Verifying QBlade-Ocean: A Hydrodynamic Extension to the Wind Turbine Simulation Tool QBlade

Behrens de Luna, Robert; Perez-Becker, Sebastian; Saverin, Joseph; Marten, David; Papi, Francesco; Ducasse, Marie-Laure; Bonnefoy, Felicién; Bianchini, Alessandro; Paschereit, Christian-Oliver

doi:https://doi.org/10.5194/wes-2023-117

Preprints

https://doi.org/10.5194/wes-2023-117

Preprints

19 Sep 2023

| 19 Sep 2023

Status: this preprint is currently under review for the journal WES.

Verifying QBlade-Ocean: A Hydrodynamic Extension to the Wind Turbine Simulation Tool QBlade

Robert Behrens de Luna, Sebastian Perez-Becker, Joseph Saverin, David Marten, Francesco Papi, Marie-Laure Ducasse, Felicién Bonnefoy, Alessandro Bianchini, and Christian-Oliver Paschereit

Abstract. To realize the projected increase in world-wide demand for floating offshore wind, numerical simulation tools must capture the relevant physics with a high level of detail while being numerically efficient. This allows engineers to have better designs based on more accurate predictions of the design driving loads, potentially enabling an economic breakthrough. The existing generation of offshore wind turbines is reaching a juncture, where traditional approaches, such as the blade element momentum theory and actuator disc models, are becoming inadequate due to the increasing occurrence of substantial blade deflections. QBlade is a tool that includes a higher fidelity aerodynamic model based on lifting-line theory, capable of accurately modeling such scenarios. In order to enable the simulation of offshore conditions in QBlade, and to make use of this aerodynamic capability for novel offshore wind turbine designs, a hydrodynamic module called QBlade-Ocean was developed. In the present work, this module is validated and verified with two experimental campaigns and two state-of-the art simulation frameworks on three distinct floating offshore wind turbine concepts. The results confirm the implementation work and fully verify QBlade as a tool to be applied in offshore wind turbine simulations. Shortcomings reported by comparable tools in the non-linear response to irregular wave excitation are confirmed by QBlade and reiterate the need for further research in this field. A code-to-code comparison with the industry-designed Hexafloat concept highlights the coupled interactions on floating turbines that can lead to large differences in motion and load responses in otherwise identically behaving simulation frameworks.

Received: 09 Sep 2023 – Discussion started: 19 Sep 2023

Download & links

Robert Behrens de Luna et al.

Status: final response (author comments only)

RC1:
'Comment on wes-2023-117', Anonymous Referee #1, 05 Oct 2023
General Comments:
This is a well-written paper describing the verification and validation work for the new QBlade-Ocean features. The example test platforms and load cases are logical and well thought out, covering a range of important physics. The comparisons with OpenFAST and DeepLines Wind show very similar performance in ability to match predictions to experimental results. The low frequency motions that OpenFAST (and other mid-fidelity codes) struggles to accurately predict are shown to be roughly equally undercalculated with QBlade, indicating that this is still a problem that needs to be addressed. The name for the potential flow models is confusing since quadratic transfer functions (or Newman’s approximation) are present for some models.
Specific Comments:
Section 3.1.2: If second order excitation forces are included in the sum and difference frequency, why is the model called linear potential flow?

Section 3.2: What kind of thruster is used to simulate the rotor loads?

Section 4.2 – last paragraph: The explanation for the difference in surge frequency focuses on the stiffness from the mooring system, but this was shown in Figure 3 to match very well for both numerical models and the experiment. Is it possible there is some difference in the tuning of the added mass for the two numerical models?

Figure 4: It would be nice if there was some difference in the markers for 2.1 and 2.2 and if the corresponding wind speeds and blade pitch values were included in the caption

Section 4.4: With these very large wave heights, the viscous forces will also be important, especially for the more slender members of the Hexafloat

Figure 7: The amplitude of the tower loads at the main wave frequency is significantly smaller for the experiment compared to both numerical models, is there any comment on the cause of this?

Section 4.4.2: As mentioned in by the authors, a key issue in the surge / pitch frequency motions is the application of the damping from the decay tests to all conditions.

Figure 13: It is mentioned that the difference in the wave frequency pitch motions between the numerical models and the experiment is on the order of a tenth of a degree, but the difference in range shown in the box-whisker plot appears larger than this. The difference doesn’t seem negligible, and may also be related to the tuning of the viscous force coefficients from the decay tests.

