Interpolation of pressure time series in an aerodynamic database for low buildings

Abstract

Database-assisted design (DAD) is a relatively new concept whereby archived pressure time series from wind tunnel experiments are used directly in structural analysis software for the design of a structure. The US National Institute of Standards and Technology (NIST) is currently creating an aerodynamic database of wind-induced pressure time series on the envelope of various low buildings. In order for DAD to be of practical use, simple interpolation schemes must be used to cover the wide range of low building geometries a designer may desire since only relatively few geometries could be included in any database. The approach investigated herein consists of using a set of re-scaled pressure time series from an aerodynamically similar building at the same wind angle and in the same terrain. The set of reference time series are corrected by adjusting their mean and rms. pressure coefficients with artificial neural network models which have captured the variability of these statistical parameters. The use of the re-scaled time series is a simple way to get the correct spatial correlations for the interpolated building without resorting to more elaborate simulation schemes. However, interpolation must be limited to situations where the spatial correlations are similar. The methodology was applied to two test cases with reasonable success.