14 May 2017

Bath Bridges: 3. Destructor Bridge

This is the third and last bridge that I visited on my recent trip to Bath.

The original Destructor Bridge was built as a railway bridge in 1870 by the Midland Railway Company, at an entirely different site. In 1905, the railway bridge was relocated to the current site, and nicknamed the "Destructor Bridge" after the nearby "Destructor", an incinerator. The present-day Midland Bridge, a road bridge, now occupies the original site.

This part of Bath is experiencing significant redevelopment, so far as I could see mainly to build new apartment blocks, the easy-money choice prevalent throughout the country. The replacement of Destructor Bridge was ordered as a planning condition for the Bath Riverside development. The new structure was designed by COWI and Knight Architects, and built by Britannia Construction, with steelwork by Cordioli.

The bridge is an absolute delight.

Its layout is asymmetrical, although this is not apparent from a distance. The 6.7m wide roadway sits between two footways, with the western footway/cycleway being wider (4m) than on the east (2.5m). Beneath each edge of the road there is a steel box girder, with the road deck supported on cross-beams laid out ladder-style. The western girder is the larger of the two, and its load capacity is augmented by a shallow steel arch. The bridge deck is constructed in orthotropic steel plate, a series of welded troughs. Anti-pigeon fillets have been added to the cross-beams.

The arch is one of the elements of the bridge that I found most pleasing. The arch rib is a super-slender steel box, more of a ribbon than a conventional box, and the arch hangers are simple flat steel plates of the same width as the arch. It's a stark and beautiful piece of architecture, and will have been very difficult for the engineers to achieve. It exposes the sheer laziness of so many similar bridges.

I suspect that it's made possible by putting more reliance on the deck girders than is immediately apparent. I think they are carrying a greater share of the load than is usually the case on an arch bridge, and most importantly the stiffness of these girders is perhaps relied upon to the restrain the arch against in-plane buckling.

Most designers put the stiffness in what they see as the main load carrying member in the arch, or share it fairly equally between the two. The precedent for Destructor Bridge's boldness might be traced all the way back to the deck-stiffened arches of Robert Maillart, and indeed the full-width hanger plates remind me of the leaf piers adopted on several of Maillart's designs.

The second box girder is therefore serving two main functions: it carries a share of the highway and footway loads, but mainly it reduces the torsional effects on the primary box girder.

Unfortunately, the arch is the source of what is at present the bridge's biggest problem. The shallowness of the rise, and the flatness of the cross-section, present a compelling invitation to anyone wishing to climb the arch, and some minimal anti-climb plates and signs fail to offer serious discouragement. Inevitably, the bridge has been plagued with visitors walking across the arches, even doing handstands, and I'm surprised not to read any mention of skateboarders and unicyclists giving it a try.

Temporary barriers have been erected around the two ends of the arch, but these were clearly ineffectual and a security guard is now present on the bridge at all times. The problem is inherent in the design, and I've seen other steel arch bridges deal with it in different ways. A bridge in Nottingham has a smoothly tapered "steeple" detail on the arch, while one in Sheffield has anti-climb "blisters". Neither solution is compatible with the Destructor Bridge design.

The same issue arose (just as predictably) on the Tradeston Bridge in Glasgow, where it seemed to be something of a novelty, a craze that just went away over time. I'm not clear whether any different anti-climb solution is under consideration on the Destructor Bridge, or whether the same process of gradual loss of interest will apply.

The low-speed nature of the road carried by the bridge has been used to good advantage to dispense with the vehicular parapets normally mandated by standards. Instead, low-profile barriers are used to keep vehicles from mounting the footways. This allows the main edge containment to be in the form of pedestrian balustrades, with the western one slightly higher than the east, as a cycle path is accommodated on this side (although not presently marked out on the paving).

The parapet design echoes the arch, with a series of flat "blades" supporting a timber handrail (with integral lighting). The taller parapet is mounted with an additional stainless steel top rail. The effect is visually attractive – transparent when viewed directly, yet solid when viewed obliquely, and I was interested to see that although visually the parapet looks quite massive at an angle it still "feels" transparent. The device of attaching the blades to the outside face of the deck edge beam is also interesting, and quite different to what is normally done to make deck edges appear thin.

I'll finish with a few of the bridge's interesting details.

The bridge abutments are massive concrete constructions, but sloped back to follow the riverbank. The corners are faceted and the front face inscribed with parallel recesses, both of which serve the purpose of breaking up what would otherwise be a large mass. The recesses also discourage graffiti.

The central part underside of the bridge is left largely untreated, but the two edge sections have been hidden behind "cheesegrater" cover plates. Service pipes run below the deck in these areas, and the covers serve both to hide them visually and to reduce the risk of vandalism.

I don't really like this detail: it makes the bridge deck seem far bulkier than it really is, and it will surely hinder future inspection and maintenance.

