When the Forum’s Akielly Hu profiled civil engineer Stephen Ressler last year, the professor’s newest online course on the Wondrium site was still a work in progress. Now available, Epic Engineering Failures—and the Lessons They Teach has several episodes with important lessons for the environmental profession in particular. These include examinations of not just the structural failures but the human and institutional failures that led to the Chernobyl meltdown in 1986, the inundation of New Orleans in 2005, and the BP oil rig blowout in the Gulf of Mexico in 2010.

I would add to those episodes two segments on railroad bridge collapses in Great Britain during some of the formative years of the industrial revolution, one over the River Dee in western England in 1847, the other over the wide Firth of Tay in eastern Scotland in 1879. Both bridges were record-breaking crossings at the time, and both disasters implicated problems with cast iron as a structural material. The events set off public clamors plus investigations and trials that led to a winding down of the metal’s use in bridge construction, as engineers shifted to newly commercialized materials that were better at responding to loads—and less troublesome in use.

The lessons these two Victorian-era disasters teach include the nascence of features in the regulation of private business that we take for granted today. Among the responses to problems in the early growth of the British railroad industry was the deep involvement of Parliament. It passed a series of rule-setting statutes during the period of expansion. In addition, major infrastructure projects required its approval. Parliament also created specialized governmental agencies to conduct industry permitting and inspections. Agency decisionmakers in turn sought input from the legal, scientific, and engineering professions. Planners took comments from stakeholders. And of special note was the role of courts in fact-finding and determining fault when a structural collapse caused damage, injury, or death.

“The earliest regulations governing the adjudication of structural failures,” according to Matthys Levy and Mario Salvadori in Why Buildings Fall Down, "are known as the Code of Hammurabi.” That is no surprise and neither are the penalties. Their book quotes from the ancient text: “If a builder build a house for a man and do not make its construction firm and the house which he has built collapse and cause the death of the owner of the house, that builder shall be put to death.”

That is not all the code has to say on the subject. It goes on to require mitigation of property harmed during construction, and damage payments “at his own expense” when a structure is deficient because the builder did “not make its construction meet the requirements.” The requirements were customs at that time and for most of the centuries since. But they became substantial by the time of Great Britain’s big railway boom and were being adopted as engineering best practices, put into building codes, or even written into laws.

Railroads were one of the key new business sectors that emerged during the industrial revolution, often simultaneously and interdependently. Thus, British coal firms began to use the newly invented steam engine, which of course ran on coal, to dewater mines. Mining businesses were simultaneously invigorated by surging demand for coke, a newly developed fuel derived from coal that burns even hotter and made possible industrial quantities of iron for the first time. And to complete the circle, steam engines were made of iron and were ultimately used to pull iron-wheeled cars on iron rails that carried coal to commercial and household customers—and of course also eventually hauled freight and passengers.

The first intercity railroad, connecting Liverpool and Manchester, opened in 1830. What followed has been termed “Railway Mania.” Within a decade and a half, almost every town in Britain was served by at least one line. Addressing issues created by the industry, Parliament passed the first law regulating these new carriers, the Railway Act of 1840, which placed the national Board of Trade—the equivalent of our Department of Commerce—in the decisionmaking process for rail projects, and created Her Majesty’s Railway Inspectorate and a position known as the Commissioner of Wrecks—the last of which Ressler tags as the forerunner to the U.S. National Transportation Safety Board.

More government action followed. “Modern industrial regulation can be traced to the Railway Regulation Act 1844 in the United Kingdom, and succeeding acts,” according to Wikipedia. This new law set passenger fares to make them affordable, ensured that there would be seats for the lower classes on some trains so they could get to their places of employment—this of course purposely pleased business interests—and established a service standard of at least 12 miles per hour.

The most important legislation may have been the Railway Regulation Act of 1846, which created a national standard for gauge—the distance between the inner edges of the two rails. Without that critical involvement of government in industrial development, an efficient, coherent, and fully national system would likely never have emerged, given competitive interests. But lines instead spread from numerous urban centers around the isle and eventually knit together into the comprehensive network that still moves goods and passengers today.

