Joe - Tech

Embodied Learning: Working through a low carbon library

Using digital and physical testing to iterate a low carbon library, and embedding this learning in the fabric of the building.

Technical Tutor: Unit 3 Tutors: Technical Coordinators:

Jason Coe

Ifigenia Liangi, Dan Wilkinson

Luke Olsen, David Storring

Joe Johnson ENVS 3046

Tuesday 24th April 2018

Word Count: 5,478

Abstract

This project proposes a library and workspace for the University of the Arts, Catania.

This technical explores how the specific conditions required by rare books, standard library books and humans can be achieved in an environmentally friendly manner, connecting the spaces wherever possible with green areas of the site. This aim is made challenging by the specific conditions each element requires, and the intense heat and diurnal temperature swings of the Catanian climate.

The project as a whole aims to make the stages of development, changes in thinking, and previous iterations of the scheme present in the final building, so that users of the library and workspace might understand how the building itself has been worked through.

The technical aims and investigations drive the iteration of the building, and the different stages of development are then woven together into spaces that acknowledge and celebrate the seams between ideas.

Both elements of the project are examined in this report.

Overview

The key technical investigation explores how clay soil construction can achieve many of the environmental requirements for the building, making use of its phase change properties.

An architectural toolkit and language is developed, using clay soil, hessian and timber, which combine to create spatially engaging, environmentally friendly, temperature and humidity controlled environments, built using extremely low cost, sustainable materials.

The document also explores the ventilation strategy the building can utilise to achieve and blend between the very specific micro-climates it requires. Analysis of the site (with a particular focus on the wind) led to the development of a thermal labyrinth technique, partially integrated within the inhabitable spaces of the building, which is tested and developed through fluid dynamics, both simulated and real.

The building is largely sunken underground to stabilise the temperatures, and this document also explores how this can be achieved structurally with timber, and how lighting can be tailored for this environment.

Detail 1 - modular timber desks

Detail 2 - compositional history door handles, cast in leather

Fragment 1 - Sunken clay shell reading room

Fragment 2 Clay shell solar shading

Fragment 3 - wind labyrinth inlet and sheltered study space

Programme requirements: A library and workroom for Catania University of the Arts

The building will house the art universities main library and rare book collections, and provide a wide range of inside and outside workspaces for students.

Each element has very specific environmental requirements which the architecture must adress:

Strategy for the project:

1. Use environmental research and testing methods to iterate the spaces and strategies used in the building, and develop a low carbon material language.

2. Reflect these changes in the final proposal, so visitors can understand how the building itself was worked through, and glimpse what might have been.

room 40 - 55 Below 25 - any temperature is suitable, as long as it is kept very stable

Sources: https://www.bl.uk/aboutus/stratpolprog/collectioncare/ publications/booklets/managing_library_archive_environment.pdf

http://glamox.com/uk/solutions/library

https://www.kellwood.co.uk/lighting/technical/cibse-recommended-lux-levels

avoiding direct sunlight wherever possible

Key technical aims:

1. Use materials with the lowest possible embodied energy.

avoiding direct sunlight wherever possible

2. Develop a passive, carbon zero ventilation strategy to achieve the programmatic requirements.

3. Develop low cost, on site construction techniques which can be expanded on into the neighbouring sites by self builders.

/ 2000 depending on the task

The research reveals that the workspaces can be dynamic in temperature, humidity and lighting.

However, the rare books room and main library must maintain a constant level for all their environmental factors, as temperature and humidity swings are most damaging to books.

4. Wherever possible embody the previous iterations and development of the building in the final spaces.

A case for self building: Construction industry corruption leaves three sites empty

The sites along Via Archimede were cleared in 1958, following the outlawing of brothels, and have remained empty since.

Mario Cucinella’s masterplan (fig. 2 -4) sought to develop the sites, however since its proposal in 2012 no work has begun.

Avoiding the industry through self-building:

Small, independent galleries and bars have begun to occupy derelict spaces elsewhere in the red light district.

The proposed library hopes to inspire the same on the other two, larger sites, by using an affordable, environmently friendly material language that can be learned, improved and expanded on.

Fig. 5-9: shots of the three empty sites.

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Central cleared site, Via Archimede Catania, Sicily

Fig. 10

Fig.

Contextual site plan

Site for self builders to expand into

Topographical site plan and site section

Site for self builders to expand into

Site

Fig.

Fig. 13

Fig.

Key environmental factors:

Extreme heat and minimal shading

Combatting the heat:

Harnessing the prevailing winds

16: Catania’s climate is very hot, consistently reaching the high 30’s in summer.

Fig. 15, 17: The site itself receives almost no shade from the surrounding buildings, making solar gain an important factor to control.

18-19: Excluding the North West, Catania receives wind from a range of directions. However, as the site is surrounded by buildings except from the East, this is the most consistent wind direction.

Fig.

Fig.

Fig. 15

Fig. 16

Fig.

Fig. 18

Fig.

Rationale for testing: Physical and virtual Combining the strengths of both for a better understanding

In my interview with fluid dynamics expert Owen Connick, he discussed how Breathing Building’s utilise Computational Fluid Dynamics and Dynamic Thermal Simulation, alongside physical water bath modelling and wind tunnel testing.

This is because no single test can answer all their questions, and each test has weaknesses - It is through conducting and comparing a range of tests that they gain an understanding.

