There is a long history of dreams of creating large enclosures to ameliorate the climate in inhospitable parts of the world. This trend started in the 18t h century when the owners of fashionable country houses built heated glazed enclosures to grow Pineapples and grapes. These stoves or conservatories were generally built of stone with the glass in conventional wooden frames. As the glass making technology improved and iron working techniques developed the greenhouses became larger. All iron glass houses appeared at the start of the 19th century. There is a classic example remaining at Bicton in Devon. This structure is more like a shell with small panes of glass set in fine wrought iron ribs. The classic examples of this type of construction were the Palm house at Kew built in 1848 and of course the Crystal Palace for the 1851 exhibition. The Crystal Palace was a relatively simple modular construction based on an 8 foot module for the floors, walls and roof and a 24 foot grid for the columns and introduced the idea of factory construction. Iron and glass structures were used for the great 19th century railway stations and increasingly for winter gardens or pleasure palaces. The largest span was 60m at St Pancras station. In the 1950s Buckminster Fuller was working on developing larger and larger geodesic domes. Climatron at St Louis was based on his ideas. One of his futuristic ideas was to build a big geodesic dome over Manhattan. To quote his reasoning at that time.
149 LARGE ENVIRONMENTAL ENCLOSURES, THE ROOF OF THE MILLENNIUM DOME Ian Liddell CBE, FREng, MA, DIC, FIStructE, MICE. Partner of Buro Happold and Visiting Professor of Engineering Design, University of Cambridge ABSTRACT INTRODUCTION This paper describes how ideas for enclosing very large areas have been around for some time though without being brought to completion. The Millennium dome is the first structure to be in this category. The paper goes on to describe the engineering and construction of the Dome. Fig 1 A Victorian Winter Garden There is a long history of dreams of creating large enclosures to ameliorate the climate in inhospitable parts of the world. This trend started in the 18 th century when the owners of fashionable country houses built heated glazed enclosures to grow Pineapples and grapes. These stoves or conservatories were generally built of stone with the glass in conventional wooden frames. As the glass making technology improved and iron working techniques developed the greenhouses became larger. All iron glass houses appeared at the start of the 19 th century. There is a classic example remaining at Bicton in Devon. This structure is more like a shell with small panes of glass set in fine wrought iron ribs. The classic examples of this type of construction were the Palm house at Kew built in 1848 and of course the Crystal Palace for the 1851 exhibition. The Crystal Palace was a relatively simple modular construction based on an 8 foot module for the floors, walls and roof and a 24 foot grid for the columns and introduced the idea of factory construction. Iron and glass structures were used for the great 19 th century railway stations and increasingly for winter gardens or pleasure palaces. The largest span was 60m at St Pancras station. In the 1950s Buckminster Fuller was working on developing larger and larger geodesic domes. Climatron at St Louis was based on his ideas. One of his futuristic ideas was to build a big geodesic dome over Manhattan. To quote his reasoning at that time. "The way the consumption curves are going in many of our big cities it is clear that we are running out of energy. Therefor it is important for our government to know if there are better ways of enclosing space in terms of material, time, and energy. If there are better ways society needs to know them. Domed cities can be illuminated by daylight without direct sunlight. That part of the dome through which the sun does not shine directly would be transparent. In summer the dome would be protected by polarised glass; during the sunny hours it would not hold heat but in winter the sun would penetrate all the dome. The atmosphere will be dust free. Controlling the environment through domes offers the enormous advantages of the extroversion of privacy and the introversion of the community" (Reference 1). Iron and glass remained the preferred materials for large environmental enclosures and are still frequently used today. Today the glass technology has developed with large panels of toughened and laminated glass supported on ever more daring steel structures but the spans and the scale of the enclosures has not increased significantly. In thel950s new translucent polymers and plastics became available. One development in particular was coated fabrics which offered new freedoms in form and span for large enclosures. Starting in the mid 1940s Walter Bird developed air inflated structures initially for