Lead Session Chair:
Stephan Barth, Managing Director, ForWind - Center for Wind Energy Research, Germany
Andre Lass (1) F P Arjun Kancharla (1) Jitendra Kumar (1) Frank-Hendrik Wurm (1)
(1) University of Rostock, Rostock, Germany
Printer friendly version: Print
Download poster(1.79 MB)
Presenter's biographyBiographies are supplied directly by presenters at OFFSHORE 2015 and are published here unedited
Mr. Laß is currently an associate research scientist at the Institute of Turbomachinery at the University of Rostock. He studied structural and fluid mechanical engineering in Rostock. Along the way he took an internship and sideline job dealing with structural mechanics at the research and development department of Suzlon, a major wind turbine company. His present research is focused on modelling the rotor dynamic behavior in diverse energy domains for rotating machines using bond graph methodology.
Unified approach for modelling of wind turbine power train using bond graph
Wind turbine is a huge, nonlinear and multidisciplinary system (mechanics, aeronautics, electrical and magnetic) that deals with various energy domains to extract electricity from wind. A unified approach for robust multi-physics modelling is needed for understanding, design, optimization, evaluation and control of wind turbine. Specialized tools have been developed in last few decades to model different physical domains individually, like finite element method (FEM) for structural analysis and computational fluid dynamics (CFD) for aerodynamic analysis. Coupling of CFD and FEM solver is achieved in various software codes but computational effort is not in acceptable range for industry applications.
Bond graph methodology is used to achieve a robust and reliable tool, which is capable of describing the dynamic behaviour of rotating components in diverse energy domains for industry applications at low computational effort. It empowers the engineer to model power exchange, energy dissipation and storage in dynamic systems of any physical domain with a unified graphical language. Coupling of dynamic of system, fatigue analysis, power optimization, damage prediction for structural components and electric system can be evaluated.
Main body of abstract
The wind turbine power train consist of several important components like blades, hub, shaft, planetary gearbox, bearings and generator. Between these components power is exchanged. Every element is attached to next element by half arrows called as power bonds considering causality of system. The blade is divided in seven flexible beam elements having six degree of freedom. Steady state or transient fluid forces can be mapped from CFD to blade nodes. Complete shaft is divided into a number of small elements and each element is modelled as Rayleigh beam. Every element is having five degree of freedom including gyroscopic effect, spinning and gravity. Main bearing is considered as ball bearing where stiffness and damping is considered in terms of geometrical descriptions like number of balls, diameter of balls, diameter of inner and outer race and many more. Gear box is modelled as a transfer function including moment of inertia of gear. Squirrel cage generator is modelled considering sinusoidal distribution of stator winding throughout the air gap. Magnetic losses, air gap and core losses are considered while modelling. Aerodynamic forces are calculated from CFD and mapped on blade nodes which creates torque and thrust on hub and deform the blade. In the integrated bond graph model torque causes rotation of shaft, which is balanced by main bearing. Rotation is multiplied by gear box to rotate generator rotor shaft which leads to an electric output.
Blade deformation, shaft vibration amplitude, power outputs and losses are calculated during transient simulation. Modal analysis is performed to extract Eigen frequency of system. The developed parametric and generic model of wind turbine can be used for any size of wind turbine to predict life and real time behaviour of different components for various boundary conditions. This integrated generic bond graph model will reduce pre-processing and new model creation time while development of new wind turbine system. Therefore it could reduce research and development cost of wind turbine power train effectively.
Possibility of physical modelling of diverse energy domains under one platform using bond graph methodology is prime learning objective of the research. It reduces modelling and simulation time for huge system. It has capability to model non-linearity of system to evaluate dynamics and rotor-dynamics for industrial applications.