In June 1996 I submitted a tender for a footbridge spanning the North Woolwich Road to the Thames Barrier Park.
Our entry consisted of a very flat suspension bridge, with the deck supported by brackets between the suspension cables.
The geometry of the suspension cables was similar to that of the Millennium Bridge, in that they were deflected in plan as well as in elevation. This geometric arrangement of cables gives wind resistance but, as we discovered, has some side effects.
The 'carrying force' of the suspension cables has a horizontal component, putting the brackets into tension.
Under symmetrical vertical loading these horizontal forces are equal and opposite.
However, when the deck is loaded with an eccentric live load, the forces in the two cable systems become different and consequently the horizontal components become unequal. If the deck is not cross braced to resist this horizontal uniformly distributed load, or if it is cross braced but is flexible, the cables must change geometry to find equilibrium. The less loaded cable must increase its horizontal sag, and the more loaded cable decrease its horizontal sag, for the horizontal forces to become balanced again.
Thus the statical logic of this cable arrangement is that a horizontally flexible deck moves sideways under any eccentricity of loading.
Calculations show that such a cable system is very sensitive to eccentric loading. Only a small proportion of the design live load placed eccentrically produces horizontal deflections of the order of several centimetres. This is not surprising as such flat suspension bridges are very flexible, with static vertical deflections due to full pedestrian loading on a 140m span being measured in metres rather than centimetres.
One would expect there to be a significant horizontal component to the dynamic behaviour of the bridge. This would be exacerbated if the cables on either side of the bridge could oscillate out of phase.
Robert Benaim (F), email@example.com