Chisholm Trail Bridge, Cambridge

This is the last of three interesting pedestrian bridges that I visited in Cambridge earlier this year, and the most recently completed of the trio.

Also known as the Abbey - Chesterton Bridge, this structure forms a key link in the "Chisholm Trail", a new walking and cycling route promoted by campaigner Jim Chisholm which will eventually extend for some 26 km in length.

The bridge was installed in November 2020 and this part of the Trail was opened to the public on 23 December 2021. The bridge was designed by Knight Architects, and various engineering firms are cited on different websites: I understand that Atkins were first involved, then Skanska, and finally Milestone Infrastructure in the eventual design-and-build contract.

The structure spans across the River Cam, and is 44m long and up to 4.9m wide. The original budget was £4.9m, although the cost went up to £6.9m, working out at around £35,000 per square metre of deck, although some of that cost will relate to ramps and landscaping at either end. The bridge reportedly weighs 140 tonnes, and it was fabricated by SH Structures.

The bridge was shortlisted for a 2022 Structural Steel Design Award, and it's easy to see why. The design is described by Knight Architects as an "ornate lattice U-beam with a structural pattern wrapped around the deck", and the combination of complexity and clarity in the bridge's appearance makes an excellent first impression.

I approached from the north side. A little to the west of Cheney Way, a footpath connects Fen Road to the riverside path. Walking eastwards along that path towards the new footbridge, the river is dominated by a lattice truss railway bridge, which carries a fully operational railway line.

The route up onto the bridge is awkward (and visible on Google maps, if my photos don't make it clear). There are steps up from the riverside path, and a cycle ramp that doubles back with a very tight bend. The proximity of a railway level crossing rules out a direct connection to Fen Road, and it seems there's no connection through what looks like new housing, which I think is unfortunate.

Passing below the bridge along the riverside, the latticework stiffening to the bridge can be seen, and is a little reminiscent of a Pier Luigi Nervi grid-shell, a contemporary echo of the adjacent rail span. The footbridge appears to be an efficient modern design broadly in a Warren truss configuration. A closer look is puzzling, as the truss has an obvious top chord, its upper boom, but no obvious bottom chord. It's unclear how the truss forces are carried, and quite how the lattice grid works structurally.

Once up onto the bridge, the answers begin to emerge. The bridge deck is segregated between cycle and footways, which is probably sensible given the way that some Cambridge cyclists like to whizz around. This explains the asymmetry in the parapet design. There are perforated stainless steel plates filling the triangular openings in the trusses, folded over at the top to provide a stiff upper edge. On the pedestrian-only side of the deck, that edge is 1.1m above deck level, and it is 1.4m on the cycle side.

My feeling of joy in the bridge design began to take a nose-dive at about this point. The first thing that drew my attention was a gap in the top chord of the truss. Err, sorry, "truss". There are regular narrow gaps in this element, with what look like little dowel bars hidden away inside. This seemed odd. For the non-engineers amongst my readers, the top chord on a single-span truss carries a compression force, while the bottom chord (mysteriously absent here), carries tension. It's not easy to carry compression if you leave a gap though!

There seem to be two possibilities for how this bridge works structurally. One is that the dowel bars do indeed transfer compression, but the gaps are needed to reduce bending restraint in the upper structural member i.e. to act as "pins" allowing local rotation. I can only say that I wouldn't do it like that, and I don't think this is the case anyway.

The other possibility is that this is not a truss at all, and that the bridge is in fact a box girder with a curved belly - and the triangulated "truss" elements are just an incredibly expensive way to support the parapet handrails. This is what I think is the case, and I suspect it's an artefact of the design-and-build process i.e., a compromise added to make the structure easier to analyse, design or fabricate. No gaps are shown on the planning consent drawings.

I wouldn't like to be the maintenance engineer charged with trying to paint the faces inside each gap at some point in the future.

The second feature of the bridge that drew my attention is the relationship between the parapet treatment within the length of the "trusses", and its treatment beyond them. The upper edge of the truss curves down at its ends, which is structurally rational for a truss and visually coherent in any case. However, it curves down below the height required for parapet containment of pedestrians and cyclists. The upper edge line of the parapets therefore passes above the upper edge line of the "trusses" at their ends, resulting in a need for parapet elements very different in form to the perforated infill panels.

