Impact of oscillation frequency on hydrodynamic coefficients and its role in ship manoeuvring behaviour

With advances in computational power, Computational Fluid Dynamics (CFD) has become a credible alternative to conventional towing tank experiments for determining ship hydrodynamic characteristics. In particular, virtual Planar Motion Mechanism (PMM) tests using CFD are increasingly employed to extract hydrodynamic coefficients for use in manoeuvring simulations. However, the sensitivity of these coefficients to key kinematic parameters—most notably the non-dimensional oscillation frequency ( ω 1 ′ )—and the implications for manoeuvrability predictions have not been thoroughly investigated. This study explores the influence of ω 1 ′ on hydrodynamic coefficients through a series of pure sway and yaw simulations using a CFD model of the HANNARA training ship—a conventionally shaped single-rudder vessel scaled at 1:16.667. The results indicate that sway-related coefficients ( Y v ′ , Y v v v ′ , N v ′ , and N v v v ′ ) best matched static drift test results within a narrow frequency range ( ω 1 ′ = 0.85–1.90), suggesting this range as a potentially optimal window for CFD-based PMM testing. In contrast, yaw-related coefficients ( Y r ′ , Y r r r ′ , N r ′ , and N r r r ′ ) exhibited significant deviations from those obtained via rotating arm tests, resulting in noticeable inaccuracies in turning indices and overshoot angle predictions . These deviations are attributed to non-linear flow behaviour at higher yaw frequencies and the challenges of capturing rotational fluid dynamics through pure yaw motion alone. Overall, the findings offer practical guidance for selecting appropriate kinematic parameters in CFD-based PMM simulations and contribute to the development of more accurate and robust methodologies for assessing ship manoeuvrability . • Sway-related hydrodynamic coefficients (e.g., Y v ′ , Y v v v ′ , N v ′ , and N v v v ′ ) showed good agreement with static drift results when using a low-frequency excitation range ( ω 1 ′ 0.85–1.90). • Yaw-related coefficients (e.g., Y r ′ , Y r r r ′ , N r ′ , and N r r r ′ ) deviated significantly from rotating arm test results across tested frequency ranges. • Using yaw-related coefficients from pure yaw tests led to large errors in manoeuvring predictions, including turning diameter and zigzag overshoot angles.