I have applied this approach to all of my own testing.

Fig. 20 - digital and physical wind tunnel testing, to optimise the wind labyrinth intake.

Fig. 21 - CFD and physical waterbath testing to simulate airflow in the inhabitable thermal labyrinth.

Fig. 22 - 1:20 and 1:1 tensioned fabric shell testing

Fig. 23 -1:20 tension shell model, loaded fabric simulation, loaded fabric ceiling proposal

Fig. 24 - Grasshopper sunlight raytracing simulation, 1:10 physical model lux level testing

Fig. 20

Fig.

Fig. 22

Fig. 23

Fig.

1 Minimising embodied energy, developing a zero carbon strategy:

Initial timber, hessian and clay material investigations

Developing a material palette that naturally cools the internal environment.

Clay precedents: Vernacular insulation and surfacing technqiues

Clay and straw are used in the vernacular architecture of many countries, and as an environmentally friendly material in eco homes.

Fig. 32: A thatched straw roof, providing effective insulation and featuring eaves which keep rain off the walls.

Clay shell precedents: Clay / fabric hybrid shell structures, by the AA visiting school, Lyon

Fig. 36: Formfinding arch based geometries which work almost entirely in compression, minimising the forces exerted on the clay.

Fig. 32 and 33: Cobb walls, which can be formed either via ramming earth into wooden formwork, creating loadbearing insulated walls, or as surface finish.

the team created a CNC cut timber scaffold, between whcih was stretched a taught fabric. This was then covered in six thin layers of earth.

Fig. 34 and 35: The traditional construction techniques of the Musgum and Maasai people also offer incredible precedents for the combination of timber and earthen materials.

Both use earth as a binding and sealing agent around bent timber frames, providing effective shade and insulation for the interior spaces.

Fig. 38: Once dry the formwork was taken out and the clay / fabric could take its own weight (although it eventually collapsed when moved, due to the lateral loads applied).

Conclusion:

This research demonstrates the possibilities of fabric / earth hybrid structures, however, I find this method very restrictive given its reliance on compression-only geometry.

Fig. 37:

Fig. 31

Fig. 32

Fig. 33

Fig. 34

Fig. 35

Fig. 36

Fig. 37

Fig. 38

Fig. 39

Fabric shell precedents: James Waller’s concrete / hessian ‘nofrango’ shells

Substituting clay and earth for concrete

1:1 Tensioned hessian test

Aims:

Observe the qualities of the material at 1:1: what geometries can be achieved, what textures are created, how does it interact with light.

Explore how easily fabric / clay shells can be fabricated.

James Waller’s cheap and effective construction method uses weighted fabric to create beautiful, structurally optimised ribbed catenary vaults.

Gauge the strength of the fabric and clay as a non structural building material - Can it support its own weight?

Method:

Fig. 41: Hessian fabric is stretched over a simple catenary steel frame, and then covered (by hand) with concrete, the weight of which pushes and stretches the hessian into a double curvature, improving its strength.

Fig. 42: The exterior of the structures are smooth, whereas the interior surface reveals the hessian, creating a beautiful, highly textural finish. I particularly like the details around the windows, where the fraying of the hesssian can be seen.

1. Hessian is stretched diagonally across a simple plywood corner, and then pulled back at various points in the middle to create surface undulations and double curvatures, increasing the strength of the structure. (Fig. 43).

2. A 50 / 50 blend of soil and clay was applied by hand in three coats, allowing the clay to dry inbetween. Given the hand application, the clay /earth thickness varies between 5mm and 10mm. However, as the material is not structural, the irregular thickness is less critical. (Fig. 45).

Fig. 44: The clay seeps through the hessian on the inside, creating a beautifully textured surface reminscent of the Waller method, clearly acknowledging where the structure used to be pinned.

Conclusions:

The shell supports its own weight, making it appropriate for non structural use (fig. 47).

Beautiful undulations can be achieved, with a pleasingly rough surface created by hand application.

The geometries that could be explored were very limited by the timber frame, highlighting the need to explore the frame more carefully.

The piece does not address the detailing of the fabric.

Fig. 40

Fig. 41

Fig.

Fig. 43

Fig. 44

Fig. 45

Fig. 46

Fig. 47

Material test 2: A first attempt at improving the frame 1:20 plaster soaked tuille over a frame

Aims:

Determine how well the loading of fabric can be tested at scale.

Determine suitable modelling materials.

Get a feel for the spatial qualities of fabric enforced shell structures.

Model material justification: Plaster in place of clay, and tuille for hessian:

At scale, clay would be far too weak, making it impractical as a testing material. The 1:1 test provides proof that the clay and earth work to rigidify the fabric, and plaster offers a similar strength and finish at scale.

The thickness of tuille is similar to that of hessian at scale, and its gridding allows it to stretch and double curve in similar ways.

Method:

1. Tuille fabric is tied to the frame using string, and then painted with polymer plaster.

2. The weight of the plaster pulls the fabric into curved geometries, as in the Waller method.

Results:

The spatial qualities seem promising.

The test is restricted by the weak, simplistic frame geometry.

The tuille fabric and plaster work well to model the larger scale clay shell.

Low carbon material justification: Clarifying why timber, clay soil and hessian have been chosen

Pine and plywood:

Avaliable cheaply and off the shelf - Accessible to future self builders.

