Correlation between wind–wave–current loads and dynamic structural responses of a novel floating-tube photovoltaic array

With the increasing deployment of offshore floating photovoltaic (FPV) systems in complex marine environments, understanding their aerodynamic and hydrodynamic characteristics under combined wind–wave–current conditions has become important for structural design and reliability assessment. In this study, a full-scale 5 × 5 modular floating-tube-type FPV array was established using a two-way coupled CFD–MBD (Computational Fluid Dynamics and Multi-Body Dynamics) method with an overset mesh technique. Two co-directional wind–wave–current conditions with different inflow directions (0° and 180°) were considered to investigate the spatial distribution of environmental loads, module dynamic responses, and load–response relationships. Time-domain, frequency-domain, and POD-based modal analyses were conducted to reveal the response characteristics and the statistical associations between dominant load and response modes. The results show that the interaction between adjacent modules leads to significant spatial non-uniformity in aerodynamic/hydrodynamic loads and structural responses. Under the 0° case, the maximum reductions in aerodynamic and hydrodynamic loads reach 77% and 69%, respectively, demonstrating a pronounced shielding effect induced by upstream modules. Under the 180° case, some modules exhibit higher loads and responses compared with the inflow-facing reference row, with the maximum relative increase in heave response reaching 485%. Frequency-domain analysis indicates that the dominant response frequencies are associated with specific load frequency bands rather than showing a direct one-to-one correspondence. POD-based correlation analysis further reveals strong temporal associations between dominant load modes and response modes. The inflow direction has a limited influence on the correlation between aerodynamic-load modes and surge response but modifies the modal associations related to heave and roll responses. Coherence analysis further identifies the frequency-dependent linear associations between the dominant load and response modes.