Back to the programme printer.gif Print

Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'Innovative concepts for drive train components' taking place on Thursday, 13 March 2014 at 11:15-12:45. The meet-the-authors will take place in the poster area.

Latha Sethuraman University of Edinburgh, United Kingdom
Latha Sethuraman (1) F P Yihan Xing (2) Zhen Gao (2) Vengatesan Venugopal (1) Markus Mueller (1) Torgeir Moan (2)
(1) University of Edinburgh, Edinburgh, United Kingdom (2) Norwegian University of Science and Technology, Trondheim, Norway

Printer friendly version: printer.gif Print


A multi-body model of a direct-drive generator for a wind turbine


In recent years, a vast majority of studies have been carried out on geared drive-trains, aimed at improving the reliability of these systems, for example [1-4]. At the same time, direct-drive generators are being considered as the alternative solution to achieving higher reliability. Yet, there isn’t enough operational experience from wind turbines with direct-drive generators, to corroborate this claim. Drive-train designers are compelled to rely on experimental testing and numerical simulation techniques for making inferences on the dynamics of the drive-train. Multi-body simulation (MBS) methods are widely used in the industry for this task [5].


In this study, an integrated multi-body model of a direct-drive wind turbine generator is presented using MBS tool, SIMPACK [6].As a first step, preliminary specifications were developed for a 1DOF torsional drive-train model of a 5MW direct drive generator and its control system. The direct-drive generator considered for this study is a 5MW low speed radial flux permanent magnet generator of the interior rotor construction developed by [7]. The generator is assumed to be driven by the NREL 5MW baseline turbine [8].

Figure 1 provides an illustration of the rotor nacelle assembly for the direct drive generator. The main properties of the generator and drive specifications for the torsional 1-DOF model are summarized in Table 1.

The drive topology is similar to the commercial MTorres design [9]. The hub was assumed to be integrated to the main shaft which carries the generator rotor. The turbine-rotor and the shaft are supported by means of two roller bearings BR1 and BR2 that are housed on generator stator support structures. A detailed description of the drive-train will be presented in the final paper.

These specifications were then used to define a multi-body model of the drive-train in commercially available MBS tool SIMPACK [6]. SIMPACK allows definition of an accurate dynamic response model that includes its aero-dynamic interaction with the wind and control system, the nacelle accelerations, the kinematic behaviour of the mechanical components and the electro-mechanical interaction at the generator. The developed model was tested for a range of wind conditions for providing insight into internal drive-train responses such as shaft displacements, bearing loading, unbalanced magnetic pull (UMP). Knowledge of these features can help verify component loading and interaction, verify the adequacy of design, predict their durability and assist in the drive-train component selection and design process.

Main body of abstract

A multi-body model of direct drive wind turbine in SIMPACK:
SIMPACK is a multi-body simulation tool that allows detailed kinematic and dynamic analysis of drive-train components by integrated wind turbine simulation, incorporating, and various force and control elements. Figure 2 shows the topology of the direct-drive generator modelled in SIMPACK.

The properties of the tower and yaw control were based on the 5MW baseline system [8].The nacelle contains the main components of the drive-element, i.e. the main shaft, generator rotor and the stator. The main shaft supporting the generator rotor and the hub is modelled as a rigid body with a 6DOF joint to account for the axial, bending and torsional loads. SIMPACK's built-in library of elements was used to model the excitations (torque/force), joints and control elements .The rotor blades were connected to the hub by a user-defined kinematic joint actuated by pitch control signal. Aerodynamic loads are generated by using the AeroDyn interface in combination with Turbsim [10] and applied using force element

Control system:
The control philosophy adopted for a direct drive wind turbine is similar to the system implemented for the gear drive system [8]; however the absence of gearbox in a direct-drive wind turbine requires a high torque operation at lower speed, suggesting different dynamics for the control action. Generator speed is measured to control the blade pitch angle and generator torque by using proportional-integral velocity controller with properties tuned to achieve particular speed-torque characteristics. The wind turbine is operated according to five different control laws depending on the measured generator speed. The blade pitch angle is regulated by measuring the generator speed error using PI controller .The controller interface DISCON is a Bladed-style dynamic Link library (DLL), similar to the developed by Jonkman[8]. The controller was programmed using Fortran subroutines, compiled and linked to SIMPACK as a user-defined force-control element. For simplicity, braking action was not modelled. It was assumed that the turbine will park while the torque controller demands no torque at wind speeds above rated.