The final detail worth noting is the benches which have been inserted between the arch hanger plates. These are a very nice feature, minimal and entirely in keeping with the geometry of the rest of the bridge. Soft lighting below the benches is an effective treatment, although unfortunately some of the lighting seems already to be broken.


Further reading:

03 May 2017

Bath Bridges: 2. Victoria Bridge

This is the second of three interesting bridges that I visited in Bath. As with Pulteney Bridge, the Victoria Bridge is also very much a heritage site, the oldest surviving bridge by brewer-turned-bridge-builder James Dredge, and a highly unusual design.

I've previously visited the Dredge bridge in Inverness, which is a shadow of its original self, and would like to re-visit the Bridge of Oich some time so I can present it here. Dredge's designs were and are unique, according to his patented "taper principle", whereby links in the supporting chains are progressively diverted downwards to pick up the deck, so that the force and capacity of the main chains both diminish with distance from the supporting towers. In this respect, it more closely resembles a cable-stayed bridge than a suspension bridge.

When I briefly mentioned Victoria Bridge in 2012, the bridge was scheduled for substantial repair and refurbishment work, on account of its poor condition. Many of the bridge's metal elements were cracked or at risk of fatigue damage, and it was both prone to vibration and at risk of damage if subjected to severe crowd loading.

Since then, the refurbishment project has been completed, and I think all involved can be congratulated. However, the works are not without a degree of significant loss, and the line between restoration and vandalism is a fine one in this case.

You can read more about the project in two thorough and informative reports by the restoration engineers, AECOM (see Further Reading for links, below). As part of the works, a thorough evaluation of the bridge's history was undertaken by local engineer Alf Perry (who sadly died in late 2016), although I believe this is so far unpublished.

I think that more than half of the original bridge ironwork has been replaced in modern steel, although actually much of the original fabric had been replaced during a series of previous interventions. All the primary load-carrying elements in the deck appear to have been replaced. Transverse tie-bar elements below the deck appear to be largely intact, although in reality these provide little support and are now largely decorative.

The structural changes both improve the bridge's strength and also reduce unacceptable vibrations, and may be justified in terms of providing a structure with greater longevity.

The deck edge stringers are connected with fishplates, which appear riveted from the outside, but which are actually connected with dome-headed tension control bolts, often used in refurbishment work to give the appearance of rivets. I'm not sure this is justified in this instance for visual reasons, but it's a perfectly acceptable engineering solution.

Large parts of the main bridge chains have been retained, but large parts have been replaced in steel. The detail at the lower ends of the inclined chain hangers is unfortunate, with the flats of the hanger links welded onto threaded rods, and some very modern turnbuckles used on the back-tie elements. Perhaps these were unavoidable, but I thought they looked inappropriate.

By far the most significant alteration to the bridge is that the hanger pattern has been substantially changed, as can be seen by looking at any images of the bridge from before its refurbishment. As built, the structure had an arrangement which had not survived on any other Dredge bridges. Where the hangers diverge from the main chain, at each chain node there were two hangers connecting to two separate deck nodes, creating a series of triangles. On later Dredge bridges, the twin hangers connected to the same deck nodes, creating a series of parallelograms. The reconstructed bridge now adopts the later Dredge design.

I don't know how contentious this change was when design proposals were presented to the local authority and to English Heritage. The engineering justification is straightforward: the original arrangement led to alternating compression and tension in each hanger, resulting in slack hangers and significant stress ranges in the metalwork, which presumably contributed over time to significant fatigue damage. In the new arrangement, nearly all the hangers remain in tension under all loads, greatly reducing the risk of further damage, and also introducing redundancy at each deck node (now supported by two hangers rather than one). However, it seems an enormous shame to have eliminated the bridge's one most unique feature.

The updated bridge now incorporates lighting in the parapet handrail (on one edge only), and seems likely to be well-set for a considerably extended lifespan.

I was able to perceive vibration in the bridge deck, but not at an unacceptable level. Pleasingly, the low headroom at the north end of the bridge over the river towpath has been tolerated without the addition of bumble-bee markings, although the transverse ties are at a level which could easily be struck by taller pedestrians.

Notwithstanding various misgivings, the refurbishment of Victoria Bridge is an impressive accomplishment, and one that those involved should be very proud of.


Further reading:

01 May 2017

Bath Bridges: 1. Pulteney Bridge

I spent some time in Bath recently to attend an IABSE conference on the subject of "creativity and collaboration". It was a very enjoyable conference, with plenty of material relevant to bridge designers, but I doubt I'll have time to capture any details here.

However, I did get time to visit three of the city's most significant bridges, which I'll feature across my next three posts.

