Back to the programme printer.gif Print

Wednesday, 12 March 2014
16:30 - 18:00 Floating wind turbines
Science & Research  

Room: Llevant
Session description

The session covers design problems related to floating wind turbines and how current research is overcoming these hurdles through innovative platform concepts, experimental methods and mooring system analysis. In particular, technical and economic studies and analyses for three different novel concepts will be presented, including one vertical axis concept, a concrete platform design and a combined wind & wave energy device. In addition, a new methodology for experimental model testing with a focus on aerodynamics and control will be presented, as well as a detailed assessment on long-term mooring system loads.

Learning objectives

  • Learn about a novel experimental methodology for floating wind turbine aerodynamic and controller testing
  • Identify challenges and benefits of an innovative vertical axis concept
  • Understand the design and potential benefits and challenges of a concrete platform
  • Assess structural fatigue damage of a multi-modal wind/wave energy device
  • Learn about a new methodology capable of reproducing life cycle mooring loads
Lead Session Chair:
Denis Matha, University of Stuttgart, Germany

Antoine Peiffer, Marine Innovation and Technology, United States
Ilmas Bayati Politecnico di Milano, Italy
Ilmas Bayati (1) F P Marco Belloli (1) Davide Ferrari (1) Hermes Giberti (1) Fabio Fossati (1)
(1) Politecnico di Milano, Milan, Italy

Printer friendly version: printer.gif Print

Presenter's biography

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

Ilmas Bayati obtained the MSc in Mechanical Engineering from Politecnico di Milano with a thesis titled “Design and validation of a Hardware-In-The-Loop experimental rig for wind tunnel tests on offshore wind turbines”. He has been awarded the prize “Bonfiglioli 2011” for the best italian thesis in the field of mechatronics. He's now a PhD candidate at the same university focusing on the hydro-aero-dynamics of wind turbines and sailing yachts, numerically and experimentally.He has been recently visiting the offshore team of NREL, working on the 2nd order hydrodinamics of Floating Offshore Wind Turbines


Wind tunnel tests on floating offshore wind turbines: design of a 6-dof robotic platform for floating motion simulation


Large-scale offshore floating wind turbines represent one of the most significant engineering challenge in wind energy at present. The tecnology is now moving toward deeper waters. Aero-servo-hydro-elastic design tools have been recently developed and compared in order to define reliable design guidelines and load cases. These codes need experimental validation in controlled environment such as wind/water basins and wind tunnels. This paper presents a task-focused 6 degrees-of-freedom robot design for providing consistent floating motions to a floating wind turbine scale model for wind tunnel tests, with the goal of testing models under a highly controlled wind flow.


Large-scale offshore floating wind turbines represent one of the most significant engineering challenge in wind energy at present. Since current fixed-bottom technology has seen limited deployment to water depths of nearly 30m (shallow waters), the tecnology is now moving toward deeper waters, where the wind resource is extremely abundant. In this regards floating wind turbine platforms are now considered for multi-megawatt wind turbines deployment [1]. A remarkable variety of floating platform concept is recently taken into account, such as tension-leg-platform, spar-bouy or barges and the dynamics of such structures are studied throughout hydro-aero integrated analysis [2]. Since the possible combinations between different wind turbine and floating platform concepts are definitley wide, a reference 5-MW wind turbine for floating offshore applications was developed by NREL and the related structural properties freely accessible [3]. This baseline wind turbine is currently taken into account for aero-hydro-servo-elastic codes comparison, for example under IEA tasks [4,5], in order to have reliable design tools for floating wind turbines [6] also for harsh load cases [7], being the operating environment of such machines potentially damaging. The reliability of such numerical codes demands for experimental validations. According to authors’ knowledge, floating wind turbine model tests were performed only within offshore water/wind basins[8],[9] so far.
This paper aims at proposing the above mentioned experimental validation within a large wind tunnel for civil-environmental applications, with a reliable control of the wind turbulence intesity [10], where the floating motion of the model is given by a custom high-performance 6 degrees-of-freedom (DOF) hexapod, capable of performing specific motion laws along such DOF or through “hardware-in-the-loop” functioning mode, so that running the hydro-platform dynamics in real-time and providing consistent displacements to the scale model (fully instrumented). This approach to floating wind turbine model testing could be integrated to water basin tests data in order to provide more accurate response regarding the aero-structure interaction and to investigate the nature of specific aerodynamic loads due to certain motion of the floater.