It's straightforward as a relationship between two intersecting lines, one curved relative to the bridge deck level, and one parallel to the bridge deck level. However, the detailing is quite awkward, and it feels like the parapet sections at the ends of the bridge are afterthoughts, with no obvious relationship to those in the middle.

Beyond the ends of the bridge, there's a classic bit of "architecture", that is the derivation of form from what is visually rational rather than structurally rational. This comes in the form of a series of concrete panels that continue the line of the "truss" curve downwards. There is no longer any force to be transmitted in the structure (we're off the end of the bridge at this point), but it results in yet another variation in the parapet form. It's lovingly detailed, yes, but a consequence of the formalist geometry of the bridge composition, not of anything else.

At each end of the bridge, there is a staircase down to the riverside footpath, wrapped around stepped landscaping. The bridge's concrete wing walls are a little austere, but they were remarkably free of graffiti when I visited.

Despite everything, I was surprised to find I still liked this bridge. Even if the structural form is seriously compromised, the visual identity remains attractive.

However, in a time when the evangelists for Net Zero are pushing us to make more efficient use of material (to "use less stuff", in the words of one), this probably isn't a design that anyone should wish to copy. One lesson that could be learned from this bridge is that a form that is visually attractive may be problematic to deliver if it isn't rooted directly in a sensible structural form (here: the lack of bottom chords to the truss).

If the Abbey Chesterton Footbridge tells us anything else about how to deliver a good quality bridge, an alternative lesson is simply about how to avoid letting the procurement process leading to compromise. With the increasing impetus towards decarbonisation and the pressing need for higher quality design more generally, it's clear we really need to find better ways to connect and incentivise the various participants in the construction supply chain.

Further information:

• Google maps
• Cambridgeshire Quality Panel Review, March 2016
• New Steel Construction, January 2021

Riverside Footbridge, Cambridge

This is another interesting pedestrian bridge that I found on a visit to Cambridge earlier this year.

The Riverside Footbridge spans the River Cam and its flood plain, a little way north of the city centre.

A design competition was held in 2003-4 for the site, with six teams shortlisted. The winner was Whitbybird with Gerry Judah, with Royal Haskoning as runner-up. The competition entries are visible on the Cambridge Cycling Campaign website (the unbuilt entry from Marks Barfield and Babtie is also here).

The bridge was opened to the public on 5th June 2008, and carries pedestrians and cyclists on segregated paths. To the west of the river, one long structure carries users across the river floodplain, while on the east, the bridge ramp runs southwards parallel to the river.

One thing I found really strange about this layout is there is no easy access to the riverside on the west, and no easy way off the bridge to the north on the east side. A couple of staircases would have solved both issues.

The design is at first glance one of a very particular type of white-painted steel arch structures popularised by Santiago Calatrava, and adopted by almost everyone else. In particular, it seems to owe something to Whitbybird's earlier Merchants Bridge, in Manchester.

On closer examination, it's quite a bit different from the "classic" tilted arch footbridge although clearly derived from that concept.

The tubular steel arch spans across the river onto thrust foundations. Below deck level, the slender arch member is welded to much larger angled plinths, which I suspect must work quite hard under varying thermal conditions.

The deck carrying the cycleway sits to the north of the arch, suspended from it by steel braces. Cranked horizontal cross-ribs correspond to each brace, running back below the arch to support the footway deck. The cycle deck is a closed steel box-girder, providing torsional restraint to the whole system, while the footway deck is an open-framed system.

The cross-ribs are suspended from steel bars which hang vertically below the arch at intervals. The hangers, struts and cross-ribs form a series of triangles spaced evenly along the length of the arch, a highly formal composition that then seems to determine the overall geometry of the structure. The cycleway deck forms a smooth curve in plan, but the footway deck does not, its geometry seeming much less logical. Together, it's visually complex, and to my eye a little uncomfortable.

The cycleway deck is fixed to the arch supports, and the ramp extensions are continuous, with no joints until their far ends. The approach ramps are supported on steel piers via sliding bearings. The piers on the west approach are "7"-shaped, and although they are pleasing enough, when seen together with the deck steelwork, the underside is rather messy.

