CFD methods are widely used to design hydrodynamically stable and efficient underwater vehicles. Maneuvering analysis involves identifying fluid-generated forces and moments, represented as hydrodynamic derivatives, which characterize translational and rotational movements across all six degrees of freedom. These derivatives are crucial for predicting maneuvering performance and developing control strategies. Commercial CFD tools like ANSYS CFX, ANSYS Fluent, OpenFOAM, and STAR-CCM+ are commonly employed for this analysis [1]. Simulation approaches for AUV flows include DNS, LES, and RANSE, with RANSE being particularly effective for simulating complex fluid flows, capturing turbulence and boundary layer behavior. By averaging the Navier-Stokes equations over time, RANSE balances computational efficiency with the ability to predict turbulent flows, making it suitable for both steady and unsteady conditions. Its robustness in estimating hydrodynamic derivatives has been validated in numerous studies [2–11]. The choice of turbulence model in RANSE simulations is critical for accurately modeling turbulent flows around AUVs [12, 13]. Two-equation models, such as standard k- \(\epsilon \) , Realizable k- \(\epsilon \) , RNG k- \(\epsilon \) , and SST k- \(\omega \) , are commonly used for their balance of accuracy and efficiency [14, 15]. The standard k- \(\epsilon \) model is efficient for preliminary analyses but struggles with complex boundary layer phenomena [16–19]. The Realizable and RNG k- \(\epsilon \) models enhance accuracy by adjusting turbulence viscosity and incorporating smaller turbulence scales, respectively [11, 19–23]. The SST k- \(\omega \) model combines advantages of both k- \(\epsilon \) and k- \(\omega \) models, excelling in boundary layer treatments and adverse pressure gradients [1, 15, 19, 24]. For unsteady flows involving AUVs, the SST k- \(\omega \) model is preferred, especially for control surfaces and flow separation issues [12, 25, 26]. Potential flow theory and panel methods are also used for preliminary estimates, while Computational Fluid-Structure Interaction (FSI) combines fluid and structural analysis for dynamic coefficients.

Computational Fluid Dynamics Based Estimation of Hydrodynamic Coefficients of AUVs

章节
发表日期：

2026年

查看全文

Xianbo Xiang,
Faheem Ahmed,
Gong Xiang

摘要

CFD methods are widely used to design hydrodynamically stable and efficient underwater vehicles. Maneuvering analysis involves identifying fluid-generated forces and moments, represented as hydrodynamic derivatives, which characterize translational and rotational movements across all six degrees of freedom. These derivatives are crucial for predicting maneuvering performance and developing control strategies. Commercial CFD tools like ANSYS CFX, ANSYS Fluent, OpenFOAM, and STAR-CCM+ are commonly employed for this analysis [1]. Simulation approaches for AUV flows include DNS, LES, and RANSE, with RANSE being particularly effective for simulating complex fluid flows, capturing turbulence and boundary layer behavior. By averaging the Navier-Stokes equations over time, RANSE balances computational efficiency with the ability to predict turbulent flows, making it suitable for both steady and unsteady conditions. Its robustness in estimating hydrodynamic derivatives has been validated in numerous studies [2–11]. The choice of turbulence model in RANSE simulations is critical for accurately modeling turbulent flows around AUVs [12, 13]. Two-equation models, such as standard k- \(\epsilon \) , Realizable k- \(\epsilon \) , RNG k- \(\epsilon \) , and SST k- \(\omega \) , are commonly used for their balance of accuracy and efficiency [14, 15]. The standard k- \(\epsilon \) model is efficient for preliminary analyses but struggles with complex boundary layer phenomena [16–19]. The Realizable and RNG k- \(\epsilon \) models enhance accuracy by adjusting turbulence viscosity and incorporating smaller turbulence scales, respectively [11, 19–23]. The SST k- \(\omega \) model combines advantages of both k- \(\epsilon \) and k- \(\omega \) models, excelling in boundary layer treatments and adverse pressure gradients [1, 15, 19, 24]. For unsteady flows involving AUVs, the SST k- \(\omega \) model is preferred, especially for control surfaces and flow separation issues [12, 25, 26]. Potential flow theory and panel methods are also used for preliminary estimates, while Computational Fluid-Structure Interaction (FSI) combines fluid and structural analysis for dynamic coefficients.