Sustainable - Grows very quickly, using land efficiently.

Easy to work with and manufacture on site.

Hessian: Low cost.

Environmentally friendly: Naturally grown, and ‘uses much less water, requires no chemical pesticides and fertilisers to grow, and replenishes much faster than cotton’. (https://theswatchbook. offsetwarehouse.com/2015/04/14/why-is-jute-sustainable-ecofriendly/)

Gridded surface enables stretching and double curvature. Clay soil:

Extremely low cost.

Readily available on site and locally.

Minimal extraction and transportation impact on the enviornment.

Phase change, cooling properties.

Key additional materials:

Steel and concrete: used as sparingly as possible. Required for bolted fixtures between pine members, pad foundations and connections between wood and concrete.

100% Recycled plastic: used for windows and superheated solar lanterns.

Fig. 48

Fig. 49

Fig. 50

Fig. 51

Fig. 52

Fig. 53

Fig. 54

Off the shelf timber precedents: The Segal Method

The scheme uses off-the-shelf softwood with shallow pad foundations, creating spacious two storey homes built by a single family without a crane or heavy machinery.

Nodal timber precedents: Matt + Fiona modular system

Conclusion:

The method demonstrates how much can be achieved without heavy machinery and only off the

timber.

The method is restricted spatially by conforming to the grid - what happens when this is broken?

Conclusion:

the project reveals a useful tactic to achieve highly specific but inexpensive structures, by

Fig. 58: The system uses standard pine lengths, in combination with plywood joints, allowing attachment in eight directions.

Fig. 59:The structures demonstrate how much can be achieved with a simple kit of parts, assembled by children under adult supervision.

Fig. 60: A graphic demonstrating how efficiently the sheet material was used, and the speed of production.

combining off the shelf timber with CNC cut modules.

Figure 55: From a visit to one of the structures in Lewisham, demonstrating how beautifully the timber has been joined, and how well it has aged.

Figure 57: Two studies I did to understand how the ceiling joints are constructed, and how the non load-bearing wall panels fit between the floor and ceiling joists.

shelf

Fig. 55

Fig. 56

Fig.

Fig. 58

Fig. 59

Fig.

Combining timber and clay through the desk

Fabricating a workdesk to develop a timber language

The tests explore how parts of the work desks might be made from unfired clay tiles, so that they could be written on, and how the armrests could contain sealed clay, taking the shape of the users.

63: The first tests revealed how carefully the unfired clay must be held to avoid cracking - The brass pins require padding to spread the load.

The tile began the same size as the timber board, revealing the shrinkage.

Fig. 61: Exploring the material language of the building through furniture, to see how the details might inspire the main structure.

Fig.

Fig. 62:

Observing clay shrinkage

Cracking resulting from too much pressure from the fixtures

Fig. 61

Fig. 62

Fig. 63

Fig.

1:1 modular timber test: Exploring a modular timber node system through the desk, to inform the structural language

This 1:1 development of the table applies what was learnt in from the Matt+Fiona project.

Key design elements:

A lazer cut module negotiates the joint between four standard lengths of pine, and further modules support the arm rest and inscribing tablet above.

Two laminated plywood frames are bolted over a block of clay and waxed canvas cover, creating a sealed, soft clay armrest.

Conclusion:

The structure is cheap, quick to assemble, environmentally friendly, lightweight and elegant.

The following investigations explore how this method could be applied to the main structure.

Laminated plywood frame

Waxed canvas

Unfired clay writing tile

Clay block

2 x 9mm lazer cut modules

Off-the-shelf pine

laminated Lazercut joint

Fig. 65

Fig. 66

Fig. 67

Fig. 68

Off the shelf timber structure: Initial iterations applying Segal’s method

Structural timber development: Blending standard lengths and smaller custom pieces

It makes use of pairs of

posts, learned from

allowing connections to slot between them. This provides many attachment points without a bracket, and provides a lot of tolerance as the joint itself is hidden.

Fig. 71: This design iteration also makes use of the Segal method, but includes a larger proportion of custom, CNC cut pieces.

Fig. 69: This first attempt at part of the library interior combines standard softwood lengths with custom, CNC cut wooden components.

Fig. 69

Fig.

Fig. 71

Fig. 70:

timber

Segal,

Achieving programmatic and environmental objectives through clay shell construction

Applying the clay shell language to address the environmental and spatial requirements of the building through speculative drawing, scale models and physical and digital testing.

Integrating the clay, hessian and timber:

Speculative drawings exploring how the clay and timber could combine into an architectural language

Integrating the clay, hessian and timber: Designing the shelves

Fig. 74: Exploring how clay shells might weave around the book shelves and books, surrounding them with a phase change material which helps to buffer fluctuations in relative humidity and temperature.

Fig. 73: These drawings speculate on how the clay might work around and within the frame, surrounding the books to regulate the environment, and opening up many sculptural opportunties for the spaces.

Fig. 74

Fragment 1: Creating a thermally insulated underground workspace

Developing

technique 1 through a 1:20 model: Using an articulated timber frame to control the fabric

This model sough to construct part of the aisle and reading room explored in the previous drawings.

The model is located underground, with the apertures at the top sitting at ground level. The model explores three work desks, sat inside pods formed beneath the roof lights, and how this area relates to the aisle of the library.

