Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Time machine

Construction

Some site teams might have wilted under the unprecedented media spotlight that was directed on the Greenwich peninsula. NCE reports on the ups, downs and eventual triumphs involved in building the Dome.

As far as most of the outside world was concerned, construction of the Dome began on 12 October 1997, when the first of the 12 bright yellow masts was raised into position. A lot had happened before then, of course, not least the driving of some 8,000 piles (see p???), welding together of the six 15m sections that made up each 90m mast, and setting up of the 10m high mast support quadropods. But the mast raising was a very visible milestone that had to be passed on time if the project was not to attract even more negative publicity.

NMEC site, structures and transport director David Trench says this was just one of the 'drop-dead dates' he inherited when he arrived on site in March 1997. 'It was a very simple, uncomplicated programme designed to ensure we would open on time.

'I decided these key dates, about 30 milestone events in all, should be set in stone, and that we would do anything, anything to meet them.'

This led to the whole project being programme rather than cost-driven. The risks of delay associated with every operation were scrutinised minutely, and plans revised without hesitation if these risks could be reduced significantly.

Originally Buro Happold had planned to attach all the cables to the multiple-flanged mastheads at ground level, then winch the masts into position from a temporary central tower before beginning assembly of the cable net proper in mid-air. This was changed after discussions with Watson Steel, the successful bidder for the cable net contract, who preferred to assemble virtually the entire spiders web at ground level to minimise the amount of work that would have to be carried out at height.

It was therefore decided to install temporary ground level anchorages in a 20m radius circle at the centre of the ring of masts and a winch at each masthead for later use in raising the net into position. A combination of permanent backstays and temporary forestays would hold each mast at the design inclination of 17degrees off the vertical until the cablenet was raised and tensioned.

Few of the press and VIPs who turned up to the first mast raising had ever been near the site before. Most found it hard to believe that the windswept hectares of grubby hardcore in front of them could be transformed into a major exhibition venue in little more than two years. The sheer size of the masts was impressive, yes. So were the two massive cranes deployed by Watson Steel, but the apparent slowness of the raising operation was far from reassuring.

But as Buro Happold project manager Tanya Ross points out: 'Everything gets quicker with practice, and we had 12 masts to learn on. The first took two days to erect - the last two went up in just one day.'

From first to last the masts were erected in only 20 days. Watson had just 16 weeks to assemble, erect and tension the cable net, all 70km of it. Produced in Doncaster by Bridon International, the galvanised cables ranged from 25mm to 90mm in diameter. Each cable had to be pretensioned before measurement to ensure accuracy of length, especially important for the paired sections.

Assembly began with the complex central ring. Templates and braces were used to maintain the spacing of the nodes and clamps. Throughout assembly of the cablenet locating the 300 permanent clamps and more than 500 nodes to the desired level of accuracy was to be a major concern.

Once the central hub was assembled radial and permanent forestay cables were attached. Lighter temporary cables then hauled the ends of the forestay cables up to the mast heads where abseilers from CAN (see box) made the connections.

A similar method was used to raise the six circumferential cable assemblies in three pairs. As each pair was winched into position abseilers connected sections of radial cable to adjoining circumferential cables. In all, 29 connections had to be made to each masthead - and 18 tie down cables had to be connected to each foot.

All connections were complete by the beginning of March. At this point the untensioned cablenet looked flimsy and insubstantial, sagging between the rigid masts. It needed very large tensile forces to transform it into a credible loadbearing structure.

So far the operation had gone smoothly. As the enormously powerful hydraulic jacks tensioned each pair of diametrically opposed radial cables and the rigging screws at the top of the perimeter masts were adjusted, it became obvious that something was out of alignment. An investigation soon revealed excessive variations in the lengths of some of the cables. Two of the circumferential rings were up to 100mm out of position, the highest point of the net was about 300mm higher than intended, the masts were just over 1degrees out of alignment.

It was a bad moment for the project team. Acrimony and a possibly unrecoverable delay might have been the outcome on a less harmonious site. But the 'Greenwich culture' (see page IV) survived the test, and the team's energies went into solving the problem rather than rehearsing excuses. Panel edge details were tweaked, some fill-in plates hastily manufactured. In the end the problem turned out not to be as bad as first feared - the same went for one of the worst moments for those working up on the net.

