Dynamic Characteristics and Wind-induced Response of a Tall Building

Abstract

The design of tall buildings requires an accurate understanding of the expected wind loads and the resulting responses. The techniques used to estimate the wind-induced response are subject to uncertainty, which can result in unsatisfactory building performance or an over-designed structure. Altering the structure to rectify unsatisfactory performance can be extremely difficult and prohibitively expensive, while an over-designed structure represents unnecessary cost to the owner. This implies that accurate estimates of wind loads and responses are crucial to tall building design. Two aspects of tall building wind-induced response estimation are investigated: the estimation of natural frequencies and damping ratios; and the understanding of mechanisms causing wind-induced responses. This was primarily conducted via full-scale testing of a tall building. The building used for full-scale measurements is Latitude tower, an office tower located in the Sydney central business district, with a height of 187m above ground and 28m of underground levels. The building has a composite design including a reinforced concrete core, and reinforced concrete floor slabs supported by steel beams spanning between the core and perimeter columns. Outriggers linking the core and perimeter columns, as well as offset outriggers at the facade, are located at mid-height. The full-scale testing was conducted in two parts: vibration testing during construction; and a two year monitoring programme commenced after construction completion. Vibration testing during construction was conducted to determine the natural frequencies and damping ratios as the structure changed. Forced vibration testing and ambient vibration testing techniques were used. The Frequency Domain Decomposition and Stochastic Subspace Identification techniques were used to estimate the natural frequencies and damping ratios from the ambient vibration test outputs. The natural frequencies and damping ratios from the forced and ambient vibration tests differed by less than 5% and 30% respectively. Changes in the fundamental natural frequencies during construction were discussed in conjunction with the structural changes to further the understanding of how changes in the stiffness and mass of a tall building influence the natural frequencies. The measured natural frequencies during the early stages of construction were used to update a finite element model representing the structure at the time of testing. The material properties and floor beams were the primary focus of the model updating. The knowledge gained from partial structure updating was applied to a model of the completed structure, and the natural frequency estimate errors improved from 17% to 7%. The fundamental mode damping ratios measured during construction changed by less than 15% between the first test, conducted when 38% of the tower height was reached, and the final test at construction completion. The wind-induced monitoring programme included the measurement of wind velocities, accelerations, and displacements at the top of the building. The peak events for southerly and westerly wind directions were discussed. It was found that the acceleration response was dominated by the fundamental vibration mode. For southerly winds this corresponded to an along-wind response, but for westerly winds this corresponds to a cross-wind response. The probability distributions of upcrossings for along-wind and cross-wind responses where not significantly different to a Gaussian distribution for both southerly and westerly winds. The slope of the linear least squares fit was greater than two in all cases, which suggested intermittent characteristics were present in the responses. The standard deviation resonant acceleration responses from a high frequency base balance wind tunnel test were within 29% of the measured values.