The Old Tacoma Narrows Bridge has been popularized in physics text books as a classical example of resonance, but this description is misleading. It is more correct to say that it failed due to the action of self-excited forces, by an aeroelastic phenomenon known as flutter. Robert H. Scanlan, father of the field of bridge aerodynamics, wrote an article about this misunderstanding.
------------
Flutter occurs when a torsional disturbance in the structure increases the angle of attack of the bridge (that is, the angle between the wind and the bridge). The structure responds by twisting further. Eventually, the angle of attack increases to the point of stall, and the bridge begins to twist in the opposite direction. In the case of the Tacoma Narrows Bridge, this mode was negatively damped (or had positive feedback), meaning it increased in amplitude with each cycle because the wind pumped in more energy than the flexing of the structure dissipated. Eventually, the amplitude of the motion increased beyond the strength of a vital part, in this case the suspender cables. Once several cables failed, the weight of the deck transferred to the adjacent cables which broke in turn until almost all of the central deck fell into the water.
The bridge’s spectacular self-destruction is often used as an object lesson in the necessity to consider both aerodynamics and resonance effects in structural and civil engineering. However the effect which caused the destruction of the bridge should not be confused with forced resonance (as from the periodic motion induced by a group of soldiers marching in step across a bridge). In the case of the Tacoma Narrows Bridge, there was no periodic disturbance. The wind was steady at 42 mph (67 km/h). The frequency of the destructive mode, 0.2 Hz, was neither a natural mode of the isolated structure nor the frequency of blunt-body vortex shedding of the bridge at that wind speed. The event can only be understood while considering the coupled structural and aerodynamic system.