Analysis of Wind Turbine Pitch 4-Point Contact Bearing

Abstract

Abstract In the field of renewable energy source, the wind power is certainly one of the most important and therefore the industrial and research community are devoting increasing attention to the design, construction, management and maintenance of wind turbines. Due to the relevance, technical, economic and social of this type of installation the correct maintenance is actually a crucial challenge. As in all mechanical engineering applications, numeric models, FEM based, are widely used during both the design step and monitoring on site. In fact, in the frame of Structural Health Monitoring (SHM), the availability of a Digital Twin is fundamental to interpret the signals coming from the sensors and to check the state of wear and the progression of any damage. In general, this numerical tool must have a good accuracy and a great rapidity of response. For this reason it is essential to develop condensed and reliable models of the main elements able to represent the real behaviour of the component without excessively increasing of computational costs. One of the most important element is the pitch variation slewing bearing, which allows the optimal wing orientation. Unfortunately, a complete numeric model of the bearing requires very large dimension and so the real time reply, necessary to SHM, is difficult. In this paper, to overcome this problem, a new simplified model is proposed to represent the array of spheres and its contact condition with the rings of the bearing. The new approach is based on the adoption of a toroid-shaped element replacing the rolling elements. To study this solution, a full-detailed model of a typical slewing bearing for pitch variation has been used to properly set up the contact model and to tune the properties, geometrical and elastic, of the equivalent toroidal body. The main feature of this approach is that, with a relatively low calculation effort, the deformation behaviour and the stress state of the elements connected to the bearing is available in an easy and accurate way. Then, the comparison in real time of the numerical results with the corresponding experimental ones from the site allows to deduce initiation and growth of a damaging process.