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

by Henry Petroski

  It is said that the first time Bouch looked across the firths of Forth and Tay, he was convinced that they could be bridged, and subsequently spent his energies trying to convince others. The economic value in opening up a continuous rail link along the eastern Scottish coast was evident to all, and the final decision turned on the uncertain balance between projected cost and potential benefit. According to one account, at least, Bouch was driven by more than economics:

  [T]he simple reason for his enthusiasm was that he was a dreamer, and the most determined type of dreamer who must build what he dreams. Perhaps in the darkness at night he already believed it built, and could have put out his hand from the sheets and touched its cold iron and masonry. All creative work has its greatest reality while it is still in a man’s mind, before he begins to execute it.

  In July 1870, “after some twenty years of hawking his dream,” Bouch learned that royal assent had been given to the bill authorizing what was effectively a North British undertaking to bridge the Tay, and one year later he watched the laying of a cornerstone. The bridge was to consist principally of latticed girder or truss spans not unlike those that engineers like Simeon Post had only a few years earlier proposed to cross the Mississippi River at St. Louis. At Dundee, however, the Tay was over a mile wide, and since the combined railroad and firth alignments necessitated an oblique crossing with a wide turn at the northern bank, the Dundee side, the full length of the bridge over water was about two miles. Because of its width, however, the waterway was generally no deeper than about fifty feet, and no more than twenty feet of sand and gravel was believed to cover bedrock.

  Until bridges are actually financed and begun, there usually remains a degree of uncertainty about the real conditions upon which physical foundations must rest, for there is seldom enough money to explore every square foot of bottom for a bridge that is as yet only some engineer’s dream. In the case of the Tay, excavation was to take place within large cylindrical caissons. Unlike those of the St. Louis Bridge, these caissons were not pressurized with air, and workers protected only by diving bells prepared the bottom for the construction of heavy brick piers. Once work was begun, however, it was soon evident that the riverbed conditions were not so substantial as test borings had indicated, and Bouch redesigned the piers to consist of groups of cast-iron columns on a wider base. This reduced the pressure on the sandy bottom to less than half of what was originally intended. The girders for the superstructure were fabricated near the shore, floated next to piers, and jacked into place. After about six years of work, in September 1877, the first train passed over the bridge.

  The Tay Bridge consisted of eighty-five individual spans, the eleven greatest of which were 245 feet in length and known as the “high girders,” so designed to allow trains to pass through rather than over, and thereby to offer the least impediment to shipping. While no one of its girders was anywhere near record length, the Tay Bridge was overall the longest bridge in the world by far when it opened officially on June 1, 1878. A year later, Queen Victoria crossed the bridge and knighted Bouch for his accomplishment. A half year after that, during a fierce storm, the high girders of the bridge fell or were blown into the Tay, carrying with them the evening train from Edinburgh to Dundee and all of its seventy-five passengers. There were no survivors.

  Though Bouch maintained that the “capsizing” of some of the cars of the train off the track and into the sides of the high girders brought them down, a court of inquiry discovered major flaws in the design and construction of the Tay Bridge. It was found, for example, that Bouch had grossly underestimated the effect of the strong winds that could develop along the firth. Upon being questioned about this during the inquiry, he gave no sign of having reconsidered the issue:

  The Tay Bridge after the collapse of its high girders, on December 28, 1879 (photo credit 3.3)

  Q: Sir Thomas, did you in designing this bridge, make any allowance at all for wind pressure?

  A: Not specially

  Q: You made no allowance?

  A: Not specially.

  Q: Was there not a particular pressure had in view by you at the time you made the design?

  A: I had the report of the Forth Bridge.