radomes but later for tennis halls and large sports halls. The main translucent fabric materials developed at this time were PVC coated polyester and later in the 1970s PTFE coated glass fibre cloth. The big breakthrough for very large covered areas was Walter Bird's low profile* cable dome (Reference 2). Fig 3 Walter Bird's Cable Dome for spans greater than 300m This was adapted by David Geiger for the US pavilion for the Osaka expo in 1970. The principle was then used in a reduced cost form for several large football stadia. Unfortunately these structures had large valleys along the cable lines which collected snow and caused local ponding occasionally leading to loss of pressure and deflation to a stable down-hanging position. These deflations caused unacceptable co-lateral damage to the fabric leading to the abandonment of the form. At the time the air supported structure seemed to point the way towards the city scale environmental enclosure envisioned by Buckminster Fuller and concepts were put forward by Walter Bird. Fig 4 WUS Pavilion at Osaka In 1970 Frei Otto and his colleagues at the IL put forward a scheme for a covered city in the arctic. The covering was to be an air-supported fabric structure 2km in diameter with an area of 3,000,000 m 2 . (Ref 3) The primary structure was to be a net of "Trevira" Polyester ropes. Fig 5 Walter Bird's Cable Dome for spans greater than 300m I helped Peter Rice with some calculations of the forces under wind load. The effects of snow were happily glossed over with the assumption that because of the smooth shape of the roof the wind flow conditions would be close to potential flow where the wind would sweep the roof clear of snow. The provision of building services and the management of an enclosure of this scale were not considered in great detail at that time. Ten years later in 1980 we in Buro Happold had the opportunity to undertake a feasibility study for covering a town in Northern Alberta. The leader of this design team was a Canadian architect called Arne Fullerton and again we worked with Frei Otto. The study was exceptionally interesting in that it included the human response of living in such a space as well as the servicing requirements and other considerations such as the impact of fire. One of the designs for this enclosure was a 150,000m 2 air-supported roof that was to have steel strand cables at 0.5m spacing and would use ETFE foil cladding. This design was taken to concept stage and was supported by calculations and reports. This time the snow loading was not glossed over but we did not know how the building would respond to the extreme snow falls for the area. Fig 6 58°N 18 Ha Air Supported Structures Experience with large air-supported roofs in North America had demonstrated how snow on such roofs could initiate ponding and cause severe maintenance problems. (Ref 4). Our roof would have had a much lower rise to the foil cladding and hence not such deep valleys to initiate snow drifting. The cushions would also have had higher pressures to support the snow. Even so there would have been a considerable risk of problems with snow. Fig 7 Chelsea and Westminster Hospital Atrium, part clear and part tinted ETFE for cushions There was no experience with the use of ETFE foil as a cladding material at that time. 10 years later in 1990 we engineered a roof for the atrium of the Chelsea and Westminster hospital with this material. Subsequently we designed a roof for a tennis hall which consisted of foil cushions on a tensioned cable structure. Now 20 years on we have built the Dome on the Greenwich peninsular for the Millennium Experience which is to be held in the year 2000. This roof is 80,000 m 2 and is the nearest structure yet to these dreams of covered urban environments. It will enable us to evaluate the performance of such a space. Fig 8 Eastleigh Tennis Centre white ETFE foil cushions on cables PRINCIPLES OF THE DOME ROOF STRUCTURE The structural concept for the roof is based on the innovatory principle of using straight tensioned cables and flat fabric for the structure rather than adopt doubly curved surfaces which had become the accepted form for such structures. In 1994 Buro Happold pointed out that there was considerable advantage in using straight tensioned cables which could carry both the uplift and down loads with resultant forces in the same direction. (Ref 5) The pretension stiffens the cables against deflection allowing high strength tensile materials to be used to create very large spans. Flexible fabric or foil can be used as cladding between the cables provided it is pre-stressed in the same way as the cables. The advantages of the arrangement are: Compared with a two-way cable net one set of cables is eliminated along with the cross clamps and terminations. Whether the load is upward or downward the cable tensions are in the same direction which can be a great advantage if the tensions are