Conditional simulation of non-Gaussian wind velocity profiles: Application to buffeting response of Vincent Thomas suspension bridge

For a proper prediction of wind-induced vibrations of a long suspension bridge, it is necessary to use a representative wind velocity profile at the bridge site. However, as the simultaneous collection of wind velocity data at closely spaced locations of the entire bridge is not a viable option, simulation has become a powerful tool for this purpose. In this study, a new iterative approach of non-Gaussian conditional simulation is proposed to conditionally simulate the wind velocity profiles by utilizing the measured wind velocities at a few locations. The approach utilizes the well-known spectral representation technique in conjunction with nonlinear Gaussian to non-Gaussian mapping technique. Focusing on Vincent Thomas suspension bridge on which three anemometers have been installed recently at three strategic locations, this study compares buffeting responses evaluated using different simulation schemes. It has been found that the bridge response evaluated using the non-Gaussian simulation scheme may be higher when compared with other simulation schemes.

► A new iterative approach of non-Gaussian conditional simulation is proposed.
► Conditionally simulate wind velocity profiles by using measured wind velocity.
► The spectral representation method is used for nonlinear Gaussian to non-Gaussian mapping.
► Buffeting response of Vincent Thomas suspension bridge is evaluated and compared.