Scholarlink

<p>The reliability of instability warning methods is closely related to the accurate detection of disturbances. Under distorted inflow conditions, disturbances display a highly non-uniform distribution, making the system particularly susceptible to false alarms. Given the potential for severe consequences, this study aims to develop a highly robust warning method in a transonic compressor. Unsteady pressure is measured on the casing wall by using a cluster of time resolved pressure transducers with axial and circumferential spatial resolution. The results reveal that under distorted inflow conditions, the maximum disturbance occurs at the exit of distorted region, resulting in a circumferential non-uniform distribution. Fast wavelet transform is employed to extract characteristic frequency bands associated with unsteady fluctuation of tip leakage flow, enabling precise disturbance quantification. By implementing a composite variable threshold strategy, the proposed method ensures a 100-revolution warning time at any circumferential location.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Spatially Adaptive Threshold Strategy for Compressor Instability Warning under Circumferential Inlet Total Pressure Distortion

  • 期刊论文

  • 发表日期：

    2026年7月24日

查看全文
  • Zhonggang Fan,
  • Xiaobin Xu,
  • Yang Liu,
  • Dun Ba,
  • Jiahui Qiu,
  • Juan Du

摘要

The reliability of instability warning methods is closely related to the accurate detection of disturbances. Under distorted inflow conditions, disturbances display a highly non-uniform distribution, making the system particularly susceptible to false alarms. Given the potential for severe consequences, this study aims to develop a highly robust warning method in a transonic compressor. Unsteady pressure is measured on the casing wall by using a cluster of time resolved pressure transducers with axial and circumferential spatial resolution. The results reveal that under distorted inflow conditions, the maximum disturbance occurs at the exit of distorted region, resulting in a circumferential non-uniform distribution. Fast wavelet transform is employed to extract characteristic frequency bands associated with unsteady fluctuation of tip leakage flow, enabling precise disturbance quantification. By implementing a composite variable threshold strategy, the proposed method ensures a 100-revolution warning time at any circumferential location.