Subterranean Currents (MSc)

SUBTERRANEAN CURRENTS

Maria Aranzales | Sherine Elabd | Orfeas Rachiotis

ARCHITECTURAL ASSOCIATION SCHOOL OF ARCHITECTURE MASTER OF SCIENCE IN EMERGENT TECHNOLOGIES AND DESIGN 2024–2025

Architectural Association, 2025 36 Bedford Square, London WC1B3ES

Architectural Association (Inc), Registered charity No. 311083 Company limited by guarantee. Registered in England No. 171402

SUBTERANNEAN CURRENTS

M.Sc Candidate

Sherine Elabd

M.Arch Candidates

Orfeas Rachiotis Maria Paula Aranzales

Founding Director

Dr. Michael Weinstock

Programme Head

Dr. Milad Showkatbakhsh

Studio Tutors

Abhinav Chaudhary Paris Nikitidis Danae Polyviou Dr. Álvaro Velasco Pérez

ARCHITECTURAL ASSOCIATION SCHOOL OF ARCHITECTURE GRADUATE SCHOOL PROGRAMMES

PROGRAMME:

EMERGENT TECHNOLOGIES AND DESIGN

YEAR:

2024 - 2025

COURSE TITLE:

MSc. Dissertation

DISSERTATION TITLE:

Subterannean Currents

STUDENT NAMES:

Sherine Elabd (M.Sc) Orfeas Rachiotis (M.Arch) Maria Paula Aranzales (M.Arch)

DECLARATION:

“I certify that this piece of work is entirely my/ our and that my quotation or paraphrase from the published or unpublished work of other is duly acknowledged.”

SIGNATURE OF THE STUDENT:

Sherine Elabd (M.Sc) DATE:

19 September 2025

ACKNOWLEDGEMENTS The research team would like to express their deepest gratitude to Dr. Michael Weinstock, founding director of EmTech, for his invaluable insight into the social and ecological dimensions of this thesis. We are equally grateful to Dr. Milad Showkatbakhsh, programme head of EmTech, for his unwavering support of the project’s vision, and to Dr. Anna Font, Head of Studies at the Architectural Association, for her incisive input, which enabled the project to expand into broader domains. We also wish to thank our studio tutors, Abhinav Chaudhary, Paris Nikitidis, Danae Polyviou, and Álvaro Velasco Pérez, for their continuous guidance throughout this journey. Finally, we extend our heartfelt thanks to our peers within the AA community for making this an unforgettable experience, and to our families and friends for their patience, encouragement, and unwavering support along the way.

ABSTRACT Desertification is expanding at an unprecedented rate, yet vast fossil aquifers lying beneath many threatened drylands still offer the hydrological capital to sustain and even rehabilitate these fragile territories. This study harnesses that potential by advancing a prototypical subterranean settlement for Egypt’s Bahariya Oasis, where the Nubian Sandstone Aquifer remains a critical but finite buffer against climate-driven aridity. This research is founded on a closed water-cycle system that combines groundwater extraction with passive atmospheric condensation, intended to catalyse urban decentralisation away from the Nile corridor. Drawing on existing knowledge of desert dwelling practices, the project situates itself at the intersection of underground structures that provide environmentally resilient spaces, clustering typologies derived from desert cities, locally sourced and developed building materials, architectural elements that reflect cultural values of privacy and self-sufficiency, and contemporary water management strategies interwoven into the urban fabric. Initial climatic, geomorphological and socio-cultural surveys of the Bahariya establish the foundations of the proposed settlement. A predictive modelling study of the Nubian aquifer informed the data-driven site selection process. The settlement is then divided into four prototypes: the cistern, housing, agriculture and condensation farm units. These are aggregated according to water consumption patterns and organised spatially to generate an experimental urban fabric, from which a complementary road and hydrological network emerges. Excavation by-products were tested as construction material, contributing to the architectural development while exploring their structural, thermal and hydrophobic potential. Each unit was further examined for environmental performance and architectural qualities: condensation towers channel airflows underground to condense water; cisterns act as social hubs and storage points; housing integrates social and environmental strategies with water networks for thermal regulation; and agricultural units combine purification processes with palm and wheat cultivation to support both the economy and a regulated water system. This study concludes by reiterating its objective of developing a subterranean settlement that harnesses underground water flows and overground atmospheric flows to sustain and propagate itself. Methods for integrating individual architectural units into a coherent urban fabric are examined as the research shifts from a computational and performance-based scale to an iterative, mega-scale project. Fig.01 Water Extraction in the Bahariya Oasis

CONTENTS 01 DESERT LANDSCAPES ..16 Arid Regions Increased Risk Of Desertification Potential Of Ground Water

02 DOMAIN

Egypt, A Nation That Follows Water The Untapped Potential of The Oasis Understanding Desert Living Construction Methodologies

..26

03 RESEARCH METHODOLGY ..70

05 DESIGN DEVELOPMENT ..130

04 RESEARCH DEVELOPMENT ..78

06 CONCLUSION

Urban Systems and Networks Architectural Development and Analysis Material Research and Prototyping

Site Evaluation and Selection Settlement Organisation Hydro-Social Network Material Development

Atmospheric Water Harvesting Water Storag and Community Spaces Residential Clusters Agriculture Planning

Discussion Appendix Bibliography List of Figures

..190

INTRODUCTION Deserts are expanding, yet the vast reserves of underground aquifers present a unique opportunity to envision sustainable models for inhabiting them. The Bahariya Oasis in Egypt exemplifies the coexistence of environmental systems on both the surface and the subterranean layers. This research proposes a prototypical subterranean settlement where groundwater is harnessed to reinvigorate the land and catalyse urban development sustained by local resources. A network of spaces and architectural systems is developed to dissect this challenge and translate it into realisable strategies. Self-sufficiency, social regulation, and water management form the core principles of the proposal. Self-sufficiency is pursued through reliance on local building materials and water resources, complemented by atmospheric water harvesting as a supplementary network designed to offset the openended extraction of the aquifer. Private and public life is regulated to preserve social cohesion and uphold the traditional practices of the local population. Water management strategies are embedded in every layer of the settlement through water generation, storage, and redistribution. Here, water flows dictate the organisation of life. Public cisterns serve as central social hubs, while water pools generated by atmospheric harvesting create invigorating atmospheres for dwellings and communal spaces. Treated wastewater sustains agricultural fields, which in turn supply both food and raw building material, supporting the settlement’s expansion. These cyclical processes form a self-regulating system for desert habitation, generated and sustained in collaboration with the subterranean currents beneath. Situated underground, the settlement anchors itself between the atmospheric flows above and the groundwater flows below, integrating both to establish a resilient and adaptive model for desert life.

Fig.02 Agriculture Field in The Bahariya Oasis

DESERT LANDSCAPES

1.1 Arid Regions Deserts and other drylands now occupy more than 30 % of Earth’s land surface and sustain the livelihoods of over one billion people1, making them the planet’s largest contiguous terrestrial biome. Although that share was fairly stable through much of the twentieth century, satellite records show that from 1982 to 2015 an additional 5.43 million km² of semi-arid land, about six per cent of the global dryland area, crossed the vegetation-loss threshold into active desertification, a shift driven primarily by anthropogenic climate change, accompanied by unsustainable land use.2

Andries Jan de Vries et al., “Breaking Rossby Waves Drive Extreme Precipitation in the World’s Arid Regions,” Communications Earth & Environment 5, no. 1 (2024): 493, https://doi.org/10.1038/s43247-024-01633-y. A. L. Burrell et al., “Anthropogenic Climate Change Has Driven over 5 Million Km2 of Drylands towards Desertification,” Nature Communications 11, no. 1 (2020): 3853, https://doi.org/10.1038/s41467-020-17710-7.

1

2

18 |Desert Landscapes

Fig.03 Aridity Index Classification

Desert Landscapes| 19

1.2 Increased Risk of Desertification Desertification, understood as the long-term loss of biological productivity in arid, semi-arid and dry subhumid zones, has moved from a distant ecological risk to an unfolding reality. The Mediterranean basin is widely recognised as a climate-change “hot-spot”, where rising temperatures and intensifying water stress render soils especially vulnerable to degradation and desert encroachment.3

Xuejie Gao and Filippo Giorgi, “Increased Aridity in the Mediterranean Region under Greenhouse Gas Forcing Estimated from High Resolution Simulations with a Regional Climate Model,” Global and Planetary Change 62 (June 2008): 195– 209, https://doi.org/10.1016/j.gloplacha.2008.02.002.

3

20 |Desert Landscapes

Fig.04 Risk of Human Induced Desertification

Desert Landscapes| 21

1.3 Potential of Ground Water Water scarcity has emerged as the defining constraint of habitability across arid and semi-arid regions. As shown in global vulnerability assessments, North Africa, the Middle East, and South Asia face the highest risk due to a combination of rising temperatures, declining precipitation, and overexploited aquifers, while southern Europe, western North America, and parts of South America are also increasingly exposed. 4 This uneven geography of vulnerability highlights the decisive role of water in shaping settlement, agriculture, and long-term resilience. Xuejie Gao and Filippo Giorgi, “Increased Aridity in the Mediterranean Region under Greenhouse Gas Forcing Estimated from High Resolution Simulations with a Regional Climate Model,” Global and Planetary Change 62 (June 2008): 195–209, https://doi.org/10.1016/j.gloplacha.2008.02.002.

4

22 |Desert Landscapes

Fig.05 Water Scarcity Vulnerabilty Classification

Desert Landscapes| 23

Against this backdrop, the Sahara, the world’s largest hot desert, encompassing roughly 9.4 million km², offers a natural laboratory for analysing enduring human adaptations to hyper-arid conditions and deriving settlement strategies capable of withstanding the heightened risks of desertification. Beneath this arid frontier, long-term habitation has depended on vast subterranean freshwater reserves. The transboundary Nubian Sandstone Aquifer System, spanning roughly 2.2 million km² across Egypt, Libya, Sudan, and Chad, stores on the order of 150,000 km³ of groundwater, while the North-Western Sahara Aquifer System) holds about 60 000 km³ yet receives < 1 km³ of modern recharge each year. These fossil stores have sustained oases and megainfrastructure such as Libya’s Great Man-Made River.5 Comparable reservoirs underpin other high-risk lands for desertification. The Guarani Aquifer System, beneath the USDA identified desertification vulnerable southern cone of South America, supplies 100 cities, but receives only 0.2 % annual replenishment. Consequently, settlements on expanding desert frontiers must couple aquifer replenishment strategies with architectural and socio-ecological practices that enable life under persistent aridity. Ahmed Mohamed et al., “The Groundwater Flow Behavior and the Recharge in the Nubian Sandstone Aquifer System during the Wet and Arid Periods,” Sustainability 14, no. 11 (2022): 11, https://doi.org/10.3390/su14116823.

5

24 |Desert Landscapes

Fig.06 Simplified Global Ground Water Resources

Desert Landscapes| 25

DOMAIN

2.1 Egypt; A Nation That Follows Water As the previous chapter highlights, water is the decisive determinant of habitability, even within the most extensive deserts. Egypt exemplifies this principle; its settlement pattern is governed almost entirely by the availability of surface and subterranean water, providing a living case study of sustained adaptation to dryland and heat. More than 95 per cent of the population is concentrated along the Nile Valley and Delta, where perennial irrigation has supported agriculture and urban growth since pharaonic times. Beyond this riparian corridor, habitation clusters around maritime water bodies, ports such as Alexandria and Port Said on the Mediterranean, and towns from Hurghada to Sharm elSheikh on the Red Sea coast.6

Manning, J. G. Water, Irrigation and Their Connection to State Power in Egypt. Working Paper, Economic Growth Center, Yale University, February 24, 2012.

6

Fig.07 Areial View of Cairo, Egypt

Fig.08 Populated land in Egypt

28 |Domain

Egypt’s renewable water availability has already slipped below the absolute scarcity benchmark of 500 m³ per capita per year. This decline is driven by rapid population growth, rising demand, and increasing pressures on the Nile, the country’s dominant but vulnerable water source. Agriculture consumes more than four-fifths of Egypt’s supply, leaving limited reserves for domestic and industrial use. Alternative sources such as desalination, treated wastewater, and deep groundwater remain marginal, while pollution and infrastructural strain compound the crisis. Together, these pressures situate Egypt as one of the world’s most water-stressed nations. Projections indicate that with a population surpassing 118 million by 2025, the gap between demand and supply will widen further. Coastal regions face salinisation and sea-level rise, while desert aquifers, though vast, are fossil reserves with little to no modern recharge. 7 Abd El-Wahed, Mohamed, Mohamed M. El-Horiny, Mahmoud Ashmawy, and Samar Abd El Kereem. “Multivariate Statistical Analysis and Structural Sovereignty for Geochemical Assessment and Groundwater Prevalence in Bahariya Oasis, Western Desert, Egypt.” Sustainability 14, no. 12 (2022): 6962. https://doi.org/10.3390/su14126962.

7

Fig.09 Water Collection Practice at a Communal Source

Fig.10 Water Consumption and Resources in Egypt

Domain| 29

2.2 The Untapped Potential of the Oases Among the archipelago of oases in Egypt’s Western Desert lies the Bahariya Oasis, spanning over 2,200 km². It is the closest major oasis to Egypt’s densely populated centres, situated approximately 350 km from Cairo and 300 km from Fayoum. HighIighted in the map, its strategic location places it directly on the axis of expansion planned by the Egyptian government for the Giza governorate, incorporating it into the national development strategies as part of desert reclamation and tourism initiatives. These conditions render the oasis an ideal site for testing the research hypothesis.8 8 General Authority for Investment and Free Zones (GAFI). “GAFI and Giza Governorate Join Forces to Promote Investment Opportunities.” February 23, 2025. https://www.investinegypt.gov.eg/English/NewsAndEvents/News/Pages/GAFIand-Giza-Governorate-Join-Forces-to-Promote-Investment-Opportunities.aspx.

30 |Domain

Fig.11 Egypt’s Map Highlighting Existing Oases and Major Cities

Domain| 31

2.2.1 The Nubian Sandstone Aquifer In addition to its accessibility, Bahariya is one of the key users of the Nubian Sandstone Aquifer System, (NSAS), the vast transboundary aquifer extending beneath Egypt, Libya, Sudan, and Chad. As illustrated in Figure 12, the NSAS underlies nearly two million square kilometres across four countries, making it the largest known fossil aquifer system in the world. Bahariya’s position at the northern margin of this immense reserve reinforces the oasis’s significance as both a vital water source for local communities and a focal point for research on water management in hyper-arid environments. Within Bahariya, around 811 shallow wells (up to 300 m deep) and 94 deep wells (reaching approximately 650 m deep) tap into this fossil groundwater reserve.9 This aquifer remains insufficiently documented, with much of the available data fragmented between geophysical surveys and limited well logs. Existing studies indicate that its morphology is shaped by a sequence of fractured limestone and Nubian sandstone layers. The shallow aquifer lies within fractured limestone at depths of 40 to 90 metres, while the deeper Nubian sandstone aquifer occurs between 800 and 1,200 metres, reaching thicknesses of around 250 metres. Flow patterns are strongly influenced by northeast- and northwest-trending fault systems, which act as conduits, most likely transferring water from the deeper Nubian sandstone into the shallower aquifer. This hydrological profile reinforces the oasis’s relevance to studies focused on water management in hyper-arid environments.10

Ali Hamdan and Rashad Sawires, “Hydrogeological Studies on the Nubian Sandstone Aquifer in El-Bahariya Oasis, Western Desert, Egypt,” Arabian Journal of Geosciences 6 (May 2011), https://doi.org/10.1007/s12517-011-0439-8. Taha Rabeh et al., “Structural Control of Hydrogeological Aquifers in the Bahariya Oasis, Western Desert, Egypt,” Geosciences Journal 22, no. 1 (2018): 145–54, https://doi.org/10.1007/s12303-016-0072-3. 9

10

Fig.12 Boundries of the Nubian Sandstone Aquifer

32 |Domain

Fig.13 Map of The Bahariya Oasis Highlighting Aquifer Qualities

Domain| 33

2.2.1 The Socio-Economic Structure of Bahariya

Fig.15 The English Volcanic Mountains, Bahariya

To evaluate the site’s potential, the research focuses on three main aspects: its socio-economic structure, climatic behaviour (solar exposure, wind, and humidity), and geological characteristics, including terrain, soil, and agricultural capacity. The current population of the oasis is estimated at around 34,000, a decline from approximately 50,000 in 1990. This demographic shift is closely tied to the retreat of agriculture; although 27% of the oasis land is suitable for cultivation, only around 10% is currently in use, largely due to water scarcity. 11 This underlines the urgency of developing sustainable methods for aquifer management, particularly as agriculture accounts for 40% of the local economy, with the remainder distributed among tourism, iron ore mining, and local services.

Fig.16 Al-Bawitie Village, Bahariya

The principal crop in the region is the date palm12, while wheat is cultivated seasonally according to annual needs in order to support self-sufficiency.13 11 Abdelhamid Elnaggar, Environmental Sensitivity to Desertification in Bahariya Oasis, Egypt, in Egyptian Soil Sci. Soc. J. (ESSSJ), vol. 16 (2014). 12 Amr Mohamed et al., “IRRIGATION WATER MANAGEMENT OF DATE PALM UNDER EL-BAHARIA OASIS CONDITIONS.,” Egyptian Journal of Soil Science 0, no. 0 (2017): 0–0, https://doi.org/10.21608/ejss.2017.1609.1123.22, no. 1 (2018): 145–54, https://doi. org/10.1007/s12303-016-0072-3. 13 Ahmed Abdalla et al., “Trends and Prospects of Change in Wheat Self-Sufficiency in Egypt,” Agriculture 13, no. 1 (2022): 7, https:// doi.org/10.3390/agriculture13010007. 14 Mohamed Abd El-Wahed et al., “Multivariate Statistical Analysis and Structural Sovereignty for Geochemical Assessment and Groundwater Prevalence in Bahariya Oasis, Western Desert, Egypt,” Sustainability 14, no. 12 (2022): 6962, https://doi.org/10.3390/ su14126962.

Fig.17Camping Sites along the Salt Mountains, Bahariya

Water and Settlement Patterns

Fig.14 The Economic Structure of Bahariya Oasis

34 |Domain

Using satellite images from Google Maps covering the period between 1985 and 2025, the research team identified active and abandoned wells, agricultural fields, and settlements across the Bahariya region. The data reveal that both agricultural activity and settlement patterns are closely tied to the accessibility of water provided by the aquifer through wells. These areas are concentrated predominantly in the northern part of the oasis, where the administrative centre of Bawiti is located, with smaller settlements in the east and south. Abandonment has occurred largely as a result of aquifer overexploitation for irrigation, which has led to water contamination through salinity and nitrate pollution.14

Fig.18 Settlement Analysis Map

Domain| 35

2.2.2 Bahariya Climate Profile Bahariya’s climate is characteristic of hyper-arid regions: Summer temperatures can reach 42°C, winter nights fall below 10°C, and diurnal variations of 20–25°C are typical. Rainfall is minimal, averaging just 4 mm annually. Humidity ranges from around 20% in the afternoon to over 50% before sunrise, and the area is subject to steady northerly winds at speeds of 3–5 m/s. 15 These climatic conditions inform the direction of the research, particularly in the development of water condensation towers and architectural typologies. 15

Elwan, A. “Quantitative Assessment of Desertification in the Bahariya Oasis, 1984–2017.” Journal of Planning & Development 2018.

Fig.19 Monthly Average Temperature and Humidity in Bahariya Oasis

36 |Domain

Fig.20 Predominant Wind Vector Direction Over the Bahariya Oasis

Domain| 37

Khamseen Winds The Bahariya Oasis, like much of Egypt, is seasonally affected by the Khamseen winds, a hot, dry, dust-laden phenomenon that typically occurs between March and May. These winds can reach speeds exceeding 40 km/h, dramatically reducing visibility and raising temperatures by as much as 10 °C within a few hours. In addition to transporting fine dust particles that accelerate erosion and infiltrate buildings, Khamseen events place significant stress on agricultural productivity and human health. Within the scope of this research, the Khamseen underscores the necessity of designing protective architectural and urban morphologies that mitigate dust infiltration, provide shaded and ventilated spaces during sudden temperature surges, and orient settlements to reduce wind-driven erosion. Strategies such as windbreak planting, buffer zones, and semi-underground construction become critical in ensuring resilience to these recurring climatic extremes. 15

Elwan, A. “Quantitative Assessment of Desertification in the Bahariya Oasis, 1984–2017.” Journal of Planning & Development 2018.

Fig.21 Sand Storm over Al-Bawitie Village, Bahariya

Fig.22 Sand Storm over Agriculture Fields, Egypt

Fig.23 Effect of Sandstorms on Urban Landscape; Cairo , Egypt

38 |Domain

Fig.24 Khamseen Wind Vector Direction Over the Bahariya Oasis

Domain| 39

2.2.3 Topographic Context of Bahariya Topographically, the oasis is a broad desert basin with its floor sitting near 73 m above sea level at Al-Qasr (NorthWest), while surrounding mesas and escarpments rise to roughly 350 m above sea level with almost 300 m of relief.17 The analysis of this topography was carried out using available DEM files for the Bahariya Oasis, processed in QGIS to generate a mesh model for further examination. The landscape’s runoff channels inform the research on where water production and consumption should be located in order to create a gravity-led hydrological system. The stepped inner slopes within the oasis provide solid, shaded ledges for the construction of new morphologies that would benefit from being partially underground by borrowing the surrounding rock’s natural 8–10 °C temperature buffer. Taken together, Bahariya’s location, climate and landform create a realistic proving ground for a “closed-loop” desert settlement, one that balances groundwater use with passive water harvesting and earth-sheltered construction. Success here would point the way for many other dryland fronts where fossil aquifers are the last buffer against advancing desertification.

15

Elnaggar et al., “Soil Classification Of Bahariya Oasis Using Remote Sensing And Gis Techniques.”

