Machine and rotor

Aerodynamics

The rotor blades of a wind turbine are exposed to massive loads. Frequent changes in the inflow cause highly fluctuating forces that can lead to material fatigue. Using numerical simulations and wind tunnel measurements, ForWind investigates the forces acting on the rotor blades due to various flows.

Numerical simulation

Computational Fluid Dynamic (CFD) methods are predominantly used to investigate flows. Basic interactions between the turbulent inflow and the wing are calculated. In order to improve time-resolved calculations, ForWind also uses new methods for this purpose.

Wind tunnel measurements

In the wind tunnel at the University of Oldenburg, ForWind measures lift and drag forces on wing segments with Reynolds numbers up to 700,000 (approx. 50m/s). The forces acting on the sash are determined without contact by pressure sensors. These measurements can be made both at fixed and rapidly varying angles of attack of the wing to the flow. In particular, the measurements under varying angles of attack exhibit brief increases in force, which are referred to as “dynamic stall”. Corresponding models to the Dynamic Stall can be created and verified based on these measurements. The accuracy of the sensors means that changes in drag due to different surface qualities of the wing can be detected during measurements at fixed angles of attack.

Furthermore, the wind tunnel offers the possibility of various measurements with open as well as closed measuring section. For example, hot-wire anemometry or laser Doppler anemometry (LDA) are available as measurement methods.

Aerolastics

Wind turbines are exposed to high dynamic loads. The precise calculation of these loads plays a central role in the design of wind turbine components. Fluctuating loads result from the interaction of aerodynamic forces and the elastic structure. Special methods, such as CFD (Computational Fluid Dynamics) and FEM (Finite Element Method), are used to look at these loads in detail. Fluid-structure coupling can also be studied using analytical models (“engineering models”) for both aerodynamic and structural behavior of the system. Analytical models play a crucial role in the aeroelastic and aerodynamic design of a wind turbine due to low-cost and resource-efficient numerical implementation.

The Institute of Turbomachinery and Fluid Dynamics at Leibniz Universität Hannover (TFD) accompanies the first development steps of wind turbines with BEM (Blade Element Momentum Theory) methods, dynamic stall models and turbulent wind simulators, which are developed in-house.

Simulations with CFD also enable accurate aerodynamic modeling of wind turbines. Turbulent simulations of wind turbines are used to calculate structural loads and to compute unsteady aerodynamic phenomena such as rotor-support structure interactions, wake dynamics, and dynamic separation. With the help of various CFD solvers as well as high and ultra-high performance computers, the institute conducts a wide range of scientific investigations.

Dynamic performance characteristics

Determining the performance characteristics of a wind turbine is essential for both certification and turbine monitoring. A power characteristic is generally understood to be the functional relationship between the speed of the wind acting on a turbine and the corresponding power output.

ForWind is working to develop rapid methods for determining power characteristics based on high-resolution wind and power data (order of magnitude 1 Hz) that require measurements of only a few days. Within the framework of a stochastic performance modeling, the reaction dynamics of the plant in connection with the turbulent structures of the incoming wind are described as a diffusion process. A mathematical analysis procedure (based on Markov processes and their description by a Langevin or Fokker-Planck equation) allows the reconstruction of the effective dynamic equation governing this process.

The method forms the basis for the determination of a so-called dynamic performance curve. This has already proven to be a very efficient alternative to the power curve according to the standard procedure (IEC 61400-12-1) and has been tested on several commercial wind turbines of the MW class.

Wind turbine regulation

Wind turbines (WTGs) are much more powerful today than they were just a few years ago. Consequently, high mechanical loads occur, such as high peak torques and dynamic load change influences, which are transmitted to the entire driveline. As a direct connection between the rotor and the generator-inverter system, the drive train must absorb both the dynamics of the wind loads via the rotor and the dynamics of the electrical grid via the generator.

This necessitates detailed investigations into the steady-state and dynamic loading of the entire drive train. For reliable predictions on the service life of individual components and thus on the service life of the entire plant, the interactions of the mechanical and electrical components must be taken into account.

The Institute for Electrical Drives, Power Electronics and Components (IALB), University of Bremen, uses a detailed wind turbine model, which includes a multi-body simulation of the drive train (MBS) in addition to the simulation of various electrical interrelationships, as a basis for this. Based on this, the scientists have developed an online observation system for recording variables that cannot be measured directly, which allows, among other things, the determination of the shaft torsional moments. Knowledge of the shaft torsional moments is a decisive criterion for the development of a minimum-load control system.

For direct experimental investigation, the drive train of a wind turbine is simulated in a test and demonstration rig at the IALB, which can be used to verify simulated findings on a real system. In addition, the test rig serves as the basis for realizing the online observer system for minimum-load control.

