Upgrading work at the Golden Valley interchange caused such a stir that the Highways Agency set up a dedicated web site (www.m5junction11.co.uk). It was seen as a trial, but in the few months since going on line it has already had more than 3,000 visitors.
But being on line is not the only thing that makes this project one of the most newsworthy and challenging jobs in the country.
Strengthening work involves an ingenious scheme to replace the support columns of the main bridge structure. The bridge decks are required to be raised nearly 200mm while keeping both the busy A40 and M5 open to traffic. The situation was further complicated earlier this year when thaumasite sulphate attack was discovered in the concrete pile caps and columns.
The Golden Valley interchange carries over 100,000 vehicles a day on its three tiers which take the busy A40 from Cheltenham to Gloucester over the M5 motorway. Between the two trunk roads is a roundabout interchange. During an inspection of the A40 structures in 1994, as part of the 40t working load assessment to bring Britain's bridges into line with European regulations that come into force on 1 January, it was discovered that the central bridge of the interchange was severely deficient in shear resistance.
'Longitudinal concrete I-beams running through the deck were found to have insufficient shear resistance over some of the support columns,' says Highways Agency project manager Martin Hobbs. 'The investigation discovered that all three bridges were deficient, but the problem was much worse in the larger central structure.'
The Highways Agency granted Gloucestershire County Council a departure from standard highway operating requirements. This limited the bridge loading to 3t vehicles and single lane traffic.
The local outcry at the thought of prolonged disruption to the interchange prompted the Highways Agency to open a dedicated project web site where visitors can find pictures of the scheme and details of work. 'The project has been in the public eye from the moment the weight limit was applied,' says Hobbs.
One option for strengthening work was to replace the structure completely. This was rejected as it would have interfered with M5 traffic, and re- routing the A40 traffic via the interchange roundabout risked creating tailbacks.
Heavy temporary props at the abutments and islands were considered but would have cost as much as upgrading the existing bridge supports and would only have provided a temporary solution. Permanent upgrading work would still be required. It was therefore decided to upgrade the existing bridge supports. Consulting engineer Halcrow was awarded the contract to design the upgrades on all three structures.
The bridges consist of post-tensioned cellular continuous concrete decks split into halves longitudinally. Each half consists of six longitudinal concrete I-beams with transverse beams every 5m. The central bridge is supported by two sets of four square vertical columns which divide it into three spans of 30m, 44m, and 30m. Two sets of 12 concrete A-frame support piers split the east and west bridges into spans of 18m, 36m and 18m.
The deck units sit on a concrete cross beam running along the top of the support columns. It is the position of this beam that led to question marks over the bridge's ability to resist shear. The outer edge of the cross beam, where there is greatest shear, did not line up with the web of the outer longitudinal beam inside the deck. This meant that the line of greatest shear ran through the flange of the deck beam rather than the shear resisting web.
'The problem is exaggerated by post-tensioning tendons running through the longitudinal deck beams,' explains Halcrow principal engineer Ron Chambers. The internal tendons resist the tensile forces generated by bending. Their profile follows the line of greatest tensile forces. This carries them through the top of the beams at the support piers where there is hogging and the bottom of the beams at mid-span where there is sagging. 'The tensioning of the tendons by pulling them at either end causes them to try to straighten. This exerts a downward force over the support columns and adds to the shear.'
On the less problematic east and west bridges the solution has been to introduce larger bearings under the deck soffit. The existing cross beam and A-frame supports were widened by 500mm at either end to carry Glacier (now Federal Mogul-Sollinger Huette) Dualign K-series rotation bearings directly under the longitudinal deck beams. Widening the piers involved breaking back the outside concrete face of the pier legs, adding reinforcement, and drilling dowels into the pad foundations and then casting a 500mm concrete thickening to the side of the legs and cross beam.
The problem was much more difficult to overcome on the central bridge, where the span is longer and the loads greater. The solution has been to reduce the shear forces at the supports by shortening the spans.
To achieve this without interfering with the M5 traffic, Halcrow came up with a design to replace the 10.5m vertical concrete columns with 12m long splayed steel support columns in the shape of a huge V. Rising from the same pile cap as the existing columns, the tubular steel piers support the deck directly under existing transverse beams inside the deck, 11.5m apart.
'The objective was to reduce the loads at the supports,' explains Halcrow resident engineer Vince Read. 'The centre span is reduced from 44m to 33m and the outside spans are reduced from 30m to 24m. Shortening the spans not only reduces the shear forces at the supports but also reduces the bending stresses.'
Two sets of six V supports replace the eight vertical columns. A level concrete cross beam along the top of the supports widens the piers so that a new bearings can be placed directly under the longitudinal deck beam webs. The maximum reaction at any bearing will be 200t.
The V supports, fabricated by Kvaerner Cleveland Bridge in Darlington, are connected by a steel tie beam at the top. Once in location they will be filled with C50 concrete to provide stiffness. They are bolted into a concrete plinth cast on the pile cap.
One of the problems raised by adding the larger 100mm-120mm bearings and columns is getting them into place. The existing 30mm gap between the deck soffit and support was too small and the bridge decks have to be jacked up. Six separate lifts of up to 190mm were needed to raise the level of the three structures, carried out at night to minimise disruption to the M5 traffic. Each lift took three nights.
Main contractor Jarvis Construction used state of the art computer controlled jacks for the operation. Each lift was carried out by specialist subcontractor Hydratight using a series of 60 pot jacks per lift. 'The movement of the jacks is continuously monitored by a computer,' explains Jarvis project manager Ahmed Esmaail. 'This regulates the jacks to within 0.1mm and ensures an even vertical lift over the whole structure.'
'The system is able to raise at a rate of 1mm per minute,' he continues. 'However because the jacking process has to be continually interrupted to shim up each jack as it reaches maximum extension, the actual jacking rate is about 10mm/hour.' The largest lifts are the two central bridge half decks, which each weigh 2,500t. Each of the jacks is doubled up in case of a failure.
'The deck halves are designed to counterbalance one another and stabilising them during jacking was a problem,' explains Esmaail. 'We used a restraining system which stopped the decks from sliding by attaching them to the abutments.' The system was designed by consulting engineer Tony Gee & Partners.
Another problem came with the discovery of thaumasite sulphate attack on the interchange pile caps and support piers. 'We found thaumasite sulphate attack on all the foundations backfilled with Lias clay,' explains Hobbs. 'The damage is variable. In some places it was down to the reinforcement.'
Jarvis recast the affected areas on the pile caps, and replaced the problematic Lias clay with a granular fill and a new drainage system. The affected A-frame supports have been exposed and checked but Highways Agency engineers have held back from carrying out remedial work while waiting for the outcome of a current government investigation into the problem (NCE 12 November). Visitors to the Golden Valley website will be watching developments with interest.