On May 24, 1847, a passenger train from Chester, England, bound for Wales crossed the River Dee on a relatively new bridge supported on cast iron at a span length never tried before. It would be a fateful day. Under the 1840 legislation the Board of Trade was tasked with evaluating bridge plans. The Chester and Hollyhead Railway had proposed a typical masonry bridge with five arched spans across the 250-foot-wide Dee—a bridge that could have been built by Roman engineers—but navigation interests objected and forced the railroad to come up with an alternative. The rail company itself was concerned that a masonry bridge would be incompatible with the river bottom at that location, where the soil of the stream bed might not support the weight of all that brick and concrete.

In effect, stakeholders objected to sharing the public channel with the new private infrastructure project, and an environmental assessment militated against the masonry bridge.

At that juncture, the railroad turned to Robert Stephenson, who along with his father, George, had a sterling reputation for designing some of the first practical steam locomotives and, importantly, the railroad bridges that made motorized transport around an island nation practical. A more trusted engineer did not exist in Victorian Great Britain.

The younger Stephenson came up with a bridge design to span the Dee that satisfied the shippers and would be able to work within the environmental constraints. His plan called for only two intermediate piers, supporting metal girders—long beams designed to carry loads. These were light enough not to cause problems with the stream bed geology, and unlike arches they provided vertical clearance for the full width of the channel, pleasing the maritime interests.

The girders had to span 98 feet between abutments on the shore and the two intermediate piers, a like distance between the piers. Stephenson decided to make the girders of cast iron, since in the 1840s wrought iron—which is stronger—would be impractical and expensive at the required size, and industrial steel—even stronger—was still decades away. Even so, no mill could make cast-iron beams long enough to span the distance. Stephenson ingeniously spliced together three cast girders end to end, long enough to reach across the gaps.

Mathematical tools to design structures were in their infancy at the time, as was the science of materials. But engineers did know from empirical testing that cast iron has an important limitation. When a girder or any beam is subject to a load, it bends. If a loaded girder is supported at each end, it sags in the middle. In the process, the top is subject to compression and shortens, and the bottom is subject to tension and elongates. Stephenson decided to use girders with a cross-section like the letter I—what today we call an I-beam. It is both more efficient in resource use and stronger than a rectangular cross-section. Much of the metal is concentrated in the top flange and bottom flange—the horizontal serifs of the I. The flanges are connected by the vertical stroke of the letter, thin material called the web. The flanges do most of the work to support the load by resisting internally the compressive and tensile forces created when such a beam bends. But cast iron’s limitation is that it is only a fifth as strong in tension as in compression. So Stephenson compensated by making the bottom flange, where the tension occurs, with five times more metal than in the top.

Alarmed by the fact that shorter bridges were already beginning to fail due to fracturing of their cast-iron girders, Stephenson decided on a measure that Ressler says today would be called “belt and suspenders engineering.” Or as Why Buildings Fall Down states, “In practice, all structural failures can be considered due to a lack of redundancy.” What Stephenson did was to use a wrought-iron chain to strengthen the girder—wrought iron is much stronger in tension than cast iron. This assembly was designed to give extra resistance when the girder was subject to the load of a passing train by offsetting the resulting tension in the bottom cast-iron flange with a metal that is much less prone to brittle fracture.

The Dee bridge was in service for eight months without incident when Stephenson became concerned that its wooden deck could be subject to fire from sparks from a locomotive—a real problem in early railroads. So he had a five-inch layer of gravel poured on top of the deck. The next train was the proverbial feather—it made it almost across the last span, when one of the girders fractured, spilling the cars into the river 40 feet below. Five of the 25 aboard died, and 17 were injured.

“At the time the incident set off a national furor,” according to “Iron, Engineering, and Architectural History in Crisis,” a paper by William M. Taylor of the University of Western Australia. He relates how the accident created a crisis of confidence among the traveling public, demanding a response. “Newspapers established an imaginative topography of risk that brought order to eyewitness accounts of the disaster. Reportage rendered the event sensational and made it subject to interpretation according to multiple and overlapping causal schema and probabilities. Questions about the performance and reliability of new or competing technologies and the competence of their proponents come to the fore.”

