No bridge collapses quickly. Just as the safe completion of a bridge is measured in years, the failure of a bridge can be reckoned in the same way. Though the chaotic physical dismemberment of the south arm of the Quebec Bridge took no more than fifteen seconds, the more orderly prelude to the catastrophe began long before.
It began in the summer of 1897, when the consulting engineer Theodore Cooper attended the annual convention of the American Society of Civil Engineers in Quebec City. A former director of the society, Cooper was one of the most respected bridge builders of the time. He made an excursion to the proposed site of the Quebec Bridge and within a week expressed an interest in giving the Quebec Bridge Company the benefit of his expertise.
Cooper's tender of interest was hardly unbidden. The Quebec Bridge Company had been sounding out American bridge engineers as consultants because its own chief engineer, Edward A. Hoare, had never worked on a bridge with a span longer than three hundred feet. Cooper was a proud, confident man, fiercely devoted to his calling. He had been graduated as a civil engineer from Rensselaer Institute (now Rensselaer Polytechnic) in 1858 at the age of nineteen. Enlisting in the Navy in 1861, he served as an assistant engineer of the gunboat Chocura for the last three years of the Civil War, then moved on to a teaching post at the United States Naval Academy. After a tour of duty in the South Pacific, he resigned from the Navy in July 1872. In May of that year, Capt. James Eads appointed Cooper the inspector of steel manufacturing for Ead's most important engineering work, the St. Louis Bridge.
If the Navy laid the groundwork for Cooper's career, the St. Louis Bridge launched it along a high trajectory. Captain Eads moved Cooper up quickly, placing him in charge of erection at the bridge, which was the most ambitious use of the cantilevered method of erection yet attempted. Cooper performed his duties admirably -- once going without sleep for sixty-five hours during a crisis, another time wiring Eads at midnight to warn him that the arch ribs were rupturing, a potentially disastrous condition that was remedied by following the instructions Eads immediately wired back. Upon completion of the work in 1874, Cooper found himself much in professional demand. By 1879, after resigning as the superintendent of Andrew Carnegie's giant Keystone Bridge Company in Pittsburgh, Cooper was able to set up as an independent consulting engineer in New York.
The projects he undertook there were notable and prestigious. His works included the Seekonk Bridge in Providence, the Sixth Street Bridge in Pittsburgh, and the Second Avenue Bridge in New York. He moved through the most rarefied atmosphere of his profession, but unlike his mentor Eads, he never oversaw a truly heroic masterwork. The Quebec Bridge, viewed in that light, was irresistible to Cooper. He said the bridge would be his last work. It would stand as the crowning achievement to an elegant career.
Almost two years would go by before Cooper's affiliation with the Quebec Bridge Company became formal. The financially troubled company had a history of moving slowly -- or not at all. Incorporated by an Act of Parliament in 1887, it had accomplished virtually nothing in its first eleven years. In March 1899 officials of the company met with Cooper in New York and arranged for him to review the bids for the long-awaited bridge contracts. All prospective contractors' plans and tenders were sent to him, as well as the clear message that he should keep in mind the weak financial position of the Quebec Bridge Company.
The Quebec Company had been in close touch with the Phoenix Bridge Company of Phoenixville, Pennsylvania, since 1897, and the Phoenix Company had already submitted preliminary plans for the bridge. Now that the bidding was open, the Quebec Company's desire to give the Phoenix Company the contract for the superstructure was barely concealed.
On April 14, 1899, John Sterling Deans, the chief engineer of the Phoenix Bridge Company, wrote to Edward Hoare, his counterpart in Quebec: "Dear Mr. Hoare--Mr.Szlapka [Phoenix's chief design engineer] and I were with Cooper the greater part of yesterday, and you will be glad to learn that there was not a single vital or important criticism or mistake found in our plans...Mr. Cooper, however, somewhat upset me, by making the following remark, which of course I understood was entirely personal and without any full knowledge of the situation. He said: 'Well, Deans, I believe that all of the bids will probably overrun the amount which the Quebec Bridge Company can raise, and that the result will be...that all of the bids will be thrown out and a new tender asked on revised specifications and plans.' Mr. Cooper undoubtedly desires to be perfectly fair, but...does not fully understand the situation. I trust, therefore, that you will give his report the most careful scrutiny, and get it in the right shape before it is submitted."