Generator reaction:
One of the most important modelling aspects for direct drive generators is the electromechanical reaction at the generator which is a function of air-gap eccentricity. Eccentricity is a measure of change in the mechanical air-gap that separates the generator rotor from the stator and is caused by several reasons such as shaft misalignment, bearing wear etc. Eccentricity results in unbalanced magnetic pull (UMP) inside the generator and is highly sensitive to bearing stiffness/shaft misalignment. This aspect will be explained in the final paper. In the SIMPACK model, kinematic sensors are used to measure the eccentricity due to shaft misalignment at every time-step and a simplified analytical model for UMP force is implemented between the stator and the rotor. This model uses a simplified linear relationship between eccentricity and UMP forces. Details of the analytical model will be presented in the final paper.

22 one-hour wind turbine simulations for the operational range of the wind turbine (4-25m/s) were carried out and the internal drive train responses were analysed. The impact of choice of the appropriate bearing stiffness on eccentricity and UMP is discussed. Figure 3 shows the time history of bearing loading and UMP at 4m/s and 12m/s wind speeds. Results of UMP, eccentricity and bearing loading for the different wind conditions will be shown in the final paper.


There is limited knowledge on the performance and reliability of direct drive generators in wind turbine drive-trains. In this study, an attempt was made to understand the dynamic behaviour of the direct-drive generator in response to various wind conditions. For this purpose, a fully integrated multi-body model of a 5MW direct drive generator for a wind turbine was developed. The drive-train model uses a radial flux permanent magnet synchronous generator of interior rotor construction. A detailed description of the model will be presented in the final paper. The NREL 5MW turbine served as the main reference system to establish the specifications of the drive-train model used in this study. The controller properties and the electromechanical reaction at the generator were incorporated to analyse the internal drive train responses.

The model was tested for the normal operating condition of the wind turbine (i.e. 4-25m/s). Results of simulations are useful to understand the sensitivities of generator response to eccentricity and provide useful information on bearing loading. The important aspects on choice of bearing stiffness for the purpose of limiting eccentricity are explained. The impact of the various wind conditions on shaft misalignment, UMP and bearing loading is understood. A general increase in UMP forces was observed with increase in wind speeds. Knowledge of these features can assist in the drive-train component selection and design process. The model may be further enhanced to understand and incorporate secondary drive-train responses such as structural deflections and also be used to predict long-term probability distribution of bearing loads and lifetime.

Learning objectives
This work aims to increase the general understanding of the dynamic behaviour of a direct-drive generator and provide details on component loading which may be important for drive-train component design.

[1] J.L.M.Peeters,D. Vandepitte and P. Sas, Analysis of Internal Drive Train Dynamics in a Wind Turbine.Wind Energy,2006, 9 (1): p. 141-161.
[2] A.R.Nejad,Y. Xing, and T. Moan, Gear Train Internal Dynamics in Large Offshore Wind Turbines, in ASME 2012 11th Biennial Conference on Engineering Systems Design and Analysis. 2-4 July,2012, Nantes, France.
[3] W. Shi, Chang-Wan Kim, Chin-Wha Chung,H.C.Park, Dynamic modeling and analysis of a wind turbine drivetrain using the torsional dynamic model. International Journal of Precision Engineering and Manufacturing, January 2013. 14(1): p. 153-159.
[4] J. Helsen, F. Vanhollebeke, F. De Coninck, D.Vandepitte,W. Desmet, Insights in wind turbine drive train dynamics gathered by validating advanced models on a newly developed 13.2 MW dynamically controlled test-rig.Mechantronics, 2011.21(4): p.737-752.
[5] B. Schlecht, T. Schulze, T. Hähnel,T. Rosenloecher. Multibody System Simulation of Offshore Wind Turbines. in EWEA 2005. Copenhagen,Denmark.
[6] Welcome to Simpack.,[Accessed, October 2013].
[7] A. Zavvos, A. S. McDonald, M. A. Mueller, Structural Optimisation tools for iron cored permanent magnet generators for large direct-drive wind turbines. Renewable Power Generation Conference (RPG 2011). 6-8 September 2011, Edinburgh, UK.
[8] J. Jonkman, S. Butterfield, W. Musial, and G. Scott. Definition of a 5-MW Reference Wind Turbine for Offshore System Development. NREL/TP-500-38060. February 2009.
[9], [Accessed July , 2013].
[10]N.Kelly,B.Jonkman,NWTC Design Codes (TurbSim),Available:,[Accessed October 2013].