A cuckoo search based neural network to predict fatigue life in rotor blade composites

In modern wind turbine blades industry, fiber-reinforced composites are mostly used for their good mechanical characteristics: high stiffness, low density and long fatigue life. Wind turbine blades are constructed in different structural elements from a variety of composite laminates, often including Unidirectional (UD) material in spars and multiple forms of Multidirectional (MD) in skins and webs. The purpose of this paper is to identify materials that have appropriate fiber orientations to improve fatigue life. By using Cuckoo Search-based Neural Network (CSNN), we have developed a model to predict fatigue life under tension-tension charges for five composite materials, with different fiber stacking sequences embedded in three types of resin matrices (epoxy, polyester and vinylester), which are all appropriate for the design of wind turbine blades. In the CSNN approach used in this work, the cost function was assessed using the Mean Square Error (MSE) computed as the squared difference between the predicted values and the target values for a number of training set samples, obtained from an experimental fatigue database. The results illustrate that the CSNN can provide accurate fatigue life prediction for different MD/UD composite laminates, under different angles of fiber orientation.