the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
OF2: coupling OpenFAST and OpenFOAM for high fidelity aero-hydro-servo-elastic FOWT simulations
Guillén Campaña-Alonso
Raquel Martin-San-Román
Beatriz Méndez-López
Pablo Benito-Cia
José Azcona-Armendáriz
Abstract. The numerical study of floating offshore wind turbines requires accurate integrated simulations, considering aerodynamics, hydrodynamics, servo and elastic response of these systems. In addition, the floating system dynamics couplings need to be included to calculate precisely the excitation over the ensemble. In this paper, a new tool has been developed coupling the NREL´s aero-servo-elastic tool OpenFAST with the Computational Fluid Dynamics (CFD) toolbox OpenFOAM. OpenFAST is used to model the rotor aerodynamics alongside with the flexible response of the different components of the wind turbine and the controller at each time step considering the dynamic response of the platform. OpenFOAM is used to simulate the hydrodynamics and the platform's response considering the loads from the wind turbine. The whole simulation environment is called OF2 (OpenFOAM & OpenFAST). The OC4 DeepCWind semi-submersible FOWT together with the NREL´s 5MW wind turbine has been simulated using OF2 under two load cases. The purpose of coupling these tools to simulate FOWT is to obtain high-fidelity results for design purposes reducing the computational time compared with the use of CFD simulations both for the rotor aerodynamics, that usually consider rigid blades, and the platform's hydrodynamics. The OF2 approach allows also to include the aero-servo-elastic couplings that exist on the wind turbine alongside with the hydrodynamic system resolved by CFD. High complexity situations of floating offshore wind turbines, like storms, yaw drifts, weather-vane, or mooring line breaks, that implies high displacements and rotations of the floating platform or relevant non-linear effects can be resolved using OF2, overcoming the limitation of many state of the art potential hydrodynamic codes that assume small displacements of the platform. In addition, all the necessary information for the FOWT calculation and design processes can be obtained simultaneously, such as the pressure distribution at the platform components and the loads at the tower base, fairleads tension, etc. Moreover, the effect of turbulent winds and/or elastic blades could be taken in account to resolve load cases from the design and certification standards.
- Preprint
(30404 KB) - Metadata XML
- BibTeX
- EndNote
Guillén Campaña-Alonso et al.
Status: closed
-
RC1: 'Comment on wes-2023-19', Anonymous Referee #1, 11 Mar 2023
This paper presents the verification of a newly developed numerical framework that couples the aero-hydro-servo-elastic simulation tool OpenFAST for offshore wind turbines with the open-source CFD package OpenFOAM. In this framework, OpenFOAM is used to perform high-fidelity hydrodynamic CFD simulations of the floating platform while OpenFAST models the aerodynamic loads on the rotor along with the flexible tower and blades. The control of the turbine is also integrated in the simulations through OpenFAST. This framework has great potential in the design and analysis of floating offshore wind turbine systems. However, some additional clarification and investigation are suggested to render the verification presented more convincing. Please see a list of comments and suggestions below:
- In Table 1, what does “deterministic” mean in the load case description?
- On page 7 line 190, please add a comma before “and two flexible modes…”
- Were wave forcing/damping zones included in the OpenFOAM/OF2 simulations?
- Please provide the total number of cells used in the OpenFOAM/OF2 simulations.
- In Figure 5, it looks like the CFD mesh does not have a boundary-layer region next to the floater surface. What is the boundary condition on the platform surface? Free slip?
- In Figures 7 and 8, the agreement between the two approaches for LC 3.1* appears acceptable. However, the results for LC 3.1 with waves show important differences. Apart from the opposite initial transient motion in surge, which the authors already pointed out and should be investigated further, there is also a large phase shift in the periodic steady-state motion. This is somewhat concerning because the same wave time series was used in both simulations. The authors are encouraged to investigate what is causing the phase shift.
- Between lines 245 and 250 on page 11, the authors state that the mean displacements of the platform are similar between the two approaches; therefore, the mean drift load is consistently captured. In my opinion, the mean displacement observed in this case is primarily driven by the wind. In fact, Figure 8b shows the two methods giving rather different mean hydrodynamic force in surge. It’s also surprising that OpenFAST appears to show a negative mean surge force, opposite of what’s expected.
- The OF2 results for pitch motion and moment appear to be slowly decaying, whereas the pure OpenFAST results do not. I’m curious if there is any potential explanation for this.
- Is the larger fluctuation of rotor speed and generator power observed in the pure OpenFAST results simply a consequence of the larger tower-top motion?
- While the hybrid approach presented in this paper is definitely more computationally efficient compared to full CFD simulations, the comparison of computing time needs to be treated with care because it heavily depends on the targeted level of numerical resolution and convergence. For example, Tran and Kim (2016) reported the use of a prism-layer mesh on the platform surface to help resolve the boundary layer. It can therefore be argued that the CFD simulations of Tran and Kim have a higher level of fidelity compared to the present study. The authors are encouraged to include discussions on these caveats in the comparison of CFD computing time.