Buro Happold partner Glyn Trippick was there when it happened, high on the central scaffolding, in May 1998. 'There was a bang, the whole net shook, and everyone dashed for safety,' he recalls. 'It turned out that the clamp connecting one of the radial cables to the edge cable hadn't been fully tightened, and it had suddenly jumped 2m along the edge cable.'

By this time, of course, the panels of PTFE-coated glassfibre fabric were being placed on the cablenet. Much to the team's satisfaction the very first 225m2 panel had been unfurled on 23 March 1998, the very day shown on the programme drawn up more than a year earlier. Another of the milestones had been passed, despite the very late change to a much stiffer and less forgiving fabric than the original. (see box).

Trials of the panel fixing techniques on a ground level mock-up close by had gone reasonably well. None the less, the team from fabric supplier Birdair's normal fixing subcontractor Danny's Construction Company found it hard going for the first few weeks. CAN's abseiling team had to be drafted in to help get the operation back on programme.

'It didn't help when we discovered that the panel edge cables had been cut in America to 75ft rather than 75m,' says Trench. 'We had to fly in replacements in a hurry.'

Part of the problem was the sheer difficulty of working at height with panels weighing more than 1t. Two rings of 72 narrow 'ribbon' panels made up the main roof, both 75m long, with the upper panel 1m wide at its narrowest and the lower panel 15m wide at its widest. Each arrived on site from America in a neat concertina fold and had to be carefully unfurled up on the cable net - wind and weather conditions permitting.

Birdair vice president of operations Stan Kopaskie says the weather restrictions worked out with the Health & Safety Executive were realistic. Panel fixing could normally continue in wind speeds of up to 45km/h. It was the fixing techniques that first caused real problems.

Rather than Buro Happold's well-tried luffgroove system (see box) the Dome panels relied on 12mm diameter secondary radial cables slipped through edge pockets on each side and clipped onto the circumferential cable rings top and bottom. These were tensioned radially first, then Tirfors were used to stretch the panels sideways until the edge cables, exposed in cut outs at 500mm centres, could be hooked over the aluminium edge connectors clipped onto the main twin radial cables.

CAN abseiliers fixed the 26,000 edge connectors well in advance of the main fabric fixing crews. With CAN working from above and Danny's construction company using two radial underslung access walkways, progress was soon back on programme. One major contributor to the whole roofing process was the record-breaking 40m by 30m by 50m high central scaffolding tower erected by Kwikform UK in just two weeks - 72km of steel tubing was needed to complete it, and the hoist running up the outside was a major boon to the fabric crews. As with most areas of Dome construction, progress on the roof was inextricably linked to other sections of the works, not least Watson Steel's construction of the six, three storey core buildings below. Work on the buildings was forbidden while men were working on the cable net: potential abrasive relationships between the building and the roof teams were minimised by the fact that Watson was main contractor for both.

Prime Minister Tony Blair attended the topping out ceremony on 22 June 1998, just 52 weeks after the first piling rigs arrived on site. Much still remained to be done, both inside the Dome and around it, but the main envelope was substantially complete.

With the prophets of doom temporarily confounded, attention switched to the 14 Zones, some of which turned out to be major engineering challenges in their own right. After handing the Dome over on time and to budget, the site team had every confidence they could do the same with the rest of the project ,and they were right.

Cover up

Specifying a polyvinylchloride-coated polyester fabric for the Dome's 90,000m2 roof was the obvious decision for Buro Happold's design team. At that point there was no confirmed need for the structure to last longer than two years, so long term durability was not a major issue (although the accepted working life of the chosen fabric was at least 12 years.) PVC-coated polyester is relatively soft and easy to handle, fixing and tensioning details are well developed, cost is lower than PTFE-coated glass fibre fabric, the only realistic alternative.

'And Buro Happold has a lot of experience with this type of fabric,' project partner Ian Liddell points out. 'Our original design featured double aluminium luff groove extrusions clipped onto the radial cables with each panel tensioned radially. But when the fabric was switched to PTFE-coated glass the supplier Birdair was unfamiliar with the luff groove system, and didn't think it would work with the stiffer new fabric.'

