Good day,
I am working on a project which is related to simulations of a hybrid wind-wave systems using OpenFAST. The reference model for the hybrid system is the Volturn US-S floater, the IEA 15MW wind turbine and three Oscillating water column (OWC) wave energy converters which are already modelled .So i have already the hydrodynamic coefficients of 4 bodies from Wamit. Body one is the platform, bodies 2-4 (modes 9,15, and 21 only heave considered) are the OWCs, simulated as flat thickless pistons at the surface level. So, the idea is to incorporate these three additional bodies into the fast simulations using the functionality available in OpenFast as per below Link: [Implementation of Substructure Flexibility and Member-Level Load Capabilities for Floating Offshore Wind Turbines in OpenFAST (nrel.gov)]. As i am a beginner user , it would be really helpful if you can guide me on the necessary changes/modification required in FST and other input files to run the simulations for the hybrid system. Also expecting your help on creating the SubDyn input file which is to be used for this particular case as i dont have any reference files for the same .Awaiting your good support and guidance as always.
Regards
Nikhil
Dear @Nikhil.Mathews,
I would say that the hydro-elastic capability of OpenFAST described in the paper you reference was developed for modeling the structural flexibility and internal member-level loading within floating offshore wind substructures, but would likely have limitations when applied to OWCs connected to an FOWT. The SubDyn capability permits the modeling of elastic modes of the floating substructure atop the 6-DOF rigid-body motion. The elasticity in SubDyn can be represented in terms of Craig-Bampton and Static-Improvement modes of a substructure composed of a combination of beam, rigid-link, and taut cable elements interconnected by cantilevered, pin, universal, and/or ball joints. I’m not sure how you’d represent an OWC with this capability. Can you clarify what you were expecting to do?
Best regards,
Dear @Jason.Jonkman ,
Thanks for prompt response.
Below is an outline of the project objective
To simulate a hybrid system consisting of an offshore floating wind turbine OWC wave energy converters in OpenFAST Software and compare the results with a previously conducted scientific study on a hybrid wind-wave system. Specific activities to meet this objective include
Define the Floating Platform Geometry: Begin by defining the geometric characteristics of the floating platform, including the OWC chamber.
Develop a numerical simulation model of (FOWT)-with oscillating water column(OWC) in WAMIT using Potential Flow theory and obtain hydrodynamic coefficients of the hybrid model.
Develop a numerical simulation model in OpenFAST that incorporates the OWCwave energy converters and FOWT as a combined platform. This will involve integrating the OWCs as additional hydrodynamic bodies within OpenFAST.
This will be done Using the new functionality and software updates in OpenFAST
(Implementation of Substructure Flexibility and Member-Level Load Capabilities for Floating Offshore Wind Turbines in OpenFAST, nrel.gov).
Validate the simulation model by comparing the hydrodynamic loads and responses of the combined FOWT-OWC platform obtained from the OpenFAST simulations with the results from an existing scientific study on a hybrid wind-wave system.
Investigate the performance and behavior of the hybrid FOWT-OWC system under various external conditions, such as different wave spectra and wind conditions. Analyze the hydrodynamic loads, power production, and structural responses of the combined platform.
Link for existing scientific study journal with which results will be compared is below
Redirecting
please note that this is a proposal that we would like to check as the previous studies on hybrid systems have not used this approach i.e to simulate the hybrid (wind and wave )system as a combined body in Openfast.
Please let me know if the concept is clear or not.
Regards
Nikhil
Dear @Nikhil.Mathews,
Thanks for summarizing your project objective, which is clear to me now.
But I’m still not sure how you intend to apply the functionality of OpenFAST to an FOWT with added OWCs, especially considering the limitations in how SubDyn treats the elasticity of the floating substructure. Can you clarify what you had in mind there?
Best regards,
dear @Jason.Jonkman ,
thanks and well received your feedback.
I am not a proficient user of OpenFast and i wanted to get your feedback on possibility of this metholodoly…I haven’t used the subdyn feature in openfast till now as same was not required in simulation of a semisubmersible floating wind turbine. This was something suggested to be my professor as to check if such a integration is possible using the said functionality in Fast. Additionally I would like to know if i can add the additional hydrodynamic bodies i.e 3 no OWC columns and modify the Hydrodyn file and see how the simulation runs. Please let me know
Best Regards
Dear @Nikhil.Mathews,
As described in the paper you referenced, the HydroDyn module of OpenFAST now supports multiple potential flow bodies and the SubDyn module supports floating substructure elasticity. I’m concerned, though, that this functionality would not be sufficient for modeling the physics of an OWC, especially considering the limitations in how SubDyn treats the elasticity. I would be curious to know from you you would like to model the physics of an OWC.
