by Roma Agrawal
As I learned later, my dad had been sitting at his desk, shouting down a poor phone line to one of his clients when a huge bang rattled the building. At first he thought an electricity generator or a large cooling unit had exploded. He jumped out of his seat, telling his staff to stay calm and remain in the office. Seconds later, however, he heard terrified people running down the stairs. Many screamed that there had been a bomb and that everybody should get out as quickly as possible. My father, uncle and their colleagues left their office, to scenes of horror.
Hundreds of people were filing down the stairs. There was barely any space to move. Head down, he focused on taking one step at a time, trying not to look at the dismembered bodies – the arms, the legs, the blood – that lay just beyond the staircase. Finally, he arrived at the ground floor. Emergency vehicles, trying to deal with the injured, blocked the street. My father and uncle fled the area and got on a bus to my grandmother’s house. About two hours after we’d come home from school – the longest two hours of my life – Pop called us to tell us they were both safe.
Years later, while studying for my masters in structural engineering, I attended a class in which we discussed how to protect towers against explosions. Suddenly, the events of that terrible day in March came rushing back. For the first time a thought occurred to me: given that it was rocked by a serious explosion right at the base of the structure, and fires broke out afterwards, why didn’t the whole Bombay Stock Exchange tower collapse?
I know now that there are two main reasons for this. The first is that engineers design certain buildings to resist explosions, so even if it is hit and damaged, it doesn’t collapse like a house of cards. There is a minimum standard of safety governing the design of all structures, but the more vulnerable ones – tall, iconic buildings, for example, or those with particularly large numbers of people inside – are designed specifically for a range of possible explosion scenarios. The second reason is that all structures should be designed to stop fires rapidly engulfing them, providing enough time for occupants to escape, and for the fire to be tackled or burn out – contained in a small area – before it causes significant structural failure.
But we didn’t start out building this way; we have learned from disasters of the past.
*
After waking early on the morning of 16 May 1968, Ivy Hodge went to the kitchen to make a cup of tea. She turned on the gas hob, struck a match – and the next thing she knew she was flat on her back, looking at the sky. A wall of her kitchen and a wall of the living room had disappeared.
In Ivy’s flat on the 18th floor of a 22-storey tower block in Canning Town, London, there had been an explosion. Occurring in peacetime in a quiet residential neighbourhood, it was an event without precedent in the city, and it profoundly influenced how we would build future structures.
The tower had been constructed quickly as part of the regeneration desperately needed in the aftermath of the Second World War. The neighbourhood had lost about a quarter of its homes to bombing, and the destruction, coupled with the large post-war population increase, meant there was a severe housing shortage. To build rapidly and efficiently, new forms of construction were being experimented with. This particular structure was the second of nine identical towers being built to create an estate called Ronan Point.
The tower had been thrown together hastily by ‘prefabrication’. Instead of pouring wet concrete on a construction site and waiting for it to solidify to form walls and floors (like most other concrete construction required), room-sized panels of concrete were made in a factory. The panels were then driven to site and lifted in to place with a crane. It was like building a house of cards: put up the walls of the ground floor, carefully place the horizontal panels on top of them to create the first floor, and so on, up and up. The panels were joined together with a small amount of wet concrete on site. The weight of the building was being channelled through these large load-bearing panels; there was no skeleton or frame. This novel prefabricated system produced lower costs, quicker construction times and required less labour, all important economic factors to consider in recovering post-war Britain.
Poor detailing, such as that used at Ronan Point, where only a small amount of wet concrete was used to join together prefabricated panels during construction.
In Ivy Hodge’s flat, gas had been leaking steadily from her recently installed but defective boiler system. The match flame had lit the escaped gas and BOOM!, the wall panels making up the corner of her flat blew out. With nothing now supporting them, the wall panels of the flat above fell, hitting the level below. One by one, each floor on that corner of the towerblock collapsed, taking a great chunk out of the structure, from top to bottom. Four people, asleep in their flats, died.
Oddly, the explosion did not perforate Ivy’s eardrums, which suggests that its force wasn’t that large – since it doesn’t require much pressure to damage them. In fact, subsequent investigations showed that even an explosion with just a third of the force of the actual event would have dislodged the wall panels. Since the panels were just sitting one on top of the other, without being tied together properly, there was little to stop them blowing out. The designers had relied on friction between the panels and the little bit of wet concrete ‘glue’ to hold them in place. It wasn’t enough. When the explosion pushed out on the wall, the force of the push was bigger than the resistance of the friction and the concrete, and it flew out. Then, because the load from the walls above had nowhere to channel itself, the walls simply fell.
The disproportionate collapse of floors following an explosion at Ronan Point, London, in 1968.
There was another unusual thing about this collapse. Normally I would expect an explosion at the base of a building to cause the most damage, because there are many storeys above it which can come crashing down. In this case, however, if the same explosion had happened at the base of the building, the collapse might not have happened at all.
