Study On Newmerical Simulation Of Wind-Load On Structures

Wind load is a kind of important loads to the buildings, and it is indispensable in the structure design. With the appearance of long-span roofs, high-rise buildings, long-span bridges, wind load plays a more and more important role in engineering, and it may become one of the control loads for structure design. So it becomes very important to measure the distribution of the wind pressure on the surface of the buildings.At present, field measurement, wind tunnel test and numerical simulation are the three major methods to study the distribution of the wind pressure on the buildings. The wind tunnel test is the most common one of them. Using atmosphere boundary layer wind tunnel test to study the wind actions on the structures in civil engineering began from the 50's. It has a history of more than fifty years. The numerical simulation was brought forward with the development of Computational Fluid Dynamics and the computer science in the 80'th, only twenty years ago. But it became the hotspot in the research of wind engineering since it was born. When compared with the traditional method in wind engineering it had its own unique advantages.This paper studied and analysed the mean surface wind pressure on the high-rise buildings and long-span roofs with complex bodily form, the mean and RMS wind pressure on a simple bodily form long-span flat roof. Methods used here were wind tunnel test and numerical simulation. The key points in the paper are the realization of the numerical simulation in the surface wind pressure evaluation on buildings, discussions on the feasibility and reliability of using this technology in the engineering. Finally shortages and the problems need be solved of this technology were pointed out.As objects of thre numerical simulation of their mean wind pressure, a typicalhigh-rise building and a long-span roof structure were chosen. The full-scaled rigid model of these structures were constructed for numerical simulation, and grids for the fluid zones were generated properly. The turbulent model used here was Realizable k-ε and non-equilibrium Wall Function which was used to solve the flow in the near wall regions. Through the numerical simulation work above, we got the distribution of the mean wind pressure on these buildings. Comparing these results with the wind tunnel test data, we found that they agreed very well with each other. For the long-span roof, we also calculated the block wind pressure coefficients using the area integral method. Comparing the block wind coefficients used in engineering to the corresponding wind tunnel test results, the two groups data got much more closer than the comparing of the mean wind pressure coefficients between measuring points.In unsteady numerical simulation, the time history of the wind pressure on a flat long-span roof was obtained using the LES method. The auto-regressive model was adopted to generate the time history of pulse wind speed. The divergence-free operation and adjustment of mass flux operation were imposed on the time series data from the AR method. The ultimate time series of pulse wind speed satisfied the target spectrum, spacial correlations and continuity condition. After these operations, the time series of pulse wind speed was imposed on the inflow boundary of the computational zone. Through the unsteady computation we got the time history of wind pressures on the whole flat roof. After the time statistic operations were imposed on the wind pressure data, the mean and RMS wind pressure were obtained. Their agreement with the corresponding wind tunnel test data proved the validity of the whole method include the generation of pulse wind speed ,the unsteady computation and the post process. All the work above might be the reference for other unsteady computation of wind pressure on more complex bodily form structures under more accurate inflow boundary conditions.