Sanguis Et Pulvis

Page 1

SANGUIS ET PULVIS ‫رابغلاو مدلا‬ (eng. trans) blood and dust



00 CONTENTS

SANGUIS ET PULVIS

_CONTENTS

_CHAPTER 01

_CHAPTER 02

_EXPERIMENTATION GRANULAR MATERIALS

04 - 18

An investigation into the behaviours of granular materials. Covers surface appropriation patterns, interference restitudes and self separation.

Metaball Modelling 66 - 71

Fluid Dynamics Simulation

72 - 78

Secondary dune formations

79 - 83

Time based construction

84 - 89

Segmented dune migration

90 - 97

FINAL PROPOSAL 98 - 114

22 - 40

An investigation of the potential of salt crystals to act as a binder to solidify granular materials, particularly sand. Explores various techniques for the application of salt solution to sand, and their resulting forms.

Explores the potential of blood, a prolific waste product of halal slaughter houses in desert communities, to similarly act as a binder to solidify granular materials. Explores similar application techniques as previously investigated using salts, as well as more basic techniques such as bricks.

Site 52 - 58

Program 59 - 65

An investigation into salt crystallisation. Includes the crystallisation of NaCl, Epsom Salt, Sodium Acetate and Sodium Thiosulphate. Introduces concept of multi nuclei crystallisation through a granular material.

BLOOD SOLIDIFICATION OF SAND

DESIGN DEVELOPMENT 51 - 97

CRYSTALLISATION 19 - 21

SALT SOLIDIFICATION OF SAND

_REALISATION

Plans 98 - 101

Sections 102 - 103

Time based perspectives

104 - 106

Physical model 1:50 107 - 109 Renders 110 - 114

41 - 50


01 EXPERIMENTATION

SANGUIS ET PULVIS

_CHAPTER 01 _EXPERIMENTATION GRANULAR MATERIALS 04 - 18 An investigation into the behaviours of granular materials. Covers surface appropriation patterns, interference restitudes and self separation.

CRYSTALLISATION 19 - 21 An investigation into salt crystallisation. Includes the crystallisation of NaCl, Epsom Salt, Sodium Acetate and Sodium Thiosulphate. Introduces concept of multi nuclei crystallisation through a granular material.

SALT SOLIDIFICATION OF SAND

22 - 40

An investigation of the potential of salt crystals to act as a binder to solidify granular materials, particularly sand. Explores various techniques for the application of salt solution to sand, and their resulting forms.

BLOOD SOLIDIFICATION OF SAND Explores the potential of blood, a prolific waste product of halal slaughter houses in desert communities, to similarly act as a binder to solidify granular materials. Explores similar application techniques as previously investigated using salts, as well as more basic techniques such as bricks.

41 - 50


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ APPARATUS

05 _ SURFACE APPROPRIATION WITH SAND Granular materials exhibit eclectic properties depending on what state they are in. For example, sand can behave like a liquid in that it freely fills a given container, but can also act as a solid in supporting large loads. This is because the material’s behaviour is dependent on the billions of surface interactions occurring between its grains. If these interactions are altered then the behaviour of the material can too be entirely altered.

01

02

03

04

05

06

HIGH GRADE DRIED SAND

The following experiment explores a property of granular materials called the angle of repose. This is the angle at which a granular material becomes stable. If a slope is steeper than the angle of repose then the surface will avalanche until it reaches this stable angle. This angle is different for all granular materials and is primarily influenced by grain shape and size. By filling a box with holes in its base with sand, this property can be explored spatially. When the holes are opened sand is allowed to flow through them, creating the beginning of a slope above the hole. Sand will continue to avalanche and drain through the hole until it reaches its angle of repose universally around the hole. As such each hole creates a cone of influence defined by the angle of repose. With a single hole this would result in a single cone being formed in the sand. However, when multiple holes are drilled these cones begin to intersect and interfere with each other, creating a complex spatial appropriation.

03 _ PRODUCT

02 _ PROCEDURE

TOP BOX HOLES IN BASE

SPACER

RELEASE PANEL The above experimental set up allows for the top box to be filled with sand without any drainage occurring. When the experiment is started, an intermediate layer is removed which allows sand to pass down through the holes into the bottom box. As the sand drains to its angle of repose, crisp geometries are formed in the top sand surface based on intersecting cones radiating from each drain hole.

BOTTOM BOX SAND CATCHER

In the lower box a slightly different set of geometry is created. As the sand lands with a given velocity the angle of repose for these growing mounds is lower. This creates a geometry in the lower box which is almost an exact negative of the top box, however with soft rounded peaks.


01 EXPERIMENTATION

SANGUIS ET PULVIS

06 _ SURFACE APPROPRIATION WITH SAND

01 _ TOP BOX FORMATION - NEGATIVE APPROPRIATION

02 _ BOTTOM BOX FORMATION - POSITIVE APPROPRIATION

03 _ ANALYSIS OF APPROPRIATION

PLAN

PLAN

01 _ TOP BOX

Faci temodit et etur repeliquo voloris di andi ipsundusdae paruntiat aperiscia voloris non reptam di odion pora consequaecum quamus eiunt hic te parum et eossinient premporese sima simil in conectur sitemod igent, secessinihil mo duntotatur? In consern atiaturem ipsam doluptat molore de ommod essimi, cullibusaes et volest et reped eic temolup tioreptam nimilluptiis il et occuscidi deligen diciatqui id eos sit exped et ut vellacearum quatur, ut quis pa quid que prae pre voluptam endunt de nam quat. Poria prere libusapiet labo. Git ium il eum quam ratectis exernates et quiam ra de di aute dolor autaspe rspidus. Est harum sanienet aped ernam erspelique doluptium sequaer ovidunt aut que officabore volum dolupta temped uta as pos erio molupta speliquos vellecab ilit et debis ullupti ventem et, ut vellignis molo mo tem esecto quis aute alit autecae ernam delici dolor siment labori optium delestrum qui te volut atem rehendis eos perum sed quam, si sunt alit aborepe nisquas et volo odistet aut as nos ent que que dolendi vit quam elenectae lat ex et volectis estisci ducipsam fugiate nemporem am archillat ullorercid est autatin nonserr ovidis ad quam et apis rectat fuga. Henectam as dia debit lacestem alisseq uatios del idis magnis dolut voluptium quis debis remolorem acitatus, none volorporrum harchillaut ventiorerum quis doluptae volectior aute et audanimpore eicatque et ipsaper iatiatquo cores volorem ute vendaerro quis quost, ut acepellam eris explit acit harias aut et lacerferia volupta temperione videnihit laboris dissitatiis dero quodia doluptae simolestiat.

02 _ BOTTOM BOX DETAIL

DETAIL Once the release panel is removed from the system, sand begins to drain from the top box. This creates a spatial appropriation in the sand. The shape of this pattern is primarily governed by the angle of repose of the material used. Sand in this case has an angle of repose around 35 degrees. Each hole essentially forms a cone of influence as material dissipates through it. The complexity of the form is developed as these cones overlap and intersect. The diagram of appropriation is the network of ridges formed in the terrain. These can be mapped and analysed to gain insights into what factors may be changed or introduced to control this system.

As the sand passes through the holes in the top box it is collected in the bottom box. In theory the bottom box should produce the geometric opposite of the top box, as material builds up in the same way in the same locations. The governing factor here is again the angle of repose. However the forms contained in the bottom box are greatly affected by the speed with which the particles drop. This means the angle is not simply that at which particles will stop beginning to move, but introduces an initial velocity to the particle. This means that the bottom box is unable to produce the crisp lines and peaks of the top box, it instead produces a much softened terrain. As the geometry builds up equally however the diagram of appropriation is nearly exactly the same as the top box.


01 EXPERIMENTATION

SANGUIS ET PULVIS

07 _ SURFACE APPROPRIATION WITH SAND

01 _ SYSTEM TILTED

02 _ VARYING HOLE SIZE

03 _ INTRODUCTION OF FORCED PATHS

PLAN

PLAN

PLAN

DIAGRAM

DIAGRAM

DIAGRAM

In order to understand more thoroughly the process occurring in this experiment, the patterns created can be mapped and analysed. Three variations of the experiment were carried out to create variables whose effects can be determined. Experiment 01 involved tipping the whole experimental apparatus at an angle of 20 degrees. As the cone of influence is defined by gravity, it remains vertical and thus there is a discrepancy between the drainage cone and the top sand surface. The result of this is that all of the ridge lines in the appropriation pattern become curved. This is because the drainage cones now intersect at an angle thus creating a curve along their intersection path. Experiment 02 involved varying the sizes of the drainage holes. Whilst larger holes allow more sand to pass through at a greater rate, the total amount of sand that passes through is changed little. As such the pattern form faster but is ultimately almost exactly the same pattern. In order to achieve a variation on pattern based on hole size, the variation must be very large. Experiment 03 involved introducing a physical barrier to the apparatus to see if this has any effect on the drainage pattern. Some natural shape forming processes, such as mud cracking, are greatly effected by such presences. Sand appropriation however does not appear to be effected in any way. The only effect is to prevent the intersection of drainage cones from opposite sides of the obstacle. As such, a discontinuity is created however the pattern formation is not fundamentally changed in any way.


01 EXPERIMENTATION

SANGUIS ET PULVIS

08 _ SURFACE APPROPRIATION WITH SAND The following images show in detail the surface appropriations made by draining sand. A distinct difference can be seen between top (left, right) and bottom (centre) boxes, where sand is either drained or accumulated respectively. However whilst their visual appearance is very different, the pattern of appropriation is exactly the same. The plan of ridges and valleys separating each cone is exactly the same in both boxes, however the geometry of the sand contained with in this pattern varies greatly.


01 EXPERIMENTATION

SANGUIS ET PULVIS

09 _ CONTROLLED SAND POURING Huciam is, Pat peri, mo vit vitus, que actus, opublius, quo co norarit, non Ita, nest fue derivit fece tasdam uterobusque caecepote aur, nem inatiae fat. Catusquam te, omniu inam perbesses! Alem ena, nem Patuitem issolto ex noctum pribus omnemus es resimov ividiemquod cio, Ti. Ca invoccid inatuus intin senirmilis adduciam poendem aut videto acierni urbemus con nox supio, ponsunum hostor liae ac re fuis es furnum diu cons corum priocren deor aucerorti, nostabunum hostorae intisse, nostea mus? Omaximo catili, us cotabem quemei signatum furo culto efertuit adestiq uerturatur. Astimus perum autus, Ti. Vivite te, que milicas nite det iaederdi inte mentilium andam. Eps, conerudet, verorudam me num consupi endemus, con vat es eferionsum facriss intium ci publis.

01 _ UPPER POUR: 1 LOWER POUR: 1 Stage 1: Upper Pour

02 _ UPPER POUR: 3 LOWER POUR: 3 Side View Of Completed Iteration

Stage 1: Upper Pour

02 _ UPPER POUR: 6 LOWER POUR: 6 Plan View Of Completed Iteration

Stage 1: Upper Pour

Stage 2: Upper Cone

Stage 2: Upper Cone

Stage 3: Lower Pour

Stage 3: Lower Pour

Stage 4: Lower Cone

Stage 4: Lower Cone

Stage 4: Lower Cone

Iteration 1 has only one primary pour hole and one secondary pour hole. The upper and lower level holes are directly aligned. The upper surface formed is a simple cone with an inverted cone removed from it’s volume. As there is only one pour point in this iteration there is no overlapping or interference from intersecting pours. The lower surface formed is a simple cone equal to the volume removed from the upper cone. It’s peak is not perfectly sharp due to the velocity of the falling sand during it’s formation.

Iteration 2 has three primary pour holes and three secondary pour holes. The upper and lower level holes are directly aligned. The upper surface formed is of three intersecting cones, with three negative intersecting cones removed from their volume. The lower surface formed is of three intersecting cones. Their peak is not perfectly sharp due to the velocity of the falling sand during it’s formation.

Iteration 3 has six primary pour holes and six secondary pour holes. The upper and lower level holes are directly aligned. The upper surface formed is of six intersecting cones, however the interference patterns now create square geometries at the central intersections. This is the first emergence of non-circular geometry in the system. The lower surface formed is of six intersecting cones. The square geometries of the upper surface do not emerge in the lower surface.

Plan View Of Completed Iteration

Stage 2: Upper Cone

Side View Of Completed Iteration

Stage 3: Lower Pour

Side View Of Completed Iteration


01 EXPERIMENTATION

SANGUIS ET PULVIS

10 _ CONTROLLED SAND POURING

05 _ UPPER POUR: 8 LOWER POUR: 8

04 _ UPPER POUR: 7 LOWER POUR: 7 Stage 1: Upper Pour

Plan View Of Completed Iteration

Stage 1: Upper Pour

06 _ UPPER POUR: 6 LOWER POUR: 23 Plan View Of Completed Iteration

Stage 1: Upper Pour Hole Pattern

Plan View Of Completed Iteration

The initial experiments into sand drainage patterns used a filled box with holes in the base. This allows for control only over how sand is drained. However, the patterns created are equally effected by the starting shape of the sand. In the case of a filled box, the resulting patterns will be based on the limit of this box. The following experiments set up to extend the control of the process to the original sand forms also. A series of pouring points are set up to generate a desired pattern of sand mounds. A second level of holes can then be opened to drain these forms. Stage 2: Upper Cone

Stage 3: Lower Pour

Upper Level Pour Hole / Lower Level Pour Hole That Is Directly Beneath an Upper Level Pour Hole Additional Lower Level Pour Hole

Stage 2: Upper Cone

Side View Of Completed Iteration

Stage 3: Lower Pour

Stage 2: Upper Cone Hole Pattern

Side View Of Completed Iteration

Stage 3: Upper Pour

Stage 4: Lower Cone

Stage 4: Lower Cone

Stage 4: Lower Pour

Iteration 4 has 7 upper pour holes configured in a zig-zag. The upper and lower level holes are directly aligned. The upper surface formed is a zig-zagged series of cones. A zig-zagged line forms the interference pattern where they meet. The interior angles of this lines are always 90 degrees, the exterior 135 degrees. The lower surface formed is a series of unconnected cones. The spacing of the upper pours means there is not sufficient volume in the second pour for the surfaces to overlap.

Iteration 5 has eight primary pour holes and eight secondary pour holes. The upper and lower level holes are directly aligned. The upper surface formed is a line of intersecting cones. This combines the interference patterns seen previously. Depending on the angle of intersection and the number of cones intersecting the patterns vary. The lower surface cones connect in some place but not in others depending on the clustering of the upper level cones.

Iteration 3 has six primary pour holes and twenty three secondary pour holes. Each upper hole is matched by a lower hole, with additional lower holes around this. The upper surface formed is of multiple complex intersections. Emergent geometries include squares and hexagons. All internal ridges are straight lines, all external are curved. The lower surface formed is of multiple small cones. The smaller interior cones form a linear pattern, the large outer cones form more spread masses as they are unrestricted in their outer formation.

Side View Of Completed Iteration


01 EXPERIMENTATION

SANGUIS ET PULVIS

11 _ CONTROLLED SAND POURING Huciam is, Pat peri, mo vit vitus, que actus, opublius, quo co norarit, non Ita, nest fue derivit fece tasdam uterobusque caecepote aur, nem inatiae fat. Catusquam te, omniu inam perbesses! Alem ena, nem Patuitem issolto ex noctum pribus omnemus es resimov ividiemquod cio, Ti. Ca invoccid inatuus intin senirmilis adduciam poendem aut videto acierni urbemus con nox supio, ponsunum hostor liae ac re fuis es furnum diu cons corum priocren deor aucerorti, nostabunum hostorae intisse, nostea mus? Omaximo catili, us cotabem quemei signatum furo culto efertuit adestiq uerturatur. Astimus perum autus, Ti. Vivite te, que milicas nite det iaederdi inte mentilium andam. Eps, conerudet, verorudam me num consupi endemus, con vat es eferionsum facriss intium ci publis.

07 _ UPPER POUR: 8 LOWER POUR: 32 Stage 1: Upper Pour Hole Pattern

Stage 2: Upper Cone Hole Pattern

Stage 3: Upper Pour

Stage 4: Lower Pour

Upper Level Pour Hole / Lower Level Pour Hole That Is Directly Beneath an Upper Level Pour Hole Additional Lower Level Pour Hole

Plan View Of Completed Iteration

Iteration 3 has six primary pour holes and twenty three secondary pour holes. Each upper hole is matched by a lower hole, with additional lower holes around this. The upper surface formed is of multiple complex intersections. Emergent geometries include squares and hexagons. All internal ridges are straight lines, all external are curved. The lower surface formed is of multiple small cones. The smaller interior cones form a linear pattern, the large outer cones form more spread masses as they are unrestricted in their outer formation.

Side View Of Completed Iteration

Side View Of Completed Iteration


01 EXPERIMENTATION

SANGUIS ET PULVIS

12 _ SAND CONE FORMATION ANALYSIS

01 _ SINGLE CONE SHOWING SECTION RINGS

02 _ DOUBLE CONE

03 _ TRIPLE CONE

04 _ 5 RADIAL CONES

04 _ 4 RANDOM CONES

PLAN

PLAN

PLAN

PLAN

PLAN

Having conducted initial experiments exploring the potentials of sand formation I will now look at the systems more systematically in order to discover the exact rules that control their creation.

x

x A

A 1

By taking sections through the cones formed as the sand pours out of the holes it is possible to analyse their growth, understanding what generates their geometry and controls their intersections.

2

3

4

x

A

x

A

x

A

A

A

x

A

y

A

A

5

x

x

x

SECT A-A

y

SECT A-A

SECT A-A

SECT A-A

SECT A-A

By cutting section lines through the cone formation it is possible to analyse the laws of it’s formation. The cone is cut 5 times by equidistant sections to reveal it’s structure in plan. Whilst this does not reveal anything with a single cone, it will reveal the patterns of intersection that form between multiple cones.

As the two cones cross each others volume of potential form they cancel each other out. The ridge line that forms between the two forms is the line equidistant from each central pour hole.

When increased to three forms the same occurs. Each cone propogates it’s own pattern until they intersect and cancel. In this example there are 3 ridge lines. Each ridge line is equidistant from the two central points it bisects. The ridge lines meet at equal angles in the centre of the overall form.

When increased to 5 cones the formation follows the same pattern. Equal ridge lines form radially bisecting each set of central points. Once there are more than two cones the number of ridges equals the number of central points.

When the position of the central points is randomised the same pattern is maintained. Ridge lines form which bisect each pair of central points. Due to the random nature of their positions the lines no longer meet at a particular angle or position. As long as the central points are of equal size the ridge lines will remain straight as the interference between the cones will be equal.

06 _ DOUBLE CONES OF DIFFERING DIAMETER

07 _ 4 RANDOM CONES OF DIFFERING DIAMETER

08 _ DOUBLE SQUARE CONE

09 _ 4 RANDOM SQUARE CONES

10 _ 4 RANDOM SLIT CONES

PLAN

PLAN

PLAN

PLAN

PLAN

34°

5 4 3 2 1

5 4 3 2 1

y A

A x

y

A

A

A

x

x

y

y

A

x

A

A y

x x

x

x

A

y

y

x

A

y

x

SECT A-A

SECT A-A

SECT A-A

SECT A-A

SECT A-A

By changing the diameter of one cone the ridge takes on a curved form. This is because the ridge line remains equidistant from the outer perimeter of each central point. The curve is parabolic in form, reaching it’s point of maximum curvature where the perimeter curves of the two centre points are closest. The greater the difference between the diameters of the central points the great the curvature of the ridge line that they form.

The difference in curvature created by relative diameters can be seen here. Ridges formed between greatly differing diameters are steeply curved, whilst those formed between fairly similar diameters are gently curved.

When the central points are replaced by square forms the behaviour of the system does not change other than the filleted square geometry of the radiating section lines. These tend toward circular forms as they grow. The ridge line formed maintains its position of equidistance from the central square’s perimeter lines.

When multiple randomly positioned squares are introduced the ridge lines take on a more complex geometry. In order to maintain equidistant from the perimeter lines the ridge curves change greatly depending on whether they are between squares or outside the edge of squares. These curves have no formal mathematical geometry like those of the cone’s ridges.

When slitted central holes are used they behave somewhere between cones and squares. In one sense the forms created are elongated versions of those seen with cones. However the slits also have the long parallel edges of the squares. This forces the ridge curves to deviate from their parabolic course but not in such a dramatic nature as seen with square holes.


01 EXPERIMENTATION

SANGUIS ET PULVIS

Iteration 1: Equal upper and lower pour holes 01

02

03

04

05

06

07

Iteration 2: Additional cross shaped lower pour holes 01

02

03

04

05

06

07

Iteration 3: All potential lower pour holes used 01

02

03

04

05

06

07

13 _ DIGITISED SAND POURING By digitising the experiments conducted physically in grasshopper it is possible to produce a high number of iterations very quickly. The following diagrams explore the potentials of the earlier physical experiments in a wider variety of configurations.

