This paper describes the Valhalla portable membrane structure. This structure uses unique methods of design, manufacture and installation to achieve the project aims of providing the first large scale portable structure to house the same events as a permanent arena.
The structure was used for the first time on new years eve 1999 to herald the new millennium. It became the largest indoor celebration in the UK for the new millennium by accommodating 28,000 people inside the structure together with staging, light and sound equipment which cost £400,000 to hire for the night. 90% of the effects were suspended from the structure with a total flown load of 400 kN.
This paper will describe the concept behind the Valhalla, (named Tensile1 by the operating company) and the processes that went into designing and manufacturing the Guinness Book of Records 'World's Largest Portable Venue'. [Ref 1].
1. VALHALLA - THE PROJECT
In Norse mythology, Valhalla was the legendary banqueting hall of Odin, principal god of the Vikings, which played host to the Einherjar, the souls of warriors who had died a courageous death in battle. The largest building in Asgard, heavenly home of the gods, Valhalla constituted one of it's 12 realms, its cavernous interior playing host to Viking heroes awaiting Ragnarok, final battle of the world. Its 540 doors, each wide enough to allow 800 warriors to enter abreast, waiting to receive the souls of the brave. This legendary building was to provide rich inspiration for the very real Valhalla project.
Conceived and designed to provide the same sense of wonder to all who enter for the first time, as that original building of Asgard, the proportions of the Valhalla structure do not shame the legend. It is 86 metres wide, 160 metres long and 25 metres high. (106 metres by 180 metres to the main anchors). The Valhalla membrane structure as conceived is, simply put, the largest practical mobile membrane structure ever built, that also has the ability to be moved quickly.
During the period 1980 - 2000 the temporary structures market in the UK grew from a situation where pole structures up to 30 metres wide, usually made from cotton canvas were the norm, to a thriving diverse business that generates nearly UK £100 million in the UK alone. This expanded market covers family, sporting, music, cultural and national events. The two largest areas are the corporate hospitality and the music/festival sectors. In the period 1980 - 2000 the span of the large tented hire structures went from a usual 35 metre maximum to the current largest span structure, the Valhalla, at 86 metres.
The structure was used for the first time on new years eve 1999 to herald the new millennium. It became the largest indoor celebration in the UK that night by accommodating 28,000 people inside the structure, in addition to staging, light and sound equipment which cost £400,000 to hire for the night. 90% of the effects were suspended from the structure with a total flown weight of 400 kN. The event generated £4 million plus on the night and the same again in CD and video sales. Almost certainly, the success of the event has led to the first ever tour by a major contemporary rock band, Radiohead, to take place exclusively in a tented structure, the Kayam concert tent, the predecessor of the Valhalla.
The project was based upon the concept designs of Rudi Enos as a result of previous projects such as the Kayam theatre membrane and the 'Walt Disney on Ice' portable structure. As with any project in the modern age, no one person is responsible for a project in its entirety. Special Structures Lab of Sheffield, UK were chosen to provide the engineering services and to prove the initial design. An architect, Bruno Postle and engineers Stuart Holdsworth and Laurent Devesne were the principle designers in addition to Rudi Enos. Bruno Postle in particular provided all of the three dimensional system geometry virtually single handed and many of the manufacturing details and parameters.
Previous large-scale structures were hampered by compromises in the specification, that generally rendered them unable to meet their tasks, so the Valhalla project laid down the most stringent design targets ever for a large scale portable tented structure:
Site wind speed=50 m/s
Snow load=0.2 kN/m2
Suspended load=200 kN per each pair of masts
Erection equipment=telescopic forklifts
16 pole structure=10 - ISO standard 12 metre containers
Modularity=12 different configurations
1.1 Operating sizes
The Valhalla is a modular membrane fabric structure designed to be operated in several different guises. It includes features such as 30 metre span between the poles, 6m high sidewalls and no intermediate poles, which are traditionally used on circus tents.
