Bach de Roda Bridge

Although it has often been suggested otherwise [16], Calatrava's bridge designs are not the direct progeny of those that proceed them. The masterpieces of 20th century engineering are antecedents to rather than progenitors of Calatrava's work. Technical elegance is of course central to Calatrava's work, but unlike Schlaich, Calatrava does not create his forms by beginning with a technical idea which is then formally refined. And unlike Maillart and Baumgart, he does not search for new structural paradigms. In fact, unlike most classic bridges, Calatrava's forms cannot be described in terms of structural typology. Using his training as both architect and engineer, and his skills as a sculptor, Calatrava brings a broad set of concerns to the problem of bridge design and produces works that transcend issues of engineering without disregarding them. His bridges are both mega-sculptures and public places, formally defined by a complex intertwinement of plastic expression and structural revelation.

Calatrava's Bach de Roda Bridge in Barcelona demonstrates this. In elevation, the choice of twin arches supporting the roadway by a set of suspender cables presented Calatrava with the structural achilles heel of this form: its susceptibility to buckling.

Bach de Roda Bridge

lateral bracing

In plan he was given the problem of resolving two arches that cross the train tracks below at a skewed (approximately 60 degree) angle. The buckling problem is often solved by a horizontally oriented truss between the main arches. [17] This solution applied to the skewed plan would certainly work structurally, but formally it would introduce a competing system, unrelated to the arches. Calatrava's solution to the structural problem was to place secondary arches of equal height next to the main arches on either side of the bridge.

The secondary arches lean inward from their base, beyond the bridge's edge, and are connected to the main arches by fins near their apex, thus bracing both arches against buckling. This resolution at once gives the bridge a unique three dimensional form, obviates the need to place a bracing truss between the two main arches, and allows the problem of skew to be solved by simply shifting the arches in plan.

night view
pedestrian walkway

The secondary arches also have a purely architectural purpose. Angled suspenders, lying in the plane of these arches, help support a pedestrian walkway at the roadway level. The walkway's edge is bowed outward in plan, reflecting the arch's elevation while creating a pedestrian plaza at the center of the bridge. The sloped suspender ropes at the walkway's edge and the main roadway suspenders themselves define the limits of the plaza in three dimensions. The concrete abutments of the secondary arches are flanked by concrete stairways, descending from the pedestrian platform to a park below. In this work, Calatrava is not just using his structural prowess and sculptural abilities to generate a provocative form. With the introduction of the pedestrian plazas and circulation system, he transforms, as Kenneth Frampton notes "the mere commission for a bridge into an occasion for creating a place." [18]

under pedestrian plaza

While the bridge's secondary arches solve a central structural problem and help to transform it into an occupiable civic icon, their proportions and detailing are less resolved than the bridge's overall concept. The cross-section of these arches, for example, seems at odds with loads they carry. Via their canted suspender ropes, they are loaded with only half the weight of the pedestrian plaza. The main arches, made of almost the same cross-section, support half of the much heavier roadway, its weightier traffic, and half of the pedestrian walk. The weight carried by the main arches is explicitly portrayed by the massive, exposed roadway girders that they support (via the main suspender rods). The lighter load carried by the canted arches is clearly alluded to in the knife-edge thinness of the pedestrian plaza's perimeter. While the large cross-section of the secondary arches leaves uncompleted the formal relationship between loads carried and structural bulk begun at the roadway-plaza level, their massiveness gives the upper level of the bridge a slightly heavy, static appearance.

concrete stair
upper level

The sloped arches also terminate abruptly at an arbitrary point a few feet above the plaza level, where they land on attenuated concrete abutments. These awkwardly detailed elements are at odds with the general dynamic expression of the bridge, which emphasizes movement in the leap of the main arches across the train tracks, and in the streamlined railing details shooting across the bridge's deck.

If the Bach de Roda bridge underscores Calatrava's synthetic use of structure to help solve both formal and spatial problems, his La Devesa pedestrian bridge in Ripoll, completed in 1991, epitomizes his structural inventiveness and his playful explorations of statics. This tilted-arch span connects the town's train station to a new park and housing development on the far side of the Ter river. Glimpsed from a distance, the Ripoll Bridge's structural system seems incomplete, and its stability precarious. Standing on its wood-planked walkway, the bridge's solidly-proportioned superstructure is reassuring, although the source of the bridge's stability remains mysterious.

La Devesa
section of tension arms

In fact, the structure's equilibrium is achieved by a complex amalgam of statical systems. Lying in the plane of the canted arch, steel tension arms pick up the walkway loads and transfer them to it. The steel arch in turn delivers these forces to the existing concrete retaining wall and new concrete pylon at either side of the river. Because they do not lie in a vertical plane, the tension arms must resist both horizontal and vertical force components to remain in pure tension.

statical systems
arch shape
(Keeping the tension arms loaded in pure tension ensures that the arch-shape is loaded in its plane, which in turn ensures that it works as an arch-structure instead of as a curved beam). The vertical force component in the tension arms comes from the weight of the walkway; the horizontal component is developed by a truss set in a plane just below the walkway. This truss keeps the walkway from translating sideways toward the arch. In doing so, it develops a horizontal force component, which is balanced by the corresponding component in the tension arms.

figure 93
concrete pylons

The bridge's tension arms restrain the arch from moving much out of its plane, which helps prevent it from buckling. They also help the arch resist buckling by dictating the way it deforms under load. As gravity loads tend to deflect the walkway and tension arms as shown in figure 93, the arms rotate the arch into a more upright position, which slightly stiffens it and further protects it against buckling. This principle - increased stiffness under increasing load - is a well known feature of suspension structures, whose cables derive their stability partially from the weight of the bridge they support. Making the same principle work for an arch - whose normal tendency is to become less stabile with increasing load - exemplifies Calatrava's unique structural ingenuity.

torque tube
concrete pylons

Because the weight of the walkway and tension arms are not centered under the arch, they tend to rotate below it as gravity loads pull them downward. This rotation is resisted by a large pipe which acts as a torque-tube, collecting torsion at each strut and delivering it to the concrete pylon and retaining wall at the arch's ends. With this move Calatrava exploits an often-neglected property of closed sections (pipes and tubes) loaded in torsion - that they, like columns, are very stiff and do not deform much under load. Using the pipe section as a torque tube, Calatrava builds on experiments he began with the Lucerne Post Office Depot and the Stadelhoffen Railway Station canopies. [19]

The Ripoll Bridge is by far the boldest of the three explorations, because torsional resistance is here a necessary feature of a major spanning element, and because the span itself is so long.

Stadelhoffen Station

Although Calatrava's use of the pipe is an elegant statement of its torsional capabilities, it was of course not employed as the most technically elegant solution to an unavoidable structural problem. The pipe celebrates a little used property of closed sections, without regard to the bridge's technical economy. Torque-tube action is also just one interesting feature of the bridge's complicated, nebulous structural system. The system eschews a clear explanation of structural behavior in favor of elision, surprise, and invention. What a contrast to the laconic expressions of structural behavior offered by the works of Maillart, Baumgart and Finsterwalder!

model of LaDevesa

The structural system of the La Devesa Bridge is so intricate as to be incomprehensible to laymen, leaving it to be appreciated purely in formal, architectonic terms. In this regard it is also quite a striking work. Discussing this bridge, Calatrava says that for him, "making a simple thing is often very difficult." Accepting that Ripoll is, at least formally, a fairly straightforward structure, it shows that Calatrava is indeed able to create a simple bridge that is both beautiful and provocative. The delicate proportioning of the structure illustrates Calatrava's strong sense of form. The carefully balanced relationships among its major components - the smooth steel superstructure, plain wood decking, and plank-formed reinforced concrete pylons - show his sophisticated articulation of systems through material choices and shapes.