Citation: https://doi.org/10.5194/wes-2023-19-RC1 - AC1: 'Reply on RC1', Guillén Campaña-Alonso, 31 Mar 2023
-
RC2: 'Comment on wes-2023-19', Anonymous Referee #2, 18 Mar 2023
This paper develop a new tool coupling OpenFAST and OpenFOAM where the hydyodynamics and dynamic reponses of platfrom is conisdered in OpenFOAM and the aerodynamics of blade and controller in simulated in OpenFAST, to obtain the high-fidelity results, especially the nonlinear hydrodynamic loads. I believe this paper more or less makes contribution to the community. However, there are some technical issues must be addressed to increase the quality of the paper. See the attached file.
- AC2: 'Reply on RC2', Guillén Campaña-Alonso, 31 Mar 2023
Status: closed
-
RC1: 'Comment on wes-2023-19', Anonymous Referee #1, 11 Mar 2023
This paper presents the verification of a newly developed numerical framework that couples the aero-hydro-servo-elastic simulation tool OpenFAST for offshore wind turbines with the open-source CFD package OpenFOAM. In this framework, OpenFOAM is used to perform high-fidelity hydrodynamic CFD simulations of the floating platform while OpenFAST models the aerodynamic loads on the rotor along with the flexible tower and blades. The control of the turbine is also integrated in the simulations through OpenFAST. This framework has great potential in the design and analysis of floating offshore wind turbine systems. However, some additional clarification and investigation are suggested to render the verification presented more convincing. Please see a list of comments and suggestions below:
- In Table 1, what does “deterministic” mean in the load case description?
- On page 7 line 190, please add a comma before “and two flexible modes…”
- Were wave forcing/damping zones included in the OpenFOAM/OF2 simulations?
- Please provide the total number of cells used in the OpenFOAM/OF2 simulations.
- In Figure 5, it looks like the CFD mesh does not have a boundary-layer region next to the floater surface. What is the boundary condition on the platform surface? Free slip?
- In Figures 7 and 8, the agreement between the two approaches for LC 3.1* appears acceptable. However, the results for LC 3.1 with waves show important differences. Apart from the opposite initial transient motion in surge, which the authors already pointed out and should be investigated further, there is also a large phase shift in the periodic steady-state motion. This is somewhat concerning because the same wave time series was used in both simulations. The authors are encouraged to investigate what is causing the phase shift.
- Between lines 245 and 250 on page 11, the authors state that the mean displacements of the platform are similar between the two approaches; therefore, the mean drift load is consistently captured. In my opinion, the mean displacement observed in this case is primarily driven by the wind. In fact, Figure 8b shows the two methods giving rather different mean hydrodynamic force in surge. It’s also surprising that OpenFAST appears to show a negative mean surge force, opposite of what’s expected.
- The OF2 results for pitch motion and moment appear to be slowly decaying, whereas the pure OpenFAST results do not. I’m curious if there is any potential explanation for this.
- Is the larger fluctuation of rotor speed and generator power observed in the pure OpenFAST results simply a consequence of the larger tower-top motion?
- While the hybrid approach presented in this paper is definitely more computationally efficient compared to full CFD simulations, the comparison of computing time needs to be treated with care because it heavily depends on the targeted level of numerical resolution and convergence. For example, Tran and Kim (2016) reported the use of a prism-layer mesh on the platform surface to help resolve the boundary layer. It can therefore be argued that the CFD simulations of Tran and Kim have a higher level of fidelity compared to the present study. The authors are encouraged to include discussions on these caveats in the comparison of CFD computing time.
Citation: https://doi.org/10.5194/wes-2023-19-RC1 - AC1: 'Reply on RC1', Guillén Campaña-Alonso, 31 Mar 2023
-
RC2: 'Comment on wes-2023-19', Anonymous Referee #2, 18 Mar 2023
This paper develop a new tool coupling OpenFAST and OpenFOAM where the hydyodynamics and dynamic reponses of platfrom is conisdered in OpenFOAM and the aerodynamics of blade and controller in simulated in OpenFAST, to obtain the high-fidelity results, especially the nonlinear hydrodynamic loads. I believe this paper more or less makes contribution to the community. However, there are some technical issues must be addressed to increase the quality of the paper. See the attached file.
- AC2: 'Reply on RC2', Guillén Campaña-Alonso, 31 Mar 2023
Guillén Campaña-Alonso et al.
Guillén Campaña-Alonso et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 776 | 281 | 21 | 1,078 | 11 | 8 |
- HTML: 776
- PDF: 281
- XML: 21
- Total: 1,078
- BibTeX: 11
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1