Damage assessment in hyperbolic cooling towers using mode shape curvature and artificial neural networks

Hyperbolic cooling towers are large thin shell reinforced concrete structures that used to remove the heat from wastewater and transfer it atmosphere using process of evaporation. During its long service life, a tower can experience damage due temperature variations, environmental degradation, or random actions such as impacts earthquakes. Such develop over time result in sudden collapse tower. To ensure operates safely efficiently at all times, is important monitor structural health. In this context, health monitoring based on vibration characteristics structure, has emerged useful method detect locate structures. towers, their particular shape, exhibit rather complex do not suit traditional vibration-based detection techniques. This paper develops applies assessment absolute changes mode shape curvature (ACMSC) conjunction with Artificial Neural Networks (ANNs) detect, locate, quantify hyperbolic towers. ANN machine learning technique predict behavioural patterns set data samples finds use quantification process. The proposed for detecting locating experimentally validated demonstrated capability accurately damage. A feed-forward network having one hidden layer Bayesian algorithm train artificial neural network. Damage indices calculated noise polluted trained then successfully assess unknown severities outcomes will enable early warning damages help towards safe operation.