【摘要】 系统综述了计算流体动力学耦合离散元法(CFD-DEM)在桥墩局部冲刷研究中的最新进展，重点论述了桥墩局部冲刷的动力学机制、CFD-DEM耦合方法的技术演进及其耦合策略等，分别从宏观和微观角度分析了CFD-DEM耦合方法在桥墩局部冲刷方面的应用，并在此基础上分析了CFD-DEM耦合方法在桥墩局部冲刷研究方面的不足。研究结果表明：CFD-DEM耦合方法能够较好地展示桥墩绕流、马蹄涡与尾涡演化过程和泥沙颗粒起动、输移、堆积之间的关系，为揭示局部冲刷机理提供了由宏观到微观的分析手段；现有研究推动了桥墩局部冲刷研究由冲刷深度预测向颗粒动力学过程分析拓展，然而高雷诺数条件下湍流结构模拟精度仍受限制，实桥尺度长时序计算难度较大，且DEM中球形颗粒假设与天然泥沙形态存在偏差，非球形颗粒在建模时仍需权衡精度与效率，CFD-DEM耦合计算成本较高；粗粒化方法虽能提高效率，却可能削弱局部流场和颗粒微观作用的表达能力，且当前多软件的耦合方式在数据交换、时间步协调和数值稳定性方面仍有改进空间；未来应重点提升并行计算效率，完善非球形泥沙颗粒动力学模型，发展高保真粗粒化方法，并推动耦合架构集成化和人工智能辅助模拟的深入应用。更多还原

【Abstract】 The recent advances in the application of coupled computational fluid dynamics-discrete element method(CFD-DEM) for studying the local scouring around bridge piers were systematically reviewed. The focuses were on the dynamical mechanism of local scouring around bridge piers, the technical evolution of CFD-DEM coupling method, and their corresponding coupling strategies. The application of CFD-DEM coupling method in local scouring around bridge piers was analyzed from both the macroscopic and microscopic perspectives. On this basis, the current limitations in this research field were identified. The research results indicate that the CFD-DEM coupling method can effectively demonstrate the evolution process of bridge pier bypass flow, horseshoe, wake vortices and the relationship between initiation, transport and deposition of sediment particles, thereby providing an analytical tool from macroscopic to mesoscopic levels for revealing the local scouring mechanisms. Existing studies propel the transition of bridge local scouring research from scour depth prediction to particle dynamics process analysis. However, the simulation accuracy of turbulent structures under high Reynolds number conditions remains limited. Long-term computation at full-bridge scale poses significant challenges. Furthermore, the spherical particle assumption in DEM deviates from the morphology of natural sediment. The modeling of non-spherical particles still requires a trade-off between accuracy and efficiency. The computational cost of CFD-DEM coupling is relatively high. Although the coarse-graining method can improve efficiency, it may weaken the representations of local flow field and mesoscopic particle interaction. Additionally, the data exchange, time-step coordination and numerical stability of current multi-software coupling frameworks require further improvement. Future research should prioritize enhancing the parallel computing efficiency, refining the non-spherical sediment particle dynamics model, developing high-fidelity coarse-graining methods, and promoting the integration of coupling architectures alongside the in-depth application of AI-aided simulation. 1 tab, 22 figs, 90 refs.更多还原