Line 485: It is mentioned that the lower mean mooring tension in the experiment results in larger motions, but I would expect this to be dependent on the change in tension with translation rather than the mean tension.

Technical Corrections:
Line 35: “to a large extend” should be “to a large extent”

Line 65: water plane area instead of water surface area?

Line 82: Section is abbreviated here but not earlier in the paragraph

Line 116: Is implantation the correct word?

Line 352: Should include in this sentence that the discussed load cases are with the DTU 10 MW RWT

Figure 11: It would be great to move the legend so that the mentioned surge discrepancy is visible

Figure 15: Again it would be better to move the legend away from the data

Line 689: Sentence should be restructured with subject and verb
Citation: https://doi.org/10.5194/wes-2023-117-RC1
- AC1: 'Reply on RC1', Robert Behrens de Luna, 15 Nov 2023
  
  First of all, thank you for your feedback and comments on how to improve the manuscript. We have considered all the specific comments and each of them has been incorporated in the updated manuscript. In order to respond more specifically to each bullet point in the reviewer's "Specific Comments" section, we address each of them in detail in the following:
  
  Response to Specific Comment on Section 3.1.2:
  
  The name linear potential flow combined with Morrison drag addressed the linearity of the governing equations required to solve in the boundary element solvers as a preprocessing step (in the BEM code WAMIT). This solver generates the damping and excitation coefficients as well as the quadratic transfer functions. The linear hydrodynamic forces (radiation and excitation) and non-linear hydrodynamic forces (slow drift and sum-frequency) are calculated from these coefficients during the simulation. We agree however, that the naming convention might be confusing and given the fact that the boundary condition of the free surface is non-linear during the calculation of the QTF, also technically not correct. We adapted the name for this modeling type in the updated manuscript to “Potential Flow plus Morison Drag” (PFMD).
  Response to Specific Comment on Section 3.2:
  
  The brand and model of the thruster (Schübeler HST) have been added to the manuscript as well as a citation to Arnal (2020) for details.
  Response to Specific Comment on Section 4.2:
  
  We agree that this explanation is misleading as stated in the manuscript since the fairlead tensions indeed are verified in Figure 3, albeit at static conditions.
  
  We can rule out the added mass modeling as the cause for the seen deviation since both QBlade and OpenFAST rely on the added mass coefficients (at infinite frequency) that are calculated in a preprocessing step within the BEM tool WAMIT instead of modeling it explicitly (as would be done in a Morison only approach) for both Softwind and OC5.
  
  The observed difference in frequency also cannot be attributed to masses and inertias since the overall masses and inertias of the Softwind and OC5 structures were defined in accordance with the rigid body masses and inertias specified in the experimental definition. Additionally, consistent mass distribution was closely adhered to for the turbine and RNA, as well as for the assembled system CoG, between the numerical codes.
  
  Hence, the dynamics from the mooring lines are the probable cause. MoorDyn (OpenFAST) models the lines as rigid cable segments that are connected with joints that neglect bending stiffness in a lumped mass approach. QBlade-Ocean instead models the mooring system using the absolute nodal coordinate transformation (ANCF) in CHRONO, a nonlinear finite element formulation that includes bending, torsion and shear deformation. As a possible result, both codes required different amounts of additional stiffness in the surge DoF of the OC5 model to tune it towards the experimental result (1e+04 N/m in QBlade compared to 5e+04N/m in OpenFAST).
  
  Supporting this thesis is the close agreement that is seen between QBlade and DeepLines Wind on the Hexafloat model (attached figure 1) with no additional stiffness included. DeepLines Wind, like QBlade, relies on cable elements to capture the mooring dynamics.
  
  The deviation in natural frequency is small and we updated the manuscript to focus less on the small difference. Moreover, it is updated to state that a possible explanation could be found in the modeling of the mooring dynamics (lumped mass and no-linear cable elements).
  Response to Specific Comment on Figure 4:
  
  Thank you for the suggestion, the markers were modified for TC 2.2 and the caption has been completed with the missing information (both for Softwind and OC5), see attached figure 2.
  Response to Specific Comment on Section 4.4:
  
  We agree, this fact has been included in this section. Viscous forces influence all three models and this information was missing.
  Response to Specific Comment on Figure 7:
  
  These load cases, according to Wendt (DOI:10.17736/ijope.2019.jc729), contained a large measuring cable bundle that was hanging from the tower top (attached figure 3). This influence was not measured and its existence needed to be accounted for by tuning preload values which were tuned according to the motion response in waves. It seems likely that this too, is the cause for the influence in tower bending moments. Additional information stating the above has been added to the manuscript in the corresponding section.
  Response to Specific Comment on Section 4.4.2:
  
