Some aspects of the behaviour in wind of large grouped chimneys

Large reinforced concrete chimneys are commonly associated with Reynolds Numbers of the order 10⁷ or higher. These values are three orders of magnitude greater than those typical of tests in boundary layer wind tunnels and an order of magnitude or more greater than that attainable in large sub-sonic tunnels such as that at Monash. If controlled wind tunnel tests cannot match the Reynolds Numbers and these large values of R(e) remain beyond the reach of computational fluid dynamics then methods of prediction based upon CFD or model testing need to be supplemented by full-scale observations. In the case of a single large chimney, the data from a variety of full-scale studies has shown that the fluctuating loads associated with vortex shedding are strongly influenced by turbulence and even at values of R(e) in excess of 10⁷, the magnitude of these forces can be more than doubled by small scale turbulence. For multiple stacks, full-scale data is extremely limited and predictions of the behaviour in groups must be viewed cautiously. The present paper deals with the study of two groups of large (of the order of 200m high) chimneys in which there were wind-induced failures due to across-wind vibration. The paper first describes the two cases and discusses both the structural mechanics and the fluid mechanics associated with each case. In both studies, the chimneys were instrumented and full-scale data gathered over a period of months. The methods of analysis of the field data are described and the data demonstrates significant magnification due to an upstream structure and also by comparatively minor departures from a circular cross-section.