First up is probably Bath's best known bridge, the Pulteney Bridge. This is one of very few historic inhabited bridges still existing in the UK, although at one time there were many, with fine examples in London, York, Durham and Newcastle, to pick just four.

The survivors include the High Bridge in Lincoln, a somewhat unprepossessing structure in Frome, and the well-known house on a bridge at Ambleside.

Approached from the south, Pulteney Bridge is an impressive structure, its three arches sitting astride the River Avon, a splendid backdrop to the crescent-shaped river weir. When first built, it connected the city of Bath to the Parish of Bathwick, an estate owned by the Pulteney family since 1726. In 1769, William Pulteney advanced plans to replace the existing Bathwick ferry with a permanent toll-free bridge, petitioning for the first of two Acts of Parliament.

The initial plan for a three-arch masonry bridge was prepared by Thomas Paty. This did not include the rows of shops which are now the bridge's most notable feature. Robert and James Adam became involved in the project in 1770, and progressed on the basis of an inhabited bridge, a concept which at the time may have been considered a relic of the past, with the well-known British examples being mediaeval in origin. Indeed, Bath Corporation objected to the proposal: "It has been for some years past an uniform practice throughout the Kingdom to avoid and condemn incumbrances of this kind".

Robert Adam's proposals owed much to an unused proposal by Palladio for the Rialto site in Venice, although more practical and less extravagant. Adam designed an entirely symmetrical structure, supporting twenty-two shops, each with an attic and some with cellars hidden within the arch spandrels (the circular cellar windows are a highly visible feature). The bridge was completed in late 1773, at a construction cost of £8,183.

Significant changes to the bridge were made in 1792-1794, to a design by Thomas Baldwin, increasing the height of the shops, knocking some shops spaces together, adding second storey windows and making other alterations. Author Eric de Maré commented on the result of this and other works that "the original house part has unfortunately been pathetically travestied by alterations".

In 1799 and in 1800, flooding cause severe damage to one of the arch piers, resulting in several of the shops being demolished. A design by Thomas Telford for replacement with a single-span cast-iron arch bridge came to nothing, and instead the north elevation of the bridge was rebuilt to plans by John Pinch, which echoed but did not precisely match Adam's design. The difference is readily visible on the road elevations.

Although the bridge was not damaged by flooding again, the rows of shops saw significant alterations during the course of the 19th century. Cantilevered structures began to appear on the external facades of the bridge, like isolated barnacles on the southern face and eventually encrusting the whole of the north face. These provided additional space for shop tenants, and echo the cantilevered structures which still adorn Italy's Ponte Vecchio, and which were common on other inhabited bridges in the middle ages. They seem to be a popular idea in Bath: I spotted them on other buildings elsewhere in the town.

Today, the entire north façade remains cantilevered, its appearance having been largely fixed since the 1870s.

Further alterations were made to the bridge in the early 1900s, with the west end of the southern row of shops being demolished and reconstructed, so that the end pavilion now sits above the arch rather than above the abutment. I doubt that many visitors spot the loss of symmetry, although Robert Adam undoubtedly would.

More recent changes have been with a conservation and restoration ethic, removing the "blisters" from the southern façade and between 1938 to 1951 reinstating parts of the elevations (with work interrupted by war). Further changes to the shopfronts were made in 1975 with a view to more closely matching the original designs.

What I find most interesting about Pulteney Bridge is the contrast and tension between the perfection of form sought by its original designer, and the loss of integrity of the original design over time. I like the extent to which this humanises the bridge; after all, we all acquire an accretion of defects, blemishes and a loss of symmetry as we age. The restorationists may not have got their hands on the jumbled north elevation yet, but clearly would like to gradually obliterate each departure from the original concept as time and funds permit. My guide here owes more to the ideas of Stewart Brand in his excellent book "How Buildings Learn" - that a successful building is one which can adapt to use.

I think the main failing of the original design is commercial - the desire to maximise income by lining every inch of each edge of the bridge with shop units. Compare the Rialto Bridge in Venice, where there are gaps in the retail arcades at the crown, and also walkways outside the shops, so that bridge users can enjoy the river more directly. Even the heavily-built Ponte Vecchio has openings from which the river can be viewed. It would be interesting to see the different reactions which would arise if it were proposed to open up any of Pulteney Bridge's retail units as public space, so that road users can also view the river.

The bridge also makes me wonder about how a modern inhabited bridge could be designed. A criticism often expressed whenever such a thing is now proposed is of the extent to which it inevitably blocks views up and downstream. The proximity of tall buildings to the riverside edge is such that Pulteney Bridge is less guilty of this than it might otherwise be. The question also arises as to how an inhabited bridge can be designed to be adaptable, with the design life of retail units being significantly shorter than the design life of the supporting bridge. Symmetry is much-prized in design, but a bridge which starts life with asymmetrical blisters and blemishes would be an interesting concept.

Further reading: