protection against environmental hazards, but its role extends to supporting psychological well-being, restoring daily routines, enabling social cohesion, and granting access to essential services (Tsioulou et al., 2020). For vulnerable groups such as the elderly, children, and low-income communities, adequate shelter is crucial for equitable recovery (Zahari & Hashim, 2018). Without it, recovery is delayed, social inequalities deepen, and vulnerabilities increase (Shelter Cluster, 2018).

Conventional sheltering progresses through four phases. Emergency shelters offer immediate but minimal protection and are often used longer than intended (IFRC/RCS, 2013). Temporary shelters provide basic services in a transitional setting (Quarantelli, 1995). Temporary housing allows daily life activities to resume and supports stability until permanent homes are built (Johnson, 2007). Permanent housing restores stable living but is often delayed by legal and logistical issues (Saraf et al., 2019).

The design of a post-disaster shelter relies on a thorough understanding of key criteria that ensure its functionality, safety, and effectiveness in addressing the needs of affected communities. An effective shelter must consider a broad range of factors that influence both the immediate well-being of its occupants and their long-term recovery. This includes not only physical aspects such as total area, height, and structural system, but also contextual and climatic elements like site conditions, local culture, traditions, and climate (Daher et al., 2015).

Clear design criteria are essential to ensure shelters meet minimum living standards and support physical, emotional, and cultural needs. Key criteria include: 

• Structural Integrity
• Lightness and Portability
• Ease and Speed of Assembly
• Habitability
• Reusability and Sustainability
• Modularity and Adaptability
• Encouraging User Involvement and Self-Recovery

A New Approach: Transitional Shelter

The concept of transitional shelter is a vital paradigm in post-disaster reconstruction, emphasizing the incremental movement from displacement toward long-term stability and independence. Despite various shelter types and criteria, a persistent limitation is the static view of shelter as a one-time product. This outlook often leads to inadequate short-term solutions or unsustainable long-term outcomes. Unlike traditional approaches that treat shelter as a final product, transitional shelter views it as a dynamic process—a series of evolving shelter phases aligned with the changing needs of affected communities. This approach spans the full spectrum of post-disaster housing, starting from temporary shelters and progressively upgrading to semi-permanent housing through community participation, professional support, and resource access (Shelter Centre, 2012).

Transitional shelters serve as a bridge between short-term emergency accommodation and permanent housing reconstruction. They provide a “temporary, but not short-term” solution that offers displaced people enhanced stability, dignity, and privacy—qualities often missing in emergency tents or collective shelters. This phased model allows shelters to adapt and improve over time, addressing immediate humanitarian needs while laying the foundation for sustainable recovery and reducing displacement disruptions (Nizam & Ikaputra, 2025). It embraces the fluid and complex nature of post-disaster recovery, supporting incremental improvements through community engagement and resource availability. By bridging emergency response and permanent reconstruction, this model promotes dignity, resilience, and long-term recovery (Wagemann Farfán, 2016).

Building on these principles, this thesis proposes the design of a transitional shelter that incorporates the essential shelter criteria discussed earlier, aiming to respond effectively to both immediate and evolving needs of disaster-affected communities.

Deployable and Demountable Structures

At the core of transformable structures lies the principle of reconfigurability—the ability of a system to change its configuration purposefully. This adaptability is often achieved through modular components that can be reassembled for different functions or kinetic mechanisms enabling transformation between distinct forms.

In literature, transformable structures are broadly categorized into two main types based on their functional objectives, design strategies, and applications (Temmermann, 2012):

• Deployable Structures
• Demountable Structures

Deployable Structures

Deployable structures are engineered to shift from a compact, stowed state to a fully functional configuration, enabling easy transport, quick assembly, and efficient use of space (Fenci & Currie, 2017). Their core functions include packaging, which involves minimizing size and volume for transportation; deployment, the process of transforming into the operational form; and stabilization, which secures the structure in place once deployed to ensure rigidity and functionality.

Common types of deployable structures (Temmermann, 2007):

• Scissor structures: Linked bars folding like scissors
• Foldable plate structures: Hinged plates folding like origami
• Tension structures: Use cables or membranes under tension
• Tensegrity structures: Combine tension and compression elements

Demountable structures

Demountable structures are systems designed for easy assembly, disassembly, and relocation, supporting sustainability and adaptability (European Commission, 2020; Kirchherr et al., 2017). They enable material reuse, minimize construction waste, and promote long product life cycles, making them highly suitable for sustainable building practices. These structures are characterized by their modularity, often employing grid-based systems that allow for flexibility in design and rapid reconfiguration. They can easily adapt to changing spatial needs, offering efficient and versatile solutions in both temporary and permanent contexts.

Common systems within demountable structures include interlocking modular systems, where components can be connected, removed, or rearranged to suit evolving requirements (Birjukov & Bolotin, 2015). Another widely used approach is grid structures, which provide stability and design flexibility. These grids can take various forms, such as orthogonal, triangular, hexagonal, radial, or diagonal grids (Temmerman et al., 2012). Each grid type offers distinct advantages in terms of load distribution, spatial organization, and construction efficiency, making demountable structures adaptable to a wide range of applications.