Fig.25 Mountain Valleys in Bahariya

Fig.26 Surface Water Within Low Terrain Areas, Bahariya

40 |Domain

Fig.27 The White Desert; Salt Mountains in Bahariya

Low Terrain

High Terrain

Fig.28 Topography Analysis on QGIS Mesh of Bahariya’s Landscape

Domain| 41

Fig.29 Run-Off Analysis on QGIS Mesh of Bahariya’s Landscape

42 |Domain

Low Concavity

High Concavity

Fig.30 Concavity Analysis on QGIS Mesh of Bahariya’s Landscape

Domain| 43

2.3 Understanding Desert Living To situate the research within its contextual foundations, the team examined both vernacular and modern practices relevant to life in hyper-arid environments. These were broadly categorised into four themes: urban methodologies, water management practices, architectural morphologies and construction methods. Across all categories, recurring concerns emerge, including the improvement of thermal comfort in indoor and outdoor spaces, the efficient use of local resources, and the calibration of water management for consumption, cooling and irrigation. Other themes address social dimensions, such as economic practices (most prominently agriculture) and architectural traditions that sustain a balance between private and public life. This contextual framework provides the research with essential tools for envisioning a prototypical settlement, which are then enhanced through modern computational methods to achieve greater efficiency.

8 General Authority for Investment and Free Zones (GAFI). “GAFI and Giza Governorate Join Forces to Promote Investment Opportunities.” February 23, 2025. https://www.investinegypt.gov.eg/English/NewsAndEvents/News/Pages/GAFIand-Giza-Governorate-Join-Forces-to-Promote-Investment-Opportunities.aspx.

44 |Domain

Fig.31 Redrawn plan of Hassan Fathy’s “Hasan Rashad House, Tanta, Egypt.”

Domain| 45

2.3.1 Urban Methodologies New Gourna, Hassan Fathy Hassan Fathy’s mid-twentieth-century schemes of New Gourna and New Baris stand as landmark attempts to modernise rural Egyptian villages while preserving vernacular construction techniques and indigenous social organisation. Each plan is ordered around a hierarchical network of roads and courtyards sized for distinct functions. In New Gourna, primary roads limited to about 10 m in width divide the settlement into three neighbourhood sectors that converge at the central public square. A secondary grid of lanes, roughly 3 to 4 m wide, links communal courtyards allocated to individual badana (extended-kin groups) and serves as a semi-public realm for clusters of one- and two-storey houses. At the most intimate scale, every dwelling contains its private courtyard, establishing a graded transition from domestic privacy to communal spaces to public life.18 18

Fig.33 New Baris Village in New Valley, Egypt

Damluji and Bertini, Hassan Fathy: Earth & Utopia, pp. 197,198, 226

New Baris, Hassan Fathy In New Baris, Fathy codified and extended the spatial logic first tested at New Gourna. Secondary lanes were limited to 300 m or less to reinforce neighbourhood cohesion, while their orientation aligned with the prevailing northwest winds to enhance thermal comfort. The primary road network followed the site’s natural undulations and culminated in a raised intersection that accommodated a consolidated administrative centre. Functions that had remained scattered in New Gourna (social-service offices, a school, a mosque, and housing for state officials) were assembled into a compact complex. These complexes were connected through a sequence of courtyards and alleys, varied in plan and section to maximise shading and cross-ventilation. 19 The design thus evolved from simply accommodating social needs to actively regulating the settlement’s microclimate, while refining pathways and communal spaces to minimise inhabitants’ exposure to the harsh environmental conditions. Fig.32 New Gourna Village in Luxor, Egypt

46 |Domain

19

Damluji and Bertini, Hassan Fathy: Earth & Utopia, pp. 228-230

Fig.34 New Gourna Village in Luxor, Egypt

Fig.35 New Baris Village in New Valley, Egypt

Domain| 47

Historical City of Yazd, Iran Yazd, Iran, offers a significant precedent of an urban environment that responds to environmental conditions similar to those of the Egyptian desert. The city is characterised by an extremely compact urban fabric connected by narrow alleyways, distinguished by wider motorways. These tortuous alleys are designed to take full advantage of prevailing wind orientations in order to generate air currents. In some cases, they are roofed with sabats (shown in Figure 38), which shield inhabitants from solar radiation and sandstorms (studies indicate a reduction of up to ~80% in direct solar radiation)19. Building orientation further enhances shading across open areas of the city. Internal courtyards function as primary exhaust spaces, and by incorporating cooling pools (covering 5–15% of their area) and vegetation (15–20%), they provide cooling, shading, humidity, and dust reduction, effectively creating a “green belt” within the urban fabric. Taken together, Yazd’s urban environment demonstrates how dense urban form, combined with integrated water features and green spaces, can generate protective microclimates for its inhabitants.20 Sadra Sahebzadeh et al., “Sustainability Features of Iran’s Vernacular Architecture: A Comparative Study between the Architecture of Hot–Arid and Hot–Arid–Windy Regions,” Sustainability 9, no. 5 (2017): 749, https://doi.org/10.3390/su9050749. Akram Ahmed Noman Alabsi et al., “Towards Climate Adaptation in Cities: Indicators of the Sustainable Climate-Adaptive Urban Fabric of Traditional Cities in West Asia,” Applied Sciences 11, no. 21 (2021): 10428, https://doi.org/10.3390/app112110428.

Fig.36 Aerial View of Yazd, Iran

Fig.37 A Close-up of Houses in Yazd, Iran

Fig.38 Narrow, Covered Alleyways in Yazd, Iran

19

20

48 |Domain

Green Belt

Climatic Orientation

Narrow Alleyways

Internal Courtyards Yazd, Iran

Compact Urban Fabric

Fig.39 Urban Analysis of Yazd, Iran

Domain| 49

2.3.2 Water Management Vernacular Methods of Water Management Water management and equitable distribution are critical issues for the survival of desert settlements. In Iran, and particularly in cities such as Yazd, the primary system of water supply since antiquity has been the qanat. Qanats are subterranean networks of gently sloping tunnels that channel groundwater from aquifers or mountain sources to lower, flatter lands, where they irrigate fields and sustain urban populations. This water network runs from 10-30 m below ground, thus delivering water at temperatures 5–7 °C cooler than ambient air.21 Operating entirely by gravity, qanats bring water to the surface without mechanical pumping; however, this also means that their discharge is subject to seasonal and hydrological fluctuations, leading to inconsistent flow rates. Despite this limitation, the qanat has long been recognised as one of the most sustainable technologies for securing water in arid environments.22 To counteract this inconsistent flow, water from that system is selectively stored in Ab-Anbars, subterranean cylindrical or polygonal cisterns covered by a dome (10–20 m deep and 300–3000 m³ in volume), lined with 2 m-thick brickwork and waterproof mortar that maintains potable water year-round in low temperatures and prevents its evaporation. 23

Fig.40 A Qanat Interior

Shakibamanesh, “Assessing the Value of Qanat System of Yazd in Promoting Urban Climate Resilience,” pp. 113-124. Ali Hamidian, Mehdi Ghorbani, Mahsa Abdolshahnejad, and Aziz Abdolshahnejad, “Qanat, Traditional Eco-Technology for Irrigation and Water Management,” Agriculture and Agricultural Science Procedia 4 (2015): 119–125, https://doi.org/10.1016/j. aaspro.2015.03.014. 23 Yousefi & Nocera, “The Role of Ab-Anbars…”, pp. 3987-4000 21

22

Fig.41 An Ab-anbar in the Central Desert City of Naeen, Iran

50 |Domain

Fig.42 The Canals of A Persian Aqueduct System in Iran

Wind Tower

Entrance

Water Storage Tank Fig.43 A Study of the Ab-Anbar System

Qanat Network

Distribution Network

Aquifer

Fig.44 Qanat System: Water Distribution from Aquifer to Household

Domain| 51

Modern Methods of Water Management More contemporary examples of water storage in arid regions include the Down to Earth project by Ruth Kedar, designed for communities in the Negev Desert. (Figure 47) The proposal developed an underground network of interconnected cisterns integrated into the foundations of houses. Each cistern is modular, constructed from precast concrete, and equipped with overflow valves that redirect surplus water to reservoirs at lower elevation points. In this way, a distributed and expandable water reserve is created, one that grows incrementally as new units are added and flexibly increases the settlement’s overall water capacity.24 24

Out of Water: Design Solutions for Arid Regions, 86–89.

Fig.45 Wadi-Hanifeh Water Distribution Project by Thomson Consultants

Significant efforts have been made to develop new methods of water generation, with atmospheric harvesting offering a largely underexplored potential in arid environments. The VENA project, designed by Ore Design + Technology, proposes a system of decentralised water-harvesting stations that rely on the temperature differential between hot daytime air and cooler subsurface conditions to trigger condensation at the dew point. Copper alloy filaments connected to an underground cistern capture airborne moisture, condense it, and channel the water downward without mechanical energy. This design is grounded in the fact that even desert air contains usable humidity; for example, one cubic kilometre of air at 50% relative humidity and 30 °C contains roughly 15,000 tons of water vapour. In practice, condensation structures such as VENA can harvest between 0.3 and 1.5 litres of water per square metre of condensing surface per night, depending on humidity, wind conditions, and surface material. While modest compared to large-scale aquifer extraction, such yields are scalable and independent of groundwater reserves, making them invaluable supplements in desert contexts.25 25

Out of Water: Design Solutions for Arid Regions, 78–82.

52 |Domain

Fig.46 Vena 1 Project by ORE, Copper-Alloy Water Condensation System

Finally, beyond water generation, distribution and storage, water budgeting and greywater treatment are essential for resilient water management in arid regions. Rather than treating water as a single-use commodity, water budgeting allocates resources strategically across a spectrum of qualities and needs: high-quality potable water is reserved for drinking and cooking, while treated effluent, brackish water, or greywater is redirected to agriculture, irrigation, or industrial uses. Projects such as the Hiriya Landfill Recycling Park demonstrate how stabilisation ponds and constructed wetlands enable communities to reclaim wastewater, with vegetation gradually transforming these facilities into public green spaces. Such cascading water cycles effectively multiply the availability of water within a settlement.26 26

Out of Water: Design Solutions for Arid Regions, 163-177

Fig.47 Schematic Diagram of Dow To Earth Cistern System

Fig.48 Schematic Diagraming of Cyclical Water Systems

Domain| 53

2.3.3 Architectural Morphologies Traditional Spatial Organzation in Egypt Architectural morphologies in Egypt’s vernacular environments centre on protecting inhabitants from harsh external conditions while upholding core societal values. Domestic architecture organises dwellings into distinct guest, family-living, and service zones. At the street edge lies the public reception room, reached through a kinked entrance corridor or separate doorway to prevent direct views into the house. (shown in Figure 51) High-level vents or small forecourts release hot air while maintaining privacy, while shaded courtyards at the centre provide light and ventilation. The sequence concludes in a service yard featuring a kitchen, store, date press, and animal stalls, all accessed separately via a side lane. Circulation progresses step by step from public to semi-private to service space, with units connected through narrow, shaded alleys that support daily movement.27 27

Fig.49 House in Dakheh Oasis, Egypt

Bassily and Refaat, “The Features and Characteristics of Desert Societies,” 270.

The traditional settlements of the Dakhleh Oasis provide additional information on methods of regulating thermal comfort and programmatic requirements. Courtyards paired with cross-ventilation shafts and perforated staircases channel fresh air through the houses, reducing heat stress. Seasonal rooms further enhance adaptability: thickwalled ground-floor “summer” rooms remain cool, whereas rooftop terraces and lighter “winter” rooms capture sunlight and breezes, offering a flexible response to the extreme seasonal variations of the desert climate. Housing units were organised in darb or hara clusters, semi-public alleys that housed extended families and created autonomous micro-communities. These develop a gradual access from private domestic courtyards to public alleys and communal facilities.28 Mashrabiyas provide an added layer of privacy and environmental comfort. These wooden lattice screens (shown in Figure 50), projecting from upper stories, filter daylight and airflow while shielding interiors from public view. Their design promotes passive ventilation, lowering indoor temperatures by up to 2.4 °C when opened compared to when closed.29 Fig.50 Wooden Lattice Screens (Mahsrabiyas)

54 |Domain

Francesca De Filippi, Traditional Architecture in the Dakhleh Oasis, Egypt: Space, Form and Building Systems, 2006. Abdullah Abdulhameed Bagasi et al., “Evaluation of the Integration of the Traditional Architectural Element Mashrabiya into the Ventilation Strategy for Buildings in Hot Climates,” Energies 14, no. 3 (2021): 530, https://doi.org/10.3390/en14030530. 28 29

Interior Courtyard

Kitchen and Dining

Bedroom

Living Room

Reception

First Floor

Second Floor

Fig.51 Floor Plans for a House in Dakhleh Oasis

Domain| 55

Enviromental Vernacular Techniques Adding to the established practices, structures that develop vertically are also employed. Hassan Fathy’s experiments in New Gourna and New Baris used vaulted and domed spaces not only as structurally efficient forms but also as climatic regulators. Vaults facilitated the upward movement of hot air to high vents, while domes diffused daylight and reduced direct solar gain, keeping interiors cooler by day and warmer by night. Fathy also used these forms to distinguish functions: guest and reception rooms were marked by higher domes for importance and ventilation, while service or storage spaces relied on barrel vaults that trapped less heat. 30 A similar sectional logic appears in Yazd, where wind towers drive air down shafts to displace warm air and cool cisterns (Figure 52 & 56), and subterranean shavadans (Figure 55) provide thermally stable chambers sunk 5–7 m into the earth, where soil inertia maintains 22–25 °C. Sunken courtyards, lowered below street level, access qanat water and employ evaporative cooling to temper surrounding spaces. 31 Hassan Fathy, Architecture for the Poor: An Experiment in Rural Egypt (Chicago: University of Chicago Press, 1973). Elisabeth Beazley and Michael Harverson, Living with the Desert: Working Buildings of the Iranian Plateau (Warminster: Aris & Phillips, 1982).

30 31

Fig.53 Hassan Fathy’s Vaulted Structures, New Baris

56 |Domain

Fig.52 Wind towers in Yazd, Iran.

Fig.54 Hassan Fathy’s Domed Housing, New Gourna

Fig.55 Section of Shavadan, Iran

Fig.56 Section of Wind Tower, Iran

Fig.57 Section of Hassan Fathy’s Housing, New Gourna

Domain| 57

2.4 Construction Methodologies The construction traditions of desert oases reveal a constant negotiation between environmental constraints, material availability, and structural performance. From the enduring use of mud brick to the more recent dominance of concrete, building practices in Bahariya illustrate both the strengths and shortcomings of vernacular and modern approaches. This section examines material context, fabrication techniques, and structural strategies that inform desert construction, with particular focus on the potential of functionally graded earthen composites, on-site production methods, and vault-based systems. Together, these investigations frame the search for an alternative construction paradigm that balances ecological responsibility with structural resilience in arid environments.

58 |Domain

Fig.58 Mud Brick Buildings in Siwa Oasis, Egypt

Domain| 59

2.4 Construction Methodologies Material Context The extreme conditions of desert oases demand adaptive construction strategies. Across the Sahara, vernacular architecture mitigated these stresses through passive design and material economy. Mud brick remains the traditional material. Hand-cast from local subsoil with fibre additions is inexpensive, low in embodied energy, and circularly sustainable: walls can be maintained through re-mudding and eventually rehydrated and reused. Yet mud brick is structurally weak, prone to cracking, and highly water-absorbent, limiting durability. 32 These shortcomings partly explain the mid-twentiethcentury shift toward modern construction. In Bahariya today, concrete dominates, supported by Egypt’s cement industry. It provides strength and rapid buildability, yet carries major drawbacks in this context: very high embodied energy, poor thermal performance due to high conductivity, and incompatibility with passive cooling strategies. Concrete dwellings often overheat and require mechanical cooling. 33 This juxtaposition frames the project: mud brick is climateresponsive but fragile, while concrete is robust but environmentally and thermally unsuitable. The challenge is to develop an alternative system that retains the lowcarbon advantages of earth while enhancing durability, water resistance, and structural performance. 32 33

Fig.59 Mud-brick Drying Field

Gernot Minke, Building with Earth. Mora-Ruiz et al., ‘Sustainable Earthen Construction’.

Fig.58 Brickmaking in the Village of New Gourna

60 |Domain

Fig.60 Excavated Mud-Brick cells, Tal Ganoub Qasr al-’Aguz, Bahariya Oasis (4th–7th c. CE)

Fig.61 Material Systems in Bahariya Oasis: Vernacular Mud Brick vs. Imported Concrete

Domain| 61

Nubian Vault Mesopotamia & Ancient Egypt

Prefabricated Bricks Eastern Mediterranean

Sail Vault Mesopotamia & Iran

Spherical Domes Ancient Rome

Barrel Vault on Beams Iraq & Egypt

Corbelled Domes Libya

Mudbrick & Stone Vault North Africa

Rib Domes South & Central Asia Iberian Peninsula

Vault Construction Vaults play a key role in desert architecture, serving both as a climate-responsive feature and a structural system ideally suited for earthen blocks. These masonry forms direct loads mainly through compression, which is suitable for soil-based materials. To ensure stability, their shape must follow a funicular curve, such as the inverted catenary, keeping thrust lines within the masonry. This way, loads are transferred in pure compression, preventing tensile stresses, cracks, or collapse. Historical precedents support this reasoning. The Nubian vault, revived by Hassan Fathy, demonstrated how unreinforced mud bricks can span space along a catenary curve. Iranian and Mesopotamian builders achieved similar forms without timber centring, while Catalan vaults proved that if the geometry is funicular, the material used can be minimal. 39 Recent research and practice expand these principles through digital and material innovation. While traditional methods often eliminate formwork, contemporary approaches usually reintroduce modular, lightweight, and reusable scaffolds to ensure safety and accuracy in more complex geometries. The ETH Zürich Block Research Group, for instance, showed how thin-tile vaults could be constructed with minimal waste using recyclable cardboard formwork, combining vernacular efficiency with digital precision.40 Such precedents highlight hybrid approaches that blend proven geometric logic with lowtech yet adaptable construction aids. 39 40

Al Asali, ‘Vaulting Cultures in the Modern Middle East’. Block Research Group, Beyond Bending.

62 |Domain

Fig.62 Comparative Taxonomy of Earthen and Masonry Vault–dome Typologies

Fig.63 Anatomy and nomenclature of centering

Fig.64 Anatomy and Nomenclature of a Masonry Arch

Domain| 63

Functionally Graded Materials Functionally Graded Materials (FGMs) are composites whose properties vary gradually within a single element, integrating strength, insulation, and durability seamlessly. Unlike layered systems, FGMs distribute properties where they are most effective, avoiding weak interfaces and improving resilience. For earthen construction, FGMs offer a way to overcome soil’s brittleness and moisture absorption while retaining compressive strength and thermal inertia. Variation within each unit can combine dense, stabilised, water-resistant zones with lighter insulating cores, reducing reliance on coatings or discrete layers. Functionally graded cementitious materials are designed with a through thickness variation in mix proportions so that properties are placed where they are most effective. Typical objectives are to improve flexural performance, crack control and durability by locating stiffer or more rigid material in high-demand zones. Common strategies include varying the water-cement ratio, aggregate type or size, and fibre volume across depth. Production is usually by sequential fresh on fresh casting with careful consolidation to maintain bond and avoid weak planes, or by controlled settling that yields a continuous gradient rather than a complex layer interface. Reviews report significant gains in bending capacity and serviceability, and reduced ingress when low-permeability mixes are placed near exposed faces. Kostas Grigoriadis, ed., Mixed Matters: A Multi-Material Design Compendium. Torelli et al., ‘Functionally Graded Concrete: Design objectives, production techniques and analysis methods for layered and continuously graded elements’. 34

35

64 |Domain

Fig.65 Vertical Functional Gradation for Earthen Masonry

Fig.66 Functionally Graded Strategies For Earthen Composites Under Bending

Domain| 65

On-Site Fabrication One of the design team’s core principles is to build desert architecture using local materials, rather than importing energy-intensive ones. Bahariya’s sandy-silt soils are ideal for construction, reflecting traditional methods of casting bricks from local earth and often reinforcing them with fibres from oasis agriculture. Vernacular practices also introduced natural additives to improve water resistance, aligning with broader research trends in bio-stabilisers and agricultural by-products that enhance soil performance, reduce carbon footprint, and lower embodied energy. Historically, oasis builders in the Sahara recycled old bricks and soils to support a regenerative building cycle. Adaptive formwork research provides further precedent. Interlocking masonry units with varied profiles can be produced using reconfigurable modular systems, reducing waste and cost. Fig 64. shows the dry-masonry interlocking blocks designed for V-INCA project designed dry-masonry interlocking blocks and a four-unit mold kit that yields multiple block types from identical inserts, enabling rapid, mortarless assembly, an example that is directly aligned with a reconfigurable formwork strategy for onsite casting.37 Circularity is central to the approach. Cast, unfired blocks have a low carbon footprint and are recyclable; these stabilised units can be remoulded, while formwork and scaffolding are designed for repeated use. This echoes historical examples of communal brickmaking yards in desert settlements, now expanded by modern fabrication tools that improve efficiency while retaining resourceconscious methods. Overall, the literature and precedents support the idea that locally sourced, on-site production can significantly cut the carbon footprint while enabling the creation of complex components. Simple methods produce complex architecture in direct dialogue with the landscape.

36 37 38

Gernot Minke, Building with Earth. Yagmur Yenice and Daekwon Park, “V-INCA: Designing a Smart Geometric Configuration for Dry-Masonry Wall,”. Gernot Minke, Building with Earth.