Main body of abstract

In order to define the hexapod requirements, in terms of maximal displacements and frequencies along the 6 degrees-of-freedom (surge, sway, heave, roll, pitch, yaw), the extreme event analysis by Jonkman [11], regarding the floating platform motion, was considered as a design reference and whose outputs consistently scaled due to the scale factor of the model.
A parallel kinematics robot type was chosen for this aim, instead of the more commercially-available serial ones, because of the need of heavily customizing the architecture and geometry in order to match the given requirements. Therefore this type of robot can be greatly task-oriented, whereas its design-phase complex and not a-priori defined. In this paper the approach leading to the drive system sizing is also reported.
This robotic device is not a simple static positioner, but a dynamic hardware-in-the-loop motion simulator, having dynamic but also geometric constraints: since such a device must be placed inside the civil-environmental test chamber of the Politecnico di Milano wind tunnel, some machine size constraints were taken into account. Among others, the most critical one concerns with the vertical dimension of the robot due to the height-limits of the test section and of the the flat-ground beneath whom all the experimental intrumentation and the robot itself must to be hidden, in order to prevent aerodynamic disturbances on the model. Limiting the height of the robot is also suitable for testing bigger models, reducing aerodynamic scale effects.
This defined requirements were transposed into the desired workspace. The tool-center-point of the robot is set to coincide with the bottom of the model, accordingly with position and orientation of the reference frame in [11]. The experimental set-up is placed on the wind tunnel turn table, that increases by one "static" large-amplitude rotational degree of freedom the whole system (wind direction not perfectly co-alingned with wave train).
The choice of parallel-kinematics robot is related to its load capacity and dynamic performance. Because of the links simultaneously connect the base of the robot to the mobile platform, loads tend to be distributed uniformly, engaging the links along their greater stiffness direction. Furthermore, the actuators occupy a position very close to the robot’s base, so that the “mobile mass” that is accelerated, in addiction to the model mass, is reduced to minimum; whereas in serial robots also intermediate actuators represent mobile masses.
These requirements led to a 6-PUS kinematic topology manipulator with parallel actuated prismatic joints, known as Hexaglide [12]. This architecture is characterized by a predominant direction of its workspace, which can be aligned with the direction of the wind, and has the capability to “stay-low” along the vertical direction, accordingly with requirements.
The parameterized dimensions of the manipulator are determined through a multi-objective optimization campaign realized with a genetic algorithm [13]. The objectives to be achieved are concerned with: (1) the coverage of the workspace, (2) the static forces multiplication, (3) the longitudinal size, (4) the interference between the links, and (5) the interference between the links and the rails. The construction of appropriate cost functions, which implements these objectives, is discussed. This approach produced several Pareto-optimal geometrical solutions which are kinetostatically analyzed and compared. Then, also the dynamic analysis was performed in order to define the actuating forces, the inverse-dynamics of the robot was solved by multi-body approach by considering reasonably inertial properties and constant-frequency sinusoidal motion laws along the whole set of degrees-of-freedom, independently. This procedure was performed starting from various static positions belonging to the workspaces and led to a ditailed map of available forces, velocities and accelerations within the workspace itself. Different Pareto-optimal solutions were compared and the mostly-matching system was selected, chosing from two main different technologies: screw-driven and belt-driven units. The inertial and frictional contributions of the linear transmissions were also added to compute the total resistance torque. Then new distribution maps were produced using the alpha-and-beta theory for checking the motor-reducer unit [14], where beta is the so called load-factor, representing the required performances. In the end, the motor-reducer unit was chosen and the links were sized for yielding and buckling resistance. The results were used for the sizing process of the components and to choose the best compromise solution.


