Full-scale measurements of wind-induced building motion

The majority of this thesis concerns the full-scale measurements of the dynamic response of four reinforced concrete (RC) residential buildings subject to wind loads, which were conducted to determine their dynamic structural characteristics. In addition to this, a purpose-built computer-controlled mechanical shaker (CCMS) was developed and used to induce measurable, vibratory responses of two of the subject buildings that could also help to characterize the dynamic structural properties. The subject buildings are of heights 256 m (Tower A), 236 m (Tower B), 218 m (Tower D) and 206 m (Tower E). Building responses were measured using accelerometers placed at the top of the highest occupied floor levels. Three significant wind events, Typhoon Imbudo (22-24 July 2003), Typhoon Dujuan (1-3 September 2003) and Typhoon Kompasu (14-16 July 2004), provided enough wind loading to generate a measurable response in the subject buildings. Using the peaks-over-threshold method, each typhoon event has an estimated 1-year return period. The peak resultant accelerations and peak standard deviation accelerations measured during the monitoring of each Tower fell below serviceability limits, particularly those set by ISO 6897. The responses of the Towers were significantly influenced by the arrangement of core structures within their respective planforms, which causes the response of the building to predominantly follow the stiffness axes of the core. Natural frequencies, damping values and peak factors were estimated for each Tower from the recorded accelerations. Measured natural frequencies were compared to those estimated by computational methods and empirically-based predictors. Measured frequencies were generally greater than those predicted computationally and empirically. These results support the notion that RC buildings in Hong Kong are stiffer than their international equivalents, which is consistent with previous full-scale measurements of structural dynamic properties conducted in Hong Kong. The Random Decrement Technique (RDT), justified as being the best available technique for measuring damping from ambient measurements of building response, was employed. The damping values of all Towers in the first two translational and first torsional modes of vibration fell between the range of 0.5% to 1.3% of critical, and were below 2% of critical, a value often cited by design Codes for RC buildings. Peak factors of the acceleration response recorded during the peak periods of response for all Towers during all Typhoons were similar to that for a Gaussian distribution. It was found that the responses of the subject buildings were generally attributed to turbulence buffeting. A CCMS was developed and used to induce measurable, vibratory responses of Tower A and Tower B. Mode shapes, natural frequencies and damping values for the first two translational and first torsional modes of vibration were obtained from the CCMS tests. It was found that experimentally measured mode shapes agreed well with those predicted computationally. Experimental damping values calculated by using the logarithmic decrement on free-decay responses at different storeys within Tower A and Tower B further support the relatively low levels of damping found by RDT on responses from ambient excitation