by Roma Agrawal
My new boss was John, a slim man of average height, with straight, short dark hair, rimless glasses and a passionate love of cricket (something that, even though I grew up in India, I couldn’t match). We went through some forms, a process made lively by his occasional ironic and funny observations; meanwhile I kept quiet about the fact that it was my 22nd birthday. Then he showed me his hand-drawn sketch of a new footbridge, made from steel, that was due to be built in Newcastle. The confident pencil marks showed that at the east end of the bridge a tall tower would support three pairs of cables. The cables in turn would hold up the main deck of the bridge. To counter-balance the weight of the bridge on this tower, a further set of cables would anchor it from behind. As I sat with John, looking at the drawings in front of me, I did a little dance inside. As far as I was concerned, this was as good a birthday present as a girl could get. I was thrilled that my first project was going to be this elegant and distinctive structure. Apart from its lovely aesthetic, however, this bridge had other nuances that made it, to my eyes, even more beautiful.
A working sketch of the Northumbria University Footbridge by John Parker.
The bridge is a ‘cable-stayed’ bridge, one famous example of which is the Millau Viaduct in France. Its gently curving deck is held in place by seven pillars, from which cables fan out in the shape of a sail, giving the impression that the bridge is floating 270m above the Tarn valley. Cable-stayed bridges have one or more tall towers to which cables are attached; the deck is pulled down by gravity, and is held in place by cables, which are always in tension. The tension forces are channelled through the cables directly into the tower. The tower in turn compresses and the forces flow down into the foundations on which the tower is supported; the foundations spread the forces out into the ground.
The Millau Viaduct in France is an elegant example of a cable-stayed bridge.
As a fresh-faced engineer, designing the cables for the Northumbria Footbridge (which were as thick as my fist) was a real challenge. If you take a metal ruler, representing the steel deck, and use three pairs of rubber bands to mimic the cables, you’ll find that you have to pull on each band just the right amount before they’re all taut and supporting the ruler evenly so it lies flat. If you pull too hard on the three bands on one side, the ruler tips over sideways. If you pull too tightly on the central pair, it bows upwards. Now imagine the same effect, but on a real, full-sized bridge.
I used software to create a three-dimensional computer model to recreate the bridge beams that run under the deck and the cables that run from the deck to the mast. Then I simulated gravity on the structure. I also had to consider the weight of all the people that would stand on the bridge, and that they might congregate on different parts of the bridge at different times. For example, during the Great North Run, in which athletes run along the motorway below the structure, cheering crowds might stand on one side as the runners approach them, then walk to the other side to watch them continue into the distance. I had to think about ‘patterned loading’ – I modelled people standing on the bridge in different configurations. No matter where people stood, the cables had to remain tight to support the deck. If the cables were not in tension they would become floppy, and the deck would lose its support. To stop this from happening, I added extra tension to the cables artificially.
Cables can be tightened up using a jack – which is a tube with clasps on each side. Each cable had at least one break in it where a jack could be installed. The clasps each held a bit of cable either side of the break. The jack can be adjusted to pull the ends closer together (to tighten the cable) or further apart (to loosen it), therefore altering the amount of force in the cable. If you look at the cables fanning out from the tower of my footbridge you’ll see that they have connector pieces – where the cables look briefly thicker than the rest of their length: those are the points at which the jacks were temporarily connected. This is like replacing the rubber-band cables in our demonstration with shorter ones, but then stretching them out to the same length as before. This puts more stretch in the rubber bands – they contain a higher pulling or tension force.
The key to building a cable-stayed bridge is balance. If you use a thin piece of card as a deck and pull on the rubber bands, the card simply lifts up. If you replace the thin card with a book, then you can pull on the bands to make them taut without deforming the book. Once the stiffness and weight of the deck and the tension in the cables are reconciled and calibrated, you can then work out what the force is in the cables. When I did the drawings of the bridge, I added notes stating how much each cable needed to be tightened to stop it going slack.