Best regards,
Dear @Jason.Jonkman ,
Good day,
Sorry for the late reply.
The Approach is to model the platform as a rigid body and OWC is to be modelled as an extra body, and we assume the body is a flat piston which is the way normally OWCs have been modelled over the years and it has been proven to be a good approach. So now we want to see if the OWCs can be approximated somehow using OpenFAST. So idea is to come up with a strategy to model the OWCs or equivalent bodies with motion in heave only and by assuming a very thin piston that has relative motion in heave DoF and see if it is a good approach or not.
Best Regards
Nikhil
Dear Nikhil,
OK, thanks for clarifying. If I understand correctly, by “very thin piston”, you mean a prismatic joint (Prismatic joint - Wikipedia). The issue is that SubDyn (which is coupled to HydroDyn to enable hydro-elastic calculations within OpenFAST) does not support this type of joint. Again, SubDyn treats the structural flexiblity of the substructure as composed of a combination of beam, rigid-link, and taut cable elements interconnected by cantilevered, pin, universal, and/or ball joints (not including prismatic joints). You would need to modify the source code to introduce such functionality. I would guess this would not be possible in SubDyn because the Craig-Bampton approach assumes linearity in deformation and requires that there are no rigid-body modes in the structural formulation.
The Structural Control (StC) submodel of ServoDyn does support the addition of members with prismatic joints (e.g., vertically oriented TMDs), but this module is not coupled to HydroDyn, so, again, a source code change would be required to introduce such functionality.
Best regards,
Dear Jason,
Thanks for your insights. So do you think source code change will be a big ask and how long a duration you think will be required on modifying the source codes.
Best Regards
Nikhil
Dear Nikhil,
OK, thanks for clarifying. If I understand correctly, by “very thin piston”, you mean a prismatic joint (Prismatic joint - Wikipedia). The issue is that SubDyn (which is coupled to HydroDyn to enable hydro-elastic calculations within OpenFAST) does not support this type of joint. Again, SubDyn treats the structural flexiblity of the substructure as composed of a combination of beam, rigid-link, and taut cable elements interconnected by cantilevered, pin, universal, and/or ball joints (not including prismatic joints). You would need to modify the source code to introduce such functionality. I would guess this would not be possible in SubDyn because the Craig-Bampton approach assumes linearity in deformation and requires that there are no rigid-body modes in the structural formulation.
The Structural Control (StC) submodel of ServoDyn does support the addition of members with prismatic joints (e.g., vertically oriented TMDs), but this module is not coupled to HydroDyn, so, again, a source code change would be required to introduce such functionality.
Best regards,
Regards
Nikhil
Dear @Nikhil.Mathews,
Yes, unfortunately, I think both options (adding a prismatic joint to SubDyn, if possible, or adding a coupling between StC and HydroDyn) would involve a rather large effort, likely requiring weeks to months of work.
Best regards,
Dear @Jason.Jonkman ,
Do you know any projects similar to such hybrid systems being discussed in this forum or in general using OpenFast.
Regards
Nikhil
Dear @Nikhil.Mathews.
I’ve certainly seen requests to add functionality to OpenFAST to support the modeling of hybrid wind-wave energy systems, but NREL has not been funded to develop such a capability yet.
Best regards,
Dear @Jason.Jonkman,
*Good day,
Thanks for your clarifications regarding the subdyn feauture. I wanted to see how the hydrodyn feature works for multiple bodies and considering the subject system. Does openfast allows to choose degrees of freedom considering a case of mutlipe bodies . I would like to consider a case where the body 1 wind turbine is subjected to all 6 degrees of freedom and body 2,3 and 4 the OWC’s is subjected to only 1 degree of freedom i.e. Heave. Also* can you please guide me as to what all changed i should consider in the Hydrodyn file as i have only used it for a single body.