Rotor modeling

The rotor of a wind turbine is subjected to strong shaking forces due to the effect of the turbulent wind field. These dynamic alternating loads can lead to faster material fatigue.

Aeroelastic simulation programs can be used to calculate the forces acting on the rotor and localize the resulting loads. For this purpose, each rotor blade is divided into a number of aerodynamically independent elements. In addition, the velocity and position of the individual blade elements and the incoming wind field are determined for each time step. The acting forces can then be calculated per element and add up to the total forces.

The goal of this research area is to integrate current results from the fields of stochastic wind field modeling as well as dynamic stall and lift augmentation into the modeling of the rotor and thus improve the calculation of dynamic loads. In particular, the occurring loads of realistic, turbulent wind fields are calculated and compared with the results of the standard method (IEC standard). A main interest is the time history of the load distributions and the alternating loads caused by the gustiness of the winds.

Projects on the research topic “Machine and Rotor

Research focus

Development of a maximum-torque synchronous generator

The development of an ultra-high-torque synchronous generator with transverse flux offers a novel alternative for wind turbines (WTGs). Compared to a high-pole synchronous generator or a geared double-fed asynchronous generator, this allows the power density of a wind energy generator to be increased with significantly smaller external dimensions. Compared to conventional generator types, weight and volume can be reduced by about two thirds. There are material savings in copper, dynamo sheet and, above all, in the steel structure of the tower and the foundation. This allows the use of smaller nacelles or larger generators.

With a transverse flux generator (TFG) as a wind energy generator, there are fundamentally new degrees of freedom for the design of high-pole generators with a small diameter. In a TFG, the magnetic flux is guided transverse to the direction of motion. By using a high-pole machine with concentrated power density at a relatively small diameter, as required for very high-power WTGs, the power-to-weight ratio is decisively reduced.

Geometry measurements on tower and rotor blades

Current developments of new wind turbines (WTGs) tend towards ever larger turbines with less stiff rotor blades. The simulation tools required for design must be validated by appropriate measurement data. Furthermore, geometric measurements are helpful to evaluate the condition of existing turbines with respect to possible pitch angle deviations between the rotor blades or possible structural damage. From this, needs-based maintenance can be derived or decisions can be made regarding continued operation.

However, the detection of geometric features or vibrations on the components of a wind turbine presents a challenge for measurement technology, especially when the turbine is in operation. Established methods either only allow point measurements or require the attachment of markers or similar. The Bremen Institute for Metrology, Automation and Quality Science (BIMAQ) at the University of Bremen has developed measurement methods that can record ground-based measurements with a laser scanner (principle of time-of-flight measurement) of the surface geometry of wind turbine components without contact and without manipulating the wind turbine. From the measurement data, for example, the pitch angles of the rotor blades, the tower clearance under different wind loads, or deformations of the rotor blades under load can be measured. On this basis, the comparisons of the measured data of the rotor blades can be used to optimize the efficiency of the turbine or to detect structural damage at an early stage.

Power flow optimization of a wind turbine

Wind fields do not offer a wind turbine (WT) a constant wind power over time, but a wind power interspersed with short-term fluctuations in the range of seconds. In the turbine concepts realized so far, the speed is only slowly adjusted to the average wind speed, so that it does not follow these short-term fluctuations. Pitch control systems equalize the loads and torques. Inevitably, however, the maximum possible power yield is not achieved, since at peaks in wind speed the rotor blades have to be adjusted.

Power output can be maximized by taking advantage of these short-term peaks. To keep the loads low, the generator speeds are more variable than in conventional systems. In this way, the rotating masses additionally act as an energy store. ForWind is investigating how all of the energy storage inherent in the wind turbine system can be used to provide the most uniform power flow and mechanical loads possible in this mode of operation. The entire mechanical and electrical drive train up to the grid is considered: the rotating masses of the rotor, machine and gear elements, the intermediate circuit of the converter and the magnetic field of the generator.

In a joint study by the departments of Electrical Machines and Drive Systems and Power Electronics and Drive Control at Leibniz Universität Hannover, the potential of the various short-term energy storage systems mentioned above is to be quantified and the basic feasibility, expense and limits for their use are to be considered. The different concepts of wind turbines (with / without gearbox) are to be comparatively included in the considerations.

Saturation behavior of doubly-fed asynchronous generators

In the wind industry, the doubly-fed asynchronous generator occupies a dominant position. With it, reactive power compensation or output is also possible via the converter supply on the rotor side, which was previously reserved for synchronous generators in the classic power supply. Since the above-mentioned operating points with power factor correction are not taken into account for asynchronous machines in the classic design without converter supply on the rotor side, the generators usually have to be overdimensioned to provide a safety reserve for the stronger saturation behavior in these operating points.

