Amey and Transport for London engineers last week finished the tricky installation and tensioning of new cables inside the stricken Hammersmith flyover, two days ahead of schedule. Mark Hansford reports from west London.
Engineers from Transport for London (TfL) and its contractor Amey breathed a collective sigh of relief this week when the last of 50 new post-tensioning cables was successfully installed and tensioned on the Hammersmith flyover.
It puts the flyover back to full loading capacity a full two months before it is called into action as a key artery on the Olympic Route Network and just five months after it was closed completely on safety grounds (NCE 12 January).
It’s been a Herculean effort by all involved.
“The engineers have done a fantastic job,” said Transport for London chief operating officer for surface transport Garrett Emmerson.
“This has been a real demonstration of what the construction industry can do,” he added. “When we leave the site in June we will have done two years’ work in six months.
“Since we took the decision just before Christmas [to close the flyover] we have had 80 people on site and probably 300 to 400 people all told working on the project 24/7.”
New cables have been installed above and below the bridge deck on the weakest spans of the 16 span structure (see box, below).
The works, which began in January, have seen around 200m of the central reservation along the flyover removed, a new structural slab and concrete barriers installed, as well as tailored anchorages for the new cables installed within the structure.
But the biggest challenge has been installing and tensioning the cables themselves.
The new cables supplement the load capacity of the existing cables, which were found to be in a critical condition where they bent as they passed over some of the bridge’s piers (see box).
Five consecutive piers towards the eastern end of the 16 span structure were found to be in need of urgent attention.
To replace lost strength 10 new bundles of cable needed to be installed over each affected pier.
The cable formation has been designed to most boost support in the immediate area around the piers so that the bundles of cable span each longitudinally – five above the deck and five below.
Of each set of five, two bundles run 33.5m along 11 of the bridge’s precast concrete deck segments, one spans 21.3m across seven segments and two span 9.1m across three segments (see drawing).
Each bundle – or tendon – is made up of 19 strands. And each had to be carefully tensioned, with tensioning of the three longer bundles carried out in overnight bridge closures so that traffic flow did not interfere with the hundreds of stress and strain gauges installed on and in the structure.
Four strands were tensioned at a time, two above and two below the deck, one on each side of the structure, to ensure the loading remained in balance.
The tensioning force that needed to be installed in the tendons was carefully calculated back at Amey’s Birmingham design centre and even more carefully applied by Armey’s post-tensioning subcontractor Freyssinet.
“We did the loading in two programmed pulls,” explained TfL project manager Chad Frankish.
The initial pull puts in 25%, or 380t, of force. “We call that the baseload and it gives us a period of assessment to ensure the flyover was responding to the loading as expected and time to make sure there is no cracking.
“We then put the remaining 75%, or 1,700t, into the structure.”
Great care is taken here to ensure the structure is not over stressed.
“From the monitoring that we were doing we knew within a really tight range how much load we have put back into each wire in each strand,” says Frankish.
“We only pulled within the modelled parameters, and nothing happened outside of those parameters.”
“And that’s why we did it in 25% and 75% pulls,” added Amey project manager Alex Gilbert. “We were checking the structure was behaving as modelled.”
Because the bridge behaved as modelled during the night time work to tension the longer cables, Amey was able to reduce the work carried out during night closures.
“Because the structure was behaving as modelled we were able to do those during the day [with traffic running],” said Gilbert. “This allowed us to speed up.”
Each strand itself is made up of seven wires, which are coated in wax oil. This will keep them lubricated and allow each and every one to be replaced should deterioration be observed in the future. “It’s a fully replaceable system; adjustment is always possible,” said Gilbert.
Tensioning of the cables finished two days early last Wednesday.
“It was a great challenge to throw at the profession,” said Emmerson. “It’s been a creative thinking challenge, a technical challenge, a construction challenge, and everybody has enjoyed that. I just hope the achievement doesn’t get lost in the melee that is the Olympics.”
The looming Olympics have certainly driven everything on this project; but costs have not been allowed to spiral as a result. The cost of the emergency repair was put at £10M at design stage, and it is set to come in near that mark; TfL is expecting an overall £13M price tag.
“I don’t think a lot of people thought we would get this done by the Olympics,” said Emerson. “It is now looking like we will.”
Emerson remains cautious because there is still a lot of work for Amey to do before the flyover can be fully reopened, and much of it is heavily dependent on good weather, which has been in short supply of late.
The work includes waterproofing of the exposed bridge deck, replacing the road surface and other ancillary works.
“Rain is now our biggest threat,” said Frankish.
Gilbert is confident though.
“We will mitigate that threat,” he said. “But we’ve still got a heck of a lot to do. We’ll tent the whole area if we need to.”
The works done to date bring the structure back to full load capacity, but TfL will tender a second phase to upgrade the post-tensioning system in the remaining 11 spans next year.
Hammersmith flyover is a precast reinforced concrete structure, with the precast concrete units held together by post-tensioned steel tendons.
The flyover is supported on a single central row of columns. It has 16 spans, 11 at 42.7m long, two of 36.6m, two of 28.8m and one of 22.6m, giving a suspended length of 622.7m. The approach ramps at either end increase the total length to 862.9m. There is only one expansion joint, 64m west of the flyover’s halfway point, which can accommodate movement of up to 360mm.
The deck is 18.6m wide overall, with a 7.3m wide carriageway and a service footway on either side of the 1.5m wide median strip.
The main structural element is a 7.9m wide hollow spine beam which varies in depth from 2m at midspan to 2.75m at the supports. It has curved undersides and was precast in units 2.6m long. Units alternate along the bridge’s length with 300mm thick transverse cantilever units. In total there are 202 units each of beams and cantilevers.
Thirteen units each make up a span, and each steel tendon runs up and down across two spans in a vertical zig-zag before anchoring back into the bridge deck.
On visual inspection, the anchorages were generally found to be in fine condition, but acoustic monitoring revealed that the critical points were just beneath this, where the tendons run in steel tubes through the upper deck. On breaking out the concrete, engineers found individual tendons were found to be completely corroded away in many places, prompting the decision just before Christmas to shut the bridge.
A temporary prop was installed to support the worst condition span in January, allowing TfL to reopen one lane in each direction to light traffic and giving Amey time to come up with a longer term solution.
That solution was to replace the strength lost with new post-tensioning cables.
The new cables are fed through specially installed concrete blocks anchored above and below the deck; in the central reservation of the highway and inside the void underneath.
To install the blocks, engineers first had to hydro-demolish the existing central reservation surface down to the top of the precast concrete segments.
Then they cast a 200mm thick reinforced concrete slab to provide a uniform surface for the concrete blocks.
At each concrete block location, 12 or 24 vertical holes were drilled, depending on the location, through the surface. Through these were threaded vertical steel bars to hold the blocks in position. In elevation, the pairs of concrete blocks were placed at each side of the piers about 2.6m, 7.2m and 13m from their centre lines, sitting on alternative precast segments adjacent to the pier.