By adding sections, it can be operated at 45 x 45m, 45 x 75m, 45 x 100m, 45 x 130m, 75 x 75m, 75 x 90m, 75 x 105m, 75 x 120m, 75 x 135m, 75 x 150m, up to the contract maximum of 20 masts or 75m x 180m (all sizes to inner walls). This increases by 5 metres all round (to the points), and 12.5 metres to the main guys. The total floor area in the contract was for a surface area of 23,456 square metres.
Part of the reasoning behind the initial decision to build the Valhalla, was to provide a structure that was not only state of the art and usable without intermediate poles, but one that was truly economic and mobile. The fact that it can be used as a one pole, two pole, three pole, four pole, six pole, eight pole, ten pole, or even twenty pole and more, allows for more efficient use of stock.
The structure is transported in 10, ISO 12 metre standard shipping containers. All structural members are designed to not only perform their task, but to fit in the containers too.
1.2 Parameters
The Valhalla uses unique methods of achieving particular objectives. Traditionally, large permanent and mobile membrane structures rely on large diameter steel cables to provide control points. As the Valhalla was to be a truly mobile membrane it was considered that extensive ground works, (or large steel cables or reinforcement), other than temporary anchors would not be suitable. Methods of providing anchorage were used that had not been considered before. So in performance terms the main points that were taken into consideration in the design process of the Tensile1 were:
1.2.1 That no steel cables were used anywhere on the structure.
1.2.2 That the surface should use the principles of true membrane technology wherever possible, although the end centres were deliberately made flatter than the ideal, to accommodate the ability to alter the span of the structure. This was only considered acceptable due to the short span of the membrane at this point.
1.2.3 That the unit be designed to be modular, and to be usable in several different guises.
1.2.4 That the structure be capable of suspending 200 kN of uniformly distributed load from each pair of masts.
1.2.5 That the structure and the supporting masts were capable of being erected without the use of cranes.
One of the main design concepts of the Valhalla is it's internal height. The steep shape works extremely well to combat the effects of suction across the roof of the otherwise flat saucer shaped structure and avoids the problems inherent with such a shape. This gives a lot of internal clearance for stage lights, trussing and rigging etc. The initial manufacturing program for the Valhalla covered an area of 20,352 m2, in which the membrane incorporates 29 sections in the roof with 16 king poles, giving a clear floor area of 30 metres by 120 metres with a load bearing capacity of an overall suggested maximum of 100 kN per king pole.
2. DESIGN
2.1 Research
The research engineer tries to develop new principles and processes by using mathematics, scientific concepts and experimentation. For instance, large computer simulations developed by our research engineers permitted the prediction of the performance of the structure prior to manufacture. We were able to predict the deformation of the membrane under varying environmental conditions using non-linear Finite Element Analysis tools, depending on the stiffness and stretch characteristics of the materials we chose.
By testing a representative sample of test pieces in the laboratory, we arrived at a reference configuration for the structural fabric. This reference configuration was modelled and simulated in software, then analysed using different parameters until the software model exhibited deformation and stress/strain results similar to the real world tests. The software could then be said to be calibrated to the material being analysed.
2.2 Development
Complex engineering systems need long periods of time for their development. It can involve the designing of components, often using new materials or new ideas, the testing of these components and then the improvement of the original ideas. This has been so with the Valhalla, the total period of the project spans from 1997 to date. All of the components must be put together to build the final engineering system. This often implies that the small-scale experiments performed by the researcher must be scaled upward to the level of industrial practice. In the case of the Valhalla, we used previous projects to provide a platform from which to start. This platform gives the project team the confidence to push the boundaries of the performance characteristics to new levels.
2.3 Design and Analysis
Coupled with and following development is design. An engineering project must not only work, but it must also be safe, economical and reliable and must meet the needs of the customer.
The specific layout of this engineering product or structure is the responsibility of the design engineering team. The following software was used on the Valhalla project.
2.3.1 Patterner for Windows -- Bruno Postle
Patterner is one of the few software packages in the world to provide the tools to create membrane structures. Developed in house at Rudi Enos Design by Bruno Postle, it is considered to be 'state of the art' for visualisation and patterning. All the production patterns for many different projects have been created with 'Patterner'.