The engineer’s job is a lot like plate-spinning. You have to plan for, and control, a multitude of problems simultaneously. Take temperature: like all structures, my bridge is affected by it. Throughout the year, to varying degrees (depending on the season), it will be heated and cooled. Steel has a ‘coefficient of thermal expansion’ of 12 x 10-6. This means that for every 1 degree of change in temperature on a piece of material 1mm long, the material will expand or contract by 0.000012mm. This may sound small, but my bridge was about 40m long and had to be designed for a temperature range of 40 degrees. The savvy among you will argue that the British summer is not 40 degrees warmer than the winter, and you would be correct, but the steel itself will get much hotter than the air as it absorbs heat from the sun. We’re looking at the range of temperature experienced by the steel, not the air, in the most extreme (but reasonable) weather we can anticipate.
This adds up to an expansion of nearly 20mm. If I fixed the ends of the bridge to stop it from expanding or contracting, a large compression force would build up in the steel deck when it got warmer, and a large tension force would build up when it cooled down. The problem is that this expansion and contraction could happen thousands of times over the life of the bridge; this constant pulling and pushing can gradually damage not only the steel deck itself, but also the supports at either end.
To prevent this, I allowed one end to move. (In larger bridges, or bridges with many supports, you can create ‘movement joints’ in multiple places. You can sometimes feel your car ‘boing’ as you drive over them.) Because the movement on this bridge was relatively small, I used a ‘rubber bearing’ to absorb it. The steel beams which made up the deck were supported on these bearings, which were about 400mm wide, 300mm long and 60mm thick. When the steel expands or contracts, the bearings flex, letting the bridge move.
I also needed to think about vibration and resonance. I’ve already explored how an earthquake can cause a building to resonate, just as an opera singer can shatter a wine glass when she hits the right note. With the footbridge, I was concerned about whether resonance could make pedestrians feel uncomfortable. Heavy bridges, like those made from concrete, generally don’t suffer from this problem because their weight stops them from vibrating easily. But the steel deck was light, and its natural frequency was close to the frequency of walking pedestrians, which meant it was in danger of resonating. So we connected tuned mass dampers with strong springs to the underside of the deck. These work in a similar way to the giant pendulum inside the Taipei tower, absorbing the sway and stopping the deck from vibrating too much. You can’t see these tuned mass dampers unless you look carefully at the bottom of the deck from the road underneath the bridge (perhaps while stretching your legs on the Great North Run). If you do, you’ll notice three steel box-like objects hidden between the bright-blue-painted beams.
A type of tuned-mass damper, similar to those used on the Northumbria University Footbridge.
Once I was sure that my bridge was stable in its final configuration, I had to work out exactly how it would be built. As it was too large to be transported to Newcastle fully constructed, I went to a steel fabricators’ factory in Darlington. Amid showers of sparks cascading from a welder’s arc, we discussed some options. We would have to bring the bridge to the site in pieces that fit on the back of lorries, so we looked at splitting i
t in various places, checking how those sections could be installed and supported safely until the cables had been tied in; like a sculpture that would need to support itself even while each piece was being placed.
We also had to consider how to cause minimal disruption to the public. Since the structure was to span a motorway, we decided the best approach was to bring it to site in four pieces, connect them together, and then use a crane to lift the assembled bridge into place. A one-of-a-kind monster crane was booked to do the job.
Months of planning went into ensuring that the bridge was hitched up without a hitch. First, the crane itself arrived in pieces at the start of a bank-holiday weekend, and roads were closed off as it was assembled by swarms of steel fixers. Meanwhile, the four steel sections of the bridge were transported from Darlington to a nearby car park, where they were joined together, like a jigsaw puzzle, to make the deck.
