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

by Henry Petroski

  Among the problems confronting Modjeski, and other Chicago bridge firms, was the design of bridges with movable roadways. One relatively new type of drawbridge was the bascule bridge, which operated on a seesaw principle, but with the counterbalancing side greatly reduced in length by the use of massive counterweights. Though these bridges had the advantage of not taking up much space in crowded cities, they had the disadvantage of requiring a lot of expensive dead weight to function properly, and the added complexities associated with mechanical movements. In time, Strauss came up with a new scheme, which employed concrete for the counterweight. However, since this relatively inexpensive material was about two-thirds less dense than the cast iron that was being used in counterweights, Strauss’s early designs had ungainly proportions. Another of his ideas, to use a new kind of trunnion bearing as the “element of movement,” was reportedly also ridiculed by Strauss’s superiors at Modjeski, and so he left.

  In 1902, Strauss thus struck out on his own as a consulting engineer, and in 1904 he established the Strauss Bascule Bridge Company. Over the next decade or so, a series of significant patents were granted for movable bridges of Strauss’s design, and his firm prospered somewhat, eventually becoming the Strauss Engineering Corporation, a consulting firm with offices in Chicago and San Francisco. Though the firm’s early specialty was movable bridges, not all of which were by any means unattractive, it also designed and built such unusual structures as the Aeroscope, portable searchlights employed during World War I, and reinforced-concrete railroad cars. Strauss would go on to serve as consulting engineer on such significant bridges as the Arlington Memorial Bridge in Washington, D.C., whose center arch is in fact a double-leaf steel bascule bridge with concealed counterweights, as well as the Bayonne Bridge and the George Washington Bridge in New York.

  O’Shaughnessy and Strauss began to talk seriously about the possibilities of bridging the Golden Gate in about 1919. The city engineer’s office provided location maps, surveys, and the like, while Strauss’s company worked on design. Soundings of the strait made by the U.S. Coast and Geodetic Survey in early 1920 were not encouraging. O’Shaughnessy shared the data not only with Strauss but also, unbeknownst to Strauss, with Frank McMath, who was with the Canadian Bridge and Iron Company in Detroit, and with Gustav Lindenthal. By mid-1921, Strauss had delivered a proposed bridge design to O’Shaughnessy who did not make it public for some time, perhaps because of its unconventional appearance. It was an ungraceful hybrid, half cantilever and half suspension bridge, but it did carry the very attractive price tag of about $17 million. O’Shaughnessy must certainly have thought that Strauss’s keen sense of economy counterbalanced at least in part the aesthetic shortcomings of the design—especially when the less cost-conscious Lindenthal eventually came in with a proposal requiring $60 million at a bare minimum, and other estimates had run in the $100-million range; McMath never did submit a plan. Thus Strauss’s design came to be used “to help stimulate public interest in building a bridge and to determine what financial and political support really existed for such a project.”

  An attractive booklet titled Bridging “The Golden Gate,” prepared by Strauss and privately printed by O’Shaughnessey and Strauss in 1921, contains a concise and convincing argument that the project was technically and financially sound. The “new cantilever-suspension type of long span bridge” that was “conceived and patented” by Strauss combined the cantilever and suspension principles to reduce the length of cable required and at the same time increase the overall stiffness of the structure. The estimated cost of the bridge, laid out succinctly on a single page, was said to be “so reasonable” that it could be paid for either directly by San Francisco, Marin County, and others who would benefit from it, or operated as a toll bridge and thus would pay for itself, according to Strauss’s estimate of revenue.

  Strauss began to talk about his proposal before small groups in Marin County and northern California. Though he was apparently a terrible speaker, his idea for a bridge that would bring clear advantages in real-estate development and commerce gained increasing support. By 1923, enough momentum had been generated so that legislation was drafted authorizing a Golden Gate Bridge and Highway District, which would create the first local tax district of any significance for the specific purpose of building a bridge. Within months, the legislation was in place. In the meantime, an engineer was added to Strauss’s staff to work out the design of a bridge larger than any the office had ever tackled.