66 |Domain

Fig.67 V-INCA Block Family Generated From A Four-Unit Configurable Mould

Fig.68 V-INCA Four-Unit Mould, Casting Sequence For 1:2 Prototypes

Fig.69 Earth Material Building Cycle - A Closed-Loop Workflow For Cast Earthen Masonry

Domain| 67

Conclusion The domain chapter of this thesis is divided into two parts. The first addresses the significance of the Bahariya Oasis as a case study. While the Egyptian government’s growing interest in its development confirms this importance, for the research team, the oasis also represents a prime example of a unique combination of parameters that can be observed in other parts of the world. The discussion above highlights the region’s harsh environmental conditions, its vast potential for harnessing local resources, and the mismanagement that has marked its recent history. The second part focuses on the strategies explored by the design team to promote development in the region. These draw heavily on vernacular traditions, as the team believes that long-tested techniques provide invaluable insight into sustainable solutions for the oasis. At the same time, this approach must respect and adapt to the social fabric of the region, which has been extensively studied to understand its characteristics and incorporate them into the design process. The following chapters present the research team’s attempts to translate these opportunities, principles and limitations into a development plan for a new prototypical settlement. Computation, material systems, water management practices and architectural design are brought together to harness these principles and merge them into a coherent system. Fig.70 Agriculture Field in The Bahariya Oasis

RESEARCH METHODOLGY

3.1Urban Systems and Networks 1. Quantum Geographic Information System (QGIS) QGIS was employed to generate a detailed topographic model of the Bahariya Oasis using a Digital Elevation Model (DEM) sourced from the United States Geological Survey (USGS). This DEM enabled the extraction of elevation data, which was subsequently used to construct a spatial mesh of the region. The mesh informed the identification of terrain variation, slope analysis, and low-lying zones critical to aquifer accessibility and gravitational water flow. This topographic groundwork established the base layer for all subsequent geospatial and environmental analyses within the research. Fig.71 Digital Elevation Model of Bahariya Oasis

2. Machine Learning and Predictive Modelling In the context of researching the local aquifer, a predictive model was employed to address the scarcity of available data concerning aquifer characteristics such as shallowness and depth. A neural network was trained using measurement points extracted from existing literature; these points included both their spatial coordinates and corresponding aquifer properties. The trained model was then used to approximate the characteristics of the aquifer beneath the site. In parallel, another neural network was developed to predict airflow behaviour within the solar updraft tower. Based on Computational Fluid Dynamics (CFD) simulations, 50 representative tower configurations were tested, and the resulting airflow velocities were used to train the model, which then extrapolated air speed predictions for the 800 parametric designs generated. This dual modelling approach produced mapping data that informed the site selection process and enabled rapid performance assessment of design variations, offering a transferable methodology for other regions with limited datasets. Fig.72 Aquifer Analysis Through Pedective Modelling

72 |Research Methodolgies

3. Site Allocation through Cellular Automata Building upon the site suitability outputs from earlier hydrological and geological analysis, the settlement area was discretised into 40 × 40 metre cells to test spatial organisation strategies using a cellular automata model. The model, inspired by Conway’s Game of Life, was customised to simulate programmatic logic across four states: void, cistern, housing, and agriculture. Transition rules were encoded to reflect functional dependencies, such as gravitational water flow from wells to cisterns to homes, as well as clustering patterns that promote spatial coherence and infrastructural efficiency. This approach allows for a generative exploration of possible settlement morphologies under both environmental and programmatic constraints. Fig.73 Cellular Automata Organization of a 1200 People Settlement

4. Shortest Walk The shortest walk is a computational method used to calculate the most efficient path between two or more points within a spatial environment. It typically employs graph theory algorithms such as Dijkstra’s or A*, which identify the shortest path based on distance, cost, or time, accounting for constraints like obstacles or boundaries. In this project, shortest walk analysis was employed to optimise both the road network and the hydro-social system by evaluating the most efficient routes between water infrastructure, housing clusters, and communal spaces.

Fig.74 Primary Water Network Generated through Shortest Walk

By simulating shortest paths within the settlement’s grid, the method enabled the design team to minimise travel distances, ensure accessibility to essential resources, and define water distribution routes based on elevation and shortest distance. Implemented through Grasshopper, the tool accounted for changes in terrain and spatial hierarchy, aligning both pedestrian and water networks with gravitational flow constraints.

5. Rhino Ecologic Rhino Ecologic is an ecological simulation framework for Rhino and Grasshopper that integrates environmental knowledge into the architectural design process. It allows users to model and analyse ecological performance in relation to architectural projects. In this study, it was used to calculate and predict the projected biomass of agricultural plots with varying morphological attributes, enabling comparative performance assessments. These assessments incorporated solar, topographical and environmental parameters. Focusing on species native to Egypt (principally palm trees and wheat), the software was further employed to evaluate optimal distribution patterns within designated plantation areas. Fig.75 Rhino Ecologic Model of an Agriculture Plot

Research Methodolgies| 73

3.2 Architectural Development and Analysis 1. Computational Fluid Dynamics (CFD) Simulations Computational Fluid Dynamics (CFD) simulations were employed to evaluate airflow behaviour and volumetric mass flow within the proposed solar updraft tower configurations. A parametric model was developed in Grasshopper to generate geometries varying in height, taper, inlet geometry, and internal bypass ratio. Selected configurations were tested under standard conditions of 5 m/s prevailing wind and a 12 °C temperature differential, representing the offset required for condensation. CFD analysis allowed for precise modelling of airflow velocity, pressure differentials, and volumetric flow rates (Q), which were subsequently used to calculate potential water yields. Fig.76 CFD Analysis of Atmospheric Water Haevesting Tower

2. Environmental Simulations Environmental analysis in this research focused on assessing and enhancing thermal comfort through the use of computational tools. Ladybug was employed to evaluate external thermal comfort conditions across the settlement, particularly in the spaces between housing clusters. This analysis informed the spatial arrangement of buildings, shading strategies, and microclimatic conditions at the urban scale. Honeybee, by contrast, was applied to the solar updraft water-harvesting tower. The tool enabled the calculation of interior temperatures based on heat gain both above and below ground, while also allowing for the allocation of material properties to the building envelope. This capability was critical for evaluating the thermal behaviour of the tower and its integration with the proposed material system. Together, these tools played a central role in improving environmental performance, supporting energy efficiency, occupant comfort, and the effectiveness of the water-harvesting strategy. Fig.77 Lady Bug Simulation of Incident Radiation on Housing Plot

74 |Research Methodolgies

3. Finite Element Analysis Finite Element Analysis (FEA) was employed to assess the structural performance of architectural and infrastructural components within the project. Using Karamba3D, the research team simulated force distributions, principal stress lines, and areas of stress concentration across different morphologies. This analysis enabled the identification of structurally vulnerable zones, informing the integration of reinforcement strategies and material gradation. The simulation outputs provided critical insight into how the geometry could be optimised to dissipate loads efficiently, particularly in components fabricated from earthen materials with variable strength. Fig.78 Karamba 3D Deflection Analysis on Vaulted Roof System

4. Multi-Objective Evolutionary Optimisation Multi-objective evolutionary optimisation, implemented through Wallacei, was utilised to evaluate competing performance criteria across architectural and infrastructural systems. The algorithm facilitated the exploration of complex design spaces by generating and iterating through a population of morphological variations. For the condensation tower, this involved balancing thermal performance, airflow velocity, structural stability, and material feasibility to identify the most efficient form for atmospheric water harvesting. At the architectural scale, Wallacei was used to evaluate housing cluster configurations, optimising for shading, spatial hierarchy, and thermal comfort. In both cases, the algorithm supported decision-making by revealing trade-offs and synergies between design objectives, enabling data-driven selection of optimal configurations under climate and construction constraints. Fig.79 Examples of Generated Housing Typologies through Wallacie

Research Methodolgies| 75

3.3 Material Research and Prototyping 1. Form-finding and discrete equilibrium Vault geometries were established whose internal thrust lines remain

with Thrust Network Analysis (TNA) to obtain funicular within the masonry thickness under self-weight and service

surfaces loading.

The resulting compression-only envelopes were then discretised into interlocking brick tessellations and validated through dynamic relaxation with Kangaroo, modelling contact, friction, and staged assembly. In combination, TNA provided the global equilibrium target, whilst Kangaroo tested the localised block-to-block equilibrium and the robustness of the topological interlock, including checks for sliding and hinge formation.

Fig.80 Vault equilibrium simulation with Kangaroo dynamic relaxation.

76 |Research Methodolgies

Fig.81 TNA for overall form, horizontal forces, and vertical forces

2. Physical Prototyping Finally, blocks.

reusable Segmental

additive-manufactured formwork and vaults were erected over modular

manual casting timber centring,

produced functionally validating the digital

graded pipeline.

The physical tests included compression testing with applied weights, thermal transference assessed through a hot plate and ice-melting experiments on the fabricated blocks, and hydrophobic performance evaluated through water submersion. This process confirmed the feasibility of the digital-to-physical workflow, which consisted of three stages: (i) mix design and physical tests, (ii) functionally graded block casting, and (iii) interlocking assembly.

Fig.82 Thermal Testing Of Block Samples Using Hot Plate Setup

Fig.83 Hydrophobicity Testing Through Water Immersion

Fig.84 Compression Testing of Sample Block Under Applied Weights

Research Methodolgies| 77

RESEARCH DEVELOPMENT

INTRODUCTION The research development engages both the macro and micro scales of the settlement as a framework for the following chapter, the design development. This dual focus ensures that the architectural morphologies are firmly anchored in the context of the Bahariya Oasis. The chapter opens with an analysis and modelling of the Nubian aquifer. As outlined in the domain chapter, the scarcity of comprehensive data prompted the design team to collate scattered information and reconstruct a digital model of the aquifer. Complementing this, a series of site analyses were conducted through a grid model and using a point-based system, the most suitable location for the settlement was identified. Water consumption and production data, combined with topographical information, then informed the horizontal organisation of the settlement. Vertical expansion, in turn, initiated the development of material and construction technologies, using soil expected to be excavated on site. This material was tested and enhanced to increase its structural strength and water resistance, ultimately resulting in a multipurpose construction medium adaptable to both wall and vault typologies. Taken together, these experiments span the urban and material scales, proposing a comprehensive system applicable to similar endeavours in comparable climates. Fig.85 Material Samples

4.1 Site Selection and Evaluation The site selection and evaluation process synthesises aquifer, soil, and contextual analyses to determine the most suitable locations for settlement within the Bahariya Oasis. Using a unified 1 km grid framework, data on aquifer shallowness and depth, soil stability, and the spatial distribution of agricultural land, settlements, and abandoned plots were systematically mapped and assessed. These parameters were then weighted according to their relative importance, with water accessibility and sociospatial context identified as the most critical factors. The resulting evaluation highlights key northern and southern areas as the most viable options, ultimately guiding the selection of a site near El Bawiti for its balance of environmental resilience and proximity to existing settlement networks.

82 |Research Development

Fig.86 Topographical Map of Bahariya Oasis

Research Development| 83

4.1.1 Aquifer Analysis and Predictive Modelling Given the scarcity of data on the hydrological character of the Nubian aquifer, the design team initiated an experiment to map its morphology, recognising it as the most critical resource for launching the design process. The assessment focused on both the aquifer’s proximity to the ground surface and its overall depth beneath the oasis. The aim was to identify the most suitable locations for establishing a settlement capable of securing sufficient water volumes. In this context, the shallowness of the aquifer was assessed to determine the accessibility of water reservoirs, while its depth was evaluated in relation to the long-term capacity for extraction. The initial stage of the experiment involved mapping the elevation points across the entire oasis. This was achieved by obtaining a digital elevation model from the United States Geological Survey.40 This data enabled the research team to assess the morphological characteristics of the area, which is situated within a depression and features locally elevated areas throughout its landscape, although it otherwise exhibits a relatively gentle gradient.

Altogether, twenty-two points were mapped onto the constructed grid, together with their respective elevation and aquifer parameters. This dataset formed the basis for the analysis of the Nubian Aquifer in the Bahariya region through a predictive model of aquifer depth and accessibility applied across the entire grid. This model utilised a resilient backpropagation algorithm to train a neural network, using the x and y coordinates of each measurement point as inputs, and the aquifer’s depth and shallowness as outputs. Through this approach, the aquifer properties of the entire grid were estimated, resulting in two distinct maps that delineate aquifer depth and accessibility, respectively.

Following the collection of elevation points across the region, an analysis of the Nubian Aquifer was undertaken. The available literature concerning the Nubian Aquifer is limited and fragmented, primarily relying on data from well production in the area, as well as geoelectric surveys carried out in selected locations. To provide an accurate assessment of the aquifer’s morphological conditions, an estimation of its characteristics was conducted using an approximate 1x1 km grid. The research concentrated on the shallowest aquifer in the region, composed mainly of fractured limestone, as it is the most accessible for water extraction. The analysis was informed by two key studies: the first by the National Research Institute of Astronomy and Geophysics in collaboration with the National Water Research Centre of Egypt, and the second by the Geology Department, Faculty of Science, South Valley University in Aswan, together with the Geology Department, Faculty of Science, Assiut University, Egypt. In the first study, a series of sections were produced using a geoelectric survey, from which measurement points were extracted alongside information concerning the distance between the shallow aquifer and ground level, as well as its overall depth. Sixteen points were identified in total, as shown in Figure 88, which correspond to the cross-sections illustrated in Figures 8 to 10 of Rabeh et al. (2018)41. The second study provided a further five measurement points for the shallow aquifer, derived from well data across the region provided by Figure 3 of Hamdan et al. (n.d.)42. U.S. Geological Survey, “EarthExplorer,” 2025, https://earthexplorer.usgs.gov/. Taha Rabeh et al., “Structural Control of Hydrogeological Aquifers in the Bahariya Oasis, Western Desert, Egypt,” Geosciences Journal 22, no.1 (2018): 145–54, https://doi.org/10.1007/s12303-016-0072-3. 42 Ali M. Hamdan et al. Evaluation of Hydrogeochemical Parameters of the Groundwater in El-Bahariya Oasis, Western Desert, Egypt. 40 41

84 |Research Development

Fig.87 Section of Bahariya Oasis Soil Layers

Fig.88 Magnetic Sounding Station Points and Well Locations Within The Oasis

Research Development| 85

The findings of this research indicate that the aquifer is most accessible in the southernmost part of the region, and to a lesser extent in its northernmost area. At the same time, the greatest volumes of the aquifer are predominantly located at the centre of the depression and, to a significant degree, in the northern areas. These results were cross-referenced with aquifer cross-sections published in Figure 4 of Sharaky and Abdoun (2020)43, which reported data accuracy levels of 80% in the southern areas, 70% in the central areas, and 85% in the northern parts of the Bahariya. Furthermore, empirical data on the locations of existing settlements and agricultural land, outlined in the domain chapter, provide additional validation for these findings. 43 Abbas M. Sharaky and Suad H. Abdoun, “Assessment of Groundwater Quality in Bahariya Oasis, Western Desert, Egypt,” Environmental Earth Sciences 79, no. 6 (2020), https://doi.org/10.1007/s12665-0208823-x.

86 |Research Development

Fig.89 3D Visualisation Of The Predictive Model Output For Aquifer Depth And Accessibility

Research Development| 87

Fig.90 Aquifer Accessibilty Gradient Map

The predictive model was applied across the constructed grid to produce gradient maps of aquifer accessibility and depth. These maps translate the interpolated data into a spatial visualisation, allowing for a comparative reading of water availability across the Bahariya Oasis. The aquifer accessibility map highlights areas where the water table is closest to the ground surface, identifying zones with the greatest potential for efficient and low-cost extraction. Conversely, the aquifer depth map illustrates the volumetric capacity of the aquifer, emphasising areas that can sustain longterm water provision despite requiring deeper wells. Taken together, the two maps provide complementary perspectives on groundwater resources: accessibility offers insight into immediate feasibility, while depth ensures resilience and capacity over time. This dual evaluation forms a critical basis for identifying viable settlement locations within the oasis. Fig.91 Aquifer Depth Gradient Map

88 |Research Development

4.1.2 Multi-criteria Site Evaluation 1. Soil Mapping and Evaluation Following the analysis of the aquifer structure, the soil conditions of the Bahariya Oasis were documented. Three principal soil types were identified: Aridisols, Entisols and rock formations, each evaluated for their suitability for construction and excavation. Although rock formations provide a stable substrate, in Bahariya they occur mainly in elevated mountainous terrain, making them difficult and costly to develop. Entisols, which include Quartzipsamments, Torrifluvents, Torriorthents and Torripsamments, are generally adequate for lightweight construction requiring minimal foundations, but are less suitable for more substantial development. In contrast, Aridisols (comprising Aquisalids, Gypsiargids, Haplogypsids and Haplosalids) present the most favourable conditions, as they can be stabilised at relatively low cost. 44 All soil types were analysed using the same 1 km rectangular grid described in the aquifer analysis, and values were assigned to each cell according to their construction potential.

Fig.92 Soil Quality Gradient Map

Elnaggar, A. A., A. A. El Baroudy, and A. A. N. Hassan. “Soil Classification of Bahariya Oasis Using Remote Sensing and GIS Techniques.” Journal of Soil Sciences and Agricultural Engineering 4, no. 9 (2013): 921–947. https://doi. org/10.21608/jssae.2013.52488 44

2. Contextual Factors for Site Selection The final analysis focused on key locations within the Bahariya Oasis, as identified in the map data shown in the Domain chapter in Figure 18. The primary information extracted concerned existing agricultural and settlement sites, abandoned plots of land, and well locations. Proximity analysis was then applied to each grid cell in relation to these features. Cells coinciding with sites already in use or previously utilised were evaluated as less suitable for new development. In addition, areas within 1.5 kilometres of abandoned plots were also considered less suitable, owing to salinisation caused by overexploitation, as discussed in the domain chapter. By contrast, areas closer to active settlements and agricultural land were considered more favourable, as they support social cohesion and an integrated urban and rural landscape. These parameters were combined using a grading system with equal weighting to evaluate the qualities of each grid cell. Fig.93 Site Integration Gradient Map

Research Development| 89

3. Weighted Analysis of Site Parameters The final stage of this experiment built on the fact that all previous analyses were conducted on the same grid format. This allowed all values to be aggregated and compared, enabling the research team to assign relative importance to each parameter.

The resulting grid, shown in Figure 94 , identifies three areas as suitable under a conservative estimation of conditions: two in the northernmost part of the oasis and several in the southernmost part.

Aquifer shallowness was identified as particularly significant, as it provides the clearest indication of water availability, which is critical in a desert context; it was therefore assigned a weight of 1.5. The agricultural and settlement context was also given a weight of 1.5, as it situates the project within its broader sociospatial setting. Aquifer depth, while important, was assigned a lower weight of 1.2, since even areas of reduced depth have sufficient capacity to support large settlements. Soil condition received the lowest weighting of 1.0, as variations in quality, apart from unbuildable rock formations, can generally be managed more easily.

The design team selected the northernmost area as the most appropriate, given its proximity to El Bawiti, the largest settlement in the region, and its stronger performance compared to the eastern site. Under less restrictive conditions, as illustrated in Figure 95, further expansion of settlements would avoid the central part of the oasis, progressing instead from the south towards the north and from the north towards the south east.

Fig.94 Weighted Criteria Gradient Map

90 |Research Development

Fig.95 Site Selection Map Showing Optimal Expansion Locations And Selected Settlement Site

Research Development| 91

4.2 Settlement Organisation The organisation of the settlement is structured through a rule-based system that integrates water management with spatial distribution. Using a 40 × 40 metre grid as the operational framework, architectural typologies are allocated through principles derived from Cellular Automata, ensuring both functional coherence and adaptability. Water consumption and distribution form the central drivers of this process, linking cisterns, wells, housing, agriculture, and condensation farms into a semi-closed loop system. Finally, these rules generate an integrated settlement fabric where resource cycles, programmatic needs, and spatial logic are resolved into a scalable prototypical model.

92 |Research Development

Fig.96 Grid Division of Proposed Site for Development

Research Development| 93

4.2.1 Water Consumption and Storage To address the challenge of developing a new prototypical settlement, the selected site is divided into a 40 × 40 metre grid, with each cell representing a potential location for one of four architectural typologies: cistern, housing, agriculture or condensation farm. Each typology occupies an area of 1,600 m², and their distribution across the grid is governed by a rule-based aggregation process informed by principles of the Cellular Automata model. Based on existing literature on water yield and the positioning of wells in the Bahariya Oasis, the design team placed two new wells within the experimental area. Their locations were derived from the best-performing positions identified through the aquifer analysis, while maintaining adequate spacing between them, greater than 1 km. Each well is estimated to yield approximately 150 m³ per hour, equivalent to 3,600 m³ of daily production under constant outflow.46 This capacity translates into 50 cistern units being supplied by each well, which in turn each provides 72 m³ per day to the settlement. The cistern units are designed to store 1,100 m³, of which 252 m³ are allocated to the water needs of around 120 people over a two-week period. 47 In addition, every cistern supports 1 acre of agricultural land cultivated with date palm trees, olive trees and wheat, which together consume on average 756 m³ during the same period. Spatially, this is expressed as three residential clusters and two agricultural units per cistern unit, each occupying 1,600 m² of the grid. Any surplus water is directed towards thermal regulation, taking advantage of evaporative cooling processes. Losses from evaporation and transfer are expected to be offset by atmospheric condensation and water purification processes. Collectively, these clusters create a semi-closed loop of water circulation that underpins the programmatic and spatial organisation of the settlement. 46 Noha H. Moghazy and Jagath J. Kaluarachchi, “Assessment of Groundwater Resources in Siwa Oasis, Western Desert, Egypt,” Alexandria Engineering Journal 59, no. 1 (2020): 149–63 47 Noha H. Moghazy and Jagath J. Kaluarachchi, “Sustainable Agriculture Development in the Western Desert of Egypt: A Case Study on Crop Production, Profit, and Uncertainty in the Siwa Region,” Sustainability 12, no. 16 (2020): 6568,

94 |Research Development

Fig.97 Grid Division of Selected Site with Optimal Well Locations

Fig.98 Allocation of Resources and Consumption to Architectural Units

Research Development| 95

4.2.2 Rule-Based Distribution Through Cellular Automata In addition to defining relational patterns between units, the aggregation system was governed by further rules to ensure functional integration. These rules established optimal elevation differences to promote gravity-fed water flow. Cisterns are positioned approximately 0.5 m higher than the housing units in order to supply them with water, while the housing units are placed about 0.5 m above the agricultural units to enable greywater distribution. Within each settlement cell, two condensation farm units are incorporated, positioned at the same elevation as the cisterns. These elevation parameters were determined according to optimal qanat incline ratios, allowing interconnection of units across the 40 m grid spacing.48 48

Ali Hamidian, Mehdi Ghorbani, Mahsa Abdolshahnejad, and Aziz Abdolshahnejad, “Qanat, Traditional Eco-Technology for Irrigation and Water Management,” Agriculture and Agricultural Science Procedia 4 (2015): 119–25

Fig.99 Cellular Automata Horizontal Placement Rules

Fig.100 Cellular Automata Vertical Placement Rules

96 |Research Development

Further rules dictate that neighbouring housing units should be clustered together, even when they belong to different organisational clusters. Agricultural units follow the same principle, although with secondary importance, in order to create continuous cultivated areas.

