the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Aug 2025
Enhanced approach to match damage-equivalent loads in rotor blade fatigue testing
Abstract. In the design process of current wind turbine blades a critical step is the certification testing to confirm design assumptions and requirements. To demonstrate reliability in fatigue testing the blade shall be loaded in all areas of interest to the load levels, which at the end of such test campaign adequately represent the blade lifetime. These loads are typically derived from aero-elastic load calculations with a combination of different design load cases in the form of accumulated bending moment distributions. The current practice includes two fatigue test sequences, which are aligned with first flapwise and lead-lag modes respectively with the aim to reach defined target bending moment distributions. These two test sequences combined may not cover all areas of interests and some areas could be tested insufficiently. Also in some areas the conventional target bending moment formulation does not represent fatigue damage of the material correctly, as it is not derived from a stresses or strain based damage calculation and does not allow for mean load correction. The aim of this work is to demonstrate these shortcomings on a particular test case and propose an enhanced method to derive representative target loads, which cover all areas of interest and are strain proportional, allowing for correct material damage accumulation and mean load correction. It is shown for the test case that compared to the conventional methods the enhanced target loads require 16 % higher test loads at certain positions along the blade within the four main load directions and even more for load directions in between.
- Preprint
(4399 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
- RC1: 'Comment on wes-2025-99', Anonymous Referee #1, 12 Sep 2025
-
RC2: 'Comment on wes-2025-99', Anonymous Referee #2, 29 Sep 2025
Overall, this study highlights the significant underestimation of fatigue damage under the current standard and suggests a more accurate approach.
- Lines 93–108, The underlying assumptions are clearly explained, and the limitations of this study are also addressed.
- Line 258, The yaw angle probabilities may represent deviations from the theoretical optimum under power production. Perhaps an explanation is needed for the choice of these specific probabilities.
Citation: https://doi.org/10.5194/wes-2025-99-RC2 -
RC3: 'Comment on wes-2025-99', Anonymous Referee #3, 29 Sep 2025
The comment was uploaded in the form of a supplement: https://wes.copernicus.org/preprints/wes-2025-99/wes-2025-99-RC3-supplement.pdf
- AC1: 'Comment on wes-2025-99', David Melcher, 04 Nov 2025
Status: closed
-
RC1: 'Comment on wes-2025-99', Anonymous Referee #1, 12 Sep 2025
This review is a compilation of several reviewers from ORE Catapult.
- Lines 70 - 80 there is a repetition of 'As the proposed approach can be used for any load direction, it enables target loads for any fatigue test method including biaxial testing. '
- Figure 1 is a good way of showing the different approaches.
- Equations 2 and 3 needn't be used, there is a matrix formulation approach which can use an arbitrary reference line and the appropriate 'bending and axial' 3 x 3 sub-matrix of the 6 x 6 section matrix to calculate strain given loads defined at an arbitrary reference point/axis. If everything is done correctly, the two approaches will give the same result so the way described in the paper is valid but not the only way (having finished the review, I now see this is addressed very well in the appendices).
- Line 250 - that's a lot of steps!
- Line 265, section 3.1. This is an interesting comparison - I'm not sure our customers would agree that 1.8% is negligible, but with our damage-based approach (which from experience of discussing this with OEMs seems to be how it is done when confidentiality is not an issue, ie. in-house) the test load (where we can't control Fz) is just increased so that the damage (which is only really coming from Mx and My loading in the test) is matching the service life damage (coming from Mx, My and Fz loading).
- This approach is good because it has the advantage of needing less customer data, and I suppose material properties for UD carbon, glass and biax and triax are available in the OPTIDAT database to get the UTS and UCS for all the common material types you'd see in the blade. The disadvantage is that you can't consider the 3 R-value diagram - when we do fatigue analyses we frequently see that carbon fails unless you can account for the very high values of m in the SN curve for compressive loading. Our test load derivation essentially continues on path 3.2 to iterate to find loads which do the same amount of damage for each cross section of blade.
- Figures 3 and 4, nice plots :)
- Figure 5 and 6 do a very nice job of showing why IEC 61400-23 needed to move away from DELs to a damage based approach.
- ...and Figures 7 and 8 do a very nice job of showing why bi-axial testing is so important.
- Line 330- add text to draw attention to the criticality of the regions missed. Along the lines of “The hatched area includes features which should be tested such as critical structural details and significant load transitions between design elements. “
- Line 340 -change “magnificent”, could be changed to either significant or magnified.
- The appendices do a very good job of defending some commonplace assumptions about the strain derivation from test loading.
- Line 353. Perhaps the conclusion could make reference to the different streams laid out in the flow diagram in fig 1.
- Figure 1: “sweep” is used, as well as in line 114, I think these have 2 different meanings, but its not clear.
- Fig 3 and Fig 4 – do these relate to the streams labelled in the flow diagram? If so, they could be included in the caption.
Overall, the paper is making a great argument for a damage-based approach to testing, which is obviously what is going to become the norm going forwards with the new release of IEC 61400-23.
Citation: https://doi.org/10.5194/wes-2025-99-RC1 -
RC2: 'Comment on wes-2025-99', Anonymous Referee #2, 29 Sep 2025
Overall, this study highlights the significant underestimation of fatigue damage under the current standard and suggests a more accurate approach.
- Lines 93–108, The underlying assumptions are clearly explained, and the limitations of this study are also addressed.
- Line 258, The yaw angle probabilities may represent deviations from the theoretical optimum under power production. Perhaps an explanation is needed for the choice of these specific probabilities.
Citation: https://doi.org/10.5194/wes-2025-99-RC2 -
RC3: 'Comment on wes-2025-99', Anonymous Referee #3, 29 Sep 2025
The comment was uploaded in the form of a supplement: https://wes.copernicus.org/preprints/wes-2025-99/wes-2025-99-RC3-supplement.pdf
- AC1: 'Comment on wes-2025-99', David Melcher, 04 Nov 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 645 | 161 | 36 | 842 | 34 | 45 |
- HTML: 645
- PDF: 161
- XML: 36
- Total: 842
- BibTeX: 34
- EndNote: 45
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This review is a compilation of several reviewers from ORE Catapult.
Overall, the paper is making a great argument for a damage-based approach to testing, which is obviously what is going to become the norm going forwards with the new release of IEC 61400-23.