A domain re-centering method for smoothed particle hydrodynamics simulation of liquid sloshing under prescribed tank trajectories

This study presents a domain re-centering method (DRM) for smoothed particle hydrodynamics (SPH) simulations of liquid sloshing under prescribed tank trajectories. At each time step, the DRM subtracts the incremental displacement of the tank center of mass from all particle positions, maintaining a compact computational domain while preserving the particle set, relative positions, inertial-frame velocities, and existing SPH governing equations. The SPH solver is validated against experimental measurements, and the accuracy and computational efficiency of the DRM are evaluated for linear transport and a 90° turn involving rectangular and cylindrical tanks with and without bulkheads. Free-surface profiles and pressure time histories closely match those obtained using an entire-domain SPH baseline with identical particle discretization. In the present cell-linked-list implementation, reduced background-cell storage and maintenance overhead yields computational-time reductions of up to 89% during curved motion. The transverse bulkhead suppresses longitudinal sloshing during acceleration and deceleration but provides limited mitigation during turning, when the dominant lateral flow is approximately parallel to its plane. These results demonstrate that the DRM reduces trajectory-dependent domain-management overhead while preserving sloshing predictions under the tested transport conditions.