resisted by a funicular arch or ring beam. Connections to the foil or fabric cladding can be greatly simplified. It was claimed that taken together these benefits would result in very economical large span roof structures. However the stressed fabric resists local loads by relatively large deflections rather than by simply increases in stresses. Concentrated snow drifts could possibly create a deflection which would be so great that water did not drain out from it. This would be potentially disastrous. The key to adopting this concept is to develop a form where there is adequate drainage to avoid ponding problems and details which will allow for the deflections. Wind stimulated dynamic oscillations are not a problem provided the fabric is fully tensioned as there is a high degree of damping from the fabric and the attached air. 152 • Fig 9 RSSB Tent, the fabric and cables can be seen responding to wind from the right In 1994 opportunities arose to utilise this concept on two structures, the Eastleigh tennis centre and a very large demountable tent of 20,000 m 2 for RSSB. The latter proved to be extremely economical and met the owner's requirement of ease of installation. THE DOME ROOF DESIGN The structural concept for the Roof of the Millennium Dome is apparently very simple. 72 tensioned steel stringer cables in pairs of 032mm steel spiral strand are arranged radially on the surface. The stringers are supported at a radial spacing of between 25 and 30m by an arrangement of upper hanger and lower tie-down cables that are arranged around the 12 100m tall primary steelwork masts. Circumferential cables keep the stingers on their radial lines. Fig 10 Dome Primary Structure The forces in the radial stringer cables are taken by a central 30mm diameter cable ring supported by forestay cables, which run through to the centre point, where they support a flying mast, which in turn supports additional radial cables which carry the cladding. The central area is formed by a cable truss connected to a 30m diameter cable ring. The level of stress within the cable ring leads to a stiff structural form. Between the cables, tensioned coated fabric is used as cladding. Both the tensioned cables and cladding carry the loads by deflection accompanied by increase in tension. This concept is simple but there are dangers associated with the deflections particularly ponding caused by snow or heavy rain. When loaded by wind or snow, the upper hanger, the lower tie-down and the stringer cables carry the loads from the fabric down to the ground. The stringer cables are restrained at the perimeter by the perimeter masts and large boundary cables attached to 24 anchor points. The vertical components on the forces at these points are resisted by ground anchors grouted into the London clay and the horizontal forces are resisted by a compression ring beam under the external wall. Tension structures rely on the shape of the stressed surface for their performance under load. Forces are resisted by the tension and the curvature, the greater the curvature the less the tension required to resist a given load. Both the radial stringer cables and the fabric are prestressed with sufficient tension to stiffen them against imposed load deflections. The prestress levels and cable geometry were selected to provide adequate deflection control, with the materials and structural sizes selected to provide high stiffness. Fig 11 Ring Beam In the event of a loss of a panel of fabric, it is important that some of the in plane prestress is maintained in the panels adjacent to prevent the deflections rising to unacceptable levels. During the normal operation of the structure, all fabric forces pass through the plane of the surface and are resisted by equal and opposite forces that arise in adjacent panels. In the event of a panel failure (or removal) this balance is upset and the radial cable connection nodes would be forced out of line. Circumferential cables through the nodes were required to maintain their spacing and resist these forces. The dome roof shape with tapering segments has an advantage in resisting ponding in that the span of the fabric panels increases as their slope increases so the fabric surface gets progressively softer. However, If the Fig 13 30m dia Central Cable runs Fig 12 Wishbones to raise circumference cables circumferential cables were in the surface of the fabric they would cause a dam at each circumferential line so an arrangement was required which would take these cables out of the surface. This was achieved by raising the circumferential cables above the surface with rigid members (wishbones) and connecting them to the nodes with criss-cross cables. Lower circumferential cables were also required to control the tiedown cables; these were also spaced off the surface but with out the criss-cross cables. It was also necessary to control the deflection of the radial cables. Their length is