The plan was to hoist the steel deck into place, and then to attach the cables. I had designed the deck such that it needed all three sets of cables to resist both its own weight and the weight of pedestrians crowded on top of it. This meant that, until the cables were in place, it needed extra support on site, so I had also calculated that the deck could stand up with a single support at its centre (it had less load on it in this configuration as the public wouldn’t have access). We erected a temporary steel column in the central reservation of the motorway.
The motorway was closed. The crane swung into action. The deck was lifted up from the car park and lowered into place, its ends held up by their permanent concrete supports, and its centre by the temporary steel one. The deck was disengaged from the crane, and the motorway reopened. This complex operation took just three days.
Over the next few weeks the rest of the bridge was assembled. The mast was lifted into place using a crane and then anchored to its concrete base with bolts. The all-important cables could then be installed in pairs starting from one end of the bridge. Every time a new pair of cables was connected, the tension was adjusted using a jack. Once the cables were all in and adjusted one final time, the road was closed again, the temporary steel column removed, and the bridge was complete.
I’m not normally excited about getting up early, but my eyes were already wide open at 5am on the day I travelled to Newcastle to visit my completed bridge, which was now ready and open to the public. After taking a first small step, which felt to me like a giant leap, I walked back and forth across the bridge a number of times. I skipped and I ran. The solid steel beams, the taut cables, the rubber bearings, the tuned mass dampers – they all reminded me of the time, only a few months ago, when I had painstakingly designed them. Details that perhaps no one would notice except me – but they made me happy.
At one end of the bridge there was a bench. I sat there, grinning, for a while, watching bleary-eyed students walking across the deck from one lecture to another, all of them oblivious to the pleasure it gave me to experience my first physical contribution to the world.
ROCK
I’ve been known to stroke concrete. Others might feel the irresistible urge to pat a little kitten or handle an object in a museum, but for me it’s concrete. It doesn’t matter if it’s a smooth, stark grey surface, or one with little stones visible, or even one left intentionally rough – I have to know what the texture feels like, how cold or warm it is. So you can imagine how I felt when I visited Rome and saw tonnes of ancient concrete above my head, but too far away to reach.
The Pantheon in the Piazza della Rotonda in Rome is one of my favourite structures. Built by the emperor Hadrian around AD 122 (at about the same time as he was building a wall to divide England from Scotland), it has stood strong ever since in a variety of guises – temple to the Roman gods, Christian church, tomb – though barbarians removed what they could and Pope Urban VIII even melted the ceiling panels to make cannons. A triangular pediment supported by a portico of sixteen Corinthian columns greets you at its entrance. Inside, the rotunda is topped with a dome punctuated by a circular opening (oculus – Latin for eye) through which streams an almost otherworldly shaft of light. It’s an atmospheric and beautifully proportioned building. I’m overwhelmed by its sheer scale when I wander around in it, bumping into people as I stare up at the beautiful roof. Even now, it’s the largest unreinforced concrete dome in the world. The Romans really honed their craft, creating an engineering masterpiece from a revolutionary material they called opus caementicium.
The giant concrete dome and oculus at the Pantheon in Rome, Italy.
For me, what’s special about concrete is that its form is indeterminate: it can be anything. It starts as rock, then becomes a lumpy grey liquid that can be poured into a mould of any shape and left while chemistry takes over, turning the liquid back into rock. The end product could be a circular column, a rectangular beam, a trapezoidal foundation, a thin curvy roof, a giant dome. Its amazing flexibility means it can be formed into any shape; because of its huge strength, and because it lasts an extremely long time, concrete is, after water, the most-used material on the planet.
If you crush most types of rock into a powder and add water, you end up with an uninteresting sludge, the two parts don’t hold together. But something strange happens when you heat certain rocks up to really high temperatures. Take a mixture of limestone and clay, for example, and fire them in a kiln at about 1,450 degrees Celsius, and they will fuse into small lumps without melting. Grind these lumps into a very fine powder and you’ve got the first ingredient of an incredible material.
