Output issue

[marisa16-daze]

Hello everyone!
I am currently conducting OpenFAST simulations, focusing specifically on the flapwise, edgewise, and torsional displacements of the blade tip.

When using ElastoDyn as the structural module, the blade tip flapwise displacement is TipDxb1, the edgewise displacement is TipDyb1, and the torsional displacement is TipRDzc1.

If I switch the structural module to BeamDyn, could you please confirm if the corresponding flapwise, edgewise, and torsional tip displacements are TipTDxr, TipTDyr, and TipRDxr, respectively?

[jjonkman]

Dear @marisa16-daze,

I agree, but the torsion should be TipRDzr in BeamDyn, expressed as a Wiener-Milenković parameter.

Best regards,

[marisa16-daze]

Dear Dr.jonkman
I have calculated and compared the output channels (including tower, blade, and platform motions) in OpenFAST for the following two grid configurations（SeaState）:
Steady Wind:10m/s
Case1.X_HalfWidth=120；Y_HalfWidth=120；NX=12；NY=12；NZ=15
Case2.X_HalfWidth=5；Y_HalfWidth=5；NX=5；NY=5；NZ=15
Contrary to my expectations, most of the output channels show negligible differences or even perfectly overlap (as shown in the attached figures). Does OpenFAST automatically calculate the dynamics for areas not covered by the grid if the grid size is too small?

I apologize for revisiting this topic, but I still have some lingering questions. Additionally, I noticed in a published paper（https://www.sciencedirect.com/science/article/pii/S0029801822004449） that the authors defined an exceptionally large SeaState grid. Is this typically done to prevent certain boundary effects?

For the IEA 15MW wind turbine on the UMaine VolturnUS-S (Semi-submersible) platform, would my current configuration (X_HalfWidth=120, Y_HalfWidth=120, NX=12, NY=12, NZ=15) be sufficient for medium-fidelity aeroelastic simulations? My computational resources are quite limited and cannot handle a larger domain or higher grid resolution. If possible, I would even consider further reducing these settings, as the current configuration still makes it very difficult to complete simulations exceeding 1000 seconds.

Size IEA 15MW wind turbine on the UMaine VolturnUS-S：

Thank you in advance for your guidance on my modeling work.
SinCerely

[marisa16-daze]

Dear Dr.jonkman
Subsequently, I conducted further tests comparing two different SeaState grid resolutions. Under a turbulent wind speed of 4 m/s, I observed more significant discrepancies in the results compared to the findings mentioned in my previous post.

Case1:X_HalfWidth=120;Y_HalfWidth=120;NX=12;NY=12;NZ=15( Total Simulation time:600s)
Case2:X_HalfWidth=95;Y_HalfWidth=95;NX=8;NY=8;NZ=15（Total Simulation time:1600s）

This is provided as an additional reference case only; please excuse the fact that I have not yet investigated these differences through statistical analysis. Meanwhile, the questions raised in my previous post still stand.How to balance computational resources and simulation fidelity when selecting an appropriate wave grid for the UMaine VolturnUS-S platform?

Best regards,

[jjonkman]

Dear @marisa16-daze,

Generally I would say that the SeaState domain should encompass the volume where the hydrodynamic analysis nodes will occupy, including from any floater motion, and be of sufficient resolution to resolve variations in wave kinematics across the domain.

The HydroDyn model of the VolturnUS-S semisubmersible provided in the IEA Wind 15-MW repository (https://github.com/IEAWindSystems/IEA-15-240-RWT/tree/master/OpenFAST/IEA-15-240-RWT-UMaineSemi) only contains a single hydrodynamic analysis node for the potential-flow solution, located in the center of the floater (without strip-theory nodes). In that case, the SeaState domain only needs to encompass any plausible floater (surge/sway) motion.

If using two different SeaState domains results in the same (or similar) response, I would say that the SeaState domain is sufficiently resolved.

Best regards,

[marisa16-daze]

Dear Dr.jonkman
Thank you for your prompt response. I will conduct further trials within the limits of my computational resources.
Best regards,

[marisa16-daze]

Dear Dr.jonkman
Please forgive my continued questions on these fundamental topics; I have one final inquiry.

To ensure a valid comparative analysis between two different platforms under the same sea state, should the wave cut-off frequencies in SeaState (WvLowCOff, WvHiCOff, WvLowCOffD, and WvHiCOffD) be kept consistent? Furthermore, how should these values be adjusted—or should they be adjusted at all—when the sea state changes?
SinCerely