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 'Advanced drive trains technologies' taking place on Tuesday, 11 March 2014 at 16:30-18:00. The meet-the-authors will take place in the poster area.

Nemanja Komatinovic Wind Energy Institute of Tokyo, Japan
Co-authors:
Nemanja Komatinovic (1) F P Yoshitaka Totsuka (1) Hiroshi Imamura (1)
(1) Wind Energy Institute of Tokyo, Tokyo, Japan

Printer friendly version: printer.gif Print

Presenter's biography

Biographies are supplied directly by presenters at EWEA 2014 and are published here unedited

Nemanja has been working in the wind industry for almost 7 years. From 2011 he works on position of senior researcher in Wind Energy Institute of Tokyo (WEIT), Japan. His current research field covers topics such as aeroleastic code development, load calculation and dynamic response of offshore structures. He obtained Master in wind energy from Technical University of Denmark (DTU) 2006. After graduation he spent 4.5 years in GH Craft/ Teijin Innovation Composite Center where he was member of the wind turbine design team and was in charge of production /lamination of medium size wind turbine blades.

Abstract

Integration of external, customized drivetrain model with fast aeroleastic code for wind turbines

Introduction

The purpose of this study is to show method and results of the coupling between complex, manufacturer-customized model of the wind turbine planetary gearbox with FAST, an open-source aeroelastic code for horizontal axis wind turbines developed by American National Renewable Energy Laboratory (NREL). The coupled model required that original low speed shaft DOF of the FAST code be replaced with the external function drivetrain module in order to simulate more realistic response that takes into account drivetrain stiffness and damping.

Approach

In order to couple customized gearbox into the overall wind turbine aeroelastic model, modular approach was necessary, so problem was solved with the help of Matlab Simulink software. For this purpose, specially developed Matlab version of the FAST code was used together with the external pitch and torque controllers. Validity of the constructed model was verified by comparing results with the default FAST model of the same wind turbine. This default FAST model was using standard internal drivetrain properties given as an additional degree of freedom of the low speed shaft.

Main body of abstract

For design and verification of wind turbines and loads that act upon them, FAST aeroelastic code developed by American National Renewable Energy Laboratory (NREL) is being widely used. It’s open source feature gives users opportunity to modify and improve code according to their needs. In the current study, WEIT and Japanese drivetrain manufacturer have developed customized version of the FAST code that successfully couples aeroleastic model of the wind turbine with custom made function that simulates wind turbine complex drivetrain response and torque control in one. For such purpose FAST’s drivetrain DOF was turned off, so time step update of the torque and drivetrain related DOF and its derivatives is done through external function and then returned back to FAST aeroelastic module for next update. Aside from torque controller implemented within customized drivetrain function, wind turbine model also incorporates external PI pitch controller.

Both controllers are based on series of measurement data taken from real wind turbine site. This system was successfully tested by simulating wind events such as steady and step up increasing of the wind speed in operational range as well as for placing wind turbine into realistic turbulent flow according to IEC-61400-1 standard. System response of high speed shaft torque and speed were similar to the ones from default FAST model with simple drivetrain properties which indicates correct methodology.


Conclusion

NREL FAST aeroelastic code was successfully coupled with the external drivetrain model that provided customized stiffness and damping properties leading to more realistic response of the gearbox and more accurate loads acting upon the transmission and wind turbine overall. This is necessary since more accurate simulations could be used for fast and reliable load estimation in the preliminary design process of the wind turbine drivetrains as well as inputs for more realistic bench-testing.


Learning objectives
Objective of the study was to show that drivetrain manufacturers can successfully use and modify open source code such as FAST and couple it with their own in-house developed model of the drivetrain in order to simulate realistic wind turbine and drivetrain responses necessary for load calculations. This method although it can’t be used for detailed drivetrain design yields at least two advantages such as low-cost and non-time-consuming modeling.