The topography and geomorphology are complex and changeable in western China, so the railway transition section is common. To investigate the aerodynamic effect of the subgrade-tunnel transition section, including a cutting-tunnel transition section, an embankment-tunnel transition section and two typical scenarios for rail infrastructures, is selected as research objects. In this paper, models of standard cutting, embankment and CRH2 high-speed train with the scale of 1:20 were established in wind tunnel tests. The wind speed profiles above the railway and the aerodynamic forces of the vehicles at different positions along the railway were measured by using Cobra probe and dynamometric balance respectively. The test results show: The influence range of cutting-tunnel transition section is larger than that of the embankment-tunnel transition section, and the maximum impact height exceeds 320mm (corresponding to 6.4m in full scale). The wind speed profile at the railway junction is greatly affected by the tunnel. Under the condition of the double track, the side force coefficient on the leeward side is negative. For embankment-tunnel transition section, the lift force coefficient of the vehicle is positive which is unsafe for operation when the vehicle is at the railway line junction.

Because of the particularity and complexity of direct air-cooling structures (ACS), wind parameters given in the general load codes are not suitable for the wind-resistant design. In order to investigate the wind loads of ACS, two 1/150 scaled three-span models were designed and fabricated, corresponding to a rigid model and an aero-elastic model, and wind tunnel tests were then carried out. The model used for testing the wind pressure distribution of the ACS was defined as the rigid model in this paper, and the stiffness of which was higher than that of the aero-elastic model. By testing the rigid model, the wind pressure distribution of the ACS model was studied, the shape coefficients of "A" shaped frame and windbreak walls, and the gust factor of the windbreak walls were determined. Through testing the aero-elastic model, the wind-induced dynamic responses of the ACS model was studied, and the wind vibration coefficients of ACS were determined based on the experimental displacement responses. The factors including wind direction angle and rotation of fan were taken into account in this test. The results indicated that the influence of running fans could be ignored in the structural design of ACS, and the wind direction angle had a certain effect on the parameters. Moreover, the shielding effect of windbreak walls induced that wind loads of the "A" shaped frame were all suction. Subsequently, based on the design formula of wind loads in accordance with the Chinese load code, the corresponding parameters were presented as a reference for wind-resistant design and wind load calculation of air-cooling structures.

In this paper a novel and efficient computational framework to estimate the stress range versus number of cycles curves experienced by a cable due to external excitations (e.g., seismic excitations, traffic and wind-induced vibrations, among others) is proposed. This study is limited to the wind-cable interaction governed by the Vortex Shedding mechanism which mainly rules cables vibrations at low amplitudes that may lead to their failure due to bending fatigue damage. The algorithm relies on a stochastic approach to account for the uncertainties in the cable properties, initial conditions, damping, and wind excitation which are the variables that govern the wind-induced vibration phenomena in cables. These uncertainties are propagated adopting Monte Carlo simulations and the concept of importance sampling, which is used to reduce significantly the computational costs when new scenarios with different probabilistic models for the uncertainties are evaluated. A high fidelity cable model is also proposed, capturing the effect of its internal wires distribution and helix angles on the cables stress. Simulation results on a 15 mm diameter high-strength steel strand reveal that not accounting for the initial conditions uncertainties or using a coarse wind speed discretization lead to an underestimation of the stress range experienced by the cable. In addition, parametric studies illustrate the computational efficiency of the algorithm at estimating new scenarios with new probabilistic models, running 3000 times faster than the base case.

Accurate numericalsimulation of wind field over complex terrain is an important prerequisite for wind resource assessment. In this study, numerical simulation of wind field over complex terrain was further carried out by taking the complex terrain around Siu Ho Wan station in Hong Kong as an example. By artificially expanding the original digital model data, Gambit and ICEM CFD software were used to create high-precision complex terrain model with high-quality meshing. The equilibrium atmospheric boundary layer simulation based on RANS turbulence model was carried out in a flat terrain domain, and the approximate inflow boundary conditions for the wind field simulation over complex terrain were established. Based on this, numerical simulations of wind field over complex terrain under different inflow wind directions were carried out. The numerical results were compared with the wind tunnel test and field measurement data for land and sea fetches. The results show that the numerical results are in good agreement with the wind tunnel data and the field measurement data which can verify the accuracy and reliability of the numerical simulation. The near ground wind field over complex terrain is complex and affected obviously by the terrain, and the wind field characteristics should be fully understood by numerical simulation when carrying out engineering application on it.

Aerodynamic configurations of bridge decks have significant effects on the aerostatic torsional divergence and flutter forsuper long-span bridges, which are onset for selection of suitable bridge decksfor those bridges. Based on a cable-stayed bridge with double main spans of 1500 m, considering typical twin-box, stiffening truss and closed-box section, which are the most commonly used form of bridge decks and assumed that the rigidity of those section is completely equivalent, are utilized to investigate the effects of aerodynamic configurations of bridge decks on aerodynamic instability performance comprised of the aerostatic torsional divergence and flutter, by means of wind tunnel tests and numerical calculations, including three-dimensional (3D) multimode flutter analysis and nonlinear aerostatic analysis. Regarding the aerostatic torsional divergence, the results obtained in this study show twin-box section is the best, closed-box section the second-best, and the stiffening truss section the worst. Regarding the flutter, the flutter stability of the twin-box section is far better than that of the stiffening truss and closed-box section. Furthermore, wind-resistance design depends on the torsional divergence for the twin-box and stiffening truss section. However, there are obvious competitive relationships between the aerostatic torsional divergence and flutter for the closed-box section. Flutter occur before aerostatic instability at initial attack angle of $+3^{\circ}$ and $0^{\circ}$, while the aerostatic torsional divergence occur before flutter at initial attack angle of $-3^{\circ}$. The twin-box section is the best in terms of both aerostatic and flutter stability among those bridge decks. Then mechanisms of aerostatic torsional divergence are revealed by tracking the cable forces synchronous with deformation of the bridge decksin the instability process. It was also found that the onset wind velocities of these bridge decks are very similar at attack angle of $-3^{\circ}$. This indicatesthat a stable triangular structure made up of the cable planes, the tower, and the bridge deck greatly improves the aerostatic stability of the structure, while the aerodynamic effects associated with the aerodynamic configurations of the bridge decks have little effects on the aerostatic stability at initial attack angle of $-3^{\circ}$. In addition, instability patterns of the bridge depend on both the initial attack angles and aerodynamic configurations of the bridge decks. This study is helpful in determining bridge decksfor super long-span bridges in future.