Stadium rockers

Imagine you are waiting in a stadium to see your favourite band. You’ve sat through the support act and the crowd begins to buzz as the stage lights come up on your musical heroes.

As the first notes blast out from the stage, thousands of people jump to their feet and begin dancing to the music. For a moment you feel dizzy and then you realise that it’s not your balance that’s off kilter. The stand beneath your feet is starting to move. All around you people start to notice the same thing and everyone has the same thought: ‘What if this thing collapses? I’m getting out of here…’

In this situation it’s likely to be panic, rather than the threat of imminent collapse, that poses the greatest danger to life and limb. That’s why Dr Paul Reynolds and his colleagues in the Department of Civil and Structural Engineering at the University of Sheffield are looking at ways to model and thereby minimise the vibration response of packed stadia.

“Although it’s a possibility we should always consider, the biggest concern is not collapse,” he says. “We need to prevent vibrations reaching a level at which people feel unsafe. They’re likely to complain or panic well before a structure becomes dangerous.”

This sort of vibration is a hot topic in the world of structural engineering, especially in the UK. That’s partly because British engineers have a penchant for steel construction, which tends to be more lively and responsive than the heavier reinforced concrete structures favoured in some other countries.

In addition, following the Hillsborough disaster the Taylor Report recommended in 1990 that stadia should be all-seated. This means that a larger seating area is needed to accommodate the same number of people at UK venues. This has led to a lot of recent stadium development, with an increased use of cantilevered tiers, which are structurally safe but can be more prone to vibration.

Bad vibrations

Although there are various mathematical models that aim to predict how a structure will behave, the big difficulty for structural engineers and designers has been a lack of data on how real-life full-scale stands behave when loaded with people.

“If you develop a numerical model of an empty structure you’ll get some dynamic behaviour, but how will it change with the presence of a crowd and their activities?” says Dr Reynolds.

Even an empty stadium will contain lots of non-structural elements, such as glazing or furniture, which affect its dynamic behaviour. But, according to Dr Reynolds, the influence of a crowd is far more complex than that of these static additions and people don’t have to be jumping around to make a difference. The dynamic behaviour of a structure will even change depending on whether people are sitting or standing.

The challenge is to develop a simplified representation of a crowd’s effect on a structure, which can then be fed into predictive models. For example, treating the crowd as a mass suspended on a spring has provided a useful approximation so far, but the Sheffield researchers believe they can do better.

Now Dr Reynolds and his colleagues have developed a low-cost data-gathering system to monitor the behaviour of a stadium at concerts and sporting events. The initial EPSRC-funded project to devise and demonstrate the system is complete and the team is now in the main, data-gathering phase of the work, again with support from EPSRC.

The team chose the Midland Road Stand at Valley Parade, Bradford, as the venue for its first remote monitoring system installation. The system was installed for a year, during which time it recorded the behaviour of the stand during 20 football matches and nine rugby matches. It also monitored the empty stadium between events in order to gauge the response of the empty structure to ambient vibrations, such as wind or passing traffic.

The readings are taken by a combination of strategically placed accelerometers, which measure the vibrations, and CCTV cameras. The images from the cameras help match the motion in the stand to crowd behaviour at specific points in time, for example, following a goal or at half time. All the data are fed to a PC, which can communicate remotely with the lab back in Sheffield for analysis.

The team has also used a similar set up to monitor several other stadia. For example, it monitored vibrations in the City of Manchester Stadium during a pop concert by the Red Hot Chili Peppers.

In general, the tendency for problematic vibrations to build up is even greater during events where music gets everyone moving rhythmically together. The researchers need to match the crowd behaviour captured on camera to different ‘modes’ of vibration in the stand.

Just as a tuning fork rings at a particular frequency, larger structures also have characteristic ‘natural frequencies’ at which they vibrate if they are struck and then allowed to move freely. Each frequency has an associated mode, which is essentially the shape of the structure as it vibrates at that frequency.

To put it simply, the lowest mode that a guitar string will form when twanged is a curve the shape of half a sine wave. The next mode would then be a full sine wave and so on.

For large, complex structures, there are an infinite number of these natural frequencies and associated modes, but Dr Reynolds says it’s the modes with frequencies below around 6 or 8 Hertz that are of the most interest.

“The lowest frequency modes tend to be the most perceptible and the most likely to be excited by human activity,” he explains.

Data for design

The need to gather more vibration data has even attracted the attention of Government. In 2000 a joint working group drawn from the Institution of Structural Engineers, the Office of the Deputy Prime Minister and the Department for Culture, Media and Sport was set up to produce a stadium-builders ‘bible’ for dynamic assessment and design.

But the working group’s guidance so far is based on limited past research and therefore uses a crude approach that is generally accepted to need improvement. The hope is that the Sheffield research will gather enough data to find a statistically important relationship between the structure of a stand, the size and behaviour of a crowd and the resulting vibrations.

“The current guidance is inadequate,” comments Dr Reynolds. “We hope our work will lead to better models in the future.”

Article reproduced from 'Newsline' by kind permission of the EPSRC.