Cable-stayed bridges are an efficient and elegant solution to bridging long spans. Though a widely used structural system, several significant issues are still unresolved regarding a cable-stayed bridge's performance, particularly with respect to the response of the cables to wind.

As tension members, stay cables have a very low diameter-to-span ratio such that the effect of bending stiffness on a cable's vibration characteristics is inconsequential. Coupled with this, stay cables have very little internal mechanical damping, thus they are quite susceptible to dynamic excitation. Several vibration mechanisms have been identified, however for some, including the potentially more damaging mechanisms, the required conditions and generating mechanisms are not completely understood.

Aerodynamically, a stay cable is simply a circular cylinder exposed to wind. A circular cylinder positioned normal to airflow is a touchstone topic in the fields of aerodynamics and fluid mechanics. A circular cylinder inclined to airflow, from a fundamental viewpoint, has been given very little attention considering the body of work concerning the former case.

The objective of this study was to determine the characteristics of the airflow about an inclined circular cylinder, and to determine and clarify necessary conditions for the generation of galloping vibrations and high-speed vortex vibrations in dry, inclined stay cables. A wind tunnel study was conducted in which a static circular cylinder was exposed, at varying relative wind-cylinder angles, to wind speeds corresponding to a Reynolds number range of 1 × 10⁵ to 6 × 10⁵ Surface pressure data was recorded to determine both the instantaneous and time-averaged pressures and forces on the section. Using the matrix computation software Matlab, programs were written to evaluate and analyse lift and drag forces, and pressure, lift, and drag coefficients. From the extensive data compiled, several significant conclusions concerning flow about an inclined or yawed cylinder and about the galloping and high-speed vortex vibration phenomena result.

Concerning flow about an inclined circular cylinder, the drag force coefficient is generally lower for lower relative inclination angles. This trend reverses over a small range of speeds within the critical Reynolds number range. Note that this and the following findings are based on the minimum inclination angle tested of 54.7°

The rate of reduction in the drag force coefficient through the critical Reynolds number range is larger for larger relative inclination angles. The drag coefficient values converge during the drag crisis, regardless of the relative inclination angle.

The net lift force that appears during the drag crisis initiates at a lower wind speed for lower relative wind-cylinder angles. The upper bound at which this net lift force disappears does not seem to be affected by inclination angle. The magnitude of this net lift force is smaller for cylinders inclined at smaller inclination angles.

As with a circular cylinder positioned normal to the flow, a Strouhal number of about 0.2 can be used to predict the Karman vortex shedding frequency for a cylinder inclined to the flow, however the inclination angle must also be considered in addition to the wind speed and cylinder diameter.

The results of this study indicate that the Glauert-Den Hartog criterion-long established as a necessary condition for galloping of bluff bodies-is applicable in predicting the conditions for the galloping of dry, inclined cables. Conditions necessary for this galloping mechanism include, but are not limited to, a relative wind-cable angle of 60°, with the airflow in the critical Reynolds number range. This form of galloping appears to be velocity-restricted and is potentially related to the flow regime characterized by the formation of a single-laminar separation bubble. A significant span-wise correlation in the dynamic lift coefficient is evident under the above conditions.

An estimate of mechanical damping required to prevent galloping vibrations was evaluated. The resulting required damping is significantly higher than levels present in typical existing stay cables, indicating susceptibility of existing prototype cables to galloping vibrations.

No indication of a relatively low frequency mechanism was found in this test to assist in further explaining the limited-amplitude, high-speed vortex excitation witnessed in a previous dynamic model wind tunnel test. As evidence of this mechanism is restricted to wind tunnel tests of dynamic models, perhaps the formation of axial vortices—thought to be the cause of this phenomenon—is dependent upon motion of the cable or model. Future wind tunnel testing concerning this phenomenon using dynamic models should include particular attention paid to scaling parameters, including Reynolds number and Strouhal number.