2.3.2 Tech-net 'EASY' membrane force and cable analysis software.
EASY analysis tools provide the most effective method of assessing the forces in the structure's membrane, it's cables and the reaction loads on the supporting steelwork. With the ability to define and model precisely the ultimate tensile strength and stiffness of the membrane and reinforcement materials and then to determine maximum loads, deflections and reaction loads from the dynamic membrane, every condition can be simulated and shown.
2.3.3 Spaceframe Analysis Software
Two main packages were used, apart from CAD (computer aided design) software to provide the level of analysis required for a project such as this, CAD Analyse and QSE Space.
2.4 King Pole and Suspension Trusses
The king poles or main masts were designed to fulfil several main objectives. Initially, the masts were required to provide the main support for the membrane which transmits the environmental loads. They were also needed to suspend a load of 230 kN, consisting of 200 kN uniformly distributed load and a truss self weight of 30 kN. The masts were restrained to an overall width of 550 mm to allow four units to fit in the width of a standard 12 metre ISO shipping container. The mast half section lengths were also designed to fit within the 12 metre length of the ISO container.
The thirty metre trusses were designed to suspend a uniformly distributed load of 200 kN of computerised lights, sound system and special effects for performance use. The trusses which in the unfolded state are roughly 30 metres by 2200 mm by 1500 mm, were designed to fold flat for transport to prevent shipping air. Shipping volumes are a much bigger problem than shipping weights in mobile structures. The trusses split into three sections of roughly ten metres and fold to 300 mm in width. The king poles and 30 metre truss were tested with 360 kN of dead load to prove the system prior to the first trial erection.
2.5 Membrane Design
The membrane was designed using current engineering practice and software. All materials used on the Valhalla complied with the appropriate National Standards or Codes of Practice, where such existed, unless otherwise specified. Membrane engineering relies on generating an equilibrium form for the surface to reduce possible pressure points and on distributing the fabric tension into perimeter cables. Where there is a valley, a cable is needed to hold down the uplift of the membrane. On the Valhalla, no steel cables have been used within the membrane as the potential for wear and rubbing is too great.
The Valhalla is a hybrid membrane structure designed to be used in many unfavourable environments. High winds, rain, and snow have been experienced in the erection process so far. The membrane has a 4,500 kilogram PVC coated webbing, radio frequency welded on each panel seam. This creates a radial net of reinforcement attached to the membrane. During the analysis optimisation process, the representative mesh was tuned to provide minimal catenary, valley belt and surface loads. A situation was modelled where a main valley belt anchor has failed, to confirm that this failure would not be catastrophic.
The structure is engineered for massive redundancy. The obvious possible failures that could be anticipated were investigated. Material tests were undertaken of the composite assembly to see if this impacted upon the analysis. Over 50 load cases were run to confirm the result of different wind directions and configurations. Wind, snow and the full suspended loads were considered. At every stage the structure met the design and performance criteria with the correct factors of safety.
2.5.1 The Webbing Subsystem
Unlike traditional membranes, which tend to use wire rope cables or structural steel members, the Valhalla uses unique methods of controlling membrane forces. In effect the membrane is supplemented by a cable net of 0.1 metre wide webbings. These are attached to the 'webbing subsystem', a grid of multi layered, 120 kN ultimate tensile strength, webbing belts which divide the membrane fields. In fact, the membrane sections do not connect to each other at any point on the structure.
The webbing sub system connects to the membrane through a system of catenaries and compression fittings, which use a maximum of nine pin connections for each membrane field. These connections are sealed by a rain cover which can be either blackout or allow light to pass, creating either a black box or a well lit exhibition space. This ability to use the same structure in either way is unique in the temporary structures field.
2.6 Guides And Codes Of Practice
The British Institution of Structural Engineers has a guideline for the use of temporary demountable structures, [Ref 2], which gives other references and requirements for marquees and tents. Other sources include the British Guild of Tent-masters Approved Code of Practice and Made-up Textiles Association Ltd (MUTA) handbook the "Use and operation of marquees".
British Standard 6399 Parts 1-3 was used for loads. This was adapted for temporary use and short venue times, i.e. up to 14 days on a site. It was assumed that the internal operating temperature would not go below 120 Celsius and that snow loads can be ignored as per British Standard 6661. DIN and ASTM codes are further developed than the current BSI codes in this regard.
