Quebec Bridge 1907 — A Compression Chord Buckled Under an Unrevised Dead Load

Summary

On 29 August 1907, near quitting time, the south anchor arm and partly built cantilever of the Quebec Bridge over the St. Lawrence River buckled and fell into the river in about fifteen seconds, killing 75 of the 86 men working on the steel that evening; only 11 survived, and roughly 33 of the dead were Mohawk steelworkers from the Kahnawake reserve near Montreal. The Royal Commission that investigated the disaster found the cause without hedging: the bottom compression chords near the main pier failed by buckling, because their latticing was too weak to make the built-up members act as a unit, and because the dead weight of the structure had been assumed too low at the start and never revised when the design grew.

The mechanism was a compression-member failure, not a fracture or a foundation movement. The lower chords of a cantilever truss carry enormous compression, and the Quebec chords were huge built-up sections — clusters of ribs laced together with riveted lattice bars rather than solid plate. The lattice was the weak link. It could not force the separate ribs to buckle together as a single column, so the ribs deflected, the latticing yielded, and chord A9L on the anchor arm folded, immediately followed by A9R. With the compression chords gone, the cantilever had no load path to the pier and the whole south arm came down.

Underneath the buckling sat a numerical error of governance. When the consulting engineer, Theodore Cooper, lengthened the main span from 1,600 to 1,800 feet to set a world record and cut construction over the deep channel, the dead-load assumptions made for the shorter, lighter bridge were carried forward almost unchanged. By the time the discrepancy was noticed, the actual weight ran well over the figures the members had been proportioned for, and Cooper had also set allowable working stresses far above contemporary practice — up to 24,000 pounds per square inch where 16,000 was customary. The chords were overloaded on paper before a single rivet was driven.

What makes the Quebec Bridge a permanent teaching case is that the structure announced its own failure for weeks and the warning was overruled. Chords already in place were measurably bending out of line that August. The inspecting engineer reported it, work was questioned, a stop order was contemplated — and the load kept going on while the responsible engineers debated whether the deflection was old or new. The chord that everyone was watching was the chord that buckled.

Timeline

1887-1900

A record-span crossing is conceived

The Quebec Bridge Company is chartered to bridge the St. Lawrence near Quebec City. To clear the deep navigable channel without a pier in midstream, the design adopts a cantilever truss — eventually the longest cantilever span in the world.

1900

Theodore Cooper appointed consulting engineer

The eminent American bridge engineer Theodore Cooper is engaged to oversee design and construction. The Phoenix Bridge Company of Pennsylvania is contracted to design and fabricate the steel; Peter L. Szlapka serves as its chief design engineer.

1900-1903

The span is enlarged and stresses raised

Cooper increases the central span from 1,600 to 1,800 feet to reduce the cost and time of working in deep water, and specifies allowable unit stresses up to 21,000-24,000 psi — well above the roughly 16,000 psi of customary practice.

1903-1904

The fatal omission

The dead-load estimate made for the earlier, shorter design is not properly revised for the heavier 1,800-foot bridge. When the increase is finally recognized, the recalculated stresses are higher, but fabrication of the under-proportioned compression chords is already advanced.

1905-1907

Erection of the south anchor arm and cantilever

The south approach, anchor arm, and cantilever arm are erected outward from the main pier on the Quebec side. The bottom chords — large built-up compression members laced together with lattice bars — carry steadily increasing compression as the cantilever grows.

early Aug 1907

Chords found out of line

Workers and the inspecting engineer Norman McLure observe that compression chords already erected have bent out of straight. The deflections in chords near the pier increase over the following weeks.

mid-Aug 1907

Warnings travel up the chain

McLure reports the growing distortions to Cooper in New York. There is dispute over whether the bend is an old erection defect or a live, worsening overload. Work continues while the question is argued.

27-29 Aug 1907

The too-late telegram

Cooper, alarmed, wires that no more load should be added to the bridge until the facts are considered. The instruction does not halt work on the structure in time; the message and the men cross paths.

29 Aug 1907, ~5:30 p.m.

The compression chords buckle

Near the end of the workday, lower chord A9L on the anchor arm buckles, instantly followed by A9R. The compression load path to the pier is lost and the south anchor arm and cantilever collapse into the St. Lawrence in roughly fifteen seconds.

