On 1 July this year, only a few weeks after the highly publicised sway problems on London's Millennium Bridge, something remarkably similar happened in Ottawa.
Huge crowds who flocked to the city's Alexandra Bridge to watch a Canada Day fireworks display were alarmed when the structure began to sway violently from side to side.
But this was no new, high tech crossing. The Alexandra Bridge is a 100 year old steel truss structure with a main span of 320m.
Originally it carried rail and tramway tracks and these days serves as a road and pedestrian crossing.
'This seems to confirm that what we have is a human behaviour phenomenon, not a structural problem, ' says Sheffield University structural dynamics expert Dr Aleksander Pavic. 'All the bridges where this lateral sway problem has been reported are very different structurally. The only common denominator is that there were large crowds of people moving across them.'
He adds: 'When this happens it appears there is a strong possibility they will start to march in phase with the bridge's lateral motion and make it worse.'
Pavic and his Sheffield team contributed to the emergency research programme launched by Arup immediately after the problem was identified. Other teams at Southampton University and Imperial College also sought to pin down the 'lock-in' phenomenon, as Arup dubbed it. This is the moment when significant numbers of pedestrians begin to march in step.
What the research was attempting to do, explains Arup director Michael Willford, was produce hard data for the design team to base its retrofit proposals on. Some of the results were surprising, he adds.
'It turned out that pedestrians are very sensitive to even small amounts of lateral movement and will start to lock-in very early - although some of those who try don't get it right, ' he says. 'If we do damp out all significant lateral movements there's a risk that pedestrians will start to lock into vertical motions instead.'
Arup now has a very good idea how the lateral forces from pedestrians change as the amplitude of motion increases, Willford says. But the university research has not been able to simulate true crowd behaviour, hence the trials with large numbers of pedestrians on the bridge proper.
Pavic believes the Arupsponsored research is a major breakthrough that will have an effect on more than just footbridge codes. Some modern grandstands are also flexible enough to trigger lock-in, he says.
'What we didn't know before is how the human-induced forces on slender structures change when humans can perceive the motion. Now we do.'