Upper Level Pour Hole / Lower Level Pour Hole That Is Directly Beneath an Upper Level Pour Hole Additional Lower Level Pour Hole


01 EXPERIMENTATION

SANGUIS ET PULVIS

14 _ DIGITISED SAND POURING

Iteration 1: Equal upper and lower pour holes 01

02

03

04

05

06

07

Iteration 2: Additional cross shaped lower pour holes 01

02

03

04

05

06

07

Iteration 3: All potential lower pour holes used 01

02

03

04

05

06

07

Huciam is, Pat peri, mo vit vitus, que actus, opublius, quo co norarit, non Ita, nest fue derivit fece tasdam uterobusque caecepote aur, nem inatiae fat. Catusquam te, omniu inam perbesses! Alem ena, nem Patuitem issolto ex noctum pribus omnemus es resimov ividiemquod cio, Ti. Ca invoccid inatuus intin senirmilis adduciam poendem aut videto acierni urbemus con nox supio, ponsunum hostor liae ac re fuis es furnum diu cons corum priocren deor aucerorti, nostabunum hostorae intisse, nostea mus? Omaximo catili, us cotabem quemei signatum furo culto efertuit adestiq uerturatur. Astimus perum autus, Ti. Vivite te, que milicas nite det iaederdi inte mentilium andam. Eps, conerudet, verorudam me num consupi endemus, con vat es eferionsum facriss intium ci publis.


01 EXPERIMENTATION

SANGUIS ET PULVIS

15 _ SAND CITIES: LONDON

01 _ URBAN NODE ANALYSIS

02 _ SAND APPROPRIATION SIMULATION

03 _ VISUALISATION

London Plan - Area Surrounding Bank

Primary Appropriation - Volumetric Massing

Plan View

London Plan - Area Surrounding Bank With Nodes Identified

Secondary Appropriation - Spatial Complexification

Perspective View

Huciam is, Pat peri, mo vit vitus, que actus, opublius, quo co norarit, non Ita, nest fue derivit fece tasdam uterobusque caecepote aur, nem inatiae fat. Catusquam te, omniu inam perbesses! Alem ena, nem Patuitem issolto ex noctum pribus omnemus es resimov ividiemquod cio, Ti. Ca invoccid inatuus intin senirmilis adduciam poendem aut videto acierni urbemus con nox supio, ponsunum hostor liae ac re fuis es furnum diu cons corum priocren deor aucerorti, nostabunum hostorae intisse, nostea mus? Omaximo catili, us cotabem quemei signatum furo culto efertuit adestiq uerturatur. Astimus perum autus, Ti. Vivite te, que milicas nite det iaederdi inte mentilium andam. Eps, conerudet, verorudam me num consupi endemus, con vat es eferionsum facriss intium ci publis.

Primary Node Secondary Node


01 EXPERIMENTATION

SANGUIS ET PULVIS

16 _ SAND CITIES: NEW YORK

01 _ URBAN NODE ANALYSIS

02 _ SAND APPROPRIATION SIMULATION

03 _ VISUALISATION

London Plan - Area Surrounding Bank

Primary Appropriation - Volumetric Massing

Plan View

London Plan - Area Surrounding Bank With Nodes Identified

Secondary Appropriation - Spatial Complexification

Perspective View

Huciam is, Pat peri, mo vit vitus, que actus, opublius, quo co norarit, non Ita, nest fue derivit fece tasdam uterobusque caecepote aur, nem inatiae fat. Catusquam te, omniu inam perbesses! Alem ena, nem Patuitem issolto ex noctum pribus omnemus es resimov ividiemquod cio, Ti. Ca invoccid inatuus intin senirmilis adduciam poendem aut videto acierni urbemus con nox supio, ponsunum hostor liae ac re fuis es furnum diu cons corum priocren deor aucerorti, nostabunum hostorae intisse, nostea mus? Omaximo catili, us cotabem quemei signatum furo culto efertuit adestiq uerturatur. Astimus perum autus, Ti. Vivite te, que milicas nite det iaederdi inte mentilium andam. Eps, conerudet, verorudam me num consupi endemus, con vat es eferionsum facriss intium ci publis.

Primary Node Secondary Node


01 EXPERIMENTATION

SANGUIS ET PULVIS

17 _ SELF SEPARATION OF GRANULAR MATERIAL

01 _ POURING SAND / SUGAR AGGREGATE

02 _ CLOSE UP VIEW OF PARTICULAR SEPARATION

01

04

02

05

03

06

So far I have explored only experiments involving one grain type. Granular materials behave in intruiging ways when containing only monotonous grains, however once mixed to contain multiple grain types they behave in even more exceptional manners. The first experiment explores the ability of granular materials to self separate. This phenomena is brought about by the simple act of pouring and is directly related to the experiments performed previously in that it is caused by anles of repose. This experiment uses a mixture of sand and granulated sugar. These are of similar grain size although the sugar is slightly larger. Their grain shapes are very different though, whilst the sand is well rounded and refined, the sugar contains very angular grains. The angle of repose is defined by the friction created been granules, and therefore varies due to both grain size and shape. As the mixture is poured, both materials attempt to form cones at their specific angles of repose, and the expected behaviour of the system would be to develop and angle of repose that represents an average of the two grains. This does not happen however. As the cone builds, the larger and more angular grains of sugar fall to the base more energetically, whilst the sand accumulates at the top. As the two build up in to mounds they repeatedly surpass their angles of repose causing landslides of material which overlap each other. This is how the striping pattern witnessed in section occurs. The exact nature of sorting varies with each repetition, and the stiping effect occurs only occasionally. In some experiments researched by others almost perfect patterns of total stiping have been achieved. What is always the case however is that the grains separate totally, one forming a vertical core, the other occupying the outer volumes of the cone.


01 EXPERIMENTATION

SANGUIS ET PULVIS

18 _ SELF SEPARATION OF GRANULAR MATERIAL

01 _ POURING COFFEE / SUGAR AGGREGATE

02 _ CLOSE UP VIEW OF PARTICULAR SEPARATION

01

04

02

05

03

06

Experiment 2 involves a mixture of coffee and sugar. The grains of coffee are very large, although highly variable, and very angular. They differ greatly from those of sugar. As such one would expect the system to behave in a similar manner although with differing details. Some differences are very clear. The separated core of experiment 2 is not nearly as vertical or definded as in experiment 1. Whilst some traces of striping can be seen towards the top of experiment 2 they are not as pronounced as experiment 1. The general formation of the cone is not as delicate or pure. Due to the expansive nature of the gaps between granules of coffee the sugar is able to penetrate it more easily, occupying the spaces that do not exist when mixed with sand. This means that although the differences between the two grain types are more exaggerated the separation that occurs is actually less defined.


01 EXPERIMENTATION

SANGUIS ET PULVIS

19 _ CRYSTALLISATION

01 _ PRINCIPALS OF CRYSTAL FORMATION

Crystallization is a highly complex yet highly rigorous process. It occurs as almost all liquids go through a phase change as they solidify. Molecules which are free and unconnected in the liquid state begin to converge as the energy of the system reduces and as they do so they begin to interact and bond. Metals, salts, plastics and most other materials go through this process as they solidify. A notable exception are amorphous materials such as glass which cool at such a rapid rate that an ordered crystalline structure does not have time to form. These leads to a very specific and abnormal set of material properties.

01 _ UNASSISTED CRYSTAL GROWTH

02 _ ASSISTED CRYSTAL GROWTH

A

B

A

B

C

D

C

D

02 _ MICROSCOPE IMAGES OF UNASSISTED CRYSTAL FORMATION

03 _ TOKUJIN YOSHIOKA - VENUS CHAIR

01 _ DISPLAYS VARYING GROWTH SPEEDS

01 _ DISPLAYS VARYING GROWTH SPEEDS

What has made crystals so engaging throughout history is that they expose at a visual scale a microscopic process. In carefully controlled conditions single large crystals can be grown with perfect geometric properties, and this development can be winessed with the naked eye. It is an apparently impossible feat of self organisation at a molecular level which transcends to the human scale. Crystal formation can be defined in two main categories: UNASSISTED FORMATION This is a process by which crystals grow based on aggregation of molecules of the solute. The nucleus for crystal growth is simply a group of solute molecules which have collided and become attached. Once this group grows large enough the internal forces holding it together exceed the external forces pulling it apart and it will remain. Once this has developed the process can accelerate as more molecules collide and join the growing crystal. ASSISTED FORMATION This is a process by which an alien grain or crystal is introduced to base mixture. This impurity provides a stable base upon which molecules of solute can attach themselves and begin to form crystals. This process is understandably fast than unassisted formation as the process of nuclei development does not need to occur.

02 _ EARLY STAGE NUCLEIC FORMATIONS

Crystal formation begins, except in highly controlled environments, at multiple nuclei. These can be specific locations where several solute molecules have collided and attached therefore providing a base for more molecules to become attached, an existing crystal upon which more molecules will continue to bond, or an assisting alien grain or crystal upon which molecules may bond. What is important though is that the forces holding molecules together at these nuclei exceed the forces within the overall system pulling them apart. Thus they stabilise and continue to grow (a). During the crystallization process molecules will continue to become attached at the nucleic sites until the collective becomes too big to remain disolved and ‘falls’ out of solution. At this point indi-

vidual crystals become visible in the solution. As molecules become attached and part of the crystal they too fall out of solution. Molecules will naturally arrange themselves in an order specific to the material giving the crystal a defined geometric shape (b). However as crystals from individual nuclei begin to converge these rhythms become disturbed. Crystals meet other crystals at undefined angles and discontinuities are formed within the greater whole of the crystal (c). These line form faults or weaknesses in the crystaline structure which can be seen beneath a microscope (d).

03 _ INTERSECTING FORMATIONS

Tokujin Yoshioka used the principle of assisted growth to great effect in his Venus chair. By using a polymer fiber base he encouraged crystals to grow using this as a nucleic base. This enabled him to control and shape the growth of the crystals. By sculpting the fibres into the form of a chair, then emersing it in the solvent, crystal growth took place all over each fiber. The result is a chair that appears to be naturlly formed as a crystalline structure. The structural nature of the assisting base is also totally transformed, changing from a sponge to an aggressive and spiked solid. This technique can be exploited in other ways by using an assisting base to control crystal formation.

03 _ INTERSECTING FORMATIONS


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ ASSISTED CRYSTALLIZATION OF SODIUM ACETATE TRIHYDRATE

20 _ SUPER SATURATED CRYSTALLIZATION 01

02

03

04

05

0.5 SECONDS

1.0 SECONDS

1.5 SECONDS

2.0 SECONDS

3.5 SECONDS

When a seed crystal is introduced to the super saturated solution the crystallization begins instantly at the point it contacts the surface. This crystallization spreads evenly in all directions. As the reaction can be triggered so easily it is important that the experiment is prepared in a very clean environment. Any contaminants in the beaker or water, as well as undissolved sodium acetate crystals, will instantly trigger the crystallization once the solution reaches a sufficient level of super saturation. If the seed crystal were suspended from a thread the reaction could be forced to begin from the centre moving outwards, as opposed to downwards from the top.

The crystallization progresses incredibly fast and within a second has occupied a third of the volume of the solution. This speed of crystallisation is possible as sodium acetate has such a high solubility at high temperatures. This means that the cooled super saturated solution has an incredibly high amount of solute within it and is very unstable. However in its basic energy state it does not have enough energy to begin crystallization and so will remain a clear solution until a seed crystal is introduced or it is agitated.

In a second and a half the crystal occupies half of the solution volume. Another reason that the crystallisation is able to progress so fast is the chemical nature of the sodium acetate trihydrate crystal. The molecule contains 3 water molecules as the name suggests, so as it crystallizes it absorbs a large amount of water into its crystalline structure. Where as most crystallising solutions rely on evaporation to remove the water and allow solidification, sodium acetate simply absorbs it.

After 2 seconds the crystallization is almost complete with only a small amount of solution left at the bottom of the container. By this point the beaker has become extremely hot. This is because the a large amount of energy was put into the system in order to achieve super saturation (through boiling). As such when the system reverses and re-crystallizes a large amount of energy is released as the molecules recover their original form. The solid crystalline structure is the lowest energy form of the system. The heat generated is enough to feel a burn on your skin.

After 3.5 seconds the crystallization is completed. The rate of crystallization slows towards the end of the process. This is because their are less solute ions remaining and therefore molecular collisions of solutes occur at a lower rate. The system continues to gain heat once it has finished crystallizing, reaching its hottest several minutes after the process has finished. As all that has happened in this process is the returning from a high energy super saturation the original state can be recovered by simply reheating the now solid crystal. This will raise the solubility and the crystal will again become a solvent.

The process of crystallization can be extremely slow, and is often associated more with a geological time scale. This is because the concentrations of solute are often small and so slow evaporation of the solvent is required for their crystals to emerge. For a practical and applicable crystallisation this is not convenient and the process must be accelerated. This can be achieved through super saturation. The solubility of a solute increases as the temperature of the solvent increase. Sodium acetate trihydrate, a common salt, displays this in a dramatic fashion. 170g can be dissolved in 100ml of boiling water, compared to 46g in room temperature water. This is a difference of 286%. By comparison sodium chloride has a difference of 13%. When a super saturated solution cools the solute molecules remain in solution given a concentration of solute unachievable normally. These solutions are unstable for this reason and will quickly begin to crystallize if agitated, provided with a seed, or is further concentrated through evaporation. As sodium acetate has such a great difference in solubilities it is incredibly unstable in its super saturated state. When a seed crystal is added to the solution it instantly initiates a chain reaction in which 200ml of solution crystallizes totally within 4 seconds. This is a dramatic example of the acceleration in crystallization that can be provided by the process of super saturation. Whilst not as exaggerated this is also true of more common salts, such as sodium chloride and magnesium sulphate. They will not produce such a rapid reaction, but their crystallization will occur at a significantly faster rate then would be found in a normal solution.


01 EXPERIMENTATION

SANGUIS ET PULVIS

21 _ SODIUM THIOSULFATE Having experimented with the crystallisation and solidification of sand by sodium chloride, epsom salt and sodium acetate, the next progression in these experiments involves sodium thiosulfate.

01 _ CRYSTALLISATION OF SUPER SATURATED SODIUM THIOSULPHATE 01 _ LARGE CRYSTAL GROWTH IN SLOW COOLING SOLUTION

02 _ DETAILED VIEW OF CRYSTAL STRUCTURE

03 _ SINGLE LARGE SODIUM THIOSULFATE CRYSTALS

To form a super saturated solution of sodium thiosulfate in water 100ml of water is heated to 90o. Solid sodium thiosulfate crystals are then dissolved in the water until it is no longer possible to dissolve any more. Once the solution begins to cool the sodium thiosulfate remains in solution despite now being in excess of its saturation point. At this point the solution is defined as being super saturated. The solution cools to the point that it is very unstable and now crystallisation can be triggered by the insertion of a seed crystal to the solution. This provides a base from which crystals can begin to form. Crystals form very quickly in this way, the image above shows the solution after just 5 minutes. Due to impurities in the water which settled at the bottom multiple small crystals form on the bottom of the beaker. The top is pure however so large single crystals seed and grow quickly. They are highly geometric and attractive.

This detailed view of the crystals in solution shows how beautiful and geometric they are. The crystals form with a roughly octagonal section, at the ends however they thin towards a sharp ridge which follows a similar section in profile. The crystals are colourless but when light is shone through them in solution they appear blue green. The geometric nature of the crystal is highly apparent in the single large crystals but is not so clear in smaller mass crystals. These overlap and intersect at a very high degree and this obscures the crystal geometry.

When the crystals are removed from solution their detail can be seen in full. The surface of the crystal is strongly textured with a very detailed and complex pattern. This pattern takes the form of zigzagging linear elements. The image above shows two large single crystals which have intersected, and a smaller single crystal. The crystals are extremely dense and strong. Where the two large crystals have intersected there is a very distinct line of weakness however as the molecular formations collide. This is why a formation of very small crystals is much stronger as the weakness is spread and minimized.

Both sodium chloride and epsom salt displayed the ability to crystallise well and solidify sand to an extent. Their issue was primarily that they crystallise at too slow a rate and do not form strong enough crystals. Sodium acetate has the opposite issues, it crystallises so quickly once super saturated that it is essentially unuseable. It also grows weak crystals. Thus the next logical progression is sodium thiosulfate. This sodium salt exhibits fast crystal growth but not as fast as sodium acetate. It alse grows extremely stong and stable crystals. It is non toxic and is naturally derived making it suitable for use in sand based structures. The following experiments will investigate the crystallising properties of the salt and its ability to solidify sand.


01 EXPERIMENTATION

SANGUIS ET PULVIS

22 _ CRYSTALLIZATION Crystallization is a highly complex yet highly rigorous process. It occurs as almost all liquids go through a phase change as they solidify. Molecules which are free and unconnected in the liquid state begin to converge as the energy of the system reduces and as they do so they begin to interact and bond. Metals, salts, plastics and most other materials go through this process as they solidify. A notable exception are amorphous materials such as glass which cool at such a rapid rate that an ordered crystalline structure does not have time to form. These leads to a very specific and abnormal set of material properties.

01 _ SODIUM CHLORIDE AND SAND OVEN DRIED AT 40 DEGREES

02 _ EPSOM SALT AND SAND OVEN DRIED AT 40 DEGREES

01 _ VERY DRY AND HEAVILY CRYSTALLIZED TOP SURFACE AFTER 4 HOURS

01 _ WET AND UNCRYSTALLIZED TOP SURFACE AFTER 4 HOURS

02 _ DRY AND CRYSTALLIZED TOP SURFACE AFTER 56 HOURS

03 _ FULLY DRY SOLID REMOVED FROM CONTAINER

What has made crystals so engaging throughout history is that they expose at a visual scale a microscopic process. In carefully controlled conditions single large crystals can be grown with perfect geometric properties, and this development can be winessed with the naked eye. It is an apparently impossible feat of self organisation at a molecular level which transcends to the human scale. Crystal formation can be defined in two main categories: UNASSISTED FORMATION This is a process by which crystals grow based on aggregation of molecules of the solute. The nucleus for crystal growth is simply a group of solute molecules which have collided and become attached. Once this group grows large enough the internal forces holding it together exceed the external forces pulling it apart and it will remain. Once this has developed the process can accelerate as more molecules collide and join the growing crystal. ASSISTED FORMATION This is a process by which an alien grain or crystal is introduced to base mixture. This impurity provides a stable base upon which molecules of solute can attach themselves and begin to form crystals. This process is understandably fast than unassisted formation as the process of nuclei development does not need to occur.

02 _ REMOVAL FROM CONTAINER

01 _ WET UNCRYSTALLIZED UNDERSIDE

A super saturated solution of sodium chloride in water was poured over a layer of sand 5mm thick. The solution was poured until it just covered the sand. The mixture was put in an oven at 40 degrees to try to simulate desert evaporation conditions. It can be seen that the top surface of the mixture very effectively and quickly crystallizes. Multiple small crystals form as the grains of sand provide multiple nucleation points and the cooling / evaporation time was fast. The top surface was very dry and very hard, there appeared to be no moisture left in the mixture. When the mixture was tipped out, it was found that the underside was still very wet and had not crystallized at all. It is possible this is because the top surface crystallized so quickly preventing evaporation from the lower levels. It is possible that the mixture was not left for long enough for all the water to evaporate. It is also possible that the mixture initially contained too much water. The reaction of the top surface in this experiment is encouraging as it was hard and fast. The procedure however needs refining.

The dried solid has two very distinct finishes depending on whether the surface was open to the air or not. The surface which formed contained by the tray has a very smooth finish. The surface formed open to the air shows free and unobstructed crystal growth so is very uneven and rough.

04 _ SOLIDIFIED SAND - CLOSED SIDE

A super saturated solution of epsom salt in water was poured over a layer of sand 5mm thick. The solution was poured until it just covered the sand. The mixture was put in an oven at 40 degrees to try to simulate desert evaporation conditions. After 4 hours of drying the surface had dried very little and there was almost no crystalisation.

05 _ SOLIDIFIED SAND - OPEN SIDE

After the surface fractured on removal from the container I selected the most complete fragment to test. This measured 140mm long and 5mm thick. It exhibited good crystallization throughout and is indicative of the surface as a whole.

06 _ SHARD MEASURING 140MM

After a further 52 hours of drying the mixture finally crystallized totally. The top surface shows an uneven crystallization with much finer sheets of crystals than shown in the sodium chloride. The crystals on the top surface flake easily.

07 _ SHARD MEASURING 5MM

A rough test was conducted to test the strength of the fragment. The piece was placed across two blocks, and two more blocks were placed on top of it. It carried this load despite being in a very weak shear position. A third block caused failure. Each block weighs 170g.

08 _ STRENGTH TESTING

On removal from the tray the surface cracked and broke. This is beacause, despite being a non stick surface, some twisting and distorting was required to remove it. The remaining fragments were completely dry and crystallized throughout.

09 _ STRENGTH TESTING


01 EXPERIMENTATION

SANGUIS ET PULVIS

23 _ SAND SOLIDIFICATION Sodium thiosulfate proved a highly suitable salt for this application through its crystallisation tests. Testing on sand will reveal if it acts as an appropriate binder.

01 _ SOLIDIFICATION OF SAND WITH SODIUM THIOSULFATE 01 _ SPECKLED APPLICATION

02 _ EVEN APPLICATION WITH ATOMISER

05 _ STRENGTH TESTING

06

For the test a super saturated solution was used, heated as previously. The sand is a fine dry sand which has been sieved. A 5mm layer of sand is spread in the baking trays before being sprayed using an atomiser.