  The Forth was, of course, the other firth that had to be crossed by a bridge, and early in his design considerations Bouch had learned indirectly from Astronomer Royal Sir George Airy how great a wind force might be expected to push against each square foot of bridge area that might be thrown across the Firth of Forth. Airy, writing from the Greenwich Observatory, acknowledged that “for very limited surfaces, and for very limited times, the pressure of the wind does amount to sometimes 40 lb. per square foot, or in Scotland to probably more.” He suggested that, on average, the entire bridge would experience a pressure of ten pounds per square foot, which led Bouch to believe he could for all practical purposes ignore the effects of the wind on any firth, even though French and American engineers of the time were assuming wind forces five times Airy’s average amount and accordingly designing bracing and connections to resist them. The Tay inquiry concluded that “the fall of the bridge was occasioned by the insufficiency of the cross bracing and its fastenings to sustain the force of the gale.”

  In addition to the inadequacy of the superstructure of the Tay Bridge to resist the wind, its piers were found to have been poorly constructed. After the accident, the iron was discovered to have been badly cast. It had tapered boltholes, resulting in loose connections; it had uneven thicknesses, causing unintended variations in strength; and it had large voids left by cast-in air pockets that were subsequently filled in at the Wormit Foundry with mixtures known there as Beaumont Egg, consisting of “beeswax, fiddler’s rosin, and the finest iron borings melted up, and a little lamp black.” The filler material was smoothed down and painted over—resulting, of course, in weak spots that were merely cosmetically sound.

  As authorized by the Regulation of Railways Act of 1871, the Board of Trade had appointed a court of inquiry comprising three members: William Henry Barlow, president of the Institution of Civil Engineers; Colonel William Yolland, chief inspector of railways; and Henry Cadogan Rothery, wreck commissioner. Only the first two were engineers, and professional loyalties appear to have surfaced when the time came to draft the final report “upon the circumstances attending the fall of a portion of the Tay Bridge.” Though there seemed to be no substantial disagreement as to contributing factors to the failure, the two engineers could only conclude that they had “no absolute knowledge of the mode in which the structure broke down,” and hence were reluctant to place blame too squarely and explicitly on Thomas Bouch. Rothery, on the other hand, had no such uncertainty, and he chose to write a separate report. (The official report, which was issued in June 1880 and circulated “to both Houses of Parliament by Command of Her Majesty,” consisted of both views.) Rothery felt it was his duty to call it as he saw it:

  We find that the bridge was badly designed, badly constructed and badly maintained and that its downfall was due to inherent defects in the structure which must sooner or later have brought it down. For these defects both in the design, the construction, and the maintenance, Sir Thomas Bouch is, in our opinion, mainly to blame. For the faults of design he is entirely responsible. For those of construction he is principally to blame in not having exercised that supervision over the work, which would have enabled him to detect and apply a remedy to them. And for the faults of maintenance he is also principally, if not entirely, to blame in having neglected to maintain such an inspection over the structure, as its character imperatively demanded.

  Sir Thomas went into seclusion and died four months later, at fifty-eight. He would be remembered in history not for his legacy of three hundred miles of railway functioning properly in England and Scotland but for the failure of his Tay Bridge. Not only a successful bridge across the Tay, but also one across the Firth of Forth were the legacies Bouch must have dreamed about. Indeed, his design for a bridge across the deeper Firth
of Forth had been under construction for over a year when the Tay fell. Bouch’s scheme for a suspension bridge of two great spans of sixteen hundred feet each might have resulted in an accomplishment that would have overshadowed the Brooklyn Bridge, then under construction in New York. With the disaster at the Tay, however, work at the Firth of Forth was suspended, and there was a general loss of confidence in the project, especially when it came out that Bouch had designed his bridges, on average, for only ten pounds of wind pressure. Even then, Bouch’s design was not formally abandoned by the Railway Board until a full year after the Tay disaster, and six months after the court of inquiry’s report, in part because there was still a great desire for bridges across the Scottish firths.