very long, 150m from the perimeter to the centre. Because of this if one 25m span were loaded the remainder of the cable in the line would act as springs so the loaded span would not be as stiff as if it was fixed at each end. The only way to gain the necessary stiffness is to use a high pretension. In fact the planned pretension in each radial line is 400kN, about 2/3 of the peak tension. The last element in preventing ponding is the patterning and prestress in the fabric panels. During the tender period some development of the design continued. We decided to change the central node for a 30m diameter cable ring. This was constructed with 12- 48mm diameter cables. Because of the redundancy implicit in the 12 cables, failure of one of these cables would not compromise the overall safety of the roof. These changes were brought in to the contract package before the contract was finally placed. STRUCTURAL DETAILING With cable structures it is essential that the details respect the system lines and system points of the cables and their intersections, as well as the likely movements of the cables at the connections. If the radial cables were continuous through the node points the flexing at those 'points would cause the cables to fail prematurely in fatigue. At every hanger location, the radial stringer cables are connected together at a node detail. This detail allows the high radial forces to pass directly thorough into the adjacent cables and allows the hangers (both upper an lower) to be connected into position. The vertically oriented connection plates allow the radial cables to rotate on their end fittings as they deflect under load. The flat top plate stiffens the node against shear forces and provides a surface that the fabric can be clamped onto to form a weather seal. DESIGN VERIFICATION As is usual for major building structures the safety of the design was verified by calculations. These relied on using our "Tensyl" program for calculating the forces in the fabric and cable structure. A Vj2 model of the cable system was modelled using Tensyl, to investigate the effects of cable prestress and the environmental imposed loadings upon the structural system. The tensyl analysis was then expanded to a V 2 model to allow us to predict the maximum cable and fabric forces, the compression loads in the supporting structures and the deflected shape of the 'total' system. Analysis of the cable system has shown that the behaviour of the structure is very sensitive to cable stiffness. The radial stringer cables rely upon a high level 154 of prestress (about 400 kN) to ensure that deflections are controlled during each load case. Under load it is critical that the stretch in the mast hanger cables is minimised. Each cable has been sized not only for strength and ultimate load capacity, but also for axial stiffness in order to ensure that the cable system does not 'go soft' when under load. This is the reason why we have used spiral strand cable, a much stiffer product than standard IWRC wire rope. Wind loads were derived initially from published data. They were then confirmed by wind tunnel testing at the BMT wind tunnel at Teddington. The safety of the components was investigated following normal design rules. Resistance of the whole structure to accidental damage is provided by redundancy, i.e. the structure can tolerate the loss of an individual component without collapse. This principle also applies to the support pyramids which are designed to withstand the removal of a leg. The 90m long masts were constructed with 8 323mm diameter tubes braced with rings at 2.5m spacing. Their overall diameter was limited by transportation requirements and a great deal of computer calculation went into verifying their load capacity. The limiting load was calculated using LUSAS in a non linear mode. Since the masts are leaning deflections under self weight and icing have to be taken into account as well as initial out of straightness. Wind loading is also significant but this does not occur with the peak down loads from snow and icing. Fig 14 Masts STEELWORK CONSTRUCTION STAGE The selected steel contractor, Watson Steel, were obliged to develop the engineers design drawings into shop drawings for the production of the components. This process involves an element of detail design of the components and connections. The shop drawings show the cutting and holing dimensions of all the plates as well as the welding and connection details. These drawings are reviewed by the engineer and architect for approval prior to the start of fabrication of each particular part. The cable work was subcontracted by Watson to Bridon Ropes of Doncaster. The cables have to be wound from wires that have been previously drawn and galvanised. For the dome project, class A galvanising, the lightest, was specified for cables which were beneath the roof and Galfan, a mixture of aluminium and zinc galvanising which