2.7 Testing
Most engineering products must be fully tested before they can be delivered to a customer. Testing may show possible failures. The product then needs to be redesigned. Development, design and testing must work closely together. It is important that the main contractor is in control of the test and the approval of finished components as soon as they are initially completed. This ensures that mistakes and inconsistencies are not repeated on further sections.
Quality control and inspection is important at every stage. The quality control engineer checks that all parts and assemblies meet technical and various other requirements. All drawings must be checked not only by the issuing designer, but also by the main and subcontractors. Any unclear or conflicting instructions must be reported to the main contractor immediately and all inspections and finished products must be approved by the main contractor.
2.8.1 Testing of Fabric Assemblies
This involves testing at the coating plant, where strict manufacturing process control is required, at independent test labs, where the material is tested for its elongation factors and flame retardancy and at the pattern cutting point, where the geometrical accuracy of the cut panels are checked. Testing must also be carried out at the fabrication and assembly shop, where general dimensional accuracy of the joined form and detail connections are inspected, which includes weld strength. Possible high temperature use requires high temperature testing of the weld.
Unless specified otherwise, fastenings were to be of the same metal as the component, with matching coating or finish. No contact was permitted between dissimilar metals in components, which were to be fixed where moisture may be present or occur. Finished components were to be rigid and free from distortion, cracks, burrs and sharp rises and moving parts were allowed to move freely and without binding on the membrane.
2.8 Lift Motors and Control Gear
Each pair of masts has three 4.5 kW motors geared to lift a maximum of 55 kN. The motors pull wire rope cables through a built in system of pulleys in the king pole top, (mast head). A pair of ropes attach to the bale, (lifting), ring, run through the mast head pulleys and link together inside the king pole. The link is connected to a single cable which runs around a winch drum geared to the lift motor. The motors are controlled by three phase switch gear which require insulation to IP 67 standards.
The controls are used to first lift the king poles, then are re-connected to the main lift motors for membrane hoist. The control gear has two hand held low voltage remote controls, each controlling eight channels. If the structure has more then eight poles, a second operator is required. Each motor is controlled by relays from the low voltage handsets. The motors have Ampere controlled overload switches to limit the maximum lifting force. These trip at the correct pre-stress load from the membrane at the bale ring, but can be over ridden if local conditions or high winds require it.
2.9 Rigging
An extensive array of wire rope cables are used to stabilise the masts during erection. All wire rope cables used for the Valhalla were IWRC galvanised and pre-stretched. Over 2,000 metres of cable were used on the structure.
3. MANUFACTURE
The making of parts and components, (sometimes with the help of robots), is usually considered a subspecialty of mechanical engineering. The knowledge of 3D CAD-CAM, engineering design, fluid mechanics, structural computation, machine design, controls, robotics, operations, maintenance and logistics are all involved in the Valhalla project. All parts are designed with the aid of computer graphics. The computer carries out all the technical computations needed to make a part meet performance requirements. This aspect of computer-aided design (CAD) is frequently coupled with computer-aided manufacture (CAM) to produce parts automatically. The scope of the works in the contract comprised the design, supply, patterning, fabrication, delivery and erection of the fabric canopy together with the associated steelwork, cables and forks, anchors and supporting steelwork. The many different engineering functions to process the Valhalla project included research, design and development.
3.1 Steelwork Assembly, Manufacture And Testing
The fabricated steelwork was created by modelling the steelwork in three dimensions and creating a profile from the assemblies, which was used to create laser cut parts. The parts were designed with slots and tongues to aid assembly which reduced the need for traditional assembly jigs.
3.2 Membrane Manufacture
Special Structures Lab Ltd were responsible for producing full cutting patterns for the membrane work, together with the detailed steelwork drawings and the true angles of corner connection plates including, information such as: -
* Drawings showing detailed arrangement of all joints and boundary details.
* Sewing and welding procedures and sequence.
* Marking and numbering for the cloths.
* Full size drawings of corner details.
* Field layout drawing showing cutting dimensions of fabric including stretch compensation and decompensation.
