29 November 2017

Spanish Bridges: 2. R-3 Footbridge, Madrid


I saw two more footbridges in Madrid. Both form part of a footway and cycleway connecting the districts of Vicálvaro and San Blas, running through the Parque de la Maceta, and crossing two major highways.

The first, and the larger of these two footbridges, spans the R-3 motorway. It is a suspension bridge with a main span of 110m across the dual 3-lane motorway and connecting slip roads, and two 40m side spans. The approach structures comprise a further 3 spans at each end.

The bridge is nicknamed by locals the "Chupa Chups" bridge as a result of its lollipop-shaped masts. It is a distinctive bridge, and unusual in many ways, not all of them good. It was completed in 2007 to a design by Leonardo Fernández Troyano of Carlos Fernández Casado S.L.

For any suspension bridge of this span, a key design issue is how to reduce bending moments and associated vertical displacement in the bridge deck. The worst case for design (as with an arch bridge) is with only half of the main span loaded, which results in S-shaped moments and deflection in the bridge deck.

The most common solution is to stiffen the bridge deck sufficiently to minimise deflections and distribute loads. The mirror-image approach is to have a slender bridge deck but to stiffen the suspension cable, but this is difficult to construct and consequently rare.

Another common approach is to use stay cables supported on the bridge towers to stiffen the deck close to the supports, with the most famous example being Brooklyn Bridge. This comes in two flavours: overlapping vertical hangers and diagonal stays, as in Brooklyn, or with the stayed section and vertically suspended section not overlapping. Saint Laurent Bridge in France is an example of the latter.

The R-3 footbridge takes a different approach, and one not commonly used (although the same design team has used it previously). Here, the stays are reversed, and rather than supporting the deck directly, they radiate upwards from the junction of the tower and the deck. These negative stays are connected to the main suspension cable, and they work by restricting movement of the main cable. This in turn reduces vertical movement of the bridge deck, and the associated bending moments.

The system can be highly effective, and without this system (or one of the alternatives) the R-3 footbridge could not have such a slender and economical bridge deck. You might ask why would a designer choose this system over the alternatives, and why is it so uncommon?

Fernández Troyano explains that it was done for reasons of construction economy. It allowed the bridge deck to be built out of repeated, identical, slender precast concrete panels. The more common positive-stayed alternative (Brooklyn et al) would have required custom deck panels to connect the diagonal stays onto.

I don't find this argument very persuasive, as I can't really believe that the complications in the deck would have been significant. Instead, the complication has been transferred to a series of customised cable clamps, each slightly different.

The real problem with this design solution is not the principle, but the detailing. At the masts, the negative stays pass through holes and are anchored in steel pipes projecting from the other side. It's an absolutely atrocious detail, a complete mess visually, and was almost certainly very awkward during construction, with the pipes so close together. It creates a length of hidden cable which can neither be inspected nor maintained and seems arranged deliberately to create water traps.

This is only one of the bridge's many flaws.

There are two basic options for a suspension footbridge when considering how to deal with the main cable as it passes over the supporting masts.

The first is to split the cable, anchoring it at the mast, as in the Nesciobrug. This allows a slender mast, but requires multiple suspension cable anchorages and potentially the need to adjust the cables during construction.

The second option, used on the R-3 footbridge, is a cable saddle, where the cable passes over the mast. The cable saddle must be large in radius, as otherwise bending stresses in the cable become unacceptable. For a footbridge, the size of saddle required is problematic, inevitably exceeding what is appropriate for a slender mast, and solutions include a fanned support (as on the Wingtip Bridge), or altering the mast to suit the saddle width (as on the Peramola Bridge, which is also negatively stayed).

For the R-3 footbridge, the designers opted for a possibly unique approach of "lollipop" mast heads, which feel like perhaps the worst possible option, visually.

The bridge's suspension cables are also anchored in a peculiar and somewhat thoughtless manner. The cables are connected to steel anchors at the top of inverted-V concrete piers. The anchors cannot be seen, as they are embedded into the pier heads, covered over in concrete and steel. There is, again, no facility for inspection or maintenance, and bituminous protection has oozed out of the anchorage and stained the facing concrete. It is a terrible detail.

Those inverted-V piers serve a dual purpose. They carry the cable anchorage forces into the ground, with the front leg in compression and the rear leg carrying the tension force.

They also act as a punctuation mark, separating the differing structural forms of the suspension bridge deck from its approach spans. The main bridge is supported on its edges, with the cable forces taken into the ground on both edges of the bridge. The approach spans, however, have a central spine beam, supported on single inclined columns. The inverted-Vs provide a visual break between the two different typologies, successfully, I think.

The approach spans are supported on further V-shaped piers, although here the designer presumably came up against difficulties in dealing with thermal movement of the bridge.

These piers have a cut-line near the base, indicating the presence of support bearings, presumably to allow the V-pier to move under thermal effects. It seems to me yet another poor detail, as the form of the pier is clearly unsuited to the loads and movements it experiences.

Seen from afar, the R-3 footbridge is an impressive and appropriate structure, with an impressively slender deck. It's unfortunate that the more closely you examine the details, the more you can see the very real flaws in the design,

Further information:

No comments: