Engineers of Dreams: Great Bridge Builders and the Spanning of America

by Henry Petroski

  But perhaps the most significant factor that widens simultaneously the communication and generation gaps between engineers of different times is the ever-present development of engineering science and the tools of analysis. No twentieth-century suspension-bridge engineer seems ever to have lost a reverence for the works of John Roebling, as epitomized in his Brooklyn Bridge. However, with the growing development of analytical tools, as embodied in the deflection theory that Leon Moisseiff seemed so effectively to have developed and applied, the methods of Roebling, which relied upon physical more than mathematical argument, appeared to have been superseded. Unfortunately, with the relegation to dusty archives of Roebling’s verbal reasoning about his concerns over stiffness and wind, the natural forces and response of bridges to them that so concerned him ceased to be a primary concern to more mathematically minded engineers, who remembered the old master’s bridges primarily as aesthetic models. The limitations of this shortsighted view of engineering history became immediately apparent in the wake of the collapse of the Tacoma Narrows Bridge, and the subsequent revitalization of the suspension-bridge form took place only in light of newly embraced aerodynamic theories and wind-tunnel testing. This new perspective led to such innovations as the winglike decks and inclined suspender cables of the Severn and Humber spans in England, the latter the longest in the world until the completion of a bridge in Denmark and the Akashi-Kaikyo Bridge across Japan’s Akashi Straits. The Severn span, however, has not been without its own problems, and it has had to be “strengthened” to carry the heavier lorries that have been allowed to use Britain’s motorways since the bridge’s original design and construction. No matter how strong the bridges are, users of the United Kingdom’s greatest spans across its wide estuaries have been buffeted by winds so forceful at times that the largest lorries have been instructed to cross in pairs, thus reducing somewhat their chances of being blown over.

  The pattern of bridge development established by Sibly and Walker suggests that in the late twentieth century there should be not only another radically new bridge type evolving toward more and more daring lengths and slenderness but also that a major failure can be expected in that type sometime around the turn of the millennium. The genre that seems so eerily poised to continue the thirty-year cycle of major bridge failures has developed from an old type that was rediscovered in Europe in response to the exigencies of rebuilding an infrastructure destroyed by the war. Though the superstructure of many bridges in Germany had been damaged, their foundations and piers were often reusable. The challenge to engineers was to design for these prewar foundations lighter bridge decks that could then carry the heavier postwar traffic. Cable-stayed bridges had been conceived centuries earlier, but they had never before been built on the scale or in the numbers that they began to be in Germany in the 1950s. For some time after that period, such bridges were thought to be the most economical and suitable choice for spans no longer than about twelve hundred feet, or somewhat shorter than the main span of the Brooklyn Bridge. By the 1980s, however, cable-stayed designs were being proposed with span lengths that had previously been thought to be in the exclusive realm of the now more conventional suspension bridge.

  The Sunshine Skyway Bridge across Tampa Bay, among the longest and most photographed of the cable-stayed bridges in the United States, has a main span of twelve hundred feet. Completed in 1987, the Florida crossing remained among the top dozen or so longest cable-stayed bridges in the world into the early 1990s, when the genre really took off to new lengths. Many long spans were completed in Japan and Canada in the late 1980s, and in 1991 the longest cable-stayed bridge finished in Europe was the Queen Elizabeth II Bridge across the Thames at Dartford, with a main span of almost fifteen hundred feet. The French, however, had already at that time under construction over the mouth of the Seine a cable-stayed bridge with a main span in excess of twenty-eight hundred feet—the magnificent Pont de Normandie—and Danish engineers let it be known that they were considering a cable-stayed span approaching four thousand feet in length to complete what is known as the Great Belt link between Denmark’s two largest islands. Though this design was eventually rejected in favor of the suspension bridge, itself remarkable with its main span of more than a mile in length, the daringness of the Danish cable-stayed proposal became the topic of some discussion among engineers.

