Background
Wind turbines are huge structures that are subjected to highly dynamic loads arising from the complex interplay between the varying wind field and wake from other turbines, the rotation of the rotor as well as the transmission and the electric generator in the drivetrain inside the nacelle. An important consequence of this is that most components in a wind turbine are experiencing highly dynamic loads resulting in wear or fatigue of the components. This can lead to premature failure of single components or even more severe breakdowns of larger parts of the wind turbine. Obviously, both wear and fatigue can result in considerable economic costs. Many of the reasons for these problems originate from how the components were manufactured in the first place. Typical manufacturing processes include casting of the rotor hub, housings and frames and possibly also the main shaft as well as forging of main bearings and gear parts. These processes inherently give rise to non-uniform internal structure and associated properties, including residual stress distributions and consequently affect the performance of components during use. For heavily loaded parts such as gear wheels it might be necessary to apply further processing such as heat treatment in order to obtain a wear resistant surface. The research centre addresses the entire chain “materials-processes-components-loading-performance” in a systematic way with emphasis on heavily loaded wind turbine components in order to be able to understand, describe and predict the components’ performance during service better than what is possible today, thus paving the road for increased reliability and durability of the entire wind turbine.
Main deliverables of the center
It is expected that the main scientific deliverables of the center will be within the following areas:
- A detailed picture of the failure mechanisms in bearings and gears in relation to manufacturing history and loading history during service. An evaluation of the potential for surface engineering as a means of improved performance of bearings and gears as well as a numerical tool for modeling the relation between surface engineering parameters and microstructure.
- Establishing the basis for material reliability and life time criteria, and to evolve the background knowledge of the materials performance. This is done by combining mechanical testing with mesomechanic, life-time and damage modeling and investigation of the microstructural features, defects and residual stresses.
- Improved numerical models for the simulation of the manufacturing of heavily loaded metallic wind turbine parts. This will encompass models for processes such as casting and forging with focus on prediction of the component’s final mechanical properties and residual stress state.
- Better understanding of fatigue failure evolution under rolling or sliding contacts, obtained through numerical modeling and experimental observations. This will result in modeling tools that can be applied to increase the life time of the mechanical components in gears and bearings.
- Improved electromechanical drivetrain loads simulation software along with validation documents. Conceptual design of the mechanical drivetrains. Improved knowledge on failure mechanisms and reliability of the mechanical components of the drivetrain.
- Improved probabilistic models for defects and damage accumulation for critical components and reliability assessment incorporating information from different types of tests and inspection during manufacturing and operation. Reliability-based decision support tools for optimizing cost of energy.
Work Packages
The center comprises six work packages:
WP 1 Materials engineering and failure analysis
The components of the drivetrain most frequently observed to fail are gears and bearings. Currently, gears are manufactured from carburized steel, i.e. the carbon concentration close to the surface is enhanced at elevated temperature, and subsequently quenched and tempered, providing a hardened phase to improve the wear and fatigue resistance, particularly sliding contact fatigue, of these components. Bearings are manufactured from classical bearing steel (100 Cr6), hardened and tempered at low temperatures to obtain wear and fatigue performance, particularly rolling contact fatigue.
WP 2 Mechanical properties and damage mechanics
The mechanical properties of differently processed materials will be investigated. Both static, fatigue, and fracture mechanics methods will be investigated under uni- and multiaxial loading conditions. In order to provide data required in WP3, material properties will also be investigated at elevated temperatures. This is essential for an accurate prediction of the stress distribution and stress accumulation in relation to the processing history of the components.
WP 3 Process modelling
Models for all the considered manufacturing processes, such as casting, forging and heat treatment, will be established. The overall objective is, based on input from WP1 and WP2, to model the multi-physics behaviour of the manufacturing processes to such an extent that the resulting state in terms of for example local distributions of mechanical properties and residual stresses can be described and used for the downstream simulations in WP4, WP5 and WP6.
WP4 Fatigue and wear in rolling and sliding contacts
Fatigue under moving contact loading, as can occur in gear teeth or in bearings, is a complex phenomenon, which relates to a number of important applications. Among these applications are fatigue (pitting failure) in roller bearings and gear contacts or fatigue in railway tracks. Of considerable importance is the subject of fretting fatigue and sub-surface failure, e.g. initiating as “butterflies” around inclusions. For gears also tooth root failure is very important, but this failure mechanism is related to the bending moment loading of the tooth.
WP5 System simulation and in-service loads
This WP will focus on the electro-mechanical interaction between the rotor and drivetrain loads and the responses of different components of the drivetrain, along with experimental testing to develop new conceptual designs of the drivetrain with improved reliability. The models will include the effect of aerodynamics and controls along with the gearbox stages and respective bearings. A suitable electrical model of the generator completes the drivetrain system. A gearbox unit will be tested under dynamic torsional load conditions to validate the software model. This will enable the prediction of the interactions between the different components and tailor the structural behavior of each component for safe and reliable conceptual designs of the whole drivetrain.
WP 6 Reliability
Based on the mathematical models developed in WPs 1-5 and test results in WP5 probabilistic models will be developed and techniques for estimating the probability of failure developed considering damage tolerant design principles and system models. The stochastic modeling will be based on theoretical considerations and the quality control in the manufacturing process and from various types of tests: coupon, subcomponent, full-scale, down-scaled and accelerated tests performed in other WPs and available in literature and other research projects.