Eads Bridge

Aesthetics and Engineering

Back    Next

Eads' title, "Engineer in Chief", is misleading. His role would today be described as "project manager" or perhaps "project architect". He established the general configuration of his bridge, largely in response to aesthetic criteria, then relied on his professional staff to figure out how to make it work. This collaboration produced an unusually graceful and expressive structure; an example of what David Billington, in his book The Tower and the Bridge, terms “Structural Art”[1] but the aesthetic payoff had a price.[2]

Carl Gayler, the last surviving member of Eads' design staff, described this dynamic in his 1909 address to the Engineer's Club of St Louis.

Arches

The fundamental decision to use arches appears to have been based on received ideas about structural efficiency and on Ead's conviction that arches would look better than a truss. Eads’ preference was apparent, even before he took on the task of designing a bridge, in his choice of Telford’s Thames River arch as precedent for long spans and in the arch-friendly language that he inserted into the model legislation prepared by his committee of the St Louis Merchant’s Exchange (Chamber of Commerce). The committee finished its work in April of 1866. By March of the following year, even before he had assembled his engineering staff, Eads was already promoting a bridge with 500 foot steel arches and agitating for the formation of the St Louis Bridge Company.[4] In April he hired Henry Flad, followed soon by Charles Pfeifer, to develop the three-arch concept into the design presented in the 1868 Report of the Engineer in Chief.[5] From the start, Pfeifer's and Flad's assignment was to design an arch bridge, there is no record that other structural systems were considered.

The 1868 Report

Eads understood that investors might not be comfortable with his perfunctory decision to use arches. He addresses the topic twice in the 1868 Report; first in the chapter titled “Arch and Truss Bridges” and again in the “Convention of Engineers” section of the report. Both sections imply but stop short of explicitly confirming that, under the particular circumstances that prevail at St Louis, an arch bridge will actually be more economical.

Arch and Truss Bridges

“Arch and Truss Bridges” explains how arches and trusses work and how an arch bridge is potentially less costly than a truss.[6] As far as it goes, the discussion is compelling but an attentive reader will notice that Eads emphasizes the efficiency of arches under ideal conditions and downplays the negative effect of asymmetric and concentrated loads. The discussion is entirely qualitive, without the numerical data needed to decide this issue.

Convention of Engineers

The “Convention of Engineers” section of the report is intended to dispel doubt that might have been inspired by the proceedings of the “Board of Civil of Engineers” that met in St Louis in August of 1867, under the auspices of the rival Illinois and St Louis Bridge Company.

The Board had nothing to say about arches but it did advise against the 500 foot option allowed by the enabling legislation. Instead, it recommended a truss bridge with eight spans, the longest being 368 feet. In his rebuttal, Eads agrees that a 500 foot truss would be prohibitively expensive but he asserts that the Board might have reached a different conclusion if they had considered an arch bridge. He “proves" this by extrapolating the cost of a 520 foot truss from data for the Board’s proposed 368 foot span and comparing this with an arch estimated using the same unit prices. Eads finds that the truss would be significantly more expensive, but this is superstructure cost only. He admits that he does not have cost data for the truss’ piers and foundations.[7] Without this information, the comparison cannot be completed.

To compensate for the missing information about piers and foundations, Eads cites the published cost of the substructures of his proposed three arch bridge and of the Board of Engineer’s eight span truss bridge. The Board’s number is higher. This comparison serves to muddy the water but it is misleading because it ignores the radically different number of spans and piers in the two designs.

The Report achieved its goal of reassuring investors. In the months following its publication, the company sold three million dollars worth of stock subscriptions, forty percent of which were paid in full. With this money in hand the board of directors authorized the start of construction for the river piers and east abutment.[8]

Economics of Steel Arch Bridges

In his 1919 study, The Economics of Steel Arch Bridges,[9] John Alexander Waddell furnishes the quantified comparison of arches and trusses that was missing in Eads’ Report. To determine their relative efficiency, Waddell undertook the enormous task of calculating stresses for a range of hypothetical arches and trusses and then estimating the probable cost of each. He found that neither arches nor trusses has an across-the-board advantage. Arches typically require less metal than a corresponding simple span truss but this economy is offset by the more costly substructure often needed for arches and by their sometimes more-expensive details.[10] Factors that influence the performance of metal arches include the ratio of rise to span, the configuration of the terrain over which they were built, and the number of hinges.

Rise

Waddell finds that the optimal ratio of rise to span is somewhere between 0.20 and 0.38. Flatter arches generate larger forces and must be more heavily-built. Taller arches, where the ratio exceeds 0.38, become less cost effective because their increased length, measured along the curve, consumes unnecessary material.[11]

Hinges

For arches supporting railway loads and with a favorable ratio of rise to run, two- and three-hinge designs are “nearly always” lighter than fixed-end arches. For railway spans Waddell suggests that two-hinge designs are the best choice because they tend to deflect less than three hinge designs. He admits that “The variations in weight among the three types, however, is never great.”[12]

Terrain

Where the ends of a span can bear directly against the rocky wall of a ravine it is possible to dispense with the massive abutments usually required to confine the ends of an arch. In this situation an arch is often more economical than a truss. An arch bridge can be competitive in flat terrain, but only where foundation conditions are ideal. Arches should not be used where “foundations have to be built on piles or on any other material that is liable to slight settlement, or when the abutments could possibly move laterally even a mere trifle”[13] [14]

Economics of Eads’ Arches

Where aesthetic refinement was at odds with an optimal structural solution, Eads was often willing to compromise structural efficiency. Foremost among such compromises was the acceptance of an extremely low ratio of rise to span.

Rise

For his bridge, Eads favored a traditional configuration in which the deck extends across the top of the arches. The height of the deck is set by the elevation of the tunnel at the west end of the bridge, limiting the rise of the arches to just ten percent of their span. More-efficient arches would have been possible if they were allowed to extend above the highway as was done at the Koblenz bridge, but this would have compromised Eads’ preferred profile.

After publication of the 1868 Report Eads revised the design to eliminate the suspended railroad deck. This adjustment improved the appearance of the bridge but further reduced the rise of the arches, from ten to nine percent of span.[15] [16] With less than half the rise of an optimal arch, Eads’ arches generate approximately twice the thrust. This is reflected in the size and cost of the arch chords and the masonry supports.

Hinges

The stronger forces produced by flat arches generate large deflections. We don’t know how much deflection the engineers were willing to accept but apparently they weren't satisfied with the results of their initial investigation. Henry Flad suggested immobilizing the ends of the arches in order to stiffen them.[17] To test this suggestion, Charles Pfeifer calculated stresses and deflections for both hinged and fixed-end arches. He found that, even after taking account of the effects of thermal expansion, with fixed ends “the sectional areas of the ribs for the greater part of the span could be reduced, if for a short distance near each pier they were increased, thus effecting an important saving of material.”[18]

After publication of the 1867 report, further revisions were made to the design. These included changing the depth of the arch from eight to twelve feet between the chords. According to Eads, twelve feet achieved the best trade-off between deflection control and thermal stress.

At first glance the decision to use fixed ends appears to be at odds with Waddell’s conclusion that hinged arches are usually lighter. The discrepancy can be attributed to the tendency of flat arches to deflect more then the taller arches favored by Waddell and therefore to benefit more from the stiffening effect of fixed ends. There is no reason to assume that Pfeifer got his sums wrong.

Terrain and Substructure

According to Waddell an arch bridge can be economical in level terrain, but only if the foundations do not extend far below the river-bed.[20] At St Louis, they extend very far below.

In the 1868 Report Eads asserts that, in order to resist the floods and ice-flows of the Mississippi, any type of bridge at St Louis, not just arch bridges, will require unusually massive piers and abutments. This requirement for oversized piers negates the advantage of a truss.[21]

Eads makes the further , somewhat contradictory, argument that because of their small size the piers of a truss bridge require more-expensive construction then is needed for the supports of an arch bridge. While the slender piers of a truss bridge must be meticulously-crafted ashlar all the way through, the generous dimensions of Eads’ piers and abutments enables them to be constructed as a shell of dressed stone enclosing a rubble core. Because the rubble masonry might cost 30 percent less per ton, the bulky substructures of Eads Bridge exact less of a cost penalty than their size might suggest.[22]

In the event, the need to oversize the piers of truss bridges was not as dramatic as Eads predicted. The next two spans constructed at St Louis after Eads Bridge, the Merchant’s and MacArthur (formerly Municipal) bridges, were both trusses. These bridges have similar spans and support heavier live loads than Eads Bridge. Their masonry piers are exposed to the same environmental threats that Eads ascribed to the river at St Louis and yet their designers were able to make them much smaller than the piers at Eads Bridge.[23] [24] Even if Eads was correct about the low cost of his masonry, his piers and abutments would have been at least as expensive because they contain 30 percent more masonry.[25] They also incur caisson and excavation costs in proportion to their greater size.

Arches vs trusses

The structural efficiency sometimes attained by arches is compromised in the case of Eads Bridge by the heavy construction required due to their low rise. Any remainig savings that might have been obtained by using arches is offset by the cost of the enormous piers and abutments required to sustain the arches. Eads' prediction, that a truss bridge would be burdened with similar massive substructures, failed to materialize.

Detailing

The type of structural system is not the only factor influencing the cost of the bridge. Eads’ idiosyncratic detailing was another.

