A turbulent/non-turbulent interface detection method is proposed based on fuzzy clustering of the instantaneous streamwise velocity field. The fuzzy cluster method overcomes the limitations of standard detection methods by removing the user bias in thresholding and window selection and not relying on the calculation of mean flow properties. The robust detection algorithm is applied to three experimental wall-bounded turbulent flows, and it is found to compare favorably against existing interface identification methods.

中文翻译：

---

湍流/非湍流界面检测的模糊聚类方法

提出了一种基于瞬时流速度场模糊聚类的湍流/非湍流界面检测方法。模糊聚类方法通过消除阈值选择和窗口选择中的用户偏差，而不依赖于平均流量特性的计算，从而克服了标准检测方法的局限性。该鲁棒检测算法被应用于三个实验性的边界壁湍流，并且被发现与现有的界面识别方法相比具有优势。