MILLENNIUM BRIDGE structural engineer Ove Arup is scouring the world for a specialist test machine that could help explain why the £18M structure is swaying far more than anticipated.
Sources close to the project told NCE that they had received reports of only two 'shaking machines' powerful enough to test the 332m long crossing.
There was still no precise knowledge of where they were located or if they were available at short notice.
One source said: 'If a machine can't be found, the only other options are dropping weights onto the deck or getting lots of people to jump up and down.'
When the bridge opened on Saturday horizontal movement on the southern span caused concern among the thousands of pedestrians who had flocked to the crossing. Arup later admitted that the lateral accelerations in particular were 'unacceptable' and the bridge was closed on Monday evening to allow tests to be carried out.
It is understood that the frequency of lateral oscillation was around 1Hz. Amplitude close to the southern shore of the Thames was measured at +/-50mm.
A shaking machine would allow Arup to carry out precise measurements of the bridge's dynamic characteristics span by span. Alternative methods, it is believed, would not yield such reliable data.
Once the source of the problem is identified, which could take several weeks of testing and analysis, the Arup team has a number of retrofit solutions available. These are believed to include active damping, which would use computer-controlled hydraulic rams to cancel out whatever frequency of swaying might develop.
This is considered to be more versatile than tuned mass dampers, which only work at one set frequency of oscillation.
Designing and manufacturing either sort of damper could take several months.
However, bridge experts consulted by NCE said that prediction of dynamic behaviour on such a long, thin lightweight structure, particularly a footbridge, was notoriously difficult.
Major bridge design consultant Maunsell's chief executive Peter Head said: 'Over a 30m40m span the effect of pedestrians on the structure is almost impossible to model. They change the natural frequency and have a marked effect on the aerodynamics of the structure.'
According to Head, wind tunnel testing, which Arup included in its design programme, is not as helpful on footbridges as it is on major road crossings. 'It's easy to put model lorries on the test model, but people behave much more unpredictably than vehicles.'
It is also hard to predict how pedestrians will react to structural movement, according to University of Sheffield structural engineering lecturer Dr Aleksandar Pavic.
'Most of the research has concentrated on the tolerance levels of stationary individuals in tall buildings, ' he pointed out. 'But we do know that acceptability thresholds go up a lot when people are walking - or if they are warned in advance that the structure is going to be lively.'
However, University of Newcastle professor of structural engineering John Knapton questioned the serviceability of the crossing. 'I'm sure it's safe, but if people are scared to cross it then we have to conclude it has failed in serviceability, ' he said.