is much more durable, for the external cables. The cable has to be pre-stretched to eliminate the construction stretch and then marked to the correct lengths under the specified pre-stress load. Most of the cables are dead length without any provision for adjustment, consequently great care has to be taken to ensure that the cables are made up to exactly the right lengths The lifting of the masts was planned by Watson with great care. This involved selecting a suitable crane and devising lifting positions which would not overstress the masts. Each mast was lifted and guyed with the two permanent backstays and two temporary forestays. There is also an intermediate position while the crane was released when only one forestay could be used and a short term guy was added from the centre of the mast to the adjacent base. While the mast is guyed with the temporary forestays the central ring is lifted by the permanent forestays. During derigging of the crane and the operations of changing the guy positions the tensions in the guys had to be carefully controlled to maintain the stability of the mast. Following the lifting of the ring, the guy system was moved so that the rest of the cable net could be assembled and lifted to its place. This was done using hydraulic cable jacks with the hoisting cables running over sheaves on the top of the masts. When the net was completely assembled and all the cable lengths checked, each of the 72 pairs of radial cables had to be tensioned. This was achieved in several steps using a 550 kN capacity 'Enerpac' pull jack in the pre-designed jacking points at the front of the perimeter masts. Because of the flexibility of the central ring and the boundary cables the tensioning of the radial cables had to be done to specified dimensions rather than to specified loads with final adjustments made at the end. SELECTION OF CLADDING AND THE INTERNAL ENVIRONMENT The roof is to provide a controlled environment for the exhibition and for what other uses it may be put. The human response preference is for a bright translucent roof with a light spectrum as close as possible to daylight. This requirement conflicts with the needs for the central show and some of the exhibits for which lower light levels would allow greater impact from exhibition lighting. Coated fabrics tend to change the spectrum to a brownish hue rather like tungsten lighting. This of course affects the perception of colours within 155 Fig 15 Perimeter masts with jacking points between rigging screws the dome and according to our researches for the 58° N project can affect the physical performance of people within the dome. It is difficult to have a translucent fabric roof with insulation but with out any insulation condensation will occur on the underside which, in certain conditions, will fall as rain. This situation would be totally unacceptable in a building that will have a lot of electrical displays. To reduce this risk a lining can he installed under the main fabric. There has been a considerable amount of experience with fabric roofs with linings where condensation has not been a problem. Checks were run on the risk of condensation as part of the environmental modelling and they demonstrated that with two membranes the risk of condensation on the underside was very low. The available materials for cladding the dome were PTFE coated glass fibre cloth, PVC coated polyester cloth or ETFE foil cushions. Our preferred material for the roof as an environmental enclosure would have been ETFE foil. This would have provided a high translucency roof with three layers of foil which would have a considerable amount of insulation effectively eliminating the risk of condensation. Unfortunately we did not consider that there was sufficient experience with detailing this material in this situation and we considered that it would be too risky to try to develop a suitable system within the very tight time scale. The necessary properties of durability and flame resistance are provided by PTFE/glass without the need for any additives. The glass fibres are not affected by UV light but they are damaged by water. The function of the PTFE coating is to protect the fibres from water and abrasion, the PTFE itself is nearly inert and is not affected by the weather. The fabric is seamed by heat sealing using a FEP interlayer which melts at a Fig 16 PTFE/Glass fabric temperature of around 350°C. A benefit of this is that the material can be repaired on site with a permanent seam that is the same as those done in the factory. With PVC/polyester the fibres are damaged by UV light and they burn so the function of the coating is to protect the fibres from UV light as well as providing the flame proofing. The PVC itself is light stable and does not burn well but it requires a number of other compounds such as pigments, UV stabilisers, plasticisers, fungicides and flame retardants to meet the functional requirements. Since 