  Cable-stayed bridge proposed over the Mississippi at Cape Girardeau, Missouri (photo credit 7.2)

  The Sunshine Skyway Bridge across Tampa Bay, shown on the cover of a brochure (photo credit 7.3)

  British engineers questioned whether it was wise for the French even to attempt a cable-stayed bridge with a main span almost twice the existing record. Questions of how the incomplete structure at the mouth of the Seine would behave in the wind were central to such endeavors, and there were warnings that scaling up in so large a leap from existing bridges was a prescription for disaster. Engineers proposing the doubling or even tripling of existing spans were confident, however, claiming that the larger bridges were “perfectly” possible because of modern computer modeling and construction techniques. Special devices were fitted to the incomplete spans to stabilize them in the wind during construction, and when the deck was finally completed, in the summer of 1994, many an engineer breathed a sigh of relief. After six years of design and construction, the span was opened to traffic early in 1995.

  Regardless of how sophisticated the computer models or construction techniques, whether the cable-stayed Pont de Normandie, with its record main span, would successfully cross the mouth of the Seine depended at least in part on luck. Excessively high or unusual winds that were not factored into the computer model could hold as much of a surprise for the engineer of the Pont de Normandie as the excessive weight of steel that was inadvertently omitted from the calculations for the Quebec Bridge held for its engineer. Any model, whether a simple equation on the back of an envelope or an elaborate numerical one in the gigantic memory of a supercomputer, is only as good as its fundamental assumptions. The Tacoma Narrows Bridge fell because the most sophisticated deflection theory used to design it did not take into account the dynamic effects of the wind. In sum, the undoing of a project will derive not so much from its size or scale as such, as from an imperfect understanding. What is an insignificant detail in a cable-stayed bridge of relatively modest size can grow to surprising importance as the size of a bridge grows. The sage advice to increase the size of bridges slowly reflects an awareness of this scale effect among more experienced engineers, but younger engineers, full of confidence in their computers, often think such caution to be a mark of excessive conservatism.

  Making great leaps in size does not, of course, doom a bridge to failure, and daring young engineers can use the historical examples of the Forth and the George Washington bridges to defend their ambitious designs. The cable-stayed bridge that might fulfill the inexorable prophecy of Sibly and Walker’s pattern of failures will not necessarily be the longest of the genre. The Tacoma Narrows Bridge had, after all, only the third longest main suspended span in 1940. But though mere size may not bring bridges down, it may often be the focus of attention with regard to warnings about their behavior. Being only the third-greatest span, with a modest traffic load and in a relatively remote location, the Tacoma Narrows was not a structure that called attention to itself—even though it was the most slender of bridges—until it began to oscillate in the wind and collapsed. It may be similar with cable-stayed bridges. The Pont de Normandie and others in the vanguard of design will be more carefully planned and watched than those that will be almost but not quite as large—those that will be, ironically, of little more than local significance or remarkableness during their design and construction. It may be from the great but not necessarily the greatest that we can expect the most unpleasant surprises.

  If a major collapse of a cable-stayed bridge is to be prevented, there must be as much attention paid to the maintenance of the engineering-design infrastructure as to that of
the physical infrastructure. This means that engineers should be as sensitive to the historical cycles of success and failure that have plagued the design enterprise as they are to the cycles of freezing and thawing that can plague their physical roadways and bridges. Care of the design infrastructure requires the maintenance of lines of communication between engineering generations, so that new tools and models are not used in ignorance of past experience. The surest way to break the vicious cycle of bridge failures identified by Sibly and Walker must certainly begin with the recognition that neglected patterns from the past become unconscious patterns for the future. Only by bringing these patterns into sharp focus and by seeing the modern engineer as reinventing, albeit with faster and more powerful tools, the bridges of the past and of different cultures, can we hope to realize dreams that do not spiral into nightmares. The bridge-and-structural engineer Henry Tyrrell articulated this almost a century ago, when he wrote the opening lines of the preface to his 1911 history of bridge engineering:

  Proficiency in any art or science is not attained until its history is known. Many a student and a designer finds, after weary hours of thought, that the problems over which he studied were considered and mastered by others, years or centuries before, perhaps with better results than his own.

