20 October 2017

German Bridges: 8. Railway bridge over the Spree at Friedrichstrasse Station, Berlin


I'm often puzzled by how people view groups of bridges, even people who nominally specialise in their study or documentation. This is the next bridge in my eastward journey along Berlin's River Spree, and it's a significant structure with a complex history. But you won't find it in a book like Berliner Brücken, nor is it listed on Wikipedia amongst fellow Spreebrücken. What crime has it committed, to be so left out? Very simply: it's a railway bridge, such a lowly beast that for some people, it seems it may as well not exist.

It's quite a significant railway bridge as well, carrying tracks westwards across the river from Friedrichstrasse railway station. What can be seen today is an amalgamation of three phases of construction. The southernmost pair of arches survives from 1882, and is in wrought iron. The northern arches date from 1923, in riveted steel. The central section of the bridge was replaced in 1997 with arches in welded steel.

The 1882 section features two-pin arched girders, constructed by lacing together the girder flanges with diagonal web members. Vertical spandrel struts are splayed out at their base to enhance stability. It's a straightforward and effective design, and in my view the most visually attractive face of the bridge.

The 1923 part is less visually appealing. The arch ribs are riveted box girders, stiffened externally. The cantilevered deck has a less "clean" appearance than on the south side, with visually obtrusive cantilevers breaking up its linearity.

A pedestrian bridge has been inserted into the north part of the bridge. This allows pedestrians to cross the river, or via a series of staircases to access the railway platforms above.

It seems like a neat idea, but in practice it's like sticking an underpass through the belly of a mechanical whale: gloomy, unattractive, and slightly squalid. Fascinating for an engineer, perhaps, but unsuitable for many potential users.



Further information:

18 October 2017

German Bridges: 7. Marshall Bridge, Germany


Here's another bridge which clearly displays the scars of its past.

The Marschallbrücke was built in 1882 as a three-span wrought iron arched bridge. The design is credited by Structurae to architect Albert Eduard Paul Gottheiner and by Berliner Brücken to engineer Friedrich Krause.

The bridge was damaged during the Second World War, and partially reinstated using steel girders afterwards. Due to its proximity to the Berlin Wall, it was not fully repaired until after the Wall came down. In 1999, the bridge was completely renewed to a design by architect Benedikt Tonon and engineer Gerhard Pichler. The reconstruction cost 12 million Deutsche Marks.

Tonon arranged for one of the original arch spans to be removed, and used as part of the Anhalter Steg, at the German Technical Museum elsewhere in Berlin.

The northernmost span was rebuilt, preserving its arches as supports to two footways on either side, thus retaining the appearance in elevation. Between these arches, the roadway is separately supported on concrete encased girders.

The southern spans were replaced with a new 39m long single-span structure, intended to improve river navigation. Again, encased girders carry the roadway, and a portal-frame structure carries the footways, reusing the original arch springings. There's clearly no chance of anyone mistaking new for old here, and that principle is central to the whole design.

The footway structures are separated from the highway by a simple metal kerb. This has presumably been judged sufficient to prevent accidental vehicle loading onto the footways.

The existing balustrades are ornate and well preserved.

The balustrades to the new span are very different, with an angular infill pattern which echoes the angularity of the structure below. There is a pattern to the lattice, with stiffer members in the form of repeating "X"s superimposed on the main lattice.

New lighting columns have been installed on the central pillars - I found these to be awkward, although perhaps awkwardness is appropriate at the juncture between old and new.

I admire the technical solution adopted, but I couldn't really bring myself to like the bridge. There is a sagging yellow line on the outer faces, perhaps to emphasise navigation requirements, and it feels like a real distraction to what would otherwise be a rigorous composition.

Further information:

17 October 2017

German Bridges: 6. Marie-Elisabeth Lüders Footbridge, Berlin


Continuing our journey eastwards along the River Spree in Berlin, the next bridge we come to is the Marie-Elisabeth-Lüders-Steg.

This twin-level bridge was designed by architect Stephan Braunfels as an adjunct to the adjacent parliamentary buildings completed in 2003. Structurae states that the wide lower deck is a 62m long prestressed concrete box girder structure; the narrower upper deck is evidently a steel truss. The actual main span is only 36m.

