Wind directionality: A reliability-based approach
Vega-Avila, Rolando E.
A methodology to reliably combine the effects of building aerodynamics and site climatology as a function of wind direction is needed to quantify the effects of wind directionality. It has been previously noted that considerations of wind directionality would result in risk-consistent, safer and more economical designs of buildings. In this doctoral exposition the author makes use of data collected at Texas Tech University to define such methodology. The West Texas Mesonet is used to define the mean and extreme climate in West Texas while the Wind Engineering Research Field Laboratory provides the aerodynamic data in representation of low-rise buildings. A novel approach to separate extremes in non-hurricane regions is presented by assuring that events are independent using atmospheric pressure data and using information from the continuous wind data sets. The aerodynamic extreme directional assessment of the low-rise building is based on estimates of pressure coefficients of building components representing the design of cladding and lateral and vertical forces representing the design of portal frames. The current standards of minimum loading of structures in the United States and Canada take into account wind directionality by stating that there is a reduced probability of the extreme winds not necessarily coming from the most aerodynamically vulnerable direction. However, no systematic reliable measure is available to-date to establish such reduced probability using extreme value distributions. In the research presented in this investigation the combination of two databases (climatic and aerodynamic) to estimate wind directionality effects corroborate the assumptions in the Standard and provide a methodology to quantify the factor in a reliable way. Results indicate that while the use of a wind directionality factor is not recommended for structural building components, if non-structural (cladding) components (which have a more pronounced directionality effect) get a discount of approximately 20% in the wind load, roughly 18% of the building population in open terrain is seeing wind loads that exceed the specified design and thus will be exposed to larger risks. This level of risk is perhaps considered acceptable but these results are based on the assumption that the code has a consistent definition of loading coefficients on the 37th percentile (or FT1 mode). Since the ASCE 7 Standard was found to possess loading coefficients with smaller percentiles (i.e. most below the 15th percentile) the risk is actually larger. Due to the large uncertainties in the wind directionality factor produced by: (1) unknown random building orientations and (2) large probabilities of exceedance in the loading coefficients specified in the standard, the true directionality issue should only be accounted through detailed analysis and not by a wind directionality reduction factor irrespective of wind direction.