Non‐Gaussian characteristics and extreme distribution of fluctuating wind pressures on large cylindrical–conical steel cooling towers

Abstract

SummaryLarge cylindrical–conical steel cooling tower (SCT) represents a new configuration of cooling tower, and its wind load distribution pattern and forming mechanism are very different from those of the traditional hyperbolic cooling towers. Large eddy simulation was used for the numerical simulation on a superlarge cylindrical–conical SCT that exceeds the specification limit, which is also the highest (189 m) SCT under construction in Asia. The surface flow field and time history of 3‐D aerodynamic force were obtained for the cylindrical and conical parts, respectively. Comparison with the measurements of other large cooling towers and the results of wind tunnel test confirmed the validity of the numerical simulation. Then, based on the probability density distribution and spatial correlation of representative measuring points, regions of non‐Gaussian distribution were identified. The forming mechanism of non‐Gaussian wind pressure distribution was revealed from the perspective of the correlation of non‐Gaussian distribution versus flow separation and eddy motion. The criteria for classifying the region of non‐Gaussian distribution for the cylindrical and conical parts were analyzed, respectively. Research shows that the wind pressures in the windward regions of conical and cylindrical parts obey Gaussian distribution; however, the wind pressures from the region of extreme negative pressure to the region of flow separation are largely non‐Gaussian and the wind pressures of the conical part are generally non‐Gaussian in the leeward region. Finally, the three algorithms for calculating the extreme values of wind pressure were used, namely, peak factor method, improved peak factor method, and Sadek–Simiu method. The distribution patterns of peak factors and extreme values of wind pressure in SCT towers were analyzed comparatively. The 2‐D formulae for fitting the extreme values of wind values for the cylindrical and conical parts were derived by nonlinear least square method. Moreover, strategy for value determination was also presented. The present research aimed to strengthen the understanding of the fluctuating wind pressure distribution and its forming mechanism for large cylindrical–conical SCT towers.