At this tense juncture, the Railway Inspectorate created by the 1840 legislation sprang into action. Captain Simmons of the Royal Engineers was named chief investigator. After extensive research, he concluded that repeated flexing of the stiff cast-iron girder had robbed it of strength—a phenomenon we now know as metal fatigue. But there was more to the story. As Professor Ressler demonstrates in his lecture on the Dee disaster, the wrought-iron chain was attached to the girder in a location that made the chain go slack when loads were applied, thus utterly failing to work as intended. It was a critical error in engineering judgment.

Another analysis of the disaster came from what was called a coroner’s inquest but was in fact a trial. The jury declined to find fault or assess liability, deciding it was an accident with many contributing factors. Stephenson was relieved; while a charge of manslaughter had been considered, the jury declined to endorse it. It did, however, find that cast iron is “a brittle and treacherous metal.”

Finally, a royal commission was appointed to investigate. The commission too condemned the design, and called on Her Majesty’s government to establish an inquiry into iron bridges and to require substantially higher safety factors in their design. Reforms followed, led mostly by the engineering profession, which developed new ways of supporting loads across long spans. Additionally, “many existing cast-iron bridges had to be strengthened,” observes Bjorn Akesson in Understanding Bridge Collapses.

In the end, according to Ressler, “Engineering is a fundamentally human enterprise, and engineers learn from failure.” The River Dee bridge was rebuilt out of wrought iron.

Three decades later, the North British Railway was interested in increasing the economic potential of its line running from Edinburgh, Scotland, to Dundee. Unfortunately, en route the rails had to cross two broad estuaries—called firths in the local language—requiring ferries to carry the trains across. A brilliant young NBR engineer named Thomas Bouch developed what Professor Ressler terms the first roll-on-roll-off ferry system.

Bouch then began to promote bridges to replace the two ferries, which were slow and also had problems in bad weather. The Firth of Forth design he came up with envisioned two majestic suspension bridges in sequence, each with what would be the longest span in the world at that time. But first he had to tackle the two-mile-wide Firth of Tay, where an 1870 act of Parliament already authorized a bridge.

Here the river was shallower, so spans would be more in line with his previous work on railroad bridges and viaducts using masonry piers supporting what is called a truss. A truss is in general an assembly of interlocking triangles joined at their vertices. It is rigid when its members are made of wood or metal because triangles retain their shape when stressed. If you want a mental image, think of the Eiffel Tower, a triumph of the wrought-iron truss. Houses and larger buildings have been using roof trusses for millennia—they show up in Roman structures—and they became popular for civil engineering works during the industrial revolution and continue so today. Trusses take relatively little material to achieve extraordinary strength, making them efficient in engineering parlance but also in financiers’ cost-benefit analyses.

A truss can be designed to function as a girder. Instead of two flanges joined by a web, as in an I-beam, a truss can be made with two strong parallel chords joined by thin diagonal members to form the required triangles. Such a truss functions under load in much the same way as an I-beam girder, but the truss uses far less material and can extend over greater distances.

Bouch’s plan for the Firth of Tay Bridge was to use such girder trusses to span a series of brick and concrete pillars built on the river bed. It was an ambitious project, requiring 86 trusses for the spans, with the longest segments, over the navigation channel, 250 feet in length—well more than twice the span of the River Dee bridge segments.

The masonry towers were to be built on bedrock below the river bed. But after 14 towers had been successfully sited, excavators could not find bedrock below the layers of clay and silt on the river bottom at the location of the 15th tower. It was a fateful development—and a sign that NBR had failed to do what we would today call an environmental assessment. Bouch was forced to improvise a less-heavy design that could be adequately supported by the existing geology. He decided to use masonry only up to the waterline. Above that would be a light tower built of iron members joined into a vertical truss.

In making a tower that functions as a truss, engineers need vertical equivalents to the chords of a girder truss. Bouch chose cast-iron pipes for this critical function. These vertical members were designed to bear the weight of the overhead structure and passing trains only in compression, making cast iron seem a reasonable decision.

To build the towers, he used a series of stackable modules that each had six such cast-iron pipes, arrayed at the vertices of an elongated hexagon. (Such a module is visible at the right side of the illustration on this page.) Joining these verticals were horizontal wrought-iron members and wrought-iron diagonals, all fastened with bolts, creating the necessary triangles. These wrought-iron members would in principle give the cast-iron columns the needed stiffness against bending.

The modules, each 11 feet high, were placed on top of each other and bolted together, with as many as necessary added to get the required height over the water. Once again, navigation interests, from the port of Perth in this case, had led the Board of Trade to require a formidable clearance, 88 feet above the water for the width of the shipping channel. To do this, Bouch had to add two extra modules to a series of 12 towers covering a total span over the channel of over half a mile.

It was an ambitious scheme. Instead of the rails passing on the top chord of the truss, as in the rest of the structure, to get the needed clearance over the navigation channel these trusses—known as the high girder—were jacked up such that trains would actually have to go through the truss, riding on the bottom chord. According to H.J. Hopkins, writing in A Span of Bridges: An Illustrated History, “From the river bed to the top of these girders was 170 feet.” According to Akesson’s detailed drawings in Understanding Bridge Collapses, the metal pillars holding up the high girder were only 21 feet wide. These tall and slender towers had to not only support the gravity loads but had to respond to lateral stresses as well from the speeding trains and from wind. That’s where the diagonal bracing was supposed to come into play and stiffen the pillars.

The bridge’s vertical trusses would function as planned if there were no flaws in their implementation—fabrication, construction, and maintenance—but there were. Cast iron is notoriously tricky to make and flaws are inevitable. The diagonal and horizontal wrought-iron braces were attached with bolts passing through attachment points that were cast into the vertical column members upon manufacturing. These lugs had to be made with the needed bolt hole. That required the molten metal to flow in the mold around a removable spike representing the bolt.

But cold-shuts where the metal doesn’t fully join around the bolt hole were common, ensuring a weak point. Further, the removable spikes were tapered to make them easy to withdraw. As Ressler demonstrates with a model, this means that all the force applied to the lug when in service—possibly tons—would be borne by just a tiny portion of that cast-iron attachment point, overstressing the connection.

In addition, the diagonals needed to be kept taut so they could do the work in tension, which was accomplished with wedges held in place by friction. But these required constant maintenance and in practice loosened and thus deprived the cast-iron verticals of the sideways support intended. It was later revealed that painters on the bridge observed horizontal deflections of up to four inches when trains passed overhead. In “Forensic Engineering: A Reappraisal of the Tay Bridge Disaster,” Peter R. Lewis and Ken Reynolds note that the workers could feel deflections when the trains were still far off, and they soon learned to secure their paint pots. Similarly, investigators determined that some passengers had complained about swaying when crossing the bridge.

It was the same engineering fault as in the Dee collapse: a result of improper fastening of wrought-iron tension members that were meant to reinforce cast-iron members against tension from bending. But the new inadequacy sprang not from the joints’ locations, as in the Dee crossing, but from their limited strength when repeatedly stressed.

However, those problems only became evident when the bridge failed catastrophically. Until then the Tay was a celebrated construction. The bridge was completed in 1878 and it passed the Board of Trade’s inspection and was put in service. It was at the time the longest bridge in the world and thus an engineering triumph for Bouch and for Britain. Indeed, when Queen Victoria crossed the Tay shortly after the bridge opened, coming back to Windsor from her summer castle in Scotland, she was so thrilled that a few days later she knighted Bouch.

But Sir Thomas’s moment in the sun was short-lived. On the night of December 28, 1879, there was a gale-force wind coming down the Firth of Tay—by some estimates perhaps 80 miles an hour. Just after 7 p.m., a northbound passenger train en route to Dundee crossed the bridge. Eyewitnesses saw sparks flying from the flanges on the wheels as they bit into the rails under the pressure of the wind. Then nothing—the train had simply disappeared. The whole 3,000-foot-long high-girder structure—including all 12 of the truss towers supporting it—had simply broken off the masonry piers and fallen over sideways, taking the train and its passengers with it. Some 75 souls died in the fall or from drowning. There were no survivors.

Pursuant to Section Seven of yet another piece of legislation, the Railway Regulation Act of 1871, the Board of Trade appointed a court of inquiry, a body with judicial powers. In Engineers of Dreams: Great Bridge Builders and the Spanning of America, Henry Petroski describes how the court “discovered major flaws in the design and construction. It was found, for example, that Bouch had grossly underestimated the effects of strong winds that could develop along the firth. Upon being questioned about this during the inquiry, he gave no sign of having reconsidered the issue.” Ressler, however, relates that Sir Thomas had sought counsel on the subject from a most authoritative source, the Astronomer Royal, but misapplied the advice he had received. In fact, Ressler says, wind was surely the proximate cause of the disaster, but other factors had weakened the structure and contributed to the collapse.

