Manchester Bridges: 17. Three bridges in Marple

It's stretching it a bit to call these "Manchester" bridges, but it's as convenient a tag as any. These are essentially rural bridges, in and around the town of Marple, near Stockport (which is part of Greater Manchester).

All three bridges, and several others, can be readily reached on a morning or afternoon's walk, and one of the links below leads to suitable walking directions. All three are Grade II Listed Buildings, and there are plenty of other Listed bridges in the vicinity, including the Marple Aqueduct.

Marple's "Roman Bridge" is a pretty packhorse-style masonry arch spanning the River Goyt, near the so-called Roman Lakes. Neither the Bridge nor the Lakes date back to Roman times, it was merely a name attached to them when they became briefly popular with leisurely Victorians. The Roman Bridge itself is believed to date from the mid 18th-century, and was at one time called Windy Bottom Bridge

The engineering perspective is hardly relevant, but it's a segmental arch with a single arch ring, with the spandrels rising to a level a little below the arch crown. It currently has lightweight iron railings, although probably had masonry parapets at one stage. It spans 13.7m and is 1.75m wide.

It's a sweet little structure in a picturesque spot. As a bridge designer, it begs the question of how modern designers could emulate its charm. Simplicity isn't enough - the attraction of a bridge like this lies in its shape, its transparency, and its texture, not all of which are easy to achieve in modern materials.

Another attractive bridge in Marple is Sutton's Lane Bridge No. 1, better known as the Roving Bridge, although it is only one of several such bridges. It spans the Macclesfield Canal at its junction with the Peak Forest Canal - both canals have further roving bridges, including a brick example nearby on the latter.

It's engineer was William Crosley, and it was built around 1831. The arch is an attractive horseshoe shape.

The edges of the arch barrel are protected with metal bars at lower level, as can be seen on the photograph to the right. These are to prevent damage to the brickwork from ropes running between horses on the towpath and barges on the canal. As can be seen in the photograph (click any image for a full-size version), there has been sufficient contact over time that even the metal bar has become grooved through repeated rubbing.

The bridge gets its name as a "roving bridge" from the spiral layout of the approach ramp on one side. This allows a horse towing a barge on one side of the towpath to cross the canal and switch to the opposite towpath without ever having to unhitch the tow-rope. It makes for a very quaint, charming bridge.

Immediately north along the Peak Forest Canal, there is a flight of locks each with a small brick footbridge, ending at a span called Posset Bridge No. 18 (or perhaps Possett Bridge, depending on the source). This was designed under the auspices of Benjamin Outram and Thomas Brown, and carries a minor highway over the canal.

It was reportedly built on behalf of the local industrialist Samuel Oldknow, and named the Posset Bridge because he arranged for the workmen to be supplied ale posset from a nearby inn.

It has been widened on its north side with a metal beam footbridge, which makes it less than attractive viewed as a whole. However, on its west side, there is a short horse tunnel, in a narrow horseshoe or egg shape, used originally to allow horses to pass below the highway while boats negotiated the adjacent lock. As at the Roving Bridge, there are clear grooves on the edges of the tunnel cause by rubbing from tow-ropes. A smaller boatman's passage also exists, providing access from the lock gates down to the lower towpath.

Further information:

• Google maps: Roman Bridge / Roving Bridge / Posset Bridge
• Bing maps: Roman Bridge (hiding beneath the trees) / Roving Bridge / Posset Bridge
• Engineering Timelines: Posset Bridge / Posset Bridge Horse Tunnel
• CanalPlanAC Gazetteer: Roving Bridge / Posset Bridge
• British Listed Buildings: Roman Bridge / Roving Bridge / Posset Bridge
• Walk Around Marple - Roman Bridge and the Lakes
• Marple Locks Heritage Trail [PDF]
• Historic Roman Bridge photos from Virtual History Tour of Marple and District
• Guide to the Packhorse Bridges of England (Hinchliffe, 1994)

The Ha-Ha Bridge, Yorkshire Sculpture Park

This was one of the earliest bridges I featured on this blog, which I recently visited at its home in the Yorkshire Sculpture Park. Designed by the artist Brian Fell, it's nearly little more than a conceptual joke - the Ha-Ha embedded in the two sides of the bridge is simply the term for the ditch below.

I still like the way each "parapet" is the reverse of the other, stencil-style. It's clear that the arch does little to support the bridge - the deck is a steel plate, supported on either edge by the lower part of the fascia plate, which returns below the bridge to form a sort of channel section. Plate diaphragms are welded at regular intervals, stiffening the deck plate and stabilising the arches.

It may only be an artist's conceit, but it still has its charm.

Further information:

• Google maps / Bing maps (aerial image is very old)
• Brian Fell
• Yorkshire Sculpture Park

Mailbox Footbridge, Birmingham

Here are some photos taken on a visit to Birmingham, of a modern footbridge across a canal. It's a rather unattractive bridge, but there are several features worthy of comment.

I haven't been able to confirm the identity of the designer or the builder, although there is a very similar design by specialist footbridge manufacturer CTS Ltd not far away.

I don't know that there's a word for this type of bridge. Most truss bridges are "through" or "half-through" structures, where the bridge deck passes between the supporting truss girders. This maximises one of the advantages of the truss form, which is its ability to accommodate a very shallow depth of structure between the footway or roadway surface and the underside of the bridge. This is often required to maintain headroom below while minimising the length of approach ramps.

This bridge, which sits just to the east of the shopping, restaurant and office development known as The Mailbox, has a three-dimensional truss supporting the deck from below. It is triangular in cross-section, with the web members arranged in Warren truss form. The deck is also fully braced, ensuring that overall the truss has sufficient torsional stiffness. As the deck is supported from its two upper chords, it is also secure against overbalancing.

However, you can see on the photo on the left that the ends of the truss terminate in a very awkward way. It's unclear quite how the truss is supported, with the bearings hidden behind a covering plate. Also, because the bottom chord is a constant section and the truss a constant depth, the entire bridge looks visually "stiff", ending abruptly where you might expect the bottom chord to curve upwards.

I believe some of this is just poor design, but some of it is due to the difficult construction history - the bridge was built before the adjacent building was started. It's far from clear how well coordinated the different construction operations and designs will have been. Whatever, the reason, the ends of the bridge look very clumsy.

The bridge parapets are in a fairly standard contemporary style, with horizontal wire infill, although the boxes for the lighting units are perhaps overly intrusive.

The decking is a cheese-grater type material, visibly perforated from below but appearing very solid from above. It will be free-draining into the canal, and may therefore be partly responsible for the poor condition of some of the steelwork below.

Underneath, the structure looks particularly unattractive, with what I presume are lighting cables and power supply units dangling from the industrial walkway material and the patchy steel frame.

There are two or three other interesting footbridges over the canals in central Birmingham, which I hope to visit and then cover here when I get a chance.

Further information:

• Google maps / Bing maps
• Canalplan gazetteer
• Construction photograph

London Bridges: 10. Chelsea Bridge Wharf Link Bridge

Nestling just below the south side of Chelsea Bridge, there's a modern walkway which carries the Thames Path past the bridge and along this bank of the river (although not very far along - the Path is currently blocked in front of Battersea Power Station by construction works).

This structure was installed in 2004. It was fabricated by Littlehampton Welding, and erected by being floated along the river. The designer was Whitbybird, now part of Ramboll. The structure is about 4m wide, 65m long, and cost £650,000.

The walkway structure cantilevers out from the bank either side of the Chelsea Bridge's abutment, assisted by two extremely slender supports founded on the river bed. The structure is designed to resist uplift in flood conditions, and so that it will not collapse even if one of the two supports is destroyed by boat impact.

The walkway itself consists of a shallow steel box girder, sufficiently shallow that vibration was readily perceptible on the structure even with only one or two people using it when I visited it.

There are holes in the edge of the deck, which puzzled me when I visited the bridge, but which I gather allow concealed lighting behind to "punctuate" the line of the bridge at night.

The parapets use alternately slanting posts in a sort of "V" configuration, which I have seen on other Whitbybird / Ramboll designs, both built and planned. Presumably this adds sufficient longitudinal stiffness to justify using more slender posts than would be possible if they were vertical.

The parapets have stainless steel hand and bumper rails, and the main parapet infill comprises tensioned wires drawn through the parapet posts. There is a clear rust spot at every intersection, which may be attributable to poor fabrication (lack of paint penetration into the hole), design detailing (how could paint ever fill the hole?), or even to bimetallic corrosion (lack of an insert to prevent contact between dissimilar metals). Whichever is the cause, it will be awkward to rectify.

There's clearly no engineering reason why the walkway has to cantilever so far from the Chelsea Bridge's abutment - this seems to me a conscious design choice to facilitate the more attractive curved alignment and create a visual distance which respects the gap in age of the structures. It also minimises the sense of it being a "pedestrian subway" on a structure in close proximity to the dark underside of the bridge above, by establishing open space around the walkway rather than clinging to a wall.

This is no major footbridge, but a positive example of simple, well-organised design appropriate to its location and pleasingly lightweight in conception.

Further information:

• Google maps / Bing maps
• Ramboll
• Metalworks magazine, Corus (2004)

London Bridges: 9. Chelsea Bridge

Chelsea Bridge is the second in a trio of structures spanning the River Thames in London, all of which can be seen from each other.