There were more collegial letters between Phoenixville and Quebec, and both Deans and Hoare stayed in close touch with Cooper. Later Cooper would maintain that no pressure had been brought to bear on him. In any event, on June 23, 1899, he sent his findings to the Quebec Bridge Company. "I therefore hereby conclude and report," he wrote, "that the cantilever superstructure plan of the Phoenix Bridge Company is the 'best and cheapest' plan and proposal."
Those three words -- "best and cheapest" -- became a touchstone for Cooper in his approach to the bridge. His subsequent letters to Quebec and Phoenixville are seasoned with references to the fiscal consequences of major design decisions. None of the parties involved ever placed costs before safety outright, but their aim was clearly to build a bridge that could bear the twin loads of its own mechanical burden and the Quebec Bridge Company's financial burden.
The Quebec Company had no cause to be dissatisfied with Cooper's scrupulous concern for its ledger books -- and had every reason to be confident of his ability to oversee the building of a good bridge. On May 6, 1900, Cooper was appointed the company's consulting engineer for the duration of the work. He had become, finally, the master builder on a project of historic magnitude.
Five days before his formal appointment on May 1, Cooper exercised his authority by recommending that the span of the bridge be lengthened from sixteen hundred feet to eighteen hundred feet. His explanation for this major design change revealed an attentiveness to both engineering and expense. Piers constructed in deeper water would be subject to the heavy ice floes of the main channel. Closer to shore, they would be less vulnerable -- any quicker to build, speeding up the completion of the entire work by at least one year. To keep down the increased cost of steel in the superstructure for an eighteen-hundred-foot span, Cooper recommended another major design change: modified specifications that would allow for higher unit stresses. [see Exhibit A1]
His recommendations were approved at Quebec almost as a matter of course. And then, for the next three years -- as work proceeded on the substructure, the anchorages, and the approach spans -- practically nothing was done to prepare for the engineering difficulties posed by the eighteen-hundred-foot span and the higher allowable stresses.
Once again, money was the root of inaction. Short of funds as usual, the Quebec Bridge Company was making no promises to anybody about its capacity to pay for the bridge's superstructure once the preliminary work was done. For all the goodwill between Phoenixville and Quebec, the Phoenix Bridge Company was politely declining to enter into a contract until payment might be assured.
And so, while the huge size of the bridge cried out for preliminary tests and research studies, none were conducted during the long slack period between 1900 and 1903. It was not in the interests of the Phoenix Bridge Company to go out-of-pocket on research costs it might never recoup, and it was plainly impossible for the Qubec Bridge Company to provide the funds. An unspoken assumption became necessary instead: Theodore Cooper's experience and authority were sufficient to confer success upon the untested work.
Then in 1903 the Canadian government guaranteed a bond issue of $6.7 million to pay for the work. With that, the torpor enveloping the project turned into humming activity. Phoenix and Quebec entered into serious contract discussions while design engineers and draftsmen struggled to meet the urgent demand for detailed drawings.
Three years of opportunity for deliberate preparation had been lost. In the rush to provide drawings so that the steel for the bridge could be fabricated with little loss of time, there was no recomputation of assumed weights for the bridge under the revised specifications. It was an oversight of critical importance, and Theodore Cooper did not intervene. He decided to accept the theoretical estimates of weight that the Phoenix Bridge Company had provided.
During the three languid years that preceded the project's lurch into progress, Cooper visited the site of the bridge three times. His third visit, in May 1903, when he was sixty-four, would be his last. After that, he would decline requests that he come to Quebec. His health was poor, he said, and his physician had advised him not to travel. From that point on, he would oversee the construction of the world's longest spanning bridge from his office in New York.
Cooper's health may indeed have been fragile, but he was hardly an invalid, commuting almost daily to his office at the foot of Manhattan Island from his home on West Fifty-seventh Street. The only specific references to illness in his letters to Quebec and Phoenixville cite "grippe" and "fatigue" as his reasons for not being able to be there.
In fact, he had never much appreciated being there in person as a consulting engineer. He regarded on-site visits as unproductive and largely devoted to atmospherics. From his earliest days in private practice, Cooper had insisted on a clause in his contract that limited his on-site responsibilities to a maximum of five days a month. When the Quebec Bridge Company's secretary, Ulric Barthe, at one point brought Cooper's attention to that understanding, Cooper replied that the five days were not an obligation but a limit, implying that it was a limit not to be abused. With his health now weakened, the five-day limit became academic.
