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

Bridges | River Usk a joint solution

river usk IMG 20170331 013524428

The bridge carrying the M4 over the River Usk in Newport, South Wales is on first sight unremarkable, but the same cannot be said about work being carried out to repair it.

The bridge is a 143m long five span reinforced concrete structure. Built in the 1960s, its looks are typical of the era, belying its role as one of the most important bridges in Wales.

The bridge is on the main arterial route into Cardiff on the M4 motorway – to close it for any length of time would be political suicide and create untold havoc on the surrounding local road network.

river usk WP 002531

river usk WP 002531

The twin deck bridge crosses the River Usk

So when problems were found to be developing with its structure, some innovative thinking was needed from consultants WSP and main contractor Costain as they sought to find a way of halting the damage.

The problem lay with the bridge’s thrust hinges, a proprietary product designed by the consultant Sir Owen Williams & Partners. The joints were put at points of contraflexure – points where there are zero moments. The joints were designed to let the deck bend and flex while carrying both thrust and shear.

To do this, when the bridge was built, 40mm wide and 430mm deep slots were cut into the top and the bottom of the deck leaving a thin strip of concrete 430mm thick in the middle. To carry the shear and the thrust, this strip was then reinforced with 32mm diameter scissor bars: two pairs of six, across the joint – three top, three bottom – with three horizontal straight bars running through the centre. In theory, when the deck hogged and sagged, the movement would be accommodated by the gap.

But due to poor construction, the slot was not wide enough in places, and as the deck flexed, the concrete at the top and on the bottom was crushed by the compression force of the two sides coming together. This in turn delaminated the concrete at the plane of the reinforcement.

As the deck flexed, the concrete at the top and on the bottom was crushed by the compression force of the two sides coming together

The resulting exposure of this reinforcement not only left it open to the elements and degradation, but more critically it meant it was not being securely held in place by the concrete, leaving it susceptible to pull out and making the joint likely to fail in the future.

A sensitivity analysis into the length of anchorage bond required was carried out by engineers from WSP. It identified that in the worst case scenario the capacity of the tapered deck support beams which span between the piers dropped rapidly after a critical length was reached.

“In the analysis, we looked at what happens when you lose cover or anchorage to that rebar,” says WSP senior engineer in civil, structural and ground engineering Richard Owen. “We found there would be a shear failure at the back of the thrust hinge detail due to that loss of anchorage.

New insitu anchor block and hangers for tendons.jpg

New insitu anchor block and hangers for tendons.jpg

New insitu anchor block and hangers for tendons

“Essentially, although the thrust hinge is strong in itself, and the beam reinforcement is quite strong, it’s the interface between the two that we’re concerned about.”

With this threshold not yet reached, the Welsh government decided to take action to prevent a future collapse of the bridge.

But repairing the joints brought about a whole new set of challenges.

Part of the problem stemmed from the bridge’s unique form. There are around 100 thrust hinge bridges in England and 15 in Wales and the deck for these structures was formed from a flat slab, with the reinforcement for the joints spread evenly across their width. In this case, the deck slab spans between tapered beams which span between the piers, so the hinges were restricted to the beam widths.

river usk hinge rebar perspective view

river usk hinge rebar perspective view

Computer generated image of hinge reinforcement

To repair the Usk bridge joint properly, concrete surrounding the damaged area had to be removed and reinstated, but this would have compromised the strength of the joint and without additional support would have caused the bridge to collapse.

With two hinge lines in the central span and one in each of the second and fourth spans, the challenge of accessing the underside of the bridge to carry out the repairs, while keeping the bridge open, had to be overcome.

Ordinarily, the job to repair the beams would have been relatively straightforward. A series of temporary props could be built from the ground underneath to provide an alternative support to the deck, then the damage to the joints could be repaired.

But this was not possible on the Usk bridge as it crosses a protected Site of Special Scientific Interest (SSSI). Building props in the river was considered to be too dangerous and intrusive. The team had to find another repair method.

Urgent work

With the added complication of the urgency of the work, conventional design timescales were dismissed. A working group comprising the Welsh government and a team of WSP structural engineers was formed to examine the feasibility of the various technical solutions. Throughout the process, consultant Atkins provided the regulatory checks on the design.

The team came up with a scheme to post tension the bridge to put an additional compression into the beams and increase the capacity of the joints while the work to repair them was underway.

“How do we repair the beam without propping it?” asks Owen. “How do we enhance the capacity of the beams at the hinges and increase the shear resistance as well?

We looked at how the forces from the anchorages would affect the structure as we were jacking tremendous forces into localised areas and how these would dissipate into the decking and the slab

Richard Owen, WSP

“We might be able to do that if we make the strut and tie component of that [the joint] much shallower so we reduce the criticality of the length of the rebar. That will allow us to adopt the traditional method of concrete removal and repair.

“The only way we could do that was to introduce external post tensioning.”

But with this solution came another raft of issues which needed to be taken into account. At this point French specialist Freyssinet came on board.

The first issue was that the post tensioning could potentially buckle the structure. “However, after looking at various modes we decided that wasn’t an issue,” says Owen. “Then we looked at the upper limit of the compressive stress over the hinge and if we’d crush or burst it.

“Then we looked at how the forces from the anchorages would affect the structure as we were jacking tremendous forces into localised areas and how these would dissipate into the decking and the slab.”

