Speaker
Description
Infrared thermography (IR) is often used in aerothermal testing of turbine components as a non-intrusive technique for accurate measurements of blade surface temperature. Subsequently, measured surface temperatures can be used to derive heat transfer coefficients, adiabatic wall temperatures, and metal and cooling effectiveness.
Recently, an IR system has been successfully deployed to measure the surface temperature on high-pressure transonic turbine blades in the Oxford Turbine Research Facility (OTRF). Rotational speeds of approximately 265 ms-1 create challenging conditions for accurate IR measurements, requiring a compromise between obtaining sufficient detector saturation (dependent on integration time) and minimising image blur [1] [2].
This paper presents methods for improving system accuracy. Firstly, a method for improving system accuracy during significant under-saturation of the detector is presented. This novel pixel-wise calibration correction reduces the discrepancy attributable to low-signal bias between the measured IR surface temperature and reference thermocouples from 17K to 2K across integration times from 1µs and 10 µs using a temperature dependent correction factor in the camera calibration curve. Secondly, measurement uncertainty resulting from motion blur is assessed across integration times from 1µs and 10 µs, and a corrective deblurring algorithm that uses directional deconvolution is presented.
These methods are applied individually and collectively to previously reported OTRF data of blade tip surface temperatures and processed to heat transfer coefficients and adiabatic wall temperatures across a range of iteration times. The overall system uncertainty for the processed quantities is evaluated, enabling the optimal detector operating conditions to be identified.
Overall, the improved calibration method reduces measurement uncertainty in surface temperature from 5% to 1% at the lowest integration time of 1µs with naturally sharper images. Additionally, image deblurring allows data collected at higher integration times to be utilised without introducing significant errors from detector under-saturation, as sharper thermal fields can be reconstructed under high blade rotational speeds.
[1] M. Sisti, C. Falsetti and P. Beard, “High speed infrared thermography to investigate heat transfer of transonic turbine rotor blades,” Measurement, vol. 256, no. 118103, pp. 1-17, 2025.
[2] S. L. Gazzini, R. Schädler, A. I. Kalfas and R. S. Abhari, “Infrared thermography with non-uniform heat flux boundary conditions on the rotor,” Measurement Science and Technology, vol. 28, no. 025901, pp. 1-15, 2017.
