Eads Bridge

The Floating Cofferdam

Back   Next

This page discusses the original scheme for constructing the piers for Eads Bridge. For an account of the pneumatic caissons which were finally used, see the Foundations and Caissons articles on the "Design and Construction" page.

Before his 1868 trip to Europe, Eads planned to install the piers for the St. Louis Bridge using an open caisson and a device he referred to as a "floating coffer dam" (cofferdam). After his return from Europe, he abandoned the cofferdam in favor of pneumatic caissons. Because the floating cofferdam didn't make it into the final design, it receives cursory treatment in most histories. This obscures the importance of the cofferdam in the evolution of Eads' thinking about the bridge.

The Cofferdam

The floating cofferdam was not a new idea. It was developed in France and England during the 18th century and had been used on a number of bridges including the original Westminster and Blackfriars bridges in London, both completed prior to 1800.

The cofferdam was a floating watertight box, similar to a drydock, in which the masonry pier of a bridge could be constructed. The cofferdam's dimensions were calibrated to provide enough buoyancy to support the partially-constructed pier, but to sink after the admission of just a few feet of water (still leaving the top of the pier dry and accessible to the masons). By this means, the cofferdam and its cargo of masonry could be lowered into a pit in the riverbed that had been prepared to receive it.

The cofferdam's walls created a "hole in the water" in which the masons would continue working until the pier had been built-up above river level. When the pier was done, the cofferdam's sea-cocks were opened, flooding it so that its walls could be detached and salvaged for re-use. The cofferdam's floor was left behind, trapped below the masonry as a permanent component of the pier's foundation.[1] [2]

The Open Caisson

When using floating cofferdams, excavation for the foundation could be problematic. Because the work was concealed under water, it was difficult to verify the nature of the material on which the pier would be founded. It could also be difficult to produce a uniform surface by means of dredging and difficult to prevent sediment from washing into the pit, ruining the carefully-graded platform before the floating cofferdam could be lowered onto it.[11]

Because Eads proposed to excavate to bedrock, the first issue ‒ the quality of riverbed material ‒ did not come into play.

The second problem ‒ slumping or washing of material into the pit ‒ was prevented by enclosing the site within a caisson, in the form of a huge iron cylinder, which would retain the sides of the hole. No attempt would be made to exclude water. Instead, the caisson would be allowed to flood to the level of the river, creating a protected pool through which steam-powered machinery could reach to dig into the riverbed. As material was removed from inside it, the caisson would settle under its own weight until it reached solid rock.

After bedrock was exposed inside the caisson, the surface would be leveled with a layer of concrete.

After the excavation and concrete base were complete, a floating cofferdam would be assembled inside the caisson and used, as described above, to lower the pier, through the caisson, onto the concrete pad. When the pier was complete, the caisson and the walls of the cofferdam would be extracted for re-use. [3]

Influence of Europe Trip

It is tempting to conclude that Eads first learned about pneumatic caissons while traveling in Europe but this is not the case. Eads was already well-informed, as is clear from the detailed description of pneumatic caissons in the 1868 Report of the Engineer in Chief, published prior to his trip. The report describes the operation of pneumatic caissons and outlines reservations about them that prompted Eads to favor the floating cofferdam.[5]

The effect of Eads' travel, his conversations with French and English engineers, and his inspection of European bridges was not to introduce him to the new technique but rather to reassure him that pneumatic caissons, of the unprecedented size required at St. Louis, would be feasible.

Transition to Pneumatic Caissons

In the September 1869 Report of the Chief Engineer, Eads described his reassessment of pneumatic caissons.

...during my visit to Europe last winter I had an opportunity of witnessing the operation of sinking masonry piers by the plenum pneumatic process, as practiced by the engineers on the continent of Europe, and I there became satisfied that with the improvements made in it by them, many of which are unpublished, I could sink the piers of your Bridge quite as safely, as expeditiously, and more economically, than by the method I had intended to use. In France and England I was so fortunate as to be permitted to exchange views upon this subject with several of the most distinguished engineers in the profession, and my opinions upon this, the most important question in the construction of your Bridge, were fully confirmed as a result of those interviews. In adopting this method of sinking the piers it was necessary to mature plans and details of an entirely new type for the necessary caissons and floating appliances, and to design and construct machinery, purchases, etc., quite different also from what had been intended.[6]

Different, but not unrelated. Eads described the pneumatic caisson as a derivative of the floating cofferdam created by extending the sidewalls of the cofferdam down to create a "diving-bell" beneath the masonry.[5]

Reservations about Pneumatic Caissons

The 1868 report identifies several issues that prompted Eads to favor the floating cofferdam. He worried about the vulnerability of pneumatic caissons to floods and ice-flows. He also worried about the difficulty of controlling the rate of descent of the caisson. Finally, he noted that pneumatic caissons were relatively untried technology that might incur unanticipated costs.