What we’ve gone for is keeping everything internal, it’s invisible to the external elevations, it’s away from the weather and it’s out of the way of shipping

Richard Owen, WSP

Positioning of the tendons locally around the joints was ruled out due to the number of anchorage points which would need to be built, as well as the risk of putting unwanted tension into the beams due to the stiffness of the piers.

Therefore the team decided to run the post tensioning tendons along almost the entire length of the bridge from pier one to pier four. Pairs of cables are being run in the three gaps created between the tapered downstand beams on the underside of the bridge. The need for tendons on the outside of the structure was avoided by putting a higher force in the pair in the outer gaps (390t) than in the middle gap (240t).

“If we put them on the outside [edge of the structure] we would have had to build some corbels on the external face and then to get the anchorage for the corbels, you’d probably have to transversely post tension through the beam which would have been a whole new design solution,” says Owen.

Usk bridge

Usk bridge

“What we’ve gone for is keeping everything internal, it’s invisible to the external elevations, it’s away from the weather and it’s out of the way of shipping.”

Areas of concrete around the new anchorage blocks had to be broken out and new reinforcement tied into the structure to ensure it was strong enough to take the required forces.

The sequence to carry out the tensioning of the cables was further complicated by the fact that the east and west carriageway are two separate superstructures but are supported by the same substructure. Therefore if one side was stressed before the other, torsional forces would have been induced leading to a potential failure of the deck.

But before the tendons could be stressed, the team hit yet another obstacle.

The problem is we’re putting in post tensioning to within an inch of its life without bursting it

Richard Owen, WSP

An expansion joint between the bridge deck and the deck supported by the approach viaduct was also not sufficient to accommodate the thermal movements of the bridge. To this point the additional compression induced in the bridge by the contact between the two decks had benefited the structure having the same effect as the post tensioning. But with the additional force from the post tensioning, the concrete in the joint could have been crushed.

“Some of the monitoring work we did a few years ago illustrated to us that the structure ceases to move freely at certain temperatures,” says Owen.

“The problem is we’re putting in post tensioning to within an inch of its life without bursting it. Then you get the glorious summer that we always have, temperatures rise, and the structure locks up and we get this additional longitudinal compression introduced into the deck and suddenly it goes pop.”

Avoiding costly temporary works

The task to widen the expansion joint was a job in itself. Because the reinforcement appeared to have been cast too close to the edge of the deck, it needed to be cut and bent back to create a new edge. But to get enough purchase on it to enable it to be bent, 250mm of concrete had to be cut away off the end leaving a large hole in the carriageway above.

“In breaking all of this out, there’s a big hole in the M4,” says Owen. “Freyssinet has brought some clever ideas into play to avoid some very costly temporary works which would otherwise disturb the traffic.”

The use of hydrodemolition for the break out work meant that contractors were confined to night works only. This programme to cut away the concrete and reinstate the joint added added another layer of complexity to the already technically challenging job.


The next challenge was access to the underside of the bridge to carry out the works. The team planned to suspend access scaffolding from the bridge, but the analysis of the bridge showed there was no spare load capacity for the works. More worryingly for the Welsh government, it also showed that technically the bridge was incapable of carrying the heaviest vehicle loads.

Again the team had to go back to first principles and prove that the factors of safety and factors applied to the materials could be reduced and therefore release more load capacity.

Using a dynamic amplification analysis which looked at the response of the structure under particular moving load cases, WSP demonstrated that the load factor could be reduced. To prove the material factors could be lowered, WSP took core samples and measured their self weights and compressive strengths. It also undertook an extensive survey of the bridge’s dimensions to ensure the elements did not differ significantly to those used in the analysis. On top of this, the team carried out a statistical analysis to prove the sample set was robust and reliable.

river usk deck night view 1

river usk deck night view 1

Work had to be carried out at night

The hard work paid off and the team was able to prove the factor could be reduced from 1.15 to 1.05, releasing enough capacity for the heaviest highways loading with enough left over for a scaffolding system to be hung from it.

Despite the additional capacity, the limits on the structure were still extremely tight. WSP designed the scaffolding system to have different zones with separate strict weight restrictions.

“The assessment that took place necessitated a strict management plan,” says Owen. “The scaffold is set up in zones which are carefully controlled. Some are personnel only and then there are some which are directly over the piers so the load path doesn’t go through the superstructure and materials can be stored there.”

Reinstating the joints

With the repairs to the concrete underneath the joints now in hand, the team had to make sure the joints were reinstated to their original design. Despite the slot running the width of the carriageway at the top and the bottom of the deck now being opened out, flexible sealants to stop water ingress into the joint could still not be installed.

“When the deck was built, little effort had been put into the keeping the gap straight, smooth or vertical,” says Owen. “We’d like to have a buried joint, but they don’t tolerate curves or differences in level. You need a nice square edge.

“Therefore we have to reconstruct it to an acceptable quality.”

With no opportunities to shut the bridge during daylight hours and no contraflow allowed due to the proximity of the Brynglas tunnel immediately west of the bridge, all work above the deck had to be carried out during overnight closures from 8pm to 6am.

Tight limits and heavy supervision are being applied to the construction of the new slot to ensure the mistakes from the original construction work are not repeated. Work was scheduled to take place along side other planned work on the adjacent Brynglas tunnels. There, Costain in partnership with Capita is in the midst of a major two year programme to bring mechanical and electrical systems up to standard. This work is due to end in January 2018. N



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.