Why not both?

As observed in the literature review and case studies, a fundamental trade-off exists between rapid deployability and structural durability in post-disaster shelter design. Structures optimized for immediate emergency response—those that are lightweight, compact, and easily transportable—often lack the resilience required for long-term habitation and may perform poorly in harsh climates. Conversely, more robust and durable shelters typically demand greater investment of time, materials, and resources for construction, even when proposed as transitional solutions. This can pose significant challenges in the immediate aftermath of disasters, particularly in complex urban contexts like Istanbul where widespread damage and infrastructural collapse are anticipated.

This trade-off presents a critical design challenge: post-disaster shelter systems must address both short-term emergency needs and longer-term living requirements. When assessing the key criteria for adequate post-disaster shelters, it becomes clear that deployable and demountable structures each fulfill distinct needs.

• Deployable structures offer rapid assembly, portability, and lightweight solutions, which are essential in emergency scenarios. However, their reliance on mechanical joints and moving parts often reduces their long-term durability and complicates maintenance.
• Demountable structures, on the other hand, usually require more time and effort to assemble, which may be a disadvantage immediately after disasters. Nevertheless, they tend to be more adaptable, reusable, and better suited to evolving, longer-term needs.

This thesis investigates the feasibility of a hybrid structural approach that combines deployable and demountable qualities to overcome the trade-offs between rapid deployment and long-term adaptability. To test this concept, a bault varrel scissor structure is chosen for the first phase deployable structure to be dismantled into a demountable structure at the second phase. For the second phase two structural options are examined in parallel:

• A rectangular shelter reinforced with diagonal struts instead of tension cables
• A 2-frequency geodesic dome

The geodesic dome was selected for its outstanding structural efficiency, thermal performance, and modularity. Its triangulated geometry ensures even load distribution, offering excellent resistance to wind and seismic forces. It’s low surface-area-to-volume ratio contributes to better energy efficiency and insulation. Beyond its technical advantages, the dome carries cultural significance, resonating with historic nomadic values of protection, flexibility, and communal living. By blending these cultural references with contemporary engineering, the geodesic dome presents itself as not only a structurally sound but also a symbolically meaningful post-disaster shelter solution.

Design Process

This research extensively employs both physical and computational prototyping to support an iterative design process. A Grasshopper script originally developed by Joseph Oster (2017) was modified to generate the scissor structure, forming the basis for parametric exploration. The initial geodesic dome geometry was sourced from SketchFab, and the majority of design iterations were conducted using Grasshopper, the Weaverbird plugin, and core modeling tools within Rhinoceros (Rhino). These digital platforms provided precise control over both geometry and structural behavior throughout the design development.

In the early stages of physical prototyping, simple, low-cost materials such as straws, needles, and tape were used to enable rapid form-finding and initial testing of geometric configurations and scissor mechanisms. These quick prototypes allowed efficient experimentation and informed the next design phases. In later stages, 3D printing was introduced to produce more precise, detailed, and structurally accurate prototypes. This transition significantly improved the evaluation of structural performance, joint articulation, and assembly behavior, leading to more refined and reliable design outcomes.

First Iteration

Following the development of the initial geometries, 3D printing became the main prototyping method for both the scissor structure and the second-phase structures.

For the scissor structure, both joints and struts were fully 3D printed, as plastic straws lacked the precision needed to model the mechanism accurately. An initial single-module prototype was created to study its kinematic behavior and structural performance.

For the geodesic dome and rectangular grid structure, a hybrid approach was used: joints were 3D printed for accuracy, while struts were made from plastic straws, as precision was less critical in the straight elements. The rectangular configuration required six different joint types, while the geodesic dome required four joint types. Additionally, both structures needed two different strut lengths to accommodate their geometric conditions.

Different strut lengths for the scissor structure were tested to optimize the system’s performance. Two sizes—170 cm and 200 cm—were prototyped and evaluated.

While the 200 cm struts were structurally sound, they presented practical limitations related to handling, transportation, deployment, and disassembly. Additionally, the larger deployed footprint was not suitable for dense urban environments like Istanbul, where space is limited. As a result, 170 cm struts were selected as the most appropriate option, offering a practical balance between structural efficiency, compactness, and usability.

For the geodesic dome, two strut lengths were required, commonly referred to as A and B struts, with a geometric relationship of:

A strut ≈ 88.4% of B strut length

Based on this ratio:

If A struts are 170 cm, B struts should be approximately 192 cm.

In the rectangular scissor grid configuration, two strut lengths were also necessary, based on the geometric relationship:

B strut = A strut × √2

Based on this:

If A struts are 170 cm, B struts should be approximately 240 cm.

This evaluation of strut lengths, joint types, and fabrication methods informed the final design decisions, balancing structural reliability, spatial efficiency, and deployability.

Second Iteration

To minimize unique components in the second-phase structures, an initial study tested whether the translational and polar units of the scissor structure could be built using struts of varying lengths. A scaled test was performed with the geodesic dome, as its strut lengths were closer. However, using different lengths prevented full collapse, reducing the compactness crucial for deployable shelters. This approach was discarded.