Tensioned fabric within the frame: Glimpsing the possibilites of the spaces

Aims:

- Test how geometry designed in Rhino can be achieved using the clay shell method, and compare their differences.

- Test how much control over the enclosures can be achieved using a more developed frame.

- Explore the qualities of the spaces formed.

- Provide a physical model for further lighting tests.

Results:

- Similar volumes can be achieved, but with a puckered character not present in the Rhino model. As precise geometry is not required in this space, this is no issue.

- Exiting spatial qualities are created, through a much more developed timber frame.

- Several details work only for model scale, and must be examined at 1:1.

be seen in the model - at 1:1 these would be less apparent, as they would be far smaller, and hessian is more opaque than tuille.

Fig. 78: Suspended tuille inside the frame, forming the volumes of space.

Fig. 79: The machine stitched seams can

Fig. 80: demonstrates how the timber frame triangulates the forces to create a very stable structure.

Fig. 75

Fig. 76

Fig. 77

Fig. 78

Fig. 79

Fig. 80

Learning from the model: Unexepected sagging of the fabric

Issue:

Fig. 83: Despite appearing taught, when loaded the fabric began to sag heavily in places, restricting the volumes. This was rectified with additonal tension strings.

Solution:

Keeping the strings adjustable would allow for the structure to be retightened and controlled.

Opportunity:

Some spaces could take advantage of this natural sagging, allowing the fabric to form itself and massively reducing construction time, as far less stringing is required.

Loaded with plaster

Un-loaded fabric

Fig. 81

Fig. 82

Fig. 83

Fig. 84

Locating the fragment: Technical section exploring the construction

2 part wooden window and glass, bolted together

Grass / soil

The wooden frame braces against the earth, using a treated plywood board sheath and waterproof membrane to protect the interior environment.

The space between the wooden frame and the fabric / clay shell can be filled with clay / straw insulation, facilitating a very controlled internal environment.

The only concrete required is in the pad foundations, which (as in the Segal method) sit 1 metre into the ground.

Treated plywood sheet

Softwood structural member

Waterproof membrane

Clay / insulation

Tension fabric clay shell

Gravel

Fabric cast concrete pad foundation

Soil

Waterproof membrane

Compacted, polished earth floor

Timber structure supporting bookshelf

Contextualising the model

Clarifying the relationship to the ground

Environmental benefits of sunken construction

Sinking large areas of the library will do help to naturally stabilise temperatures and cool the building.

Fig.

Although the data is from Virginia, it demonstrates how underground temperatures are cooler and far more stable, becoming increasingly stable the further down they are measured.

87:

Fig. 85: Fragment 1 is one of several underground workspaces located around the site, as demonstrated in the massing model.

Fig. 86: A cast piece relating the model to the ground.

Fig. 85

Fig. 86

Fig. 87

Fig. 88

Foundations for timber frame: Fabric formwork development through door handles

Detail 1: Casting previous iterations into the building

Developing a fabric formwork technique through 1:1 door handles

Environmental issue:

Some concrete is required for the foundations.

Aims:

Minimise the environmental impact of the concrete through fabric formwork.

Take advantage of the sculptural opportunity to create cast elements in the buiulding, and the detail that can be captured.

Precedent: Mark West

Mark West and C.A.S.T. create beautiful fabric cast pieces, which minimise the formwork required, and reduce costs, using only basic timber elements and a cheap geotextile fabric.

Fig. 94: I am particularly interested in the incredible detail captured in their work, as the porous textile allows air to escape from the concrete, and creates a perect surface finish.

Material: Reclaimed Leather

Reclaimed , recycled leather comes at no additional cost to the environment or the animals.

Fig. 96 and 97: The leather can be embossed and etched, allowing for information to be written in to its surface, and very accurately transferred to whatever is being cast.

Fig. 92

Fig. 93

Fig.

Fig. 95

Fig.

Fig.

Digital palimpsests: Projecting previous drawings onto the formwork

Method:

1. Each handle is designed by drawing onto the leather, embossing the formwork with the development drawings.

2. The drawings from previous handles are scanned, vectorized, and projected in Rhino onto the unrolled formwork. Sketches that land on the formwork are lazer etched into the leather.

Surface logic:

The older the drawing, the further back it falls into the cast. This is created by extruding some elements of the drawings and layering the formwork (fig. 100), thus creating channels cast into the handles.

1:1 formwork development

Lazer etching and cutting leather patterns

Simple, bolted connections between the sheets, allowing the casts to be reused.

Result:

The handles have the developmental drawings of themselves and their predecessors etched into their surface.

Thus, they reveal their compositional history through their texture, allowing visitors an insight into how the pieces were worked through.

Fig 101: Etched leather patterns

Fig.102:

Fig. 103: Removing the cast from the formwork

Fig. 98

Fig. 99

Fig.

Fig. 101

Fig.

Fig.

Second Handle:

Embedded with drawings from handle 1

Fig. 104

Fig. 105

Fig. 106

Third Handle:

Embedded with drawings from handles one and two

Fig. 107

Fig.

Fig.

The armatures that hold dinosaur bones grip them to avoid damage, and make explicit where the artefact ends and the support begins.

1. Sections are cut through the object at regular intervals.

2. Single curved surfaces are lofted between the curves.

3. These panels are unrolled into the pattern pieces.

4. These pieces are given a margin and attachment holes, to allow them to connect together.

Fig. 111: The handles are held in a similar waythrough photogrammetry accurate armatures can be constructed to grip the pieces with no permanent fixtures, and they use a language of 3D print and brass to differentiate them from the handles.

113: The weight of the concrete would bulge the formwork, resulting in a smooth, curved geometry akin to the original digital version.

Holding the embodied learning artefacts: A material language that respects where learning occured

Applying the technique to the foundations: Developable srf study

1.

2.

3.

4.

Fig.

Fig. 110

Fig.

Fig. 112

Fig. 113

Fig. 114

Fig.

Learning from the model: Responding to deflection and lighting issues

The model revealed a need for widened apertures, additional lighting, and cross bracing for the structure.

These additions must be handled to clearly show the seams between the new and old design iterations.

Additional rooflights:

Acknowledging the model and boosting lux levels

Fig. 126: Incorporating additional rooflights into the space, following the geometry created through the deflection of the model. Thus, additional light is brought into the spaces, in a manner which refers to the model and the learning that occurred through it.

Fig. 125: Enlarging some rooflights to improve lux level, but acknowledging the previous iteration through the ventilation openings.

Fig. 124: Analysing the amount of deflection, and highlighting the new areas opened up in the structure.

Fig. 124

Fig.

Fig. 126

Fig. 127

Combining iterations: Blending between ideas through architectural insertions

1. An earth shell entrance from an ovular aisle configuration, designed to blinker views of the space until visitors passed through the threshold - concealing the secondary circulation routes.

First drawn: 12th of February

1. An earth shell entrance from an ovular aisle configuration, designed to blinker views of the space until visitors passed through the threshold - concealing the secondary circulation routes.

February 22nd

February 22nd

First drawn: 22nd of February

26th of February

2. A sunken staircase and concrete pile, modelled soon after the books were relocated underground, to help stabilise the environmental conditions.

The piles were intended to be cast in the same manner as the door handles, and potrude above the ground to receive the timber frame, revealing their patternation to visitors as they navigate the books.

2. A sunken staircase and concrete pile, modelled soon after the books were relocated underground, to help stabilise the environmental conditions.

The piles were intended to be cast in the same manner as the door handles, and potrude above the ground to receive the timber frame, revealing their patternation to visitors as they navigate the books.

First drawn: 26th of February

1.

2.

2.

1.

2. 3

3. Development of the timber frame and pile system, to facilitate the tensioned fabric shell structure (developed initially through the threshold in fig. 1). This pile would also be cast in fabric formwork, similarly to the door handles, and accomodates two angles of the frame with a seat inbetween.

Simulation 1: Formfinding with weighted fabric

Simulating the loaded fabric to glimpse its potential

Test:

Use Kangaroo, a physics engine for grasshopper, to apply gravitational loading to a surface, simulating the bowing experienced unintentionally in the fragment 1 model.

Aims:

Quickly explore a range of formal possibilities, with more accuracy than a physical test.

Understand how to consider forces in the 3D model, to bridge the gap between the digital and physical elements of the project.

Successes:

Techniques were discovered and fed into the 3D modelling development of the rare books room, giving the digital proposal far more validity.

Weaknesses:

I did not incorporate specific material qualities or forces into the script, so the simulations are not tied to the properties of hessian and clayit can be made more or less flexible, and thus is not always accurate to real life.

However, it is very useful for modelling the difference created by adjusting the tension of the fabric.

1. Fabric without loading

2. Loaded fabric

Fig. 128

Fig. 129

Fig. 130

Fig.

Developing a weighted fabric ceiling: A hanging, insulated ceiling for the rare books room

Digital form finding experiments, exploring how a weighted hessian ceiling might behave. A variety of different tensions were explored.

Ideas sparked from the simulation:

Pinching the fabric to pucker the ceiling.

- Letting the bulging around the rooflights focus light downwards. (fig. 136).

Letting the roof fall and became partitions for the floor below. (fig. 137).

Fig. 132: Locating the apertures in plan.

Fig. 133 and 134: Using the boundary of the room, developed through plans a quad based mesh was developed, fixed around the edges and window openings.

Fig. 135 to 137:

Fig. 132

Fig. 133

Fig. 134

Fig. 135

Fig. 136

Fig. 137

Clay shell Solar Strategy: Restricting solar gain by adapting the technique for the outdoors

Fragment 2: Applying the clay / hessian method: Diffusing light for the rare books room

Rare books room lighting requirements:

The building needs to let in light from the roof, however direct sunlight would damage the books.

Thus, solar shading on the rooflights is required to reflect indirect lighting into the space, and prevent the windows from superheating in the sun.

Fig. 140: A speculative drawing based off the simulated formfinding, exploring the atmosphere the diffused rooflighting could create.

Fig. 138 Fig.

Fig. 140

Rainwater protection: Test confirming the shell quickly degrades in rain

Fig. 141: Although for much of the year Catania is very dry, it does occasionally receive heavy rainfall.

Additive design:

Building up from past iterations, and differentiating with colour

Revealing the window’s progression through colour

Fig. 142: A five stage test of the 1:1 wall fragment, in which water was gradually misted over its surface, revealed the structure quickly lost its integrity and crumbled - there is a clear requirement to shield them from the rain.

Fig. 143: To allow the rooflight to discuss how the idea progressed in its final appearance, it was designed additively, with nothing of the original removed.

Each layer of design is colour coded, celebrating the different stages and blending of ideas.

The design includes a rain cover, and fig 144 discusses the aim to reflect diffused light into the rare books room via the rooflight and ceiling.

1.

2.

3.

4.

Fig. 141

Fig.

Fig. 143

Fig.

Testing the additive design: Using grasshopper to confirm the correct reflection of light

Test:

Using raytracing to bounce light into the diffuser, and observe whether it is reflected into the space.

Physical model: Constructing the window at 1:10

Result 1

The diffuser deflects light from the sun early in the morning or late in the evening straight into the space, as the heat is less intense.

Result 2

The light is reflected twice before entering the space, reducing its intensity and the heat brought in.

Figures 147 to 148: Forming the diffuser, using the plywood strip as a guide.

Figure 145: Sun low in the sky

Figure 146: Midday sun

Fig. 145

Fig.

Fig. 147

Fig.

Clay / fabric formwork: Developing technique 2 through the model

Limitations of technique 2:

Although the plywood smoothes the geometry of the tensioned fabric, it is not accurately tied to the 3D model.

Therefore, there is no guarantee it will reflect light in the same way.

Technique 3, developed in the next model, would have provided more accuracy.

Flexible plywod guides to smooth the geometries

Fig. 149

Fig.

Lux

level

test 2:

Testing the

performance

of the rooflight diffusers

Aim:

Determine how effectively light is reflected.

Method:

- Placing the model in a black out box, and measuring how much daylight light is diffused into the space.

- One side of the model was kept unpainted and installed with a dark, absorbant shade, to measure against the plaster painted shell.

Result:

The shade was very effective at diffusing light, suggesting a number of rooflights would bring sufficient indirect light into the rare books room.

Improving light reflection: Coating the solar shades in plaster

Fig.

Fig.

Test 1: direct sunlight onto the diffuser

Test 2: Overcast sky

Results:

The diffuser was effective at bringing light into the space, vastly out performing the absorbant test.

Issue:

There is no control over the light level, which could result in overheating on certain days.

Concluson:

These devices would be very useful in the rare books room and main library, bringing in ample light without the risk of direct sunlight on the books, with an added layer of control.

Fig. 153

Fig. 154

Fig. 156

Fig. 157

Fig. 158

Fig. 159

Camera settings are kept consistent for the three DSLR photos, to give an experential sense of the difference combined with the lux reading.

Could be resolved via a heat proof film on the windows, or articulation of the opening from within.

Developing a passive ventilation strategy through inhabitable thermal labyrinths

Exploring the massing of the building to achieve a naturally ventilated library environment, applying the material language already developed.

Precedent 1: Wind labyrinth passive cooling Introduction and precedents

Precedent 2: Stack ventilation in Libraries

St. John’s College library, Hereford Archive

St John’s College Library

Cullinan Studio

Strategy:

- Natural Stack ventilation

- High thermal mass from masonry construction.

- Super heated lantern and fan, generating draw through the library.

Hereford Archive

Architype

Strategy:

- High thermal mass for temperature stability.

Premise:

Subturranean air intakes use lower ground temperatures to chill the air over a long distance, bringing cool air into the building withput the need for air conditioning.

On still days, or if the labyrinth is particularly long, fans are required to draw the air through.

- Thermal isolation from the office space.

- Low ACH (as the repository is not often occupied) protecting the space from external temperatures.

- Buffering: Complete seperation from the office space

Fig. 160

Fig. 161

Fig. 162

Fig. 163

Fig. 164

Locating the Labyrinth inlets: Simulated wind analysis of the site

Fig 165: Drop in wind speed from the South, caused by surrounding buildings.

Fig 166: Uninhibited wind approaching from the East.

Fig. 167: Wind tunnel software used for testing

Fig 168: Diagram showing the primary intake direction, and location of the labyrinth

Fig. 165

Fig. 166

Fig. 167

Fig. 168

Precedent 3: Earthship construction Using the ground to cool and regulate the building

Similarly to the wind labyrinth, air is taken in through the ground to lower its temperature.

Glazed facades are superheated, drawing this cold air through the building.

Massing iteration 1: Applying the precedents

Initial massings sought to isolate the rare books room in the centre, with an occupiable thermal labyrinth cooling the main library and workspaces.

All the spaces are sunken underground, for temperature stability and cooler air, aside from the lobby, which is a tented space with superheated rooflights to draw the air through.

Precedent summary:

- Thermal isolation is best for rare books, to ensure no temperature swings.

- Stack ventilation with superheated elements is best suited to humans, as temperature swings are not damaging, and the spaces receive cool air with high ACH rates.

Fig. 169

Fig. 170

Fig.

Fig.

Inlet 1: Creating shelter behind the inlet

The inlet is effective at gathering the wind into the labyrinth, and creates much lower air velocities behind it.

Low pressure area created behind the labyrinth.

through the labyrinth.

Fig. 177: velocity simulation

Fig. 178: Pressure simulation

Fig. 179: Simulated wind flowing

Long Section

Front Elevation

Fig. 175

Fig. 176

Fig. 177

Fig. 178

Fig. 179

Inlet 2: Adding a fin to increase the impact

Fig. 182 and 183: Velocity simulation The added fin is effective increating more low wind speed area behind the inlet, and forcing fast moving air over the top.

Long Section Front Elevation

Fig. 180

Fig. 181

Fig. 182

Fig. 183

Inlet 3: Adding seated space behind the inlet

Modifications:

The inlet is raised off the ground, to clarify that it is not a passageway.

However, this negatively impacts the performance, and could be done with a grill instead.

Limitations:

The Seated area is much larger than the inlet itself - the design could be reconfigured with a larger inlet.

Only stepped seating is provided, with no desks, cooling or rain protection elements.

Fig. 186 and 187: pressure and velocity simulations. The seated area behind the inlet is sheltered effectively.

Front Elevation

Fig. 184

Fig. 185

Fig. 186

Fig. 187

Inlet 4: Widening the opening to increase intake

Taking advantage of different wind directions: Creating two further openings, running either side of the shelter

Opportunity:

Fig. 191: Adding the seating to the inlet requires the protected area to be longer. Rather than add sides, two additional inlets can be added.

This allows the inlet to pick up more irregular winds from the North and South, with the seating area nested between them.

Improvements:

The inlet has been widened to increase intake.

Limitations:

A tongue was added to try and improve the wind intake, however fig. 190 reveals the high pressure area beneath it - wind is still running under the inlet and not into it.

Front Elevation

Long Section

Fig. 188

Fig. 189

Fig. 190

Fig. 191

Inlet 5:

Noticing issues with the added inlets through wind simulation

Issue observed:

Figures 194 and 195 demonstrate how wind enters through the Eastern inlet, but can immediately exit from the North and South.

This renders this iteration useless, and is subsequently addressed.

Front Elevation Long Section

Fig. 192

Fig. 193

Fig. 194

Fig. 195

inlet 6: Refining the triple inlet geometry

Improvements:

Both iterations move the point at which the inlets join underground, and direct them so they join the wind direction from the eastNo wind is lost by flowing back out.

Front Elevation

Long Section

Fig. 196

Fig. 197

Fig. 198

Fig. 199

Confirming suitability of inlet 6: A final round of simulated wind testing

Fragment 3: 1:20 physical model: Developing technique 3 to accurately achieve the inlet geometry

Aims:

- Develop the formwork to accurately construct the digitally simulated inlet.

- Provide a sheltered workspace.

Fig. 202

Fig. 200: Velocity test highlighting the low speed / pressure area created behind, and the high pressure on the shell in

Fig. 201: High pressure on the shell demonstrates how the air is forced down into the labyrinth.

Fig. 200

Fig. 201

Housing the wind inlet: Developing armatures to achieve the geometry

Developing the timber frame: Triangulating the joints and relying less on ‘model solutions’

Fig. 205: Further development on the Segal joinery method.

Fig.203 and 204 - Curved plywood elements control the fabric on model scale.

fig. 205 - 1:1 detail of the formwork.

Fig. 203

Fig. 204

Fig. 206

Fig. 205

Model assembly: Rehearsing the construction

Stringing up the fabric formwork and creating the volumes

Figures 208, 209 and 210: The three inlet volumes and the merging point is clearest at this stage.

Fig. 208

Fig. 209

Fig. 210

Technique 3: Achieving the simulated geometry Solidfying the shells

Successes:

The model manages to closely recreate the modelled geometry, using formwork to guide the fabric.

Weaknesses:

The detailining is at model scale, which must be addressed at 1:1.

Fig. 211

Fig. 212

Fig. 213

Revised attachment: Structural glulam ribs

Section of fully detailed clay shell formwork

Steel attachment plate

Softwood beam

Bolted connection

Hessian, clamped between the ribs

Bolted connection Attachment plate, flush

Concrete pad founda-

Fig. 218: Steel attachment plate, sandwiched by timber ribs.

Fig. 219: Concrete pad foundation and steel attachment plate, connected to the concrete via cast in threads.

Glulam ribs

Steel attachment plate

Concrete pad foundation

Steel attachment plate

with concrete

Fig. 218

Fig.

Fig. 220

Wind tunnel testing: Aims, apparatus and precedents

Figures 223 and 224: Wind tunnel testing can be conducted at large or small scales.

Smoke introduced into the tunnel acts as a passive tracer, allowing the behaviour of the wind to be studied.

The critical difference is at small scales it is far harder to get the airflow laminar (flowing in parallel lines). This makes results more difficult to observe, but is more true to the turbulent winds that effect buildings.

Physical wind test 1:

Confirming the volumes are open and airtight

Result:

all three were open and airtight.

Side intake 1

Contraction speeds up air due to Bernoulli’s principle

Dry ice as passive tracer

Nozzle directing dry ice more accuractely

Side intake 2

Main intake

Fig. 223

Fig. 224

Fig. 225

Fig. 226

Fig. 227

Bladeless fan

Flow Straightener

Simple tunnel enclosure

Model

Wind Tunnel test rig I constructed

WInd tunnel test 1: Examining how smoke responds to the geometry at high speeds

Limitations:

Fig. 229: The turbulent air made results difficult to observe, as the smoke would disperse quickly.

The test only explores performance on windy days.

Aims: Compare the physical performance of the geometry against the digital.

Result:

The labyrinth was effective in taking in the wind, confirming the digital simulation.

Fig. 228

Fig. 229

WInd tunnel test 2: Still day

Ascertaining the impact of additional suction through the inlet

Result:

Even the slightest suction makes a dramatic improvement on a still day.

No suction from the model

Light suction from the model

Fig. 230

Still day wind testing 2: Exploring the impact of gentle draw through the inlet

Issue revealed in CFD consultation:

‘The worst case scenario is a warm day with no wind - you still need to ventilate the building, so you need a buoyancy system to work aswell.’

Aim: Explore how the air in the chamber responds to the suction.

Result: The inlet created circular movements in the air, and was effective in bringing the smoke into the labyrinth.

Fig.

Fig. 232

Iterations of the plan, in response to consultation and testing

Iteration 1: Slow gradation of temperature through inhabitable wind labyrinth

Consultation feedback: Too much fluctuation

‘...if you want too keep the cool air in during the day, you need to close it, to maintain a stable, constant temperature. You could use your concept, but it comes at the sacrifice of not maintaining such a controlled environment.’

Iteration 2: Responding to the Consultation

Introducing a third space, blending between the workrooms and rare books room

Solution:

- Use the concept from iteration 1 for the workspaces, allowing an open air, thermally dynamic environment.

- Create a sealed library aisle wrapping around the rare books room, with doors helping to stabilise the environments.

Main library and workroom

Sealed rare books room

Labyrinth inlet

Fig. 238: workspace / library corridor

Fig. 239: Tented atrium

Fig. 236

Fig.

Water bath testing precedent:

Breathing Building’s model for the Bloomburg office, explained by its creator

Water bath ventilation testing: Diverging the path of the thermal labyrinth

Owen Connick: ‘This is the most complicated water bath model ever made’.

Aims: Observe how air flows from the inlet through the space.

Method:

- A scaled model is placed in a water bath, filled with hot water, or heated by wires inside the model.

- Ink is injected as a passive tracer to make the water movement apparent.

Result:

The model confirmed that the stack ventilation would be effective, and the hot air would move up through the atrium.

Basic physics of fluid dynamics, and its similarity to airflow:

Hot wires heat the water, causing it to expand fall in density, and rise through the spacwe due to buoyancy.

Air movement is also government by buoyancy, rising at it is heated and its density falls.

The movement of ink in water happens much faster than air in real life, as the lengths are far less.

Aims:

Explore if the air could move through the thermal labyrinth and ventilation tube into the rare books room naturally, or if a fan is required.

Advantage over CFD:

A number of iterations can be tested more quickly.

Boundary conditions do not have to be known.

Method:

1. Three 1:10 simple acrylic models of the space (exploring different lengths and entry geometries) are placed in a hot water bath.

2. Ink is introduced, and its movement is observed.

Fig. 243

Fig. 244

Fig. 245

Fig. 246

Fig.

Test 1: Shortest path

Conclusion:

The ink preceded equally through all parts of the model, suggesting an even airflow.

Fig.

Conclusion:

Water moved much slower through the diverging path, suggesting a weak airflow.

Fig.

Conclusion:

Similar to test 2, the ventilation route received less airflow.

Fig.

Confirming the water bath results with CFD simulation: Highlighting the need for a mechnaical draw

Diverging ventilation route experiences lower wind velocity, suggesting a mechanical fan would be required.

Using the results to develop the architecture

Detailing the wind labyrinth

Work in Progress:

Applying the strategy learned from the waterbath and CFD testing to the occupiable wind labyrinth and rare books ventilation

Sealed doorway from open air workspace

A stair way from a previous design iteration becomes the desk and the opening for the wind inlet.

Opening for the rare books room ventilation Clay / straw insulation seperating the spaces thermally, aside from the ventilation duct

Rare books shelving

Ventilation duct sunken into the ground, containing a solar powered fan for additional draw.

The ventilation rejoins the main labyrinth

Locating the diverging ventilation in the floor of the rare books room.

Fig.

Fig.

Conclusion

I began this project with a determination to create a building that tailored itself specifically to the Catanian climate and impact as minimally as possible on the environment, and sought to use these as the criteria to develop engaging and exploratory architecture, rather than tacking them on at the end.

Through careful examination of precedents I have developed my own take on clay shell construction, explored through a range of 1:1 and scaled prototypes. I believe they offer exciting spatial possibilities at extremely low environmental and financial cost, and hold possibilities far beyond what has been reached in this report.

Continuous use of testing, both real and digital, and at a variety of scales, has allowed me to hone my proposal with each iteration, shown most fully through the development of the labyrinth inlet: Digital geometry informed by wind simulation was translated into a physical model, which in turn was tested, and revealed the need for mechanical ventilation support, adressed through detail drawings. I am excited to use the information from the weighted fabric simulations and water bath testing to develop another 1:20 model, as currently they have only been able to inform how the building is drawn.

Given the complexity of fluid dynamics and ventilation, it is harder to determine the success of the strategy I have developed, but through testing and consultation it has developed in sophistication considerably. The building applies the precedents I have researched and advice from Owen Connick, however to fully understand whether the scheme would work a full, professional CFD study would be required, alongside 1:1 testing, as seen in the Bloomberg building.

The secondary aim of the project - to reveal the all the iterations and embed them in the final proposal - has featured at points, in my studies for the concrete piles, door handles and roof lights. My aim will now be to embed all the iterations and progress developed through this study in my final proposal, so this exploration can be shared with the visitors of the building through their experience, and not just confined to a report.