The input from the sculptor, Gerry Judah, is minimal, and seems to mainly consist of two circular platforms semi-attached to the western ramp, like pimples (the designers called them "buds"). I say "semi"-attached, as they are mostly structurally independent except for small cleats which I'm guessing contribute to their lateral stability.

They feel a little like an afterthought, and the space they provide, with piddly little circular benches and pointless masts, isn't sufficiently pleasant to justify the investment that clearly went into detailing them.

If the underside of the approach ramps is untidy, the underside of the main span is just plain unsightly. From down here you get the best view of a tie bar that connects the ends of the arches and which just adds to the visual clutter.

One day, people will realise quite how bad these white tubular steel bridges look after a few years if allowed to accumulate dirt rather than being kept clean.

It's clearly a difficult site for a footbridge, with the design challenges including the floodplain, the constrained approach directions, and the need to find something for the artist to contribute. I'm quite sure I could have designed something a lot worse. But I didn't admire this bridge: there's too much going on, and the desire for crisp detailing is let down by the compromises made to hold it all together.

Further information:

• Google maps
• Steelconstruction.info
• Structurae
• Capturing Cambridge

Garret Hostel Bridge, Cambridge

I was in Cambridge earlier this year. Unfortunately, there was no public access to the city's most famous span, the Mathematical Bridge, but I did get the chance to visit two other interesting structures.

The Garret Hostel bridge is a modernist classic, completed in 1960 to replace William Chadwell Mylne's cast-iron arch (completed 1837), which had fractured due to settlement. That in turn was only the latest in a long line of bridges built across the River Cam at this site, the first supposedly a timber bridge from the late 15th century. One of the bridges, built in 1769 by James Essex, was named the Mathematical Bridge, similar in appearance to its more famous cousin a short distance along the river.

The cast iron bridge can be seen in an illustration in the collection of the National Trust. The "mathematical bridge" of 1769 is depicted in an illustration on the Queens' College website.

The bridge is Listed Grade II, which is unusual for such a modern structure. It was reportedly designed by Timothy Morgan of the architects Guy Morgan and Partners, who died in the same year that his bridge was opened, according to the bridge's Listing details. According to Atlas Obscura, the architect was intead a Timothy Guy Morgan, an undergraduate at the local school of architecture. The National Trust Book of Bridges instead claims it was Guy Morgan, born 1902. I don't know which of these is correct! The bridge was built by J. L. Kier & Co, who are also cited as the engineer. That bit seems clear.

It is an immediately attractive bridge, from almost any angle, with its striking curved concrete underside and central crease-line. Getting up close and seeing how rough the bush-hammered concrete surface is just adds to its charm. The bronze handrails are also a lovely feature.

Light on the surface of the river reflects in patterns on the underside of the bridge.

The bridge has the visual form of a very shallow arch, albeit one that is steep to climb and clearly from the days before modern accessibility requirements! However, its true form is hidden within the stone-clad abutments at either end.

It is a post-tensioned concrete girder bridge, in a shallow portal frame arrangement supported on traditional concrete hinges at its west end, and on some form of metal bearings at the east end. The prestressing cables were tensioned from the west end only, after the concrete was cast, and if you look closely at my photos, you can see the manhole cover which provides access into the hidden jacking chamber.

It is relatively early for a post-tensioned bridge in the UK, completed 3-4 years after Cavendish Bridge in Derbyshire, and in the same year as structures such as Queen's Bridge, Perth, and Bridstow Bridge.

None of this is visually apparent, but fortunately details of the bridge were published in the book Modern British Bridges in 1965, and I've reproduced them here as they are quite informative.

The drawings show the soffit of the bridge to be parabolic in curvature, and make clear how the cross-section varies in depth from the crown to the supports. The prestressing cables are arranged to resist sagging at the crown, and hogging at the ends.

The superstructure is supported on huge reinforced concrete abutments, which sit on raked concrete-filled steel tubular piles.

As is often the case with bridges pushed to the limits of slenderness, the material is simply shifted from one place to another, in this case towards the ends and into the foundations. The span-depth ratio that results is 48:1 (85 ft span vs 1.75 ft depth at crown), which is visually attractive but not especially exotic.