The question of his health also caused Cooper to offer what amounted to a pro forma resignation in 1904. On a visit to New York, S.N. Parent, the president of the Quebec Bridge Company, asked Cooper when he might see him in Quebec again. Cooper's answer was never. He then asked to be relieved of his responsibilities, but Parent would not hear of it. A short time later Cooper made the same offer to John Deans of the Phoenix Bridge Company, who also refused to treat it seriously. The matter was laid to rest, and Cooper refrained from pressing it. Feeling, as he later said, "a pride and a desire to see this great work carried through successfully, I took no further action".
In the summer of 1903, while Cooper was still well enough to travel, his pride in the great work took him to Ottawa. He was incensed. Collingwood Schreiber, the chief engineer of the Department of Railways and Canals, had suggested that the department hire its own consulting bridge engineer to review and correct the detailed drawings of the Quebec Bridge -- after Cooper had seen them -- and then submit them to Schreiber for final approval. Robert Douglas, an engineer in Schreiber's department, had reviewed Cooper's new specifications for the eighteen-hundred-foot-span bridge and had criticized the high unit stresses. "Considering that the American government in several cases appointed four or five engineers to consider and determine unit stresses of unexampled magnitude," Douglas would say later, "I thought that this matter was too important to be left to the judgment of Mr. Cooper." But confidence in Cooper was the byword just then, and foresight was at a premium.
Upon learning of Schreiber's proposal, Cooper wrote angrily to Quebec: "This puts me in the position of a subordinate, which I cannot accept." His brisk discussions with Schreiber in Ottawa yielded a decidedly one-sided compromise, much to the relief of Cooper's worried colleagues in Quebec and Phoenixville. It was agreed that plans and specifications would pass from Cooper to Schreiber for final approval; as it would turn out in practice, Schreiber's initials could just as well have come from a rubber stamp.
"I think," Cooper wrote to Hoare upon his return from Ottawa, "this will allow us to go on and get the best bridge we can, without putting metal where it will do more harm than good." By now, whether he wanted it or not initially, Cooper had attained virtually absolute authority over the engineering of the Quebec Bridge. He would say later that the burden had been imposed upon him by the circumstances, and that it was an onerous one. But in 1903 he had journeyed in haste to Ottawa to block Schreiber's attempt to have drawings independently reviewed' in 1904 he had quickly acceded to the protestations of Deans and Parent that he not resign; and in 1905 he insisted that a young, recently graduated engineer be installed at Quebec to serve, in effect, as his eyes and ears on the bridge.
The young engineer's name was Norman McLure, and though nominally he answered to both Cooper and Hoare, he was in fact Cooper's personal representative, communicating with Cooper frequently. McLure's intelligence, energy, and loyalty suited Cooper well. He was well trained and well recommended, and had enough technical competence to keep Cooper accurately informed and to execute Cooper's instructions, but not nearly enough experience to act without Cooper's authority.
The practical effect of all this, after the contract between Quebec and Phoenix was signed and erection of the superstructure got under way late in the summer of 1904, was to leave the day-to-day, hands-on building of the most technically ambitious bridge project in the world to a group of men utterly unprepared to grasp the scope of the work. No one at the site knew enough about what he was doing to act with authority. Everything of import was referred to Cooper.
The best opportunity for the critical computation of weights, during the waiting period from 1900 to 1903, had long since been missed, but early in 1905 the shop drawings of the south anchor arm were practically complete, and it would have been possible to recompute the weight of the arm to within a few percentage points of its actual weight. Neither the Phoenix Bridge Company nor Theodore Cooper bothered to do it--now for the second time.
On February 1, 1906, they began to pay the price. Cooper received a report from E.L. Edwards, the Phoenix Bridge Company's inspector of materials, revealing that the actual weight of steel put into the bridge had far exceeded the original estimated weight. (By June the projected weight for the complete structure would have to be raised from sixty-two to seventy- three million pounds.) Cooper concluded that the increase in the already high stresses, due to the error reported by Edwards, was between 7 and 10 percent.
By this time the south anchor arm, tower, and two panels of the south cantilever arm had been fabricated, and six panels of the anchor arm were already in place. Cooper decided that the increase in stresses was safe, and he permitted work to continue. The only alternative would have been to start building the bridge all over again.