But to characterize this bridge as simple, even expressionistically, is to ignore the complex web of relationships among components, and to neglect their simultaneous functions as compositional tools, spatial delineators, and structural subsystems. The seemingly straightforward reinforced concrete pylons, for example, serve three interrelated functions

section of concrete pylons
concrete pylons

Rising from the park, one of these canted stalagmites serves as an abutment, creating a landing point for the steel arch at the same elevation as on the far bank. The second pylon anchors a flanking concrete stair (similar to the abutment employed less decisively on the Bach de Roda bridge), used to descend from the bridge's walkway to the park below. Taken together, the pair of pylons also formally resolve the difficult problem of terminating a symmetrical structure at asymmetrical end conditions.

connection to station side

The most unfortunate feature of this bridge is the awkward concrete appendage that receives the arch and walkway on the station side of the bridge. This crude corbel was invented by Calatrava's Spanish structural consultant to accommodate a field change in the location of the twin pylons on the opposite bank. Local construction authorities insisted on moving the pylons a couple of feet back from the edge of the river, making the bridge's span a couple of feet longer. Rather than re-proportioning the arch or pylons to accommodate the new length, Calatrava allowed the problem to be solved by the hastily designed, box shaped cantilever. In accepting this ad-hoc solution, Calatrava exposes an impatience with the vagaries of construction. This is particularly unfortunate in a bridge relying so heavily on harmony among all its pieces.

This lack of follow-through can sometimes have technically deleterious consequences. In an effort to cut costs, Calatrava's pile-supported foundation design for the Bach de Roda Bridge was changed by local authorities to a spread footing scheme underneath the side-span supports. To Antonio Carreras, Calatrava's associated engineer for the bridge, this change was responsible for the minor settlement problems evident on portions of the bridge today. Though the settlement is slight, it could have been avoided altogether by a greater insistence on the use of piles.

Bach de Roda

Calatrava followed the design and construction of the Alamillo Bridge more closely. This is of course partly because stakes were higher here - the bridge was the subject of an enormous amount of technical scrutiny and features the longest clear span Calatrava has constructed to-date. The result of Calatrava's persistence is a bridge bearing a remarkable resemblance to the model he produced in 1987, featuring technical details that have silenced his critics.

Alamillo Bridge

More than most of Calatrava's bridges, the form of the Alamillo bridge is comparable to modern structures designed by conventional methods. Since the roadway is supported periodically along its length by a series of suspender cables, which are in turn tied back to a tower, the bridge can be classified as a cable-stayed structure. Louis Wintergast's bridge over the Rhine River near Speyer in Germany, for instance, epitomizes contemporary, asymmetrical cable-stayed forms. [20] [21]

Bridge over the Rhine River

But unlike Wintergast's bridge, how the Alamillo Bridge works is not easily perceived. Lacking the backstays that stabilize the tower of Wintergast's bridge, the Alamillo bridge suggests incipient movement more than static support. Like Calatrava's bridge in Ripoll, Alamillo seems to be held in static equilibrium by sleight of hand. On one hand, without back-stays, why isn't Alamillo's tower bent toward the water by the pull of its roadway stays? On the other hand, considering the somewhat massive profile of its tower, why doesn't the whole structure tip over backward? Answers are provided by the tower's mass itself, which plays the role of the backstays used on most cable-stayed bridges.

diagram of tower

The tower's own weight pulls it downward, and by virtue of its backward cant, counteracts the tendency of the roadway-stays to bend the tower toward the water. The downward and backward pull of the tower is delicately controlled by its tonnage, which is calculated so that it will not pull on the cables with enough force to tip the bridge over backward.

Using the weight of the tower to counter the pull of the cable-stays is an interesting inversion of a principle of Gothic construction, where the weight of vertically projecting pinnacles is used to counter the outward thrust of flying buttresses. Another interesting feature of this seemingly meta-stable form is that the horizontal force that the stay cables put into the roadway deck is equilibrated by the analogous force they put into the tower. These two forces equilibrate each other where they meet at the base of the tower. The net result is that (unlike most asymmetric cable-stayed bridges) the foundation of the Alamillo bridge must resist only vertical loads, and not also horizontal ones. Although Calatrava's structural system draws on both cable-stayed and gothic structural paradigms, it is not simply an amalgam of the two and the resulting bridge does not reflect any clear structural type.

The structural principle of the Alamillo's tower also shows how Calatrava uses structural experiments with small sculptures as inspirations for his bridges. Calatrava often explores "toys and games that can give plastic expression to the principles of statics," and goes on to say "exercises such as these are the generators of many of my ideas about formal language which later, as full sized structures, inhabit a real landscape." [22] His 1979 toros sculpture shows how this process helped generate the form of the Alamillo Bridge. [23]

Tower Toros Sculpture

The sleek appearance of the Alamillo Bridge contrasts sharply with Wintergast's structure, which emphasizes clearly articulated components and the structural actions of thrust, counter-thrust, compression and support. Calatrava's structural components are melded together by smooth, doubly curved forms to emphasize the bridge's flowing silhouette rather than clearly delineating their structural purpose. The intersection of the bridge deck, tower and abutment, for example, is made with a smoothly sculpted, integrally cast moment connection, without changing materials or using discrete pieces. The result is a continuous transition zone, which emphasizes the continuity and plasticity of the bridge without visually explaining the transfer of forces among the components it connects.

intersection of bridge deck

Bent Steel Chair

Designs by Norman Bel Geddes

Overall, the appearance of Calatrava's bridge is that of a precisely machined form, more reminiscent of Jean Prouve's bent-steel furniture designs than most civic engineering structures. [24] The bridge's plastic, finely sculpted image (which is echoed in the guardrail and traffic bollards shown in the presentation drawings and model of the Bach de Roda Bridge) also bears a close resemblance to the manufactured metal fins of Detroit's automobiles or the streamlined designs of Norman Bel Geddes. [25] But unlike these machined precedents, that evoke the image of motion while concealing the guts of the machines they house, the Alamillo Bridge's form not only houses the structure, but is the structure itself. In this way, Calatrava avoids the treatment of structure as a hidden armature evident in much postmodern architecture, while he simultaneously provides an alternate to the so called "high-tech" expression of Michael Hopkins, Norman Foster or Renzo Piano.

The execution of the Alamillo Bridge demonstrates Calatrava's approach to construction. Borrowing the technique pioneered by Baumgart and Finsterwalder, Calatrava proposed building the Alamillo Bridge using a slip-formed, successive cantilever method. This would have allowed the bridge to be erected without scaffolding, by building it simultaneously upward and outward from the base of the pylon. Each new segment of the tower would balance a new segment of roadway; stay cables would connect the two sections and balance the loads between them. This scheme was abandoned by the bridge's contractors, who were relatively unfamiliar with slip-formed techniques and had their own ideas about how to build the bridge. [26] Without objection by Calatrava, the contractors used a huge crane to lift sections of the tower into place, and they built the roadway on top of false-work. By allowing this change, Calatrava shows himself more interested in getting his work built than how it is built.

false work for Bach de Roda

Blimp Hangar at Orly

This pragmatic attitude was also applied to the construction of the Bach de Roda Bridge. The bridge's contractor erected both the main and secondary arches using abutment-to-abutment false-work. The arches were installed in segments, using a crane to place them onto the scaffolding. While this brute-force method of construction worked, it is not technologically refined. The structures of Eugene Freyssinet, a French contemporary of Maillart's, demonstrate developments in construction methods made by structural engineers more concerned with assembly than Calatrava. Over the course of his career, Freyssinet developed many elegant construction schemes that are still admired today for their simplicity and economy. To solve the problem of constructing an arch form under similar circumstances as at the Bach de Roda Bridge (over land), Freyssinet developed a scaffold-less method to build his parabolic blimp hangars at Orly. [27] Construction of the hangers proceeded by construction machines that crawled up the unfinished edges of the vaults, extending the structure as they went. This system allowed the unfinished vaults to be braced with single struts, rather than with continuous false-work.

Although Calatrava's interest in construction is demonstrated by the erection methods he proposes for many of his projects, he is not obsessed with elegant assembly. Unlike Freyssinet, he is not trying to innovate in the field of construction technology. And unlike Finsterwalder, he does not employ construction techniques as a primary inspiration for formal expression. Instead of trying to maximize technical elegance in assembly, Calatrava is content to use any reasonable construction techniques to produce his work; instead of refining a form derived from a particular construction process, Calatrava employs various erection methods (that are often not reflected in the final form of this work) to help realize a particular artistic vision. In this way, Calatrava uses construction technology as a means rather than an end, reflecting his stronger interests in utility, program and expression.

East London River Crossing
site plan

In a proposed bridge over the Thames in London, Calatrava shows one way he would pursue his bridge-design ideals at an even grander scale than at Alamillo. His 380 meter long East London River Crossing would open a new major thoroughfare between Thamesmead and Docklands, while standing as a monumental gateway to the entire city.


As in almost all of Calatrava's bridges, a sense of movement is immediately conveyed by the bridge's elevation. But unlike Ripoll and Alamillo, the motion suggested here is derived from a few strong, sweeping gestures, instead of the brinksmanlike testing of statical limits. The high elevation of the roadway and low profile of the arch are responses to meeting the clearance requirements of boats below and planes above. The side-span pylons are easily identifiable as a variation of the V-shaped twin pylon form first employed at Ripoll. But here the pylon system works a bit differently. On this bridge, each set of pylons acts with the roadway girder as a huge cantilevered truss, supporting the traffic above them and the gravity reactions of the center-span arch.

free body diagram

Each truss is anchored in the river at the base of its twin, canted pylons, and on each riverbank under the large, vertical pylons. The weight of the center-span arch is delivered to the trusses as concentrated loads at their tips. Working like a diving board with a swimmer at its end, each truss delivers this load to the base of the twin pylon, while the vertical pylon beyond pulls downward, acting as a tension strut to resist the tendency of the truss to tip over. The ends of the center-span arch are tied together by the roadway girder below it, which prevents the arch's horizontal thrust from being transferred to the pylons. Although the center strut connecting the roadway and the arch's apex is first noticed as a formal marker for the center of the span, and for the rhythm it establishes with the vertical side-span pylons, it also serves a central structural role. Connected rigidly to the roadway girder, it acts as a beam to resist horizontal movement of the arch and thereby stiffens it against buckling. This element highlights Calatrava's continued search for novel solutions to the problem of arch buckling that he began with the Bach de Roda and Ripoll bridges.

East London River Crossing

Despite its intriguing structural system, the East London River Crossing's form does not focus attention on how it works. As with Alamillo, one reads the bridge's plastic shapes as parts of a continuous, single object. Visually, the bridge appears as a sculptural statement rather than a structural diagram. Calatrava's emphasis on the bridge's overall form rather than its structural details stems from his gestural design methods. Upon obtaining a commission (or upon entering a competition), Calatrava first visits the site. After this he produces a series of sketches, often rendered in watercolor, representing trial bridge designs. The margins of these sketches are sometimes filled with structural calculations, reflecting Calatrava's dual search for a particular visual gesture and structural viability. [28] Settling on a form, Calatrava passes copies of the final sketches to his staff architects and engineers. The architects develop the profile into presentation drawings while the engineers analyze the design's structural performance. In using this method for the London Crossing design, Calatrava established a strong visual reading of the bridge before the structural details were fleshed-out. The inevitable result is a bridge emphasizing its form over its structural detail. Calatrava's design method works opposite from that used to produce most major bridges, in which the choice of structural system is decided first, and is immediately followed the development of the structure's technical details. This method produces a bridge whose silhouette emerges primarily from the details of its structural design.

Calatrava's presentation models also underscore his emphasis on overall form over structural elaboration. The models are crafted with exceptional precision, yet their monochrome palette and mirrored ground planes make them seem unreal or otherworldly. These miniature forms do not reveal structural particulars or materials as much as they evoke an abstracted, aesthetic vision.


Calatrava's highly conceptual models leave much open to imagination, and heighten one's anticipation about the tectonic details of his full-size structures. Viewing a model before visiting the completed bridge, one wonders about the actual structure's materials and construction as much as one anticipates the experience of observing and moving through it at full-scale.

model of Bach de Roda

Although not discernable from his models, Calatrava develops a clear idea of the materials for his commissions long before construction begins. Calatrava feels that one cannot design a structure without at least postulating what it is made of and why. [29]

Kuwait Pavilion

Kuwait Pavilion

Even for his most schematic competition entries, he decides what his design will be made of. In his small commissions, material choices often make a strong contribution to the work. Their particular contribution varies widely among projects, depending on the specific materials employed. Calatrava's precast concrete piers supporting the roof of his Kuwait Pavilion at Seville's Expo 92, [30] for example, contribute to the expression of the pavilion as much by their impeccably smooth finish as by their specific form. These virtually identical piers echo the smooth, repetitive forms of the wood finger trusses they support. By contrast, the rough texture and idiosyncratic individuality of the plank-formed, cast in place concrete of the Bach de Roda Bridge serves as a foil for the smooth, machined steel superstructure above it.

