Particle image velocimetry (PIV) is a widely used experimental technique in ocean engineering, for instance, to study the vortex fields near marine risers and the wake fields behind wind turbines or ship propellers. However, the flow fields measured using PIV in water tanks or wind tunnels always have low resolution; hence, it is difficult to accurately reveal the mechanics behind the complex phenomena sometimes observed. In this paper, physics-informed neural networks (PINNs), which introduce the Navier–Stokes equations or the continuity equation into the loss function during training to reconstruct a flow field with high resolution, are investigated. The accuracy is compared with the cubic spline interpolation method and a classic neural network in a case study of reconstructing a two-dimensional flow field around a cylinder, which is obtained through direct numerical simulation. Finally, the validated PINN method is applied to reconstruct a flow field measured using PIV and shows good performance.

This work experimentally investigated the performance of hydrodynamic forces on a semi-submerged cylinder under an oscillatory flow. To generate the equivalent oscillatory flow, the semi-submerged cylinder is forced to oscillate in several combinations of different periods and amplitudes. The mean downward lift force was observed to be significant and the fluctuating lift forces show dominant frequency is twice that of oscillatory flow and amplitude that is the same as the mean lift force. Based on this main hydrodynamic feature, a novel empirical prediction formula for the lift forces on semi-submerged cylinder under oscillatory flow is proposed where the lift forces expression is proportional to the square of oscillatory flow velocity. This novel empirical formula directly assigns the fluctuating lift force with frequency twice of oscillatory flow and the amplitude that is the same as the mean lift force. This assignment of empirical lift force formula reduces parameters required to determine a dynamic lift force but is demonstrated to well predict the fluctuating lift force. The lift coefficient can reach 1.5, which is larger than the typical value 1.2 of the drag coefficient for a fully submerged cylinder with infinite depth. Moreover, relationships among hydrodynamic coefficients, Keulegan-Carpenter (KC) number, Stokes number and Froude number are studied. With the increase of KC number, the Froude number has a more significant influence on the distribution of hydrodynamic coefficients. As Froude number is increasing, the drag coefficient shows a nonlinear decay (KC < 20) but a linear increase (KC > 20), while the added mass coefficients show a nonlinear (KC < 20) and a linear (KC > 20) increase trend. The present work can provide useful references for design of the relevant marine structures and serve as the useful guideline for future research.

In this paper, the vortex-induced force characteristics of a flexible riser with buoyancy modules and strakes were experimentally investigated. The strain information caused by vortex-induced vibration (VIV) at different cross-sections on the flexible cylinder was measured in the model test. Combined with the modal superposition method and inverse analysis method for vortex-induced force, the vortex-induced forces on the flexible cylinder were inverted, including the excitation force in phase with velocity and added mass force in phase with acceleration. Then, the excitation coefficients and added mass coefficients at each time step were identified via a least-squares method. The results show that the added mass coefficient and excitation coefficient of the flexible riser covering buoyancy modules present a “sawtooth” distribution along the length of the pipe, while the total mass (sum of structural mass and added mass) is consistent along the pipe. The strake has little effect on the added mass coefficient of the buoyancy module and the bare pipe section, but when it covers the bare pipe section, the excitation coefficient of the bare pipe section decreases to a negative value, showing a damping effect. In addition, when the bare pipe section of the flexible riser is completely covered with helical strakes, only the vortex-shedding frequency component of the buoyancy modules remains in the VIV response of the pipe. The reason for this phenomenon is that helical strakes completely suppress the vortex shedding of the bare pipe section, resulting in the reduction in the VIV excitation energy of the original bare pipe section to 0.