This work represent a new tool for testing scale models of floating wind turbines. As the experimental validation of aero-hydro-servo-elastic numerical codes is requested, a 6 degrees-of-freedom robot platform was designed with the goal of simulating the floating motions of real floating wind turbines. The goal of this project is to have the ability of testing floating wind turbine scale models within a civil-envirnoment wind tunnel facility with the aim of pay greater attention to the aerodinamics of such muchines and to the highly controlled quality of the wind flow acting on the model. The choice of an hexapod robot is consistent with the need of investigating the aerodynamics on the blades, tower and structures due to specific motions, or combinations, of the floating system. This robot will be able to reproduce certain motion laws along the 6 degrees-of-freedom or providing standalone consistent displacements to the base of the tower, operating on a “hardware-in-the-loop” mode. The design of such a motion simulator was developed with particular attention to the requirements, in terms of displacements and frequencies, given by previous experimental and numerical studies on floating wind turbines and consistently adopted due to the model dimensions (scale factor) that the target wind tunnel facility (Politecnico di Milano) is capable to contain. The parallel-kinematic hexaglide robot family was chosen for its capability of matching the kinematic/dynamic requirements. Different Pareto-optimal solutions were compared and the mostly-matching system was selected, chosing from two main different technologies: screw-driven and belt-driven units. The motor-reducer unit was chosen and the links were sized for yielding and buckling resistance.

Learning objectives
This work represents a new way of testing scale models of floating wind turbines with a greater attention to the aerodinamics of the structures due to motion of the floating machine. This approach will allow designers to validate numerical codes and implement new formulations based on measured data under highly controlled wind flow.

[1]S.Butterfield, W. Musial, J.Jonkman, P.Sclavounos, Engineering Challenges For Floating Offshore Wind Turbines, NREL/CP-500-38776, Sept.2007.
[2]J. Jonkman, D. Matha, Dynamics Of Offshore Floating Wind Turbines Analyisi Of Three Concepts, Wind Energy, Volume 14, Issue 4, pages 557–569, May 2011.
[3]J. Jonkman, S. Butterfield, W. Musial, G.Scott, Definition Of a 5MW Reference Wind Turbine For Offshore System Development, NREL/TP-500-38060, Feb 2009.
[4]P.Passon,M. Kuhn, S.Butterfield, J.Jonkman, T.Camp, T.J, Larsen, OC3 Benchmark Exercise Of Aero-Elastic Offshore Wind Turbine Codes, The Science of Making Torque from Wind, Aug.28–31, 2007.
[5]F. Vorpahl, M.Strobel, J.Jonkman, T.Larsen, P.Passon, J.Nichols, Verification Of Aero Elastic Offshore Wind Turbine Design Codes Under IEA Wind Task XXIII, Wind Energy, DOI: 10.1002/we.1588.
[6]A.Cordle, J. Jonkman, State Of The Art In Floating Wind Turbine Design Tools, 21st International Offshore and Polar Engineering Conference, June 19 – 24, 2011.
[7]Nguyen_Manuel_Jonkman_Veers_Simulation Of Thunderstorm Downbursts And Associated Wind Turbine Loads_SOLARENERGY
[9]H.H.Nguyen, L.Manuel, J.Jonkman, P.S. Veers, Simulation Of Thunderstorm Downbursts And Associated Wind Turbine Loads, Journal of Solar Energy Engineering, Vol. 135, Issue 2, January 25, 2013.
[10]G.Diana, S.De Ponte, M. Falco, A.Zasso, A new Large Wind Tunnel For Civil Environmental And Aeronauitcal Application, Journal of Wind Enegineering and Industrial Aerodynamics, 74-76 (1998), 553-565.
[11]J.M. Jonkman,Dynamics Modeling and Loads Analysis Of an Offshore Floating Wind Turbine, NREL/TP-500-41958, Nov. 2007.
[12] Honegger, M., Codourey, A., Burdet, E., Adaptive Control of the Hexaglide, a 6-DOF Parallel Manipulator, Proceedings of IEEE Int. Conference on Robotics and Automation, ETH - Zürich. 1997.
[13] Kalyanmoy Deb, Multi-Objective Optimization using Evolutionary Algorithms, Wiley, 2001.
[14] Giberti, H., Cinquemani, S., Legnani, G. A Practical Approach to the Selection of the Motor-Reducer Unit in Electric Drive Systems. In: Mechanics Based Design of Structures and Machines, vol. 39(3) (2011), pp. 303 - 319.