New fixing details eventually meant new fabric tensioning procedures had to be developed. 'It also had a serious effect on the patterning of the individual panels,' Liddell reports. 'With the software we've developed over the years for PVC-polyester fabric structures we could produce patterns for the entire roof in less than a week.

'When the switch was made, it took us two months to hand calculate the new patterns.'

Increased durability came at a price. Overall cost went up £8M, and light transmission dropped from 15% to 8%. Against that could be counted the 'self-cleansing' nature of the PTFE (Teflon) coating, a significant consideration once the Dome was seen to have a potential long term future at Greenwich.

Officially, the fabric change was a result of the new Labour Government's decision that the Dome should have the durability to last at least 25 years - although the option of dismantling the entire structure and re- erecting it on another site was never ruled out.

But the fact that it effectively defused a high-profile, scientifically dubious, Greenpeace campaign against the 'toxic Dome' was a not unexpected bonus.

Hanging loose

One of the most spectacular sights during the construction of the Dome was the team of abseilers dangling from the giant cable net. What turned out to be the largest abseiling contract ever for specialist contractor CAN began in January 1998 during the construction of the cable net, where CAN was acting as a subcontractor to Watson Steel. In the end it stayed on site throughout the remainder of the project.

CAN's abseilers were originally called in to connect the circumferential and radial cables that form the net once it had been hoisted into the air and suspended from the masts. A team of 20 abseilers was used during the operation to attach the permanent clamps.

This should have been the end of CAN's involvement, but the decision to switch from the original PVC-coated polyester roof fabric to Teflon- coated glass changed everything (see box).

As a result, an even larger team of abseilers worked on the net - at first just placing the hooped cladding clamps, which lock onto the double cables.

'We had guys sliding down the wires on ropes fixing the clamps on as they went down,' explains CAN's site supervisor Wal Thompson. 'They started at the top and worked down, with someone on the deck sending the clamps up.'

Following behind the clamp fixers was another CAN team working with the fabric erector to fix the panels. Thompson says: 'It was quite an awkward place to work. Just keeping vertical was a problem because there was nothing fixed to work from. We had to fix ropes all over the place.

'It's also very big. Once the fabric started going on it was just like a snow field with no distinguishing marks to tell you where you were.'

The value of CAN's contract with Danny reached seven figures, and it became the biggest abseil job ever undertaken, accounting for up to 85 abseilers, and 22km of safety rope.

The companies now work together regularly, and CAN has opened a specialist division for working on fabric structures.

Breathing space

Before the Dome began to take shape the elegant curves of the 35m high Blackwall Tunnel vent dominated the Greenwich site. Located within the Dome's footprint, owned by the Highways Agency and impossible to relocate, the vent threatened to compromise the purity of Buro Happold's structural concept. Its dual function as exhaust outlet and fresh air intake meant it had to be left open to the atmosphere, which in turn implied the designers had to provide a 50m diameter circular opening in the Dome's canopy.

Four main radial cables had to cross the vent opening. Without the balancing transverse forces from the missing fabric canopy panels the radials each side of the opening would be pulled outwards, relaxing the tension in the adjacent panels. To maintain the symmetry of the transverse, Buro Happold inserted an orthogonal grid of 8mm diameter cables at 1m centres, which effectively mimicked the missing panels.

By comparison, the vertical enclosure looked relatively straightforward. Unlined panels of PTFE-coated glassfibre - from Taconic in Ireland - would be tensioned vertically between vertical aluminium luff groove extrusions. But there were complications. Principal of fabric subcontractor Landrell Architectural Design Martin Hall says the biggest problem was estimating the maximum likely relative movement between the roof proper and the top fixed rail of the vent surround support structure.

'Under extreme weather conditions this could be more than 1m,' Hall reports. 'Patterning the fabric was tricky as well. The surround, a chamfered fluted cylinder, is far from symmetrical, because the opening in the roof isn't concentric with the circular enclosure boundary at ground level.'

Landrell arrived on site with its own team of fabric erectors in late 1998. Hall comments: 'Four years ago Landrell was far too small to consider bidding for the main Dome fabric contract. If anything similar comes up again, we would be ready for it.'

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Please note comments made online may also be published in the print edition of New Civil Engineer. Links may be included in your comments but HTML is not permitted.