Wind load prediction methodologies and application to full scale wind data

Date

2005-12

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The wind-induced damage to the low-rise buildings in the United States and other countries in this world is considerable. For the wind damage deduction, a reliable and economical wind load design code, it is necessary to enhance the understanding the wind load effects on the low-rise building. The wind load time series simulation and the estimation to the equivalent static wind loads are two of the major efforts to reach this goal.

Quasi-steady theory is used to investigate wind speed and area-averaged wind pressure data. Field data obtained at Texas Tech University¡¯s Wind Engineering Research Field Laboratory (WERFL) is used for this investigation. There are two variations of quasi-steady theory used in this investigation: 1) quasi-steady theory with linear assumption and the first order term of wind speed; 2) quasi-steady theory with the second order term of wind speed and the nonlinear approximation of the mean wind pressures. The quasi-steady theory is shown to adequately simulate the 15-minute time series of the net shear forces at the WERFL site. The variation of quasi-steady theory with the second order term of wind speed and the nonlinear approximation provides a better prediction than the quasi-steady theory with linear approximation.

The quasi-steady theory generally provides the out-prediction to the net shear force induced by wind. This out-prediction is attributed to the low-pass filter effect which is expressed by the aerodynamic admittance function.

The state-space model, based on the rational function approximation to the WERFL aerodynamic characteristics, is developed to simulate the wind-induced shear force time series. This state-space model, as a linear system model, predicts the RMS of wind loads precisely, and has the similar level of accuracy for the peak wind load prediction compared with the variation of quasi-steady theory with the linear approximation.

AutoRegressive eXternal (ARX) models with order 3 are shown to simulate the wind-induced shear force time series in this dissertation. The ARX model parameters are shown to be linearly related to the mean wind speed.

The equivalent static wind load (ESWL) estimation is a simple and direct approach for design purposes. It describes the dynamic wind loads in an equivalent static wind load envelope. Gust response factor (GRF) is currently widely-used in many codes and standards. Load response correlation (LRC) is an alternative which provides the ESWLs based on the correlation between the wind loads and structural response. In this dissertation, gust load envelope (GLE), is presented as an approach to estimate the ESWLs. After investigation using WERFL full-scale data, GLE provides a very simple and physically meaningful ESWL for the dynamic wind loads on the low-rise structure.

The procedures and results presented in this dissertation are useful to wind tunnel modelers, computational wind engineers and structural analysts solving nonlinear wind-induced structural dynamics problems. The quasi-steady variation, state-space model, and ARX model provide the possibility to simulate the wind loads time series. These models are useful to numerical modelers to verify the accuracy of their numerical models. The structural analysts can easily generate simulations of wind load time series using these models, and then the GLE method will give the ESWLs for their design needs.

Description

Keywords

Equivalent static wind load (ESWL), Time-series, Prediction, Wind-load

Citation