  Joseph Strauss’s 1921 proposal for a bridge across the Golden Gate (photo credit 5.14)

  This engineer, Charles Alton Ellis, was born in Parkman, Maine, in 1876. His formal education consisted of an A.B. in mathematics and Greek from Wesleyan University, and he learned engineering on the job at the American Bridge Company. He learned so well that in 1908 he became an instructor in civil engineering at the University of Michigan, and he was a professor of structural and bridge engineering at the University of Illinois when Strauss hired him, in 1922, to work on the Golden Gate Bridge. While Ellis did the essential design work, Strauss continued to promote his bridge scheme, which had encountered the familiar challenge of presenting an obstacle to navigation. Final War Department approval was to come slowly, and in the meantime Strauss engaged Professor George F. Swain of Harvard University and Leon Moisseiff to serve on a board of consultants. Both endorsed Strauss’s plans as workable, if not graceful, and at his request Moisseiff prepared plans for a comparable suspension bridge for the site. His design, with a four-thousand-foot center span, represented more than a doubling of the suspension-bridge length, itself a record, that was soon to be realized in the Moisseiff-designed Delaware River Bridge, and was to be an even greater leap in the state of the art than Ammann’s Hudson River Bridge, then proceeding within the secure financial base of the Port of New York Authority. Though cost estimates for both Strauss’s and Moisseiff’s versions of a Golden Gate Bridge remained attractive at under $20 million, some supporters were beginning to have second thoughts about providing financial guarantees to the project and going ahead without further studies. By the end of 1928, however, the Golden Gate Bridge and Highway District was incorporated.

  A page from a promotional booklet showing Joseph Strauss’s concise cost estimate for a Golden Gate Bridge (photo credit 5.15)

  Estimated Cost, Proposed Golden Gate Bridge

  With incorporation, the district could get down to business, which meant fixing on a design that would be approved by the War Department, going to the voters for their approval, issuing bonds to raise capital, and letting contracts for the actual construction. Although Strauss was indeed chief engineer of the district, his or any other design still required the approval of the politically appointed members of the district’s technical board. Among the engineers who were approached anew for designs in 1929 were Ammann, Lindenthal, and Modjeski. Needless to say, Strauss was not in their league when it came to long-span suspension bridges. Indeed, he had not designed or built a single one. Furthermore, questions that had arisen about foundation conditions made it increasingly doubtful that the heavier cantilever design, even if modified, could be supported. In order to bolster his credibility for the job, Strauss had engaged Ammann and Moisseiff as consulting engineers to him as chief engineer. They were already engaged in the building of the George Washington Bridge, of course, which not only put them in the forefront of bridge building but also made their full availability for a West Coast project somewhat questionable. Ammann had accepted Strauss’s offer reluctantly, in part because he was unsure about the organizational structure that had him working for the chief engineer rather than for the district board, and he did not want to appear to be competing with consulting engineers in private practice. No doubt he avoided quitting the project outright because he must have realized that Strauss was not only determined but also politically effective enough to get the world’s largest bridge built, and he did not want to be left out of participating in it. In the final arrangeme
nt that was worked out, Ammann and Moisseiff continued to be associated with Strauss, as members of a board of engineers appointed by the Bridge and Highway District, and he was to be associated with them as consulting engineer on the Bayonne and George Washington bridges. The Golden Gate board was chaired by Strauss, who was in fact also chief engineer of the project. O’Shaughnessy, who was so instrumental in getting the bridge idea off the ground in the first place, was disappointed, to say the least, at being left out entirely; he would eventually join the forces completely opposed to the bridge.