  Thank you for the feedback, as noted by the reviewer, this information is already in the manuscript and thus no changes are necessary.
  Response to Specific Comment on Figure 13:
  
  We agree that the formulation in this paragraph is inviting for misinterpretation, thank you for pointing this out. The description that is mentioned by the reviewer is supposed to address the comparison between QBlade and OpenFAST to each other and not, as was interpreted, between the numerical codes and the experiment. We agree with the reviewer that the IQR in the box-plot that corresponds to the experiment is wider compared to the numerical codes, this is a result of the non-linear motions too. This is stated just below the sentence that is addressed in the review with QBlade and OpenFAST underestimating the IQR by 16% and 10% respectively.
  
  A time series of the corresponding test case that proves the statement concerning the difference being in the order of tenths of a degree between QBlade and OpenFAS is shown in attached figure 4. The paragraph is reformulated in the manuscript to clarify this point.
  Response to Specific Comment on Line 485:
  
  After revisiting Figure (13d), we agree that the explanation given in line 485 is not sufficient to explain the increased response towards the lower end of the linear wave frequency range seen in the PSD of the fairlead tension in experiment. After consultation with the ECN researchers involved in the Softwind SIL setup, it is believed that the cause could be an incorrect estimation of the hydrodynamic damping on the mooring lines, which is applied by the Morison drag. This damping term often is sea-state dependent and can cause these discrepancies. The reason why this difference in the fairlead tension is not visible to the same degree in the surge PSD could be argued by the fact that the linear wave frequency range can be considered as a high frequency for the surge motion since the natural frequency in this DoF is well below this range. As a consequence, the response of the platform at wave frequencies is dominated by wave forcing and inertia. Hence, mooring stiffness has negligible effect. The manuscript is modified to include this information.
  Comment on the Technical Corrections:
  
  All Technical Corrections suggested by the reviewer are accepted and the modified manuscript will include the corresponding changes.
  
  We thank the reviewer for the constructive comments and we appreciate the detailed revision that has led to a significant improvement of this work.
  
  Citation: https://doi.org/10.5194/wes-2023-117-AC1
RC2:
'Comment on wes-2023-117', Anonymous Referee #2, 05 Oct 2023
The paper presents a code-comparison of the newly implemented hydrodynamic module in QBlade which is compared to model test data and simulations with openFAST and DeepLines. Three different cases are considered: OC5, Softwind and Hexafloat where both comparison with OpenFast and experimental data was performed for the first two cases, and only code comparison was done for the latter case. The paper is very well written with clear objectives and discussions, but maybe a bit too long. The validation is performed in a structured and stepwise manner where the complexity is increased for each step. However, the paper itself is just another code-comparison paper and does, as such, not bring any new scientific value to the community. Methodologies and models implemented and discussed are well established, and any shortcomings of these methods have been well-known for decades. For instance, it is well-known that second order models obtained from potential flow codes tends to fail predicting the low frequency motions and that this phenomena goes further back than the OC5 studies. There are several papers already addressing this issue. To increase the scientifc significance of the paper, a minimum should be to make an attempt to improve the prediction of slow drift motions using state-of-art approach or introduce a new methodology - not only doing a pure code-to-code comparison. Applying LLFVW in global analysis of FOWTs is to some extend of newer relevance, but not really the main objective of the paper.
Does the paper address relevant scientific questions within the scope of WES?
Only partly, as they do not attempt to solve any shortcomings with the methodologies, e.g. improved prediction of slow drift motions.

Does the paper present novel concepts, ideas, tools, or data?
No

Is the paper of broad international interest?
Only partly, there is some interest for code-comparison even if new methodologies are not introduced.

Are clear objectives and/or hypotheses put forward?
yes

Are the scientific methods valid and clear outlined to be reproduced?
yes

Are analyses and assumptions valid?
yes

Are the presented results sufficient to support the interpretations and associated discussion?
yes

Is the discussion relevant and backed up?
Yes, but does not bring any new knowledge

Are accurate conclusions reached based on the presented results and discussion?
yes

Do the authors give proper credit to related and relevant work and clearly indicate their own original contribution?
No, additional references (outside of the OC-community) where challenges with slowly varying motions should be given

Does the title clearly reflect the contents of the paper and is it informative?
Yes,

Does the abstract provide a concise and complete summary, including quantitative results?
Yes

Is the overall presentation well structured?
yes

Is the paper written concisely and to the point?
The paper is very comprehensive and well written, but is considered to be a bit too long to be concise in this respect.

