A New Generation of Tools for the Design of Tall Buildings Subjected to Wind Loads

Current approaches to the estimation of wind-induced wind effects on tall buildings are based largely on 1970s and 1980s technology, and were shown to result in some cases in errors of up to 40%. Improvements are needed in: (i) the description of direction-dependent aerodynamics; (ii) the description of the direction-dependent extreme wind climate; (iii) the estimation of inertial wind effects induced by fluctuating aerodynamic forces acting on the entire building envelope; (iv) the estimation of uncertainties inherent in the wind effects; and (v) the use of applied wind forces, calculated inertial forces, and uncertainty estimates, to obtain via influence coefficients accurate and risk-consistent estimates of wind-induced internal forces or demand-to-capacity ratios for any individual structural member. Methods used in current wind engineering practice are especially deficient when the distribution of the wind loads over the building surface and their effects at levels other than the building base are not known, as is the case when measurements are obtained by the High-Frequency Force Balance method, particularly in the presence of aerodynamic interference effects due to neighboring buildings. The paper describes a procedure that makes it possible to estimate wind-induced internal forces and demand-to-capacity ratios in any individual member by: developing aerodynamic and wind climatological data sets, as well as aerodynamic/climatological directional interaction models; significantly improving the quality of the design via rigorous structural engineering methods made possible by modern computational resources; and properly accounting for knowledge uncertainties. The paper covers estimates of wind effects required for allowable stress design, wherein knowledge uncertainties pertaining to the parameters that determine the wind loading are not considered, as well as estimates required for strength design, in which these uncertainties need to be accounted for explicitly.