  The Tay Bridge had, after all, operated successfully for more than a year, and it had demonstrated the great economic value of bridges across the firths. It should not be surprising, then, that a new Tay Bridge, which would have two tracks rather than one and would accordingly be wider and more substantial, was soon proposed. Less than eighteen months after the first Tay Bridge collapsed, plans for a new one were submitted for parliamentary approval, which was required for all civil-engineering works. The new undertaking was overseen by the firm of Barlow, Son & Baker, and the engineer was to be William H. Barlow himself, with his son, Crawford Barlow, as his assistant. The second Tay Bridge was to be constructed sixty feet upstream from the first, with piers spaced the same distance apart so that the sound girders, which did not fall, could be easily reused. The high girders were redesigned to much more substantial standards than Bouch’s, including the ability to resist a wind pressure of fifty-six pounds on every square foot, and all cylinders of the supporting piers were to be tested well beyond the load they were expected to carry. There was naturally great local interest in the design of the replacement structure, and the piers of the “New Viaduct” and the “Old Bridge” were contrasted in the October 18, 1881, issue of the Dundee Advertiser as follows:

  The massive character of the new structure as compared with the old is obvious at a glance, especially (1) the greater lateral stability from the substitution of twin piers for the single pier below, and the increased width for the double line of rails above; and (2) the greater vertical stability from the diminished height of the superstructure and the arched formation at the upper junction of the piers.

  As in the wake of all major failures, the design of the new structure had improved features that far exceeded correcting the deficiencies of the old. With favorable public opinion thus assured, tenders were invited, and that of William Arrol & Company of Glasgow was accepted late in 1881. Construction began the next year and was completed in 1887. The stumps of the original bridge piers remain in place today, serving as tidal breakwaters for the much more substantial piers of the second Tay Bridge, and as a stark reminder of the accident and its victims.

  2

  When the time came, in 1881, to have a new bridge designed for the Firth of Forth, proposals were invited by the Railway Board from its consulting engineers, Sir John Fowler, William Barlow, and Thomas E. Harrison, who had been set up as a panel to reexamine Bouch’s scheme. John Fowler, then approaching sixty-five years of age, was younger by almost five years than Barlow and by almost ten years than Harrison. Fowler had begun his training as a civil engineer at sixteen, given evidence before Parliament by the time he was twenty-one, and been in charge of a railway-construction project at twenty-two. In his mid-forties, he had been involved in seventy to eighty “major schemes” a year, and it has been estimated that he must have been involved in over a thousand jobs over the course of a professional career that spanned more than sixty years, working with at least fifty different assistants. Talented assistant engineers were clearly essential to someone like Fowler, and his assistant on the Firth of Forth project, Benjamin Baker, was among the best. Baker, thirty years Fowler’s junior, began work in his London office on the Metropolitan Railway project, the first link in the London underground when it opened in 1863, but preferred designing long-span bridges.

  The high girders of the rebuilt Tay Bridge, as they stand today, with the stumps of the original bridge still visible in the water (photo credit 3.4)

  A truss- or girder-bridge design was not appropriate for the Firth of Forth, because the many piers that would have had to be sunk in deeper water would have presented an engineering challenge and an unwanted expense. Besides, it was a girder bridge across the Tay that had failed, so adverse public opinion would have had to be overcome. Suspension bridges had long been suspect in Britain for rail traffic, but John Roebling’s successful one over the Niagara Gorge had put the form in a new light. However, the problems with wind, and the fact that the now abandoned design of Bouch had been of the suspension type, again cast it in disfavor. Fowler and Baker were thus inclined, for both technical and nontechnical reasons, to look to different bridge forms entirely.

  The location that Bouch had identified was ideal, in that the firth was relatively narrow there, albeit relatively deep, and in that about midway between the shores, at Queensferry and South Queensferry, was an island, or “garvie” in Scots. It was said to be named Inchgarvie because its representation on a scale map was an inch long; coincidentally, its shape also resembled that of a small herringlike fish called a garvie. According to the engineer Baker, who later lectured on the design of the bridge, the area “should be well enough known to every reader of fiction,” for it was the setting of Robert Louis Stevenson’s Kidnapped, whose hero was taken “at the very spot where the bridge crosses.” However, what presented themselves here to Baker and Fowler were not fictional settings but physical conditions for a bridge with piers only on the island and on or near the shores, which thus required two free spans each on the order of seventeen hundred feet. No such bridge had ever been built anywhere in the world.

