Optimal structural design searching algorithm for cooling towers based on typical adverse wind load patterns

Abstract

This investigation aims at typical adverse wind load patterns under tower group interference and overall structure optimization of large cooling towers considering multiple geometric parameters, focusing on potential unfavorable load patterns. Wind tunnel tests were performed on typical six-tower groups with various central distances, incoming flow directions and arrangement types, and external wind pressures were measured under interference conditions. Adverse wind load patterns are summarized with the help of interference criteria for three parameters: aerodynamic load, structure response and weighted internal force combination. Furthermore, overall structure optimization is achieved through the hybrid of grid search and gradient descent method, taking into account various possible wind loading distributions. Total 13 geometric parameters are adjusted to seek an optimized design that satisfy the requirement of structure strength, stability and economy under a particular wind load pattern. Crossover checks are then introduced to select a final optimized design scheme that is universally applicable to different adverse wind load patterns. The optimization procedure proposed in this study shows that adverse wind load patterns of grouped towers has distinct features comparing with wind load codes. It is necessary to consider the influence of multiple typical adverse wind load patterns for structural design, and the final optimization scheme should be determined through crossover checks to guarantee structure safety, stability and economy of cooling towers. • Typical adverse wind load patterns under complex six-tower group are proposed. • A hybrid optimization algorithm of the grid search and gradient descent is proposed. • Multi-parameters for cooling tower structure design are optimized. • Crossover check is used to ensure structural safety for multiple wind load patterns.