Non-stationary Analysis Oil Field Measured Wind Data And Wind-induced Responses Of Tall Buildings

With the rapid development of the world economy and construction technology,many tall buildings emerge all over the world. The application of light weight andhigh strength materials produce many tall buildings with more flexible and lowdamping characteristics, which make these tall buildings more sensitive to the wind.Field measurement is a direct and effective method for evaluating the wind effect onsuch tall buildings. Taking Guangzhou West Tower as an example, field measurementon the wind effect on this super tall building was conducted. The comprehensiveanalysis on the wind-envirmnent and wind-induced response on this tall building wasstudied based on the field measured signals. The main content of this research studymainly includes the follwoing:As the field measured wind speed and wind-induced response behavenon-stationary characteristics, an effective method for analyzing non-stationaryrandom signal was first proposed in this study. The new deveopled Hilbert HuangTransform (HHT) method, which mainly includes Empirical Mode Decomposition(EMD) and Hilbert Tranform algorithms and is very suitable for analzingnon-stationary data, was introduced in this study.The algorithm and numerical processtechnique of HHT was described in great detail in Chapter2. A simulated signal wasadopted for specifying the whole processing of HHT mehtod, its effect on describingtime-frequency characteristics of non-stationary data was discussed.Secondly, taking Guangzhou West Tower as an example, a structural healthmonitoring programm was established for this super tall building. The wind speed,wind direction, wind-induced acceleration and wind-induced dispalcement weremeasured synchronously atop of this super tall building under moderate monsoonwind. The mean wind speed and direction, turbulence intensity, gust factor, powerspectral density and turbulent integral scale for fluctuating wind speed were firstanayzed by ordinally vector decomposition method. In order to capture thenon-statioany characteristics of measured wind speed signal, the EMD method wasadopted for analzying the same record. The time-varying mean wind speed wasobtained by EMD algorithm. Comparing with the vector decomposition method,analyzed results from EMD method could obtain the inherent information aboutmeasured wind records. The probability density distribution and power sprectral density of flunctauting wind speed obtained from EMD method also verify thevalidity and effectiveness of the proposed EMD method. The concept of time-varyingturbulence intensity was then introduced in this study. In the second part of thischapter, strong wind data near the sea and ground level was measured during typhoonPaiBiAn passing Maoming. The EMD method was used for analyzing this measurednon-stationaty wind data to obtain the wind characteristic of strong typhoon. Thetime-varying mean wind speed, turbulence intensity and power spectral density offlunctuating typhoon wind in both time and frequency domain were ontained andcomapred with those analyzed results for monsoon wind measured atop of GuangzhouWest Tower.In the final part of this study, time-frequency analysis was conducted on themeasured wind-induced acceleration and displacement response of Guangzhou WestTower. The EMD method was first adopted to extract the IMF components whichwere related with wind-induced response in the specified vibration modes in eachmain axis of tall building. The Random Decrement Technique (RDT) was utilized toobtain the free decay curve of free vibration response for the specified mode. HilbertTransform was then used for the free vibration response to obtain the amplitude andphase functions with each time step. The curve fitting of the logarithm of amplitudeand phase curve were conducted by least square method to obtain the naturalfrequencies and damping ratios for the first vibration mode. Meanwhiletime-frequency analysis was also conducted for the measured wind-induced responseby the Hilbert spectrum in HHT method. The participtation of each vibration mode inwind-induced response could be expressed in both time and frequency domain. Thenon-stationary characteristics of wind-induced displaceemnt could be describedclearly by the proposed HHT method in this study.