  What was being constructed in recent years in Europe and America, however, was a somewhat new type of bridge that was being used to span increasingly greater distances with as few supports in the water as possible. One of the earliest of these bridges was completed in 1869 over the River Main at Hassfurt, Germany, by the Bavarian engineer Heinrich Gerber, who a few years earlier had been granted a patent for the design, which came to be known as a Gerber bridge. In his bridge, which looked not unlike the high dark bridge that today carries the Pulaski Skyway over New Jersey marshland, the girder depth varied along the length of the bridge, which was articulated at strategic locations in order to simplify design calculations and to allow for minor settlement of the piers without exerting undue stress on the superstructure.

  Gerber’s concept had considerable appeal to engineers, and many other “Gerber bridges” were built according to similar principles, in part because bridge engineering generally had evolved to the point where this bridge type was a natural solution for supporting the increasingly heavy loads of commerce. Thomas Bouch had, in fact, designed and built one of the first such bridges in England in 1871, at Newcastle. This was apparently forgotten when the Tay disaster occurred; otherwise Bouch’s involvement might have given this genre a bad name as well, at least among the general public and politicians.

  In America, the expanding railroads presented many opportunities to build new bridges, and by 1877 a Gerber-type bridge with three 375-foot spans was carrying the Cincinnati Southern Railway over the Kentucky River. The engineers of this bridge were Louis G. F. Bouscaren, who had been an assistant engineer on the Eads Bridge, and Charles Shaler Smith, who was responsible for the Illinois approach to that bridge. Shaler Smith was born in Pittsburgh in 1836 and was to die in St. Louis fifty years later. As a child, he attended private school, but he apparently had no formal engineering training. Nevertheless, his work with Gerber-type bridges was instrumental in introducing the form to America, and his writings on comparative analyses of different truss types and on wind pressure on bridges were important contributions to bridge design in America.

  In 1883, as construction was beginning on the Firth of F
orth bridge, a Gerber-type with a 495-foot span was completed, built for the Michigan Central and Canada Southern railways. It was almost 240 feet above the Niagara Gorge and just south of Roebling’s suspension bridge, also known as the Grand Trunk Bridge. The chief engineer was Charles Conrad Schneider, born in 1843 in Saxony, where he was trained and practiced as a mechanical engineer before coming to America in 1867. He began work here, as many immigrant engineers then did, as a draftsman. His early work with the Rogers Locomotive Works in Paterson, New Jersey, led to his involvement with railroad companies, and before long he was in charge of engineers in the New York office of the Erie Railroad, whose chief engineer was Octave Chanute. Born in Paris in 1832, Chanute moved to America with his family when he was eight years old, attended private schools in New York, became an assistant chainman on the Hudson River Railroad while still a teenager, and worked his way up the railroad ladder. In the process, he had gained enough experience in matters of the maintenance of ways to become involved in extending existing railroads. From 1867 to 1869, Chanute designed and oversaw construction of the first bridge over the Missouri River, at Kansas City. Later in life, he became interested in the nascent field of aviation, which led ultimately to the association of his name with an air-force base in Illinois.

  In Chanute’s office, Schneider had the responsibility for checking the bridge plans submitted by bridge companies, a practice not generally carried out by railroads at that time. In particular, Schneider was responsible for checking the strain sheets, which showed the portion of the load that each member of a bridge was designed to carry. It was a natural development for bridges also to come to be designed in railroad offices. Once designs became specified by the railroad, rather than as part of a lump-sum contract that included everything from design to construction by a bridge-building company, the Erie Railroad began to buy its bridges by weight, for the price of materials was the chief determinant of cost. The practice of letting bridge contracts by the pound soon became widespread.