Abstract. The premature failure of wind turbines due to unknown loads leads to a reduction in competitiveness compared to other energy sources. Especially the failure of main bearings results in high costs and downtimes, as for an exchange of this component the rotor needs to be demounted. Load monitoring systems can make a significant contribution to understand and prevent such failures. However, most load monitoring systems do not take into account the main bearing loads in particular as there is no commercially applicable measuring system for this purpose. This work shows how main bearing loads can be estimated using virtual sensors. For this purpose, several regression models are trained with test bench data considering strain and displacement signals. It is investigated with which combination of signal type and regression model the highest accuracy is achieved. The results show that for either using strain or displacement signals an appropriate accuracy can be achieved. In particular, it is shown that a linear regression with interactions already achieves good accuracy and that further increases in regression model complexity do not add significant value.
A significant factor for success of energy production based on renewables is the expanded use of wind energy. For this reason, the wind energy will remain a central element of renewable power generation. As a result of the increased demand on wind energy, a trend towards higher power and torque densities of wind turbine drivetrains can be observed in the current and future development of wind turbines. These trend results in a more compact design of wind turbine drivetrains. Thus, drivetrains become more complex and have stronger interactions between the individual components and therefore the future design of wind turbines will rely more and more on sophisticated simulation models of whole assemblies. Consequently, a general and objective method is needed to quantify the quality of these models and their sub-models. This paper will introduce a modelling quality parameter (MQP) which allows the objective quantification of model quality by comparison between measurement and simulation data. The MQP highlights insufficient sub-models and their need for further improvements.
The continued expansion of fluctuating energy sources such as wind turbines and solar systems will increase the demand for more flexible operation modes of power plants. Especially steam turbines with all their components will have to sustain a higher amount of start-stop cycles in order to compensate for variations in wind and solar radiation. Besides the rotor, inner casings are an example for main steam turbine components which are strongly loaded by thermal cycles at each start and shut down procedure. A precise prediction of the attainable number of start-stop cycles enables a more flexible operation within the guaranteed lifetime. However, this would require time-consuming FE calculations for each power plant due to their specific steam parameters. In this paper, a physics based surrogate model is discussed for a fast prediction of permissible start-stop cycles at plant specific steam parameters. The correlation between the physical properties from the surrogate model (wall temperature difference and the resulting stresses) and the attainable number of start-stop cycles from the FE model is determined. A validation with a different inner casing design within a usual wall temperature range confirms the high accuracy level of the surrogate model compared to uncertainties like material scatter or casting tolerances. With the provided approach typically a higher number of starts can be efficiently calculated in the bidding phase compared to assuming only one conservative value for each turbine type or size. Furthermore, the steam parameters can be optimized for increasing the number of starts to the required value without additional and time-consuming FE calculations.
The continuously rising demand for renewable energies leads to increased installations of wind turbines with higher power. While the current power-to-weight ratio of up to 20 metric tons of cast iron per megawatt is stagnating, cast iron components of modern wind turbines are facing new challenges in terms of weight, manufacturability, and castability. These challenges can be addressed by systematically using multi-domain optimization approaches to reduce component weight and increase local component utilization.In order to meet the requirements for modern cast iron components, this multi-domain approach must employ methods from casting simulation, micromechanical analysis, topology optimization, and strength assessment. Here, casting simulation is used to determine local microstructure descriptors, which are subsequently used in micromechanical shakedown analysis to estimate the local microstructure-dependent fatigue strength. In parallel to the fatigue strength estimation, topology optimization is performed iteratively in combination with a castability analysis. The component strength is evaluated using a strength assessment approach based on the previously determined local material properties in combination with the topology optimized component.In this study, the overall concept of the proposed multi-domain approach is presented and requirements for the application of such an approach are formulated. The use case of this study is a planet carrier of a wind turbine gearbox manufactured from austempered ductile cast iron ADI-GJS-1050‑6. For this use case, a weight reduction of 17% was achieved while maintaining the required stiffness, such that the microstructure variance along the component was significantly reduced. Furthermore, the potentials and limitations of the presented approach are outlined and discussed in the context of the design of heavy-section castings.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.