John Kowenhoven quotes Carl Gayler about the design of the arch chords. According to Gayler, Charles Pfeifer “had made a sketch of what he considered a proper cross-section for the chords of the arches, all in the European style: plates and angles riveted together.”

Eads promptly vetoed it. “Tubes, safely enveloped, had been one of his earliest conceptions... Eads just loved this part of the work and all the minutest details of the tubes with their couplings and pin connections, the skewbacks and anchor bolts, all of it to the last 1/8 of an inch are the work of Eads, of course always subject to Pfeifer's established effective areas."[26]

Eads domesticated Pfeifer’s characteristically European design by interpreting it as a pin-connected assemblage of tubular struts and eye-bars, not unlike the patent trusses usually provided by American builders. In theory, the circular cross section of his chords provides the best resistance to buckling. It also creates a dramatic distinction between the primary members of the arches and other components, a contrast which may have appealed to Eads’ aesthetic sensibility.

Although attractive, Eads’ design required complicated shaping and joinery, compounding the difficulty of working with the unfamiliar material, steel. The curved staves for the chords proved difficult to roll and the forgings for the couplings were nearly impossible to make.[27] [28] Months were lost before the metallurgic problems were finally resolved and then, when usable blanks began to arrive at Keystone’s shop for finishing, progress was further impeded by the extensive machining and close tolerances required for the couplings. It took two and one-half work-days to turn and grove the ends of a single tube, a similar amount of time to plane the mating surfaces, bore out the interior, and cut matching groves in each coupling, and yet more time to complete the pins and prepare the holes to receive them. With 1,036 tubes and 1,012 couplings to produce, the bridge company was forced to pay for night work and to finance the acquisition of additional machine tools for Keystone. Even with these added resources, a steady flow of finished tubes and couplings did not begin to arrive at the bridge until June of 1873, two years after Butcher’s first attempts at forging and a year after the first blanks were delivered to Keystone.[29]

Charles Pfeifer’s riveted box-sections might have required slightly more metal and might have been less dramatic than Eads' tubes and eye-bars, but they would have eliminated the curved staves, the forgings, and most of the machine-shop work.

Cost

When discussing construction costs it is customary to distinguish between “hard costs”- the cost of design and construction, and “soft costs” which include real-estate and the cost of financing. The 1868 Report estimated that the bridge would be completed in 1872 for a “hard” cost, including approaches but excluding the tunnel, of about $4,100,000.[30] By the time the project was completed, two years behind schedule in 1874, the cost had grown by sixty-three percent, to about $6,700,000.[31] Part of the increase was due to scope changes such as increasing the width of the highway deck and extending the east abutment to bedrock. A significant share can be attributed to the design decisions discussed above.

The cost overrun was certainly unwelcome but it was the delay that proved fatal to the bridge company. While construction dragged on at St Louis, the national economy slipped into recession, sapping the railroad’s enthusiasm for pioneering new routes through St Louis. At the same time, new crossings at Keokuk, Iowa and Hannibal, Missouri and the introduction of ferryboats with roll-on-roll-off capacity for rail cars compromised the bridge’s monopoly on trans-Mississippi traffic. J. Edgar Thompson’s death, in May of 1874, deprived the bridge of its strongest advocate on the board of the Pennsylvania Railroad. Absent the momentum that might have prevailed a few years earlier, railroads terminating in East St Louis, many of them with interests in the ferry and transfer companies, were able to enforce a boycott of the bridge.[32]

The bridge company was organized on the principal that, even if profits were small at first, income from bridge tolls would at least cover the interest on the company’s bonds. The boycott prevented this. The size of the company's debt ensured that it would burn through its reserves quickly, hastening its collapse.

Aspiration

The priority assigned to aesthetics is re-litigated whenever a major infrastructure project is planned. Does fiscal propriety demand the cheapest and most expedient solution or is something more required of a highly visible structure that will endure for decades? Eads felt that more is required. The bridge would be a monument to the vision and capacity of the city of St Louis...

When all of the many difficulties that have retarded this great work shall have at last been surmounted, and the Bridge becomes an accomplished fact, it will be found unequaled in the important qualities of strength, durability, capacity and magnitude, by any similar structure in the world. Its great usefulness, undoubted safety and beautiful proportions will constitute it a national pride, entitling those through whose individual wealth it has been created to the respect of their fellow men; while its imperishable construction will convey to future ages a noble record of the enterprise and intelligence which mark the present times.[33]

Unfortunately for investors, Eads neglected to include this manifesto in the 1868 Report of the Engineer in Chief where it might have informed their decision to buy shares. Instead it was published in the 1871 Report, as consolation for the growing cost and delayed completion of the bridge.

---

Footnotes

Back    Next