29 Aug 1907, evening

75 dead

Of 86 men on the steel, 75 are killed and 11 survive. About 33 of the dead are Mohawk ironworkers from Kahnawake. It is among the worst bridge-construction disasters in history.

30 Aug 1907

Royal Commission appointed

A Royal Commission of inquiry is established within a day, chaired by Henry Holgate, to determine the cause.

1908

The verdict

The Commission reports that the chords failed by buckling, that the latticing was inadequate, that the dead load was assumed too low and never revised, and that the unit stresses were too high — attributing the failure principally to errors of judgment by Cooper and Szlapka.

The Record Span and the Laced Chords

The Quebec Bridge was built to break a record. A cantilever truss reaches across a gap from two arms balanced on piers at either side, the arms growing outward until they meet in the middle. The deep, fast-flowing St. Lawrence ruled out a pier in the navigation channel, so the design committed to the longest cantilever clear span in the world — 1,800 feet between piers after Cooper enlarged it from the original 1,600. Lengthening a cantilever does not add load gently. It increases the bending and the compression in the bottom chords near the piers disproportionately, because those chords carry the entire reaching arm. The enlargement made the most heavily loaded members much more heavily loaded, and that is exactly where the failure occurred.

The bottom chords were the structure's compression backbone, and they were enormous: built-up members assembled from multiple steel ribs set side by side and tied together not with continuous plate but with riveted lattice bars — diagonal lacing across the open faces. Lacing is lighter and cheaper than plate, and on a column it has one job: to force the separate ribs to act as one stiff member so the whole section buckles as a unit rather than rib by rib. The Quebec lacing could not do that job at the loads imposed. The Commission and the contemporary analysis by C. C. Schneider found the latticing insufficient to make the parts act together, so the ribs were free to deflect individually. A built-up compression member is only as strong as the lacing that unites it; here the lacing, not the steel area, set the capacity, and the lacing had been treated as a detail rather than as the governing element.

How an Unrevised Dead Load Overloaded the Chords

The buckling was the visible event; the overload was the cause. Compression members are proportioned against the load they must carry, and the load that governs a long cantilever's chords is overwhelmingly the dead weight of the bridge itself. That weight was assumed early, for the smaller scheme, and the assumption was not properly carried forward when Cooper extended the span and the design grew. By the time anyone reconciled the figures, the real structure weighed materially more than the chords had been designed for — the recomputed stresses were higher, and the weight ran well above the original allowance, but the chords had already been fabricated to the optimistic numbers.

Cooper compounded the deficit at the other end of the calculation. He had specified allowable working stresses far above the practice of the day — figures reaching 24,000 psi against a customary ceiling near 16,000 — partly to keep the record-span steel from becoming impossibly heavy. High allowable stress and an underestimated dead load push the same way: both shrink the margin between the working condition and the buckling condition. The chords ran with little reserve from the day they were erected, and every panel added to the growing cantilever drove their compression higher. The bending observed in August was not a cosmetic flaw. It was the member telling its inspectors that it had reached the load it could not carry. The lattice yielded, the ribs deflected past recovery, and A9L folded. There was no redundancy: when a primary compression chord buckles, the cantilever has no other path to the pier, and total collapse follows in seconds.

The Reckoning: Errors of Judgment, Codified

The Royal Commission, reporting in 1908, performed the autopsy with unusual candour for its era. It located the initial failure in the lower chords of the anchor arm near the main pier; it identified the latticing of those chords as too weak to make the members act as a unit; it found that the dead load had been assumed too low and not afterward revised, and that the specified unit stresses were higher than sound practice supported. It assigned the responsibility plainly: the failure could not be attributed to any cause other than errors of judgment on the part of the two engineers responsible for the design, Theodore Cooper and Peter Szlapka. No defective steel, no freak load, no foundation movement, no act of God appeared in the finding.