Friction depends on weight. The heavier the load acting at the junction between two surfaces, the greater the friction. Close to the top of the tower (where Ivy was), there were only four storeys of weight at the junction between wall and floor, so the friction was low. The pressure of the explosion overcame the friction and sent concrete panels flying. But at the base of the tower, the weight of more than twenty storeys of panels created greater friction between wall panels (it’s the reason why pulling a magazine out of the base of a stack is much harder than extracting one from higher up). So counter-intuitively, the explosion near the top was the event with disastrous results. This is not a very common occurrence now – especially because, as we’ll see, buildings aren’t built like this anymore.
The debacle at Ronan Point had two important lessons for future construction. Firstly, it was vital to tie structures together, so that if a wall or floor panel were pushed with a force bigger than expected, the ties would stop the panels from sliding out. (At Ronan Point, steel rods, for example, tying together the prefabricated wall panels between floors could have helped the building withstand the blast; variations of this tie-system are used in modern prefabricated buildings.) Even for structures built in a more traditional way, with all the concrete poured, or steel being fixed on site, it is essential to make sure that the beams and columns have robust connections. In the case of steel frames, the bolts used to join pieces of steel together should be strong enough not only to resist normal loads exerted by wind and gravity, but also to keep the structure bound together.
Secondly, engineers had to prevent disproportionate effect. At Ronan Point, a single explosion on the 18th floor caused the corner of the tower to collapse at all levels. This domino-effect was disproportionate to the cause, and a new term, disproportionate collapse, was born. If an event like an explosion happens, then of course damage will occur, but the effect of an explosion on one storey shouldn’t propagate throughout the structure. The problem at the Canning Town tower block was that the loads didn’t have anywhere to go. So
the key is to ensure that the forces have somewhere to go, even if part of a structure disappears. It’s like sitting on a stool: theoretically, only a quarter of your weight is transmitted through each of the four legs. But if, like many people, you’re inclined to tilt the stool so all your weight is going down only two legs, you’ve just doubled the load the leg is designed for – the legs fail, you hit the ground, and you bruise your backside. But if structural engineers anticipate this sort of behaviour and design every leg for double the load, then you’re safe.
Thus the idea of consciously creating new paths for loads to travel through was born. In my computer model, I will delete a column, record the larger forces in neighbouring columns, and design for this higher load. Then I know that even if that column is gone, its neighbours will do its job. Then I put that column back in and remove another one, trying different combinations to check my structure is stable in the face of explosions. Never challenge a structural engineer to a game of Jenga: we know which blocks to remove – how to take chunks out of a structure so that it doesn’t crash.
*
Throughout history, engineers and civic authorities have been engaged in battle – against the fires that threaten to raze our towns and cities to the ground. Roman houses were often made with timber frames, floors and roofs, which caught alight easily, and fires were common. The Great Fire of Rome in AD 64 laid waste to two-thirds of the city. Originally, timber was not protected with anything to resist fire like it is now, and walls were made from wattle and daub. Wattle, a lattice woven from narrow wooden strips which looked a bit like a straw basket, was coated – daubed – with a mixture of wet soil, clay, sand and straw. Such a construction was highly flammable, enabling fire to spread quickly. The narrow streets aggravated the situation because flames could easily jump the small distance between one building and another.
In the first century BC, Marcus Licinius Crassus was born into the upper echelons of Roman society. He grew up to become a respected general (he helped quash the slave revolt of Spartacus) and a notorious businessman. Crassus was a man who spotted opportunities: observing the devastation caused by Rome’s fires, he created the world’s first fire brigade, made up of over 500 slaves who were trained to fight fires. He ran it as a private business, rushing his team to burning buildings, where they intimidated and drove away rival firefighters, then stood about until Crassus had negotiated a price to put out the fire with the building’s distraught owners. If no deal could be reached, the firefighters simply allowed the structures to burn to the ground. Crassus would then offer the owners a derisory sum to purchase the smoking site. This meant that he quickly managed to buy up much of Rome, and amassed a fortune as a result. Fortunately, modern-day fire brigades work on a more honest basis.
After the Great Fire of Rome, Nero ordered several changes to the city. Streets were made wider, apartment buildings limited to six storeys, and bakers’ or metal workers’ shops separated from residential units, using double walls with air gaps. He proclaimed that balconies should be made fire-proof to make escape easier, and invested in improving the water supply, so it could be used to extinguish fires. The Romans learned from tragedy, and we too have benefited from that hard-won wisdom. Thousands of years later, these simple principles – separating rooms, flats and buildings with fire-resistant materials and installing air gaps – are still used to prevent fires ravaging modern structures.
*
On 11 September 2001, the world watched in horror as two planes collided with the World Trade Center towers in New York. I was in Los Angeles on holiday before starting at university, and was scheduled to fly to New York the next day. Paralysed, I sat watching the news, shocked as the towers collapsed an hour after being hit. A few days later, I went directly back to London, already feeling part of a changed world.
Looking at the events from an engineer’s point of view, the events of that appalling day had a ripple effect on the design and construction of skyscrapers. Reading about the structural failures that led to the collapse of the towers, I was surprised to learn that it wasn’t just the impact of the planes that caused the devastation, it was also the fires that followed.
New York is filled with spectacular skyscrapers, yet the World Trade Center’s twin towers (opened in 1973) were among the city’s most iconic symbols. Visually, each of the towers was very simple – a perfect square from a bird’s-eye view, 110 storeys high. Each had a large central core made of steel columns. But this spine wasn’t responsible for keeping the towers stable: they used the ‘turtle-shell’-style exoskeleton instead.
Vertical columns, spaced just over a metre apart all around the perimeter of the square, were joined up at each storey with beams. The beams and columns together formed a robust frame, similar to the construction of the Gherkin we saw earlier, but with giant rectangles instead of triangles. The connections between the beams and columns were very stiff. This external frame kept the building strong against the force of the wind.
When the planes crashed into the towers, giant holes opened up in the exoskeleton. A number of columns and beams were destroyed. Engineers had in fact planned for the possibility of some form of impact by aeroplane. They had studied what might happen if a Boeing 707 (the largest commercial aircraft in operation at the time of construction) hit the building, and they had designed accordingly. The beams and columns had been constructed with extra-strong connections tying them together, so even though some of the structure was gone, the loads found somewhere else to go: they flowed around the hole (using the principle of preventing disproportionate collapse, which engineers had learned from Ronan Point).
Loads within a building find new routes as the forces are channelled through alternative load paths.
The planes that hit the twin towers were not the Boeing 707s that engineers had planned for nearly 30 years earlier; they were larger 767s, carrying more aviation fuel. On impact, the fuel caught fire, and the conflagration of the fuel, aircraft parts, desks and other flammable material inside the building made the steel columns very hot. When steel gets hot, it behaves badly: the tiny crystals which make up the material become excited, vibrate and begin to move around, and the normally strong bonds between them are loosened. Loose bonds mean soft metal. So hot steel is weaker than cold steel, and cannot bear the same load. On 9/11, the columns just next to the holes were supporting a larger load than usual, because they were channelling not just their own forces but also those their neighbours had once carried. The steel columns and floor beams had been sprayed with a special paint mixed with mineral fibres, designed to insulate the steel from the heat of a fire and prevent it from getting too hot. But the crash of the plane and the projectile debris had chipped away areas of the protective paint, leaving big patches of exposed steel. The temperature of the columns around the perimeter of the tower rose ever higher.
The steel columns which made up the core also became unnaturally hot. Two layers of gypsum board (a panel made of gypsum plaster pressed between two thick sheets of paper) separated the core from the rest of the building. The idea was that a fire in the office space couldn’t infiltrate the core past these boards, so people could run into this safe zone and to the stairs to escape. But this board was damaged, leaving the core columns susceptible and the intended safe passage exposed.
The columns became weaker and weaker, and as temperatures reached about 1,000° Celsius, they gave up. They couldn’t carry the forces any more and they bowed.
In the end, the columns failed completely and the structure above it was then left vulnerable to the effects of gravity. The floor above the failed columns came crashing down. But the level on which it landed wasn’t strong enough to resist the falling load and it too failed. One after the other – in a domino-effect reminiscent of the Canning Town disaster but on an even more shockingly huge scale – all the floors failed and the towers came down. The fire protection – paint and boards – was no match for the size and intensity of the fire.
The way we design
skyscrapers has changed since that day. Now, we make sure that escape routes are protected more robustly. The easiest way to do this is to build the core in concrete instead of steel, so that instead of weak gypsum boards standing between the fire and safety, you have a solid wall of concrete.
Concrete is not a good conductor: it doesn’t transmit heat well, which means it takes longer to heat up. To strengthen concrete, however, we insert steel reinforcement bars into it; these are excellent conductors of heat, which creates a problem for the engineer. In a fire the steel bars heat up, and the heat energy spreads quickly through their length, while the surrounding concrete heats up slowly. The hot steel expands more quickly than the colder concrete, causing the outer layers of concrete to crack and burst off. This is similar to how thick glass tumblers crack if you pour hot water into them: the inner layers of the glass get very hot and expand, but the outer layers remain cold because glass, like concrete, is not a good conductor of heat. As the inner layers expand against the colder outer layers, the outermost cracks.
Through testing and experimentation, we know how long it takes for concrete to conduct heat to steel bars, and then for the steel bars to heat up and make the concrete burst. So we bury the steel deep enough in the concrete to ensure that the fires can be put out before the outer layer of concrete is damaged. This buys enough time for people to leave the building through the concrete core, or for firefighters to get the flames under control, without the structure collapsing. The taller or larger the building, the longer it takes to escape, so the deeper the steel is embedded in the concrete. Just a few centimetres make a tremendous difference.
So concrete cores perform a dual function: keeping the building stable against wind loads, and forming a protected escape route for the occupants. Today, even if we use an exoskeleton to resist wind (which means we don’t need an internal core), we still often install concrete walls to safeguard escape routes. And the protection for steel columns and beams against fire has also been improved dramatically: fire-resistant boards and intumescent paint (which expands when heated and insulates the metal) are much more robust now than ever before. They stop steel getting too hot too quickly, so it remains strong.
