Unexpected discrepancy in pitch decay period and wave‑induced motion of a FOWT between OpenFAST and OWS (OpenFAST+WEC‑Sim) using identical WAMIT hydrodynamic data.

[CUI20210930]

Summary

Unexpected discrepancy in pitch decay period and wave‑induced motion of a FOWT between OpenFAST and OWS (OpenFAST+WEC‑Sim) using identical WAMIT hydrodynamic data.

System & Version Information

• OpenFAST version: v3.5.4
• WAMIT version: v7.5

Issue Description

I am performing coupled aero‑hydrodynamic simulations of a floating offshore wind turbine. To validate the numerical setup, I compare two modelling approaches:

1. OpenFAST alone – solving both aerodynamics (FAST.Farm / AeroDyn) and hydrodynamics (HydroDyn).
2. OWS (OpenFAST + WEC‑Sim)– using OpenFAST for aerodynamics and WEC‑Sim for hydrodynamics.

Both approaches use the identical set of WAMIT hydrodynamic coefficients (v7.5) obtained from a single WAMIT run. The same viscous correction (drag coefficients) is applied, and the mooring system is modelled via MoorDyn in OpenFAST with identical parameters. In OpenFAST, PtfmRefxt in HydroDyn is set to the centre of mass of the WINDMOOR platform (a deliberately offset centre of gravity).

I observe an unexpected discrepancy between the two methods:

Test case	OpenFAST result	OWS result
Pitch free‑decay (still water)	natural period ≈ 27.4 s	natural period ≈32 s
Regular wave test	motion amplitude 28 % larger	reference baseline

The discrepancy is consistent and significant in both frequency (decay test) and amplitude (forced response).

---

What I Have Verified (Identical between the two simulations)

Quantity	Status
Wave excitation forces (1st‑order)	identical (from same WAMIT data)
Radiation damping matrix	identical
Added mass matrix	identical
Platform mass & inertia	identical
Mooring properties (MoorDyn input)	identical
Viscous drag coefficients	identical
Hydrostatic restoring terms?	suspect discrepancy

Given the above, the only remaining suspect is how the hydrostatic or coupling stiffness terms (e.g., the (3,5) term or other off‑diagonal restoring coefficients) are treated in the two time‑domain solvers, or whether they are incorrectly transferred from the WAMIT output.

---

Hypothesis / Suspected Root Cause

I suspect the problem may arise from:

• Different handling of the reference point offset between the platform CoM (where PtfmRefxt is set) and the point used by WAMIT for its output (e.g., SWL, point 0,0,0).
• Incorrect treatment of certain off‑diagonal coupling terms in the hydrostatic stiffness matrix when the WAMIT data is transformed to a different reference point.
• Discrepancy in how radiation damping / added mass matrices are interpolated or applied when the platform undergoes large pitch motion.

---

Questions for the Community / Maintainers

1. Are there known pitfalls when WAMIT v7.5 output is used with a platform that has a significant vertical offset between the centre of mass and the WAMIT reference point?
2. Could this be related to how off‑diagonal hydrostatic terms (e.g., coupling between heave & pitch) are imported/transformed in HydroDyn vs. WEC‑Sim?
3. Is there any known interpolation issue in applying radiation damping / added mass matrices when pitch motion is large?
4. Would switching NBodyMod (in HydroDyn) affect the handling of coupling stiffness terms in such a scenario?

Any insights or suggestions would be greatly appreciated. I am very much looking forward to any guidance you may offer.

Attachments

• OpenFAST_inputs.zip – HydroDyn, ElastoDyn, MoorDyn, AeroDyn input files
• WAMIT_inputs.zip – original WAMIT v7.5 input file and output files

---

Thank you for your time and support!

OpenFAST_inputs.zip

WAMIT_inputs.zip

[andrew-platt]

Dear @CUI20210930,

Could you try running this case with a newer version of OpenFAST such as v4.2.1 or v5.0.0 and see if the issue persists there? We unfortunately don't have the resources to fix issues in the v3.5.x versions of OpenFAST.

Regards,

[jjonkman]

Dear @CUI20210930,

Based on your summary, I suspect the issue is that your WAMIT analysis is based on a reference point of (0,0,0), while the reference point you set in HydroDyn (PtfmRefxt, PtfmRefyt, PtfmRefzt) is set to the center of mass. The reference point set in HydroDyn should match the WAMIT reference point.

I'm not sure what you mean by OpenFAST + WEC-Sim, as I'm not aware of any coupling between these tools.

Best regards,