To simplify production, a standard strut length was proposed. By extending the joint arms of the longer strut connections, varying edge lengths could be achieved using identical struts. For the geodesic dome, the longer B strut (originally 192 cm) was adjusted by extending joint ends to compensate the extra length while using 170 cm standardized struts.

A similar approach was tested for the rectangular structure, where the diagonal B strut should be about 240 cm. Adjusting this with joint extensions would require additional joint types and complex front-face geometries, increasing fabrication complexity and reducing modular efficiency. This option was found impractical.

Cables were then explored as multi-purpose components—used both for stabilizing the scissor structure during deployment and as diagonal bracing in the second-phase shelters. This dual-use reduced material variety and improved assembly efficiency.

A full arc of the scissor structure was fabricated using 3D-printed joints and struts, assembled with M4×16 mm bolts for secure, repeatable connections.

Two second-phase shelter options were then developed and compared:

The rectangular form offers intuitive space division and compatibility with conventional furniture but demands more structural material and a heavier, more complex assembly.

The geodesic dome, while challenging in interior planning, offers superior lightness, portability, habitability, structural safety, and adaptability. Its radial clustering echoes traditional communal spatial organizations, fostering social cohesion.

Due to these advantages, the geodesic dome was selected as the second-phase structure for this transitional shelter system, best addressing the diverse demands of post-disaster environments.

Final Proposal

Following the selection of the second-phase structure, both the scissor structure and the geodesic dome were fully constructed using 3D-printed components. This approach eliminated the inaccuracies and tolerances present in earlier straw prototypes, enabling a more precise evaluation of structural behavior.

The joints were designed for modularity and ease of maintenance. Their lower components can be detached or replaced without affecting the alignment or integrity of the surrounding structure. All connecting elements were pre-drilled to support quick, accurate, and efficient on-site assembly.

Aluminum was selected for the shelter’s structural system due to its optimal balance of strength, lightness, corrosion resistance, and ease of manual assembly—key factors for post-disaster deployment. It is lighter and more portable than steel, more affordable and practical than carbon fiber, and highly recyclable, making it both efficient and sustainable. For the wall and floor modules, fiberglass reinforced plastic (FRP) sandwich panels with mineral wool insulation were chosen. FRP offers lightweight durability, excellent thermal and acoustic insulation, high corrosion resistance, and improved fire safety when combined with intumescent coatings. Its prefabricated, customizable surfaces—including options like wood texture—support rapid assembly, user engagement, and cultural relevance, making it an ideal choice for adaptable, humane shelter environments.

To ensure ease of maintenance and long-term adaptability, the joints were specifically engineered so that their lower components can be detached or replaced without compromising the structural integrity or alignment of the surrounding elements. Both connecting elements are pre-drilled to facilitate efficient and accurate assembly on site.

Due to the limited dimensions of the triangular panels forming the dome, a standard door could not be integrated directly. Consequently, a dedicated door module was developed, comprising two pentagonal panels—one attached to the dome structure and the other serving as the entrance. This module also functions as a connector between separate units and can be attached to any of the five sides of the dome, enabling modular expansion and spatial flexibility.

The wall system is divided into two categories: facade modules and interior modules, with facade modules designed to meet higher requirements for water resistance and thermal insulation. The exterior modules are designed to sit atop the structural struts and are secured using aluminum L-profiles. Each module consists of a sandwich panel composed of two 10 mm layers of fiber-reinforced plastic (FRP), with a 50 mm core of mineral wool providing thermal insulation. Additional edge-sealing components enhance the water resistance of the facade system. Four types of wall modules are available:

• Fully closed
• Half-height window (top)
• Full-height window
• Half-height window (centered)

The interior modules use a modular design for flexible space configuration. They attach to structural struts via pre-drilled holes and screws for easy installation and reconfiguration. Each triangular wall panel has a grid with holes at every node, allowing components to be mounted along any edge. While pre-fabricated modules are provided, the open-source system encourages user customization to fit individual needs.

The space is intended to encourage users to explore, adapt, and personalize it according to their needs. However, to provide initial guidance and demonstrate the range of possible configurations, several example layouts are proposed as starting points. The four proposed configurations each have distinct characteristics tailored to different stages and needs in post-disaster sheltering.

The overall planning of the settlements follows a modular approach, developed in accordance with UNHCR standards for emergency and transitional shelter layouts. In addition to the standard modular arrangement, a small-scale community cluster model was introduced, grouping 4 to 6 families around a shared open space. The residential units are positioned radially along the perimeter of this central area, creating a communal environment inspired by traditional Turkish nomadic practices, where tents were arranged around a common space and daily activities took place in front of the tent entrances. This configuration fosters a sense of community, encourages outdoor interaction, and promotes both physical and mental well-being among residents. by reinforcing social ties and providing access to open, shared environments.

Grounded in democratic and open-source ideals, the design uses a hybrid deployable-demountable system for compact storage and easy deployment. This shelter not only responds to emergencies but supports recovery by bridging short-term relief and long-term habitation.

In summary, the thesis offers a user-centered, structurally focused shelter solution that emphasizes shelter as a process, rather than a static product.