  The earlier years of the cable-stayed bridge genre brought to the fore such individual engineering personalities as the German Fritz Leonhardt, who practiced in Stuttgart; the greatest spans today are being designed by firms that carry the names, but not necessarily the personalities, of the older generation. As the presence of Ammann and Steinman continues to be felt through the firms of Ammann & Whitney and of Steinman, Boynton, Gronquist & Birdsall, commonly referred to simply as Steinman, so does that of Leonhardt in the firm of Leonhardt, Andrä und Partner, to which in the early 1990s the patriarch still contributed his philosophy, if not his daily presence. However, not all of the most recent record spans of cable-stayed structures have been designed by firms that tie themselves in name to great engineers of the past. The Danish firm with the trendy name CowiConsult, “one of the world’s leading bridge designers,” was the one designing, in its bridge department in Copenhagen, the world’s longest cable-stayed span. Other great bridge-designing firms, such as Britain’s Acer-Freeman-Fox, and the American Sverdrup Corporation, still tie themselves explicitly to the names of their entrepreneurial ancestors, but the increasingly used anonymous collective designations of “partners” and “corporation” indicate the trend away from the small partnership team or the dominant personality of an individual consulting engineer.

  Whether their designs are attributed to forceful individuals or to anonymous corporations, cable-stayed bridges are not the only uniquely twentieth-century bridge type, nor is steel the only twentieth-century bridge material. Among other notable categories that have been very successful, not only as great structures but also as works of art, are the concrete bridges of the Swiss engineers Robert Maillart, in the first half of the century, and Christian Menn, after midcentury. Both have concealed steel, as reinforcement and as cables, in their bridges, which are primarily concrete structures. David Billington has described both Maillart and Menn as structural artists whose works are monumental pieces of sculpture as well as utilitarian works, and he has written in illuminating detail especially about Maillart’s great concrete bridges.

  As concrete has challenged steel, so have new materials challenged them both. Advanced composite materials made of glass, carbon, and polymer fibers, originally developed for the aerospace industry, are being introduced into experimental bridges, enabling them to weigh as little as one-tenth as much as conventional steel or concrete designs. One such bridge, a 450-foot-long cable-stayed road bridge over Interstate 5 in San Diego, has been aided in the materials-testing-and-design phase by a grant from the Federal Highway Administration. Because the cost of the new stuff is as much as twenty times that of conventional materials, such bridges can generally be expected to need this kind of financial support, and to remain in the experimental category until the materials become economically competitive. That will not prevent engineers from dreaming of using the newer materials to span such continuing challenges as the Strait of Messina and the Strait of Gibraltar.

  In the meantime, others continue to work with conventional materials but in unconventional forms. Among the most talked-about individual bridge designers of the late twentieth century, in any material or form, is Santiago Calatrava, whose training as both an engineer and an architect has given him and his work a special cachet. Calatrava was born in the second half of the twentieth century, in 1951, in Valencia, Spain; he studied architecture there before going to the Swiss Federal Institute of Technology in Zurich, where he became also a civil engineer. Since opening a practice in Zurich in 1981, he has been responsible for structures of dramatic space and volume, mostly connected with transportation, throughout Europe. His best-known structures tend to be his bridges, however, including the canted-towered, cable-stayed Alamillo Bridge, built for the 1992 Expo in Seville, and the Bach de Roda Bridge in Barcelona, whose inclined arches and suspender cables enclose pedestrian paths that broaden midspan to plazas to create a secure yet open space that is both protective and inviting. The engineer-architect Calatrava has been accused of being more the latter than the former, however, for he has said that he wants “to win back engineering objects like bridges for architecture.” Never mind that such talk can reopen old wounds and spark debate between the professions; in the final analysis, Calatrava’s work will be judged by the standards of both, and there are indications that he forgets some of the fundamental principles of structural-engineering art in his pursuit of appearances.

