Question about comparing pitch hydrodynamic moment between OpenFAST and CFD for the OC4 DeepCwind semi-submersible in regular waves

[Lin.Yang2]

Dear Jason,

I am working on a CFD validation study of the OC4/DeepCwind semi-submersible floating wind turbine. I am trying to compare the pitch hydrodynamic moment （My）from OpenFAST and CFD under a regular wave condition（only wave，no wind）, but I found that the results are difficult to interpret.

In my CFD model, the floating platform is solved in 6DOF motion. The global pitch motion response from CFD agrees reasonably well with OpenFAST, with a difference of around 10%. However, when I compare the pitch hydrodynamic moment, the difference is very large.

In OpenFAST, I sum the wave-excitation moment (B1WvsMyi), radiation moment (B1RdtMyi), hydrostatic restoring moment (B1HDSMyi), and additional damping (B1AddMyi), i don’t use Morrison viscous correction. In CFD, I use the OpenFOAM to integrate pressure and viscous loads on the platform surface. I output the pitch moment about the platform center of gravity.

Before comparing the pitch moment, I had already verified the hydrodynamic loads Fx and Fz in the surge and heave directions. The results from OpenFast and CFD were not significantly different. I know that OpenFast calculates the moments and responses based on the reference point (0,0,0), while CFD calculates them based on the center of gravity. But since the motion response results are not significantly different, why is the pitch moment so nearly twice as much? Is there any OpenFAST output channel that represents the true total hydrodynamic pitch moment in a form more directly comparable to CFD surface integration?

Best regards
Lin.Yang

[Jason.Jonkman]

Dear @Lin.Yang2,

You mention the use of different reference points ((0,0,0) versus center of gravity); are you transforming the outputs from one solver (OpenFAST, CFD) to the same reference point of the other for consistency?

Regarding HydroDyn, does your OpenFAST model only have potential-flow enabled, without strip-theory elements? Do you only have a single potential-flow body (NBody = 1)? Have you enable second-order terms?

Best regards,

[Lin.Yang2]

Dear Jason,

Thank you for your comments. I have checked this point again, and I think I have found the main reason for the discrepancy.

In the above results, the reference points for the pitch moment that I compared were different. The reference point for the moment in OpenFAST was (0,0,0), while in CFD it was the center of gravity (0,0,-10.2). I attempted to convert the force moment in CFD to the fixed point (0,0,0), but the result still showed a significant difference compared to OpenFAST.

In OpenFAST, I only used potential flow and did not use strip-theory elements. For the viscous effect of water, I represented it in the motion equations using a viscous damping matrix. I have only one floating buoy, and I didn’t use second-order terms for this regular-wave case.

My current understanding is that the OpenFAST body reference point is not a fixed point in the global coordinate system during the platform motion. It is the body reference point of the floating platform. In the OC4 model, this point is initially located at the global origin ((0,0,0)), but when the platform translates in surge, sway, and heave, this body reference point also moves with the platform. Therefore, for the CFD result, I should not transform the hydrodynamic moment to the fixed global point ((0,0,0)). Instead, I should transform it to the instantaneous body reference point. In my CFD model, the instantaneous center of gravity is

G(t) = (x_G(t), y_G(t), z_G(t))

The initial vertical distance between the OpenFAST body reference point and the center of gravity is 10.20 m. Therefore, the instantaneous body reference point used for the CFD transformation is

P(t) = G(t) + (0, 0, 10.20)

After transforming the CFD surface-integrated moment to this moving reference point, the pitch hydrodynamic moment becomes much closer to the OpenFAST result.

Please let me know if my thought is correct, or if there are other details in HydroDyn that I should consider.

Best regards

Lin. Yang

[Jason.Jonkman]

Dear @Lin.Yang2,

Thanks for clarifying. Your calculation of P(t) is missing the contribution of rigid-body rotation (from floater roll, pitch, yaw) with the -10.2-m offset between (0,0,0) and the center of gravity.

Best regards,

[Lin.Yang2]

Dear Jason,

Thank you for pointing this out! I updated the transformation by including the rigid-body rotation of the offset between the CG and the OpenFAST body reference point. After applying this correction, the CFD pitch moment agrees much better with the OpenFAST result.

Thanks again for the helpful suggestion!

Best regards,

Lin. Yang