The strong coupling effect of two-phase flow and ice accompanies the ice accretion process of aircraft and wind turbine in damp and cold environment. A method based on the Eulerian two-phase flow, domain discretization of finite volume method (FVM) and finite element method (FEM), and fluid–solid coupling for numerical simulation of ice accretion is presented in this paper. In addition, the icing process of two-dimensional (2D) ice accretion on airfoils is investigated. It is found that the difference between simulation results and experimental data comes from the phase changes of local collection efficiency [Formula: see text], which is a function of local vortex strength and changes with time. Furthermore, with the ice accretion, it is easy to generate a negative or distortion grid because of the inappropriate mesh boundary merge and reconstruction, leading to an inaccurate ice prediction result. Based on the influence region of icing accretion and capturing the complex flow characteristics of the near-wall region, a dynamic and partition adaptive grid reconstruction strategy between macroscopic ice layer and microscopic ice crystal growth process is established. This is then used to build the ice accretion step adjustment strategy for multi-step simulation method under different icing conditions. The droplet collection efficiency distribution is modified by considering the vortex structure in the near-wall region. For verification purposes, multi-step simulation results for ice accretion of NACA0012 airfoil and large camber and strong separation airfoil under specified icing conditions are compared with the corresponding experimental data and some previously predicted results. The results of quantitative comparison of ice shape indicate that the current calculation method has better consistency with the experimental data, especially for glaze ice. The similarity of ice shape is more than 20% higher than the previous prediction results, showing that the time-space adaptive adjustment strategy has good robustness and accuracy for ice prediction.

中文翻译：

---

自适应积冰模拟的方法和准则：网格边界合并和重建

两相流与冰的强耦合作用伴随着飞机和风力发电机在潮湿和寒冷环境中的积冰过程。本文提出了一种基于欧拉两相流、有限体积法（FVM）和有限元法（FEM）的域离散化以及流固耦合的积冰数值模拟方法。此外，研究了机翼上二维（2D）积冰的结冰过程。发现模拟结果与实验数据的差异来自于局部收集效率的相位变化[公式：见正文]，它是局部涡流强度的函数，随时间变化。此外，随着积冰，由于网格边界合并和重构不当，容易产生负网格或变形网格，导致冰预测结果不准确。基于结冰影响区域，捕捉近壁区域复杂流动特性，建立宏观冰层与微观冰晶生长过程之间的动态分区自适应网格重构策略。以此构建不同结冰条件下多步仿真方法的积冰步长调整策略。通过考虑近壁区域的涡流结构来修改液滴收集效率分布。出于验证目的，将NACA0012翼型和大弯度强分离翼型在特定结冰条件下的积冰多步模拟结果与相应的实验数据和一些先前的预测结果进行了比较。冰形定量比较结果表明，目前的计算方法与实验数据具有较好的一致性，特别是对于釉面冰。冰的形状相似度比以往的预测结果提高了20%以上，表明时空自适应调整策略对冰的预测具有良好的鲁棒性和准确性。冰形定量比较结果表明，目前的计算方法与实验数据具有较好的一致性，特别是对于釉面冰。冰的形状相似度比以往的预测结果提高了20%以上，表明时空自适应调整策略对冰的预测具有良好的鲁棒性和准确性。冰形定量比较结果表明，目前的计算方法与实验数据具有较好的一致性，特别是对于釉面冰。冰的形状相似度比以往的预测结果提高了20%以上，表明时空自适应调整策略对冰的预测具有良好的鲁棒性和准确性。