Damage Identification For Glass Curtain Wall Based On Hilbert-Huang Transform And Transmissibility Function

Glass curtain wall is a new type modern wall, which unites the architectural aesthetics, building function, building energy conservation, and architectural construction. The building presents diverse hue from different angle; and shows people dynamic beauty in the sunshine, moonlight or lighting. There are many magnificent glass curtain walls all over the world, such as World Trade Center, Amoco Building, Sears Tower, Hong Kong Bank of China Tower, Beijing Great Wall Sheraton Hotel, and Shanghai Union Building. Most of the Chinese curtain wall projects constructed in 1980s reach the expiration of their period of design. Nevertheless, there are no prevention measures for most glass curtain walls against shedding or sudden exploding, which lead to hidden safety trouble. Because of the various types of potential danger existed in current glass curtain wall structures, it is of great necessity to provide an evaluation system for the safety of the glass curtain walls. This dissertation, which is based on structural dynamic testing theory, applied the Hilbert-Huang transform signal analysis method and vibration transmissibility function damage identification index to the damage identification of glass curtain wall. The main works of this dissertation are as in the following:The basic theory of vibration transmissibility function was introduced, which was applied into damage identification for glass curtain wall. Both the vibration transmissibility function and the frequency response function can apply into damage identification for glass curtain wall, and the vibration transmissibility function is superior to the frequency response function in some degree. A novel signal analysis method, Hilbert-Huang Transform, which is suitable for processing nonlinear and non non-stationary signal and was widely applied to earthquake motion, biomedicine, and structural health monitoring etc., was introduced. The numerical simulation showed that using Hilbert-Huang Transform processing non-stationary or nonlinear signal is feasible, and it is superior to other methods. A novel dynamic damage identification method was proposed. Compared with traditional methods, this method can identify the structural damage without excitation measurement. The severity of damage can be reckoned through the damage parameters.