1987 several of these compounds especially fungicides and heavy metal stabilisers have been banned and this has led to an increase in problems of fungal growth in the yarns which severely discolours the cloth. This situation has recently been improved by the use of anti-wicking treatments to the yarns. The other big problem with PVC coatings has been dirt retention. The PVC coating is porous and the plasticisers absorb dirt. This has recently been improved by the use of fluoropolymer surface lacquers which give it a durable sealed surface. After investigating the products of the three best coaters in Europe an outer fabric was selected which gave 15% translucency and an inner lining fabric which gave 75%. The combination gave the highest translucency, about 12%, and a good colour rendering. The fabric selection was changed to PTFE/glass after a political decision to build the dome with a long life. FABRIC WORK The contractor who had made the best offer for the PTFE/glass material was Birdair from Buffalo, New York State. They have been producing structures in PTFE for over 20 years including some 12 covered stadiums of approximately half the area of the dome. The fabric patterning and attachment details had to be modified to accommodate this alternative material and since time had been lost in the programme, this had to be done in a very tight time scale. Because of the 156 arrangement of the panels within the cable net, and the fact that the cloths were to be fitted in to dead lengths, the patterns had to be extremely accurate. Since the warp direction of the panels of the outer fabric ran radially on the roof with 25m long cloths, it was necessary to model the fabric as an equal mesh net to represent the warp and fill lines of the cloth. This was a much more time consuming method than the standard method of representation using triangular elements. These basic geometry patterns were converted by Birdair into cutting patterns. They also built in the stretch compensations, which were agreed after biaxial tests on the actual production cloth and added in all the edge details.which were agreed after biaxial tests on the actual productioncloth and added in all the edge details. Fig 17 The fabric attachment detail proposed by Buro Happold , was a double luff groove extrusion fitted onto the radial cable pairs to accept a roped edge on the fabric. Birdair proposed a 12mm edge cable in the fabric which would hook into special clamps fixed to the cables. The clamps were developed into a two part extrusion cut into 50mm lengths and retained by two 12mm bolts. Fabric sealing flaps were closed over the top of the site joints and sealed together using a hot iron at 380°C and an fep inter-layer Fig 18 TUNNEL VENT AREA A 50m diameter hole was required in the roof around the Blackwall tunnel vents, to accommodate the 'air rights' of the ventilation structure. After considering a number of ways of leaving a hole in the fabric, Buro Happold adopted a net of 8mm cables at lm spacing which would replicate the stress-carrying capacity of the fabric but would allow the vent air to pass through. The cable net was attached to the fabric with clamp bars at the edges. The net patterns were developed directly from the typical fabric patterns with the boundary line being defined to align with the top of the enclosure. The net arrived on site in rolls and was erected in the same way as the fabric using the same extruded hooks modified with a steel plate to which the cable terminations were attached. LEARNING FROM THE DOME There are two questions; is a large environmental enclosure of 10 or 15 hectares feasible. If so would it be energy efficient and provide a improvement it living conditions in extreme climate zones. The zones would be either semi-arctic or desert requiring either raising or lowering the internal temperature. Form The selection of the form of the Dome as a shallow spherical cap is beneficial for wind and snow loading. The smooth profile generates a smooth airflow over the surface with of local turbulence. It also reduces the uplift pressures so reducing the tensions in the cables Snow will always be a problem on transparent or translucent roofs If not from the load effects then because it excludes the light. In windy conditions the snow will mostly blow off. Dividing the roof surface up into individual panels with valleys between has been shown to give trouble because of the concentration of snow which can build up there. The smooth Dome form avoids this, it also has the useful property of having increasing slope with increasing fabric span that reduces the risk of ponding. This form could be increased to say double the area without compromising the structural behaviour. Unfortunately we may have to wait several years to get a few snow storms to prove or disprove the behaviour of the Dome under snow loading. Internal Air The biggest problem with the internal environment