仿生水下推进技术在水下航行器设计与智能机器制造中具有重要应用前景。本论文围绕仿生推进中的复杂流动问题，重点探索了浸没边界法在复杂水动力学环境中的应用。通过将隐式力法-浸没边界法与大涡模拟相结合，提出了一种新的仿真框架，成功解决了高雷诺数流动、复杂几何形状和动态边界条件下的仿真问题。结合生物仿生学原理，本文深入探索了企鹅等水生生物推进机制的复杂性，还从流体动力学的角度揭示了推力生成、涡流动力学、流场结构以及水下生物运动特征的相互作用，提供了仿生推进系统设计优化的重要理论依据。

本文的的工作分为算法提升和仿生应用两方面，主要内容包括：

算法提升：

• 基于复合线性插值的浸没边界法：在浸没边界法中，固体拉格朗日网格对流体欧拉网格的切割作用会产生不同的切割网格类型，这些差异会影响计算精度和效率。本文在详细分析不同切割网格类型特点的基础上，提出了一种基于双线性和三线性插值技术的复合线性（Multi-linear）插值方法，该方法结合二次多项式修正项，采用较大插值模板来细化原始插值结果，显著减轻了切割网格对收敛速度和计算精度的不利影响。

• 虚拟流动壁模型：仿生推进的复杂几何和运动壁面为高雷诺数壁面湍流模拟提出了巨大挑战。浸没边界法多用于直接数值模拟，无法有效解决高雷诺数下的湍流计算。本研究提出了一种虚拟流动壁模型，用以模拟高雷诺数下复杂流动的边界效应。该方法引入线性切应力假设，通过在模化层外部边界获取外区解析物理量，根据虚拟速度剖面函数获取壁面模化的涡粘系数，进而在壁面附近构造等粘性区，结合浸没边界法实现了粗糙网格下对高雷诺数复杂流动的高效模拟，避免了传统壁面模化方法在复杂几何体下的适应性问题。

仿生应用：

• 顺桨运动对扑翼推进性能的影响：扑翼-顺桨运动是水下仿生扑翼推进的基本运动形式，企鹅的运动是该类运动形式的典型代表。本文建立了高保真二维企鹅翼模型，系统研究了扑翼-顺桨运动模式下的非定常水动力学特性。分析表明，顺桨运动是企鹅扑翼产生推力的主要机制，最佳顺桨幅值能够最大化推力系数，同时使推力和升力与运动周期同步，过大的顺桨幅值则可能导致负推力。在固定斯特劳哈尔数下扑翼幅值的变化对最佳顺桨幅值的确定无显著影响。引入推力角的概念，发现推力角能够有效地反映顺桨运动对推力的决定性作用，为分析和优化扑翼推进性能提供了新的理论工具。
• 高雷诺数下身体与翅膀的交互机制对扑翼推进性能的影响：体翼交互作用是生物体增强自身推进性能的重要机制，但水下扑翼推进中身体和翅膀交互机制的影响尚不明晰。本文进一步建立了高保真企鹅体翼三维运动模型，系统分析了不同雷诺数和不同运动参数下的企鹅水动力学性能，并探讨了身体与翅膀之间的相互作用对推进力生成的影响。研究表明，高雷诺数下流动由粘性主导转向惯性主导，推力增加、负推力区域减小，整体推进效率显著提升。身体模型在扑翼下冲程增强推力而在上冲程削弱推力，因此选用非对称扑翼可能更有利于提高企鹅的整体推进性能。适当的顺桨幅值下，身体模型的存在增强了翅膀的推进力和推进效率，同时，翅膀运动导致的非定常流场也减小了身体受到的阻力。身体和翅膀之间具有推力增益与阻力减缓的双重效应，但身体阻力的减缓占主要地位。此外，身体模型会带来推力不稳定性，该不稳定性随顺桨幅值和后掠角的增加而愈加明显。企鹅的高效推进需要选择合适的顺桨幅值，对于后掠角的考察中未发现其对推进性能的积极影响。