Perhaps one of my more knowledgeable readers can tell me whether it is Britain's only self-anchored suspension bridge, or whether there are other examples? I'm not thinking of bridges like the Roxburgh Viaduct Footbridge, or the Royal Albert Bridge, although they are indeed self-anchored bridges with suspension systems, but structures which resemble the more conventional suspension bridge in form.

In a conventional suspension bridge, there are two (or rarely, more) towers, from which suspension cables are hung, with the bridge deck passing below and supported by (usually) vertical hanger cables. The main suspension cables are anchored into foundations which may either consists of ground anchorages, or massive blocks constructed with sufficient weight to restrain the pull of the cables. There are two significant advantages. The first is that the main structural elements are purely in tension, which allows a much lower weight of material than for elements subjected to compression and the attendant risk of buckling. The second is that the span can be constructed without the use of any temporary supports below the deck, minimising both cost and disruption to the obstacle spanned.

The self-anchored suspension bridge dispenses with the need for foundations to anchor the main cables by anchoring them instead to the bridge deck. The cost and scale of anchorage foundations can be considerable, so this seems to be a sensible approach, since the deck has to be present in any case to carry traffic, and might as well perform a second function. In practice, however, the self-anchored option is rarely, if ever, the best engineering solution. The main cable forces must be exactly balanced by a compression force in the deck, necessitating a much heavier deck than is required in the conventional option. More significantly, the cables cannot be erected until the deck is available to provide their anchorage, which in turn means that the deck must be built using extensive temporary support from below.

Megalomaniac megaprojects like the San Francisco Oakland Bay Bridge aside, this means that self-anchoring is rare, particularly for spans of any significant length. The disadvantages generally outweigh the fairly limited benefits.

Chelsea Bridge was built in 1937 to a design by Rendel, Palmer and Tritton (now High Point Rendel), with the architects George Topham Forrest and E. P. Wheeler. It replaced an earlier and much more ornate wrought iron suspension span designed by Thomas Page and opened in 1858.

The bridge is 213m long, with a 107m main span, matching the span arrangement of the nearby Grosvenor Bridge such that both bridges are easily navigated by river boats. The deck is 25m wide, with the footways cantilevering beyond the suspension cables and their towers. The entire bridge, which is now Grade II Listed, is built of riveted steel.

The main suspension cables consist of 37 locked-coil ropes tied together in a hexagonal arrangement, which is not a system I've seen previously. The deck hangers are clamped to these cables with bolted fittings. The main cables disappear into shrouds at the ends of the deck, so the way in which they are connected to the deck is not visible, which is a shame as I can't quite imagine how it works - the force from each of 37 individual cables has to be transmitted into the end of the main deck girders.

The main towers are in riveted steel box construction, tapering towards the top, where there are exposed saddles. They have something of the shape of Cleopatra's Needle. Most suspension bridges require cross-bracing between the tower legs to provide stability, but the span of Chelsea Bridge is short, and hence the towers aren't tall enough to require this. I think it looks good and wonder what size of suspension bridge renders it impractical.

The towers are hinged at their base, using heavily stiffened rocker bearings, which again is not something normally associated with suspension bridges. On a conventional suspension bridge, the towers have to be stable to support the main cables before the deck is in place, and hence are cantilevered rather than hinged.

I think the largest suspension bridge with hinged towers may be the Florianopolis Bridge, which has a 340m main span. There, the designer, David Steinman, saw the use of hinges as "the most economical and scientific design for suspension bridge towers", because of the reduced bending stresses in the permanent situation, but although Chelsea Bridge adopted the same choice ten years after Florianopolis, it's not a form that has prospered.

From below the deck, its structural form can be seen clearly. The main girders are formed of paired girders, with crossbeams and cantilevers at regular intervals, and extensive lattice-member bracing connecting the other members.

This is the below-stage machinery supporting the dramatic performance above, not really intended for public viewing. It all looks surprisingly well-maintained.

The good level of maintenance is evident throughout the bridge. The paintwork is all good, including the red on the cables and the elements of blue on the parapet infill (an improvement for local Chelsea residents over the previous red-and-white scheme, which reminded them uncomfortably of the colours of rival London football team Arsenal).

The cables and towers are studded with funfair-style lightbulbs, and the main streetlamps are in an unusual arrangement, with the posts carefully set around hanger cables, as can be seen in the photo above right if you look carefully (as always, click on any image for the full-size version).

In his book Cross River Traffic, Chris Roberts describes Chelsea Bridge as "a very striking, if odd, combination of Thunderbird One and seaside pier", which isn't far off the mark.

Further information:

• Google maps / Bing maps
• Wikipedia
• Structurae
• Engineering Timelines
• British Listed Buildings
• Where Thames Smooth Waters Glide (includes images of the previous bridge on the site)
• Civil Engineering Heritage: London (Denis Smith, 2001)
• Cross River Traffic: A History of London's Bridges (Roberts, 2005)
• An Encyclopaedia of Britain's Bridges (McFetrich, 2010)

London Bridges: 8. Grosvenor Railway Bridge

The first of a trio of bridges all within spitting distance of each other is the Grosvenor Bridge, which carries the railway lines out of Victoria station across the River Thames in London.

