Conductor gallop is the high-amplitude, low-frequency oscillation ofoverhead power lines due to wind.[1] The movement of the wires occurs most commonly in the vertical plane, although horizontal or rotational motion is also possible. The natural frequency mode tends to be around 1 Hz, leading the often graceful periodic motion to also be known asconductor dancing.[2][3] The oscillations can exhibit amplitudes in excess of a metre, and the displacement is sometimes sufficient for thephase conductors to infringe operating clearances (coming too close to other objects), and causingflashover.[4] The forceful motion also adds significantly to the loading stress oninsulators andelectricity pylons, raising the risk of mechanical failure of either.
The mechanisms that initiate gallop are not always clear, though it is thought to be often caused by asymmetric conductoraerodynamics due to ice build up on one side of a wire. The crescent of encrusted ice approximates anaerofoil, altering the normally round profile of the wire and increasing the tendency to oscillate.[3]
Gallop can be a significant problem fortransmission system operators, particularly where lines cross open, windswept country and are at risk to ice loading. If gallop is likely to be a concern, designers can employ smooth-faced conductors, whose improved icing and aerodynamic characteristics reduce the motion.[4] Additionally, anti-gallop devices may be mounted to the line to convert the lateral motion to a less damaging twisting one. Increasing thetension in the line and adopting more rigid insulator attachments have the effect of reducing galloping motion. These measures can be costly, are often impractical after the line has been constructed, and can increase the tendency for the line to exhibit high frequency oscillations.[5]
If ice loading is suspected, it may be possible to increase power transfer on the line, and so raise its temperature byJoule heating, melting the ice.[3] The sudden loss of ice from a line can result in a phenomenon called "jump", in which thecatenary dramatically rebounds upwards in response to the change in weight.[1][2] If the risk of trip is high, the operator may elect to pre-emptively switch out the line in a controlled manner rather than face an unexpected fault. The risk of mechanical failure of the line remains.[6]
The earliest studies of long wires embedded in a moving fluid motion dates to the late 19th century, whenVincenc Strouhal explained "singing" wires in terms ofvortex shedding.[7][8] Gallop is now known to arise from a different physical phenomenon:aerodynamic lift. Ice accumulated on the wire destroys thecircular symmetry of the wire, and the natural up-and-down "singing" motion of a wire changes theangle of attack of the iced wire in the wind. For certain shapes, the variation in lift across the different angles is so large that it excites large-scale oscillations.[9]
Mathematically, an unloaded extended wire in dead air can be approximated as a massm suspended at heighty by aspring withconstantk. If the wind moves with velocityU, then it makes angleα with the wire, where
At large wind velocities, the lift anddrag induced on the wire are proportional to the square of the wind velocity, but the proportionality constantsCL andCD (for a noncircular wire) depend onα:
whereρ is the fluid density andl the length of the wire.[10]
In principle, the excited oscillation can take three forms: rotation of the wire, horizontal sway, or vertical plunge. Most gallops combine rotation with at least one of the other two forms. For algebraic simplicity, this article will analyze a conductor only experiencing plunge (and not rotation); a similar treatment can address other dynamics. From geometrical considerations, the vertical component of the force must be
keeping only terms first-order in the regimeẏ ≪ U.[10] Gallop occurs whenever thedriving coefficient1/2ρlU · (CD + ∂CL/∂α)|α = 0 exceeds the naturaldamping of the wire; in particular, anecessary-but-not-sufficient condition is thatThis is known as theDen Hartog gallop condition, after the engineer who first discovered it.[9][10]
At low wind velocitiesU, the above analysis begins to fail, because the gallop oscillation couples to thevortex shedding.[10]
A similar aeolian phenomenon isflutter, caused byvortices on theleeward side of the wire, and which is distinguished from gallop by its high-frequency (10 Hz), low-amplitude motion.[2][3] To control flutter, transmission lines may be fitted withtuned mass dampers (known asStockbridge dampers) clamped to the wires close to the towers.[5] The use of bundle conductor spacers can also be of benefit.