Is the language fluent, precise, and grammatically correct?
Language is fluent

Are the figures and tables useful and all necessary?
Yes, but some figures could be larger to be easier to read.

Are mathematical formulae, symbols, abbreviations, and units correctly defined and used according to the author guidelines?

Not checked

Should any parts of the paper (text, formulae, figures, tables) be clarified, reduced, combined, or eliminated?
Paper should be shorten

Are the number and quality of references appropriate?
Some additional (classical) reference should be added

Is the amount and quality of supplementary material appropriate and of added value?
N/A
Citation: https://doi.org/10.5194/wes-2023-117-RC2
- AC2:
  'Reply on RC2', Robert Behrens de Luna, 15 Nov 2023
  Thank you for the constructive feedback, we appreciate the suggestions and aim to include them in the revised manuscript to improve its scientific relevance. To reply to the review, we structured the response in three sections. First, we would like to respond to the comments in the written paragraph with a general statement. Second, we break down the written comment by the reviewer to its main points of criticism and specifically address each one of them and detail the corresponding changes made to the revised manuscript. Finally, the numbered comments from the reviewer that have not already been addressed in sections I and II are commented on. We reviewed the manuscript with a focus on the scientific value and its improvement.
  Written Paragraph (Section I):
  
  From a general standpoint, to us, the scientific value lies in the fact that even though QBlade makes use of several advanced modeling techniques compared to DeepLines Wind and OpenFAST, the overall influence on global loads and motions in the very controlled and idealized environment that we investigated is only limited. This, while opposing our expectation going into this project, is also a result that is worth communicating with the community in our opinion. Having this in mind, the validation and verification of the hydrodynamic module of QBlade is a necessity to continue in this direction of research and to have a more detailed look into the possible advantages that the increased fidelity of the structural and aerodynamic methods contribute in more realistic environmental scenarios that should increasingly pose challenges for the BEM method.
  
  A subsequent study to the present manuscript analyzes the influence of the increased fidelity in more realistic met-ocean conditions. The considered conditions include misalignment, shut-down, ETM, etc. in a set of DLCs (760 simulations). The models that were used in this subsequent study rely on the QBlade (with LLFVW), OpenFAST and DeepLines Wind models that are presented and validated in detail in the present manuscript and hence builds on the results and findings of this work. A manuscript of this subsequent study in realistic met-ocean conditions is also currently in the discussion phase of WES (https://doi.org/10.5194/wes-2023-107).
  
  We appreciate the suggestion made by the reviewer that the scientific significance of this work can be increased by applying recently published methods to increase the accuracy in the prediction of slow drift motions. Following this recommendation, studies on improving the accuracy of slow-drift motions were considered for QBlade. The approaches introduced by Li and Bachynski-Polic (https://doi.org/10.1016/j.oceaneng.2021.109165) and Wang et al. (https://doi.org/10.1016/j.renene.2022.01.053) were investigated and the latter approach was identified as the one with a physically sound and applicable solution to the problem.
  
  Hence, the OC5 QBlade model in this study was updated concerning its drag coefficients close to the sea surface level. Moreover, the code itself was modified to optionally apply the axial drag to act on heave plates only on the faces experiencing negative flow with a filter that is applied to the vertical velocity component, allowing to partly separate the damping caused by the heave plate on the pitch and heave DoFs. The same approach is then partially (due to the lack of heave plates) applied to the Softwind model to investigate the transferability of the approach to spar-type floating structures. The improvement on the slow drift response for the OC5 model can be seen in the attached figure.
  Main Points addressed in the written statement of the review (Section II):
  
  More specifically we address the points that were made by the reviewer in the written paragraph in the following:
  i) “but maybe a bit too long”
  
  We agree that the manuscript could be shortened. To streamline the conveyed message, we will transfer some of the figures of the first load cases to the Appendix and only reference them briefly in the main body of text. To further cut some length, the frequency analysis from the regular wave section (sect. 4.4) is removed as this mostly relates to the importance of using the exact wave from the experiment as an input.
  ii) “the paper itself is just another code-comparison paper and does, as such, not bring any new scientific value to the community”
  
  We understand that with this paper the hydrodynamic module that matches the current state of the art is validated and verified extensively with experiments and other simulation tools and that this by itself doesn’t introduce new methodology or advance the state of the art with regards to the modeling of hydrodynamic forces, etc.
  