East London River Crossing

Calatrava's design for the East London River Crossing shows that specific material choices are less important to him in his larger commissions. This is particularly true when a change in materials will improve the constructability of the work. Calatrava envisioned the pylons of the East London Crossing in steel, but he would allow them to be built of reinforced concrete if a contractor claimed this would significantly reduce its cost. Similarly, the Alamillo bridge's tower was designed in cast-in-place concrete, but at the contractor's behest, Calatrava allowed it to be built of a concrete-encased steel shell instead.

Alamillo Bridge
steel shell

Except for some minor changes in geometry at the base of the tower, this change did not affect the general form of the bridge. However, its close-up, tactile relationship to pedestrians passing over the bridge was rather drastically altered by this move. The slightly rough, honeycombed surface characteristic of slip-formed concrete has been replaced by smooth, slightly wavy sheets of steel. The boundary between sheets is clearly marked by groove welds that have not been ground flush. Although the result is not objectionable, it is very unlike the original design.

Calatrava's attitude toward the materials used in his large commissions contrasts strongly with some of his contemporaries, most notably Peter Rice. In describing Renzo Piano's De Menil Museum in Houston, Rice, who designed the building's structure and natural light control systems, reports

We wanted to use ferro cement as the principle element in the design of the light louvers which were a part of the roof. This decision in turn gradually determined and dictated the design of the building... After we made the decision to use ferro-cement, we went to Houston to find out how we would make it...the typical reaction from manufacturers was "Well, we can do it out of pre-cast and it will only be 1/4 of an inch thicker. Isn't that going to be sufficient?" We couldn't find anyone who understood the tactile quality we were seeking from these pieces." [31] [32]

The De Menil Collection

In the end, Rice and Piano found the material they wanted at a small manufacturing plant in England. In this design Rice and Piano used a particular material technology to at least frame, if not determine, the form of the work. By contrast, Calatrava's Alamillo and East London Crossing bridges show that he is willing to sacrifice tactile detail to ensure the realization of his primary gestures - at least in his larger commissions.

the scintillating essay continues...

or the essay from the beginning

Footnotes and Figures

[16]. For Example:
Dennis Sharp, "Felix Candella and Santiago Calatrava," World Architecture. Issue 13, 1991, p 35.
"Expressive Engineering", Architectural Review. September 1987, p 51.

[17]Arched Roadway Bridge Showing lateral bracing system between arches.

[18].Werner Blaser, ed, Santiago Calatrava: Engineering Architecture. Page 17.

[19]Stadelhoffen Station. Steel and glass canopy over train platform.

[20]Bridge Over the Rhine River near Speyer, Germany. Louis Wintergast, 1968.

[21].Fritz Leonhardt, Bridges - Aesthetics and Design. Deutsche Verlags-Anstalt, Stuttgart, 1984. pp 257 - 278.

[22].Bridging the Gap, Page 123.

[23]Tower Toros Sculpture.

[24]Bent Steel Chair Design. Jean Prouve.

[25]Streamlined Designs. Norman Bel Geddes.

[26].Interview by the author with Dr. Calatrava. Zurich, June, 1992.

[27]Blimp Hangar at Orly, Eugene Freyssinet. Construction photo.

[28]Bridge, Santiago Calatrava. Early Sketch. (Image currently unavailable).

[29].Interview by the author with Dr. Calatrava. Zurich, June 1992.

[30]Kuwait Pavilion. Precast concrete piers and wood finger trusses.

[31].Bridging the Gap. Pages 92 - 94.

[32]The De Menil Collection, Renzo Piano and Peter Rice. Ferro cement light louvers.