City Research Online

Abstract

With recent construction practice, modern tall structures tend to be lightweight with low structural damping which can give rise to the structure being sensitive to dynamic excitation by the wind, with potential problems with local or overall structural fatigue failure or discomfort to occupants.

At present the UK Code of Practice for wind loading of structures provides no method for determining dynamic wind loads for building design. Codes of Practice in other countries which do present methods for determining these dynamic loads tacitly assume that the fluctuating pressures on the windward and leeward face of the building are perfectly correlated. Further, it is suggested that the spatial correlation of fluctuating pressures on the building faces are identical to the spatial correlation of atmospheric turbulence. It is also often assumed that the effects of turbulence scale on the flow around bluff bodies such as buildings is of secondary importance.There is very little experimental evidence to substantiate such assumptions and the principal objective of the study was to examine their validity and also provide data with which to validate design calculations and refine current procedures.

The experimental programme to investigate in considerable detail and provide quantitative data on the effects of turbulence scale on the cross-correlations of fluctuating pressures on and between the windward and leeward faces of rectangular section prisms was conducted in the 3m x 1.5m industrial aerodynamics wind tunnel. To isolate the effects of turbulence scale the velocity profile of the atmospheric boundary layer was not modelled. Turbulence was generate dusing bi-planar grids and longitudinal integral scales, Ly» of 58, 100 and 198 mm were achieved with an intensity of 10%. Cross-spectral analyses of fluctuating pressures on models with breadth/ depth ratios B/D of 1, 2 and 4 and an aspect ratio of 6 show that the coherence between the windward and leeward face pressures has a constant value of about 0.3 for wavelengths 1-100 times model breadth. On the windward and leeward faces the coherence data is not accurately represented by the simple exponential function y(n) = exp [-c + nr/U] but found to be a function of both nr/U and r/L.s where c is a decay constant and r is the spatial separation. For the windward face values for the decay constant which give the best fit to the data have been determined and found to be a function of separation r. For lateral separations c ranged in value from about 2 to 6 corresponding to r/B=0.2 to 0.8. These are smaller than the generally accepted values of about 6-8 which are based on measurements of spatial correlations of atmospheric turbulence. The variations of these measurements with turbulence scale are found to be significant and application of the results to design concepts is briefly discussed along with areas for future research.

Publication Type:	Thesis (Doctoral)
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) T Technology > TL Motor vehicles. Aeronautics. Astronautics

Export

Downloads