Efficient single-phase RANS modelling of sloshing in cylindrical tanks: Applicability and limits

Sloshing in partially filled cylindrical tanks remains a critical problem in ocean engineering and cryogenic fuel containment, where predictive accuracy must be balanced against computational efficiency. While interface-resolving two-phase methods provide high fidelity at significant cost, efficient single-phase RANS formulations offer an attractive alternative whose domain of validity remains unclear. This study evaluates OpenFOAM’s potentialFreeSurfaceFoam solver against the CCP-WSI Blind Test Series 5 and extended frequency-response simulations across multiple fill levels and excitation amplitudes. The results show a bounded domain of validity governed primarily by geometry. At moderate amplitudes and fill levels h / D ≥ 0 . 5 , the solver captures resonant frequency and bulk centre-of-mass motion with competitive accuracy. Near resonance at low fill levels, however, the linearised hydrostatic boundary condition enforces amplitude-independent stiffness, suppressing frequency softening and peak–trough asymmetry. Errors are dominated by geometric amplification at the curved contact line rather than turbulence modelling, which has only secondary influence in non-breaking regimes. A consistent hierarchy emerges between global and local metrics: bulk mass redistribution is predicted more reliably than peak sidewall run-up. The work defines limits for fixed-mesh single-phase sloshing solvers and provides guidance for cylindrical tank applications where efficiency must be balanced against interface fidelity. • Efficient single-phase RANS modelling of cylindrical tank sloshing. • Geometry-controlled applicability limits in fixed-mesh free-surface RANS. • Bulk mass redistribution predicted more reliably than local wall run-up. • Linearised free-surface formulation suppresses frequency softening. • Orders-of-magnitude mesh reduction compared to interface-resolving CFD.