Damage Detection in Wind Turbine Towers using a Finite Element Model and Discrete Wavelet Transform of Strain Signals

Abstract

Wind turbine support towers at heights in excess of 90m are nowadays being formed in steel, concrete and hybrid concrete and steel structures. As is the case for all towers of this height, the towers will be assembled using a number of segments, which will be connected in some way. These local connections are to be viewed as areas of potential local weakness in the overall tower assembly and require care in terms of design and construction. This work concentrates on identifying local damage which can occur at an interface connection by either material or bolt/tendon failure. Spatial strain patterns will be used to try to identify local damage areas around a 3 dimensional tower shell. A Finite Element (FE) model will be assembled which will describe a hybrid tower as a continuum of four-noded, two-dimensional Reisser- Mindlin shell elements. In order to simulate local damage, an element around the circumference of the tower interface will be subjected to a reduced stiffness. Strain patterns will be observed both in the undamaged and damaged states and these signals will be processed using a Discrete Wavelet Transform (DWT) algorithm to investigate if the damaged element can be identified.