Fig.101 Final Aggregation of Units based on Stated Rules

These rules aim to promote coherent agricultural and urban zones, thereby strengthening overall consistency in the settlement fabric. Additional rules position wells as attractor points, drawing the surrounding typologies to organise around them. Finally, the geometrical centre points of each cluster are interlinked, after which neighbouring grid points are adjusted to their closest available and most optimal position. This process establishes the spatial framework for a road network that interconnects all cluster typologies of the settlement. The final aggregation, as illustrated in Figure 100, represents the outcome of an iterative process informed by Cellular Automata principles. Each cell is assessed individually for its optimal positioning within the global rules of the system until collectively resolved into the final settlement typology. Finally, a region of the settlement catering to 1200 inhabitants is showcased in Figure 101 to provide insight on the scale of the project. Fig.102 Settlement of 1200 People

Research Development| 97

4.3 Hydro-Social Network The organisation of the settlement extends beyond the allocation of units to encompass the infrastructural systems that interconnect them. Central to this is the hydro-social network, which represents a series of topological connections between the architectural units introduced in the first chapter. Generated by identifying the most proximate links between different unit combinations and isolating the shortest paths, the network establishes optimal connections across the settlement. Developed through this logic, the hydro-social network produces both road and water systems that align with the distribution of cisterns, wells, housing, and agricultural units.

98 |Research Development

Fig.103 Primary Water Network of Generated Settlement

Research Development| 99

4.3.1 Road Network Organisation The initial experiments focused on generating the main road network. Conceived at a higher level than the predominantly subterranean settlement, this network emerged by linking the centres of each cluster and carving the shortest distances between neighbouring nodes. This process established the framework for the primary road system, functioning as the motorway that interconnects the settlement. A secondary network was then produced by applying the same method to the cistern and housing units, the principal social and habitable spaces. These routes are embedded into the architectural typologies, aligning with their spatial organisation and eliminating the need for further alterations to the overall configuration. Paths are placed along the peripheries of each unit: housing clusters maintain open-air routes, while cistern units incorporate shaded passages that seamlessly integrate with their internal functions. These architectural properties will be examined in greater detail in the design development chapter.

Fig.105 Axonmetric Diagram of Road Network

Fig.104Road Networks

100 |Research Development

4.3.2 Water Network Organisation The water distribution network of the settlement follows the same organisational principles as the road network. The primary supply network is created by linking each cistern unit to its most proximate well along the shortest possible path, thereby establishing the backbone of water provision for the settlement. A secondary network interconnects clusters of five neighbouring cisterns to form semi-autonomous water infrastructures that provide redundancy in the event of local malfunctions. Finally, tertiary supply lines link cistern units with housing and agricultural units respectively, ensuring integrated circulation throughout the settlement. These networks are conceived through the same logic of interconnection, but their organisation also incorporates a vertical dimension. Wells are located at higher elevations, which allows the primary network to be established at the deepest level of the hydrological system, following the qanat principles, developing at approximately –7.50 m. The secondary network interconnecting cisterns is set at –7.00 m, aligning with the elevation of the cistern reservoirs. From there, the supply lines branch upwards: at –6.50 m for the housing units, whose lowest reservoir levels are at –5.50 m, and at –6.00 m for the agricultural units, whose lowest levels are at –5.00 m. In conclusion, this multilayered hydro-social network establishes both the horizontal and vertical organisation of the settlement’s water infrastructure. It provides a coherent system of interconnection while avoiding spatial conflicts between overlapping networks.

Fig.107 Axonmetric Diagram of Water Network Fig.106 Water Networks

Research Development| 101

4.4 Material Development The material development phase bridges the literaturebased framework with material experimentation and prototyping. It aims to test how theoretical strategies, such as functional grading, fibre reinforcement, and biostabilisation, can be applied to locally available soils and adapted for desert architecture. Through a sequence of design, testing, and fabrication stages, the process establishes clear performance benchmarks and progressively translates them into architectural applications.

102 |Research Development

Fig.108 Material Samples

Research Development| 103

Excavation The excavation required for subterranean archetypes produces large volumes of sandy-silt soil, which in this research is treated not as waste but as the primary construction resource. This approach reflects the vernacular logic of desert settlements, where building material was traditionally taken directly from beneath the architecture itself.

Excavation thus becomes the starting point of a circular system: recovered soil is combined with agricultural by-products to enhance cohesion, tensile resistance, and water resistance. The act of carving space below ground simultaneously generates material for construction above, linking site formation and resource supply in a single integrated process.

Fig.109 Excavation-led circular resource model.

104 |Research Development

On-Site Fabrication Hub To transform the excavated soil into architectural components, an on-site fabrication system is proposed as a contemporary analogue of the communal brick-making yards once common across Saharan oases. At its core, a temporary hub is organised into sequential stations for soil and fibre mixing, modular casting of interlocking blocks, CNC and 3D printed reusable formwork, curing, and pre-assembly.

Circularity defines the process: cast, unfired blocks have a low carbon footprint and remain recyclable, while modular formworks and scaffolding are designed for repeated use. Co-locating these processes reduces transport, manages soil variability, and enables rapid iteration. The result is a framework of low-tech precision, where simple means deliver complex architecture in dialogue with the landscape.

Fig.110 On-site fabrication hub for earthen FGM blocks.

Research Development| 105

Material Properties

Sand

The material system is conceived as a functionally graded composite, transitioning from a dense, structural exterior with high thermal mass to an interior tuned for hydrophobicity and crack resistance. Prior research indicates that gradual gradients can reduce weak interfaces and enhance integrity. Following this principle, the outer blends prioritise compressive strength and thermal inertia, while inner blends target water resistance. Each component, fibres, xanthan gum, lime, magnesium oxychloride, sodium silicate, polymers, or oils, was selected from previous studies for a specific role. In hyper-arid climates, massive earthen walls moderate indoor temperatures through thermal inertia,49 but unreinforced soil is brittle and weak in compression. The integration of fibres, such as straw or date palm, improve tensile resistance and reduce shrinkage.50 And, using xanthan gum, for instance, increases soil strength by more than twice, and in combination with fibres, up to fivefold.51 While stabilisers like lime or biopolymers significantly raise compressive strength,52 mineral stabilisers extend this logic: magnesium oxychloride cement (MOC) and sodium silicate are both noted for providing high compressive strength and water resistance.53 Water resistance remains critical for subterranean and condensation-exposed structures. Vernacular builders employed organic additives and coatings, milk, cactus juice, or oils, while casein paints created insoluble compounds that clogged pores.54 Contemporary research expands this with polymer emulsions, which drastically reduce absorption while improving strength, and drying oils such as boiled linseed oil, which polymerise into effective hydrophobic barriers.55 These findings demonstrate that both natural and synthetic additives can successfully impart water resistance when integrated within the mix, avoiding the fragility of applied coatings. Gernot Minke, Building with Earth. S Nasla et al., ‘An experimental study of the effect of pine needles and straw fibres on the mechanical behaviour and thermal conductivity of adobe earth blocks with chemical analysis’. 51 Pouyan Bagheri et al., ‘Effects of Xanthan Gum Biopolymer on Soil Mechanical Properties’. 52 Sanket Rawat et al., ‘Mechanical Performance of Hybrid Fibre Reinforced Magnesium Oxychloride Cement-Based Composites at Ambient and Elevated Temperature’, 53 Gobinath et al., ‘Banana Fibre-Reinforcement of a Soil Stabilized with Sodium Silicate’. 54 Przemysław Brzyski et al., ‘The Infl uence of Casein Protein Admixture on Pore Size Distribution and Mechanical Properties of LimeMetakaolin Paste’. 55 Lee et al., ‘A Study on Water Repellent Effectiveness of Natural Oil-Applied Soil as a Building Material’. 49 50

106 |Research Development

Clay

Lime

Fibres

Sand

Clay

Sodium Silicate

Clay

MgO + MgCl2

Fibres

Sand Fibres

te

Xanthan Gum

Sand

Clay

Lime

Casein

Clay

Lime

Polymer Emulsion

Clay

Lime

Linseed Oil

Fibres

Xanthan Gum

Sand Fibres

Xanthan Gum

Sand Fibres

Fig.111 Constituent palette and pairing matrix for graded earthen composites

Research Development| 107

Material Experiments

Material Design Structural Strenght

Main Composition +

+

+ Binder

Material Workflow The research followed an iterative workflow structured in three stages: Material Experiments, Architectural Application, and Fabrication Process.

Mixing Proportions

Material Samples

Thermal Mass

Ph Hydrophobic

+ Additives

ρ

Physical Test

Stage 01. Material Design: Candidate mixes were developed from sand, clay, fibres, and stabilisers, then cast into samples for physical testing. Each formulation was assessed for compressive strength, thermal mass, and hydrophobic behaviour. Results informed refinements in binder proportions and additive combinations. Stage 02. Architectural Application: Selected Functionally

Evaluation

Graded Material (FGM) samples were integrated into vault design studies. Using computational tools such as Thrust Network Analysis (TNA), Kangaroo and Karamba 3D, the topological interlocking logics were tested to ensure equilibrium and constructability. These simulations linked material performance to structural form.

Stage 03. Fabrication Process: Parameters derived from

testing and simulation informed the casting of interlocking units. Reusable formwork was developed through CNC milling and additive manufacturing, while blocks were manually cast and assembled into prototypes. This stage validated both the fabrication workflow and the scalability of the material system.

108 |Research Development

STAGE 01 STAGE 02

Fabrication Process Architectural Application Parameters Setup Form finding

Physical Test FGM Sample

Tesselation design Structural Analysis

Formwork Additive Manufacturing

Software </> Hardware

Evaluation

Fabrication Parameters

Final Prototype

Manual Casting

Evaluation

Evaluation

2 STAGE 03

Fig.112 Material Research workflow

Research Development| 109

Xanthan Gum 2%

Water 5%

Xanthan Gum 2% MgCl2 5% MgO 8%

Lime 13% Fibre 5%

Sand 50% Clay 25%

Fibre 5%

Water 5%

S

Sand 50% Clay 25%

Exp 1. Lime

Exp 2. MgO + MgCl2

γ = 9.09 kg/m3 σ = 0.01 kN/ cm2

γ = 13.12 kg/m3 σ = 0.0154 kN/ cm2

Stage 1.1 Initial Experimentation Six candidate formulations, combining sand, clay, fibres, and different stabilisers, were initially tested to evaluate feasibility. Half-cube specimens (100 × 100 × 50 mm) were cast with minimal water for workability and left to air-dry for one week. The trials assessed drying behaviour, shrinkage/crack formation, and early strength.

110 |Research Development

High Thermal Mass Material

Xanthan Gum 2%

Water 5%

Sodium Silicate 13%

ass

Water 7%

Casein 3%

Sand 50% Clay 25%

Fibre 5%

Sand 50% Clay 25%

Water 7%

Linseed Oil 3% Lime 10%

Lime 10%

Lime 10%

Fibre 5%

Water 7%

Polymer Emulsion 3%

Fibre 5%

Sand 50% Clay 25%

Fibre 5%

Sand 50% Clay 25%

Exp 3. Sodium Silicate

Exp 4. Casein

Exp 5. Polymer Emulsion

Exp 6. Linseed Oil

γ = 12.44 kg/m3 σ = 0.0146 kN/ cm2

γ = 13.01 kg/m3 σ = 0.0096 kN/ cm2

γ = 16.81 kg/m3 σ = 0.0164 kN/ cm2

γ = 12.96 kg/m3 σ = 0.0154 kN/ cm2

Water Resistant Material

Fig.113 Initial screening of stabiliser systems

Research Development| 111

Drying/Shrinkage & Cracking

15

20

25

3 0

%

Fig.114 Volumetric shrinkage after 7-day air curing of Exp. 4 Casein

112 |Research Development

Oil treated earth

Polymer enh anced earth

Casein lime b inder

Silicate stab iliz ed earth

Sorel cement

Lime stab iliz ed earth

0

5

10

During curing, most samples showed only superficial surface crazing. The casein blend developed a circumferential crack, while the lime-stabilised mix exhibited excessive shrinkage.

Fig.115 Volumetric shrinkage (%) at day-7 for six formulations

/

Oil treated earth

Polymer enhanced earth

Casein lime binder

Silicate stabilized earth

Sorel cement

Lime stabilized earth

0.000

0.002

0.004

0.006

0.008

0.010

0.012

0.014

0.016

Simple hand-press compression at day 7 indicated that the remaining formulations, MOC, sodium silicate, polymer emulsion, and linseed oil, exceeded the threshold set for continuation.

kN cm2 0.018

Compressive Strenght

(kN/cm) 2

Fig.Hand-press compression set-up

Fig.117 Compressive strength (kN/cm²) for the initial six formulations

Research Development| 113

γ = 9.09 kg/m 3 σ = 0.01 kN/ cm 2

Polyme

%

Fib

res

γ = 9.09 kg/m 3 σ = 0.01 kN/ cm 2

Stage 1.2 Material Experimentation Twelve candidate formulations were cast into 60 × 60 × 60 mm cubes and cured for 14 days under controlled conditions. Each sample was then subjected to a compressive strenght, water absorption and thermal conductivity test.

114 |Research Development

γ = 9.09 kg/m 3 σ = 0.01 kN/ cm 2

γ = 9.09 kg/m 3 σ = 0.01 kN/ cm 2

% Sodium Silicate

Oil Emulsio

γ = 16.81 kg/m3 σ = 0.0164 kN/ cm2

γ = 12.96 kg/m3 σ = 0.0154 kN/ cm2

Polymer

es % Fibr

Linseed

γ = 12.44 kg/m3 σ = 0.0146 kN/ cm2

n

γ = 13.12 kg/m 3 σ = 0.0154 kN/ cm 2

γ = 13.12 kg/m 3 σ = 0.0154 kN/ cm 2

γ = 12.44 kg/m3 σ = 0.0146 kN/ cm2

Sodium Silicate

MgO + MgCl 2

γ = 16.81 kg/m3 σ = 0.0164 kN/ cm2

γ = 12.96 kg/m3 σ = 0.0154 kN/ cm2

% MgO + MgCl 2 Fig.118 Parametric mix matrix for graded earthen composites

Research Development| 115

Thermal Conductivity

Fig.119 Experimental setup and qualitative visualisation

116 |Research Development

12% SodiumSilicate+2% F ib res

12% SodiumSilicate+5% F ib res

10% SodiumSilicate+2% F ib res

10% SodiumSilicate+5% F ib res

15% Sorel Cement +2% F ib res

15% Sorel Cement +5% F ib res

12% Sorel Cement +2% F ib res

12% Sorel Cement +5% F ib res

0

5

10

15

20

25

30

35

40

45

50

Thermal conductivity was assessed with a transient plate method on flat samples, supported by qualitative ice-melt trials to visualise heat transfer.

Fig.120 Thermal conductivity of candidate stabilised-earth mixes

5

10

15

20

Water absorption was measured through 1-hour and 24-hour submersion tests, recording weight uptake.

25

%

Water Submersion

Sodium Silicate + Linseed Oil

Sodium Silicate + Polymer Emulsion

Sorel Cement + Polymer Emulsion

Sorel Cement + Linseed Oil

0

Fig.121 Water submersion test

Fig.122 Water absorption (%) after 1-h and 24-h submersion

Research Development| 117

Compressive Strength Compressive strength was evaluated by loading the cubes to failure, providing a benchmark for structural capacity.

Fig.123 Hand-press compression set-up

118 |Research Development

Fig.124 Compressive strength (kN/cm²) for all the formulations

Proposed Material The formulation selected are: 1. High thermal mass grade: Stabilised with magnesium oxychloride cement (MOC), fibres, and a sand–clay matrix. 2. Water-resistant grade: Combining MOC and linseed oil within the sand–clay base.

These two composites are integrated into a functionally graded block, where the material transitions from a dense structural outer layer to a hydrophobic inner surface. This gradient minimises weak interfaces while consolidating the advantages of both mixes into a single unit, enabling controlled moisture protection, passive climatic regulation, and structural reliability.

Functunally Graded Blocks

Composition MgCl2 4%

Water 8%

Linseed Oil 3% MgCl2 4%

MgO 8%

Water 8%

MgO 8%

Fibre 5%

Fibre 2%

Sand 50%

Sand 50% Clay 25%

Clay 25%

MOC High-mass Grade

MOC Water Resistant

Properties Shrinkage / Cracking

Shrinkage / Cracking

Local Availability

Compressive Strenght

Local Availability

Compressive Strenght

Embodied Energy

Water Uptake

Embodied Energy

Water Uptake

Thermal Conductivity

Thermal Conductivity MOC High Mass Grade

Concrete

MOC Water Resistant

Mud Brick

Fig.125 Proposed Material

Research Development| 119

Stage 1.3 Functionally Graded Material Fabrication To combine the two mixes into a single graded element, three wet-on-wet casting protocols were explored, drawing on established methods from functionally graded cementitious material (FGCM) research. Literature identifies sequential casting with intermixing, vibrationassisted settlement, and static compaction as the main strategies for achieving through-thickness gradients while avoiding weak interfaces.56 Three casting strategies were tested to establish a reliable method for producing graded blocks. In the first, a transitional region was created by placing the second mix while the first remained plastic and lightly shearing the interface; this generated a gradient but risked partial layer separation. The second approach used table vibration to promote interpenetration and air release; while bonding improved, excessive mixing often blurred the gradient, aligning with studies warning that vibration must be tightly controlled. The third approach, static compaction, consolidated layers under constant pressure, producing stable and continuous gradients without delamination. Across trials, static compaction proved the most consistent and robust method, and was therefore selected as the primary protocol for graded prototypes. 56

Yu, R., and P. Kabele. “Functionally Graded Concrete: A Review of Materials, Production Methods, and Structural Applications.”

120 |Research Development

Pouring (Baseline)

Transitional Region

- Uncontrolled - Weak

- Partial layer Separa - Poor Interface

50% MOC High Mass 50% MOC Water Resistant

33% MOC High Mass 33% Transitional Layer 33% MOC Water Resista

Table Vibration

Static Compaction

paration

+ Mixed Well - Gradient Disrupted

+ Stable Gradient + Best Result

ss er esistant

50% MOC High Mass 50% MOC Water Resistant

50% MOC High Mass 50% MOC Water Resistant 80% MOC High Mass 20% MOC Water Resistant 20% MOC High Mass 80% MOC Water Resistant

Fig.126 Material Samples

Research Development| 121

Stage 2 Form Finding This stage establishes the vault geometry based on funicular principles. Using Thrust Network Analysis, the vault forms were digitally generated to ensure that thrust lines remain entirely within the masonry, enabling compression-only behaviour.57 The diagrams illustrate the equilibrium of horizontal and vertical forces, confirming that the proposed forms follow stable catenary logic. This method provided a rigorous structural basis while allowing exploration of complex free-form geometries adapted to site conditions.

57

Block, Philippe, and John Ochsendorf. “Thrust Network Analysis: A New Methodology for Three-Dimensional Equilibrium.”

Form Diagram

Form Diagram

Horizontal Force Diagram

Horizontal Force Diagram

Vertical Force Diagram

Vertical Force Diagram Fig.127 Thrust Network Analysis of a funicular vault under variant boundary conditions.

122 |Research Development

Following the TNA stage, the vaults were modelled in Kangaroo, the physics engine for Grasshopper. Here, a spring-particle system was used to simulate real-time equilibrium of discrete block assemblies. This allowed testing of the interlocking brick tessellations under self-weight, validating that the vault could stand without mortar once the keystone was placed. The simulations further helped assess local deformations, confirming that the proposed tessellation maintained stability while accommodating soil variability.

Fig.128 Discrete arch under self-weight: thrust-line verification.

Research Development| 123

Interlocking and topological optimization The blocks were developed as interlocking units, where geometric profiles enhance mechanical stability and guide assembly without reliance on mortar. The topological optimization of the joint geometry ensures that thrust is safely transferred across units, improving the vault’s resistance to sliding and shear. This, in parallel with the strategy of creating a graded composition for the material performance, defines the basis for scalable vault construction in desert conditions.

Fig.129 Interlocking voussoir family and tiling logic for a graded vault

124 |Research Development

Fig.130 Dry interlocking earthen units: joint engagement test.

Research Development| 125

Stage 3 Fabrication The fabrication stage translates the experimental findings on material performance into full-scale construction strategies. Having established the high-thermal-mass and hydrophobic composites, as well as the casting protocol for producing stable gradients, the focus shifted to scaling production methods and integrating them with architectural form. This process combines material grading, interlocking geometry, and modular formwork into a single workflow designed for on-site implementation.

126 |Research Development

Fig.131 Stacked graded-earthen prototype: dry interlocking under vertical load

Research Development| 127

Form Diagram

Horizontal Force Diagram

Fabrication trials explored both additive and subtractive techniques, including CNC- and 3D-printed reusable formwork for complex vault components, manual casting of graded units, and small-scale prototyping of interlocking assemblies. The objective was not only to demonstrate constructability, but also to ensure that the chosen methods align with the principles of circularity, low-tech precision, and adaptability to desert conditions. Through this stage, fabrication becomes the link between material research and architectural application, testing how laboratory-scale composites can be systematically transformed into viable building systems.

Fig.132 Reusable modular formwork set for casting graded interlocking blocks

128 |Research Development

Fig.133 Erection sequence for a dry-joint vault.

Fig.134 Quarter-scale vault prototype on centring

Fig.135 Inclined roof-panel prototype with dry interlocking units

Research Development| 129

DESIGN DEVELOPMENT

INTRODUCTION The design development chapter examines the architectural dimensions of the typologies outlined in the research phase. The urban, material and environmental framework established earlier anchors these typologies as iterative instances that respond to social, programmatic and infrastructural parameters. In this way, the typologies do not merely act as stepping stones towards an optimal outcome but also as discrete cases that can be dissected and reconfigured in further iterative processes. While computational logic is employed to integrate these parameters, the research extends beyond a purely calculative approach. Architectural design remains a constant effort of the design team to interpret computational results, situate them within real conditions and enhance what has already been identified as optimal by the digital process. The prototypes presented in the following chapter embody this intention, translating abstract data into architectural space. Fig.136 Housing Unit Visualisation

5.1 Atmospheric Water Harvesting The initial experiment in this chapter repurposes the concept of the solar updraft tower, traditionally designed as a renewable energy device, into an atmospheric water harvesting system. Rather than generating power, the tower’s geometry is employed to intensify air suction, diverting part of the airflow into subterranean cooling chambers. Within these chambers, vapour condenses on cooled surfaces and is channelled into the settlement’s water network. The study investigates how climatic conditions, geothermal properties and tower morphology shape water yield, employing parametric modelling, CFD simulations and optimisation to define an efficient prototype for desert settlements.

134 |Design Development

Design Development| 135

5.1.1 Solar-Induced Airflow Dynamics The solar updraft tower (SUT) functions through the combined greenhouse and chimney effects, a concept first proposed in early theoretical studies58 and later demonstrated at the Manzanares pilot plant in Spain. 59

By ensuring a steady and powerful flow, the SUT concept provides an effective mechanism to maximise the amount of atmospheric moisture brought into contact with cooler surfaces.

A transparent collector heats the air beneath it, which rises through a tall chimney, creating a pressure differential that draws in cooler air from the perimeter. The tower’s taper further accelerates the flow, sustaining a powerful suction effect.60 For water harvesting, this principle is significant not for energy generation but for maximising air throughput, thereby increasing the amount of atmospheric moisture available for condensation.

The greater the air throughput, the higher the potential rate of condensation, making this principle fundamental to designing an efficient atmospheric water harvesting system.

58 Haaf, W. (1984). Solar chimneys: part II: preliminary test results from the Manzanares pilot plant. International Journal of Solar Energy, 2(2), 141–161. 59 Schlaich, J., Bergermann, R., Schiel, W., & Weinrebe, G. (1995). Design of commercial solar updraft tower systems—utilisation of solar induced convective flows for power generation. Journal of Solar Energy Engineering, 127(1), 117–124. 60 Krisst, R. (1983). Solar chimney for power generation. Applied Energy, 13(2), 83–100.

Fig.137 Solar Updraft Tower Mechanism Diagram

136 |Design Development

5.1.2 Design Concept & Condensation Performance In the proposed design, solar radiation is absorbed by a collector made from the earthen composite material developed in the previous chapter, enhanced with volcanic ash to improve heat retention. This raises the temperature of incoming air, which rises through the tower to create suction, while a portion of the flow is channelled underground into chambers cooled by surrounding soils, typically 12 °C lower than surface conditions. Additional cooling is achieved by the integrated qanat network, which circulates aquifer water to dissipate heat. Within these subterranean chambers, condensation occurs on corrugated panels formed from the hydrophobic earthen material developed in this research, whose corrugated surfaces enhance droplet formation and whose hydrophobicity prevents absorption. The collected water is directed into reservoirs for distribution within the settlement or recharge of the aquifer. System performance is assessed by calculating dew point thresholds using the Magnus–Tetens approximation, which relates ambient temperature and humidity to the conditions required for condensation.

Fig.139 Proposed Atmospheric Harvesting System

The performance of the system is evaluated by calculating the temperature differential required for condensation and by estimating potential water yields. The dew point temperature is determined using the Magnus–Tetens approximation, which relates ambient air temperature and relative humidity to the threshold at which vapour condenses.61 For Bahariya’s summer conditions, with an average air temperature of 36 °C and a relative humidity of 53%, the required cooling is approximately 12 °C. Once this threshold is reached, water yield can be derived from airflow calculations, where volumetric flow is expressed as the product of inlet area and air speed. The yield rate is then estimated through a mass balance approach that links airflow volume and the change in humidity ratio to the total amount of water condensed.62 Lawrence, M. (2005). The relationship between relative humidity and the dew point temperature in moist air: A simple conversion and applications. Bulletin of the American Meteorological Society, 86(2), 225–234. 59 Klemm, O., et al. (2012). Fog as a fresh-water resource: Overview and perspectives. Ambio, 41(3), 221–234. 61

Fig.138 Condensation and Water Yield Calculations

Design Development| 137

5.1.3 Experiment Set-up The initial tower morphology was developed to generate strong suction and maintain high airflow velocities through its shaft while allowing redirection underground. Computational fluid dynamics (CFD) simulations were used to evaluate both airflow intensity and stability within the structure. Results demonstrated that the design sustained an average velocity of 6 m/s, confirming consistent performance throughout the shaft. Importantly, the diverted airflow retained sufficient speed as it entered subterranean chambers, ensuring high air volumes for condensation. These outcomes validated the viability of the initial geometry and established a foundation for subsequent optimisation.

Fig.140 Intial Tower Morphology Development

138 |Design Development

Fig.141 CFD Preliminary Test

Fig.142 Parameters and Objectives Set for Optimization of Proposed Tower

Following this proof of concept, the design was systematically iterated through a parametric framework to be optimised through Wallacie. Variables, specified in figure 142 included shaft height, base and outlet radii, inlet size and number, and underground chamber dimensions. The optimisation was guided by four key objectives: maximising heat gain at the collector surface, achieving minimum subterranean temperature for condensation performance, maximising airflow speed across different morphologies, and minimising structural deflection using finite element analysis in Karamba 3D. To extend beyond the limits of direct simulation, a neural network algorithm was trained on 50 CFD test cases with varied morphological inputs, learning the relationship between geometry and airflow dynamics. Once validated, the predictive model was integrated into the Wallacei optimisation process, enabling it to estimate airflow speeds for any possible morphology without requiring full CFD simulations for each case. This allowed Wallacei to evolve and evaluate 800 phenotypes efficiently, significantly reducing the time needed to explore the design space. Design Development| 139

Wallacei Optimisation Settings 1. Population Generation size: 40 Generation count: 20 Total population size: 800

2. Algorithm Parameters Crossover probability: 0.9 Mutation probability: 1/n Crossover distribution index: 20 Mutation distribution index: 20 Random seed: 1

3. Simulation Parameters Number of genes (sliders): 13 Number of values (slider values): 238 Number of fitness objectives: 4 Size of search space: 2.1 × 1015

Fig.143 A Selection of Pareto Front Phenotypes

140 |Design Development

5.1.4 Results & Post Analysis The optimisation process produced a broad variation of tower morphologies, each balancing the four fitness criteria differently. Successive generations demonstrated measurable improvements, with the SD graphs showing how populations gradually converged towards higher-performing solutions, and the parallel coordinate graphs visualising the trade-offs between objectives. From this set, the most optimal morphology was selected from the Pareto front. The chosen outcome was then cross-referenced with a final CFD simulation, which confirmed the machine learning model’s prediction with nearly 90% accuracy. Using the previously established condensation equations, the system’s performance was estimated to yield an average of 2,800 litres of water per day, based on an assumed efficiency of 50% and six hours of daily production under adequate thermal conditions. This validates both the optimisation framework and the integration of computational and environmental parameters in shaping an efficient waterharvesting prototype.

Fig.144 Selected Morphology

Fig.145 CFD Analysis on Selected Morphology

SD - Graphs

FO-1

FO-2

FO-3

FO-4

Parallel Coordinate Graph

Design Development| 141

5.1.5 Architectural Development The atmospheric water harvesting tower is scaled into a condensation farm, each occupying a 40 × 40 metre grid cell within the settlement framework. Organised as clustered arrays, these farms supplement cistern storage by offsetting evaporation losses and reinforcing the semi-closed water cycle of the settlement. Their operation contributes not only to the water supply but also to agricultural irrigation and evaporative cooling. The cooled subterranean chambers help regulate the local microclimate, lowering ambient temperatures in adjacent housing clusters and cultivation areas. Each farm is strategically positioned in relation to wells and cisterns, aligning with the spatial and hydrological principles outlined in the research development chapter.

142 |Design Development

Fig.146 Schematic Plan and Section for Condensation Farm

Design Development| 143

5.2 Water Storage & Community Spaces The cistern unit is conceived as both the hydrological core and the principal social hub of the settlement. Designed on a 40 × 40 metre grid, it integrates water storage with spaces for communal gathering, aligning infrastructural necessity with social programme. ts dual role is articulated through a vaulted roof system and subterranean platforms that enclose the water bodies, creating a spatial organisation capable of balancing environmental performance with cultural practices of gathering and privacy. Optimised through computational modelling, the cistern demonstrates how storage, shading, ventilation, and enclosure can converge into a single architectural typology that sustains both the settlement’s water cycle and its communal life.

144 |Design Development

Design Development| 145

5.2.1 Programmatic Distribution This architectural typology is intended to serve as the principal social and hydrological hub of the proposed settlement, taking the form of a cistern arranged on a 4 × 4 grid of 40 × 40 metres, covering a total area of 1,600 m². The layout of the programme was initiated by determining water storage requirements, accounting for a buffer period alongside a constant supply from local wells, and proportionally allocating the remaining space to accommodate the range of social activities outlined in the programme. As outlined in the Settlement Organisation chapter, each cistern is designed to supply water to three residential units, serving a total of 120 inhabitants, and to 1 acre of agricultural land divided into two units, for a two-week buffer period. Although a constant flow from the wells is anticipated, allowances were incorporated for potential pumping malfunctions or maintenance. This requirement translates into 1,100 m³ of water storage, defined as a pool of 320 m² with a depth of 3.5 metres. The remaining 1,280 m² of the cistern footprint is allocated to social spaces, equating to approximately 10 m² per person. These areas are structured into public, semi-public and private zones, each accommodating different functions and grouped according to their use. Their relationship to ground level is determined by the desired degree of connection with the immediate surroundings.

146 |Design Development

Fig.147 Programmatic Distribution for Cistern Units

Design Development| 147

5.2.2 Morphological Framework and Experimental Set-up The morphological attributes of the cistern are defined by two primary archetypes: an overground vaulted roof (the components of which were developed in the Research Development chapter) and an underground platform aggregation. Both are generated and iterated through concentric subdivisions of the rectangular grid, which are subsequently divided into two diagonal segments in varying orientations. For the vaulted structures, a system of arches is applied diagonally to the grid, aligned with its subdivisions, to produce clearly segmented roof partitions. These are spanned by vaults at intervals of 5.7 metres. The roof segments are inclined according to the length of each segment in relation to its subdivision group. Adjustable parameters include the overall height of the system and the width of the arches, allowing variations in proportion and spatial quality. The platform configuration begins with the random selection of grid parts within the two innermost concentric subdivisions to serve as water bodies that meet storage requirements. Maintenance facilities and promenade decks are then positioned around these water bodies, while remaining programmatic functions are distributed across a further random selection of grid rectangles. Once this distribution is set, the grid is extruded according to the dimensions of its subdivisions, generating the final three-dimensional arrangement.

Fig.148 Varying Design Parameters for Cistern Units

148 |Design Development

Performance Objectives and Optimisation The computational model of the cistern underwent multi-objective optimisation using Wallacei. The first objective was to minimise displacement in the roof structure, thereby ensuring stability under self-weight and wind loads. This was assessed through Finite Element Analysis conducted with Karamba 3D, incorporating the material properties obtained from the testing phase of this research. The second objective sought to minimise water evaporation by reducing exposure to solar radiation. Sun-ray vectors were extracted using Ladybug Tools for the Bahariya Oasis between April and September, and the proportion of the water surface under occlusion was set as a parameter to be minimised. The third objective was to maximise the compactness of each functional cluster, measured by calculating the mean distance between all elements within each cluster and minimising the overall sum. The final objective ensured minimal distances were maintained between clusters of the same programme, promoting spatial coherence across the system.

Fig.149 Objectives Set for the Optimisation of Cistern Units

Design Development| 149

Wallacei Optimisation Settings 1. Population Generation size: 50 Generation count: 100 Total population size: 5000

2. Algorithm Parameters Crossover probability: 0.9 Mutation probability: 1/n Crossover distribution index: 20 Mutation distribution index: 20 Random seed: 1

3. Simulation Parameters Number of genes (sliders): 9 Number of values (slider values): 12125 Number of fitness objectives: 4 Size of search space: 1.6 × 10¹8

Fig.150 A Selection of Pareto Front Phenotypes for Cistern Units

150 |Design Development

5.2.3 Optimisation Results and Morphology Selection Standard deviation graphs confirmed convergence across the four optimisation objectives. Structural stability (Objective 01) showed reduced variability as stable configurations emerged. Minimisation of solar exposure (Objective 02) displayed similar convergence. Compactness (Objective 03) began with broad variability, reflecting exploratory iterations, but stabilised as compact clusters were achieved. Spatial coherence (Objective 04) also narrowed progressively, confirming improved organisation of related functions. Collectively, the results demonstrate that the optimisation effectively refined the design towards balanced trade-offs across structural, environmental and spatial criteria. The most optimised morphology was selected by evaluating the Pareto front solutions, with particular emphasis on structural stability and the compactness of the zoning programme. In the selected configuration, the roof is divided into three zones, with the central one designed to have a shorter span. At ground level, two platforms are positioned at the outer perimeter of the building, while the remainder of the plan is maintained at lower elevations, enclosing the water bodies integrated into the design.

Fig.151 Selected Morphology for Cistern Units

SD - Graphs

FO-1

FO-2

FO-3

FO-4

Parallel Coordinate Graph

Design Development| 151

5.2.4 Post-Analysis Post-analysis confirmed that the structural deflection of the roof is maintained at an average minimum of 2 cm, demonstrating that the vaults provide an effective means of spanning large areas with materials working predominantly in compression. Furthermore, CFD analysis verified that the orientation of the structure plays a critical role in mitigating the impact of incoming winds from the south-east, which carry sandstorms, while simultaneously allowing prevailing winds from the northwest to enter, ensuring adequate ventilation within the cistern. Fig.152 Evaluation of Vault Deflection Using Karamba 3D

Fig.153 CFD Simulation longitudinal section run on North-West winds

Fig.154 CFD Simulation cross section run on North-West winds

152 |Design Development

The construction sequence of the cistern vault combines discrete earthen blocks with temporary timber centering trusses. The vault surface is assembled from functionally graded blocks, designed with dry interlocking joints to ensure stability without the need for mortar. During construction, a lightweight timber truss framework acts as centering, guiding the precise placement of the blocks and distributing loads until the vault is self-supporting. Once the structure stabilises under compression, the temporary framework is dismantled, leaving behind a durable, modular system capable of spanning large areas efficiently.

Discrete Functionally Graded Block

Dry Interlocking Joints

Temporary Timber Centering Truss

Fig.155 Construction Sequence for Vaults

Design Development| 153

5.2.5 Facade Development The architectural typology explored in this chapter develops a vaulted roof geometry combined with platform ensembles that operate as a unified spatial proposal. Yet, the connective tissue required to render this construct fully enclosed remains unresolved. To address this, the design team introduced mashrabiyas across all available openings of the structure, including entrances at ground level and vault apertures at roof level. The design principle involves subdividing each opening into a grid of smaller apertures, whose widths can be iteratively adjusted to regulate the amount of void within the structure. Collectively, this produces a mashrabiya system capable of fluctuating in void ratio across its surface, thereby offering a flexible means of modulating enclosure, light, and ventilation. Fig.156 Parameters for Mashrabiyas Variation

154 |Design Development

Following the modelling of the mashrabiyas, a multi-objective optimisation experiment was conducted using Wallacei. The objectives of this process were to minimise direct solar penetration into the interior, maximise cross-ventilation throughout the system, and secure the highest possible level of outward visual perception from eye level. The aperture of the sub-openings of the mashrabiyas was used as the iterative morphological parameter. The interplay of these parameters resulted in the cistern being articulated with optimised mashrabiya openings. As illustrated in Figure 157, the apertures vary visibly across different sectors of the surfaces, adapting their size and density according to orientation and elevation to accommodate the defined performance requirements.

Fig.157 Selected Facade

Fig.158 Optimisation Parameters for Mashrabiyas

Design Development| 155

5.2.6 Architectural Development The morphological format generated through the experiments was further developed into a fully articulated design proposal. The research team’s intention is to demonstrate that this typology can accommodate functional spaces that promote social interaction and foster a sense of community by creating a hospitable environment around the vital function of water storage in a desert context. The vaulted roof unifies all functions of the cistern into a cohesive whole, reinforcing this communal sentiment while providing a flexible ceiling height that was adjusted to emphasise either privacy or collective engagement, depending on programmatic requirements. The cistern unit is thus suggested as the principal public space of the settlement.

156 |Design Development

-5.00 m Lower Ground Floor Plan

+0.00 m Ground Floor Plan

Fig.159 Floor Plans for Optimised Cistern Morphology

Design Development| 157

158 |Design Development

Fig.160 3D Section Through Selected Cistern Morphology

Design Development| 159

5.3 Residential Clusters The housing units form the core of the settlement’s residential fabric, designed as clustered ensembles that integrate environmental performance with social organisation. Structured on a 40 × 40 metre grid, each cluster combines subterranean cooling, condensation towers, and courtyard typologies to balance private, communal, and environmental needs. The clusters are generated through voxel-based aggregation, allowing flexible adaptation to programmatic requirements while preserving a coherent spatial logic. Multi-objective optimisation guided the design process, refining solar shading, ventilation, compactness, and privacy to achieve a morphology that reinforces both climatic adaptation and community life.

160 |Design Development

Design Development| 161

5.3.1 Programmatic Distribution The residential units of the settlement are organised into clusters of five housing units, each composed of voxels measuring 2 × 2 × 5 metres and arranged on a 40 × 40 metre grid. The first floor of each unit is positioned 5 metres below ground level, while the second floor begins at ground level. Each cluster incorporates a condensation tower, designed to sustain a series of water bodies that provide cooling through evaporation. The housing units are divided into four distinct zones: the private quarters, a private courtyard enclosed within the geometry of the cluster, the public quarters, and a public courtyard. The aggregation of public courtyards from multiple clusters forms a larger communal public space. A peripheral circulation route is situated between the retaining wall and the clusters, facilitating access to the private quarters and enhancing air circulation. Two housing typologies are developed, accommodating either 4–5 residents or 6–8 residents, allowing flexibility for different family sizes. 162 |Design Development

Fig.161 Programmatic Distribution and Massin Logic for Housing Clusters

Design Development| 163

5.2.2 Parameters & Experiment Set-up Each cluster is generated by adding voxels according to programmatic requirements for one of the two typologies. Voxels are placed in unoccupied grid areas, excluding those reserved for the condensation tower and its associated water pools. The grid is first established at level -5 m and then repeated at ground level, with all positions available for voxel placement except those intersecting the tower’s vertical projection. Once the arrangement of water bodies and clusters is defined, two additional processes take place. First, five voxels within the interior of each cluster are removed to create private courtyards. Second, neighbouring voxels between clusters are removed to form narrow alleys linking the main public courtyards with the peripheral route. The second floors of the clusters are configured to cover at least 50% of unused lower-level areas, providing additional shading. Openings are introduced on the outermost faces of each cluster at both levels to improve natural lighting and ventilation. Finally, private spaces are assigned through a random allocation process, in which a defined number of voxels from the lower level and the entire upper level are designated for private use.

Fig.162 Design Parameters for Housing Clusters

164 |Design Development

Performance Objectives and Optimisation The computational model of the residential clusters underwent multi-objective optimisation using Wallacei, guided by six criteria. The first objective minimised solar insolation on water bodies, assessed through sun-ray vectors for April to September, generated with Ladybug Tools. The second applied the same method to residential clusters to maximise self-shading. The third addressed the orientation of openings to capture prevailing north-westerly winds, with simulations optimising window alignment for crossventilation. The fourth maximised the number of openings to enhance ventilation and indirect daylight penetration. The fifth minimised the mean distance between voxels within each cluster to ensure compactness. The sixth minimised unobstructed lines of sight between windows of private quarters, thereby maintaining privacy.

Fig.163 Optimisation Objectives for Housing Clusters

Design Development| 165

Wallacei Optimisation Settings 1. Population Generation size: 50 Generation count: 100 Total population size: 5000

2. Algorithm Parameters Crossover probability: 0.9 Mutation probability: 1/n Crossover distribution index: 20 Mutation distribution index: 20 Random seed: 1

3. Simulation Parameters Number of genes (sliders): 7 Number of values (slider values): 106007 Number of fitness objectives: 4 Size of search space: 1 × 1023

166 |Design Development

Fig.164 Pareto Front Solutions for Housing Clusters

Design Development| 167

5.2.3 Optimisation Results and Morphology Selection Standard deviation graphs demonstrated steady convergence across all six objectives. Solar exposure on water bodies and residential clusters (Objectives 01 and 02) was progressively reduced. Opening orientations and the number of openings (Objectives 03 and 04) improved consistently, while compactness and privacy (Objectives 05 and 06) also displayed clear refinement. Collectively, the optimisation confirmed the model’s ability to balance environmental, spatial, and social performance. The final cluster morphology, selected from the Pareto front solutions, exhibited particularly strong performance in compactness, privacy, and self-shading. The water network was designed to maintain direct contact with all residential units, while courtyard typologies offered varying degrees of enclosure, from fully private to indirectly connected spaces. The condensation tower was placed at the south-east corner, where it provides shading for the structure and enables the full extension of the water bodies. A gradient of privacy was achieved, with private spaces concentrated along the outer perimeter and public functions situated at the centre of the aggregation.