The first attempt to spray the sodium thiosulfate solution produced an unexpected result. The solution was allowed to cool to room temperature before it was sprayed, this meant that it was too unstable and like the sodium acetate crystallised too quickly. This quickly clogged the atomiser, and altered what should have been a fine mist into a powerful jet. On striking the sand this had the effect of causing it to ball and cluster. These clusters were however touching so the crystallisaton process bound them together. This gave a very delicate and weak, but visually complex and interesting, shell of sand. The perforated structure hints towards might be possible with more adventurous production techniques, however this particular strategy is very uncontrollable and unreplicable. Once the basic process of solidification is understood however further experiments will be conducted exploring these concepts.

The solidified sand was very stable and did not break or crack whilst being removed from the tin. Both the internal and external faces had a similar appearance of roughly textured sand, unlike the smooth finish found previously. Whilst dense and strong it had an immediately brittle feel. 03 _ SOLIDIFIED SAND FACE

In experiment 2 the solution was sprayed whilst still hot and this prevented immediate crystallisation. The solution was sprayed over a 5mm thick layer of sand and left to dry at room temperature. The hot solution evaporated very quickly accelerating drying and crystallisation time and the panel appear dry after only 20 minutes. By comparison the previous successful solidification required several days to dry and crystallise.

04 _ SOLIDIFIED SAND EDGE

A rough test was performed to examine the strength of the solidified sand sample. The sample was placed across two blocks leaving an open gap of 15cm. A large tub of wood filler was placed on the sample and it was strong enough to take this load. A large book was then placed on top of the filler, and this proved too much for the sample. The sample sheet was very thin at 5mm and due to the early nature of the experiment did not undergo as much crystallisation as hoped. This is apparent in the lack of a white surface on the sample. It was however fairly strong. The nature of the material is that the sand grains, which are close packed in compression, are held together by the crystals. As such when place in compression the load is passed through the sand grains whilst they are held in place by the salt. The salt meerly acts as a form work to retain the sand. However when placed in tension the sand grains are pulled apart and therefore all the stress is placed through the crystalline joints of the salt. These joints are comparatively weak and so break under little loading. As such this material will always be considerably stronger when placed in compression, and is therefore well suited to domed or arched construction.

07


01 EXPERIMENTATION

SANGUIS ET PULVIS

24 _ CRYSTALLIZATION Sodium thiosulfate has so far proved a suitable salt for this process. It has been tested on small flat surfaces but must now be tested on larger and more complex geometries.

01 _ FORMATION OF SUPER SATURATED SOLUTION

02 _ POURING AND SOLIDIFYING

03 _ REMOVING THE SHELL

01 _ LAB SETUP

01 _ SAND CONE IS POURED

02 _ AFTER SPRAYING THE SURFACE SOLIDIFIES

03 _ A HOLE IS MADE TO REMOVE INTERNAL SAND

04 _ MOST INTERNAL SAND IS REMOVED AND THE STRUCTURE STANDS

The first test involves attempting to solidify a single cone, the most basic sand geometry. Important factors and the degree to which the spray disturbs the surface of the sand, the speed of crystallisation and the strength of the produced shell. Once these factors have been established and tested practically more complex geometries and systems can be experimented with.

02 _ SOLID CRYSTAL MASS

03 _ DISSOLVING CRYSTALS

The apparatus required for the process are an electric hob, a variety of lab quality glass beakers, a stirring rod, a spraying box and an atomiser (preferably several). All surfaces should be protected as once the salt crystallises it is hard to remove. Absorbent paper towel works well as the crystals form on the towel rather than the wood. Safety goggles and protective gloves can be worn but are not necessary. It is however unadvisable to contact the salt crystals with the eyes. The solution is sprayed at near boiling temperature to prevent premature crystallisation, as such gloves may be required for the spraying procedure. Having sprayed the solution the atomiser should be sprayed with hot tap water to clear the insides of solution and avoid crystallisation and blocking. If this is not done the atomiser can only be used once. Even doing this each atomiser only lasts around 4 or 5 sprays.

01 _ The shell is carefully lifted trying to avoid breaks


01 EXPERIMENTATION

SANGUIS ET PULVIS

25 _ THIN SAND SHELL The final thin shell produced is very fragile and only around 3mm thick. Although built up over three layers of spraying and adding sand this has not added enough thickness to make the shell strong. The finish is of a patchy white colour on the external surface where the salt crystals are visible. The internal finish is a very rough textured sand which appears bobbled. The shell is strong enough to support its own weight but the feet are very weak and tend to crack.


01 EXPERIMENTATION

SANGUIS ET PULVIS

26 _ DUNE SHELL In order to create the solidified sand shell that forms the structure of the pavilion it is first necessary to pour and artificial dune. This is done using a bottle with a fine opening in one end. The dune can be sculpted to a desired form by moving the pour point during the pour.

01 _ SOLIDIFICATION OF THE DUNE

03 _ REMOVING THE SHELL

03 _ THIN SAND SHELL 2

01 _ IMMIDIATELY AFTER SPRAYING

01 _ The shell is carefully lifted trying to avoid breaks

01 _ The shell is strong and stable but remains brittle

The solidified sand shell is formed by 5 layers of sand with 4 sprays of salt solution between each sand application. The sand dune geometry is now much larger than previous experiments, around three times the size of experiment 1, and as such requires more layering to add strength. This will be a more true test of the strength of the material created. The shell formed in this model was significantly stronger than the earlier experiments but it shared some of the same issues. It was heavy and solid, could easily carry its own weight, and whilst not tested to breaking point feels as though it is highly stable. It is however still brittle, an issue shared with the first thin shell created. Two methods could be adopted to avoid this brittle characteristic: use more salt solution to achieve more crystallisation and stronger joining, or use additives as in reinforced concrete to give tensional strength. These possibilities will be explored in the next experiments.

02 _ AFTER TWO MINUTES


01 EXPERIMENTATION

SANGUIS ET PULVIS

27 _ DRAINING THE DUNE INTERIOR

01 _ DRAINING OF THE INTERIOR VOLUME

02 _ SPATIAL CONDITION CREATED IS BASED ON THE ANGLE OF REPOSE

01

04

02

05

03

06

Once the outer surface of the dune has been solidified, as in the previous model, the interior volume can be drained to create the inhabitable space of the pavilion. In this model this has been achieved by drilling holes in the base of the container, although in full scale this would need to operate differently. As the sand pours through the holes it forms to its natural angle of repose, creating bucket like spaces within the shell. The size of these spaces is related to the spacing and diameter of the drainage holes. Once the interior space has been drained it can be stabilized with salt solution to create a habitable surface. Holes can be created in the upper shell surface to allow light to penetrate to the volume below, creating a high contrast lighting atmosphere. Due to the thermal properties of the sand shell the space beneath will remain extremely cool despite the hot external temperature of the desert. It will also remain relatively warm through the cold desert nights.


01 EXPERIMENTATION

SANGUIS ET PULVIS

28 _ DUNE INTERIOR SPATIALITY


01 EXPERIMENTATION

SANGUIS ET PULVIS

29 _ DRIP FORMING

01 _ DRIP FORMING OF SAND SOLUTION ON PVC FORMER

02 _ DRIP FORMED SAND DOME

01

04

02

05

03

06

The techniques of dune solidification explored thus far have been uniform and uncontrolled. This has resulted in the creation of even and uniform shells with little variation over their surface. The interior and exterior effect of this is monotonous and singular, creating cave like interior spaces. Light penetrates from perforations only and spreads within the dark space of the shell. Whilst this is desirable in some conditions and where this spatial atmosphere is desired, it is also necessary to develop alternative techniques that provide a rich and varied atmosphere. To achieve this a structure must be created that is not of uniform thickness or application. Light must be allowed to penetrate through the structure itself rather than only the openings. This will then provide an interior condition of variance. The first experiment in achieving this explored drip forming. In this process the saturated salt solution is pre-mixed with sand to create a viscous liquid. This is then dripped on to a form work in a random manner until the desired coverage has been achieved. The sand is then allowed to set through the crystallisation of the salt. As the salt is a hydrous crystal it absorbs water as it crystallises, this gives a very quick drying time of around 30 minutes. Once dried the structure can be removed from the former. The resulting structure is suprisingly strong but has a feel of inherent fragility due to its multiple perforations. Due to the process of forming some areas of the sand are very thin, making them translucent. This gives a highly variable light condition in the interior space.


01 EXPERIMENTATION

SANGUIS ET PULVIS

30 _ DRIP FORMED SAND DOME INTERIOR ATMOSPHERE


01 EXPERIMENTATION

SANGUIS ET PULVIS

31 _ CREST POURING

01 _ CREST POURING AND SAND REMOVAL

02 _ CREST POURED SAND STRUCTURE

01

04

02

05

03

06

The second experiment of crest pouring aims to achieve similar goals to that of drip forming. The goal is to create a nonuniform sand structure with a dynamic spatial affect. Crest pouring uses the same viscous liquid of sand and salt solution as drip forming. In this process a sand dune is used as the former however. The mixture is poured along the crest of the dune, building a reservoir along this line. As this reservoir overflows the mixture cascades down the side of the dune until they meet the ground plane. Once dried these arms form the structural elements which support the crest line. Small secondary pours are made at lower levels to link these arms and ensure the structure is laterally stable. Once the structure has dried and become solid, which takes around 30 minutes, the sand can be blown away using an air gun. This is an accelerated simulation of the way a real sand dune would migrate away from the solidified structure. The movement of sand reveals an increasingly large volume within the solidified structure. Some of the sand in unable to fully escape the structure on the leeward side which gives the space a specific directionality relative to the wind. The final structure mimics the overall geometry of the original dune, however its form is so different that they appear unrelated. The highly skeletal structure appears to have been dug up from within the sands of the desert, imposing a sense of timelessness. The structure is strong and stable but needs to be tested at a larger scale in order to be analysed correctly.


01 EXPERIMENTATION

SANGUIS ET PULVIS

32 _ CREST POURED SAND STRUCTURE EXTERIOR FORM


01 EXPERIMENTATION

SANGUIS ET PULVIS

33 _ CREST POURED SAND STRUCTURE EXTERIOR FORM


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ STAGES OF DUNE PAVILION CONSTRUCTION

34 _ PROPOSED CONSTRUCTION TECHNIQUE 01 START DIGGING BASE HOLE

02 BUILD PAVILION FLOOR

03 COMPLETE PAVILION FLOOR (CHECK HOLES)

04 START POURING DUNE

05 COMPLETE POURING DUNE

CONSTRUCTION DAY 1

CONSTRUCTION DAY 2

CONSTRUCTION DAY 2

CONSTRUCTION DAY 3

CONSTRUCTION DAY 3

The hole above which the pavilion will sit is dug first. This supplies the required volume of material for the construction of the pavilion’s artificial dune.

Once the hole has been dug a suspended floor will be constructed. This involves a grid of vertical supports which are driven into the sand and a network of horizontal beams above them. Lateral bracing is provides by sheets of mdf which form the top surface of the floor.

The complete floor surface will contain a pattern of holes. Once the dune is poured it will continuously pour through these holes back into the hole beneath. They are sized such that it will take several days for all the sand which can pass back to do so. A such the physical nature of the pavilion, in terms of its massing, volume and spatiality, will be constantly evolving through the course of the draining, reaching a point of equilibrium only at the end.

Once the floor surface is complete the dune can begin to be poured. This is done by the crane lifting the Bulk Bags of material and opening their bottom panel, allowing the material to flow out.

Once the dune has reached its designed scale and form the pouring is ceased. Some sand may need to be added as the sand settles and the geometry alters. As the sand pours through the hole beneath at such a slow rate it will not greatly affect the dune in this construction phase.

The following technique describes how a pavilion scale building may be constructed in the desert using the technique of crest pouring using super saturated salt solution. The desert in which this is proposed is the Black Rock Desert in Nevada, USA. This is a highly seasonal desert in which a playa lake is formed during winter. The presence of this evaporative lake creates a highly saline environment, and is therefore highly suitable as a source of salt for use in the construction. A pavilion scale structure could be constructed over a period of 3 days in the desert. This period would consist of four definable phases: digging, flooring, pouring, solidifying. Digging involves the removing of sufficient material for the construction of the artificial dune. Flooring involves the construction of a supported floor plane which upon which the pavilion will sit. This floor will be penetrated with carefully positioned and sized holes to allow the passage of material from the dune back to the hole beneath. In this way the pavilion will slowly drain over the course of several days creating a dynamic and evolving space that mimics the passage time in the desert.

The pavilion requires 480 m3 of material. 1 man day is equivalent to 20 m3 of material, meaning a team of 24 is required to dig the necessary material in 1 day. Sand will be placed into Bulk Bags with an openable bottom panel which can be lifted out by the crane. A mini crawler crane and forklift will be brought by the team to assist with the construction. These are necessary for the movement and pouring of the sand.

The load that will be applied to the floor will vary greatly across its area. At the edges there will be very little dead load, whilst at the centre where the dune is higher the load will be higher. A maximum dead load of 3.2 tonnes / m2 will be reached whilst the average load will be around 1.1 tonnes / m2. This load will universally reduce as the festival goes on and sand drains through to the hole beneath. The structural members and surface of the floor will need to be analysed and specified accordingly.

The crane must follow the crest line of the designed dune in order to create the correct geometry. The outline of the floor can be used to judge how the geometry is growing and sand added where it is required.


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ STAGES OF DUNE PAVILION CONSTRUCTION

35 _ PROPOSED CONSTRUCTION TECHNIQUE 06 START POURING SALT SOLUTION

07 COMPLETE POURING SALT SOLUTION

08 VOLUME DRAINAGE STAGE 1

09 VOLUME DRAINAGE STAGE 2

10 VOLUME DRAINAGE STAGE 3

CONSTRUCTION DAY 3

COMPLETION

DRAINING DAY 1

DRAINING DAY 4

DRAINING DAY 7

In order to create the rigid structure of the dune a mixture of super saturated sodium salt solution and sand must be poured over the crest of the dune. This mixture will fall down the sides of the dune to the base.

The sodium salt used forms an aqueous crystal, meaning that 5 molecules of water are chemically bound within each crystal molecule. This means that where most solutions rely on drying to crystallise fully, this does not. As crystallisation commences water is absorbed into the crystals, drying it from within. This gives a very fast solidification time of only several hours at this scale.

At the beginning of the festival the pavilion will be an almost solid dune. Very little interior space will exist. This will promote a very different kind of interaction with those who experience it. They will be able to climb the dune and scale the structure. The can nestle into the small voids that begin to emerge beneath the structural elements.

After a day the sand will have half passed through back to its original hole. At this point the outlines of drainage patterns will be emerging in the dune, an indicator to the process that is occurring. The spaces within the structure will have become more expansive and occupiable. The structure can still be scaled, with the soft bed of sand beneath to fall into. The light condition will vary towards the edge of the pavilion. The central area will be most shaded beneath the more dense elements along the dune’s crest. The outer areas will be more exposed and hot during the day.

After several days the full drainage pattern will have been achieved and the pavilion for the first time will become static. At this point the drainage pattern will have created pockets of space, occupiable buckets partially separated from each other. The variation in the location of the drainage holes creates a great variety of spaces, some barely having change through the course of the drainage and others having dropped several meters. At this point the crest of the structure will be high above the floor of the pavilion and less open to climb. The lower elements will however still be highly accessible.

The final phases of the pavilion construction are pouring and solidifying. Pouring involves the formation of an artificial dune that will form the basis of the pavilion. The material which was removed in the digging phases is poured back on to the site in a specific manner to construct a dune as designed. This dune features a long curved crest which rises to one end. Solidifying involves the creation of a rigid structure over the poured dune using a mixture of sand and super saturated sodium salt solution. The super saturated solution can be created by using the desert sun to heat a black metal container. Once the solution reaches around 50 degrees it is able to become super saturated. This is then mixed with sand and poured over the crest of the dune. Once crystallised, the mixture becomes a stable solid and forms the load bearing section of the pavilion.

The first pour should stick to the ridge of the dune, forming the main structural elements. Additional pours can be made where necessary to create additional support. Whilst the pour elements will naturally branch and create cross bracing between themselves, further pours can be made to add diagonal bracing where necessary.

The surface texture and colour can be used as indicators for the stage of the crystallisation process. Once the crystals have fully formed and dried the surface will become white and roughly textured.


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ STAGES OF DUNE PAVILION CONSTRUCTION

36 _ PROPOSED CONSTRUCTION TECHNIQUE 11 REMOVE REMAINING SAND

12 REMOVE FLOOR STRUCTURE

13 STRUCTURE LEFT BEHIND

14 PLAYA LAKE BED FLOODS

15 STRUCTURE DISSOLVED

DECONSTRUCTION DAY 1

DECONSTRUCTION DAY 2

DECONSTRUCTION DAY 3

AUTUMN

LATE AUTUMN

The remaining sand can be swept through the holes in the floor to reveal the bare floor construction. At this point the hole beneath will nearly be filled back to its original volume.

Once the remaining sand has been cleared the floor structure can be removed. The floor panels can be removed first, followed by the joists. Care must be taken during this phases not to critically damage the structure by colliding machinery with it.

At this stage, only the salt solidified sand structure will remain. A small recess will be left beneath it due to the material held in the structure not having been returned to ground yet.

In Autumn the playa floods due to the influx of rain water from the surrounding mountains. In a very wet season the lake reaches around a meter deep, in dry seasons only a few hundred mm.

After the collapse of the structure the material will become totally reabsorbed to the landscape. The currents of the playa lake will redistribute the material evenly leaving no trace of the pavilion when the lake dries up the following spring.

The deconstruction phases of the pavilion would occur over a long time period extending into the desert winter. Initially the floor structure of the pavilion can removed, leaving only the salt solidified sand structure. This is how the pavilion will be left. Once the seasonal playa lake forms in Autumn the water will be absorbed into the structure, destabilising the bonds as the salt crystals dissolve. The structure will eventually collapse and be absorbed into the playa floor.

The structure will soak up this water, dissolving the crystal bonds which stabilise it. This will critically weaken the structure leading to its collapse.


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ CALIBRATION OF CAMERA, LASER AND BACKGROUND

02 _ LIVE 3D SCANNING OF MODEL AS LASER SWEEPS

37 _ BUILDING A 3D SCANNER

03 _ RESULTANT MESHES 01 _ OUTER SURFACE

The nature of the construction process and material explored means that reproducing realistic 3d models of it is very challenging. The process of pouring carves tiny channels in the sand and builds small ridges at the edges of the flows. The paths that the pours take are complex and meandering, not straight and uniform. All of these things are extremely difficult to replicate in CAD. As such the most appropriate method for quickly digitising the design is through 3D scanning. Whilst commercial 3D scanners are extremely expensive it is possible to construct one using a background, web cam and a laser line. A stand alone application called David Scanner is able to convert the input of these into an accurate and detailed 3D mesh which can be used for further development and rendering.

02 _ INNER SURFACE

The 3d scanning system uses a web cam to analyse the displacement of a laser line as it passes over the surface of the model. The camera and laser must first be calibrated. This is done with an empty corner. The dots of the background sheets allow the camera to calculate its position and scale relative to the background and the intersection angle with the laser. Once this has been done the model can be placed in the corner. It is vital that once the calibration has been completed the camera and background corner do not move at all. It is important that the frame rate of the camera is sufficient to capture the motion of the laser. A very slow laser motion can be achieved by using a mechanical rig, however when passing by hand this is not possible. As such the camera is not used on its highest resolution, but is rather set to a low resolution that has a higher frame rate. The resolution of the scan is directly related to the resolution of the camera so this reduces the quality of the scan but is the best set up for hand passed lasers. The lights must be turned off and the camera setup so that only the laser line is visible to achieve a clean scan.

Once the scan starts the laser must be slowly and evenly passed over the surface of the model. In order to achieve the smoothest possible pass the laser has been attached to a tripod which can provide a smooth motion compared to the bare hand. As the laser passes over the model the camera is able to read the distortion of the line and transfer this into 3d geometry. The slower and more evenly the laser is passed the more accurate and clean the produced mesh will be. Multiple passes are not necessary and can damage the mesh by producing overlying anomalies. As the pass is being made the depth field of the scan is shown in real time on the computer monitor. As such if any areas have been missed or appear inaccurate due to a jump of the laser or and inconsistency in the pass speed the pass can be aborted and the scan restarted. In order to scan a full model it is necessary to perform several scans from different angles. The software contains an automated stitching application which is able to align and weld the multiple meshes produced into a single solid mesh. Due to the complexity of the model being scanned this did not work and the meshes had to be combined by hand in Rhino.





01 EXPERIMENTATION

SANGUIS ET PULVIS

41 _ WASTE BLOOD AS A BINDER 01 _ EQUIPMENT LAYOUT WHY BLOOD? The previously explored techniques of sand solidification have relied on sodium thiosulphate salt as a binder. This has several limitations as a binding material. Firstly, it is a salt which is only readily available in very few desert environments. This means that it is not widely applicable as a binding material. Secondly, it is water soluble, meaning that it is only suitable for temporary structures. Due to this, for the continued research of sand solidification as a building technique it is necessary to explore a new binding material. This material should be readily available in desert environments and should be water resistant, allowing the creation of more permanent structures. The first material I will explore in line with these requirements is animal blood. Animal blood is the readily available waste product of halal slaughter houses, the practice of slaughter adhered to by desert communities throughout the Sahara. In the recent past, animal blood was widely used as the basis of natural glues known as albumin glues. These utilised the protein albumin, which is abundant in blood, to create a powerful water resistant glue. This glue was dominant in the production of plywood and other manufactured timber products before the widespread adoption of synthetic glues. As such, blood based binders form a highly suitable starting point in the search for a suitable binder.