Therefore, software was developed at the Institute of Propulsion Systems and Power Electronics (IAL) at the University of Hannover as part of a ForWind project together with a major manufacturer of doubly-fed generators. This can calculate the magnetic circuit calculation of the machines also in the points with reactive power output.

Extensive measurements were carried out in the manufacturer’s test field on
different generators were carried out in order to compare the classical calculation approaches with those of the newly implemented method. It has been shown that the measured currents of the generators are larger, especially in the points with reactive power output, compared to the classical calculation approach.

With the newly implemented method, which records the saturation, in particular of the rotor yoke, much better in the case of double-sided feeding, the measurement results and the calculation correlate much better for the generators investigated, so that in future it may be possible to dispense with overdimensioning the generators.

Early damage detection on rotor blades

Rotor blades, especially of offshore wind turbines, are dynamically highly stressed structures. Damage occurs time and again, necessitating longer downtimes or even the replacement of entire blades. Typical damage represents delamination and fatigue of the composite structures, e.g. in the root area of the rotor blades.

The common practice of damage detection is still visual inspection, which is relatively easy to practice for onshore turbines, but is no longer calculable for offshore turbines due to the weather factor. A permanent recording of the vibration behavior of the rotor blades via a special measurement concept makes it possible to automatically detect damage to the structures at an early stage.

Within the scope of this development project, the monitoring concept is to be investigated with regard to its suitability for detecting damage typical of rotor blades. The monitoring concept is based on a method that simultaneously records strains and accelerations of the structure via integrated sensors. The vibration velocity and the stress are determined from these measured variables and are included in the proportionality relationship used.

The advantage of this method is the parallel consideration of two proportional amplitudes, which very soon lose their proportionality to each other in case of damage. Model tests have shown that this effect can be measured much sooner than, for example, the change in the natural frequencies of the structure.

The investigations within the scope of the project provide the result at which positions the sensors have to be installed in order to be able to provide the most reliable results. For this purpose, it is necessary to adapt the proportionality relation, originally designed for continuous beam structures, to the complex structure of a rotor blade.

Furthermore, critical damage variables are determined in cooperation with manufacturers. Experience data on damage growth serves as a source for limit value determination, which is integrated into the system.

Sensor development for condition monitoring systems

Various condition monitoring systems (CMS) have already established themselves on the market. However, most drive components cannot currently be monitored during operation. Thus, only a few conditions, such as deformations and temperature changes on the housing or noises or vibrations, can be observed from the outside. However, the causes of problems that occur (e.g. mechanical stresses) are usually located inside the gear unit housing.

New in-process measurement methods allow measurements in the gearbox during operation with a measurement uncertainty down to the sub-micrometer range. Most of these sensors exist today only as laboratory samples and must be further developed for space- and weight-saving long-term applications within wind turbine gearboxes.

In addition to micro-engineered sensors for measuring forces, moments and vibrations, as well as radiometric and thermographic sensors, fiber sensors are also to be developed. These can be integrated into the material of the rotor blades, for example, to measure strains

Flow visualization

The evaluation of flow conditions on wind turbine (WTG) rotor blades is important for optimal turbine operation and early detection of surface damage. For this purpose, the Bremen Institute for Metrology, Automation and Quality Science (BIMAQ) at the University of Bremen has developed various methods based on thermographic detection of the rotor blade surface during operation of the WT. The different heat transfers of different flow states (laminar, turbulent, dissipated) cause a thermal fingerprint even at the smallest temperature gradients between the rotor blade and the surrounding air, which correlates with the local flow conditions.

Local thermal differences can be detected from the ground or from drones using high-resolution thermographic cameras. Various evaluation algorithms allow automatic detection of the laminar-turbulent transition, areas of turbulent detachment, and damage that creates a turbulence wedge. By means of geometric assignment, the measured positions can be output directly in the coordinate system of the rotor blade.

Gear measurement

Gearbox damage often causes malfunctions and failures in wind turbines (WTGs). Previous experience with the use of gearboxes cannot be directly transferred due to the special requirements of the large WTG gearboxes. The laboratory for large gear measurements at the Bremen Institute for Metrology, Automation and Quality Science (BIMAQ) at the University of Bremen is expected to provide new insights here.

The heart of the air-conditioned laboratory is a coordinate measuring machine (measuring volume 2.5 m x 2 m x 0.7 m; length measurement deviation E < 1.6 µm + L/400 µm/mm; 18-22°C). This is used to measure new and defective wind turbine gear wheels in order to investigate the relationships between design, manufacturing, quality and functional properties of large gears in wind turbines – with effects on wear, service life, type of damage and noise development, among other things.

The laboratory will also be used to close the calibration chain for large components in gear and toothing production: In cooperation with the Physikalisch-Technische Bundesanstalt (PTB), reference standards are calibrated at the BIMAQ, i.e. standard references are defined for production.