The bridge doesn't appear to have an official name, borrowing its normal title from the adjacent Marie-Elisabeth-Lüders Haus, home to German parliamentary offices, meeting rooms and library. (Wikipedia suggests it may be officially titled the Jakob-Mierscheid-Steg).

The lower walkway is open to the public, but the upper floor is private.

According to Berliner Brücken, the columns which appear to hold up the upper truss are not actually connected to the lower bridge in any structurally meaningful manner. That sounds unlikely, and my photographs don't bear it out: the columns are concrete, of course they sit on the lower span.

However, the whole bridge strikes me as a poor example of design. The truss members in the upper walkway are angled as if the bridge was designed to be entirely supported at its ends: they are not the correct way round for a truss span supported also in the middle.

I wonder whether it was originally intended as a single span, and then the additional columns inserted only when it became apparent the buildings either side couldn't hold it up.

I find little to like about the bridge, other than perhaps the paving and the uplighters.

I think it's a little depressing that such a muddled, dispiriting design could be sufficiently well-regarded to be built at such a nationally significant civic site.

Further information:

16 October 2017

German Bridges: 5. Crown Prince Bridge, Berlin

Ok, with the BAMPOTs out of the way, it's back to Berlin, and continuing steadily eastwards along the River Spree (the next bridge to the west is the Gustav Heinemann Bridge).


Santiago Calatrava has designed two bridges in Berlin. The Kronprinzenbrücke was completed in 1996, and is the result of a 1991 design competition. I may cover his other bridge later on in this series.

The previous bridge at this site had been demolished to reduce the number of East German refugees fleeing into West Germany. The Crown Prince Bridge was funded following German reunification and was presumably quite a significant symbol of the need to rebuild cultural and physical connections.

It's not a huge bridge - it's only 74m long, with a main span of just 44m. It carries a highway and walkways across the River Spree.

Calatrava's design owes something to his earlier (unrealised) Wettstein Bridge. It gives the appearance of being a skeletal steel arch bridge, while in reality being something different. The main bridge span is supported via lateral steel beams onto two "arches" running beneath the deck, and inclined outwards.

There's no a priori reason not to use vertical "arches"; the tilt is just typical Calatrava playfulness, part of an effort to generate a visually more dynamic geometry for the steel skeleton.

I put "arches" in quotes because on the face of it this actually appears to be a cantilever bridge, with Vierendeel trusses spanning outwards from the support piers, and given the illusion of arches by the adoption of shallow arch curvature. If these were true arches, the shallow curvature would lead to very high longitudinal thrust loads, and the support piers are not arranged in such a way as to resist those loads.

Instead, the piers are arranged to resist lateral thrusts, in line with the river. This is purely a function of the arch tilt, which is severe enough to put the piers into considerable lateral tension. It appears to be resisted by the very visible "knee" elements, but these just carry the loads into the interior of the lower concrete part of the piers. It's not visible to the observer, but inside the concrete there are large steel portal frames, which act as ties to restrain the lateral forces.

It's a "plinth" bridge of sorts, a visually attractive and interesting superstructure which is perched upon rather than integrated with its supports. The obvious way to resist the lateral forces at the piers is through a horizontal tie at springing level, but this spoils the purity of the superstructure's conception. The result, as with many Calatrava bridges, is that the substructure is forced to work unusually hard to allow the upper parts to remain unsullied.

As with Calatrava's better designs, much of the detailing of the bridge has been very well done. However, it's hard to tell how much of that is down to Calatrava and how much to other engineers charged with realising his design. Several details in the completed bridge differ from those in original design drawings (included in Frampton's book, linked below).

Look closely at how the arch spandrel elements are connected both to the arch and to a tubular deck girder, or at the shaping of the thrust supports on the bridge piers. The tilted parapets are attractively assembled, and the parapet ends are quite gorgeous, finely sculpted blocks which put so many other bridge designs to shame.

Shaped arch elements on the face of the abutments give an indication of load paths - abutments are too often blank, blocky and unattractive, but not here. This is particularly significant for a bridge which is experienced at close hand from the river side paths.

Not everything is great: the underside of the bridge deck is given texture and form by the exposure of a large number of ribs and service pipes, but it feels over the top to me. The upper chords of the Vierendeel trusses are also absurdly large compared to the lower chords - this appears to be solely so that pipes can be hidden inside.