  The value we see in publishing this study in WES however, is that QBlade is established as a wind turbine simulation tool capable of performing floating offshore wind calculations. Through its advanced LLFVW capabilities and accurate multi-body structural model, it surpasses the current state of the art in those domains. By introducing a hydrodynamic module, the enhanced aerodynamic and structural capabilities can be used for FOWT technology research. In turn, validating the hydrodynamic module in the complex scenarios of FOWT requires a detailed analysis in a controlled environment with trusted reference results to increase the trust of the community. Furthermore, our findings indicate that in a controlled environment, such as the one used in our simulations, there is limited deviation resulting from increased aerodynamic fidelity. This has motivated us to continue our research and analyze the effect in more realistic conditions (as is described in detail below point iv)).
  iii) “To increase the scientific significance of the paper, a minimum should be to make an attempt to improve the prediction of slow drift motions using state-of-art approach or introduce a new methodology - not only doing a pure code-to-code comparison”
  
  As indicated in the general comment above, with the implementation of the approach laid out by Wang et al. (https://doi.org/10.1016/j.renene.2022.01.053), a recent proposal to improve predicted response at low frequencies in engineering tools is applied on the OC5 semi-submersible in QBlade. We saw that a different parameter setting of the cut-off frequency and blending parameter as compared to the cited reference was required in our cases. The method and the process to find the parameter settings are described in the revised manuscript. The latter information will provide valuable input for the applicability of this method to other tools and environmental conditions. Furthermore, the approach is partially transferred to the spar-type Softwind model. The improved results are added to the existing figures (irregular wind and wave cases) and show the improvement of this methodology with respect to the experimental result.
  iv) “Applying LLFVW in global analysis of FOWTs is to some extend of newer relevance, but not really the main objective of the paper.”
  
  As indicated in the general comment, a study on this topic has been carried out by the same group of authors which is also in the preprint phase within WES (https://doi.org/10.5194/wes-2023-107). An exhaustive amount of simulations of the three FOWTs presented in the current manuscript were carried out in QBlade, DeepLines and OpenFAST in realistic met-ocean conditions. DELs and ultimate loads are extracted allowing a comparison between the BEM and LLFVW methods under such conditions as well as the analysis of their influence on these metrics. The present manuscript and the abovementioned preprint should be seen as parts I and II that rely on each other’s findings. We will modify the titles and link both manuscripts to each other as a multi-part paper in order to highlight the connection and reliance of these papers to another. Of course, both manuscripts are revisited to emphasize the connection between them.
  Numbered Comments not already addressed in the replies above (Section III):
  10. The manuscript is revised to include a broader set of references concerning non-linear hydrodynamic excitation outside of the OC-community.
  
  14. As described under i), the manuscript will be streamlined by removing the frequency analysis of the linear wave cases and by shortening the sections discussing the natural frequencies and static conditions
  
  16. Considering the length of the manuscript and the requirements defined in the WES template regarding maximum allowed width, it is difficult to increase the figure size. Axis limits and figure sizes are revised to improve the readability of the figures containing multiple subfigures.
  
  19. Classical references are added to the manuscript at the corresponding sections that describe the basic theory behind a given modeling approach (e.g., Newman’s approximation, Morison’s equation, etc.)
  
  Citation: https://doi.org/10.5194/wes-2023-117-AC2

Robert Behrens de Luna et al.

Viewed

Total article views: 471 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
331	128	12	471	4	3

HTML: 331
PDF: 128
XML: 12
Total: 471
BibTeX: 4
EndNote: 3

Views and downloads (calculated since 19 Sep 2023)

Month	HTML	PDF	XML	Total
Sep 2023	110	59	5	174
Oct 2023	165	55	7	227
Nov 2023	56	14	0	70
Dec 2023	0

Cumulative views and downloads (calculated since 19 Sep 2023)

Month	HTML	PDF	XML	Total
Sep 2023	110	59	5	174
Oct 2023	165	55	7	227
Nov 2023	56	14	0	70
Dec 2023	0

Viewed (geographical distribution)

Total article views: 456 (including HTML, PDF, and XML) Thereof 456 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 03 Dec 2023

Short summary

In this paper, a novel hydrodynamic module of QBlade is validated and verified on three floating offshore wind turbine concepts using two experimental campaigns and two state-of-the-art simulation tools. With three very distinct substructure designs, various modeling approaches could be verified. Based on this study, the hydrodynamic capability of QBlade is matching the current state-of-the-art. This allows its advanced aero-elastic capability to be used in offshore conditions.


Total:	0
HTML:	0
PDF:	0
XML:	0