  Santiago Calatrava’s Alamillo Bridge in Seville, Spain (photo credit 7.4)

  In his Alamillo Bridge, for example, Calatrava employed a massive counterweight under the roadway to add enough tension to the longest cable so that it would be taut and not sag under its own significant weight. This added considerably to the cost of the bridge, and at the same time sacrificed structural honesty. Calatrava may see such compromises as necessary to win back bridges for architecture, yet the great engineers have never felt they were deliberately wresting bridges from architects. Even if modest spans like the ones Calatrava has designed may fairly be viewed as pieces of sculpture as well as utilitarian crossings by those who commission them, it is not necessarily obvious that bridges of record span, which will always remain first and foremost engineering problems requiring engineering solutions, should be saddled with significant extra weight, be it physical or metaphorical, for the sake of appearance alone.

  The city of St. Paul, Minnesota, recently commissioned not an engineer but a sculptor, James Carpenter, “to conceive the form for a bridge” across the Mississippi River. Among the forms developed by the New York artist, who worked in conjunction with a German engineer, was a skewed-decked, cable-stayed bridge supported by a V-shaped pylon located on an island six hundred feet from either shore. Though a computer-generated image of the bridge inserted electronically into photos of the site showed the structure to be a striking design, with elements clearly intended to make the bridge distinctive, the cost would have been more than twice that of a conventional span. It fell upon the City Council to decide whether to spend an extra $15 million dollars, even though the bulk of it might have to be requested from federal bridge-replacement funds, for a structure whose principal characteristic might be said to be difference for the sake of being different. The city’s Public Works Department was instructed to work out estimates for a cable-stayed bridge with its two spans in alignment, but the cost for such a structure was still considered prohibitive, because it was learned that “federal funds would not be available for anything beyond the least expensive design.”

  Computer-generated image of Calatrava’s unrealized East London crossing of the River Thames (photo credit 7.5)

  Though the mayor clearly preferred the artist’s signature design, political and economic rea
lities led him to lean toward a less expensive double-arch option, also produced by the artist, which was not unlike an unrealized Calatrava proposal for a single-arch crossing of the Thames in East London. In fact, this medium-priced design was the public’s choice, according to opinion surveys, but in the end the mayor recommended a third type of bridge to cross the river at Wabasha Street. This was the least dramatic, least distinctive, and least expensive of the original options—a box-girder bridge that could be built for $20 million and could include “pedestrian amenities such as ornamental lighting, windshields, pedestrian outlooks, and stair and elevator towers” to the island in the river. As in so many other cases regarding bridges and their appearance, in the final analysis the politicians and citizens of St. Paul had to settle for what they could afford. The dream of an artist, neither more nor less than that of an engineer, was alone insufficient to dictate reality.

  A public tension between means and wants often only highlights a more constant tension between function and form. Though it ebbs and flows as surely as do the waters over which many a monumental bridge is built, the ongoing push and pull between designer and financier, between engineer and architect, between engineering and art, come to our attention mainly when a question of design bobs to the surface. Disagreements over form will no doubt remain as long as there are engineers and artists who see their objectives as different. Bridge design is among the most visible and vulnerable arenas in which such competition has taken and will continue to take place, and it is wise to recall that the greatest bridge engineers have always taken an equal interest in the structural and artistic values of their designs, the mediators more often than not being safety and economics. Not that engineers are more willing than architects to sacrifice beauty for brawn, or looks for lucre; the greatest bridges that engineers have built are clearly the ones that unite and achieve both structural and aesthetic goals, and often with striking strength and economy in their context. Above all, however, engineers know that, first and foremost, their bridges must stand into the future against weight and wind and want. The most beautiful bridge, when negelected in structural design and maintenance, can become, fallen, the most ugly pile of concrete and steel. That is not bridge building.