Regards
Nikhil
Dear @Nikhil.Mathews,
Each potential-flow body in HydroDyn is assumed to have 6 modes; that is, each body is assumed to have motion in 6 degrees of freedom (3 translation, 3 rotations) and hydrodynamic loads are calculated in 6 directions (3 forces, 3 moments). The actual structural degrees of freedom are not defined within HydroDyn, rather they are defined within the structural modules coupled to HydroDyn (ElastoDyn or SubDyn), but again, neither of these structural modules currently supports prismatic joints.
Unfortunately, the HydroDyn documentation on readthedocs (4.2.8. HydroDyn User Guide and Theory Manual — OpenFAST v3.5.0 documentation) was not updated when the multi-body functionality of potential-flow was added to HydroDyn. Nevertheless, hopefully the slightly outdated documentation on readthedocs together with the theory documentation on the multi-body functionality (https://www.nrel.gov/docs/fy20osti/76822.pdf) will be useful to use the multi-body functionality. Basically, though, to add multiple potential-flow bodies requires that you set NBody > 1 with NBodyMod set appropriately (see report 76822 for the definition of NBodyMod), and then set multiple values (one for each body) of PotFile, WAMITULEN, PtfmRefxt, PtfmRefyt, PtfmRefzt, PtfmRefztRot, PtfmVol0, PtfmCOBxt, PtfmCOByt, AddF0, AddCLin, AddBLin, and AddBQuad.
Best regards,
Good day Jason,
Attached is the Hydrodyn file I have modified for multiple bodies. I am also sharing Hydrodynamic coefficients obtained from Wamit for all the bodies. Please note for bodies 2 -4 only heave mode i.e. 9,15 and 21 is considered.
Can you please let me know the modifications made as per the requirements? Also, I would like to have a clarification for the quadratic drag matrix as to what coefficients or values they represent. I have inputted the same values which I have of a reference 15 MW standard semisubmersible floating platform without the OWCs.
Also, I am sharing my Elastodyn input file. I would like to if we can input the above files directly to Fast since for bodies 2 to 4 only hydrodynamic coefficients in heave are obtained or do we have to specify or modify any other input files(.1,.hst or .3 files) to get simulations running with such a requirement.
regards
Nikhil
(Attachment IEA-15-240-RWT-UMaineSemi_ElastoDyn.dat is missing)
(Attachment cd.hst is missing)
(Attachment cd.3 is missing)
(Attachment cd.mmx is missing)
(Attachment IEA-15-240-RWT-UMaineSemi_HydroDyn.dat is missing)
(Attachment cd.1 is missing)
Dear Jaison,
Thanks for the explanation and i will make the changes as advised. I have few more queries as to regarding the PtfmRefxt , PtfmRefyt, PtfmRefzt ,PtfmCOBxt, PtfmCOByt values .Referring to github example file u shared for NBody 4 , does these values correspond to centre of buoyancy values which are obtained from Wamit output .mmx file. I have shared this file also in the previous conversation . Also i wanted to know does quadratic drag value corresponds to any coefficient obtained as Wamit outputs or how those values are determined to be exact.Becoz i will have to make a 24 x 24 matrix for the same
Dear @Nikhil.Mathews,
HydroDyn inputs PtfmRefxt, PtfmRefyt, and PtfmRefzt should be used to locate the reference point for motions and the calculation of hydrodynamic loads on each potential-flow body. These should match what you’ve specified as XBODY in WAMIT for each body.
HydroDyn inputs PtfmCOBxt and PtfmCOByt should be used to locate the center of buoyancy of each potential-flow body. These should match the WAMIT output.
AddBQuad should represent the viscous drag of each body, not otherwise captured by the strip-theory (extended Morison) solution. In your HydroDyn model, I don’t see that you’ve defined any strip-theory members, so, all viscous effects should be represented via AddBQuad. WAMIT does not account for viscous effects, so, WAMIT cannot calculate AddBQuad. AddBQuad should be derived from knowledge of viscous effects for the specific floater you have; if that is not available, AddBQuad could be calibrated to match known CFD solutions or experimental results that account for viscous effects. Setting AddBQuad to zero would neglect viscous effects.
Best regards,
HI Jason,
i have modified the Hydrodyn input and Wamit data as advised. However I m an getting an error as attached. I have shared the updated files with you as per the below link
Regards
