Neural network-based calibration for simultaneous measurement of blade tip clearance and rotor axial displacement with skewed-dual-light-probe

Abstract The accuracy of dynamic blade tip clearance (BTC) measurements in aero-engines is often compromised by rotor axial displacement (RAD), which remains a pervasive challenge requiring effective solutions. Additionally, the confined space within aero-engines necessitates the development of sensors with expanded functionalities. This study investigates the skewed-dual-light-probe (SDLP), a commonly used sensor for BTC measurements. The underlying causes of accuracy degradation due to RAD are analyzed, and a neural network-based calibration method is proposed to enhance the SDLP’s resistance to RAD effects while enabling simultaneous measurement of RAD. Numerical and experimental validations were conducted to assess the performance of the proposed method. The results demonstrate that the method improves the SDLP’s BTC measurement accuracy by nearly 50% under axial displacement conditions, with an RAD measurement error of less than 4%. This novel calibration method significantly enhances the SDLP’s robustness against interference, extends its functional capabilities, and provides valuable insights for improving the calibration methods of other BTC measurement sensors.