Parametric study to investigate span-wire traffic signal system performance during tropical storms

In Florida, many highway intersections use span-wire systems to support traffic signals, making their ability to withstand extreme wind events, like hurricanes, crucial for safety. These systems’ responses can be significantly influenced by installation parameters such as wire sag and tension, which vary across intersections. The goal of this research was to better understand how these factors impact the system’s wind response. To achieve this, an experimental program was developed to measure the effects of catenary cable sag, messenger cable pretension, and the location of cable end supports using load cells, accelerometers, and inclinometers. These data were then used to calibrate numerical models to assess the system’s response with more refined setup parameter modifications. Experimental results showed a strong correlation between installation parameters and the system’s performance under wind loading. For instance, increasing the value of catenary sag (from 5% to 7%) reduced drag forces and the root mean square (RMS) of accelerations, giving the system a more advantageous aerodynamic response to wind forces. Numerical models for long- and short-span-with-springs models were developed to quantify and evaluate span-wire assemblies, comparing them with full-scale experimental results. First, long- and short-span-with-springs model results (i.e., total mean drag, total mean lift forces, mean inclinations, and wire deflections) were compared and were similar. This proved that the short-span-with-springs model can be used to get comparable results to long spans. The short-span with-springs model can simulate various span lengths found in the field by adjusting the longitudinal spring stiffness that corresponds to a particular lateral flexibility, normal to the plane of cables. In addition, results obtained experimentally from long- and short-span full-scale assemblies were compared to each other and to their corresponding model results and were similar as well. Similarly, the short-span full-scale assembly was installed with coil springs at both ends of the catenary and messenger cables. This facilitated the simulation of various span lengths by attaining the same lateral stiffness properties of span lengths found in the field. Comparison of experimental and model results served as a validation for the numerical models. These models helped to assess forces, inclinations, and wire deflections undergone by span-wire systems. A theoretical buffeting analysis of a span-wire traffic signal system was also performed. This analysis evaluated the signal assembly’s response to fluctuating wind speeds, providing RMS, mean, and peak values for acceleration and deflection in the along-wind direction. This analysis was conducted for a one 5-section and two 3-section configuration span-wire traffic signal assembly and carried out for winds that are perpendicular to the frontal area of the traffic signal system. The mass and frontal area of this system was taken as one lump mass and area. These analytical results were compared to the full-scale experimental data, providing a good agreement between experimental results and numerically obtained estimations. However, there were some deviations between analytical and experimental results. For instance, experimental values for rms of acceleration and deflection are a bit lower than the analytical values in the lower wind speeds. This is because some cross-wind response took place at lower wind speeds during testing of the assembly which decreased some response in the along-wind direction. Buffeting analysis does not reproduce the effect of cross-wind response.