Eads Bridge

Strength of the Bridge

Back    Next

The article "Strength of the Bridge," in Eads’ 1868 Report of the Engineer in Chief, reassures investors that...

On the following page we learn that the bridge is designed to support 7.2 tons per lineal foot including its own weight. This works out to 1.8 tons per foot of each of the four arch ribs, the load that Charles Pfeifer cites in his appendix to the Report.[2] Of this total, 0.8 tons per foot represents the “moving load”.

The preceding describes the schematic design. There is some ambiguity about the live load used for the final version. Woodward repeats the 0.8 tons per foot number[3] but he appears to be quoting the 1868 report not describing the as-built condition. Later writers, among them Carl Gayler, tell us that the bridge was designed for a live load of 1 ton per foot of rib or 8,000 lbs per foot of bridge.[4] [5] According to Gayler, 2,000 pounds of this are allocated to each track leaving 4,000 pounds per foot for the highway deck (74 pounds per square foot on the 54 foot 2 inch wide deck).

Despite Eads’ hyperbole about wall-to-wall locomotives 2,000 pounds per foot of track was not especially heavy, even in 1868. The committee of engineers that met in St Louis in 1867 recommended 2,184 pounds per foot of track and an average of 90 pounds per square foot for roadway and sidewalks.[6] For his Ohio River bridge of 1865, Jacob Hayes Linville assumed a “rolling load” of of 3,000 pounds per foot of track.[7]

Railroad technology evolved rapidly in the last years of the 19th century and trains soon became heavier than Eads Bridge could handle. The framing for the railroad deck was especially problematic. With no provision for spreading out wheel loads, floor beams were subjected to extreme load cycles. This contributed to rapid deterioration and in 1888 it was necessary to completely rebuild the deck.[8] In 1902 further work was performed to reinforce the new floor beams.[9] These improvements increased the capacity of the railroad deck but nothing could be done to improve the strength of the arches themselves.

Operational Constraints

According to a report prepared in 1922,[10] the reinforced deck could accommodate Cooper E-36 loads. This was being generous. The same report notes that main-line locomotives were banned and only small switch engines allowed on the bridge, an admission that the bridge could not support the “locomotive surplus” specified by E-36. Because switch engines could only handle 950 tons on the ramp at the east end of the bridge, heavy trains had to be broken into smaller units that could be shuttled across.

Disregarding the locomotive surplus, E-36 loading is equivalent to about 0.9 tons per foot of arch rib. This would soak up almost all of Gayler’s one ton per foot design load leaving no capacity for traffic on the highway deck. This suggests that the bridge may have sometimes been overloaded.

1928 Design Review

Questions about the bridge’s capacity persisted and were addressed in 1928 in a thorough review of the bridge conducted by consulting engineers John N. Ostrom and J. M. Johnson.[11]

Ostrom and Johnson make no mention of E-36 loading. Instead, they report that the TRRA (Terminal Railroad Association) had an operational rule limiting the weight of trains to 4000 pounds per foot of track. This was enforced by inserting lighter cars between heavy cargoes to control the average weight of trains. They also report that trains were limited to 15 miles per hour and that only TRRA’s switch engines were permitted on the bridge. Even with these restrictions, 4000 lb / ft on each track, in addition to the weight of traffic on the upper deck, is clearly more than the bridge’s designers intended.

Load

To verify that the bridge was safe for the increased loads, stresses were recalculated based on a live load of 10,000 lbs per foot of bridge. This accommodates the TRRA’s 4,000 lb per foot limit for trains on the two tracks while reserving 2000 pounds per foot for the upper deck.

Allowable Stress

To verify the composition of the steel, samples of metal were collected from the bridge and subjected to chemical analysis. The analysis confirmed that the staves of the arch tubes are chrome steel of consistently high quality. Based on this information and reassured by the 60 years of experience with steel gained since the construction of the bridge, Ostrom and Johnson adapted an allowable compressive stress of 40 ksi in place of the 30 ksi assumed by Eads’ staff. Allowable stress in tension was raised from 20 to 30 ksi.

For the strength of the wrought iron links that brace the arch ribs, the study adapted an allowable stress of 22 ksi as suggested by the 1921 AREA manual. This is a considerable improvement on the 10 ksi assumed by Charles Pfeifer in his calculations.[12] [13]

Conclusions

Based on these revised criteria, the arch tubes were judged to be safe but parts of the diagonal bracing were marginal. The worst case was a wrought iron eyebar brace subject to 17.7 ksi, well in excess of Pfeifer’s 10 ksi[14] and approaching the 18 ksi that had been administered in testing when the eyebars were made. This stress was computed, however, without taking credit for the cross-section of the tee-irons riveted to the inside face of the eyebars. With the area of the tee bars added, the braces and the bridge were deemed safe.[15]

Interventions in the bridge after 1928 took care to avoid further increasing the loads it must carry.

1947 Deck Replacement

In the late 1940s an inspection of the highway deck revealed that its supports were badly corroded. “Portions of the floor beams were remnants of those originally placed in the bridge; considering their loss of sectional area, they could carry a 20-ton truck only out of habit.” The wind truss had suffered too and was so damaged that “its value was highly questionable.” The wooden deck itself, which had last been replaced in 1934, was due for renewal.[16]

Work to provide a permanent roadway was carried out between 1946 and 47. The wood deck was replaced with a concrete-filled steel grating and the wrought-iron deck beams with welded steel framing. To offset the weight of the grating, the wind truss, originally located below the deck, was removed. In its place the inherent rigidity of the grating provided lateral bracing for the bridge. At the same time, the camber of the upper deck was reduced by cutting off the top of the posts that support it. This lowered the deck by 39 inches at the center of the bridge. Ostensibly this was done to improve sight distances for motorists but it also saved weight. More weight was saved by using lightweight concrete, created by substituting shale clinker for the usual stone aggregate, to fill the cells of the grating.[17]

Removal of the wind truss and careful design of the new structure resulted in an overall reduction of the dead-load on the bridge. It also added 18 inches of clearance above the tracks. It was hoped that better clearance would mitigate the “blast effect” of the exhaust of steam locomotives, which tended to strip the paint from overhead structures.[18]

2003 Deck Replacement

The 1947 deck lasted more than 50 years. It was replaced in 2003 as part of the renewal of the bridge after it was transferred to Metrolink. In place of the concrete-filled steel grating the new deck used an innovative “Exodermic” composite steel and concrete assembly that weighed less and could span further. The improved span capability allowed the removal of every other floor beam, further reducing the weight of the bridge.

Like the grating that it replaced, the composite deck provided lateral bracing for the bridge. Prior to rehabilitation, wind loads were transferred from the grating to the piers and abutments through finger-plate expansion joints. The new design uses a tongue and socket device located below the deck, reminiscent of the detail at the ends of the original wind truss. Because the expansion joints are no longer required to transfer wind load, the finger plates could be replaced with neoprene expansion devices that better accommodate the movement of the arches and which prevent water from draining from the roadway onto the structure below.[19]

Present Condition

Eads Bridge is subject to HS-20 highway loading.[20] For a four-lane bridge HS-20 stipulates a "lane load" of 480 pounds per foot or 1,920 pounds per foot for all four lanes.[21] This is approximately the live load assumed by Ostrom and Johnson. The weight saved when the decks were replaced frees capacity to accommodate the concentrated loads specified by the HS-20 standard.[22]

The railroad deck is no longer exposed to the punishing loads that were imposed by steam locomotives. Metrolink's commuter cars weigh less than half as much as the switch engines described in the 1928 study and they spread their weight over much more track.[23] [24]

---

Footnotes

Back    Next