There were three members of the court: “William Henry Barlow, president of the Institution of Civil Engineers; Colonel William Yolland, chief inspector of railways; and Henry Cadogan Rothery, wreck commissioner,” writes Petroski. “Only the first two were engineers, and professional loyalties appear to have surfaced when the time came to draft the final report.”

The three men did agree on the central conclusion: “We find that the bridge was badly designed, badly constructed, and badly maintained and that its downfall was due to inherent defects in the structure which must sooner or later have brought it down.” However, the engineers were reluctant to assign the blame to Bouch, because in their view there were too many unknown factors about what caused the collapse. They also felt they lacked the legal authority to determine fault, even though previous Section Seven inquiries had done so.

But Rothery, a lawyer, submitted a minority report. He placed the blame squarely on the engineer for a faulty design, sloppy construction, and inadequate inspection and maintenance, all of which fell under Sir Thomas’s responsibility.

The inquiry did make some unanimous findings concerning the causes of the collapse. It concluded that “the fall of the bridge was occasioned by the insufficiency of the cross bracing and its fastening to sustain the force of the gale.” Photographs it commissioned show broken-off pieces of bolt-and-lug assemblies littering the tops of the masonry piers that had supported the iron towers; numerous lugs had failed in tension, where cast iron is weakest, and many showed signs of cold-shuts. Ressler points out that fatigue no doubt played a role in the failure of the cast-iron lugs, as they were repeatedly stressed when—because of the insufficiently taut diagonals—the towers swayed with each passing train. He also notes that as the lugs failed and the towers broke off and toppled into the firth, many pillars simply pulled out of the bolts anchoring them to the masonry foundations.

As to the role of the wind in precipitating the collapse, Yolland and Barlow said, “There is no requirement issued by the Board of Trade respecting wind pressure, and there does not appear to be any understood rule in the engineering profession regarding wind pressure in railway structures; and we therefore recommend the Board of Trade should take such steps as may be necessary for the establishment of rules for that purpose.”

Rothery again dissented, insisting it was up to the engineering profession to devise safety standards for wind loading. The board compromised—it empaneled a five-person expert commission to consider the issue. The commission established standards for wind loading and safety factors to be applied to large projects.

Sir Thomas died just a few months after the inquiry’s report, a broken man. As Ressler states, “In a sense, he was the 76th victim of the Tay Bridge disaster.”

The death knell for cast iron as a structural metal came not from government or the courts, however, but from the engineers. Its time as a major construction material had passed. According to Understanding Bridge Collapses, “The lessons learned from the Tay Bridge disaster were adapted to the full in the construction and design of both the New Tay Bridge and the Forth Bridge. Maximum stability was ensured and the materials used were tested meticulously—nothing was left to chance.”

Bouch had become so disfavored that his plan for the double suspension bridge over the Forth was canceled and was replaced with a structure of immense proportions, of a radically new design, the cantilever truss. (See photo.) It crossed the deep estuary with two great leaps of 1,600 feet each, the longest spans in the world at the time. And its towers were sloped outward so that they had a squat cross-section, making them virtually invulnerable to the wind. Finally, in a thorough repudiation of past practice, it was made of steel, which had just become available in industrial quantities, and which is stronger—in both tension and compression—even than wrought iron. The Firth of Forth Bridge, now 133 years old, ended up costing ten times more than the original Tay Bridge, but for North British Rail and the riding public, it was an all-too-reasonable expense.

In replacing the Tay Bridge, however, Scotland showcased the traditionally thrifty character of its renowned engineering fraternity. Whereas the original bridge carried a single track, the new one promised to make more money by carrying two tracks on much broader wrought-iron towers, presenting a wider stance to the wind. Second, the railroad salvaged the locomotive—and waggishly renamed it The Diver—returning it to service. Finally, the company recycled the wrought members of the old bridge, reusing them in the immensely strong new structure still in service today.

Notice & Comment is the editor’s department and represents his views.