SD - Graphs

Parallel Coordinate Graph

168 |Design Development

FO-1

FO-2

FO-3

FO-4

FO-5

FO-6

Fig.165 Selected Phenotype for Housng Cluster

5.2.4 Post-Analysis Studies

Fig.166 Incident Radiation Simulatin

Fig.167 Direct Sun Hours Simulation

Further post-analysis using CFD and Ladybug Tools, based on the EPW weather file for Bahariya Oasis, assessed incident solar radiation and direct sun hours during peak summer months. Results demonstrated that the proposed aggregation significantly reduced solar exposure on key façades and roof surfaces while ensuring shaded outdoor and semioutdoor spaces. CFD simulations confirmed that the spatial configuration supported steady airflow, enhancing passive cooling potential and validating the environmental performance of the residential clusters. Fig.168 CFD Simulation Plan View

Design Development| 169

5.3.5 Architectural Development Having consolidated the housing cluster’s morphology, the design team advanced its architectural development into a fully functional residential cluster. The spaces were first organised into continuous vault geometries, which segmented the interior into distinct private and public quarters. This decision-making process was guided by the privacy index, which indicated the most appropriate positioning for each quarter, as illustrated in Figure 164. The vaults of the first floor were cast in alignment with the footprint of the upper floor using the “casting on vault” technique outlined in the Research Development chapter. Communal spaces, including the water surfaces, were programmatically interwoven with the public quarters of the residences in the form of reception spaces, thereby reinforcing the communal aspect of the cluster. Careful measures were taken to isolate these areas from the private spaces of the houses through the use of corridors, walls, and other architectural elements, thereby preserving their seclusion. The private courtyards of each residence were further refined, ensuring accessibility from all quarters and incorporating additional water pools. Mashrabiyas were applied to all openings identified by the iterative optimisation process, providing environmental comfort and privacy. Peripheral access to the unit was enhanced by staircases descending on all sides, which also created space for circulation at both underground and overground levels. As highlighted in the section shown in Figure 170, the vertical organisation of the cluster introduces urban filters that provide separation while also creating openings for communal activity.

170 |Design Development

-5.00 m Lower Ground Floor Plan

+0.00 m Ground Floor Plan

Fig.169 Floor Plans for Optimised Housing Cluster Typology

Design Development| 171

172 |Design Development

Fig.170 3D Section of Selected Housing Cluster

Design Development| 173

5.4 Agriculture Planning The agricultural units of the settlement are envisioned as multifunctional landscapes that merge food production, ecological restoration, and public amenity. Designed on a 40 × 40 metre footprint, each unit operates as both farmland and water infrastructure, recycling greywater from adjacent housing clusters through a stepped purification system. By combining plantations, water features, and pedestrian routes, these units extend beyond their productive role to form visitable green spaces, reinforcing the settlement’s self-sufficiency while fostering ecological resilience and social interaction.

174 |Design Development

Design Development| 175

5.4.1 Integrated Agriculture and Greywater Reuse The agricultural units are conceived as hybrid spaces that integrate productive landscapes, visitable green spaces, and water management systems. They are arranged on a 4 × 4 metre grid within a 40 × 40 metre footprint. Greywater produced by neighbouring housing units is directed into these clusters, where it undergoes purification through a sequence of five pools filled with gravel and papyrus plants. Each pool is set 0.5 metres lower in elevation than the previous one, forming a gravity-led qanat system. This natural filtration process provides an additional water source for date palm and wheat plantations, enabling the reuse of household effluent while generating ecological value. The morphology of each agricultural unit is dictated by the orientation of the inflowing greywater, which in turn structures the organisation of plantations and water features.

Housing Units

Agriculture Units

Grey Water Inflow Fig.171 Section of Urban Plan Illustrating Water Filteration Network

176 |Design Development

Fig.172 3D Section of Proposed Agriculture Plot

Design Development| 177

5.4.2 Parameters & Experiment Set-up A multi-objective optimisation experiment was conducted using Wallacei, with objectives divided into two categories. The first addressed topological parameters for optimising the flow and distribution of water through the purification pools. This was achieved by regulating the distances between pools to secure the optimal incline of the qanat system, while simultaneously minimising variation in spacing to ensure even distribution. The second category focused on ecological performance, evaluated using Rhino Ecologic. Here, parameters included solar exposure, site topography, and projected growth patterns over five years, to maximise biomass production. Rhino Ecologic was applied to the surfaces surrounding the purification pools, rendering them suitable for plantation. Its analysis produced a single value, the Flora Density Index, which quantified the biomass potential of each configuration; maximising this index was set as the experiment’s primary ecological objective.

Fig.173 Design Parameters for Agriculture Units

178 |Design Development

Performance Objectives and Optimisation

Fig.174 Optimisation Objectives for Agriculture Units

Design Development| 179

5.4.3 Optimsation Results and Selection The process yielded four principal typologies, each generated by optimisation runs corresponding to the four orientations of the incoming water flow. Among the Pareto front solutions, the outcomes with the highest Flora Density Index and the most cohesive water purification system were selected. The allocation of wheat and palm tree plantations in each typology was guided by the Flora Density Index, with high-biomass species such as palm trees occupying the upper ranges and wheat fields assigned to the lower ranges. Papyrus plantations remained concentrated within and around the purification pools. North - Flow Experiment SD - Graphs

FO-1 Plant Volume

FO-2 Optimal Water Flow

FO-3 Even Water Flow

Selected Phenotype

FO-4

South - Flow Experiment SD - Graphs

FO-1 Plant Volume

Selected Phenotype

180 |Design Development

FO-2 Optimal Water Flow

FO-3 Even Water Flow

Palm Trees Wheat Fields Papyrus Plantations Flora Density Index Low Density

High Density

East -Flow Experiment SD - Graphs

FO-1 Plant Volume

FO-2 Optimal Water Flow

FO-3 Even Water Flow

FO-1 Plant Volume

FO-2 Optimal Water Flow

FO-3 Even Water Flow

Selected Phenotype

West - Flow Experiment SD - Graphs

Selected Phenotype

Design Development| 181

5.4.5 Architectural Development Pedestrian pathways were introduced along the water flow, weaving circulation into the ecological infrastructure. This design approach enables agricultural units to function not only as productive fields but also as accessible green spaces, enhancing the social and environmental fabric of the settlement. As illustrated in Figures 175, this agricultural prototype represents a landscape design strategy in which water management practices actively shape the terrain and integrate purification, production, and public space into a unified outcome.

182 |Design Development

Fig.175 Section for Agriculture Plot

Design Development| 183

184 |Design Development

Design Development| 185

186 |Design Development

Design Development| 187

188 |Design Development

Design Development| 189

190 |Design Development

CONCLUSION Design Development| 191

6.1 Discussion This research set out to construct a diverse array of typologies, developed through multiple iterations, in order to address the urgent environmental challenges and cultural dimensions of desert dwelling in Egypt. The catalogue of units presented in Figure 176 provides an insight into the morphological, programmatic, and material explorations undertaken by the research team. Cisterns, housing clusters, agricultural fields, and condensation farms together form the mosaic that constitutes the backbone of the proposed prototypical settlement. Each unit and its successive iterations throughout the experimental process provide the architectural substance which, through the design team’s intent, is transformed into a realisable proposition.

192 |Conclusion

Fig.176 Catalogue of Units

Conclusion| 193

The urban-scale studies undertaken in this research aim to address the challenges of continuity that inevitably arise from a fragmented methodology such as the one employed in this study. Despite the considerable effort invested in shaping an integrated and cohesive urban landscape, tensions persist in terms of form and coherence. Why should housing units remain divided when they stand side by side? Should housing and agricultural fields not intertwine, binding property and labour into a seamless whole? What new possibilities might unfold when two social spaces, such as cisterns, are placed in close proximity? Many scenarios can be imagined. When housing units converge, they may generate larger courtyards, accommodate additional towers, or consolidate into more centralised clusters. When adjoining agricultural fields, the towers could extend their service to the plantations, allowing gardens to emerge from this union. When cisterns are paired, they form larger reservoirs, accompanied by more expansive communal spaces and more monumental structures rising above them.

194 |Conclusion

Fig.177 Clash Instances

Conclusion| 195

The clash catalogue in Figure 178 presents selected instances of collision and integration between different units under specific aggregation methods. Yet nothing prevents the conception of a more expansive matrix capable of accommodating larger and more complex ensembles. Such instances may be applied selectively according to environmental, social, economic, or productive needs. In this way, the manner of aggregation and the deliberate intent to merge become tools in the designer’s hands. For the design team, this serves a greater purpose: to frame the prototypical settlement not as a fixed scheme but as a transferable model. It embodies both the building blocks and the design strategies necessary to reconfigure and reconstruct the system into ensembles imbued with intent. Within the scope of this dissertation, that intent is clearly defined: a plan to resettle the deserts, emerging from the subterranean currents of the Bahariya Oasis.

196 |Conclusion

Fig.178 Clash Catalogue

Conclusion| 197

198 |Conclusion

Conclusion| 199

6.2 Appendix Appendix 1 Site Selection Training Data Aquifer Depth

Aquifer Shallowness

X Value

Y Value

220

20

-166142.7411

57501.95992

240

110

-156670.6981

67169.62038

220

70

-144401.6989

100204.6409

340

60

-149860.1678

109306.9157

140

30

-146210.9925

43060.58

120

80

-143556.724

45699.55556

150

45

-140588.9277

48258.91125

170

40

-138250.8195

50819.52814

150

40

-141970.1304

82799.31276

120

80

-139257.0798

80532.76978

140

110

-137435.2077

78957.19873

120

40

-134044.8086

76762.06016

100

50

-131138.3717

74852.54396

110

50

-128204.4244

72908.50069

180

25

-156974.321

94412.52891

140

35

-153712.8057

94375.67563

120

40

-150027.4777

94449.38219

100

50

-147042.362

95333.86092

70

55

-143596.5802

94430.95555

140

30

-140132.3719

93270.07721

120

20

-138787.2271

96936.97862

120

30

-135709.9782

96807.99214

200 |Conclusion

Appendix 2 Settlement Organization C# using System; using System.Collections; using System.Collections.Generic; using Rhino; using Rhino.Geometry; using Grasshopper; using Grasshopper.Kernel; using Grasshopper.Kernel.Data; using Grasshopper.Kernel.Types; using System.Linq; /// <summary> /// This class will be instantiated on demand by the Script component. /// </summary> public class Script_Instance : GH_ScriptInstance { #region Utility functions /// <summary>Print a String to the [Out] Parameter of the Script component.</summary> /// <param name=”text”>String to print.</param> private void Print(string text) { /* Implementation hidden. */ } /// <summary>Print a formatted String to the [Out] Parameter of the Script component.</summary> /// <param name=”format”>String format.</param> /// <param name=”args”>Formatting parameters.</ param> private void Print(string format, params object[] args) { /* Implementation hidden. */ } /// <summary>Print useful information about an object instance to the [Out] Parameter of the Script component. </ summary> /// <param name=”obj”>Object instance to parse.</ param> private void Reflect(object obj) { /* Implementation hidden. */ } /// <summary>Print the signatures of all the overloads of a specific method to the [Out] Parameter of the Script component. </summary> /// <param name=”obj”>Object instance to parse.</ param> private void Reflect(object obj, string method_name) { /* Implementation hidden. */ } #endregion #region Members /// <summary>Gets the current Rhino document.</ summary>

#region Members /// <summary>Gets the current Rhino document.</ summary> private readonly RhinoDoc RhinoDocument; /// <summary>Gets the Grasshopper document that owns this script.</summary> private readonly GH_Document GrasshopperDocument; /// <summary>Gets the Grasshopper script component that owns this script.</summary> private readonly IGH_Component Component; /// <summary> /// Gets the current iteration count. The first call to RunScript() is associated with Iteration==0. /// Any subsequent call within the same solution will increment the Iteration count. /// </summary> private readonly int Iteration; #endregion /// <summary> /// This procedure contains the user code. Input parameters are provided as regular arguments, /// Output parameters as ref arguments. You don’t have to assign output parameters, /// they will have a default value. /// </summary> private void RunScript(List<Point3d> P, List<int> S, List<Point3d> AttractorPoints, double WellNumber, double MarginTolerance, double Range1, double Range2, double Range3, double Margin1, double Margin2, double Margin3, bool Reset, ref object S_out) { if (P == null || S == null || P.Count != S.Count || WellNumber < 1) { S_out = null; return; } if (Reset || persistentStates == null || persistentStates.Count != S.Count) { persistentStates = new List<int>(S); S_out = new List<int>(persistentStates); return; } // Start fresh List<int> states = new List<int>(persistentStates); for (int i = 0; i < states.Count; i++) states[i] = 0; bool[] assigned = new bool[states.Count]; int wellCount = 0; Conclusion| 201

// --- NEW LOGIC BRANCH --// If Attractor points are provided, use the attraction logic. // Otherwise, fall back to the original farthest-point logic. if (AttractorPoints != null && AttractorPoints.Count > 0) { // 1. Score all potential well points based on proximity to the nearest attractor List<Tuple<int, double>> wellCandidates = new List<Tuple<int, double>>(); for (int i = 0; i < P.Count; i++) { double minAttractorDist = double.MaxValue; foreach (Point3d attractor in AttractorPoints) { double d = XYDistance(P[i], attractor); if (d < minAttractorDist) minAttractorDist = d; } wellCandidates.Add(new Tuple<int, double>(i, minAttractorDist)); } // Sort candidates by distance to an attractor (closest first) wellCandidates.Sort((a, b) => a.Item2.CompareTo(b. Item2)); // 2. Iterate through sorted candidates and try to build clusters foreach (var candidate in wellCandidates) { int wellIdx = candidate.Item1; if (assigned[wellIdx]) continue; // Skip if this point is already used // Find up to 5 unassigned Cisterns for this Well List<Tuple<int, double>> cands = new List<Tuple<int, double>>(); for (int i = 0; i < states.Count; i++) { if (assigned[i] || i == wellIdx) continue; double xyDist = XYDistance(P[i], P[wellIdx]); double dz = Math.Abs(P[i].Z - (P[wellIdx].Z + Margin3)); if (xyDist <= Range3) cands.Add(new Tuple<int, double>(i, dz)); } cands.Sort((a, b) => a.Item2.CompareTo(b.Item2)); List<int> cisterns = new List<int>(); for (int i = 0; i < 5 && i < cands.Count; i++) { cisterns.Add(cands[i].Item1); } if (cisterns.Count < 5) continue; // skip this well, cannot form a valid cluster // For each Cistern, assign 6 unassigned Housing and 2 unassigned Agri List<int> hTotal = new List<int>(); List<int> aTotal = new List<int>(); bool validCluster = true; foreach (int cisIdx in cisterns) 202 |Conclusion

{ // Housing List<Tuple<int, double>> hCands = new List<Tuple<int, double>>(); for (int i = 0; i < states.Count; i++) { if (assigned[i] || i == wellIdx || cisterns.Contains(i)) continue; double xyDist = XYDistance(P[i], P[cisIdx]); double dz = Math.Abs(P[i].Z - (P[cisIdx].Z + Margin2)); if (xyDist <= Range2 && dz <= MarginTolerance) hCands.Add(new Tuple<int, double>(i, dz)); } hCands.Sort((a, b) => a.Item2.CompareTo(b.Item2)); List<int> hChosen = new List<int>(); for (int i = 0; i < 6 && i < hCands.Count; i++) hChosen.Add(hCands[i].Item1); if (hChosen.Count < 6) { validCluster = false; break; } hTotal.AddRange(hChosen); // Agri List<Tuple<int, double>> aCands = new List<Tuple<int, double>>(); for (int i = 0; i < states.Count; i++) { if (assigned[i] || i == wellIdx || cisterns.Contains(i) || hTotal.Contains(i)) continue; // Check hTotal now double xyDist = XYDistance(P[i], P[cisIdx]); double dz = Math.Abs(P[i].Z - (P[cisIdx].Z + Margin2)); if (xyDist <= Range2 && dz <= MarginTolerance) aCands.Add(new Tuple<int, double>(i, dz)); } aCands.Sort((a, b) => a.Item2.CompareTo(b.Item2)); List<int> aChosen = new List<int>(); for (int i = 0; i < 2 && i < aCands.Count; i++) aChosen.Add(aCands[i].Item1); if (aChosen.Count < 2) { validCluster = false; break; } aTotal.AddRange(aChosen); } if (!validCluster) continue; // skip this well, couldn’t build a full cluster // If we got here, the cluster is valid. Assign it. states[wellIdx] = 1; assigned[wellIdx] = true; foreach (int ci in cisterns) { states[ci] = 2; assigned[ci] = true; } foreach (int hi in hTotal) { states[hi] = 3; assigned[hi] = true; } foreach (int ai in aTotal) { states[ai] = 4; assigned[ai] = true; } wellCount++; if (wellCount >= WellNumber) break; // We have enough wells } }

else {

} cands.Sort((a, b) => a.Item2.CompareTo(b.Item2)); List<int> cisterns = new List<int>(); for (int i = 0; i < 5 && i < cands.Count; i++) { cisterns.Add(cands[i].Item1); } } else break; }

// --- ORIGINAL LOGIC (Farthest Point) --// 1. Farthest Point Well Placement List<int> wells = new List<int>(); List<int> pool = new List<int>(); for (int i = 0; i < P.Count; i++) pool.Add(i); Random rand = new Random(); if (pool.Count > 0) { int first = pool[rand.Next(pool.Count)]; wells.Add(first); pool.Remove(first); while (wells.Count < WellNumber * 2 && pool.Count > 0) // Find more potential wells { double maxMinDist = -1; int bestIdx = -1; for (int i = 0; i < pool.Count; i++) { int idx = pool[i]; double minDist = double.MaxValue; for (int j = 0; j < wells.Count; j++) { double d = XYDistance(P[idx], P[wells[j]]); if (d < minDist) minDist = d; } if (minDist > maxMinDist) { maxMinDist = minDist; bestIdx = idx; } } if (bestIdx != -1) { wells.Add(bestIdx); pool.Remove(bestIdx); } } else break; } } // 2. Greedily Assign Clusters from the potential wells for (int wIdx = 0; wIdx < wells.Count; wIdx++) { int wellIdx = wells[wIdx]; // Find up to 5 unassigned Cisterns for this Well List<Tuple<int, double>> cands = new List<Tuple<int, double>>(); for (int i = 0; i < states.Count; i++) { if (assigned[i] || i == wellIdx) continue; double xyDist = XYDistance(P[i], P[wellIdx]); double dz = Math.Abs(P[i].Z - (P[wellIdx].Z + Margin3)); if (xyDist <= Range3) cands.Add(new Tuple<int, double>(i, dz));

} // 2. Greedily Assign Clusters from the potential wells for (int wIdx = 0; wIdx < wells.Count; wIdx++) { int wellIdx = wells[wIdx]; // Find up to 5 unassigned Cisterns for this Well List<Tuple<int, double>> cands = new List<Tuple<int, double>>(); for (int i = 0; i < states.Count; i++) { if (assigned[i] || i == wellIdx) continue; double xyDist = XYDistance(P[i], P[wellIdx]); double dz = Math.Abs(P[i].Z - (P[wellIdx].Z + Margin3)); if (xyDist <= Range3) cands.Add(new Tuple<int, double>(i, dz)); } cands.Sort((a, b) => a.Item2.CompareTo(b.Item2)); List<int> cisterns = new List<int>(); for (int i = 0; i < 5 && i < cands.Count; i++) { cisterns.Add(cands[i].Item1); } if (cisterns.Count < 5) continue; // skip this well // For each Cistern, assign 6 unassigned Housing and 2 unassigned Agri List<int> hTotal = new List<int>(); List<int> aTotal = new List<int>(); bool validCluster = true; foreach (int cisIdx in cisterns) { // Housing List<Tuple<int, double>> hCands = new List<Tuple<int, double>>(); for (int i = 0; i < states.Count; i++) { if (assigned[i] || wells.Contains(i) || cisterns.Contains(i)) continue; double xyDist = XYDistance(P[i], P[cisIdx]); double dz = Math.Abs(P[i].Z - (P[cisIdx].Z + Margin2)); if (xyDist <= Range2 && dz <= MarginTolerance) hCands.Add(new Tuple<int, double>(i, dz)); } hCands.Sort((a, b) => a.Item2.CompareTo(b.Item2)); List<int> hChosen = new List<int>(); for (int i = 0; i < 6 && i < hCands.Count; i++) hChosen.Add(hCands[i].Item1);

Conclusion| 203

if (hChosen.Count < 6) { validCluster = false; break; } hTotal.AddRange(hChosen); // Agri List<Tuple<int, double>> aCands = new List<Tuple<int, double>>(); for (int i = 0; i < states.Count; i++) { if (assigned[i] || wells.Contains(i) || cisterns.Contains(i) || hTotal.Contains(i)) continue; double xyDist = XYDistance(P[i], P[cisIdx]); double dz = Math.Abs(P[i].Z - (P[cisIdx].Z + Margin2)); if (xyDist <= Range2 && dz <= MarginTolerance) aCands.Add(new Tuple<int, double>(i, dz)); } aCands.Sort((a, b) => a.Item2.CompareTo(b.Item2)); List<int> aChosen = new List<int>(); for (int i = 0; i < 2 && i < aCands.Count; i++) aChosen.Add(aCands[i].Item1); if (aChosen.Count < 2) { validCluster = false; break; } aTotal.AddRange(aChosen); } if (!validCluster) continue; // skip this well // Assign Well and its cluster states[wellIdx] = 1; assigned[wellIdx] = true; foreach (int ci in cisterns) { states[ci] = 2; assigned[ci] = true; } foreach (int hi in hTotal) { states[hi] = 3; assigned[hi] = true; } foreach (int ai in aTotal) { states[ai] = 4; assigned[ai] = true; } wellCount++;

// Assign State 5 (max 2 per well) List<int> finalWells = IndicesByState(states, 1); foreach (int wellIdx in finalWells) { List<Tuple<int, double>> candidates = new List<Tuple<int, double>>(); for (int i = 0; i < states.Count; i++) { if (states[i] != 0) continue; double xyDist = XYDistance(P[i], P[wellIdx]); double zDist = Math.Abs(P[i].Z - P[wellIdx].Z); if (xyDist <= Range3 && zDist <= Margin3) { candidates.Add(new Tuple<int, double>(i, xyDist)); } } candidates.Sort((a, b) => a.Item2.CompareTo(b.Item2)); int assignedCount = 0; foreach (var candidate in candidates) { if (assignedCount >= 2) break; int candidateIdx = candidate.Item1; if (states[candidateIdx] == 0) { states[candidateIdx] = 5; assignedCount++; } } } persistentStates = new List<int>(states); S_out = states; } // <Custom additional code> // <Custom additional code> static List<int> persistentStates = null;

if (wellCount >= WellNumber) break; } } // --- POST-PROCESSING (Applies to both logic paths) --// Remove orphans (housing/agri not near any cistern) List<int> finalCisterns = IndicesByState(states, 2); for (int i = 0; i < states.Count; i++) { if (states[i] == 3 || states[i] == 4) { bool nearCistern = false; for (int j = 0; j < finalCisterns.Count; j++) { if (XYDistance(P[i], P[finalCisterns[j]]) <= Range2) { nearCistern = true; break; } } if (!nearCistern) states[i] = 0; } } 204 |Conclusion