has proved to be the dust. Firstly that generated by the construction operations which began immediately the roof was completed and continued for a year. We think that the ongoing running operation of the dome will create a dusty environment generated by the visitors Fig 19 moving around and by the cleaning operations with dry brushing. There is also the dust and dirt coming in with the external air in the polluted environment. The solution would be to hose the surfaces down. This is the method adopted in large covered stadia but of course it causes condensation on a single skin roof. The special qualities of fresh air are not well understood but dust and pollutants are known to make it unpleasant. In a covered environment a plentiful supply of outside air is normally required to keep it smelling sweet. If the temperature outside is sub zero any cold air introduced will flow across the ground floor tending to defeat the object of the covering. It would be interesting to investigate to what extent fountains and "rain" would clean the air so that air changes can be reduced. Heating In winter the warm air immediately migrates to the top of the space until it looses heat so heating the air is not very energy effective. A better way is to heat the ground to improve the local comfort and let the air look after its self. This is more or less the approach adopted in Victorian stoves where the heat was introduced via pipes in ground trenches. Ground heating can be done with low grade heat supplimented on sunny days by radiant heat from the sun and heat from internal buildings. Because the covered ground area is so large little heat will be lost to the outside although it might be absorbed into the ground at a low temperature. Heat exchange to the ground will change its temperature very slowly and the deep ground temperature of 8 or 10 deg C will provide limits. Cooling In desert conditions the enclosure would need to provide shading. This can be provided with a single skin of Teflon/glass fabric. The problems of condensation would be very much less than in cold climates. Cooling at ground level is easier to achieve since the cold air tends to stay on the ground. To gain a benefit from the enclosure it would be necessary for the ground to act as a coolth sink. Would this happen by natural means or will mechanical cooling systems be required? Cladding In the case of the Dome the dust has made the lining fabric unacceptably dirty. This is largely due to the particular material supplied by Chemfab. Called Fabrasorb It is marketed as a sound absorbent material because of its porosity. In reality it is glass fibre cloth barely coated with teflon and the coating is easily damaged by handling. The porosity improves the Fig 20 translucency at first but the result is that the dirt gets into the fibres and cannot be removed. There are now more translucent PTFE/glass fabrics around, the problem is that the inner layer may not bleach out because of the lack of UV penetrating the outer skin. The porous lining will also allow water vapour to pass through and condense on the outer skin. The solution for a habitable enclosure has to be to treat the roof as for a swimming pool. This means that there should be sufficient insulation to prevent condensation. The best way to acheive this is to use three layer ETFE foil as was originally proposed for 58 deg North. The benefit of this approach is that a vapour barrier would not be required on the ground so it could be treated is it is outside. Plants could be grown anywhere, water features could be used and the interior could be cleaned with a hose with water soaking into the ground and evaporating from it. The Dome and other large enclosures such as the Eden project will enable us to obtain data to evaluate the benefits of such structures and define the limits to the size. This may be determined by fresh air, roof drainage and access requirements as much as by the structure. REFERENCES 1. R Buckminster Fuller and Robert Maries, The Dymaxion World of Buckminster Fuller, 1960, Anchor Books Edition, 1973. 2. Walter W Bird, The History of the Air Structures in the USA. IL16 Zette, publication of the Institute for Lightweight Structures, University of Stuttgart. 3. IL2, Publication of the Institut for Lightweight Structures. University of Stutgart, 1971 4. Liddell W I, 'Minnesota Metrodome' A study on the behaviour of air supported roofs under environmental loads. Structural Engineering Review 1994, Vol 6, No 3-4, pp.215-235, Pergamon. 5. C Gill, I Liddell, C Schwitter, Straight cables for tension structures. Procedings IABSE symposium, Birmingham 1994 6. Liddell W I, 'Creating the Dome' The 1997 Royal Academy of Engineering, Hinton Lecture published by the RAEng. 7. Liddell WI, Miller PW, 'The design and construction of the Millennium Dome', The Structural Engineer, Vol77, No7, 6 April 1999 8. Elizabeth Wilhide, The Millennium Dome Ted Smart 1999.