A bridge was first built here in 1860 as part of the Victoria Station & Pimlico Railway. Designed by John Fowler, the bridge cost £84,000 and carried two railway tracks over four 53.3m river spans (plus a number of approach spans). Masonry-faced brick piers carried segmental wrought iron arches, with six arched ribs in each span.

As the railways rapidly expanded, so did the bridge, with a first widened section completed in 1866 (designed by Sir Charles Fox), adding another 5 tracks on the east side. The structure was widened again in 1907, adding a further 2 tracks on the west side, using mild steel arches.

Due to the cost of maintenance, it was decided in 1958 to replace the bridge, with work taking place between 1963 and 1967, to a design by Freeman Fox (now part of Hyder, and founded by the same Charles Fox mentioned above). The bridge now consists of ten parallel bridge structures supported on common piers. The arch profile and general elevation of the bridge has remained largely unaltered since the original construction, although reconstruction and underpinning of the piers has shortened the spans to 50m each.

Throughout construction of the new bridge, the contractor had to maintain eight live tracks of railway traffic, and two of the four river spans also had to remain open to boat traffic at all times. This led to a very complex construction sequence.

The foundations consisted of a concrete raft (original bridge), four cast iron caissons filled with concrete (the first widening), and a single cast iron caisson filled with concrete (the second widening). These supported piers variously faced in limestone (orginal) and granite (widenings). During the reconstruction, the existing foundations were isolated within a cofferdam, and widened in reinforced concrete. This had to be completed in small sections as access became available.

A temporary steel truss was erected above the bridge, and used both to dismantle the original arches and erect the new arches, lifting the elements onto and off of pontoons. This worked one span and one track at a time, which is why the new bridge is built of several separate decks rather than one continuous structure.

The current arches are two-pinned welded steel box girders, each 1.13m deep and 0.61m wide, with steel plates up to 31mm thick. There are two arch ribs for each bridge deck. Tubular steel spandrel posts carry an orthotropically stiffened steel deck. Further details, construction photographs and technical drawings are available in the book Railway Bridge Construction by F.A.W. Mann.

It's an essentially modern design, but I doubt that any casual observer is aware of it. The retention of the original open-spandrel arch elevations means that the bridge appears at first glance to be an unaltered historic structure.

The visual design of the piers is difficult on a bridge like this, particularly where they have been widened from the original and risk looking very bulky in comparison to the steelwork they carry. They also have to look appropriate at widely varying water levels (the Thames is a tidal river at this point). I think the chess-piece shaped profile works well, particularly the way the pier face curves outwards at the top to meet the arch springings.

The spandrel columns are remarkably slender, but the transparent effect you would see on a simple elevation of the bridge is nowhere apparent, due to the sheer number of columns and the shading effect from the deck.

Further information:

• Google maps / Bing maps
• Structurae (not all details given are correct)
• Wikipedia (similarly, much of the article is incorrect)
• Engineering Timelines
• Where Thames Smooth Waters Glide (includes archive photos and drawing)
• Railway Bridge Construction (Mann, 1972)
• British Railway Bridges and Viaducts (Martin Smith, 1994)
• Civil Engineering Heritage: London (Denis Smith, 2001)
• An Encyclopaedia of Britain's Bridges (McFetrich, 2010)

Footbridge Awards 2011 - technical above 75m span

Right-o, it's time to (briefly) cover the final batch of designs shortlisted for this year's Footbridge 2011 awards, this time the long-span bridges in the technical stream. Thanks once more to Frame and Form for posting all the images, and as on previous posts, you can visit their website to see the pictures, which I won't repeat. You can also now simply click here to see all my posts on the 2011 shortlist.

I've previously commented on four out of the six bridges, as they also featured in the aesthetics category: Center Street Bridge, Des Moines; College Bridge, Kortrijk; Passerelle la Defense, Paris; and Grimburg Harbour, Gelsenkirchen. Of those, the College Bridge and Passerelle la Defense are the technical standouts, I would say.

The other two candidates are the Forthside Bridge, Stirling, and the Kurilpa Bridge, Brisbane. Both are a tangle of masts and cables, and essentially a variation on the classic cable-stayed bridge. Forthside is what is popularly called an "inverted Fink truss", while Kurilpa is a simplified tensegrity structure, but both rely on the successive cantilevering of cables from main support pylons.

I've discussed both Forthside and Kurilpa here before. Both are technically impressive, although to some extent only in how they resolve the problems of their own making, as the architectural concept determines the structural form as much as arising from it.

The winners in each category will be announced on 6th July.