Final Report of IEA Wind Task 29 Mexnext (Phase 3)
2018
This report describes the results from the third phase of IEA TCP Wind Task 29 Mexnext, denoted as Mexnext-III, in which 22 parties from 10 countries participated. The main impact of the Mexnext-III project lies in the insights offered in the accuracy of design models by which it is possible to assess design calculations in a more thorough way. Moreover improved design models and recommendations on model use are provided which will reduce uncertainties in design calculations. Based on the project results it can be stated that current state of the art design models provide accurate aerodynamic load predictions for design conditions in constant axial and uniform inflow, while more scatter is observed in modeling the turbulent wake state, separated flow, 3d effects and tip effects. A prerequisite for accurate results is a careful consideration of input parameters and output post-processing. This especially holds for CFD simulations, of which the results are highly dependent on the large number of input- and output related choices. Guidelines to reduce the associated uncertainties have been given. Improvements have been made to engineering models describing amongst others yawed and dynamic inflow effects. However, it is shown that still large uncertainties exist in oblique, dynamic and/or nonuniform inflow conditions. Following this result it is recommended to further improve unsteady aerodynamic modeling in the fourth phase of IEA TCP Wind Task 29. Dedicated aerodynamic measurements, mainly from the New MEXICO experiment have been used to validate, improve and understand aerodynamic models. In the New MEXICO experiment pressure and flow field measurements were taken on a 4.5 meter diameter rotor in the large German Dutch Wind Tunnel, DNW. The New MEXICO experiment is a follow-up of the MEXICO experiment which was the subject of analysis in the first phase Mexnext-I. In Mexnext-I significant and unexpected discrepancies were found between simulations and measurements in terms of blade loads and induced velocities. Resolving the reason for these discrepancies and reducing them has been accomplished in Mexnext-III. Thereto, the entire New MEXICO database has been documented and delivered to the Mexnext-III consortium. With this database a large comparison exercise has been performed, featuring six simulation cases in axial and yawed inflow conditions, containing results of over 20 codes ranging from BEM to CFD. More than 10 different variable types ranging from lifting line variables to pressures, loads and velocities have been compared for the different conditions, resulting in over 250 comparison plots. Apart from benchmarking much effort was spent on understanding the the flow physics of various phenomena. The result is a unique insight in the current status and accuracy of rotor aerodynamic modeling. More specifically it can be concluded that most questions from Mexnext-I have been answered and so a much better understanding of the aerodynamics of the MEXICO rotor in terms of blade aerodynamics as well as near wake aerodynamics has been gained. This is illustrated by the better agreement between calculations and measurements compared to the agreement in Mexnext-I. This is true for pragmatic engineering models but also for high fidelity models. Opposite to the situation after the first MEXICO campaign, the relation between loads and velocities now fulfills the momentum balance, not only in axial direction but even in the tangential direction.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
13
Citations
NaN
KQI