The Commission also named a failure of oversight. The most heavily loaded members in a record-setting structure had been designed to allowable stresses no one had tested at that scale, on a dead-load figure no one had rechecked, under a consulting engineer working remotely from New York who had not insisted on independent verification of the controlling assumptions. When the chords began to bend, the apparatus for stopping work was too slow and too uncertain to act before the load that was already on the bridge finished the job. Cooper was not prosecuted; the failure was professional, not criminal — which is precisely why it entered the permanent curriculum of structural engineering as the textbook case of an under-designed compression member and an unmanaged change of scope.

Contributing Factors

01

The dead load was assumed early and never revised when the design grew

The governing load on a long cantilever's chords is the bridge's own weight, and that weight was estimated for a smaller scheme and carried forward when the span was enlarged. The members were proportioned to a figure that no longer described the structure. A change in span, depth, or member size invalidates every load assumption downstream of it; the controlling loads must be recomputed in full, not scaled by memory, whenever the design changes.

02

The latticing, not the steel area, governed the compression capacity

The bottom chords were built-up members whose strength depended entirely on lacing forcing the ribs to buckle as one. The lacing was insufficient, so the section buckled rib by rib far below the strength implied by its gross area. In a built-up compression member, the connecting lattice or batten is a primary structural element, not a detail; its adequacy must be checked against the same buckling that governs the member, because it sets the member's true capacity.

03

Allowable unit stresses were pushed above proven practice

To keep record-span steel from becoming unbuildably heavy, the allowable working stresses were raised well above the customary ceiling. High allowable stress shrinks the margin to buckling at the same time the underestimated weight was eating into it. Unprecedented spans demand more conservative allowables, not less; the further a structure departs from validated practice, the larger the reserve it requires against the assumptions that precedent has not yet tested.

04

A primary compression chord had no redundancy, so its buckling ended the structure

A cantilever arm reaches the pier through its bottom chords and nothing else. When A9L buckled, there was no alternate path and the collapse propagated to total loss in about fifteen seconds. A member whose failure brings down the structure must be designed and reviewed as a fracture-critical, non-redundant element, with margins and independent checks proportioned to the fact that it has no backup.

05

Visible distortion was debated rather than treated as a stop condition

The chords bent measurably out of line for weeks, were reported up the chain, and were argued over — old defect or live overload — while erection continued. A compression member deflecting out of straight under increasing load is reporting incipient buckling, and ambiguity about its cause is itself the signal to stop. Observed deformation in a critical member is a halt-and-verify condition by default; the burden is on proving it safe before loading resumes, not on proving it dangerous before stopping.

Aftermath

The collapse killed 75 men, devastated the Kahnawake Mohawk community that supplied much of the ironworking crew, and erased the first Quebec Bridge into the river. The Royal Commission's 1908 report became a landmark document in the history of structural engineering: a clear, public attribution of a major disaster to under-design — an underestimated and unrevised dead load, latticing too weak for the compression it carried, and allowable stresses beyond sound practice — rather than to bad luck or bad steel. The bridge was redesigned from scratch with far heavier members and rebuilt; a second tragedy struck in 1916 when the central span fell while being hoisted into place, killing 13 more, before the structure was finally completed in 1917. The 1907 disaster hardened the profession's insistence on independent checking of governing assumptions, conservative allowables for unprecedented spans, and proper design of built-up compression members and their lacing. In Canada it became bound up with the founding ethos of professional engineering and the Ritual of the Calling of an Engineer, and it stands to this day as the byword for a compression chord overloaded by an arithmetic no one went back to redo.

Lessons

Recompute every governing load in full whenever the span, depth, or member sizing changes; never carry forward a dead-load estimate made for a smaller design, because the loads that buckle a chord scale faster than intuition.
Treat the lacing or batten of a built-up compression member as a primary structural element and check it against buckling, because the connection that unites the ribs, not the steel area, sets the member's real capacity.
Carry larger reserves, not smaller, the further a structure departs from validated practice; raise the conservatism of your allowables for record spans rather than raising the allowables to make the record buildable.
Design and review any non-redundant primary compression member as fracture-critical, because its buckling has no alternate load path and ends the structure in seconds.
Stop work and verify whenever a critical member visibly deforms under increasing load; deflection out of straight is incipient buckling, and the burden is on proving it safe before reloading, never on proving it dangerous before halting.