The powder is called cement. It’s a dull grey colour and might not look particularly impressive. But because it’s been burned at very high temperatures, the parent materials are chemically changed. If you add water to this powder it doesn’t turn into a sludge – instead, a reaction called hydration begins. The water reacts with the calcium and silicate molecules in the lime and clay to create crystal-like rods or fibres. These fibres give the material a jelly-like structure – a matrix – which is soft but stable. As the reaction continues, the fibres grow, and they bond to each other. The mixture becomes thicker and thicker until, ultimately, it solidifies.
So water + cement powder = cement paste. Cement paste hardens into a rock incredibly well, but it has its drawbacks. For a start, making it is expensive. The process also uses a large amount of energy. And importantly, hydration releases lots of heat. Once the chemical process finishes, the cement cools down, and as it cools it shrinks. And cracks.
Fortunately, engineers realised that cement paste binds solidly to other rocks, and began adding aggregates (small, irregular pieces of stone and sand of varying sizes) to the mixture. The aggregates help to reduce not only the amount of cement powder being used (and hence the amount of heat being released), but also the energy consumption and hence cost. The cement paste undergoes the same chemical reaction, creating fibres that in turn bind strongly to other fibres and the aggregates – and the whole mass solidifies to give us the concrete we are familiar with today. So water + cement powder + aggregate = concrete.
To make good concrete, the proportions of this mix need to be right: too much water, and not all of it will react with the cement powder – and the concrete will be weak. Too little water and all the powder doesn’t react and, again, the concrete ends up weak. For the best result, all the water needs to react with all the cement powder. And the mixing itself needs to be right too: concrete can end up poorly if it isn’t stirred properly. The larger, heavier stone aggregates settle to the bottom, leaving the fine sand and cement paste at the top, making the concrete inconsistent and weak. That’s why concrete trucks have giant rotating drums – the mixture is continually sloshed around so that the aggregates are nicely distributed throughout.
Ancient engineers didn’t have such trucks, but their formula for concrete was pretty similar to ours. They too burned limestone, and powdered it then added water to create a paste with which to bind stones, bricks and broken tiles. However, their mixture was much lumpier and thi
cker than ours is today. But then the Romans found something even better. In the land around Mount Vesuvius was an ash they called pozzolana. Instead of using burnt limestone as a cement, they tried this ready-made ash. When they mixed it with lime, rubble and water, their resulting concrete hardened as they expected. But this mixture also hardened underwater. That’s because the pozzolanic chemical reaction did not need carbon dioxide from the air to help it along: the mixture could harden without it.
To begin with, the Romans didn’t appreciate the amazing potential of the material they had made, and they only used it in small structures in a tentative way. They used it to strengthen the walls of their homes and monuments – sandwiching a layer of concrete between two layers of brick. After all, how did they know that it wouldn’t crack and crumble in a few years like plaster did? As the years passed, of course, they realised that this incredibly resilient substance was nothing like plaster, and concrete became a commonly used material. And because it solidified underwater, they could build concrete foundations for bridges in rivers, solving the problem they’d had so far in trying to cross vast stretches of water.
A Roman concrete sandwich. In Roman construction, the concrete wall was faced with a brick layer on both sides.
The Romans frequently used arches in their constructions, and concrete is a good material for arches. For one thing, it is incredibly strong. If a standard brick made from fired clay can carry the weight of five elephants, a similar brick made from relatively weak concrete can carry fifteen. In fact, a brick made from one of the stronger mixes of concrete can carry 80 elephants. And its strength can be changed, depending on the exact proportion of ingredients you add to the mix. Unlike bricks and mortar – where mortar is usually weaker than brick and more susceptible to crushing – concrete is cast monolithically (in large continuous chunks) and doesn’t have weak links in the same way: its strength is maintained uniformly across its whole body. Ultimately, of course, if the compression load is large enough, concrete will crush and crumble, but it takes a lot of load (or a good number of elephants) to get to this point.