Learning from Structural Failures

In 1968, a gas explosion in a corner flat on the eighteenth floor of Ronan Point, a newly built precast-concrete tower in East London, blew out a single load-bearing wall panel. That one panel had been holding up the corner of every flat above it. With its support gone, the flats above fell onto the flats below, and the impact of each falling floor overloaded and collapsed the next, so that an entire corner of the twenty-two-storey building peeled away from top to bottom. Four people died. The building was barely two months old.

The failure mode is progressive (or disproportionate) collapse: a local failure that spreads, through a chain reaction, into a collapse grossly out of proportion to the original cause. The precast panels had been connected in a way that gave the structure no alternative load path and no ability to bridge over a lost element - remove one, and there was nothing to arrest the cascade. The deeper lesson was about robustness: a building must be able to survive the loss of a single element without collapsing disproportionately, through redundancy, continuity and ties that let load find another way down.

Ronan Point transformed the codes. It gave us the modern requirement for structural robustness and tying - the rules that a structure must be tied together horizontally and vertically so that local damage stays local, and that key elements either be protected or the structure be able to lose them safely. Every architect designing anything tall or repetitive inherits this lesson: the question is never only 'is each element strong enough?' but 'what happens if one of them is suddenly gone?'

In 1981, during a crowded tea dance in the atrium of the Hyatt Regency hotel in Kansas City, two suspended walkways stacked one above the other collapsed onto the crowd below, killing 114 people and injuring more than 200 - at the time the deadliest structural collapse in United States history. The cause has become the single most-taught case study in engineering, because it is so simple, so avoidable, and so human.

The original design hung both walkways from a single set of long steel rods running from the roof: the upper walkway's beam and the lower walkway's beam were each to be supported on the same continuous rods. During fabrication, this was changed - for practical, build-ability reasons - to two shorter rods: one from the roof to the upper walkway, and a separate one from the upper walkway to the lower. It looks like a trivial detailing change. But it doubled the load on the upper walkway's connection: instead of that beam carrying only its own walkway, it now carried the entire weight of the lower walkway hanging beneath it, all through one box-beam connection that had never been designed for it. Under the crowd, the connection tore through the beam and both walkways fell.

The failure mode is a connection failure, but the true lesson is procedural: a structural change was made and nobody with the competence and responsibility independently re-checked it against the design. This is exactly the scenario the previous lesson warned of - the 'minor' change, the missing second pair of eyes. Hyatt Regency is why proof checking, formal review of shop-drawing and design changes, and clear responsibility for connections are now treated with such seriousness. A change is never too small to check when it lies in a load path.

In 1940, only months after it opened, the Tacoma Narrows Bridge in Washington State began to twist and undulate in a moderate wind of around 65 km/h - far below any wind the bridge was 'strong' enough to resist - until the oscillations grew so violent that the deck tore itself apart and dropped into the water. Famously captured on film, 'Galloping Gertie' is the most vivid lesson in structural dynamics ever recorded.

The bridge did not fail because it was too weak in the ordinary sense; it failed because of aeroelastic flutter, a dynamic interaction between the wind and the structure's own motion. The slender, solid-sided deck shed vortices and coupled with the wind in a way that fed energy into its natural twisting mode faster than damping could remove it, so each oscillation grew larger than the last - a self-reinforcing spiral of resonance until the material could take no more. The static strength of the deck was almost irrelevant; what mattered was its dynamic behaviour, its shape in the airflow, and its frequencies.

The lesson reshaped how we design anything slender and exposed - long-span bridges, tall towers, cable structures, large roofs. It established that dynamics and aerodynamics matter as much as static strength, and it is the origin of routine wind-tunnel testing and dynamic analysis for major structures, and of shaping decks and towers to be aerodynamically stable rather than merely strong. For the architect, the enduring point is that a structure is not just a static object resisting forces but a dynamic system with frequencies of its own, and that wind, footfall and earthquakes are dynamic actions that can find and amplify those frequencies. Strength alone never guarantees safety.

Not all lessons come from spectacular one-off events; many of the deadliest failures are depressingly ordinary buildings that were simply not detailed to survive the forces they met. In the 2011 Christchurch earthquake in New Zealand, the CTV Building, a six-storey reinforced-concrete office block, collapsed almost completely, killing 115 people - the largest loss of life in that earthquake. The investigation pointed to a combination of non-ductile detailing, inadequate connections between elements, and vulnerabilities that current understanding of seismic design would not permit. It is a sobering reminder that a building can be legal when built and still be lethal when the ground moves, and that seismic knowledge - and the codes that carry it - advance precisely because of such events.

India knows this failure mode intimately. The country's recurring building collapses - during the 2001 Bhuj earthquake, and in the steady toll of non-earthquake collapses of poorly built apartment blocks in its growing cities - repeatedly trace to the same causes: non-ductile reinforced concrete, soft or open ground storeys (a weak, column-only ground floor under heavier storeys, which concentrates all the earthquake demand into a floor with no walls to resist it), poor-quality materials and workmanship, unauthorised extra floors, and construction that departed from any competent design. These are not exotic engineering mysteries; they are failures of detailing, quality and enforcement. It is exactly why IS 13920 (ductile detailing) and the seismic provisions of IS 1893 and the National Building Code exist and must be applied, and why the enforcement gap between a good code and a badly built building is one of the most important safety issues in Indian construction.

The common thread across CTV and the Indian cases is ductility and honest execution. A structure survives extreme events not by being merely strong but by being able to bend, absorb energy and warn before it fails - and by actually being built the way it was designed. The everyday collapse teaches the same lesson as the famous one: the details and the discipline are the safety.

Step back from the individual cases and a small set of recurring causes appears, repeated across decades and continents. Failures cluster around: a change that was not re-checked (Hyatt); a lack of robustness or alternative load path (Ronan Point); ignored dynamics - resonance, flutter, vibration (Tacoma); non-ductile detailing that fails suddenly rather than bending (CTV, Indian RC collapses); poor connections, which are so often the true weak link; and the human layer beneath all of them - communication breakdowns, unchecked assumptions, commercial pressure, and warnings that were noticed but not acted upon. Very few great failures are caused by a single heroic villain; most are the alignment of several small holes - a flawed detail, a missed check, a pressured schedule, a silent assumption - lining up until the failure passes straight through, the 'Swiss cheese' model of accidents.

What makes the profession remarkable is the learning loop it built in response. A failure is followed by rigorous, often independent forensic investigation to establish the true root cause, not a scapegoat. The findings become changed codes and changed practice - robustness ties after Ronan Point, mandatory checking after Hyatt, wind-tunnel testing after Tacoma, ductile detailing after every major earthquake. And those lessons become shared professional memory, taught to every student and written into the standards, so the whole discipline gets safer over time. This is why structural engineering, uniquely among the building arts, treats its failures as public property to be studied openly rather than as embarrassments to be buried.

For the architect, the practical takeaways are clear and permanent. Ask always 'what if one element is gone?' (robustness). Never let a structural change go unchecked, however small it seems. Respect dynamics, not just strength. Insist on ductile detailing and honest construction in seismic zones. And read failures deliberately throughout your career - they are the most concentrated structural education available, and every one of them was paid for in advance by people who did not walk out of the building.