Some considerable effort has gone into detailing the highway face of the bridge, with bespoke kerb lighting units and lighting columns. However, the curse of poor maintenance has left these looking forlorn, and in some cases damaged and corroded.

Nonetheless, I think this is on balance an interesting and attractive bridge, lacking in the overpowering and inhuman scale that ruins many of Calatrava's later projects.

Further information:

14 October 2017

Bridge Awards for Mediocrity and Plain Old Terribleness - The Winner!

I announced the Bridge Awards for Mediocrity and Plain Old Terribleness (the BAMPOTs, for short) back on 3rd September. Twenty-two bridges were nominated by the blog's readers, and passed to a panel of bridge experts to produce a shortlist of the worst of the worst, which was announced on 1st October. The shortlist was opened up to a final public vote.

This convoluted process is of course just a way of getting myself off the hook ... I didn't nominate, shortlist or vote for any of these bridges, so don't blame me!

Anyway, the votes are in, the counting is done, and we have a winner!

Seventy-three people voted. In reverse order, the votes are as follows:
  • Arch footbridge near Exeter, UK: 7 votes (9.6%)
  • Cumberland River Bridge, Nashville, USA: 8 votes (11.0%)
  • Lucky Knot Bridge, China: 13 votes (17.8%)
  • Seabraes Footbridge, Dundee, UK: 15 votes (20.5%)
  • Millennium Bridge, Ourense, Spain: 30 votes (41.1%)
So what can be said about a bridge which won over two-fifths of our discerning audience (and, incidentally, which was unanimously the worst rated bridge by the judging panel)? (All photos courtesy of Victor Hermida Prada).


The nominator was anonymous, and said: "My vote goes to: Millenium bridge... in Ourense, Spain. I don't think it needs much explanation, just type 'puente del milenio ourense' in Google and the images will make clear why!"

A comment on Twitter is also illuminating regarding the context for the bridge:

The judges said: "I will never understand the 'creative mind' that came up with this. It has more than a smack of 'emperor's new clothes' about it" ... "Exquisite in the dictionary sense meaning 'piercing, excruciating, agonising, harrowing, tortuous, tormenting'" ... "It has everything: Absolute lack of respect for the valuable collection of historic bridges in the city, a deck depth that makes any other structural help unnecessary, lots of unnecessary structural help, inclined pylons (verticality is overrated) that lead to clumsy over-designed piers that try to relate to the inclination."


I think it is undeniably an awful bridge, and well deserving of recognition as this year's top BAMPOT. I'd echo all the comments from the third judge above - its largest offence is its ungainly combination of multiple geometries, none of them designed well in their own right, and worse when plastered on top of each other. The walkway gimmick gives the initial impression of being a suspension bridge, but instead it's the killer garnish that overpowers everything else.

It's a hard bridge to look away from, but only in the same sense as a car crash. I'd certainly love to visit it, and indeed the Ourense tourist office promotes it as a significant attraction, along with their other bridges.


I'd like to thank everyone who nominated or voted for the BAMPOTs, and particularly the judging panel who showed sound judgement and also impressive endurance.

What do others think? Is the Ourense bridge a worthy winner? Do you think one of the other bridges should have won, and if so, why? Please use the blog page comments feature to provide feedback!

04 October 2017

German Bridges: 4. Humboldt Port Railroad Bridge, Berlin


Taking a little detour, we continue east along the River Spree from the Gustav Heinemann Bridge, and then turn north into the Berlin-Spandau Ship Canal. The channel opens out here into the wide Humboldt Harbour, and is spanned by several railway tracks just to the east of Berlin Central Station.

The Humboldt Port Railroad Bridge, or Eisenbahnüberführung Humboldthafen, as Structurae has it, was built in 1999. It carries a significant stretch of railway - from here you can apparently travel to Paris in the west, or to Moscow in the east.

The design contract was awarded to Architekten von Gerkan, Marg und Partner, working with engineers Schlaich Bergermann. The contractor's engineer was Leonhardt Andrä und Partner, while Krätzig und Partner acted as expert review engineer.

The 240m long bridge supports six tracks, and varies in width from 40m to 66m to accommodate the track layout on the station approach. This section is part of a longer series of railway viaducts around 1km long in total.