CREATING BLOOD GLUE The main raw material of the glue in fresh bovine blood. This was collected directly from slaughtered bullocks as they were hung in the abattoir. 40 litres of blood were collected from 4 bullocks, as the collection process involved a lot of wastage. In a facility specifically designed for blood collection it is estimated that 30 litres could be collected from each bullock.

The second vital additive is a preservative. Preservatives stop the blood from decomposing and prevents bacteria growing in the blood, and in the resultant glue. The preservative added in this process is sodium azide, and was added in the amount of 5g / litre of blood.

MIXING BUCKETS

TARPAULIN

When the blood is collected it is essential that an anticoagulant is added immediately. Without this the blood will turn to a jelly like substance within 5 minutes. The anticoagulant used in this process is EDTA, and was added in amount 10% w/w of the resultant volume.

FRESH COW’S BLOOD

KILN DRIED SAND

GLOVES

2L BEAKER

SODIUM AZIDE

EDTA

ACETIC ACID

SIDE BOARDS

BRICK FORMER

02 _ BLOOD GLUE PREPARATION PROCESS 01

02

03

04

05

06

FRESH BOVINE BLOOD

BLOOD VOLUME CALCULATION

PRESERVATION USING EDTA ANTICOAGULATION USING SODIUM AZIDE

MEASURING ACETIC ACID

MEASURING BLOOD MIX

ADDING ACETIC ACID AND MIXING

Once these additives have been mixed in thoroughly the blood is stable and can be left until the glue is required to be made. To make the glue a monobasic acid must be added to the mixture. The monobasic acid added in this process is glacial acetic acid. This acid reacts with the albumin protein in the blood and creates a glue. According to US Patent 4180412 this acid should be added in the amount of 50ml / litre of blood. However, it is stated that there is a minimum level of acid which is required. Once this amount is reached the maximum glue strength and drying time is reached by adding more acid up to a limit where it becomes weakened. As such to explore these limits 4 test batches were made using varying amounts of acid. By testing 4 test batches it can be calculated which has the most desirable properties and this formula can be subsequently used for the following models.

DIVERTING COMPONENTS


01 EXPERIMENTATION

SANGUIS ET PULVIS

42 _ FRESH BLOOD GLUE TESTING

01

02

03

04

01 _ LIQUID GLUE POURING

01 _ LIQUID GLUE POURING

01 _ LIQUID GLUE POURING

01 _ LIQUID GLUE POURING

Test mixture 01 used the minimum quantity of acid specified in US Patent 4180412. This involved adding 10ml acid to 250 ml blood. This mixture was stirred well for 20 seconds and then poured over the sand former and left to dry. The mixture did not thicken noticeably during the mixing time and was very runny when poured. It retained a deep red colour. Test mixture 02 used a large quantity of acid. This involved adding 25ml acid to 250 ml blood. This mixture was stirred well for 20 seconds and then poured over the sand former and left to dry. The mixture thickened to the point that it was no longer liquid during the mixing time and had to be scraped out of the beaker. It adopted a very dark colour nearing black. Test mixture 03 and 04 sit between 01 and 02. For test mixture 03 this involved adding 15ml acid to 250 ml blood. This mixture was stirred well for 20 seconds and then poured over the sand former and left to dry. The mixture thickened a little during the mixing time and ran more slowly down the surface of the sand. It retained a very deep red colour. Test mixture 04 used a little more acid than 03. This involved adding 20ml acid to 250 ml blood. This mixture was stirred well for 20 seconds and then poured over the sand former and left to dry. The mixture thickened to the texture of honey during the mixing time and was viscous when poured. It retained a tone of red but was very dark, approaching black.

02 _ DRIED GLUE SURFACE

250mm

ANTICOAGULATED BLOOD

250mm

ANTICOAGULATED BLOOD

250mm

ANTICOAGULATED BLOOD

250mm

ANTICOAGULATED BLOOD

10ml

GLACIAL ACETIC ACID

10ml

GLACIAL ACETIC ACID

10ml

GLACIAL ACETIC ACID

10ml

GLACIAL ACETIC ACID

20 sec

MIXING TIME

20 sec

MIXING TIME

20 sec

MIXING TIME

20 sec

MIXING TIME

6 hours

DRYING TIME

6 hours

DRYING TIME

6 hours

DRYING TIME

6 hours

DRYING TIME

02 _ DRIED GLUE SURFACE

02 _ DRIED GLUE SURFACE

02 _ DRIED GLUE SURFACE


01 EXPERIMENTATION

SANGUIS ET PULVIS

43 _ FRESH BLOOD GLUE TESTING The test pours were left for 6 hours to dry before the sand former was removed from beneath them.

01 _ REMOVING SAND FORMER

02 _ TEST POUR 01 SURFACE

TEST POUR 01

01 _ EXTERNAL SURFACE

Test pour 01 appeared to have dried well, however when the sand was removed from beneath it, it was apparent that only the very top surface had set and the lower mass of the glue remained very soft. As such the structure quickly sagged and collapsed.

Test pour 02 was not able to be excavated due to its nature. Test pour 03 appeared to have dried well. The top surface had blistered slightly in the sun and small edges had begun to curl upwards slightly. However when the sand was removed from beneath it, it was apparent that only the very top surface had set and the lower mass of the glue remained very soft. As such the structure quickly sagged and collapsed. Test pour 04 appeared to have dried well. The top surface was very blistered from the sun and its edges were curled upwards a lot. However when the sand was removed from beneath it, it was apparent that only the very top surface had set and the lower mass of the glue remained very soft. As such the structure quickly sagged and collapsed.

TEST POUR 03

CONCLUSION The conclusion that can be drawn from these tests is that the glue produced by this process dries by water loss rather than by chemical reaction. Due to the nature of the structural process it is necessary for the glue to dry by chemical reaction, as a very thick body of glue is required to set strongly all the way through. This can never be achieved with a glue that dries by water loss as once the outer surface sets it prevents any water loss occurring from the centre. A new glue formula must be tested which will create a chemically setting glue. TEST POUR 04

02 _ INTERNAL SURFACE


01 EXPERIMENTATION

SANGUIS ET PULVIS

44 _ BLOOD GLUE BRICK TESTING The glue mixture used in test pour 04 dried the most throughout the mixture. As such this glue was recreated to test its drying when mixed with sand and cast as a brick.

01 _ MIXING AND FORMING TEST BRICK

02 _ EXAMINING THE DRIED BRICK

01 _ SAND ADDED 400% BY WEIGHT TO GLUE

01 _ BRICK SURFACE WHEN CAST

02 _ BRICK SURFACE WHEN DRIED

03 _ DRIED BRICK

To make the glue 20ml acid was added to 250ml blood, and was stirred well for 10 seconds. 1 kg of sand was added to the glue and mixed through thoroughly. This created a very thick aggregate. This aggregate was scraped into a former created in the sand and left to dry for 6 hours.

When the brick was uncovered it appeared much the same as the test pours attempted previously. The outer surface was of a very dark red colour and was hard to touch. However when the brick was lifted from the former it was clear that it was not strong and solid as the outer surface appeared. The brick was broken in two and the centre revealed the same issue as was found with the test pours. Only the very thin outer surface of the brick had set, creating a water tight skin around the brick. This prevented an water loss occuring from the middle of the brick and as such no setting occured. 02 _ MIXTURE PLACED INTO SAND FORMER

CONCLUSION The conclusion that can reached from this test brick is the same as was found with the test pours. The glue produced by this process dries by water loss rather than by chemical reaction. Due to the nature of the structural process it is necessary for the glue to dry by chemical reaction, as a very thick body of glue is required to set strongly all the way through. This can never be achieved with a glue that dries by water loss as once the outer surface sets it prevents any water loss occurring from the centre. A new glue formula must be tested which will create a chemically setting glue.

04 _ DRIED BRICK SECTION


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ BLOOD GLUE CREATION

45 _ BLOOD BRICK MAKING PHASE 2 01 WEIGH 100 PARTS DRIED BLOOD AND 140 - 180 PARTS WATER

02 MIX AND LEAVE TO SIT FOR 2 HOURS

03 FULLY MIX BREAKING DOWN LUMPS

04 ADD 5 1/2 PARTS AMMONIUM HYDROXIDE AND 8 PARTS PARAFORMALDEHYDE TO PRODUCE FINAL GLUE

Blood brick making phase 2 is based on US Patent 1,459,541. The issue experienced in phase 1 was that the glue created dried by water loss, and as such was unable to solidify beneath the surface. The glue described in the above patent is able to set under water, proving that it sets by a chemical reaction and not water loss. This will allow the centre of very thick pours to set. The process of obtaining and treating fresh blood, as used previously, was very time consuming and problematic. As such for the next series of experiments I will use dried blood, which acts in the same way once mixed with water to create a solution. The active element of the blood in glue manufacture is the albumin proteins. When blood is dried, the solid elements left behind include this protein. Accordingly, a dried blood solution behaves in the same way as fresh blood. Another advantage is that it is easier to control in an experimental environment. When using fresh blood, one has to guess as to the solid content and adjust the other additions accordingly. When dealing with dried blood however, only the solid content is left, making accurate repetition and variation possible. A safe experimental procedure is essential to follow the above patent as it involves the chemical paraformaldehyde. This is a corrosive, toxic, flammable, carcinogenic and blinding chemical. It is absolutely essential that it be worked with in isolated conditions, in this case within a fume cupboard. Once the paraformaldehyde reacts with the proteins in the blood it is chemically altered and is no longer dangerous in any way. As such the resulting glue is safe. In fact, this preparation of glue was used on an industrial scale in the production of various manufactured wood boards for over 50 years, proving its ultimate safety.

CONCLUSION: FAIL All quantities used in this glue formula must be measured carefully by weight. In order to test a variety of glues, the quantity of water was varied. The effect of this is to create glues of differing consistencies and adhesive properties. Three different preparations wer made, using 140, 160 and 180 parts water. The blood used in this experiment is dried pigs blood, usually used for black pudding manufacture. The advantage of using dried blood is that it can be stored and purchased easily. Dried blood is created by evaporating fresh blood to leave the solid elements. This must be done at a fairly cool temperature to prevent rendering the proteins insoluble. Once mixed with water its properties should be very similar to fresh blood. The water added to the blood must be room temperature. Cool water will not create a solution easily, and hot water will alter the proteins that are vital to the glue.

The blood and water can be mixed gently initially however the blood will not go into solution readily. Instead it forms and dense, sticky paste. Some mixing can occur, but then the mixture must be left to rest for 2 hours. Over this period the blood softens and becomes easier to mix.

After resting for two hours the blood can be easily stirred into the water to form a dark red / brown mixture. The consistency of this solution varies depending on water content. The batch with the least water has a consistency like thick glue, and the most is closer to single cream. All lumps should be stirred out at this stage.

The final stage of the glue creation is the addition of ammonium hydroxide and paraformaldehyde. Paraformaldehyde acts as a bio-connector and bonds the proteins in the blood together to form a glue. Paraformaldehyde is extremely dangerous (flammable, toxic, corrosive, carcinogenic, blinding) and as such this stage must be carried out in a fume cupboard with great care.

This process failed to create a workable blood glue. The patent formula followed stated that after adding the formaldehyde the mixture would thicken to a jelly like mass, but that after resting for an hour would return to a workable consistency. However, once thickened the glue did not return to a workable consistency. It remained as a solid mass, only getting harder with time.

Once mixed the glue must be left in the fume cupboard for several hours to allow any unreacted paraformaldehyde to vaporise. No risks can be taken with this.

I conclude that this is due to the type of paraformaldehyde used. The patent specifies slow reacting type paraformaldehyde, however the chemical is no longer categorised like this and is available only in one form. I believe that the type used was in fact fast reaction, and that the firming and releasing process described occurred in a matter of minutes rather than hours. This means this glue can never be used, as it must be left to vaporise for at least an hour. Also, due to the dangerous nature of paraformaldehyde, it is not suitable for use in a remote desert environment. Another process must be used.


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ COAGULATED BLOOD BRICKS

46 _ BLOOD BRICK MAKING PHASE 3 01 MIXING BLOOD / WATER SOLUTION WITH SAND (50% w/w)

02 THE BRICK MOULDS IN A DRYING OVEN AT 70 oC

03 THE BRICK MOULDS FULLY SET AND REMOVED FROM OVEN

04 ATTEMPTING TO RELEASE THE BRICK

After the failure of the formula used in phase 2, a different direction was required. Firstly, the use of exotic and dangerous chemicals to create a glue makes little sense in a remote and poor society. Secondly, the attempts to manufacture glue from blood have proved largely unsuccessful and there is no current body of knowledge to consult on these processes. I was aware that blood and water on their own form a glue of significant adhesion, but how could a mixture of just blood and water be made to set fully throughout its depth? When blood and water set to form a glue, what occurs is the coagulation of the blood due to loss of water. Once the blood has become coagulated it forms an insoluble solid which binds strongly. I had found this from the dried blood found on the glass containers used in previous experiments. This is not the only way in which blood will coagulate though. It will also do so if heated over 64 oC. This means that once placed into moulds the mixture could be placed in an oven and heated to this temperature, resulting in the setting of the blood throughout its depth. This allows for the creating of a blood brick without the addition of any chemicals at all. Whilst this may seem like a return to traditional brick making techniques, the temperature required is very very low by comparison. This temperature could be achieved by either using solar ovens, or laying the bricks in the desert sun.

CONCLUSION: PARTIAL SUCCESS The mixture of blood and water is mixed with sharp sand at a ratio of 50% w/w. The resulting mixture is very thick and granular. Two batches of mixture were produced using different quantities of water. As before this would allow the comparison of the two bricks for optimisation of the process.

The mixtures are placed into steel brick moulds which have been preheated in a drying oven to 70 oC. This is the temperature required to coagulate or clot blood. When blood clots it essentially acts as a glue, bonding strongly to anything it is in contact with and becoming insoluble in water. In this case, it will bond to the sand creating a solid brick. The moulds are left in the drying oven for 3 hours to allow the full depth of the brick to reach sufficient heat. The process of heating to this temperature is very sustainable in the desert, as solar ovens can easily reach this temperature even on cloudy days. On hot days this temperature would even be reached by simply leaving in the sun.

During the heating process the mixture swells slightly, giving the brick a domed top surface. The colour of the brick also becomes significantly brighter, and more red. The moulds are left to cool for 2 hours before they are opened.

Before pouring the blood mixture into the moulds, each one was coated with a thin layer of oil. This oil is used when casting concrete bricks to prevent adhesion to the mould wall. Unfortunately, we did not anticipate that the temperature of the preheated moulds would burn off the oil before the mixture could be poured. As a result there was in fact no release agent left to prevent adhesion to the mould. The blood mixture stuck to the mould casing with incredible strength, at least proving the concept of the process. However, it also made removal impossible. The strength of the created brick was such that it took over an hour to remove it from the mould casing.

Whilst not a complete success, this experiment at least proved the concept of the process. The experiment was let down by a procedural error, which with revision should produce bricks over high quality and strength. Due to this error however, it was not possible to test the strength of the bricks. The brick produced was extremely hard on its outer surface, and became weaker towards the centre. This is most likely due to water content at the centre which is unable to escape. As such, for further experiments a blood mixture with minimum water content will be produced. The brick produced was very attractive, having a rusty red colour. The colour also varied beautifully through the depth of the brick.


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ COAGULATED BLOOD BRICKS

47 _ BLOOD BRICK MAKING PHASE 4 01 FINAL BRICKS IN THE DRYING OVEN The final blood brick making phase uses the same experimental procedure as phase 3. A mixture of blood and water is added to sand in a ration of 50% w/w. This mixture is then poured into cling film lined mould casings, to allow for easy removal, and left to solidify in a drying oven.

The previous process was repeated but using cling film to line the mould casings. This would allow for a clean release of the bricks after solidifying. In order to allow the drying of the maximum amount of water, the bricks were left in the drying oven for 1 hour to solidify sufficiently to remove the mould casing. They were then returned to the oven without the casing to allow maximum drying and thus maximum strength. The bricks produced were uniform in colour and texture. They were extremely hard on all faces and felt strong under compression, easily holding my weight without denting. The only irregularity is domed top formed by expansion in the mould. Whilst this could be eradicated by compression during drying, I feel it adds a desirable texture to the brick and differentiates it. The bricks would be layed with this face as the external surface, giving the building an uneven and textured surface.

02 FINAL BRICK SHOWING DOMED SURFACE

03 STACKED BRICKS WITH ROUGH EXTERNAL SURFACE


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ COAGULATED BLOOD BRICKS

48 _ BLOOD BRICK TESTING 01 FINAL BRICKS IN THE DRYING OVEN

02 FINAL BRICK SHOWING DOMED SURFACE

In order to gain some quantitative data on the brick material created, compression testing was conducted. Three different bricks were made using different ratios of blood and water to allow a comparison to be made.

SPECIMEN 1: 100 parts blood, 180 parts water, 50% w/w sand MAX LOAD: 4333 N

The brick was placed between the steel compression plates of the machine, before a small pre-load is applied to make sure that the compression surfaces are flat and even. The machine then gradually increases the load applied to the brick, measuring the resistance provided by the brick.

SPECIMEN 2: 100 parts blood, 160 parts water, 50% w/w sand MAX LOAD: 4895 N

SPECIMEN 3: 100 parts blood, 140 parts water, 50% w/w sand MAX LOAD: 5348 N

The graph shows that whilst all of the bricks behaved in a similar manner, specimen 3 withstood the greatest load. Specimen 2 behaved in a very similar way, however specimen 1 was significantly weaker. These results corollate to the water content of the blood mixture used in the brick. According to this limited testing, it appears that the lower the water content of the brick the greater its compressive strength. Page 2 of 2

CONCLUSION It is important to note that the testing conducted here is very limited and therefore is not decisive. It does however suggest that by lowering the water content of the brick mixture a stronger brick is produced. I would predict that this is because when the brick dries some moisture is trapped within the brick making it weaker. If there is less water present in the original mix, then there will be less water present in the solidified brick therefore making it stronger. There could however be other factors effecting this, and much more extensive testing is required with more variables. 18 May 2012 The main issue of using such a test with this material is that the machine used is designed for testing materials which are extremely rigid and brittle, such as concrete or standard brick. As such, when the machine detects any significant deflection in the material it halts the test. The blood bricks created so far have a very hard external surface however there interior is more like a hard sponge. This means that they squash under loading, which the machine interprets as failure. As such none of the bricks could be tested to their true point of failure. As such without a revised testing procedure these results can not be reliably compared to those of standard brick or concrete.


01 EXPERIMENTATION

SANGUIS ET PULVIS

01 _ BLOOD CREST POURING

49 _ CREST POURING TESTING 01 THE SMALL TEST DUNE COVERED WITH DIVERTING COMPONENTS

02 CREST POURING OF BLOOD BINDER

03 COMPLETED CREST POUR

04 BLOOD STRUCTURE AFTER DRYING AND SETTING

When completed, nearly the whole surface of the dune is covered by blood material. Some areas remain untouched behind components, however in general the surface adhesion of the blood material makes it cling to the diverting components. This means that the material hugs the component and follows it around to the shadow side. This fills in the intended clear zone behind the component with material. This process will need to be revised slightly to achieve the results created through Real Flow simulation.

When removed from the oven the blood has completely set and has a black / red colour. Unfortunately, some large cracking is visible in the surface of the structure. This is because the material contracts slightly when it coagulates. Whilst this was not an issue when making bricks with material, as they are so thick, with the thin shell structure created here it causes the material to crack at certain points.

After the failure of the formula used in phase 2, a different direction was required. Firstly, the use of exotic and dangerous chemicals to create a glue makes little sense in a remote and poor society. Secondly, the attempts to manufacture glue from blood have proved largely unsuccessful and there is no current body of knowledge to consult on these processes. I was aware that blood and water on their own form a glue of significant adhesion, but how could a mixture of just blood and water be made to set fully throughout its depth? When blood and water set to form a glue, what occurs is the coagulation of the blood due to loss of water. Once the blood has become coagulated it forms an insoluble solid which binds strongly. I had found this from the dried blood found on the glass containers used in previous experiments. This is not the only way in which blood will coagulate though. It will also do so if heated over 64 oC. This means that once placed into moulds the mixture could be placed in an oven and heated to this temperature, resulting in the setting of the blood throughout its depth. This allows for the creating of a blood brick without the addition of any chemicals at all. Whilst this may seem like a return to traditional brick making techniques, the temperature required is very very low by comparison. This temperature could be achieved by either using solar ovens, or laying the bricks in the desert sun.

Blood mixed with sand is poured on to the crest of a test dune. This mixture is liquid but has a very thick consistency meaning it flows slowly over the surface of the dune. This diverting components channel the flow to create openings in the surface of the structure. As the blood flows over the surface it sinks through the sand to some degree thickening the structural depth of the shell created. The components are coated in petroleum jelly beforehand to allow for easy removal. This prevents the blood adhesive from bonding to the plastic. If the components can not be easily removed then there is a risk that the structure will be broken during the removal process.

Having been poured, the mixture must be dried at around 70 oC to coagulate the blood. This sets the blood / sand mixture into a solid material. The mixture requires around 2 hours in the drying oven to achieve this coagulation through its thickness.


01 EXPERIMENTATION

SANGUIS ET PULVIS

50 _ BLOOD BRICK MAKING PHASE 4 01 EXTERNAL SURFACE

02 INTERNAL SPACE

03 WIDESPREAD STRUCTURAL CRACKING AND FAILURE


02 REALISATION

SANGUIS ET PULVIS

_CHAPTER 02 _REALISATION DESIGN DEVELOPMENT 51 - 97

Site 52 - 58

Program 59 - 65 Metaball Modelling 66 - 71

Fluid Dynamics Simulation

72 - 78

Secondary dune formations

79 - 83

Time based construction

84 - 89

Segmented dune migration

90 - 97

FINAL PROPOSAL 98 - 114

Plans 98 - 101

Sections 102 - 103

Time based perspectives

104 - 106

Physical model 1:50 107 - 109 Renders 110 - 114


02 REALISATION

SANGUIS ET PULVIS

52 _ SIWA OASIS, WESTERN DESERT, EGYPT

01 _ THE WESTERN DESERT OASIS, EGYPT

Sanguis et Pulvis seeks to reestablish the dissipating autonomy of the Siwa Oasis in Egypt’s Western Desert. Throughout its history the Berber settlement has remained isolated and self sufficient; its traditional economy based on agriculture, its vernacular based on mud brick construction. However, two environmental phenomena have jeopardised this ancient way of life. Firstly, the encroaching sands of the Sahara have significantly reduced the farmable land surrounding Siwa. This has forced the town to focus its economy increasingly on the unreliable and highly seasonal tourism industry. This has greatly eroded both the towns economic and agricultural self sufficiency. Secondly, a series of rare rainstorms in the 1920s and 1930s caused great damage to the towns historic centre, effectively melting the salt rich mud bricks. This destroyed faith in the material on which the towns vernacular was based despite thousands of years of reliance on it. Since this, the towns development has been almost entirely in concrete, reducing traditional structures to tourist memorabilia. The project addresses these issues in three ways: it creates a new local construction material based on slaughtered animal blood, it creates a new economic base through the low-tech generation of solar electricity, and it enables large scale agricultural land reclamation by forming a protective barrier from the Sahara. Material The project incorporates several halal abattoirs which harvest the waste blood from slaughtered cattle. The waste blood (mixed with a preservative and an anticoagulant to allow storage) is mixed with sand to create the base of the material. This can then either be poured over a sand dune using the crest pouring process to continue the Sanguis et Pulvis project, or can be cast as bricks for use by the local community. In order to become a solid structural material it must be heated sufficiently to coagulate the blood. This can be achieved simply by leaving in the desert sun, or by setting in solar ovens. Generation Desertec is a multi-billion dollar project of a scale never seen before in the energy industry. It’s mission is to exploit the fact that an area of the Sahara the size of Wales receives enough solar energy to theoretically power the entire of Europe and North Africa. It aims to create a vast international network of high voltage power lines across Europe and North Africa called MedGrid, linking solar farms, wind farms and geothermal, biomass and hydro power stations, and capable of supplying 40% of Europe and North Africa’s electricity. Energy would be produced throughout the network, in multiple countries, and flow internationally in an open manner. This decentralisation of generation is the key to the implementation of an effective renewable energy network. When one source fails, multiple others are available to compensate. It is however a project tinged with strong intimations of neocolonialism; of resource exploitation of Africa by Europe. Desertec is mainly funded by German companies, and most of the cost of Desertec will go to German engineering companies who manufacture the complex equipment used. The sun is the most valuable resource possessed by Saharan countries and must become the economic foundation of their emerging communities in a manner that is locally implementable and beneficial. As such the project uses fresnel lense solar concentrators and Stirling engines to generate electicity. Both of these technologies are fairly simple to construct and can be manufactured locally. Reclamation The project is realised in full over many years, eventually creating a long barrier protecting the town of Siwa from the Sahara. This barrier will prevent the further encroachment of the Sahara and allow agricultural land reclamation to occur in the shadow of the building. This will allow the reclamation of up to 14000 acres of land for agricultural use. This will boost the historic agricultural economy of the town, as well as providing the necessary grazing area for increasingly large cattle herds.

02 _ SIWA OASIS

03 _ SIWA


02 REALISATION

SANGUIS ET PULVIS

53 _ SIWA OASIS

01 _ SIWA OASIS

Siwa oasis is an incredibly fertile area and because of this has always based its economy on the production of dates and olives. The oasis has 400,000 date and olive trees which are communally owned by the inhabitants, who share the income from their sale. The rich agricultural lands are being encroached upon by the Great Sand Sea however, damaging the towns agricultural output and jeopardising their future. The panoramic image opposite illustrates clearly the relationship between the desert and the town. The historic settlement of Shali is at the very centre of the plantation area and sits on top of a large rock outcrop. Lake Siwa is a large but incredibly salty body of water which can not be used for agricultural purposes. Siwa contains over 200 fresh water springs which produce water from the abundant aquifer beneath the town. Many of these springs drain into the lake, where the water becomes so salty that nothing can survive due to the mineralogy of the soil. This is a key element in the agricultural future of the town, in that these drainage routes should be redirected into irrigation. The saltiness of the soil which creates the salt lake was also responsible for the destruction of the town of Shali. The mud bricks used naturally contain a large amount of salt, and when a freak rain storm hit the town early in the 20th century this salt dissolved and lost its structural integrity. This is one of the key aims of the project, to reintroduce a local structural material to the community in order to revive an individual desert vernacular.

THE GREAT SAND SEA

THE PALM MARSHES

SIWA LAKE

OLD MUD BRICK HOUSE RUINS

SHALI GHADI

THE LIBYAN PLATEAU


02 REALISATION

SANGUIS ET PULVIS

54 _ SHALI - A DESERT VERNACULAR The opposite map shows the old town settlement of Shali. This has been the centre of Siwa since it was first occupied in around 500BC. Due to the towns position on the trade routes between Egypt and Libya it was regularly attacked, and as such the town was constructed on top of a large rock outcrop and the entire town was contained within its fortified walls. At this time it was a very small settlement and this was possible. As the town began to grow it spread beyond the walls of Shali and became the town of Siwa.

01 _ SHALI (OLD TOWN), ORIGINAL WALLS

02 _ THE ‘MELTING’ OF SHALI 01 _ SHALI, 1923

03 _ SHALI’S ‘MELTED’ WALLS

02 _ SHALI, 2012

04 _ AIT BENHADDOU, MOROCCO

However, in 1924 and again in 1928 freak rainstorms occured in the region, which is normally classified as super arid (less than 5mm rainfall per year). This rainfall dissolved the salt present in the mud bricks created from the local soil and ‘melted’ the towns buildings and fortifications. Whilst the historic town of Shali has been left untouched as a ruin, the rest of the town was knocked down and reconstucted using generic concrete based structures. This destroyed the individual feel and atmosphere created by the unique vernacular that existed previously, and broke the desert tradition of building with the material readily available in the environment. The project aims to address this development by creating and reintroducing a locally supplied material to reestablish a unique vernacular in the town. This material will utilize one of the most abundant waste products of North African countries - blood.

05 _ ANCIENT GERAMA, LIBYA


02 REALISATION

SANGUIS ET PULVIS

55 _ SAND DUNE FORMATION AND MIGRATION

01 _ VARIATIONS OF SAND TRANSPORTATION

02 _ FORMATION OF SAND DUNES

03 _ SAND DUNE INTERNAL STRUCTURE

Whilst sand dunes can grow to be hundreds of meters high and many kilometres long, their behaviour ultimately comes down to the interaction between individual sand grains. This is why dune behaviour is so complex to understand and is still to a large extent unproven, as within a dune there are many billions of grain interactions occurring simultaneously.

01 _ CREEP (COARSE PARTICLES)

01 _ BARRIER CREATES LEEWARD DEPOSITION

01 _ DEEP RADAR IMAGE OF LARGE AND OLD SAND DUNE

WIND

The way in which sand dunes form and migrate is in particular governed by these grain interactions as described in the following diagrams. AIR CONVERGES AND VELOCITY INCREASES

02 _ SALTATION (MEDIUM PARTICLES)

AIR DIVERGES AND VELOCITY DECREASES

BARRIER

DEPOSITED SEDIMENT

WIND

02 _ DUNE ENVELOPES BARRIER

03 _ SUSPENSION (FINE PARTICLES) AIR CONVERGES AND VELOCITY INCREASES

AIR DIVERGES AND VELOCITY DECREASES

WIND UPWIND SLOPE

SLIP FACE

BARRIER PARTICLE TRANSPORTATION

DUNE FORMATION

DUNE STRUCTURE

Sand dunes rely on particle transportation for their formation and sustenance. There are three types of transportation related to dunes: creep, saltation and suspension. These act upon coarse, medium and fine particles respectively. Creep is a very slow process, and suspension involves particles too fine to play a key role in dunes. Therefore saltation is the key process in relation to dunes.

Dunes form as a result of the saltation process. As sand travels it is carried in high energy bursts by the wind. Therefore when wind encounters a barrier, and thus a pocket of low pressure is created behind it, sand is deposited in this region. Once sand begins to accumulate, it acts as an absorber to other sand particles, as landing in a pile absorbs more energy and slows transportation. From this humble beginning dunes can grow over long periods of time to be up to 300m high and 50km long.

As dunes are formed in such a slow and progressive way, traces of their formation are left in their structure. As sand blows up the windward face and deposits on the slip face, layers of sediment begin to become defined. These layers are most apparent in sand stones which display the clear patterning, and can show the wind direction because of this.

In saltation, particles are given large amounts of energy through collisions with other particles. This forces them up off the ground only a few inches, but high enough to be carried a short distance by the wind, in which they gain energy. When they land this energy is transferred to the particles with which they collide, forcing them up into the wind. In this way particles can move a long way relatively quickly.

The section shows the typical dune shape, with a low angle upwind slow and a steep angle slip face. The slip face defines the angle of repose for the material. As wind acts upon the sand, it is blown up the windward slope (angle defined by the energy of the system), reaches the crest and cascades down the slip face into the low pressure zone.

The above image is a deep radar from a large old dune. The layering apparent here shows the stages of growth the dune went through and also shows that the shape is partly due to the underlying topography. This topography may well be an older incarnation of the dune which became stabilised and solidified, before reactivating in a more arid time. The most clear layering is at the surface, with the lower levels more distorted. This is due to failures and avalanches of the dune.


02 REALISATION

SANGUIS ET PULVIS

56 _ DUNE TYPES Sand dunes vary greatly in shape and scale, and this variation is largely defined by the wind and the availability of sand in the environment. A continuous wind direction will create different dunes to a varying wind direction. Equally, an environment with abundant sand will create different dunes to an environment with scarce sand.

01 _ BARCHAN DUNES

02 _ TRANSVERSE DUNES

03 _ LINEAR DUNES

04 _ STAR DUNES

As the project intends to use sand dunes as the basis for creating structures it is vital to understand the range and characteristics of the various dune formations. It is also important to understand the way in which dunes of different types migrate, as the project is a long time based construction process and this will inevitably effect the design. Sand dunes exist as 4 main types: barchan, transverse, linear and star. Whilst some other types of dune may appear to exist, they are in fact usually amalgamations of these types. Each of these dune types is formed in a different manner, exists due to a different set of conditions, and migrate in a different manner.

D

WIN

WIN

D

IN W

WIN

D

ND

D

WIN

WI

D

WIND

WIND DIRECTION: CONSTANT SAND AVAILABILITY: SCARCE

WIND DIRECTION: CONSTANT SAND AVAILABILITY: ABUNDANT

WIND DIRECTION: BIDIRECTIONAL SAND AVAILABILITY: SCARCE

WIND DIRECTION: MULTIDIRECTIONAL SAND AVAILABILITY: ABUNDANT

Brachan dunes are the most common form of dune found in deserts. They are typically formed when there is relatively little sand available and when the wind blows from a constant direction.

Transverse dunes usually occupy vast areas as continuous fields. They are often created when fields of brachan dunes become inundated with additional material, enabling them to join together in long bands. The long bands then cease to function as individuals but act as a mega dune instead, often many kilometres long and hundreds of meters high. Like brachans they also rely on a constant wind direction to form, however require much more sand material to be available.

Linear dunes are formed when there is relatively little sand material available and the wind blows from two slightly different directions. This alternation of wind direction effectively causes the creation of a windward face on both sides of the dune as material is blown up towards the crest. This means that the dunes are generally smaller and have a lower angle.

Star dunes are formed from any availability of material, but vitally require winds to blow from multiple directions. As the winds shift around the dune, material is gathered from all directions and blown together to form a central mound. Material is blown up various windward faces and cascades down various slip faces, resulting in highly complex and varied geometries and scales.

As wind strikes the dune it is diverted around it. As the wind passes around the dunes edges and over its ridge, areas of low pressure are created on the insider of the dunes wings and on the slip face. This means that wind speed slows and any sand which was being carried is suspension is dropped in the low pressure area. It also means that rather than blowing out into the open expanse found between dunes, it generally remains as part of the dune, being sucked back into its concave side. This is how the dune achieve their distinctive form. Brachan dunes migrate relatively quickly and can move up to 100m / year. As they do so at different rates, dunes inevitably cross each others paths, combining and amalgamating. In this way complex arrangements and geometries can be formed which occasionally resemble transverse dunes.

As macro variations in wind occur, different sections of the dune will advance at different rates. This gives the dunes their distinctive waving ridge line. Due to the scale of the dunes their windward face is often very shallow and long, creating a dramatic contrast with the slip face. The scale of transverse dunes is often considered as their wavelength. This is the distance between crests in the field. The wavelength of dunes can reach many kilometres, in which case there are often desert pavements found between dunes. These area areas where no sand is present and fine particles have been blown away, leaving only coarse material. This forms a stable surface through the desert, and are often used to traverse large expanses of dunes.

Linear dunes do not migrate as brachans and transverses do, but more elongate or stretch. Material can pass down the channels between them and become attached at some stage further down the dune, growing the end of it.

As star dunes are central formations, they do not migrate. Their extremities may stretch and distort but their main mass will remain fairly constant. As such star dunes do not join up with each other, they do already however resemble multiple joined dunes. In areas of large material availability star dunes can become enormous, as they only ever accumulate more material. They can reach up to 200m high and be several kilometres across.


02 REALISATION

SANGUIS ET PULVIS

57 _ DUNE TYPES OF THE GREAT SAND SEA

01 _ NORTHERN REGION OF THE GREAT SAND SEA

The factors previously described which control sand dune formation vary considerably across the vast expanse of the Great Sand Sea. In the middle of the sea, where there are very few obstacles or geometric irregularities, the wind behaves very differently to the edge of the sea, where the Libyan plateau effects wind flow and creates an aeolian micro climate which differs from the regional average.

SIWA

This boundary area of the Great Sand Sea is where the project will be sited, and as such these are the dunes which will be solidified by it. The character, scale, and geometry of these dunes will have a huge impact on the design of the project. The Libyan Plateau causes the predominant wind to be pulled around into the low pressure zone it creates. The way in which this occurs depends on the exact wind direction and its interaction with the plateau, but generally the result is a less uniform wind direction towards the north east. This relative irregularity in wind direction has a clear effect on the dunes present. Whilst in the middle of the sea extremely regular and massive scale transverse dunes occur, at the edge condition the dunes become smaller and mixed in character. The general type however is a combination of barchan and transverse dunes. When combined this creates what looks like a very elongated barchan with a very irregular crest line. These dunes tend to dissipate to nothing at their tales as with regular barchans.

18 km

18 km

TRANSVERSE DUNES: MEGA SCALE

TRANSVERSE DUNES: LARGE SCALE

The transverse dunes of the central region of the Great Sand Sea are of an enormous scale. Individual dune ridges can run for up to 50km, the wavelength between dunes can be as much as 5km and dune heights reach 200m.

The transverse dunes of the northern region of the Great Sand Sea are of a large scale. Individual dune ridges can run for up to 15km, the wavelength between dunes can be as much as 2km and dune heights reach 100m.

Being positioned at the centre of the sea there is an abundant supply of sand which enables these dunes to form and sustain themselves. The dune fields migrate slowly as the dunes are so big that the low pressure on the slip slope is very big. This means that sand does not break free on the slip slope and move quickly across the break to reach the dune in front.

Being positioned at the north of the sea there is still an abundant supply of sand but not as large or as stable as the central region. The dune fields migrate at a medium pace, as the dunes are big the low pressure on the slip slope is quite big. This means that sand struggles to break free on the slip slope and move quickly across the break to reach the dune in front.

Between the huge ridges of the transverse dunes, smaller patterns of dunes form. These are akin to large scale ripples, forming in the turbulent winds that come off the dunes. When viewed in detail these may appear to be formations of their own, but the overview clearly shows the dominant pattern of dune formation.

Between the ridges of these transverse dunes, a very small pattern of dunes form. These are akin to large scale ripples, forming in the turbulent winds that come off the dunes. When viewed in detail these may appear to be formations of their own, but the overview clearly shows the dominant pattern of dune formation.

5 km BARCHAN DUNES: LARGE SCALE STAR DUNES: SMALL SCALE Barchan dunes are rare in the Great Sand Sea as they are generally associated with a scarcity of sand. They do however form at the very northern edge next to Siwa. This is partly due to the reduced availability of sand at the edge condition but is also due to the effect of the Siwa depression on local winds. The strongest winds blow around Siwa because of the large cliffs that sit to its north, therefore the majority of sand material is blown into the central regions. The barchan dunes found here are very large. Their ridges can be 600m long and they can be 50m high in their central region. Barchan dunes migrate relatively quickly and often engulf each other of join together. This results in complex dune geometries which are in constant flux as they are shaped by the wind. Small star dunes can also be seen between barchans. This suggests that winds varies in direction. Whilst this occurs enough to form small stars, it does not occur enough to greatly infulence the barchans.


02 REALISATION

SANGUIS ET PULVIS

58 _ WINDS OF THE GREAT SAND SEA 01 _ WINTER WIND ANALYSIS, SIWA, EGYPT Dune migration is dictated by wind direction and regularity. If a strong and very constant wind exists then migration in that environment will be very fast. Conversely if there is a strong but greatly varying wind pattern then star dunes will form which do not migrate. From the wind data for Siwa it is clear that there is a fairly strong predominant wind from the south west. This varies a little between summer and winter but is stable enough to create constant and moderately fast dune migration.

02 _ SUMMER WIND ANALYSIS, SIWA, EGYPT


02 REALISATION

SANGUIS ET PULVIS

59 _ GLOBAL INSOLATION

01 _ AVERAGE YEARLY GLOBAL INSOLATION

This map illustrating global solar insolation levels shows clearly why the Sahara is such a desirable location for solar electricity generation. It receives higher average insolation levels than any other area on earth. Due to this a solar panel in the Sahara will produce around 3 times as much power as one in the UK during the summer. Other areas with very high insolation include Australia, South Africa, the Middle East and Chile. In general this pattern matches that of areas with very little rainfall. This is because these areas have exceptionally low cloud cover, which combined with desert conditions amounts to the maximum possible solar insolation.

2500 kWh / m2

2500 kWh / m2

2500 kWh / m2

2500 kWh / m2

2500 kWh / m2


02 REALISATION

SANGUIS ET PULVIS

60 _ GLOBAL ELECTRICITY NETWORKS

01 _ GLOBAL HVDC ELECTRICITY TRANSMISSION NETWORK

02 _ DESERTEC / MEDGRID INTERNATIONAL HVDC ELECTRICITY TRANSMISSION NETWORK

The project seeks to establish a new and sustainable economic base for the town of Siwa which will be able to support its continued growth and development. This is to be achieved through the establishment of Siwa as a Desertec site, selling electricity via MedGrid to the single European energy market. The Sahara is the most abundant source of solar energy on the planet. An area the size of Wales in the Sahara receives enough solar energy in one year to power the entire of Europe and Africa. The issue currently is that the infrastructure is not in place to enable the sale of electricity from the Sahara to Europe. Without this infrastructure and the guaranteed sales provided by the European market, the vast investment required to create solar farms is not available. However this is set to change imminently with the progression of the Desertec Industrial Initiative. This initiative is driving the creation of an international supergrid which will facilitate the international sale of Saharan energy. The maps opposite illustrate the intentions of the super grid and how it will connect Europe, North Africa and the Middle East into a seamless open energy market. The project in Siwa will become a key node on this new grid.

GENERATION POINT USAGE POINT

HVDC CABLES

SOLAR THERMAL

BIOMASS

DESERTEC/MEDGRID HVDC CABLES

WIND

PHOTOVOLTAIC

POTENTIAL HVDC EXTENSIONS

HYDRO

GEOTHERMAL


02 REALISATION

SANGUIS ET PULVIS

61 _ LOW TECH SOLAR CONCENTRATION Desertec in unfortunately a project tinged with strong intimations of neocolonialism; of resource exploitation of Africa by Europe. Desertec is mainly funded by German companies, and most of the cost of Desertec will go to German engineering companies who manufacture the complex equipment used. The sun is the most valuable resource possessed by Saharan countries and must become the economic foundation of their emerging communities in a manner that is locally implementable and beneficial. As such the project uses fresnel lense solar concentrators and Stirling engines to generate electricity. Both of these technologies are fairly simple to construct and can be manufactured locally. As illustrated opposite, fresnel lenses act to bend light toward a focal point behind the lense. This is the opposite of a parabolic trough mirror which reflects light toward a focal point in front of the mirror. Parabolic troughs, the favoured technology of Desertec, are extremely complex to manufacture and are only produced by 2 companies in Germany. They are also not ideally suited to the desert environment as even a thin covering of dust massively reduces their reflectivity and thus their power output. Fresnel lenses however are very simple and cheap to manufacture so can be produced locally, driving the investment capital of the project into the local economy. They are also more appropriate for the desert as, because they bend rather than reflect, a thin layer of dust over their surface does not have nearly a big an effect on their power output. This makes them the most suitable concentration equipment for this project. Desertec uses, in conjunction with parabolic troughs, steam turbines to generate electricity. This is not appropriate for rural desert communities as steam turbines are extremely expensive to install and require extensive maintenance. Also the abrasive sandy environment is likely to damage the turbines. The project proposes the use of an alternate generation technology: Stirling Generators. Stirling generators convert a heat differential directly into work. They are completely sealed to the external environment, are the most efficient technology available, and are low tech in terms of manufacture, again allowing local production and investment. In this arrangement, a fresnel lense is used to heat a transfer fluid, which heats the hot chamber of the Stirling generator.

01 _ SOLAR CONCENTRATION TECHNIQUES

01 _ PARABOLIC TROUGH EXPENSIVE HIGH MAINENANCE SPECIALIST MANUFACTURE

02 _ CONVENTIONAL LENSE EXPENSIVE THICK HEAVY LOW MAINTENANCE

03 _ PROPOSED SYSTEM (FRESNEL / STIRLING)

04 _ THE STIRLING GENERATOR

Piston flexures - support and spring the piston / alternator assembly

Parabolic troughs are the current favoured technique of Desertec developments. They operate by reflecting sun light towards a central point in order to heat a transfer fluid. Parabolic troughs are expensive and difficult to manufacture, and require extensive maintenance as a very thin layer of dust will have a drastic effect on their reflectivity.

Linear alternator, magnets, stator and coils remain stationary and produce power Linear alternator mover - reciprocates with piston to produce power

Conventional lenses operate by bending sun light towards a central point in order to heat a transfer fluid. Their advantage of parabolic troughs is that they do not reflect but bend, and therefore a thin layer of dust on their surface has a less significant effect on their performance. They are however extremely thick, expensive and heavy due to the extremity of their optical nature.

Power piston - reciprocates and drives the alternator

Close clearance seals - isolates gas working space

Displacer rod - drives the displacer Cooler - transfers heat from the working gas

03 _ LENSE WASTAGE

Regenerator - improves the engine cycle efficiency

Much of the thickness of the lense is however redundant. The significant surface is the curved surface, and the rest of the mass simply allows this curvature.

Displacer - reciprocates, shuttling fluid between expansion and compression spaces Heater - transfers heat to the working gas

04 _ FRESNEL LENSE CHEAP THIN LIGHT LOW MAINTENANCE

By keeping only the affective part of the lense a much thinner, lighter and cheaper lense is created, known as a fresnel lense. These have the advantages of a lense without any of the draw backs.

Displacer flexures - support and spring the displacer

A low tech generation system is created by using a fresnel lense and a Stirling generator. A Stirling generator works by converting a temperature differential into work. This reduces the complexity of the system as the hot side remains as hot as possible while the cold side remains as cold as possible. This removes the need for complex heat exchangers within the system. The fresnel lense heats a transfer unit which heats a transfer fluid. This fluid is used to maintain a very high temperature at one end of the Stirling engine. The cool end of the engine is maintained either through a simple radiator cooling system or through an active cooling system. The greater the temperature differential the greater the power output of the engine, thus an active efficient cooling system is preferable. The motion produced by the engine is cyclic and linked directly to a generator creating AC power. An efficient Stirling generator can operate at up to 30% efficiency and requires very little maintenance through its life cycle. The key components of the lense and Stirling generator are low tech and can be easily manufactured locally.

A Stirling engine directly converts a heat differential into work via the expansion and contraction of a working gas. The heat input can be as small as the body heat from a persons palm, however for large powerful engines a temperature differential of several hundred degrees is required. This can be easily provided by solar concentration lenses. The hot side of the engine heats the working gas as it flows into the hot chamber. This causes the working gas to expand and drive the displacer. As the dispacer moves it forces the working gas into the cool chamber where it contracts, driving the displacer back again. The cycle alternates continuously, and is extremely efficient as the hot and cold sides remain distinct. The regenerator recovers energy as the working gas passes from chamber to chamber to increase efficiency. The motion of the displacer drives the power piston which moves a coil within a magnetic field to generate. As the coil alternates through the magnetic field an AC current is generated.


02 REALISATION

SANGUIS ET PULVIS

62 _ THE HALAL ABATTOIR The proposal of a new locally produced material for Siwa begins with an abundant waste product in North Africa: blood. Counties which follow halal meat processes produce vast amounts of blood through the complete draining of carcasses before they are hung. Typically this blood is incinerated, although occasionally it is used to make fertilizers. Blood has exceptional adhesive qualities however, and was used for a long time to create albumin adhesive for use in the manufactured timber industry. In the same way as salt was utilized as a binding agent in brief 2, blood can be used here.

01 _ WHY WASTE BLOOD?

02 _ ABATTOIR PROCESSES

AVERAGE FEMALE COW 500 KG MASS

AVERAGE FEMALE COW 40 LITRES OF BLOOD

40 LITRES OF BLOOD PER COW

ALL BLOOD DRAINED IN HALAL PROCESS

01 _ BLESSING AND BLEEDING

02 _ SKINNING

03 _ EVISCERATION

04 _ SPLITTING AND COLD STORAGE

BLESSING AND BLEEDING

SKINNING

EVISCERATION

SPLITTING AND COLD STORAGE

This is the most distinctive and important stage of the halal meat process. All animals are blessed in the name of Allah before having their throat deeply slit.

Once the carcass has hung for a sufficient period for all of the blood to drain from its body, it moves to the first stage of the linear abattoir system: skinning.

For ease of storage and sale the carcass is split directly down the middle. This produces slabs of meat which hang neatly and efficiently.

The animals are hung whilst all of the blood drains from their body as the Quran stipulates Muslims must not eat blood.

Using a sharp knife the bind between the skin and flesh is cut carefully. The skin is sold for leather and other uses, so it is essential that it is removed cleanly without damaging it any way. Once the skin has been removed the bare flesh of the carcass is left exposed and cleanliness is absolutely vital. Because of this, abattoirs use a system of overhead runs from which the carcasses are hung by the legs. The carcasses can then be moved from station to station without ever touching the ground and therefore remaining as clean as possible.

Evisceration involves the removal of the internal organs of the carcass. This is very important and must be done very soon after the death of the animal as they will quickly begin to taint the flavour of the meat if they are left in for too long.

An average cow produces 40 litres of blood upon slaughter. This can be collected and mixed with preservative and anticoagulant to allow storage and accumulation. When sufficient blood has been stored it can be crest poured to solidify a sand dune into a structural shell. This process relies upon the heat of the desert sun to set the material. The blood must be heated to above 53 degrees in order to clot and set. This temperature will be easily achieved in the searing irradiation of the Sahara where even the air temperature regularly exceeds this. The abattoirs will also produce enough meat to make the town self sufficient. The enlarged cattle herds that will be required can be sustained on the reclaimed agricultural land created by the project’s effect on dune migration towards the town.

MIXED WITH SAND AND POURED TO CREATE STRUCTURE

MIXED WITH PRESERVATIVE AND ANTICOAGULANT FOR STORAGE

1:5 SAND

This process creates the vast amount of waste blood which this project utilizes as a component for new building material.

A deep cut is made down the belly of the carcass the length of the body. From this cut the organs are removed and stored in a special chilled by-products area. These parts of the meat can often be sold for subprime meat products such as animal food. Once evisceration has taken place the rapidly perishable elements of the carcass have been removed. This enables the carcass to be hung and matured.

The carcasses can then enter the hanging phase which is required for the meat to mature in texture and flavour. Depending on the quality of meat and the end purpose this process can last for up to a month. For lower end products though they may only be hung for several days. Cold storage constitutes one of the most expensive elements of the abattoir. The energy required to reduce the temperature of a large carcass from 25 degrees to 5 degrees is large and as such the load on cooling systems is great. This is a large challenge for any abattoir, and especially one based in the desert.


02 REALISATION

SANGUIS ET PULVIS

63 _ PROGRAMMATIC AREAS

01 _ PROGRAMMATIC AREAS

Having established the program for the building and the areas on which it must focus it is possible to begin diagramming the spatial requirements of this program. Once rough figures have been calculated for the various programmatic elements a diagrammatic plan can be formed to guide the design process.

CATTLE SHEDS 2200 M2 BANKED BEDDING

SOLAR COLLECTION / GENERATION 800 M2 FEEDING

281 M2 ACCESS ALLEY

Cattle associated buildings will occupy the majority of the gross area. This is to be expected given the area required to keep and manage livestock. Sheds provide soft bedding and standing areas for cows to spend long periods of time. Bedding areas connect directly to feeding areas, where troughs provide a constant availability of feed. As such livestock feed freely and pass between these areas as they wish. During the cooler parts of the day livestock will be kept in either enclosed pens of taken to open grazing. In this time the bedding and feeding areas can be cleaned and maintained. These areas should ideally exist either side of an access alley. This alley provides access for a tractor to carry out cleaning and maintenance as well as restocking feed troughs. The alley may also be used for the moving of cattle in and out of the sheds. Grazing lands are also required on top of the areas shown here, however these areas are so disproportionately large that they have not been represented here. Ideally both small enclosed and large open external areas are required.

1600 M2 FRESNEL LENSE CONCENTRATOR

OFFICE 100 M2 STIRLING GENERATORS

70 M2 CONVERT/ TRANSFORM

The majority of the area associated with solar generation is required for concentration lenses. These lenses concentrate the energy received from the sun across a large area to a very small point. This enables the heating of a transfer fluid to several thousand degrees. Stirling generators are fairly small as there are many of them, as opposed to one very large generator. The area they occupy is very small compared to the concentrators required to power them. In order to sell electricity to the MedGrid as well as use it locally, a series of converters and transformers will be required. Stirling generators created AC current, which can be used locally once stepped to the correct voltage. However MedGrid operates a high voltage DC network, as such electricity will need to be converted before sale to the grid.

550 M2 OFFICE

A fairly small amount of office space (for 70) is provided. This will provide an administrative base for all the programs associated with the building.

CHEMICAL STORAGE / MANUFACTURE 300 M3 BLOOD STORAGE

300 M3 GLUE MANUFACTURE/ STORAGE

ABATTOIR 180 M3 SLURRY STORAGE

Due to the complex closed loop systems used in the building a large amount of storage as well as chemical production facilities is required. Much of this can be placed beneath ground and concealed. However some aspects, such as blood storage and blood glue manufacture may wish to become focal public elements of the building. In this case they will need to be sited above ground, and will require pumping to.

200 M2 CIRCULATION

SERVICES 120 M2 CHILLING/ STORAGE

120 M2 BYPRODUCTS STORAGE

60 M2 KILLING/ DRESSING

40 M2 GENERAL STORAGE

25 M2 COLD DOCKS

25 M2 REST

The abattoir will handle the halal slaughter of livestock reared by the facility, as well as the slaughter of any animals being brought to the facility. Efficient modern abattoirs tend to adopt a linear format with each bay leading to the next for a clean and quick process. The abattoir is one of the most energy demanding elements of the program as it requires cooling, especially the cold storage area. As such it is likely that much of this program will exist beneath ground to provide the maximum protection from the harsh desert environment.

60 M2 AMMONIA COOLING SYSTEM

60 M2 BIOMASS BURNER

The buildings services consist of a biomass burner and an endothermic reaction cooling system. These provide heating and cooling to the stirling generators and cool the abattoir spaces.


02 REALISATION

SANGUIS ET PULVIS

64 _ DIAGRAMATIC PLAN

01 _ DIAGRAMATIC PLAN

01 _ COW ACCESS / SHEDS / TRANSFER / GRAZING

This diagrammatic establishes the key relationships and adjacencies that must be developed in the design for the building. These relationships are most vital in the cow sheds and abattoir as a detailed succession of spaces is required which must meet the various standards applying to food production. The diagrammatic plan splits the project into two sides, an agricultural side and a generation side. This also allows the splitting of public and private access to keep the rough agricultural side distinct.

ROUTE TO OPEN GRAZING

The program of cow sheds and abattoirs will exist in secondary dunes extending perpendicular to the main central dune which will contain public circulation and the Stirling generators.

A

ROUTE TO OPEN GRAZING

ATR CSH

SECURE OURDOOR PEN

ABT

ATR

ATR

CSH

SECURE OURDOOR PEN

ROUTE TO OPEN GRAZING

ROUTE TO OPEN GRAZING

ABT

ABT

AA/VC

MCS CHB

MCS CHB

OFF

ABT

AA/VC

OFF PUBLIC CIRC.

STIRLING GEN. PLATFORMS

02 _ ABATTOIR / COLD STORAGE / CHILLED LOADING

AA/VC

MCS CHB

OFF

A

CSH

SECURE OURDOOR PEN

SECURE OURDOOR PEN

AA/VC

MCS

CSH

CHB OFF

03 _ PUBLIC OBSERVATION / SOLAR CONCENTRATORS / STIRLING GENERATORS / OFFICES


02 REALISATION

SANGUIS ET PULVIS

65 _ DIAGRAMATIC SECTIONS

01 _ DIAGRAMATIC SECTION

The following diagrammatic sections have been developed in conjunction with the plan. Again these describe a series of key adjacencies which must be maintained through the design phase. The most important of these from a performance perspective is the maintaining of a cool zone based around the meat cold storage. Any spaces which require moderate cooling, such as offices or chilled loading bay, should be positioned around the cold storage sectionally. This will make the cold storage more efficient by it not only being underground but also surrounded by cool spaces.

FRESNEL LENSE CONCENTRATORS

THERMODYNAMIC STIRLING GENERATORS

CIRC.

OFFICE

OFFICE

CIRC.

OFFICE

CHILLED BAY

ACCESS ROAD

ANIMAL ARRIVALS / VET CHECK

COW SHEDS

A

A HEATING

COOLING

MEAT COLD STORAGE

ABATTOIR

SLURRY PIT CHEMICAL PROCESSING

02 _ THE SECTIONAL COLD ZONE THE COLD ZONE One of the most expensive elements of an abattoir to run is the cold storage. Cold storage must be maintained around 5 degrees and this is extremely hard to achieve in a desert. As such it is vital that the cold storage is afforded maximum protection from the external desert environment. To resolve this the cold storage is placed underground, providing great thermal mass to insulate it. It is also in this way protected from the sun and external heat. To maximise the efficiency of cooling throughout the whole building, any other spaces which require cooling are the placed around the cold storage, this includes the abattoir, chilled loading bay and offices. These will receive a cooling effect from the cold store and maintain a habitable climate. Finally, spaces where cooling would be beneficial but not necessary are placed beyond these. In this way the circulation and veterinary check areas will receive some cooling.

03 _ KEY SERVICE LINK KEY SERVICE LINK A key sectional connection is created by the services used in the building. A closed loop system of heating and cooling is created, which uses the faeces and urine create by cows to create biomass for burning and ammonium nitrate for use in an endothermic reaction cooling system. This cooling system provides cooling for the cold storage and other spaces, as well as cooling the cold side of the stirling generators. The biomass burner is used to heat the hot end of the stirling generators at night when there is no sun, in order to maintain 24 hour power generation. Faeces and urine are collected in the cow sheds and stored in the slurry pit. This faeces then must be processed to achieve the desired forms for both systems. This final product must then reach the heating and cooling units and the generation side of the building. This connection links the two extremities of the building and is key to the buildings workings.


JACK S MUNRO 01 DIGITISED EXPERIMENTS SANGUIS ET PULVIS 02 REALISATION

DS10

66 _ DUNE SOLIDIFICATION MODELING 81 For pavilion designed briefmolupta 2 a laser scanner was used in order digitise thesiminis physiEsti the omnient, core repercilinintis eritam voluptat oditasp eribustonimolupta cal created. However, in order to fullyinctiistrum develop a quid proposal at asdaesequi large scaleomnim and inquassi more aut model aut officatur antur ab iliam am veribus quatibu detail, technique be developed for tumendit modellingaut such occatesa sinvenit raemust culluptiis sus explita quistructures. quam utae. Ita doluptat iur, odionsectis autatibus dipicabore, odi comnihil explat aut modi cone quate dolorru ptatius My first attempt technique involves using Grasshopper analyse the andanditio maio at eadeveloping dionsenimethis nonsed que cor sitiam, voluptatiisi blanis to volecat imusanpaths of least resistance down dune molest face, these represent that liquid would dae maximpeliti unt quam serothe voluptio volore volupturthe autpath ex es prothe et quatinima flow down. The paths eturis can then be bifurcated to create finer network These volorpo repersperum ut provit vel illit volore eosamolorepudit ut of autcurves. ut dolla si iliqui curves can be thickened using to create a curve wire frame of the am reperiberum facepe nos nethree vendisaxis es of dismetaballs di re, omniaep udignim simus. desired is exploded to give a point which can beonseque meshed Et ipictegeometry. voluptaquiThis temwireframe evel minimaionem quatus exerias delcloud, ea diam et velesti using rotating algorithm. molumMeshLab’s eiunt optatquas niscubes doluptat ommodition res dolupit aspersped mi, con parciis quissim aute none verempo reptaquae sum secaess itataspicia consed quae. Caborrum de Whilst this ipsum technique not idealculparum and has many flaws is a good point from nonectum hiciaisinvenima quatur remitimos aut utstarting venda nusci unt. which I can developquid more detailed and accurate techniques. Soluptisim quam quiassimi, sant lignimet quam quid ullenim re excerro commoss imaionsequi omnihitas min nectioratur, omni quam re consecte et repe volecup taquam harum incit apit, ut et que liquianda et quate volutempor reicabor maxima cor ate vella is quiasim voluptat ratur, se veliquis autatur? Tiorum rersperum quatio. Upis incimus a corposam esto quid quae omnimi, et od eos est, tet fugitibus et, voluptatur, illuptiis maximax imolorem archicto occus incimpo resciis tiandus. Recta verspiciis illuptat. Gendita dusdae comnihi ciisquos maior simin niet in nos non cones ma etur ma nobitaqui quae miliandis senis doluptur? Facerum haruptur? Accae. Et quis es quibus moluta ea nonseriae nulliquatum eserchil modit et praes accuptatis ad modis apit ma delibus a quaest vendam, ium re, simolore cor sitis ipsum quam, id min nihilla ciurere ptatem vit od ut labo. Nam cus, oditasin conse maximet elest, con est offic te cusandus audis sim essimodi vit et aliciistis nonsectotate ipsam, tem dolupid untiscimus estenis deliquia sint autationecab ipsus, to maxima doloremporit volora illaboritae nonsedisti ilitatur maiorerum hil in conectur sum venim vollam audis dolorum quibus aut vel et fuga. Ut dion re, simolup taquos aut latest, quam nos doluptus nullecumquam fugiam lia preped mini deliquatur sum et ommod ullis accus si core, occulluptas alignis re, te nobitiam etur asperitassum ipsantis evelit litaspicatus mossi tem vollenis estis erum ne nos ute dolo quas repuda quam labore, voluptatecus eostis maximporem versperum, officat emolupt aquaessi ulparum quidella praturiorum explam sa non est et quiaspe rferspellaut eos exceped quam nonet volupta tatquae. Et as doluptat ditia volupta volorem num velibus exerovid moluptibusa ditiat quam sequis sitatquiate quia peria nus ea andisqui dolorepe doloriaturit evel evelige ndamus, ut que voluptur? Otatur sitet lab ipic to minveni tibeatent. Ut veligenisto omnis et officitia dusam, od ut eturibeatum et omnimaio. Ut ut unt eribus dolor sequost, que aut quae. Et veniendae voluptat audignam qui que velent ea voluptis asperiatur sequatur alibusciet aut omnimilles et endita perovidit quibere peritio intur? Quidis

01 _ TIME LAPSE SERIES OF BARCHAN MIGRATION (STARVED) 01

02

03

The base boundary and the crest of the dune are defined by unbroken curves. The crest rises to a high point in the centre and tails of to the edges, simulating the experimental dune geometry.

Each side of the dune is lofted individually to create two separate surfaces. The nearest side is the windward face, and as such has a much more shallow angle. The slip face has an angle of 34 degrees, the angle of repose for dry sand.

Pour points are defined along the crest of the dune. From these points a script is used to calculate the path of least resistance down the slope. This simulates how a liquid, such as the glue binder, would pass down the surface.

04

05

06

Having defined the main structural lines it is now possible to add diagonal bracing elements. These will support the main elements and reduce lateral buckling.

Using a metaball function in x, y and z gives an energy field representation of the lines defined so far. This describes an outline of the desired geometry.

These boundary curves can be divided to give a dense point cloud which accurately depicts the desired geometry. This is exported as an .asc file and meshed using a Rotating Sphere algorithim in MeshLab.


02 REALISATION

SANGUIS ET PULVIS

67 _ DESIGN DEVELOPMENT

01 _ GROUND LEVEL PLAN

The first plan developed for the building is really just an extension of the diagram created previously. It features a large main dune space which houses the fresnel lenses and Stirling generators which is open to the public and forms the main attraction space. The cow sheds and abattoir are housed in extending secondary dune arms which are created by the sand removed from the main dune through dune migration. These will form as extending fingers off the main body of the dune.

CATTLE STALLS

CATTLE STALLS CATTLE STALLS

CIRCULATION CIRCULATION

CIRCULATION VETINARY / EXAMINATION

CATTLE ENTRY

CATTLE ENTRY

ABATTOIR

ABATTOIR ABATTOIR CATTLE ENTRY

CHILLED LOADING MAINTENANCE / STORAGE

COLD STORAGE

CHILLED LOADING

COLD STORAGE

CHILLED LOADING

PUBLIC CIRCULATION PUBLIC CIRCULATION

COLD STORAGE


02 REALISATION

SANGUIS ET PULVIS

68 _ DESIGN DEVELOPMENT

01 _ ROOF PLAN

CATTLE STALLS

CATTLE STALLS CATTLE STALLS

CIRCULATION CIRCULATION

CIRCULATION VETINARY / EXAMINATION

CATTLE ENTRY

CATTLE ENTRY

ABATTOIR

ABATTOIR ABATTOIR CATTLE ENTRY

CHILLED LOADING MAINTENANCE / STORAGE

COLD STORAGE

CHILLED LOADING

COLD STORAGE

CHILLED LOADING

PUBLIC CIRCULATION PUBLIC CIRCULATION

COLD STORAGE


02 REALISATION

SANGUIS ET PULVIS

69 _ DESIGN DEVELOPMENT

01 _ SECTION

FRESNEL LENSE CONCENTRATORS

HEAT TRANSFER UNITS

CHILLED LOADING

LARGE SCALE STIRLING GENERATORS

CATTLE ENTRANCE

CHP

LINEAR ABATTOIR COLD STORAGE

MAINTAINANCE / ACCESS

COOLING SYSTEM

CATTLE SHED


02 REALISATION

SANGUIS ET PULVIS

70 _ DESIGN DEVELOPMENT

01 _ INTERNAL RENDER SOLAR CONCENTRATORS


02 REALISATION

SANGUIS ET PULVIS

71 _ DESIGN DEVELOPMENT

01 _ EXTERNAL RENDER ENTRANCE RAMP


02 REALISATION

SANGUIS ET PULVIS

72 _ DIVERTIVE COMPONENT FLOW CONTROL

01 _ UNOBSTRUCTED POUR PATH 01

The second experiment of crest pouring aims to achieve similar goals to that of drip forming. The goal is to create a nonuniform sand structure with a dynamic spatial affect. Crest pouring uses the same viscous liquid of sand and salt solution as drip forming. In this process a sand dune is used as the former however. The mixture is poured along the crest of the dune, building a reservoir along this line. As this reservoir overflows the mixture cascades down the side of the dune until they meet the ground plane. Once dried these arms form the structural elements which support the crest line. Small secondary pours are made at lower levels to link these arms and ensure the structure is laterally stable.

02 _ OBSTRUCTED POUR PATH

In previous experiments of crest pouring the liquid binder has been allowed to flow down the face of the dune unobstructed. This allows the liquid to follow whichever course it wants, typically the path of least resistance. This provides a raw and beautiful geometry however allows the designer no control other than where the pour points are located. For realisation in a real building a much greater degree of control is required.

01

Once the structure has dried and become solid, which takes around 30 minutes, the sand can be blown away using an air gun. This is an accelerated simulation of the way a real sand dune would migrate away from the solidified structure. The movement of sand reveals an increasingly large volume within the solidified structure. Some of the sand in unable to fully escape the structure on the leeward side which gives the space a specific directionality relative to the wind. The final structure mimics the overall geometry of the original dune, however its form is so different that they appear unrelated. The highly skeletal structure appears to have been dug up from within the sands of the desert, imposing a sense of timelessness.

03 _ DIFFERENT COMPONENTS FOR DIFFERENT FACADES

By placing a barrier in the path of the liquid as it pours down the dune face it can be forced to change course in a predictable manner. The more obstacles that are placed the more complex the sequence of diversions becomes. This provides a low tech solution to the problem presented: a lack of control. By manipulating the positions, shapes and scales of the barriers the liquid can be controlled to move in whatever way the designer wishes, and as such can be tailored to the needs of the building. This changes an interesting and beautiful process into a realistic building technique.

SECTION SHOWING INDIVIDUAL FACADE REQUIREMENTS

SOUTH DUNE FACE

NORTH DUNE FACE

N

EN REV

02

02

T CEN

ON RC

A SOL

P ORS RAT

T

GH T LI

CHI REA

OR

ERI

T G IN

LIG

HT

MU

ST

BE

DI

FF

US

ED

FR

OM

NO

RT

H

The structure is strong and stable but needs to be tested at a larger scale in order to be analysed correctly.

SOUTH FACADE REQUIREMENT

03

NORTH FACADE REQUIREMENT

03

The diverting component used on the south facade is required to create a clear unobstructed circular opening in which the fresnel lense can be mounted. In addition to this it must cause a thickening of the top section of the profile, as additional strength will be needed here to hold the weight of the lense frame and mechanism. This will give the face the structural path of a series of stacked arches.

As the solar concentrators will prevent any light entering the interior space, the north facade is required to diffuse light into the space. In order to do this the component must create a large flat lower surface which is able to reflect light back into the interior. This is harder to achieve than thickening the top edge as poured material will naturally move away from the bottom edge of the diverting component. A two piece ‘scoop’ design will need to be incorporated to achieve this.


02 REALISATION

SANGUIS ET PULVIS

73 _ SOUTH FACADE DIRECTIVE COMPONENT

01 _ CIRCLES CLOSE ARRAY

02 _ CIRCLES MEDIUM ARRAY

KEY FEATURE: VERY SMALL OPENINGS BETWEEN FORMING COMPONENT AND FLUID 01_ 3D MESH

03 _ CIRCLES WIDE ARRAY

KEY FEATURE: MEDIUM OPENINGS BETWEEN FORMING COMPONENT AND FLUID 01_ 3D MESH

KEY FEATURE: LARGE ELONGATED OPENINGS BETWEEN FORMING COMPONENT AND FLUID 01_ 3D MESH

The south facade requires a structural thickening on the top edge to create structural arches through its length. As such a component must be designed that creates this effect. When the material flowing down the dune surface collides with the component it will bunch and thicken on this edge provided it creates sufficient resistance to the flow of the material. As such, components appropriate for this surface will have a wide and fairly flat top surface. The most simple component that fulfils this requirement is the circle. The following tests compare the effects of different spacings of circular components on the structural geometry created by the flow around them. Whilst closer spacing creates more frequent arches and a stronger looking structure, it also creates significantly more self weight and therefore reduces the structure’s strength. A compromise must be reached between these two factors to optimise the component.

02_ POUR PHASE 1

03_ POUR PHASE 2

02_ POUR PHASE 1

03_ POUR PHASE 2

02_ POUR PHASE 1

03_ POUR PHASE 2

04_ POUR PHASE 3

05_ POUR PHASE 4

04_ POUR PHASE 3

05_ POUR PHASE 4

04_ POUR PHASE 3

05_ POUR PHASE 4


02 REALISATION

SANGUIS ET PULVIS

74 _ SOUTH FACADE DIRECTIVE COMPONENT

01 _ OVAL ARRAY

02 _ TRIANGLE ARRAY

KEY FEATURE: SMALL OPENINGS BETWEEN FORMING COMPONENT AND FLUID 01_ 3D MESH

03 _ REVERSED TRIANGLE ARRAY

KEY FEATURE: OPENINGS AND COMPONENT HAVE SIMILAR ANGLES, CREATING RECIPROCAL FORMS 01_ 3D MESH

KEY FEATURE: SQUARED OPENINGS BETWEEN FORMING COMPONENT AND FLUID 01_ 3D MESH

The south facade requires a structural thickening on the top edge to create structural arches through its length. As such a component must be designed that creates this effect. When the material flowing down the dune surface collides with the component it will bunch and thicken on this edge provided it creates sufficient resistance to the flow of the material. As such, components appropriate for this surface will have a wide and fairly flat top surface. The most simple component that fulfils this requirement is the circle. The following tests compare the effects of different spacings of circular components on the structural geometry created by the flow around them. Whilst closer spacing creates more frequent arches and a stronger looking structure, it also creates significantly more self weight and therefore reduces the structure’s strength. A compromise must be reached between these two factors to optimise the component.

02_ POUR PHASE 1

03_ POUR PHASE 2

02_ POUR PHASE 1

03_ POUR PHASE 2

02_ POUR PHASE 1

03_ POUR PHASE 2

04_ POUR PHASE 3

05_ POUR PHASE 4

04_ POUR PHASE 3

05_ POUR PHASE 4

04_ POUR PHASE 3

05_ POUR PHASE 4


02 REALISATION

SANGUIS ET PULVIS

75 _ NORTH FACADE DIRECTIVE COMPONENT

01 _ CIRCLES WITH SEMI CIRCLE SCOOP

02 _ TEARDROPS WITH QUARTER CIRCLE SCOOP

KEY FEATURE: MEDIUM THICKENING OF UPPER AND LOWER EDGE WITH INTERESTING SPILL OVER 01_ 3D MESH

03 _ OFFSET TRIANGLES WITH LINEAR SCOOP

KEY FEATURE: EXAGGERATED THICKENING OF LOWER EDGE

KEY FEATURE: VARIED THICKENING OF LOWER EDGE AT DIFFERENT ANGLES. COMPLEX OVERALL GEOMETRY.

01_ 3D MESH

01_ 3D MESH

The north facade component presents more of a challenge as it is required to create a thickened lower edge. As described previously, flowing material naturally will want to bunch and collect on the top surface and therefore thicken in the opposite manner. This can be seen in the first simulation, in which an equal thickening is observed on both edges. The component is optimised to feature a pointed top edge, splitting the material flow and preventing it from thickening at this point. The component then uses two catching elements which redirect the divided material flow towards the back edge of the component. This allows for a second collision at this point creating thickening of the bottom edge. The success of the effect can be clearly seen in test 02. This component provides the most successful result.

02_ POUR PHASE 1

03_ POUR PHASE 2

02_ POUR PHASE 1

03_ POUR PHASE 2

02_ POUR PHASE 1

03_ POUR PHASE 2

04_ POUR PHASE 3

05_ POUR PHASE 4

04_ POUR PHASE 3

05_ POUR PHASE 4

04_ POUR PHASE 3

05_ POUR PHASE 4


02 REALISATION

SANGUIS ET PULVIS

76 _ DIRECTIVE COMPONENTS TESTING

01 _ PREPARATION OF EXPERIMENT APPARATUS

02 _ POURING THE RESIN / SAND MIXTURE

03 _ CLOSE UP VIEW OF RESULTANT FORM

01

01

02

03

04

05

06

The following physical experiment aims to clarify some of the findings of the Real Flow simulations. It uses an angled base board and a series of wooden diverters. The binding material used in this case is epoxy resin. This is used as it exhibits a smooth transition from liquid to solid as it dries, unlike many other resins which remain very liquid before suddenly becoming a jelly. The findings of the experiment were largely as expected, although the apparatus has some major limitations that hinder its effectiveness. The most easily observed difference between the simulated structure and the one realised here is that in real life the binding material clings to the diverting component much more. This means that a clean structural shadow is not formed behind the components, instead it becomes partially in filled with binding material.

02

03

04


02 REALISATION

SANGUIS ET PULVIS

77 _ DIRECTIVE COMPONENTS TESTING The dried resin and sand mixture is attractive in finish as it has a fairly rough visual texture but a very smooth resin like physical texture. The voids created by the diverting components can clearly be seen and these provide easy and regular mounting points in which to fit the fresnel lenses.

02 _ RESIN / SAND POUR EXTERNAL SURFACE

01 _ RESIN / SAND POUR INTERNAL EFFECT


02 REALISATION

SANGUIS ET PULVIS

78 _ MAIN DUNE CREST POURING

01 _ STAGES OF REALFLOW CREST POURING

02 _ FINAL UNEDITED MESH FROM REALFLOW

01

04

02

05

03

06

Having tested the crest pouring simulation process in Real Flow it is now possible to pour a structure of the entire dune. This involves a much large pour area within Real Flow which generates many more particles within the simulation. This makes the simulation dramatically slower and the following pouring simulation took around 18 hours to calculate.


02 REALISATION

SANGUIS ET PULVIS

79 _ SECONDARY DUNE FORMATION

01 _ WIND PATTERNS WHICH FORM SECONDARY DUNES 01 _ VASARI WIND VELOCITY ANALYSIS

02 _ AREAS OF SAND MOVEMENT AND DEPOSITION

Creation Of High And Low Velocity Channels

Areas Of Erosion And Deposition

As the structure of the main dune interferes with the wind flow through it, high and low velocity / pressure channels are created. These channels dictate the movement and deposition patterns of the loose sand within the dune.

The areas hatched in blue are the areas of high velocity wind and therefore erosion of sand material. These areas will be similar to the ‘desert roads’ found in large desert dune fields where all of the surface sand has been blown away to leave a hard base surface.

When the main dune has been stabilised through the crest pouring technique, it will encompass the sand mass or the original dune within it. As the wind blows against the mass of sand it will slowly migrate with the wind flow, creating open space within the structure. As the structure begin to open up wind becomes able to flow through the structure and at this point the relative portions of solid and open structure begin to define the wind flows that remove the sand. Four areas on the north facade are not treated in the same component based way as the rest of the facade. Instead they are cast to become completely solid. These are the four areas in which the internal loading bay will be located and where the cow sheds are required. These solid areas of the facade prevent wind flow through portions of the facade and this causes the acceleration on wind flow in the areas between them. This can be seen clearly in figure 01. As has been described previously in the analysis of sand dune formation, sand dunes are formed when sand particles which are being carried in a fast wind flow are deposited in a neighbouring area experiencing low wind flow. This is because the particles are sucked into the area of low pressure created by the slow velocity. In the natural desert environment this is found when the wind blows against a large rock. Sand does not build up against the face of the rock as might be expected as this is the high velocity and high speed face. Instead it builds up behind the rock in the low velocity and low pressure space. This same process will happen in the formation of the buildings secondary dunes. Sand will be removed from the main dune in the high velocity channels shown in orange. This material with then be deposited in the low velocity / pressure channels shown in blue. These areas have been identified in figure 02. As the sand is removed from the main dune this process will result in the creation of long secondary dunes which are roughly perpendicular to the main dune. These dunes can then be solidified through the crest pouring technique to create the cow sheds and abattoir entrance.

The areas hatched in red are the areas of low velocity wind and therefore deposition of sand material. Secondary dunes will form in these areas as sand deposits in the low pressure pocket behind the solid structural elements. As these dunes grow, more sand will become deposited in them as when migrating sand grains meet a dune their kinetic energy is absorbed and they become part of the dune mass. This is why sand separates into areas of huge dunes and barren roads in the desert environment.


02 REALISATION

SANGUIS ET PULVIS

80 _ SECONDARY DUNE FORMATION

01 _ SECONDARY DUNE FORMATION TIME BASED SECTION PHASE 01

PHASE 02

Immediately After Crest Pouring

5 Months After Crest Pouring

Immediately after the main dune has been solidified the wind flow is forced over the geometry of the dune as it did prior to solidification. In this condition sand material is picked up by the wind along the windward face where wind velocity his high. When the wind carried material clears the crest of the dune it is deposited in the low pressure zone behind the leeward face. This begins the process of secondary dune formation. The removal of sand from the dune will be at its slowest at this phase as the wind can not yet penetrate the structure of the dune.

As material is removed from the top of the dune the wind is allowed to penetrate the interior of the structure. This drastically accelerates the rate at which material is removed from the main dune as the wind accelerates through the openings on the windward face and carries material out of the openings on the leeward side. This material is deposited in the low pressure zone behind the solid part of the structure. This forms a secondary dune which extends out behind the main dune. As was illustrated in the previous wind analysis plan, channels of high velocity wind exist between the solid structural sections. Material removed in these channels will also be deposited in the low pressure zone behind the solid structure accelerating its growth.

Having established how the wind flow patterns through and around the building setup the creation of secondary dunes in plan, the following diagrams illustrate how the secondary dunes are formed in section. The formation of secondary dunes is a slow time based process. I estimate that it will take around 10 months for the secondary dunes to be created naturally by the wind. This is based on the dune migration speed of the Northern Great Sand Sea which is around 70m / year. The migration distance required in this project is significantly less than 70m but the migration rate may be slowed by the interferences of the structure with saltating sand grains. This scale of the affect the structure will have on migration could be simulated to an extent through fluid dynamics modelling but this is exceptionally complex and the results would still be highly unreliable. Large scale testing would be required to form an accurate prediction of migration speed. As the wind flows over the outer surface of the dune, sand material is picked up in the current. This material is deposited in the low pressure zone behind the dune. As more material is removed from the dune the wind will be able to penetrate the structure and this will speed sand removal significantly. Wind will be channelled through the structure removing sand through openings in the leeward face. As this material is removed it is deposited in the low pressure zone, extending the secondary dune. Once this process has been completed the internal volume of the main dune will be nearly empty. The secondary dune will be fully formed from the removed material with the interior element sculpted into a steep windward slope by the wind flow. These secondary dunes can now be solidified to form the structure for the cow sheds and abattoir. The interior of the main dune can now be dug down in order to construct the element of the building which sit below ground level.

PHASE 03

34 o

34 o

10 Months After Crest Pouring As wind blows through the structure the interior portion of the secondary dune is sculpted into a steep windward face. This is the steepest angle that sand will form to in a stable manner and becomes this steep because of the solid structure behind it preventing its migration. The secondary dune is now fully formed with all of the material from the main dune having been removed and redeposited in the low pressure secondary zones. The secondary dunes can now be crest poured to form the structure of the cow sheds and abattoir. Now that the main dune interior space has been emptied, digging can begin to construct the volumes which sit below ground level.


02 REALISATION

SANGUIS ET PULVIS

81 _ DESIGN PLAN

01 _ PLAN

This plan shows a slightly more developed stage than the previous diagrammatic plan. The spatial and programmatic allocations have now been established as well as key circulation routes. This plan functions well as a diagram and can now be worked up in more detail

CATTLE TRANSFER CATTLE STALLS

CATTLE STALLS

VET. / EXAM.

ABATTOIR

CATTLE TRANSFER

BRICK LOAD

VET. / EXAM. ABATTOIR CATTLE STALLS BRICK LOAD

CHILLED LOADING

CHILLED LOADING

CATTLE TRANSFER CATTLE STALLS VET. / EXAM.

OFFICE

OFFICE PUBLIC CIRCULATION

ABATTOIR

LARGE SCALE STIRLING GENERATORS

BRICK LOAD

VET. / EXAM.

CHILLED LOADING OFFICE

CHILLED LOADING OFFICE

LARGE SCALE STIRLING GENERATORS

ABATTOIR


02 REALISATION

SANGUIS ET PULVIS

82 _ DESIGN ROOF PLAN The main development at this stage of the project is the formation of the secondary dunes. From the wind analysis conducted previously the shape of these dunes can be predicted, but this is a fairly inexact process. Many components are used in the solidification of these secondary dunes to provide the maximum ventilation possible but without sacrificing too much shade.

01 _ ROOF PLAN


02 REALISATION

SANGUIS ET PULVIS

83 _ DESIGN SECTIONS

01 _ SECTION A-A

In section the primary relationship remains that of the projects cooled spaces sitting around each other to maintain an efficient cool pocket. This has been achieved successfully in the current plans, with the cold storage, loading bays, abattoir and offices all sharing walls or floors with the cold storage area.

02 _ SECTION B-B

FRESNEL LENSE CONCENTRATORS HEAT TRANSFER UNITS BLOOD BRICK LOADING LARGE SCALE STIRLING GENERATORS

BLOOD / CHEMICAL STORAGE MAINT./ ACCESS

BLOOD BRICK CASTING

BLOOD BRICK STORAGE


02 REALISATION

SANGUIS ET PULVIS

84 _ RECLAMATION OF SIWA Whilst Siwa does not experience dangerous levels of dune migration that would threaten the existence of the town, over the last century much of the towns southern agricultural land has been lost to the Great Sand Sea.

01 _ LONG TERM EVOLUTION OF SIWA 01 _ EXISTING CONDITION

03 _ PHASE 2

As can be seen from the aerial image of the existing condition, large areas have been overrun by sand and most of the peripheral areas area becoming thinned by the incursion of sand. This loss of agricultural land has reduced the towns output of dates and olives harming its economy. It has also reduced the potential grazing lands required to manage large herds of livestock, which is necessary for the proposed program. The proposed buildings will solidify the large dunes which sit to the south of Siwa at the edge of the Great Sand Sea. This will prevent the migration of sands towards the town from the South and enable the reclamation of lost agricultural land. As the full proposal is built phases by phase, increasing amounts of the southern periphery will be solidified, gradually increasing the amount of land that will become suitable for farming as the project expands. This gained area will provide the grazing lands required by the cattle who inhabit the buildings and provide the raw building material for the community. This will change the identity of Siwa greatly. The town has always been nuclear, focusing around the old town settlement of Shali. From Shali the agricultural fields have spread outwards until they lose their battle with the encroaching sands. The establishment of this southern barrier now gives the town a very bold outer definition, shifting focus from the central area. This opens up the potential for new extended development in the area between and releases the town from its stifling centrality.

PREVIOUSLY AGRICULTURAL LAND COVERED BY THE WESTERN DESERT

02 _ PHASE 1

04 _ PHASE 3


02 REALISATION

SANGUIS ET PULVIS

85 _ CONSTRUCTION SEQUENCE PART 1

01 _ AERIAL VIEW OF CONSTRUCTION SEQUENCE 01

02

03

04

ORIGINAL DUNE TO BE POURED

PLACEMENT OF STEEL DIVERTING COMPONENTS

CREST POURING OF THE DUNE WITH BLOOD GLUE / SAND MIXTURE

SOLIDIFICATION OF BLOOD / GLUE AND COMPONENT REMOVAL

ORIGINAL DUNE TO BE POURED

ORIGINAL DUNE TO BE POURED

ORIGINAL DUNE TO BE POURED

ORIGINAL DUNE TO BE POURED

The construction process used for the project is completed over a long period of time and in multiple stages. As such it is important to understand it as a time based process, in much the same way as the design for the Time Pavilion in brief 2. The following timed based construction perspectives illustrate the construction process as the building evolves from dune to structure.


02 REALISATION

SANGUIS ET PULVIS

86 _ CONSTRUCTION SEQUENCE PART 2

02 _ AERIAL VIEW OF CONSTRUCTION SEQUENCE 05

06

07

08

MIGRATION OF DUNE AND EARLY FORMATION OF SECONDARY DUNES

FULLY FORMED SECONDARY DUNES

CREST POURING OF SECONDARY DUNES AND SHAPING OF INTERIOR FLOOR

INSTALLATION OF FUNCTIONAL COMPONENTS

ORIGINAL DUNE TO BE POURED

ORIGINAL DUNE TO BE POURED

ORIGINAL DUNE TO BE POURED

ORIGINAL DUNE TO BE POURED


02 REALISATION

87 _ PLAN LEVEL 00

SANGUIS ET PULVIS


02 REALISATION

88 _ SECTION AA

SANGUIS ET PULVIS



02 REALISATION

SANGUIS ET PULVIS

90 _ SEGMENTED DUNE MIGRATION TESTING The project deals with dune solidification on a very large scale, however uses a material which can only be accumulated at a rate related to the size of livestock herds in the town. This means that it will not be possible to produce sufficient structural material to solidify the whole dune in one pour. As such, the dune will instead be solidified segment by segment, with sufficient time in between pours to allow for the collection of blood. This creates an interesting opportunity to fully engage with the time based nature of the construction process as during the waiting period the dune will migrate, segmenting its body.

01 _ REMOVING SAND FORMER 01 _ SAND DUNE

02 _ DUNE MIGRATION ANALYSIS 04 _ DUNE MIGRATION PHASE 01

01 _ ORIGINAL DUNE OUTLINE

The original dune was formed in a standard transverse shape. In a dune of this type the geometry is fairly regular along its length and therefore the action of wind against it is fairly uniform. Slight differences will of course occur creating moderate high and low velocity channels. This is what gives the dunes their characteristic wave shape. No large scale deformation of the standard geometry is created however. One end of the dune was solidified using a binder of PVA glue. This suddenly creates an irregularity in terms of the way the wind is able to act on the dune. This irregularity will cause the dune to migrate in a nonuniform manner.

Normally a dune would be expected to migrate at a fairly even rate along its length. This is because dunes tend to be unobstructed and so the wind is able to act on the whole dune equally. However, once a section of the dune has been solidified this effects the wind flow over the dune in this area. To fully understand the effect this obstruction has on the dune’s migration, and therefore the geometry of the building, physical tests were carried out on a small test dune solidified in part using PVA glue.

02 _ APPLICATION OF BINDER

05 _ DUNE MIGRATION PHASE 02

02 _ NEW DUNE OUTLINE

As can be clearly seen from the outline of the created dune, an acceleration in migration has occurred on the end of the dune which collides with the solidified structure. This occurs because the solidified structure prevents the material with it from migrating as quickly as the rest of the dune. This creates an offset between the main dune body and the solidified end section. This offset essentially creates a wall which catches the wind and accelerates it in this region. It also channels the wind into this specific location. This causes an increase in the speed at which sand is moved locally and this creates an extended tail on the leeward side of the dune.

03 _ SOLIDIFICATION OF BINDER

06 _ DUNE MIGRATION PHASE 03

This local acceleration of the wind also causes the rapid removal of sand from the solidified section once an offset has been created. Whilst the dune is aligned the wind can not penetrate the solid section, but the creation of the offset allows the wind to penetrate the open side of the solidified section and blow material out of the openings on the leeward side. This equally allows the wind the reach the ground behind the solidified section preventing the build up of sand in this area. Instead sand that would normally settle in this area amalgamates with the extending tail, further extending it away from the main mass of the dune. Accurate dune migration modelling is very hard to carry out due to the scale of the experiment behaving very differently to real world scale. When this experiment is left for longer the sand is all blown quickly away. At full scale however I predict that the tail seen forming here would become elongated further as material is added to it from the draining solidified section.


02 REALISATION

SANGUIS ET PULVIS

91 _ SEGMENTED DUNE MIGRATION TESTING In these detailed images of the solidified section the impact of the solid segment on wind flow can clearly be seen. The solidified section acts as a solid wind diverter causing flow to be channelled through the are in front of it. This causes accelerated migration in this area due to accelerated wind flow, and creates an extending arm at one end of the dune. The longer this process is allowed to continue the longer this arm becomes under the newly disrupted wind flow.

01 _ THE DUNE OFFSET WHICH AFFECTS AIR FLOW

02 _ THE EXTENDING TAIL WHICH FORMS BEHIND THE SOLIDIFIED SECTION


02 REALISATION

SANGUIS ET PULVIS

92 _ SECONDARY DUNE FORMATION

01 _ SECONDARY DUNE TAIL FORMATION 01 _ SMALL TAIL CREATED BY SOLIDIFIED END SECTION

From the physical tests conducted on segmented dune migration it is clear that when a segment of the dune is solidified it has the effect of concentrating wind flow along its edge . This is because the solid wall of the solidified segment essentially acts as a wind catcher, channelling and accelerating the wind. This increased wind flow causes an extending tail to begin forming as it causes accelerated sand movement. As further segments are solidified the wind will be channelled by each one, causing the dune to be separated into multiple smaller dunes acting under localised wind flows. When these dunes become split the way wind acts upon them becomes very different. Previously the dune was long and perpendicular to the wind direction making it a transverse dune. However, as is illustrated in the opposite diagrams, when the dune becomes split suddenly the wind is acting parallel with it and blows slightly differently on each side. This converts the dune to a transverse type, and as such it will migrate in a transverse manner. This means that the dune will become slowly extended into a long thin form as material is blown up and along its crest. In this way, secondary dunes will be created forming through the back edge of the original dune, and they will be orientated parallel to the wind direction.

02 _ SIMILAR LINEAR SECONDARY DUNE FORMATION OCCURRING BEHIND STABILISED DUNES IN THE ALGERIAN SAHARA

WIND FLOW

WIND FLOW

02 _ WIND FLOW BECOMES SPLIT BY SOLIDIFIED SECTIONS EXTENDING TAIL

This secondary dune formation effect can be seen in natural landscapes also. The image opposite shows a secondary dune forming perpendicular to a large transverse dune. In this case the transverse dune has become stabilised due to a long period of slow movement. This stabilised dune now acts as a wind diverter, splitting the wind into separate flows. This causes sand to accumulate in the wind shadow formed and elongate into a secondary dune. This is the exact same process that will occur with the building as it is formed in segments.

WIND FLOW

WIND FLOW

WIND FLOW 03 _ LINEAR SECONDARY DUNES FORMED BY CONVERGING WIND FLOWS

WIND FLOW

WIND FLOW


02 REALISATION

SANGUIS ET PULVIS

93 _ SEGMENTED DUNE MIGRATION

01 _ TIME BASED PLAN OF SEGMENTED MIGRATION (PLAN CUT AT + 0.00 M) 01 _ DUNE STAGE 01

The following diagrams describe in full the formation process of the building. As each segment is solidified, the wind flow acting upon the dune is effected and disturbed. This causes the acceleration of the wind along certain channels which causes the segmentation of the dune into multiple elongating linear dunes. The actual process of dune segmentation is extremely complex, and it should be noted that this is an approximation of a natural process. It is almost impossible to simulate the migration pattern that would occur with this process as it depends upon the complex interactions of billions of sand grains acting under a constantly evolving wind pattern. The only way to simulate the process would be to conduct large scale and long time span testing in the desert to test how real dunes would migrate.

WIND FLOW

02 _ DUNE STAGE 02

WIND FLOW

03 _ DUNE STAGE 03

WIND FLOW

04 _ DUNE STAGE 04

WIND FLOW

05 _ DUNE STAGE 05

WIND FLOW

06 _ DUNE STAGE 06

WIND FLOW

07 _ DUNE STAGE 07

WIND FLOW


02 REALISATION

SANGUIS ET PULVIS

94 _ TIME BASED PLANS As the building forms over a long time period its plan is only realised segment by segment. As such, each segment of the building contains all the necessary programmatic elements to operate fully without the other segments. In this way the building can be used in the most efficient manner throughout the construction process. Accordingly, each segment consists of solar concentration lenses, Stirling generators, cow sheds, abattoir and brick line. The first section does not feature cow sheds, an abattoir or brick line. This is because these functions are housed in the secondary dunes, however the first section does not create a secondary dune. This means there is not sufficient space to house these functions within the limited first section. Upon the completion of this section the building will commence power generation but cattle will be housed temporarily in sheds in the old town of Siwa. Cattle will then be transferred after the completion of the second section. As the building progresses, the village which will build up around it gradually grows. Initially, only enough buildings will be required to house the limited construction team, however as internal excavations commence, power generation begins, and the abattoir and brick line become active, the population of the village working in the building will grow dramatically. In addition to this, it is likely that the building will become a large scale tourist attraction. As the building and surround village grows, it will also develop hotels and other tourist related infrastructure.

01 _ DUNE STAGE 01

01 _ DUNE STAGE 01

01 _ DUNE STAGE 01

01 _ DUNE STAGE 01


02 REALISATION

9501 _ DUNE STAGE 01

SANGUIS ET PULVIS 01 _ DUNE STAGE 01

01 _ DUNE STAGE 01

01 _ DUNE STAGE 01

01 _ DUNE STAGE 01


02 REALISATION

SANGUIS ET PULVIS

96 _ FLUID DYNAMICS SIMULATION

01 _ SEGMENTED FLUID DYNAMICS MODELLING 01 _ SEGMENT 01

The design uses a complex construction technique based around the liquid flow of a binding material over sand. This is not possible to model using conventional techniques as the outcome is not predefined. Instead, the layout of diverting components and the geometry of the dune is predefined and the form of the building is allowed to emerge from the interaction of the binder with this condition. The construction technique itself is a generative process and vital to the design. In order to model the building this process must be copied in a computational manner, and this involves fluid dynamics simulation. Fluid dynamics simulators specialise in animating the behaviour of liquids acting in complex ways. They can powerfully and accurately simulate the flow of liquids and offer a high degree of control over the liquid properties such as surface tension, viscosity and adhesion. The most widely used fluid dynamics simulator is called Real Flow and this is the software I will use to ‘pour’ my building computationally. First the basic dune is created, and vitally joined with a base plane ensuring that no holes or naked edges exist through which liquid could leak. Diverting components are then modelled across the surface of the first segment using Grasshopper and Rhino. A fluid emitter is then created following the crest line of the dune, and is activate to pour the binding material. This material is set with a very high viscosity and surface tension to simulate the properties of a thick glue like liquid. The program animates the particle cloud by single frames and can be stopped when it reaches the desired resolution. The particle cloud can then be meshed using Real Flow’s powerful marching cubes based meshing algorithm. The dune is then replaced by the phase 2 dune which has migrated onwards. The mesh created in segment 1 is loaded into Real Flow as a collision object, meaning that when segment 2 is poured it will interact not only with the dune and components but also segment 1. This allows for the complex intersections of the segments which are vital to the design. This process is repeated for all 5 segments of the design, exporting each segment’s mesh to Rhino for amalgamation. The meshes are formed in immense detail and require significant reduction in Rhino. On average each original mesh consists of around 800,000 mesh faces, however this can be reduced to around 50,000 without significant loss of quality.

02 _ SEGMENT 02

03 _ SEGMENT 03

04 _ SEGMENT 04

05 _ SEGMENT 05


02 REALISATION

SANGUIS ET PULVIS

97 _ CONSTUCTION OVER TIME Sanguis et Pulvis engages with time in terms of both its construction and habitation. As the building progresses and extends over time, so too does the program and surrounding community. Indeed, even after the completion of the building, construction simply moves to the next dune in the continuing barrier across the desert. As such the building has no definitive end, it is constantly evolving a changing. After the completion of the main structures the interior spaces and excavations will constantly develop too as sand is blown through the building and they are re-dug.


02 REALISATION

SANGUIS ET PULVIS

98 _ ROOF PLAN

KEY PLAN

100m

75m

50m

40m

30m

20m

10m

0m


02 REALISATION

SANGUIS ET PULVIS

99 _ GROUND LEVEL PLAN

KEY PLAN

100m

75m

50m

40m

30m

20m

10m

0m


02 REALISATION

SANGUIS ET PULVIS

100_ GROUND LEVEL PLAN

0.00m

0.00m

0.00m 0.00m

-3.75m

0.00m

-1.25m

-2.50m

0.00m

-3.75m

-1.25m 0.00m

KEY PLAN

-2.50m -3.75m

0.00m -1.25m

-2.50m 0.00m

-3.75m

-1.25m -2.50m

100m

75m

50m

40m

30m

20m

10m

0m


02 REALISATION

SANGUIS ET PULVIS

101_ BASEMENT PLAN

KEY PLAN

100m

75m

50m

40m

30m

20m

10m

0m


02 REALISATION

102_ SECTION AA

SANGUIS ET PULVIS


02 REALISATION

103_ SECTION BB

SANGUIS ET PULVIS


02 REALISATION

SANGUIS ET PULVIS

104_ TIME BASED CONSTRUCTION PERSPECTIVES PHASE 01

00 MONTHS

A suitable dune is selected in the Great Sand Sea near the southern edge of Siwa. The dune should be sufficiently large to house the required program, but small enough to be viably solidified.

PHASE 02

0.5 MONTHS

PHASE 03 01 MONTH

Diverting components formed from thin steel sheet are arrayed across the surface of the dune. Only a small section of the windward tail of the dune is covered, representing the area pourable with the amount of blood stored.

Large storage tanks of blood / sand mixture are transported to the dune from the temporary abattoirs in Siwa. This mixture is poured over the crest of the dune using a concrete pump and allowed to flow down its surface.

A small settlement is created in the shadow of the dune to house workers during this limited phase of construction. At this stage accommodation for around 10 -15 people would be required.

As the material flows over the surface it will sink through the loose sand, thickening the section. The diverting components control the flow of material to create openings for fresnel lenses to be mounted in. On the north face openings are created to allow diffuse lighting and ventilation.


02 REALISATION

SANGUIS ET PULVIS

105_ TIME BASED CONSTRUCTION PERSPECTIVES PHASE 04

06 MONTHS

The poured material is coagulated by the heat of the desert sun, creating a solid structural mass. As the desert winds act upon the dune, the unsolidified sand continues to migrate, segmenting the dune. This new segment of dune is prepared as before with diverting components and is crest poured with blood / sand mix. The time gap between phases 3 and 4 allows for the storage of sufficient blood to carry out this pour. As the dune migrates the interior volume of the first section becomes evacuated. As this happens the interior levels can be dug upon which the generators are mounted.

PHASE 05 12 MONTHS

PHASE 06 18 MONTHS

The solidified section of dune has the effect of concentrating wind flow on the eastern edge of the loose dune. This causes accelerated migration along this edge and creates a secondary dune extending perpendicular to the original. This secondary dune will house the cow sheds and abattoir. The secondary dune and the third section can both be prepared and cast at the phase.

Having solidified the secondary dune, digging can commence on the underground sections of the building. This includes the abattoir and brick manufacturing facilities. A channel is dug beneath the secondary dune to a depth of 4m, and brick tunnels are then built along this channel using blood bricks. The channel is calculated to the angle of repose for sand, meaning that no sheeting or retaining walls are required.

Fresnel lenses and generators have now been installed in the first section allowing for the commencement of electricity production. This will require a much larger work force to be on site 24hrs and thus the community surrounding the building grows accordingly.

By this time the fourth section of the main dune is being prepared for pouring as the building nears its final form. The installation of generating equipment to the second section again expands the requirements of the associated village.


02 REALISATION

SANGUIS ET PULVIS

106_ TIME BASED CONSTRUCTION PERSPECTIVES PHASE 07

22 MONTHS

Once the tunnels have been completed the channel is back filled with the dug sand. This recreates a flat ground level above the tunnels. Access is maintained via a ramp at either end as well as stairs from the central circulation space.

PHASE 08 25 MONTHS

PHASE 09 27 MONTHS

As the final secondary dune is prepared for solidification the building reaches its final form. The following migration sections will be too small to be worth casting.

The completed building houses 5 electricity generation zones, 4 cow sheds, 4 abattoirs and 4 brick manufacturing lines. The electrical output of the building now forms the communities primary economic base. The meat from the abattoirs makes them more than self sufficient. The hugely enlarged cattle herds can be grazed on the reclaimed agricultural land that results from the buildings sand shadow. The blood stored from the abattoirs is used to create the next building in the continuing barrier.

At this stage all but the final generation phase and cow sheds are finished, meaning that a large number of workers are ever present on the site maintaining the generating machinery, cleaning lenses, redigging sand as it moves, operating the abattoir and tending the live animals. The growing village swells to house this growing workforce. As tourist begin to be attracted to the building, hotels and other amenities will begin to emerge in the village.


02 REALISATION

SANGUIS ET PULVIS

107_ SECTIONAL PHYSICAL MODEL 1:50 The sectional model is constructed using the same technique that is proposed for the building. A dune is formed over a base, diverting components are arrayed over it, and a binding agent is crest poured to solidify the dune. In this case the binding material is not blood, as I have found that blood cracks during drying at this scale as it is very thin. Instead a polyurethane varnish is used as the binding agent. This behaves in a similar way to epoxy resin and will set right through to the centre as blood would.

01 _ THE PROCESS

02 _ AFTER FULL CURING OF BOTH SEGMENTS

01 _ MDF AND CLAY BASE

04 _ STRUCTURAL SEGMENT 01

02 _ CREATION OF DUNE 01

05 _ CREATION OF DUNE 02

03 _ CREST POURING DUNE 01

06 _ CREST POURING DUNE 02

Once the first dune segment has been cast, the sand is drained from the box and the dune is re-poured further along the base. This simulates the migration of the dune over a 6 month period and allows for the pouring of a second intersecting segment. The second segment is again arrayed with diverting components and the crest poured with binding agent. The liquid material collides with the first segment as it flows down the dune surface, and passes through the openings in it to reach the ground plane. Once this second segment has cured the sand can again be drained and the complete interior spaces revealed.

01 _ THE FINISHED INTERSECTING SEGMENTS


02 REALISATION

SANGUIS ET PULVIS

108_ SECTIONAL PHYSICAL MODEL 1:50 The finished physical model gives a fairly accurate representation of the building’s aesthetic and atmosphere. As it is made using the same process, many of the distinctive characteristics would remain. The most notable of these is the very rough and sandy interior surface of the structure. This forms because the binding material soaks down through the sand as it sets, meaning the inner most surface will always consist of loosely bonded sand. The interior space shown here is significantly smaller than it should be according to the designs. This is because the binding material penetrated the sand much more than expected and therefore created a shell that is much thicker than necessary. Because of this the interior spaces feel a little more restrictive than those produced through fluid dynamics simulation. This modelling inaccuracy is easy to correct through a slight modification to the construction process.


02 REALISATION

SANGUIS ET PULVIS

109_ SECTIONAL PHYSICAL MODEL 1:50


02 REALISATION

SANGUIS ET PULVIS

110_ EXTERIOR RENDER - THE BLOOD VILLAGE

KEY PLAN


02 REALISATION

SANGUIS ET PULVIS

111_ INTERIOR RENDER - CENTRAL DUNE SPACE

KEY PLAN


02 REALISATION

SANGUIS ET PULVIS

112_ INTERIOR RENDER - BRICK LINE

KEY PLAN


02 REALISATION

SANGUIS ET PULVIS

113_ INTERIOR RENDER - ABATTOIR

KEY PLAN


02 REALISATION

SANGUIS ET PULVIS

114_ INTERIOR RENDER - GENERATOR PLATFORMS

KEY PLAN


Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.