List<int> IndicesByState(List<int> states, int target) { List<int> idxs = new List<int>(); for (int i = 0; i < states.Count; i++) if (states[i] == target) idxs.Add(i); return idxs; } double XYDistance(Point3d a, Point3d b) { return Math.Sqrt((a.X - b.X) * (a.X - b.X) + (a.Y - b.Y) * (a.Y - b.Y)); } // </Custom additional code> // </Custom additional code> }

Appendix 3 Material Mixing Quantities Methodology

MIX

Weight per cube (g)

Sand 432 Clay 225 Fibre 45 1 · Lime + Xanthan Lime 90 Xanthan 18 Water 90 Sand 428 Clay 223 2 · Sorel cement Fibre 45 (MgO + MgCl₂) + MgO 7 MgCl₂ 45 Xanthan Xanthan 18 Water 71 Sand 432 Clay 225 3 · Sodium-silicate + Fibre 45 Na₂SiO₃ 117 Xanthan Xanthan 18 Water 63 Sand 432 Clay 225 Fibre 45 4 · Lime + casein Lime 90 Casein 18 Water 90 Sand 432 Clay 225 5 · Lime + polymer Fibre 45 emulsion Lime 90 Polymer solids 27 Water 81 Sand 432 Clay 225 Fibre 45 6 · Lime + linseed oil Lime 90 Linseed oil 27 Water 81

Preparation protocol (immediately before charging the earth blend) Hydrate XG (2 min high-shear, Make 2 wt % xanthan sol (20 10 min stand). 1:1 lime putty (90 g L¨¹ at ~65 °C). Dose 0.90 L g Ca(OH)₂ + 90 mL H₂O), mellow to supply 18 g XG (~0.88 L of 10–15 min. Blend XG-sol into this is water counted within putty while stirring; keep pH≈12. the 90 g water). Fold into the dry earth; compact. If using 20 wt % MgCl₂ brine: Hydrate XG and dissolve into 45 g MgCl₂ needs 225 g brine, which brings ~180 g the MgCl₂ water. Sprinkle dry water do not add free water MgO into the liquor; stir ~60 s (work to consistency). Prefer (exotherm). Charge and compact solid MgCl₂ flakes to keep within ~8 min (rapid MOC set). water at ~71 g. Hydrate XG (2 wt %). For Option A, dry-blend powder silicate Anhydrous Na₂SiO₃ powder with earth; add XG-water as mix 117 g. add no free water water. For Option B, premix XGand expect a wetter mix. sol with diluted water-glass and fold into earth; no extra water. Disperse casein as 10 wt % in lime-water: 180 g dispersion Dissolve casein at pH≈10 (≈45 supplies 18 g casein (~162 °C). Make lime putty; blend cag is water, counted inside sein dispersion into putty; mix and the 90 g target reduce free compact. water accordingly). Using 50 %-solids latex: Mellow lime putty 10–15 min. weigh 54 g emulsion ( 27 Add latex slowly into putty (never g polymer + 27 g water). lime into latex). Fold into earth; Subtract 27 g from the 81 g avoid high-shear once latex is free-water allowance add present. 54 g extra water. Practical stock solution (count its carrier water)

Weigh 135 g emulsion ( 27 g oil + 108 g water) and add Pre-emulsify with ~0.5 wt % mild no free water (accept a wet- surfactant; count its water in the ter mix or offset with a little budget; mix gently. dry sand).

Conclusion| 205

Stage 01.1 Results

MIX

Initial Volume (cm3)

Final Volume (cm3)

Surface Area (cm2)

1 · Lime + Xanthan

500

371.07

39.06

344

40

500

424.175

44.65

568

70

500

465.696

47.04

591

70

4 · Lime + casein

500

419.616

44.64

557

43.75

5 · Lime + polymer emulsion

500

397.575

41.85

682

70

6 · Lime + linseed oil

500

428.64

44.65

567

70

2 · Sorel cement (MgO + MgCl₂) + Xanthan 3 · Sodium-silicate + Xanthan

MIX

F1 (KN)

Weight Force (KN/m3)

F2 (KN)

Stress (KN/cm2)

1 · Lime + Xanthan

0.0033712

9.09

0.392

0.0100

0.0055664

13.12

0.686

0.0154

0.0057918

12.44

0.686

0.0146

4 · Lime + casein

0.0054586

13.01

0.42875

0.0096

5 · Lime + polymer emulsion

0.0066836

16.81

0.686

0.0164

6 · Lime + linseed oil

0.0055566

12.96

0.686

0.0154

2 · Sorel cement (MgO + MgCl₂) + Xanthan 3 · Sodium-silicate + Xanthan

206 |Conclusion

Material Weight External Weight (g) (kg)

Volume m3)

Stage 01.2. Results

MIX

Sorel cement

Silicate stabilized earth

Polymer enhanced earth

Sand/ Gravel (%)

Clay (%)

Fiber (%)

Sorel cement

208 |Conclusion

Silicate stabilized earth

Polymer enhanced earth

Oil treated earth

initial Volume (cm3)

Additives (wt%)

50%

30%

5%

50%

30%

2%

50%

30%

5%

50%

30%

2%

50%

30%

5%

50%

30%

2%

50%

30%

5%

50%

30%

2%

50%

30%

5%

50%

30%

5%

50%

30%

5%

50%

30%

5%

Final Volume (cm3)

Surface Material Area (cm2) Weight (g)

External Weight (kg)

F1 (KN)

187

44.65

250

70

0.00245 13.10

0.686

0.0154

189

47.04

258

70

0.00253 13.38

0.686

0.0146

188

44.64

248

70

0.00243 12.93

0.686

0.0154

186

41.85

252

70

0.00247

13.28

0.686

0.0164

178

44.65

568

70

0.00216

12.11

0.686

0.0154

177

47.04

591

70

0.00235

13.29

0.686

0.0146

164

44.64

557

40

0.00245 14.94

0.392

0.0088

158

41.85

682

63.5

0.00225

14.27

0.6223

0.0149

178.695

44.65

258

70

0.00253 14.15

0.686

0.0154

188.328

47.04

234

70

0.00229

12.18

0.686

0.0146

166.32

44.65

221

70

0.00217

13.02

0.686

0.0154

154.548

47.04

200

70

0.00196

12.68

0.686

0.0146

Oil treated earth

MIX

Binder (%) 8% + 4% MgO + MgCl2

Sodium Silicate

2%

216

2%

216

2%

216

10% + 5%

2%

216

10%

2%

216

2%

216

2%

216

2%

216

3%

216

3%

216

2%

216

2%

216

8% + 4% 10% + 5%

10%

Xanthan Gum

Xanthan Gum

12% 12%

MgO + MgCl2 Sodium Silicate MgO + MgCl2 Sodium Silicate

8% + 4% 12% 8% + 4% 12%

Polymer Emulsion

Linseed Oil

So

Si ea

Po en

O

Weight Force Stress F2 (KN) (KN/m3) (KN/cm2)

Conclusion| 207

Appendix 4

Road Network C# using System; using System.Collections; using System.Collections.Generic; using System.Linq; using Rhino; using Rhino.Geometry; using Grasshopper; using Grasshopper.Kernel; using Grasshopper.Kernel.Data; using Grasshopper.Kernel.Types; public class Script_Instance : GH_ScriptInstance { #region Utility functions private void Print(string text) { /* Implementation hidden. */ } private void Print(string format, params object[] args) { /* Implementation hidden. */ } private void Reflect(object obj) { /* Implementation hidden. */ } private void Reflect(object obj, string method_name) { /* Implementation hidden. */ } #endregion #region Members private readonly RhinoDoc RhinoDocument; private readonly GH_Document GrasshopperDocument; private readonly IGH_Component Component; private readonly int Iteration; #endregion private void RunScript( List<Point3d> P, List<int> S_in, double StepDistance, double DetourTolerance, ref object S_out, ref object C_out) { // 1. INPUT VALIDATION if (P == null || S_in == null || P.Count != S_in.Count) { Component.AddRuntimeMessage( Grasshopper.Kernel.GH_RuntimeMessageLevel. Warning, “Input P and S_in must be valid and have the same number of items.”); return; } if (DetourTolerance < 1.0) DetourTolerance = 1.0;

208 |Conclusion

if (StepDistance <= 0) StepDistance = 40.0; // 2. INITIAL SETUP List<int> newStates = new List<int>(S_in); List<Curve> connectionCurves = new List<Curve>(); List<int> state1Indices = GetIndicesByState(newStates, 1); List<int> state0Indices = GetIndicesByState(newStates, 0); if (state1Indices.Count < 2) { Component.AddRuntimeMessage( Grasshopper.Kernel.GH_RuntimeMessageLevel. Remark, “Fewer than two ‘State 1’ points found. No curves to create.”); S_out = newStates; C_out = connectionCurves; return; } // 3. FIND CLOSEST NEIGHBORS AND CREATE CONNECTIONS var createdConnections = new HashSet<Tuple<int, int>>(); foreach (int startIdx in state1Indices) { var neighbors = new List<Tuple<int, double>>(); foreach (int candidateIdx in state1Indices) { if (startIdx == candidateIdx) continue; double dist = P[startIdx].DistanceTo(P[candidateIdx]); neighbors.Add(new Tuple<int, double>(candidateIdx, dist)); } neighbors.Sort((a, b) => a.Item2.CompareTo(b.Item2)); int neighborsToConnect = Math.Min(2, neighbors. Count); for (int i = 0; i < neighborsToConnect; i++) { int goalIdx = neighbors[i].Item1; var connectionPair = new Tuple<int, int>( Math.Min(startIdx, goalIdx), Math.Max(startIdx, goalIdx)); if (createdConnections.Contains(connectionPair)) continue; createdConnections.Add(connectionPair); // Pathfinding Logic (Direct vs. Detour) Point3d ptA = P[startIdx]; Point3d ptB = P[goalIdx]; double directDist = ptA.DistanceTo(ptB); int bestDetourIdx = -1; double bestDetourDist = double.MaxValue; if (state0Indices.Count > 0) { foreach (int idx0 in state0Indices)

{

} double detourDist = ptA.DistanceTo(P[idx0]) + P[idx0].DistanceTo(ptB); if (detourDist < bestDetourDist) { bestDetourDist = detourDist; bestDetourIdx = idx0; }

} // 5. PROCESS STATE CHANGES AND REPLACEMENTS if (pointsToEliminate.Count > 0) { foreach (int idxToEliminate in pointsToEliminate.Keys) { newStates[idxToEliminate] = 0; }

} } // Choose detour if shorter (within DetourTolerance) if (bestDetourIdx != -1 && bestDetourDist <= directDist * DetourTolerance) { Polyline detourPolyline = new Polyline { ptA, P[bestDetourIdx], ptB }; connectionCurves.Add(detourPolyline.ToNurbsCurve()); } else { connectionCurves.Add(new LineCurve(ptA, ptB)); } } }

List<int> availableReplacementIndices = GetIndicesByState(newStates, 0); foreach (var eliminated in pointsToEliminate) { int eliminatedIdx = eliminated.Key; int originalState = eliminated.Value; Point3d eliminatedPt = P[eliminatedIdx]; if (availableReplacementIndices.Count == 0) break; int replacementIdx = -1; double minReplacementDist = double.MaxValue; foreach (int availIdx in availableReplacementIndices) { double dist = eliminatedPt.DistanceTo(P[availIdx]); if (dist < minReplacementDist) { minReplacementDist = dist; replacementIdx = availIdx; } }

// 4. COLLISION CHECK ALONG CURVE PATHS Dictionary<int, int> pointsToEliminate = new Dictionary<int, int>(); foreach (Curve curve in connectionCurves) { double curveLength = curve.GetLength(); for (double d = 0; d <= curveLength; d += StepDistance) { Point3d testPoint = curve.PointAtLength(d); int bestCandidateIndex = -1; double minDistance = double.MaxValue; for (int i = 0; i < P.Count; i++) { int currentState = newStates[i]; if (currentState == 3 || currentState == 4 || currentState == 5) { double distToTestPoint = P[i].DistanceTo(testPoint); if (distToTestPoint < minDistance) { minDistance = distToTestPoint; bestCandidateIndex = i; }} } if (bestCandidateIndex != -1 && !pointsToEliminate.ContainsKey(bestCandidateIndex)) { pointsToEliminate.Add(bestCandidateIndex, newStates[bestCandidateIndex]); }

if (replacementIdx != -1) { newStates[replacementIdx] = originalState; availableReplacementIndices.Remove(replacementIdx); } } } // 6. ASSIGN OUTPUTS S_out = newStates; C_out = connectionCurves; } private List<int> GetIndicesByState(List<int> states, int targetState) { List<int> indices = new List<int>(); for (int i = 0; i < states.Count; i++) { if (states[i] == targetState) indices.Add(i); } return indices; } } Conclusion| 209

Appendix 4 Atmospheric Water Harvesting Training Data

Gene | Phenotype Phenotype 01 Phenotype 02 Phenotype 03 Phenotype 04 Phenotype 05 Phenotype 06 Phenotype 07 Phenotype 08 Phenotype 09 Phenotype 10 Phenotype 11 Phenotype 12 Phenotype 13 Phenotype 14 Phenotype 15 Phenotype 16 Phenotype 17 Phenotype 18 Phenotype 19 Phenotype 20

210 |Conclusion

Gene 00 (radius) 2.1 2.5 2.7 2.9 2.1 2.6 2.6 2.7 2.8 2.8 2.5 2.8 2.6 2.1 2.6 2.9 2.1 3 2.8 2.6

Gene 01

Gene 02 (height)

Gene 03

8 6 10 6 5 6 9 10 5 11 7 9 10 6 4 4 6 11 4 4

30 32 32 31 28 32 24 25 26 31 21 30 30 25 26 22 26 31 26 27

0.64 0.46 0.63 0.52 0.42 0.47 0.69 0.47 0.68 0.63 0.45 0.63 0.41 0.56 0.5 0.66 0.47 0.56 0.42 0.41

Gene 04 0.024 0.025 0.01 0.015 0.018 0.01 0.014 0.001 0.014 0.015 0.024 0.007 0.025 0.007 0.025 0.017 0.025 0.022 0.017 0.001

Gene 05

Gene 06

-5 -4 -4 -5 -4 -5 -4 -4 -4 -5 -4 -4 -6 -4 -5 -5.5 -5.8 -3.2 -5.4 -3.8

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0. 0.0

ne 06 0.011 0.012 0.012 0.015 0.002 0.007 0.012 0.013 0.004 0.005 0.011 0.009 0.009 0.011 0.007 0.008 0.006 0.007 0.01 0.011

Gene 07 0.56 0.63 0.41 0.54 0.7 0.41 0.59 0.7 0.54 0.6 0.42 0.43 0.45 0.61 0.4 0.46 0.49 0.65 0.47 0.55

Gene 08 (width)

Gene 09 (height)

0.38 0.42 0.43 0.44 0.32 0.36 0.26 0.41 0.46 0.22 0.24 0.24 0.5 0.41 0.36 0.36 0.28 0.46 0.42 0.25

0.7 0.7 0.8 0.8 0.9 0.7 1 0.5 0.7 0.6 0.5 0.7 1 0.6 0.5 0.6 0.7 0.6 0.6 0.5

Gene 10 Gene 11 3 2 5 4 3 1 0 2 2 4 0 2 4 4 3 5 5 4 2 5

2 0 3 3 1 1 2 2 5 1 3 2 2 2 2 2 0 1 1 3

Gene 12 0.57 0.4 0.38 0.5 0.43 0.5 0.41 0.32 0.45 0.42 0.41 0.58 0.5 0.42 0.5 0.61 0.53 0.39 0.43 45

CFD RESULT 4.5 m/s 5.5 m/s 4 m/s 3.5 m/s 6.5 m/s 1.5 m/s 2.5 m/s 1.2 m/s 3 m/s 6 m/s 1.2 m/s 3.5 m/s 6.8 m/s 3 m/s 4.5 m/s 2.5 m/s 4 m/s 6.7 m/s 2.5 m/s 3.5 m/s

Conclusion| 211

6.3 Bibliography Abdalla, Ahmed, Till Stellmacher, and Mathias Becker. 2022. “Trends and Prospects of Change in Wheat Self-Sufficiency in Egypt.” Agriculture 13 (1): 7. https://doi.org/10.3390/agriculture13010007. Alabsi, Akram Ahmed Noman, Yue Wu, Auwalu Faisal Koko, Khaled Mohammed Alshareem, and Roknizadeh Hamed. 2021. “Towards Climate Adaptation in Cities: Indicators of the Sustainable Climate-Adaptive Urban Fabric of Traditional Cities in West Asia.” Applied Sciences 11 (21): 10428. https://doi.org/10.3390/app112110428. Al Asali, M. Wesam. 2022. “Vaulting Cultures in the Modern Middle East.” ANTA: Archives of New Traditional Architecture 3 (Spring): 88–95. Bagasi, Abdullah Abdulhameed, John Kaiser Calautit, and Abdullah Saeed Karban. 2021. “Evaluation of the Integration of the Traditional Architectural Element Mashrabiya into the Ventilation Strategy for Buildings in Hot Climates.” Energies 14 (3): 530. https://doi.org/10.3390/en14030530. Bagheri, Pouyan, Ivan Gratchev, and Maksym Rybachuk. 2023. “Effects of Xanthan Gum Biopolymer on Soil Mechanical Properties.” Applied Sciences 13 (2). https://doi.org/10.3390/app13020887. Bassily, Eman Faiez Maher, and Tamer Refaat. 2021. “The Features and Characteristics of Desert Societies in Spontaneous Oasis Architecture: A Case Study of Al-Bawiti Low Surface, Bahariya Oasis.” International Journal of Applied Science and Research 4 (5): 268–87. Beazley, Elisabeth, and Michael Harverson. 1982. Living with the Desert: Working Buildings of the Iranian Plateau. Warminster: Aris & Phillips. Block, Philippe, and John Ochsendorf. 2007. “Thrust Network Analysis: A New Methodology for Three-Dimensional Equilibrium.” International Journal of Solids and Structures 44 (16): 4342–58. Block Research Group, eds. 2017. Beyond Bending: Reimagining Compression Shells. Munich: DETAIL. https://doi. org/10.11129/9783955533915. Brzyski, Przemysław, Zbigniew Suchorab, and Grzegorz Łagód. 2021. “The Influence of Casein Protein Admixture on Pore Size Distribution and Mechanical Properties of Lime-Metakaolin Paste.” Buildings 11 (11). https://doi.org/10.3390/ buildings11110530. Burrell, A. L., J. P. Evans, and M. G. De Kauwe. 2020. “Anthropogenic Climate Change Has Driven over 5 Million Km² of Drylands towards Desertification.” Nature Communications 11 (1): 3853. https://doi.org/10.1038/s41467-020-17710-7. Damluji, Salma Samar, and Viola Bertini. 2018. Hassan Fathy: Earth & Utopia. London: Laurence King Publishing. de Vries, Andries Jan, Moshe Armon, Klaus Klingmüller, Raphael Portmann, Matthias Röthlisberger, and Daniela I. V. Domeisen. 2024. “Breaking Rossby Waves Drive Extreme Precipitation in the World’s Arid Regions.” Preprint, February 2024. Elnaggar, Abdelhamid. 2014. “Environmental Sensitivity to Desertification in Bahariya Oasis, Egypt.” Egyptian Soil Sci. Soc. J. (ESSSJ) 16. Elnaggar, Abdelhamid, Kh El-Hamdi, Abdelaziz Belal, and Mohamed Kafrawy. 2013. “Soil Classification of Bahariya Oasis Using Remote Sensing and GIS Techniques.” Journal of Soil Science and Agricultural Engineering, Mansoura University 4 (September): 921–47. https://doi.org/10.21608/jssae.2013.52488.

212 |Conclusion

Fathy, Hassan. 1973. Architecture for the Poor: An Experiment in Rural Egypt. Chicago: University of Chicago Press. Filippi, Francesca De. 2006. Traditional Architecture in the Dakhleh Oasis, Egypt: Space, Form and Building Systems. Gao, Xuejie, and Filippo Giorgi. 2008. “Increased Aridity in the Mediterranean Region under Greenhouse Gas Forcing Estimated from High Resolution Simulations with a Regional Climate Model.” Global and Planetary Change 62 (June): 195–209. https:// doi.org/10.1016/j.gloplacha.2008.02.002. Gobinath, Ravindran, Isaac Ibukun Akinwumi, Olaniyi Diran Afolayan, et al. 2020. “Banana Fibre-Reinforcement of a Soil Stabilised with Sodium Silicate.” Silicon 12 (2): 357–63. https://doi.org/10.1007/s12633-019-00124-6. Grigoriadis, Kostas, ed. 2016. Mixed Matters: A Multi-Material Design Compendium. Berlin: Jovis. Haaf, W. (1984). Solar chimneys: part II: preliminary test results from the Manzanares pilot plant. International Journal of Solar Energy, 2(2), 141–161. Hamdan, Ali, and Rashad Sawires. 2011. “Hydrogeological Studies on the Nubian Sandstone Aquifer in El-Bahariya Oasis, Western Desert, Egypt.” Arabian Journal of Geosciences 6 (May). https://doi.org/10.1007/s12517-011-0439-8. Hamidian, Ali, Mehdi Ghorbani, Mahsa Abdolshahnejad, and Aziz Abdolshahnejad. 2015. “Qanat, Traditional Eco-Technology for Irrigation and Water Management.” Agriculture and Agricultural Science Procedia 4: 119–25. https://doi.org/10.1016/j. aaspro.2015.03.014. Krisst, R. (1983). Solar chimney for power generation. Applied Energy, 13(2), 83–100. Klemm, O., et al. (2012). Fog as a fresh-water resource: Overview and perspectives. Ambio, 41(3), 221–234. Lawrence, M. (2005). The relationship between relative humidity and the dew point temperature in moist air: A simple conversion and applications. Bulletin of the American Meteorological Society, 86(2), 225–234. Lee, Jong-Kook, Jee-Eun Lee, Seong-Cheol Park, et al. 2015. “A Study on Water-Repellent Effectiveness of Natural Oil-Applied Soil as a Building Material.” Open Journal of Civil Engineering 5 (January): 139–48. https://doi.org/10.4236/ojce.2015.51014. Manning, J. G. n.d. Water, Irrigation and Their Connection to State Power in Egypt. Margolis, Liat, Aziza Chaouni, and Rahul Mehrotra, eds. 2014. Out of Water: Design Solutions for Arid Regions. Baden: Lars Müller Publishers. Minke, Gernot. 2006. Building with Earth. Basel: Birkhäuser. https://doi.org/10.1007/3-7643-7873-5. Moghazy, Noha H., and Jagath J. Kaluarachchi. 2020a. “Assessment of Groundwater Resources in Siwa Oasis, Western Desert, Egypt.” Alexandria Engineering Journal 59 (1): 149–63. https://doi.org/10.1016/j. aej.2019.12.018. 2020b. “Sustainable Agriculture Development in the Western Desert of Egypt: A Case Study on Crop Production, Profit, and Uncertainty in the Siwa Region.” Sustainability 12 (16): 6568. https://doi.org/10.3390/su12166568.

