CFD commercial software FLUENT, based on solving RANS equations, was used to calculate propeller hydrodynamic performance and exciting force. Sliding grid technique was applied to simulate the propeller rotation. Different advance coefficients and oblique inflow angles were set as different working conditions through adjusting inflow magnitude and direction. DTMB4679 propeller as the research object, hydrodynamic performance of this propeller in oblique flow was first calculated, and agreed well with test results. It verifies the accuracy of calculation method. The time-domain data of propeller exciting force under different working conditions was monitored, and then spectrum curve was obtained through fast Fourier transform. The results show that propeller load in oblique flow will aggravate, and transverse velocity component of oblique flow will cause great transverse force generated by propeller. Moreover, the smaller advance coefficient causes the greater peak of propeller fluctuating pressure. Additionally, oblique flow angle has a greater influence on unsteady bearing force than fluctuating pressure.

Solving partial differential equations (PDEs) is essential in scientific forecasting and fluid dynamics. Traditional approaches often incur expensive computational costs and tradeoffs in efficiency and accuracy. Recent deep neural networks have improved the accuracy but require high-quality training data. Physics-informed neural networks effectively integrate physical laws to reduce the data reliance in limited sample scenarios. A novel machine-learning framework, Chebyshev physics-informed Kolmogorov–Arnold network (ChebPIKAN), is proposed to integrate the robust architectures of KAN with physical constraints to enhance the calculation accuracy of PDEs for fluid mechanics. We study the fundamentals of KAN, take advantage of the orthogonality of Chebyshev polynomial basis functions in spline fitting, and integrate physics-informed loss functions that are tailored to specific PDEs in fluid dynamics, including Allen–Cahn equation, nonlinear Burgers equation, Helmholtz equations, Kovasznay flow, cylinder wake flow, and cavity flow. Extensive experiments demonstrate that the proposed ChebPIKAN model significantly outperforms the standard KAN architecture in solving various PDEs by effectively embedding essential physical information. These results indicate that augmenting KAN with physical constraints can alleviate the overfitting issues of KAN and improve the extrapolation performance. Consequently, this study highlights the potential of ChebPIKAN as a powerful tool in computational fluid dynamics and proposes a path toward fast and reliable predictions in fluid mechanics and beyond.

In marine environments, ships are bound to be disturbed by several external factors, which can cause stochastic fluctuations and strong nonlinearity in the ship motion. Predicting ship motion is pivotal to ensuring ship safety and providing early warning of risks. This report proposes a real-time ship vertical acceleration prediction algorithm based on the long short-term memory (LSTM) and gated recurrent units (GRU) models of a recurrent neural network. The vertical acceleration time history data at the bow, middle, and stern of a large-scale ship model were obtained by performing a self-propulsion test at sea, and the original data were pre-processed by resampling and normalisation via Python. The prediction results revealed that the proposed algorithm could accurately predict the acceleration time history data of the large-scale ship model, and the root mean square error between the predicted and real values was no greater than 0.1. The optimised multivariate time series prediction program could reduce the calculation time by approximately 55% compared to that of a univariate time series prediction program, and the run time of the GRU model was better than that of the LSTM model.

This paper establishes a numerical model for simulating submarine surfacing through the ice. The numerical model is simplified to be the interaction between a submarine and an ice sheet. The ice sheet is discretized into a number of ice material particles, and the motion and the failure of them are calculated by the peridynamic method. A contact detection method is used to determine the interpenetration between ice material particles and the submarine. To verify the feasibility of peridynamics in modelling ice failure, the two cases of three-point bending of columnar ice and vertical penetration of ice sheet for a cylinder are simulated. Convergence studies are taken with respect to the particle spacing size for submarine surfacing through ice. The surfacing process of the submarine has been simulated, and the calculation results are discussed. The calculated ice failure feature is similar to the full-scale observation.

The ship powering performance can be predicted based on model tests or simulations. However, differences exist between the full-scale ship performance and extrapolated model data. Therefore, research on full-scale ship performance simulation has important scientific significance and engineering value. This study used the REGAL general cargo vessel to perform full-scale ship resistance and self-propulsion simulations for various grid numbers, time step sizes, and wall Y+ values and compared the calculation and empirical results. The resistance components, waveform, Courant–Friedrichs Lewy (CFL) number, pressure field, and wake flow field of various resistance simulation conditions were analysed. The wall Y+ value was observed to affect the capture of boundary layer flow, ship pressure field evolution, and waveform, which subsequently affect the ship resistance. A Y+ value within 200 was considered to be appropriate when simulating the powering performance of a full-scale ship. The waveform evolution was significantly affected by the time step size, which should meet the requirement that the CFL number on the free surface is less than 1. When simulating the self-propulsion of a full-scale ship based on sliding mesh, the CFL number on the interface should also be less than 1. Additionally, to obtain the correct propeller excitation force, it was recommended that the time step size be maintained below 1°/step (0.000076 L_PP/v).