the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Numerical Analysis of Transonic Flow over the FFA-W3-211 Wind Turbine Tip Airfoil
Abstract. Modern wind turbines are the largest rotating machines ever built, with blade lengths exceeding one hundred meters. Previous studies demonstrated how the flow around the tip airfoils of such large machines reaches local flow Mach numbers (Ma) at which the incompressibility assumption might be violated, and, even in normal operating conditions, local supersonic flow could appear. In the present study, a numerical analysis of the FFA-W3-211 wind turbine tip airfoil is performed. The results are obtained by means of the application of numerical tools: (1) Xfoil with the Prandtl-Glauert compressible correction and (2) Computational Fluid Dynamic (CFD) simulations, where an Unsteady Reynolds-Averaged Navier-Stokes (URANS) model is used. A preliminary validation of the latter CFD model is performed to demonstrate that the URANS approach is a viable method for predicting the aerodynamic performances in compressible and transonic flow that provides additional and more reliable information compared to the classical compressibility corrections. From this study, three key findings can be highlighted. Primarily, the main transonic features of the FFA-W3-211 wind turbine tip airfoil have been assessed, selecting specific test cases of particular industrial interest. Then, the threshold between subsonic and supersonic flow is provided, considering also an increase of the Reynolds number (Re) from a characteristic value used in the wind tunnel experiments to the one realistic for large rotors. A strong dependence on this quantity is observed, revealing that, for the same Mach number, also the Reynolds number plays a crucial role in promoting the occurrence of transonic flow. Finally, the possible presence or absence of shock waves was investigated. The results indicate that the appearance of transonic flow is a necessary but not a sufficient condition to lead to shock formation.
- Preprint
(2544 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on wes-2024-47', Anonymous Referee #1, 15 Jun 2024
This article provides an investigation of the effects of compressibility on 2D airfoil flows in the context of wind turbine blade application at high Reynolds numbers and realistic Mach numbers for modern wind turbines.The manuscript is clearly and well-written. The methodology and scientific content appear correct and error-free. Only in a few places, the reviewer would like some clarifications and small corrections, as reported below.
Introduction, l.35-40: It is reported that 1) Yan and Archer found that neglecting compressibility could result in wind farm power production overestimation, 2) Campobasso et al and Ortoloni et al found that compressibility produces an increase of peak rotor power. This appears contradictory, and the authors should comment here if there are different mechanisms at stake, or at least point out this contradiction.
Section 2.2, l. 133: The authors refer to "dynamic stall". There seems to be some confusion here. The present calculations are done at constant angle of attack, and the authors probably means "stall" or "separation". Dynamic stall refers to the dynamic behavior of airfoil separation and stall when the angle of attack is varying.
Fig 6, caption: "Twoo"
Section 3.3: The interpretation of Fig. 14 needs clarification. If the reviewer understands well, all symbols are URANS calculations. When a supersonic region is present, the grey symbols are replaced by red ones, which are in turn replaced by green ones when a shock is detected. It is not clear why the authors did not conduct calculations for an angle of attack of 2 degs, M=0.6 and 0.65, for all Reynolds numbers, in order to refine the threshold curve shapes in this region, as illustrated by the horizontal line in Fig. 13 in this region.
Furthermore, it is noticed (l.244-245) that increasing Reynolds number promotes the appearance of transonic flow. Couldn't it be surmised here that this is caused by the decreasing viscosity? Or could there be another explanation?
Finally, it could be mentioned somewhere in the manuscript that tip effects (where the compressibility effects are maximum) may also play a role in the scenarios presented in this paper.
In conclusion, the paper is worth publishing and only minor revisions are suggested.Citation: https://doi.org/10.5194/wes-2024-47-RC1 -
RC2: 'Comment on wes-2024-47', Anonymous Referee #2, 19 Jun 2024
The article addresses the possibility of transonic/supersonic flow occurrence in the blade tip region of large wind turbines. The manuscript is well organized and well written, and relevant to the field. An interesting cross comparison of high- and low-fidelity aerodynamic methods to foresee the possible occurrence of transonic/supersonic flow is also presented.
Some additional comments are as follows. I would recommend the authors addressing them in the revised manuscript.
Mesh refinement should be addressed, indicating if or providing evidence that the presented parametric analyses provide mesh-independent results.
Could authors indicate the turbulence intensity of the results in Fig. 1? I would imagine the probability of transonic flow occurrence increases with the turbulence intensity, that tends to be low offshore. This I would assume that the occurrence of transonic conditions is more likely onshore. Can authors comment on this in the paper?
Page 4. Article reads: ‘The aim of the first part of this study is to prove that transonic flow can appear even in normal operation conditions’. Please define ‘normal operating conditions’.
Page 4, article reads: ‘Also, to save computational resources, wall functions are used to model the boundary layer region’. AoAs of \pm 15 degrees are quite high. Is there not the chance that shock-induced separation due to boundary layer/shock interactions may occur? In this circumstance, use of wall functions may induce significant errors. Can authors please comment on this matter in regard to the presented results?
Fig. 3, Table 1. This type of analysis should be performed with the AoA pf \pm 15 degrees, since pressure perturbations on either airfoil size are more likely to extend farther away from the airfoil, potentially leading to spurious reflections. I am not asking to redo this analysis, but some comments on the choice of 7.99 degrees may be helpful.
Fig. 4c: is it not possible that additional reason for the discrepancy OpenFOAM/experiments could also be the use of wall functions in separated flow region?
Perhaps figures 11 and 12 could be slightly improved? For example, indicating LE is at s=1 (I suppose). Indication of upper and lower side in cf graphs may also help quicker readability.
Citation: https://doi.org/10.5194/wes-2024-47-RC2 - AC1: 'Comment on wes-2024-47', Maria Cristina Vitulano, 29 Jul 2024
Status: closed
-
RC1: 'Comment on wes-2024-47', Anonymous Referee #1, 15 Jun 2024
This article provides an investigation of the effects of compressibility on 2D airfoil flows in the context of wind turbine blade application at high Reynolds numbers and realistic Mach numbers for modern wind turbines.The manuscript is clearly and well-written. The methodology and scientific content appear correct and error-free. Only in a few places, the reviewer would like some clarifications and small corrections, as reported below.
Introduction, l.35-40: It is reported that 1) Yan and Archer found that neglecting compressibility could result in wind farm power production overestimation, 2) Campobasso et al and Ortoloni et al found that compressibility produces an increase of peak rotor power. This appears contradictory, and the authors should comment here if there are different mechanisms at stake, or at least point out this contradiction.
Section 2.2, l. 133: The authors refer to "dynamic stall". There seems to be some confusion here. The present calculations are done at constant angle of attack, and the authors probably means "stall" or "separation". Dynamic stall refers to the dynamic behavior of airfoil separation and stall when the angle of attack is varying.
Fig 6, caption: "Twoo"
Section 3.3: The interpretation of Fig. 14 needs clarification. If the reviewer understands well, all symbols are URANS calculations. When a supersonic region is present, the grey symbols are replaced by red ones, which are in turn replaced by green ones when a shock is detected. It is not clear why the authors did not conduct calculations for an angle of attack of 2 degs, M=0.6 and 0.65, for all Reynolds numbers, in order to refine the threshold curve shapes in this region, as illustrated by the horizontal line in Fig. 13 in this region.
Furthermore, it is noticed (l.244-245) that increasing Reynolds number promotes the appearance of transonic flow. Couldn't it be surmised here that this is caused by the decreasing viscosity? Or could there be another explanation?
Finally, it could be mentioned somewhere in the manuscript that tip effects (where the compressibility effects are maximum) may also play a role in the scenarios presented in this paper.
In conclusion, the paper is worth publishing and only minor revisions are suggested.Citation: https://doi.org/10.5194/wes-2024-47-RC1 -
RC2: 'Comment on wes-2024-47', Anonymous Referee #2, 19 Jun 2024
The article addresses the possibility of transonic/supersonic flow occurrence in the blade tip region of large wind turbines. The manuscript is well organized and well written, and relevant to the field. An interesting cross comparison of high- and low-fidelity aerodynamic methods to foresee the possible occurrence of transonic/supersonic flow is also presented.
Some additional comments are as follows. I would recommend the authors addressing them in the revised manuscript.
Mesh refinement should be addressed, indicating if or providing evidence that the presented parametric analyses provide mesh-independent results.
Could authors indicate the turbulence intensity of the results in Fig. 1? I would imagine the probability of transonic flow occurrence increases with the turbulence intensity, that tends to be low offshore. This I would assume that the occurrence of transonic conditions is more likely onshore. Can authors comment on this in the paper?
Page 4. Article reads: ‘The aim of the first part of this study is to prove that transonic flow can appear even in normal operation conditions’. Please define ‘normal operating conditions’.
Page 4, article reads: ‘Also, to save computational resources, wall functions are used to model the boundary layer region’. AoAs of \pm 15 degrees are quite high. Is there not the chance that shock-induced separation due to boundary layer/shock interactions may occur? In this circumstance, use of wall functions may induce significant errors. Can authors please comment on this matter in regard to the presented results?
Fig. 3, Table 1. This type of analysis should be performed with the AoA pf \pm 15 degrees, since pressure perturbations on either airfoil size are more likely to extend farther away from the airfoil, potentially leading to spurious reflections. I am not asking to redo this analysis, but some comments on the choice of 7.99 degrees may be helpful.
Fig. 4c: is it not possible that additional reason for the discrepancy OpenFOAM/experiments could also be the use of wall functions in separated flow region?
Perhaps figures 11 and 12 could be slightly improved? For example, indicating LE is at s=1 (I suppose). Indication of upper and lower side in cf graphs may also help quicker readability.
Citation: https://doi.org/10.5194/wes-2024-47-RC2 - AC1: 'Comment on wes-2024-47', Maria Cristina Vitulano, 29 Jul 2024
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
445 | 147 | 25 | 617 | 21 | 21 |
- HTML: 445
- PDF: 147
- XML: 25
- Total: 617
- BibTeX: 21
- EndNote: 21
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1