Floods and Ice

The first reservation ‒ about exposure to floods and ice ‒ was corollary to Eads' assumption that work would proceed more rapidly using the floating cofferdam. Because the open caisson would be accessible for powered excavation machinery, the pit could be prepared and the pier completed in a single summer and fall, avoiding winter ice and spring floods. Work inside the air chamber of a pneumatic caisson would be mostly hand-labor and would take longer.

The speed advantage of the floating cofferdam scheme was never tested. By the time construction started, the project schedule had been revised to require both piers to be installed in a single year-round campaign. Instead of scheduling the work to avoid dangerous seasons, the hazards of flood-born debris and winter ice were addressed head-on, by constructing massive timber and riprap "ice breakers" upstream of the piers.[7]

The speed of construction using the floating cofferdam was probably overstated. It may indeed have been possible to dig to bedrock faster, but the floating cofferdam couldn't be assembled inside the caisson until after excavation was complete and then the masonry would take just as long to finish in the floating cofferdam as in the upper compartment of a pneumatic caisson. With a pneumatic caisson, excavation proceeds at the same time as the masonry. Even though hand digging would be slow, the masonry not the excavation would determine the time needed to complete the pier.

Controlling the Descent

The second reservation in the 1868 report ‒ about possible difficulty controlling the descent of pneumatic caissons ‒ was a reaction to published accounts of European caisson work, at Kehl and at Koningsberg, Prussia.[8] The weight of the Prussian caissons and their payload of masonry was suspended, during their entire descent, from jack-screws attached to cribbing erected around the caisson. The caissons had an unpleasant tendency to hang up due to friction with the surrounding sand, even after the screws had been slacked off, and then to drop abruptly, testing the capacity of the screws and cribbing and endangering workers in the air-chamber below.[12]

Originally cited as an argument in favor of the floating cofferdam, the problem of support and control had to be addressed when Eads decided to use pneumatic caissons. Eads avoided trouble with jack-screws by adapting a different method. While the St. Louis caissons sank through the water they were suspended from screws, but after they landed on the riverbed their weight was transferred to sill-plates bearing directly against the sand. After the weight was transferred to the sills, the jack-screws served no purpose and were removed.

Possibly because their great size altered the ratio of surface to mass in their favor, the St. Louis caissons did not experience the problem with friction that had plagued the Prussian examples. They settled smoothly. The descent was managed from inside the air chamber by digging trenches adjacent to the sill plates. The pressure of the sill caused the sand to flow into these trenches, gently lowering the caisson. Progress could be accelerated by slowly releasing a few pounds of air pressure. This caused water to seep under the toe of the caisson's wall, liquefying the sand beneath the sill and causing it to flow into the trench more quickly.[9]

Cost

The third reservation cited in the 1868 report ‒ possible unforeseen costs of untried technology ‒ was allayed by Eads' trip to Europe. After seeing them in action, Eads was confident that pneumatic caissons were no longer "untried technology" and that any problems would be surmountable. Woodward reports that Eads was more enthusiastic than his staff, no doubt because he saw the similarity between caissons and the diving bells and other underwater apparatus with which he had long been familiar through his work in the salvage industry.

The decisive factor in the decision to abandon the floating cofferdam was a cost of a different sort. In the spring of 1869, in an effort to reduce the burden of interest on the company's bonds, the bridge company's board ordered that the construction schedule be accelerated. Under the new schedule both piers would be constructed at the same time, requiring duplication of equipment. The combination of floating cofferdam plus open caisson was a much larger and more costly structure than a pneumatic caisson and was justified only if it could be reused. If both piers were to be constructed at the same time, pneumatic caissons were the lower-cost approach.[10]

---

Footnotes
Back   Next