The bridge comprises prestressed concrete "T-beam" decks, 1.7m deep and spanning up to 25m onto steel column supports. Over the harbour there is a single 60m span, comprising steel arches. There are three viaducts running alongside each other, each curved in plan.

The arch ribs are steel tubes 660mm in diameter, made from 100mm thick plate. The detailing where the tubes support the deck is odd: there seems to be a thin concrete platform immediately below the main deck girders, with this platform element compressed by the arch tubes so that it forms part of the arch structure.

The bridge has pinned articulation at the supporting arch nodes, and also some of the tubular column supports.

The connections between steel tubes are made in cast steel, something rarely seen in bridge construction, and claimed here to be a first on a rail bridge (its use on a footbridge can be dated back at least a decade before).

Cast steel is usually seen as expensive, although is increasingly common in large buildings where the structural members are visible. It has considerable advantages over welded steel with regard to fracture resistance and fatigue, especially as sharp corners can be easily eliminated.

For the Humboldthafen Bridge, finite element analysis was validated with physical testing at the University of Karlsruhe, using tubes up to 508mm diameter and with walls 50mm thick.

I don't think it's an especially beautiful bridge. There are individual elements with nice detailing, but the junction of the arch and deck is very awkward, the two structural forms are superimposed rather than integrated. The hinges at the column bases and arch springings are not well expressed, and there are inconsistencies in the different types of cross-bracing applied. I'm not sure the cast nodes add a great deal visually, either.

Where it is successful is in how it unifies the arch and beam/column parts of the viaduct, a combination of systems which could otherwise have been very disjointed visually.

Further information:

02 October 2017

German Bridges: 3. Gustav Heinemann Bridge, Berlin


For now, I'll keep heading east along Berlin's River Spree. After the Moltke Bridge, the next structure is the Gustav Heinemann Bridge (Gustav-Heinemann-Brücke), which was built in 2005.

The design was selected as the winner of a competition. The bridge was designed by architect Max Dudler and engineers Grassl and KLW Berlin. Lighting was designed by Designplan Leuchten. The bridge connects the forecourt of Berlin's central railway station to the various central government precincts around the Spreebogenpark. It is named after former President of West Germany, Gustav Heinemann.

The design seems to be a celebration of austerity. The two edge girders are Vierendeel trusses of constant depth and bay spacing, and the use of constant H-sections for all the members adds to the visual simplicity. The bridge is 5m wide, with wooden decking supported on steel cross-members.

The truss girders are 2.25m tall, but the centre-to-centre dimension is reported as 1.82m. The spans are respectively 9m, 66m and 13m, giving an effective span-to-depth ratio of 36:1, which is high even for a pedestrian bridge.

The two relatively short end spans help to "clamp" the centre span. The bridge deck is therefore tied down at its ends using hinge links, and sits on the intermediate support piers via elastomeric bearing pads. Tuned mass dampers are hidden below the deck to reduce vibration, and the bridge was vibration-tested by groups of up to 25 people during construction.

The deck level is raised relative to the bottom of the girders, presumably because otherwise the upper chord of the girders would be visually intrusive. The walkway is separated from the girders by a balustrade in the form of a metal grille.

The H-section configuration of the girders means that the outer and inner surfaces are a series of recessed boxes defined by the overlap of the chord and web member flanges. On the inner face, the lighting units are tucked away into the recess at the top of the girder webs.

The detailing is exceptionally simple, but I found it considerably more satisfying than the more normal hollow-section Vierendeel truss footbridge, which presents a featureless face to the world. The arrangement of the H-sections is logical in resisting in-plane bending moments.

For me, the colour is a slightly odd choice, being a grey-green with a slight hint of the military to it (see, for example, "Panzer Olive Green"). It's preferable to the conventional pale grey of so many bridges, but only just.

I can't quite work out what's going on with the timber flooring. There are five rows of bolt-heads clamping these to steel stringers running below. there are also two steel durbar plates, one on each edge. Can anyone tell me what these are for?

On the whole, I find this a bridge to admire rather than necessarily to like. Rigour and logic always have a certain appeal, but it's a blank, steely-eyed, inexpressive sort. I don't think a spot of real colour would have gone amiss, and the bridge geometry would have lent itself to something subtly Mondrianesque, perhaps.


Further information: