Aerodynamic Loads on Tall Buildings: Interactive Database

Under the action of wind, tall buildings oscillate simultaneously in the alongwind, acrosswind, and torsional directions. While the alongwind loads have been successfully treated using quasi-steady and strip theories in terms of gust loading factors, the acrosswind and torsional loads cannot be treated in this manner, since these loads cannot be related in a straightforward manner to the fluctuations in the approach flow. Accordingly, most current codes and standards provide little guidance for the acrosswind and torsional response. To fill this gap, a preliminary, interactive database of aerodynamic loads is presented, which can be accessed by any user with Microsoft Explorer at the URL address http://www.nd.edu/;nathaz/. The database is comprised of high-frequency base balance measurements on a host of isolated tall building models. Combined with the analysis procedure provided, the nondimensional aerodynamic loads can be used to compute the wind-induced response of tall buildings. The influence of key parameters, such as the side ratio, aspect ratio, and turbulence characteristics for rectangular sections, is also discussed. The database and analysis procedure are viable candidates for possible inclusion as a design guide in the next generation of codes and standards. DOI: 10.1061/~ASCE!0733-9445~2003!129:3~394! CE Database keywords: Aerodynamics; Wind loads; Wind tunnels; Databases; Random vibration; Buildings, high-rise; Turbulence. 394 / JOURNAL OF STRUCTURAL ENGINEERING © ASCE / MARCH 2003 tic model tests are presently used as routine tools in commercial design practice. However, considering the cost and lead time needed for wind tunnel testing, a simplified procedure would be desirable in the preliminary design stages, allowing early assessment of the structural resistance, evaluation of architectural or structural changes, or assessment of the need for detailed wind tunnel tests. Two kinds of wind tunnel-based procedures have been introduced in some of the existing codes and standards to treat the acrosswind and torsional response. The first is an empirical expression for the wind-induced acceleration, such as that found in the National Building Code of Canada ~NBCC! ~NRCC 1996!, while the second is an aerodynamic-load-based procedure such as those in Australian Standard ~AS 1989! and the Architectural Institute of Japan ~AIJ! Recommendations ~AIJ 1996!. The latter approach offers more flexibility as the aerodynamic load provided can be used to determine the response of any structure having generally the same architectural features and turbulence environment of the tested model, regardless of its structural characteristics. Such flexibility is made possible through the use of well-established wind-induced response analysis procedures. Meanwhile, there are some databases involving isolated, generic building shapes available in the literature ~e.g., Kareem 1988; Choi and Kanda 1993; Marukawa et al. 1992!, which can be expanded using HFBB tests. For example, a number of commercial wind tunnel facilities have accumulated data of actual buildings in their natural surroundings, which may be used to supplement the overall loading database. Though such HFBB data has been collected, it has not been assimilated and made accessible to the worldwide community, to fully realize its potential. Fortunately, the Internet now provides the opportunity to pool and archive the international stores of wind tunnel data. This paper takes the first step in that direction by introducing an interactive database of aerodynamic loads obtained from HFBB measurements on a host of isolated tall building models, accessible to the worldwide Internet community via Microsoft Explorer at the URL address http://www.nd.edu/;nathaz. Through the use of this interactive portal, users can select the Engineer, Malouf Engineering International, Inc., 275 W. Campbell Rd., Suite 611, Richardson, TX 75080; Fomerly, Research Associate, NatHaz Modeling Laboratory, Dept. of Civil Engineering and Geological Sciences, Univ. of Notre Dame, Notre Dame, IN 46556. E-mail: [email protected] Graduate Student, NatHaz Modeling Laboratory, Dept. of Civil Engineering and Geological Sciences, Univ. of Notre Dame, Notre Dame, IN 46556. E-mail: [email protected] Robert M. Moran Professor, Dept. of Civil Engineering and Geological Sciences, Univ. of Notre Dame, Notre Dame, IN 46556. E-mail: [email protected]. Note. Associate Editor: Bogusz Bienkiewicz. Discussion open until August 1, 2003. Separate discussions must be submitted for individual papers. To extend the closing date by one month, a written request must be filed with the ASCE Managing Editor. The manuscript for this paper was submitted for review and possible publication on April 24, 2001; approved on December 11, 2001. This paper is part of the Journal of Structural Engineering, Vol. 129, No. 3, March 1, 2003. ©ASCE, ISSN 0733-9445/2003/3-394–404/$18.00. Introduction Under the action of wind, typical tall buildings oscillate simultaneously in the alongwind, acrosswind, and torsional directions. It has been recognized that for many high-rise buildings the acrosswind and torsional response may exceed the alongwind response in terms of both serviceability and survivability designs ~e.g., Kareem 1985!. Nevertheless, most existing codes and standards provide only procedures for the alongwind response and provide little guidance for the critical acrosswind and torsional responses. This is partially attributed to the fact that the acrosswind and torsional responses, unlike the alongwind, result mainly from the aerodynamic pressure fluctuations in the separated shear layers and wake flow fields, which have prevented, to date, any acceptable direct analytical relation to the oncoming wind velocity fluctuations. Further, higher-order relationships may exist that are beyond the scope of the current discussion ~Gurley et al. 2001!. Wind tunnel measurements have thus served as an effective alternative for determining acrosswind and torsional loads. For example, the high-frequency base balance ~HFBB! and aeroelasgeometry and dimensions of a model building, from the available choices, and specify an urban or suburban condition. Upon doing so, the aerodynamic load spectra for the alongwind, acrosswind, or torsional response is displayed along with a Java interface that permits users to specify a reduced frequency of interest and automatically obtain the corresponding spectral value. When coupled with the concise analysis procedure, discussion, and example provided, the database provides a comprehensive tool for computation of the wind-induced response of tall buildings. Wind-Induced Response Analysis Procedure Using the aerodynamic base bending moment or base torque as the input, the wind-induced response of a building can be computed using random vibration analysis by assuming idealized structural mode shapes, e.g., linear, and considering the special relationship between the aerodynamic moments and the generalized wind loads ~e.g., Tschanz and Davenport 1983; Zhou et al. 2002!. This conventional approach yields only approximate estimates of the mode-generalized torsional moments and potential inaccuracies in the lateral loads if the sway mode shapes of the structure deviate significantly from the linear assumption. As a result, this procedure often requires the additional step of mode shape corrections to adjust the measured spectrum weighted by a linear mode shape to the true mode shape ~Vickery et al. 1985; Boggs and Peterka 1989; Zhou et al. 2002!. However, instead of utilizing conventional generalized wind loads, a base-bendingmoment-based procedure is suggested here for evaluating equivalent static wind loads and response. As discussed in Zhou et al. ~2002!, the influence of nonideal mode shapes is rather negligible for base bending moments, as opposed to other quantities like base shear or generalized wind loads. As a result, base bending moments can be used directly, presenting a computationally efficient scheme, averting the need for mode shape correction and directly accommodating nonideal mode shapes. Application of this procedure for the alongwind response has proven effective in recasting the traditional gust loading factor approach in a new format ~Zhou et al. 1999; Zhou and Kareem 2001!. The procedure can be conveniently adapted to the acrosswind and torsional response ~Boggs and Peterka 1989; Kareem and Zhou 2003!. It should be noted that the response estimation based on the aerodynamic database is not advocated for acrosswind response calculations in situations where the reduced frequency is equal to or slightly less than the Strouhal number ~Simiu and Scanlan 1996; Kijewski et al. 2001!. In such cases, the possibility of negative aerodynamic damping, a manifestation of motion-induced effects, may cause the computed results to be inaccurate ~Kareem 1982!. Assuming a stationary Gaussian process, the expected maximum base bending moment response in the alongwind or acrosswind directions or the base torque response can be expressed in the following form: