Abstract: Objectives The flow field around underwater vehicles is highly complex, making the accurate prediction of hydrodynamic characteristics a significant challenge in computational fluid dynamics. To address this issue, this study aims to develop an effective approach suitable for numerical simulation and engineering prediction of complex flows. Methods Based on the conventional SST turbulence model, an anisotropic SST turbulence model (ASST) is proposed by introducing nonlinear correction terms into the Reynolds stress formulation. To assess its applicability and predictive capability, systematic numerical validations are conducted using several benchmark cases at moderate to high Reynolds numbers (Re = 3.6×10⁴–1.4×10⁷), including backward-facing step flow, the Rood wing–body configuration, the A-airfoil, and the SUBOFF model. Results The results indicate that the ASST model can effectively capture complex flow phenomena over a wide range of Reynolds numbers. In quantitative comparisons, the SST model generally underestimates Reynolds stress levels, whereas the ASST model effectively mitigates this issue through nonlinear corrections, resulting in improved predictive accuracy. Conclusions This study provides an important methodological reference for improving the prediction accuracy of hydrodynamic characteristics of underwater vehicles.