  The newly printed letterhead of the Office of the Chief Engineer listed Strauss in that position, with Charles Ellis next in line as designing engineer. Three consulting engineers were also listed: Ammann, Moisseiff, and, as the local representative, Charles Derleth, Jr., dean of the College of Engineering at the University of California at Berkeley. Derleth had been chief engineer of the Carquinez Strait Bridge at Vallejo, California, designed by David Steinman and William Burr, and had also been a critic of Strauss’s early design. Dominating the letterhead, however, was not the names of the engineers but the seal of the district, showing a combination cantilever-suspension bridge that strongly resembled the chief engineer’s original design. Yet Strauss’s ambition was greater than his pride: he came to acknowledge that a suspension bridge of the kind suggested by Moisseiff not only would be lighter and less expensive to build, but also could be completed in less time. Because of the Depression, cost estimates for construction had fallen even since Strauss’s original estimate, and he had made his bid for the job more attractive to the Bridge and Highway District by agreeing to terms that represented “the lowest ever written for a bridge job of such magnitude.” His firm’s fee was to be 4 percent, which compared rather unfavorably with the 7 percent recommended by the American Society of Civil Engineers. Though this meant that Strauss would have to run a bare-bones engineering-design office for the project, he was apparently willing to compromise in many ways to get a chance to pursue his dream bridge. (Strauss had set his fee with the understanding that the bridge plans would be prepared sequentially as construction progressed. When the district directors ordered the plans prepared all at once, in order to go for lump-sum bids, Strauss sued for an additional 1 percent fee and won.)

  Charles Ellis was put in charge of the detailed design at the Strauss Engineering Corporation, with Moisseiff checking all the calculations. Early plans called for the Marin pier to be two hundred feet out in the water, but a decision was made to put it adjacent to the shore, thus reducing costs while at the same time increasing the length of the main span of the bridge to forty-two hundred feet. The south pier and anchorage were to be the focus of controversy, however. Berkeley Professor Andrew C. Lawson had been retained as consulting geologist, and he had serious doubts that rock conditions near the San Francisco shore could support a tower of the size required. This sparked prolonged debates about how sound the rock really was and how the bridge would survive movement of the tower and anchorage during an earthquake. (The span was to be built parallel to the San Andreas Fault, only six miles away.) It was Ellis who assured critics on the latter point. The major design challenge Ellis faced, however, was that of the towers themselves. The decision was that they should be of steel and that their design was of prime importance not only structurally but also aesthetically. Heavy diagonal bracing was to be confined to below the roadway, with the upper portions of the towers braced only with horizontal members. Strauss suggested that the skyscraper towers have “the stepped-off type of architecture” that was so prominent a feature of the contemporary Chrysler and Empire State buildings, and this was followed. The first consulting architect retained by Strauss was John Eberson, an Austrian-born electrical contractor who had taught himself architecture and become a prominent theatrical-set designer. He produced a design that had archlike openings above the roadway and at the tower top, thus concealing with architectural embellishments two of the four strong horizontal structural members that Ellis and Moisseiff knew were needed to provide the proper stiffness.

  As the design of the bridge proceeded, it was learned that the San Francisco approach roads had been realigned by federal highway authorities, requiring a new arrangement for the toll plaza and thus further architectural work. Since the New Yorker Eberson had proved to be too expensive for Strauss’s budget, and since Strauss knew that a locally prominent architect might be more helpful in gaining political and financial support for the bridge, he hired Irving F. Morrow as the new consulting architect. Born in Oakland, California, in 1884, Morrow studied architecture at Berkeley and at the Ecole des Beaux-Arts in Paris. Though he had no prior experience in bridge work, he was a daily commuter by ferry between his office in San Francisco and his home in Berkeley, and thus had observed the ever-changing conditions of light, shadow, and color in the Golden Gate. It would be Morrow’s Art Deco design treatment of the towers, incorporating such façade details as the vertically faceted aluminum fascia panels to cover the heavy steel structural horizontals, that would give the bridge much of its dramatic effect in the changing light. He would also play a key role in helping to decide on such details as bridge railings and paint colors that would make the Golden Gate Bridge one of the most distinctive in the world.

