A series of wind tunnel aerodynamic experiments were conducted to investigate the wind loads on the low-rise buildings with arched roofs. The wind loadings were calculated from the simultaneous measured wind pressures data over all of the models surfaces in a simulated open terrain exposure atmospheric boundary layer. With these data the equivalent static wind loads (ESWL) of fluctuating wind pressures acting on buildings frames were evaluated and the results of equivalent static wind loads were compared with the 2006 version design codes of wind loadings on buildings in Taiwan. The design codes suggested the direction of wind acting on the low-rise buildings roofs were normal to the ridges of roofs. In this study results shown, the equivalent static wind loads caused by the approaching turbulent flow may exceed the suggested values by codes. So the peak response of structural frames should be considered carefully.

Series of numerical simulations were conducted to investigate the effects of backyard depth to the wind environment of the atrium. In the first stage of this study it was found that the original designed narrow backyard is not conducive to induce wind flow into atrium. A CFD software was adopted to simulate the wind distribution with variant backyard depth. And the result shown, if the depth of backyard might allow the wake vortex to shed from the main house roof and formatted well then the velocity in the atrium would be increase. Not only the mean velocities were checked, also the uniformity of velocity in the atrium had be inspected. We found that the deeper backyard would increase the mean velocity contemporary . In the real situation of building sites, there might not be enough space for the backyard. In this study one more aisle was put beside the main house to check its effects . And the calculation results shown, that the two ailses will improve the flow rate from atrium to backyard even in the layout of original design and the wind environments also be improved. With this results it is evident that will also good for the natural ventilation and pollutant matter dispersed.

A series of aero-elastic experiments were conducted to identify the aero-elastic parameters of 2-D square prisms under the interference effects of the tandem arranged upstream prism wake. The wake induced across-wind vibration of springs supported downstream prism was measured close to the lock-in wind speed in smooth flow. To identify the aero-elastic parameters, such as Y1,  and Y2, the trace-to-resonance method was adopted with the decay-to-resonance rules. It is well contributed to build up the predicting model of the aerodynamic loadings with the measured results of the vibration of prism under the wake interference effects with these parameters. With the aero-elastic parameters we collected in the various experiments cases, an empirical predicting model with the neural network predicting technique were constructed in this study. It is found that more clear trend of aero-elastic responses will be arose from this empirical model.

The dynamic wind loadings of solar panel mounted on flat roof building model were checked with a series of wind tunnel aerodynamic test. In this study the location and wind angle are controlled factors. The results shown that for the panel located near downstream area and its high end is facing the inclined oncoming flow might be the extreme condition, in which the panel is suffered by the highest up-lift forces. Beside the wind pressures coefficients, we have also calculated the co-spectrum of pressures signals at variant faces of panel and roof edges. With the results we found the high fluctuating and extreme wind dynamic loading on panel might come from the well correlation of building roof-top shedding vortices to the panel surface, not the separation shear layers by panel itself.