* Indications of places where smoothing curves were required.
* Positions of pre punched holes for all site fitted clamps.
* A schedule of all shackles, ropes and fittings and a marking system for the same.
CAM (computer aided manufacturing), was used to plot the panels used to assemble the sections. This involved the plotting of individual panels and the cutting of all left and right hand versions of the panel. This part of the assembly process was totally automated. Fabric panels were marked out along and at right angles to the centre line of the roll to an accuracy of 2 mm, making sure that if the panel was laid out on the floor prior to marking there was no diagonal distortion of the weave. The fabric was then cut along the marked lines to an accuracy of +/- 2 mm. Cable cuffs were bias cut strips of fabric with the yarns at 45 degrees and were pre-fabricated with reinforcing where required.
4. ERECTION CONCEPTS AND PROCEDURES
The Valhalla can be erected with a twelve man team in three days using only telescopic forklifts when transported by road. All lift motors are built into the king poles and all rigging travels with them. The king poles are laid out and attached to steel base plates, which accept horizontal loads of some 80 kN during initial lift of the king poles and downward pressure of 170 kN under maximum load. In firm soil the standard base plates can accept these loads.
The structure is erected by first laying out and erecting the king poles, then laying out and assembling the membrane and attaching the rain covers. The bale, (lifting), rings are raised up the king poles to 50% of the height of the masts. The 'A' frames, (outer masts), are connected to the membrane and partly raised into the air with mechanical winches. When the membrane has been checked to ensure that all connections are correct the 'A' frames can be raised to the service position. The membrane can be raised to 95% of the prestress form and checked once again.
The membrane is then raised to its service height and all tie backs and tensioners checked for correct tension. At this point the tie back tensions are checked against the figures arrived at through analysis and corrected if necessary. At this point the lifting gear is made fast at the built in points and all electrical gear disconnected. The structure is now ready for interior fit out after which the walls can be installed.
4.1 Ground Anchors
Despite the efforts of the design team to minimise them, anchor forces on the Valhalla can be as high as 280 kN. Previously unused methods of temporary anchors were required and techniques developed to deploy them. Two metre lengths of screw anchor are used with a tourque meter on a hydraulic motor with shear pins. When tourque readings or lack of shear in the shear pin show that the anchor will not meet the expected load, an additional two metre length of anchor is inserted and screwed into the ground until the required readings are attained. The inserted anchors have a representative 20% test applied to them with a 600 kN portable test rig. If any anchor fails, all anchors are then tested. This allows the operator of the structure to be absolutly sure that the structure can meet performance targets.
Another concern is the possibility of the main masts sinking into the earth under high load. The following table gives guidance as to the allowable bearing pressures on certain soil types and the size of spreader required. Extracted from British Standard 8004. The soil loading data and information was arrived at by on site testing and compiled using data and research undertaken by Ovesen and Stromann [Ref 3].
Type of soil
Allowable bearing pressure
(kN/m2)
Spreader plate size
(maximum req. for this bearing pressure)
Loose sand
90
2.1m x 2.1m
Medium dense sand
100
2m x 2m
Firm clay
75
2.31m x 2.31m
Stiff clay
150
1.65m x 1.65m
The structure is one of the largest portable buildings ever built. While smaller structures have been larger when connected together and larger structures with lower wind speed capabilities have been built, there has never been a structure with the presence, the design standards, the performance capabilities and the sheer size, height and load carrying capabilities of this ground breaking structure.
It defines a new paradigm for portable structures.
References
1. Guinness World Records, Guinness Superlatives, 1999, London
2. Institution of Structural Engineers, 1995, Temporary Demountable Structures, Guide to Procurement, Design and Use, SETO, London, ISBN 1 874266 17 4
3. Ovesen, NK and Stromann, H, 1972, Design Methods for Vertical Anchor Slabs in Sand, Proceedings, Speciality Conference on Performance of Earth and Earth-Supported Structures, American Society of Civil Engineers, Vol. 2.1, pp.1481-1500
Overview - Valhalla Membrane Structure - Tensile 1
Enos, P R - ABSTRACT