Biomimetics underwater propulsion technology holds significant application prospects in the design of underwater vehicles and intelligent machine manufacturing. This paper focuses on the complex flow issues in biomimetics propulsion, with particular emphasis on exploring the application of the Immersed Boundary Method (IBM) in complex hydrodynamic environments. By combining the implicit force method-immersed boundary method with large eddy simulation (LES), a new simulation framework is proposed, which successfully addresses the simulation challenges encountered under high Reynolds number flows, complex geometries, and dynamic boundary conditions. This framework not only ensures high computational accuracy but also significantly enhances computational efficiency, overcoming key bottlenecks faced by traditional immersed boundary methods in biomimetics propulsion applications. Integrating biological biomimetics principles, this study not only clarifies the complexity of propulsion mechanisms in aquatic organisms such as penguins but also reveals, from a fluid dynamics perspective, the interactions between thrust generation, vortex dynamics, flow field structure, and underwater biological movement characteristics, providing essential theoretical guidance for the design and optimization of biomimetics propulsion systems.
The contributions of this paper are divided into Algorithm Improvements and biomimetics Applications, with the main content outlined as follows:
Algorithm Improvements:
• Multi-Linear Interpolation-based Immersed Boundary Method: In the immersed boundary method, the cutting of the fluid Eulerian grid by the solid Lagrangian grid results in different types of cut grids, which can affect both computational accuracy and efficiency. This paper provides a detailed analysis of the characteristics of different cut grid types in the immersed boundary method, and proposes a multi-linear interpolation technique based on bilinear and trilinear interpolation methods. By combining quadratic polynomial correction terms and a larger interpolation template, this approach refines the original interpolation results and significantly mitigates the adverse effects of cut grids on convergence speed and computational accuracy.
• Virtual Flow Wall Model: The complex geometries and moving boundaries inherent in biomimetics propulsion pose substantial challenges for simulating wall turbulence at high Reynolds numbers. The immersed boundary method is frequently employed in direct numerical simulations, but it struggles with turbulence calculations under high Reynolds number conditions. In response, this study introduces a virtual flow wall model to simulate boundary effects in complex flows at high Reynolds numbers. The method introduces a linear shear stress hypothesis, retrieves external analytical physical quantities from the modeled boundary, and determines the wall-modeled turbulent viscosity coefficient based on a virtual velocity profile function. By constructing an equi-viscous region near the wall, this model, combined with the immersed boundary method, enables efficient simulations of high Reynolds number flows under coarse grids, overcoming the adaptability limitations of traditional wall modeling methods in complex geometries.
Biomimetics Applications:
• Impact of Feathering Motion on Flapping Propulsion Performance: The flapping-feathering motion is the fundamental movement pattern for underwater biomimetics flapping propulsion, with penguin motion serving as a representative example. This paper establishes a high-fidelity 2D penguin wing model to systematically study the unsteady hydrodynamic characteristics of this combined flapping-feathering motion. The analysis demonstrates that feathering motion is the primary mechanism behind thrust generation in penguin flapping, with the optimal feathering amplitude maximizing the thrust coefficient while synchronizing the thrust and lift with the motion cycle. However, excessive feathering amplitude can result in negative thrust. The variation of flapping amplitude under fixed Strouhal number has no significant effect on the determination of optimum feathering amplitude. The concept of angle of thrust (AoT) is introduced, showing that the angle of thrust effectively reflects the decisive role of feathering motion in thrust generation, offering a new theoretical tool for analyzing and optimizing
flapping propulsion performance.
• Impact of Body-Wing Interaction at High Reynolds Numbers on Flapping Propulsion Performance: Body-wing interaction is an important   mechanism for organisms to enhance their own propulsion performance, but the influence of body-wing interaction in underwater flapping-wing propulsion is still unclear. In this study, a highfidelity three-dimensional model of penguin body and wing motion is developed. The hydrodynamic performance of the penguin is systematically analyzed under various
Reynolds numbers and motion parameters, and the influence of body-wing interaction on thrust generation is examined. The results show that at high Reynolds numbers, the flow transitions from being dominated by viscosity to being dominated by inertia, leading to an increase in thrust and a reduction in the negative thrust region, significantly improving overall propulsion efficiency. The body model enhances the thrust in the downstroke of flapping and weakens thrust in the upstroke, so the choice of asymmetric flapping may be more favorable to improve the overall propulsive performance of the penguin. The presence of the body model enhances the mean propulsive force and efficiency at the appropriate feathering amplitude, while the unsteady flow field induced by wings motion also reduces the drag experienced by the body. The interaction between the body and wings demonstrates a dual effect of thrust augmentation and drag reduction, with the reduction in body drag being the more significant factor. Additionally, the body model introduces thrust instability, which becomes more pronounced with an increase in feathering amplitude and sweep angle. Efficient penguin propulsion requires selecting an optimal feathering amplitude, and no positive effect of sweep angle on propulsion performance is observed
in this study.