The drawings make clear that the stone-clad abutments, which look so nice, are essentially fake. Their front faces incline backwards away from the river, perpendicular to the soffit of the bridge to give the visual impression that they resist the thrust of an arch, while in reality doing nothing of the sort.

My view is that if fakery is to be the approach taken in bridge design, this is a great example of how it should be done well. It's a beautiful bridge, and sits nicely amongst its surroundings.

Futher reading:

• Google maps
• Wikipedia
• British Listed Buildings
• Structurae
• CanalPlanAC
• Modern British Bridges (Henry and Jerome, 1965)
• Bridges of Britain (de Mare, 1975)
• The National Trust Book of Bridges (Richards, 1984)
• Civil Engineering Heritage: Eastern & Central England (Labrum, 1994)
• An Encyclopaedia of British Bridges (McFetrich, 2019)

Welsh Bridges: 23. Dernol Footbridge, River Wye

This was the third of three bridges that I tried to visit on the River Wye in Powys in August this year.

The first was barely a ghost, and the second a literal washout. Would it be third time lucky?

I parked up in a layby on the A470, north-east of the bridge, and followed a public right of way through the fields and downhill towards the site of the bridge.

A bridge! A palpable bridge!

Alan Crow's book Bridges on the River Wye indicates that a suspension footbridge was built here in 1975 by the Newbridge-on-Wye firm N.R. Hope. From the description, it is the same bridge that can be seen today, except that all or much of the fabric may have been renewed. Powys County Council tell me that the bridge was "refurbished" in 2018.

The bridge is of essentially the same design as the now-destroyed Cwmcoch footbridge, constructed by the same builder in 1967.

Timber towers sit on concrete footings, and support a steel tube or roller over which the main cables pass. Additional stay cables are attached to the rear face of the towers.

The main cables are anchored to steel cantilever beams at each end of the bridge. Three parapet wires on each side of the bridge are also tensioned against these cantilevers, and the tower stay cables are spliced into the lower of these parapet wires.

The bridge floor comprises three timber planks running longitudinally, sitting on transverse timber members. Every third cross-member also provides the support for braced parapet posts.

Below the bridge decking, six parallel wires run longitudinally, and these provide substantial support to the deck along most of its length, as there are no vertical hangers as you would expect on a conventional suspension bridge. The deck wires are anchored to steel cross-members attached to the main anchor stanchions.

The main suspension cables only directly support the middle part of the deck. The two cables pass below grooved longitudinal timber members on each edge of the bridge.

In the absence of the under-deck wires, this would be quite an unstable arrangement, as the central portion would "rock" longitudinally under load. The deck wires are therefore essential both to prevent that rocking, and to support the deck between the points where the main cables connect.

I've included a few more photographs to illustrate both the details and the setting of this attractive, economical bridge, and a video to show how much it moves under a single person load.

Further information:

• Google maps
• Bridges on the River Wye, Alan Crow, 1995

Welsh Bridges: 19. Lôn Las Ogwen Footbridge

Not far from the Britannia and Menai Bridges, the dedicated pontist may happen upon this lesser-known footbridge.

It carries the Lôn Las Ogwen, a walking and cycling route, over the A4244 highway. The trail follows the line of the former Penrhyn Quarry Railway, which was closed in 1962.

The footbridge diverts from the original line of the railway, presumably to allow a small railway junkyard to be preserved on the south abutment of the original railway bridge.

I don't know who designed the bridge, possibly local consultancy YGC, but it was fabricated by D. Hughes Welding and Fabrications, and built by contractor Mulcair Ltd. At a guess, the main span probably doesn't exceed 20m.

At first glance, it's a steel arch bridge with a rather chunky looking parapet, decked out in the patriotic Welsh colours of green, white and red.

A second look makes clear that it is, as the fabricator says on their website, "a Vierendeel Construction with a Decorative Arch".

Opinions on this may vary. Some may note that it is just another in a long line of fake arch bridges, and hardly as egregious as some examples. Others may wonder if the emphasis on superficiality over substance combines with the colouring to act as a sly post-modern comment upon the inherent hollowness of nationalism.

I'm not sure I would go that far, but I can say that I don't like it.

Further information:

• Google maps
• D. Hughes Welding (including lots of construction photographs)