  Before the bridge could be decorated, however, it had to be financed and built, and chief engineer Strauss presented his structural and architectural report to the Bridge and Highway District Board on August 27, 1930, just two weeks after a construction permit was signed by the secretary of war. A referendum was announced, and on November 4 the voters approved a $35-million bond issue by more than three to one. The successful outcome was due in no small part to the fact that the construction of the George Washington Bridge had gone so smoothly, technically and financially, its opening now only a year away. Even before that occurred, however, Engineering News-Record would remark that “the Golden Gate Bridge is a fact today because the Fort Lee bridge was built yesterday. It was the latter project that attuned the public mind to the possibility of financing such huge enterprises.” The sale of the Golden Gate bonds would go no more smoothly than the continuing design, but construction would begin within three years.

  The greatest tensions in the bridge’s design proved to be not in the cables or anchorages but in the interpersonal relationship between Strauss and Ellis. With the bond issue passed, there was great optimism and pride in the project, and there were audiences wanting to hear all about it. Among these was the first West Coast conference of the National Academy of Sciences, to be held on the Berkeley campus. The university’s president, Robert G. Sproul, proposed to Dean Derleth that the scientists be told about the significant engineering problems that were being overcome in designing the local bridge, and Derleth arranged for Ellis to make the presentation. Though he acknowledged that Strauss was the boss of the project, when it came to the design of the bridge itself, Ellis stated, “Mr. Strauss gave me some pencils and a pad of paper and told me to go to work.” Derleth himself, at a luncheon hosted by skeptics about the design where Strauss expected him to silence the critics, appealed to the authority not of Strauss but of Ellis, “who stands high as a structural bridge engineer.” The dean went on to describe how Ellis was in charge of the first design and how Moisseiff’s theory was used to lighten the trusses employed. Finally, Derleth assured the audience that neither Ellis nor Moisseiff had made any errors in their lengthy and complex mathematical calculations, “even if Mr. Strauss and Mr. Ammann and I do not know anything about the subject.” Thus revealed to be just another manager, Strauss was not happy.

  Strauss had indeed put Ellis in charge of the design, and the staff of Strauss Engineering was to assist him, especially with calculations for the tower. But since Ellis had found at least one of Strauss’s staff unprepared to work on “a problem of this nature,” and others were busy on a potentially very profitable bascule-bridge job, he took on much of the work himself, having it checked by Moisseiff and his own staff. As the bri
dge seemed to be associated more and more with Ellis, Strauss tried to rein him in by pushing for milestones and deadlines. Ellis defended his reluctance to promise when things would be completed or to delegate calculations by describing the structure as one requiring extrordinary research and theoretical work. Yet times were tough; Strauss’s corporation apparently felt financial pressures, and he was little interested in paying for work he thought to be a frivolous academic exercise. In late 1931, Strauss wrote to Ellis that he should turn the work over to his assistant, Charles Clarahan, Jr., and take a vacation. This left one person in the office working on the bridge; before his vacation was over, Ellis received a letter from Strauss effectively laying him off and instructing him to turn over the job to Clarahan. According to Strauss, “the structure was nothing unusual and did not require all the time, study and expense” that Ellis had been devoting to it.

  It was the Depression, of course; Ellis remained out of work until he joined the faculty of Purdue University in 1934. In the meantime, he continued to work on the tower design, writing to the consulting engineers about his concerns. Eventually, Ellis “vexed” even Moisseiff, who wrote to Derleth that he and his staff found nothing wrong with the methods used by the Strauss Engineering Corporation. The tower design had been verified by means of a stainless-steel model at Princeton University, and at a program there to describe the tower design and testing, Moisseiff described how the stiff frame design, “without diagonals, was adopted for the sake of appearance and how this complicated the analytical design.” Though he also commented on the “ingenious method of analysis” worked out by Ellis, he credited Frederick Lienhard of his own office with the computations that checked with the model tests.