Rolling bearings in the drive train

The drive train of wind turbines (WTG) is exposed to a wide range of stresses, such as unsteady load and wind fields. This includes short-term events concerning the electrical system, such as brief power failures, as well as loads on the rotor caused by gusts of wind. The precise dimensioning of mechanical components such as gearboxes and bearings is therefore an indispensable prerequisite for the safe and reliable operation of WTGs. In this way, temporary costly failures can be avoided. The existing service life calculations do not sufficiently take into account the special conditions during operation of a wind turbine. There is a lack of explanations for the large number of failures observed in practice.

The Institute of Machine Elements, Design Technology and Tribology (IMKT) examines the service life of rolling bearings subjected to high special loads experimentally and theoretically. The knowledge gained will enable a more accurate lifetime prediction. High dynamic loads and slippage between rolling elements and rings are taken into account. Throughout the investigations, condition monitoring plays an essential role. Methods are being developed to detect damage to rolling bearings at an early stage. Transferring these methods to wind turbines serves to improve monitoring and sustainably increases their availability.

One focus of the IMKT is the lubrication and sealing of rolling bearings, especially under extreme conditions. Lubricating greases at extremely low temperatures are a research focus. Studies have analyzed the frictional behavior of rotary shaft seals (RWDR) for these applications. Test rigs are available at the IMKT, ranging from test rigs for model investigations to large-scale test rigs for the investigation of a life-size wind turbine. In addition, it is possible to examine the most important components under climatic conditions, starting from low to very high temperatures.

Reliability of power semiconductor devices

The demands on the power components are high, as they are exposed to extreme requirements in terms of temperature and power cycles, and wind turbine generators (WTGs) are designed for a service life of 25 years. The thermo-mechanical degradation and failure mechanisms of the bond joints and chip soldering that accompany operation are investigated within the accelerated test method of the so-called Power Cycling Test (PCT) at the Institute for Electrical Drives, Power Electronics and Devices (IALB), University of Bremen.

The standardized accelerated test procedure with respect to corrosion and other moisture-driven degradation mechanisms is the so-called High Humidity High Temperature Reverse Bias (H³TRB) test, under which a service life of 25 years can be simulated within 1,000 hours. The IALB specializes in individual H³TRB tests with more stringent conditions.

Rotor blade reliability

The reliability of rotor blades is of particular importance for the safe and economical operation of wind turbines (WTG). Over the past decades, wind turbine manufacturers have gained a lot of experience with rotor blade design and damage during operation. Nevertheless, cracks in the rotor blade structure are still a cause of costly repairs and operational downtime. This shows that there are still knowledge gaps in the design of rotor blades with respect to fatigue, which is especially (but not exclusively) true for rotor blade bonding.

In order to close these knowledge gaps and to increase the reliability of rotor blades, the Institute of Statics and Dynamics at Leibniz Universität Hannover is working on the development, characterization and validation of simulation methods and models to describe the (fatigue) damage behavior of (short-fiber-reinforced) bonded joints in rotor blades. The focus here is on trailing edge bonding, as this is particularly susceptible to damage. This research makes significant contributions in the area of advanced materials testing and non-destructive imaging for short fiber reinforced adhesives. In addition, we are concerned with modeling the spatial distribution of fiber orientation in short-fiber-reinforced trailing-edge bonded joints, as well as continuum mechanical and energy-based modeling of fatigue damage behavior in short-fiber-reinforced bonded joints.

Inverter reliability

As the number of installed wind turbines (WTGs) increases, so does the number of power converters in the field. These are based on power semiconductor devices, in most cases IGBTs (Insulated Gate Bipolar Transistor) and diodes. Fluctuations in temperature and humidity due to operating and environmental conditions will cause the semiconductors to age over time.

Plant failures due to degraded converter components result in significant costs. While predictive maintenance is already state of the art for powertrains and other mechanical parts, it is not yet available for power electronics. The aim is to develop a system for predicting the remaining service life or online monitoring of the electronics.
This requires, on the one hand, precise measurements of the relevant variables of a wind turbine in operation and, on the other hand, appropriate models for service life prediction. The latter already exist in rudimentary form and take into account both thermomechanical and electrochemical aging mechanisms. For the purpose of recording the required field data over a long period of time, a flexible condition monitoring system has been developed at the Institute of Electrical Drives, Power Electronics and Devices (IALB). The plant is monitored at high resolution up to 200 kHz, 24 hours a day, 365 days a year. Currently and in the future, data is being collected from WTs at active commercial wind farms to identify the events and conditions that have a major impact on degradation.

Based on the measured data and reliability models, a real-time algorithm can estimate the current lifetime consumption and predict the remaining available lifetime.