Wind-induced torsional loads on buildings

Abstract

Listen

Translate article

This paper examines the mean and peak torsional wind loads on tall buildings using two data bases of torsion measured experimentally in wind tunnel tests: the first, a large data base of mean torsional loads; and the second, a smaller data base of peak torsions. Although the mean load constitutes only a part of the total peak load required for design, it provides considerable insight into the aerodynamics of torsion, while improvement in its estimation also improves the estimation of the total peak load, using empirical gust factor methods. Comparisons between experimental results and the corresponding provisions of the 1985 National Building Code of Canada and Commentary indicate that, while the NBCC is a good estimator of mean shear loads, it significantly underestimates the mean torsional loads on tall buildings. The experimental data are further analysed to provide an improved estimation method for both the mean and the peak torsion. For mean torsion, this involves evaluating various definitions of the torsion coefficient and classifying building shapes in order to decrease the variability of the associated coefficients. This process leads to some notion of those shapes susceptible to large torsional loads and the most important building parameters on which to base predictions. This insight, along with the data base of peak torsion, is used to simplify and improve an existing method for estimating peak torsion, which was developed using a smaller data base. Key words: torsion, wind loading, codes, wind tunnel tests, tall buildings.