Directional wind characteristics analysis in the mountainous area based on field measurement

Abstract

Wind characteristics are one of the most important factors when designing long-span bridges. In a typical deep-cutting gorge, the wind field is influenced by local terrain and shows dependence on wind direction according to preliminary numerical simulations. Field measurement is then employed to evaluate the impact of wind direction further. Instead of the common method, which splits samples into sectors by referring to the canyon direction, a qualitative computation approach is adopted. Specifically, marginal probability distribution functions (PDFs) of various wind characteristic parameters are explored initially, followed by binary and ternary joint PDFs associated with wind direction using the modified Bernstein copula. The directional wind characteristics are then examined utilizing conditional PDFs and mathematical expectations for all wind directions. The results show that both the distribution and mean value of wind parameters in severe winds exhibit a substantial association with wind direction. The statistical properties change significantly and cannot be captured using the common method when airflows enter from downstream of the canyon and are blocked by the ridge near the bridge. When designing long-span bridges sensitive to wind action, the impact of wind direction should be carefully factored, especially when local terrain blocks the frequent incoming flow.