Conclusion| 213

Mohamed, Ahmed, Ezzat Ahmed, Fahad Alshehri, and Ahmed Abdelrady. 2022. “The Groundwater Flow Behavior and the Recharge in the Nubian Sandstone Aquifer System during the Wet and Arid Periods.” Sustainability 14 (11): 6823. https://doi. org/10.3390/su14116823. Mohamed, Amr, Ali Ali, and Ahmed Abd El-Ghany. 2017. “Irrigation Water Management of Date Palm under El-Baharia Oasis Conditions.” Egyptian Journal of Soil Science 0 (0): 0–0. https://doi.org/10.21608/ejss.2017.1609.1123. Mora-Ruiz, Viviana, Jonathan Soto-Paz, Shady Attia, and Cristian Mejía-Parada. 2025. “Sustainable Earthen Construction: A Meta-Analytical Review of Environmental, Mechanical, and Thermal Performance.” Buildings 15 (6): 918. https://doi. org/10.3390/buildings15060918. Nasla, S., K. Gueraoui, Mohammed Cherraj, et al. 2021. “An Experimental Study of the Effect of Pine Needles and Straw Fibres on the Mechanical Behaviour and Thermal Conductivity of Adobe Earth Blocks with Chemical Analysis.” JP Journal of Heat and Mass Transfer 23 (June). https://doi.org/10.17654/HM023010035. Rabeh, Taha, Said Bedair, and Mohamed Abdel Zaher. 2018. “Structural Control of Hydrogeological Aquifers in the Bahariya Oasis, Western Desert, Egypt.” Geosciences Journal 22 (1): 145–54. https://doi.org/10.1007/s12303-016-0072-3. Rawat, Sanket, Paul Saliba, Peter C. Estephan, Farhan Ahmad, and Yixia Zhang. 2024. “Mechanical Performance of Hybrid FibreReinforced Magnesium Oxychloride Cement Schlaich, J., Bergermann, R., Schiel, W., & Weinrebe, G. (1995). Design of commercial solar updraft tower systems—utilisation of solar induced convective flows for power generation. Journal of Solar Energy Engineering, 127(1), 117–124. Based Composites at Ambient and Elevated Temperature.” Buildings 14 (1):. https://doi.org/10.3390/buildings14010270. Sahebzadeh, Sadra, Abolfazl Heidari, Hamed Kamelnia, and Abolfazl Baghbani. 2017. “Sustainability Features of Iran’s Vernacular Architecture: A Comparative Study between the Architecture of Hot–Arid and Hot–Arid–Windy Regions.” Sustainability 9 (5): 749. https://doi.org/10.3390/su9050749. Shakibamanesh, Amir. 2017. “Assessing the Value of Qanat System of Yazd in Promoting Urban Climate Resilience.” International Journal of Architectural and Environmental Engineering 11 (1): 113–24. Torelli, Giacomo, Mar Giménez Fernández, and Janet M. Lees. 2020. “Functionally Graded Concrete: Design Objectives, Production Techniques and Analysis Methods for Layered and Continuously Graded Elements.” Construction and Building Materials 242 (May): 118040. https://doi.org/10.1016/j.conbuildmat.2020.118040. Yenice, Yagmur, and Daekwon Park. 2019. “V-INCA: Designing a Smart Geometric Configuration for Dry-Masonry Wall.” Proceedings of the 39th Annual Conference of the Association for Computer Aided Design in Architecture (ACADIA). Yu, Rongjin, and Pavel Kabele. 2019. “Functionally Graded Concrete: A Review of Materials, Production Methods, and Structural Applications.” Cement and Concrete Composites 104: 103–22. https://doi.org/10.1016/j.cemconcomp.2019.103122.

214 |Conclusion

Conclusion| 215

6.4 List of Figures Fig.01 Water Extraction in the Bahariya Oasis; Photographed by DPA Fig.02 Agriculture Field in The Bahariya Oasis; Photographed by Ems-Foster Productions Fig.03 Aridity Index Classification; Created by Author; According to the Map Published by the UN Environmental Programme Fig.04 Risk of Human-Induced Desertification; Created by Author; According to the Map Published by the US Department of Agriculture Fig.05 Water Scarcity Vulnerability Classification; Created by Author; According to the projections of the IWMI for 2050 (increase in temperature 2.8-4.6 °C by 2100) Fig.06 Simplified Global Groundwater Resources; Created by Author; adapted from UNESCO–IAH groundwater atlas Fig.07 Aerial View of Cairo, Egypt; Photographed by James-L-Stanfield Fig.08 Populated land in Egypt; Created by Author; Adapted from Google Earth Pro Data Fig.09 Depletion of Water Sources in Egypt; Photographed by Green Prophet Fig.10 Water Consumption and Resources in Egypt; Created by Author; According to data obtained from the Egyptian Central Agency for Public Mobilisation and Statistics (CAPMAS) Fig.11 Egypt’s Map Highlighting Existing Oases and Major Cities; Created by Author; Adapted from Google Earth Pro Data Fig.12 Boundaries of the Nubian Sandstone Aquifer; Created by Author; Modified from Thorweihe (1990) Fig.13 Map of The Bahariya Oasis, Highlighting Aquifer Qualities; Created by Author; According to data obtained from Rabea (2018) Fig.14 The Economic Structure of Bahariya Oasis; Created by Author; based on data from Central Agency for Public Mobilisation and Statistics (CAPMAS), 2022 Fig.15 The English Volcanic Mountains, Bahariya; Extracted from Flickr; Photographed by Unknown Fig.16 Al-Bawitie Village, Bahariya; Extracted from Encyclopedia Britannica; Photographed by Unknown Fig.17 Camping Sites along the Salt Mountains, Bahariya; Extracted by TripAdvisor; Photographed by Unknown Fig.18 Settlement Analysis Map; Created by Author; Adapted from Google Earth Pro Data Fig.19 Monthly Average Temperature and Humidity in Bahariya Oasis; Created by Author; based on data from the Egyptian Meteorological Authority Fig.20 Predominant Wind Vector Direction Over the Bahariya Oasis; Created by Author, referencing Wind Map Data Service. Accessed [2025]. https://www.windy.com Fig.21 Sand Storm over Al-Bawitie Village, Bahariya; Extracted from sawti.com; Photographed by Unknown Fig.22 Sand Storm over Agriculture Fields, Egypt; Extracted from natureunfolding.wordpress.com; Photographed by D Varro Fig.23 Effect of Sandstorms on Urban Landscape; Cairo, Egypt; Extracted from byozarab.media; Photographed by Unknown Fig.24 Khamseen Wind Vector Direction Over the Bahariya Oasis; Created by Author, referencing Wind Map Data Service. Accessed [2025]. https://www.windy.com Fig.25 Mountain Valleys in Bahariya; Photographed by “Egypt Travel Gate” Fig.26 Surface Water Within Low Terrain Areas, Bahariya; Extracted from marsaalamtours.com; Photographed by Unknown Fig.27 The White Desert; Salt Mountains in Bahariya; Extracted from viator.com; Photographed by Unknown Fig.28 Topography Analysis on QGIS Mesh of Bahariya’s Landscape; Created by author based on the NASA SRTM DEM (30 m resolution, 2000) Fig.29 Run-Off Analysis on QGIS Mesh of Bahariya’s Landscape; Created by author based on the NASA SRTM DEM (30 m resolution, 2000) Fig.30 Concavity Analysis on QGIS Mesh of Bahariya’s Landscape; Created by author based on the NASA SRTM DEM (30 m resolution, 2000) Fig.31 Hassan Fathy’s “Hassan Rashad House, Tanta” Floor Plan; Created by author based on Steele, An Architecture for People: The Complete Works of Hassan Fathy (1997) Fig.32 New Gourna Village in Luxor, Egypt; extracted from researchgate.net; Photographed by Unknown; Source: Wael A Yousuf, The challenge of sustainability in developing countries and the adaptation of heritage-inspired architecture in context Fig.33 New Baris Village in New Valley, Egypt; extracted from researchgate.net; Photographed by Unknown; Source: Ayah Ramadan, Towards Low Energy Buildings through a Prototype of Desert Rural House in Alwadii Algadid in Egypt Fig.34 New Gourna Village Plan in Luxor, Egypt; Created by author based on the site plan by Hassan Fathy, ca. 1946–52; Digital Collections, The American University in Cairo Fig.35 New Baris Village Plan in New Valley, Egypt; Created by author based on the site plan by Hassan Fathy, 1967; MIT Libraries Visual Collections Fig.36 Aerial View of Yazd, Iran; Extracted from istockphoto.com; Photographed by zanskar Fig.37 A Close-up of Houses in Yazd, Iran; Extracted from Encyclopedia Britannica; Photographed by Unknown Fig.38 Narrow, Covered Alleyways in Yazd, Iran; Extracted from researchgate.net; Photographed by Unknown; Source: Ahad Mohammadi et al., The most suitable walking route for tourism through the historical-cultural fabric of Yazd City Fig.39 Urban Analysis of Yazd, Iran; Created by the author based on Google Earth Pro imagery Fig.40 A Qanat Interior; Extracted from flickr.com; Photographer Unknown 216 |Conclusion

Fig.41 An Ab-anbar in the Central Desert City of Naeen, Iran; extracted from researchgate.net; Photographed by Unknown; Source: Fatemeh Afsahhosseini, The impact of Iran’s urban heritage on sustainability, climate change and carbon zero Fig.42 The Canals of A Persian Aqueduct System in Iran; Extracted from orienttrips.com; Photographed by Unknown Fig.43 A Study of the Ab-Anbar System Fig.44 Qanat System: Water Distribution from Aquifer to Household; Created by author based on descriptions and diagrams of Ab-Anbar cistern and qanat systems in Iran Fig.45 Wadi-Hanifeh Water Distribution Project by Thomson Consultants Fig.46 Vena 1 Project by ORE, Copper-Alloy Water Condensation System Fig.47 Schematic Diagram of Dow To Earth Cistern System; Created by author, adapted from Out of Water (2015) Fig.48 Schematic Diagram of Cyclical Water Systems; Created by author, adapted from Out of Water (2015) Fig.49 House in Dakheh Oasis, Egypt; Extracted from flickr.com; Photographed by Unknown Fig.50 Wooden Lattice Screens (Mashrabiyas); Extracted from x.com/MuslimCulture; Photographed by Waheed Sobhi Fig.51 Floor Plans for a House in Dakhleh Oasis; Created by author based on field plans in Traditional Architecture in the Dakhleh Oasis, Egypt: Space, Form and Building Systems (2011) Fig.52 Wind towers in Yazd, Iran; Extracted from flickr.com; Photographed by Krzysztof Nitj’ Sefni Fig.53 Hassan Fathy’s Vaulted Structures, New Baris; Extracted from sensesatlas.com; Photographed by Unknown Fig.54 Hassan Fathy’s Domed Housing, New Gourna; Extracted from see.news; Photographed by Ahmed Yasser Fig.55 Section of Shavadan, Iran Fig.56 Section of Wind Tower, Iran; Created by author based on descriptions and diagrams of Shavdans and Wind Towers in Iran Fig.57 Section of Hassan Fathy’s Housing, New Gourna; Created by author based on Steele, An Architecture for People: The Complete Works of Hassan Fathy (1997) Fig.58 Brickmaking in the village of New Gourna; photographed by Aga Khan Trust for Culture. “The Mud Brick Manual,” EastEast. CC BY-NC 4.0 Fig.59 Mud-brick drying field; photographed by Earth Architecture, “Gaza Mud Brick Houses as Inverted Tunnels,” 30 July 2009 Fig.60 Excavated mud-brick cells, Tal Ganoub Qasr al-’Aguz, Bahariya Oasis (4th–7th c. CE); photographed by Egyptian Ministry of Tourism and Antiquities (MoTA), press release, 13 March 2021. © MoTA Fig.61 Material Systems in Bahariya Oasis: Vernacular Mud Brick vs. Imported Concrete. Techniques, Composition, and Properties; Created by author; data compiled from Minke (2006), Houben and Guillaud (1994), Walker et al. (2020), Mehta and Monteiro (2014), Neville (2011), and Hammond and Jones (2011) Fig.62 Vertical functional gradation for earthen masonry; Created by author; adapted from Koizumi, “The Concept of FGM.” Ceramic Transactions: Functionally Gradient Materials (1993) Fig.63 Functionally graded strategies for earthen composites under bending; Created by author; adapted from Koizumi, “The Concept of FGM.” Ceramic Transactions: Functionally Gradient Materials (1993) Fig.64 V-INCA block family generated from a four-unit configurable mould; Created by author; adapted from V-INCA: Designing a Smart Geometric Configuration for Dry-Masonry Wall Fig.65 V-INCA four-unit mould, casting sequence for 1:2 prototypes; photographed by Unknown Fig.66 Earth Material Building Cycle - A closed-loop workflow for cast earthen masonry; Created by author; based on… Fig.67 Comparative taxonomy of earthen and masonry vault–dome typologies; Created by author; based on Vaulting Cultures in the Modern Middle East (2022) Fig.68 Anatomy and nomenclature of centring; Created by author; adapted from Auroville Earth Institute. Building with arches, vaults and domes (1996) Fig.69 Anatomy and nomenclature of a masonry arch; Created by author; adapted from Auroville Earth Institute. Building with arches, vaults and domes (1996) Fig.70 Agriculture Field in the Bahariya Oasis; photographed by Unknown Fig.71 Digital Elevation Model of Bahariya Oasis (Created by author) Fig.72 Aquifer Analysis Through Predictive Modelling (Created by author) Fig.73 Cellular Automata Organisation of a 1200 People Settlement (Created by author) Fig.74 Primary Water Network Generated through Shortest Walk (Created by author) Fig.75 Rhino Ecologic Model of an Agriculture Plot (Created by author) Fig.76 CFD Analysis of Atmospheric Water Harvesting Tower (Created by author) Fig.77 Lady Bug Simulation of Incident Radiation on Housing Plot (Created by author) Fig.78 Karamba 3D Deflection Analysis on Vaulted Roof System (Created by author) Fig.79 Examples of Generated Housing Typologies through Wallacei (Created by author) Fig.80 Vault equilibrium simulation with Kangaroo dynamic relaxation (Created by author) Fig.81 TNA for overall form, horizontal forces, and vertical forces (Created by author) Fig.82 Thermal Testing of Block Samples Using Hot Plate Setup (Created by author) Fig.83 Hydrophobicity Testing Through Water Immersion (Created by author) Conclusion| 217

Fig.84 Compression Testing of Sample Block Under Applied Weights (Created by author) Fig.85 Material Samples (Created by author) Fig.86 Topographical Map of Bahariya Oasis (Created by author) Fig.87 Section of Bahariya Oasis Soil Layers (Created by author) Fig.88 Magnetic Sounding Station Points and Well Locations Within The Oasis (Created by author) Fig.89 3D Visualisation Of The Predictive Model Output For Aquifer Depth And Accessibility (Created by author) Fig.90 Aquifer Accessibility Gradient Map (Created by author) Fig.91 Aquifer Depth Gradient Map (Created by author) Fig.92 Soil Quality Gradient Map (Created by author) Fig.93 Site Integration Gradient Map (Created by author) Fig.94 Weighted Criteria Gradient Map (Created by author) Fig.95 Site Selection Map Showing Optimal Expansion Locations and Selected Settlement Site (Created by author) Fig.96 Grid Division of Proposed Site for Development (Created by author) Fig.97 Grid Division of Selected Site with Optimal Well Locations (Created by author) Fig.98 Allocation of Resources and Consumption to Architectural Units (Created by author) Fig.99 Cellular Automata Horizontal Placement Rules (Created by author) Fig.100 Cellular Automata Vertical Placement Rules (Created by author) Fig.101 Final Aggregation of Units based on Stated Rules (Created by author) Fig.102 Settlement of 1200 People (Created by author) Fig.103 Primary Water Network of Generated Settlement (Created by author) Fig.104 Road Networks (Created by author) Fig.105 Axonometric Diagram of Road Network (Created by author) Fig.106 Water Networks (Created by author) Fig.107 Axonometric Diagram of Water Network (Created by author) Fig.108 Material Samples. Photographed by the author Fig.109 Excavation-led circular resource model. Created by the author Fig.110 On-site fabrication hub for earthen FGM blocks. Created by the author Fig.111 Constituent palette and pairing matrix for graded earthen composites. Photographed by the author Fig.112 Material Research workflow. Created by the author Fig.113 Initial screening of stabiliser systems. Photographed by the author Fig.114 Volumetric shrinkage after 7-day air curing of Exp. 4 Casein. Photographed by the author Fig.115 Volumetric shrinkage (%) at day 7 for six formulations; Created by the author, based on physical tests Fig.116 Hand-press compression set-up. Photographed by the author Fig.117 Compressive strength (kN/cm²) for the initial six formulations; Created by the author, based on physical tests Fig.118 Parametric mix matrix for graded earthen composites. Photographed by the author Fig.119 Experimental setup and qualitative visualisation. Photographed by the author Fig.120 Thermal conductivity of candidate stabilised-earth mixes; Created by the author, based on physical tests Fig.121 Water submersion test. Photographed by the author Fig.122 Water absorption (%) after 1-h and 24-h submersion; Created by the author, based on physical tests Fig.123 Hand-press compression set-up. Photographed by the author Fig.124 Compressive strength (kN/cm²) for all the formulations; Created by the author, based on physical tests Fig.125 Proposed Material. Created by the author, based on physical experiments Fig.126 Material Samples. Photographed by the author Fig.127 Thrust Network Analysis of a funicular vault under variant boundary conditions. Created by the author using RhinoVault Fig.128 Discrete arch under self-weight: thrust-line verification. Created by the author using Kangaroo Fig.129 Interlocking voussoir family and tiling logic for a graded vault. Created by the author Fig.130 Dry interlocking earthen units: joint engagement test. Photographed by the author Fig.131 Stacked graded-earthen prototype: dry interlocking under vertical load. Photographed by the author Fig.132 Reusable modular formwork set for casting graded interlocking blocks. Created by the author Fig.133 Erection sequence for a dry-joint vault. Created by the author Fig.134 Quarter-scale vault prototype on centring. Photographed by the author Fig.135 Inclined roof-panel prototype with dry interlocking units. Photographed by the author Fig.136 Housing Unit Visualisation (Created by author) Fig.137 Solar Updraft Tower Mechanism Diagram (Created by author) Fig.138 Condensation and Water Yield Calculations (Created by author) Fig.139 Proposed Atmospheric Harvesting System (Created by author) Fig.140 Initial Tower Morphology Development (Created by author)

218 |Conclusion

Fig.141 CFD Preliminary Test (Created by author) Fig.142 Parameters and Objectives Set for Optimization of Proposed Tower (Created by author) Fig.143 A Selection of Pareto Front Phenotypes (Created by author) Fig.144 Selected Morphology (Created by author) Fig.145 CFD Analysis on Selected Morphology (Created by author) Fig.146 Schematic Plan and Section for Condensation Farm (Created by author) Fig.147 Programmatic Distribution for Cistern Units (Created by author) Fig.148 Varying Design Parameters for Cistern Units (Created by author) Fig.149 Objectives Set for the Optimisation of Cistern Units (Created by author) Fig.150 A Selection of Pareto Front Phenotypes for Cistern Units (Created by author) Fig.151 Selected Morphology for Cistern Units (Created by author) Fig.152 Evaluation of Vault Deflection Using Karamba 3D (Created by author) Fig.153 CFD Simulation longitudinal section run on North-West winds (Created by author) Fig.154 CFD Simulation cross section run on North-West winds (Created by author) Fig.155 Construction Sequence for Vaults (Created by author) Fig.156 Parameters for Mashrabiyas Variation (Created by author) Fig.157 Selected Facade (Created by author) Fig.158 Optimisation Parameters for Mashrabiyas (Created by author) Fig.159 Floor Plans for Optimised Cistern Morphology (Created by author) Fig.160 3D Section Through Selected Cistern Morphology (Created by author) Fig.161 Programmatic Distribution and Massing Logic for Housing Clusters (Created by author) Fig.162 Design Parameters for Housing Clusters (Created by author) Fig.163 Optimisation Objectives for Housing Clusters (Created by author) Fig.164 Pareto Front Solutions for Housing Clusters (Created by author) Fig.165 Selected Phenotype for Housing Cluster (Created by author) Fig.166 Incident Radiation Simulation (Created by author) Fig.167 Direct Sun Hours Simulation (Created by author) Fig.168 CFD Simulation Plan View (Created by author) Fig.169 Floor Plans for Optimised Housing Cluster Typology (Created by author) Fig.170 3D Section of Selected Housing Cluster (Created by author) Fig.171 Section of Urban Plan Illustrating Water Filtration Network (Created by author) Fig.172 3D Section of Proposed Agriculture Plot (Created by author) Fig.173 Design Parameters for Agriculture Units (Created by author) Fig.174 Optimisation Objectives for Agriculture Units (Created by author) Fig.175 Section for Agriculture Plot (Created by author) Fig.176 Catalogue of Units (Created by author) Fig.177 Clash Instances (Created by author) Fig.178 Clash Catalogue (Created by author)

Conclusion| 219

220 |Conclusion

Conclusion| 221

222 |Conclusion