Development, calibration, and testing of thin-film heat flux sensors for wall heat flux measurements of hydrogen and hydrogen-methane flames

Abstract High spatial and temporal resolution heat flux measurements serve as an important tool for engine development. Traditionally, heat flux gauges rely on fast-response surface temperature measurements of a thermo-physically well-characterized substrate. However, previous work has shown that the thermal product (TP) of the bulk substrate may not adequately represent the effective TP of the gauge—thereby limiting the accuracy of the subsequent heat flux calculations. Accurate dynamic calibration of individual sensors is therefore essential. This work presents the development of painted platinum heat flux gauges for wall heat flux studies of hydrogen flames, describing in detail a new calibration methodology—substantially increasing the understanding of sensor calibration method, its limitations and the relative contributions to the overall measurement uncertainty arising from individual elements of the calibration. The method employs a conventional double electrical discharge technique combined with a novel approach to the identification of the linear region in the temperature response dataset, which was shown to improve the consistency of the derived TP. Significant sensitivity to the estimated starting time of the linear temperature response region was identified. Nevertheless, lower uncertainty in the TP values of individual sensors than has previously been reported in the literature was achieved. The calibrated sensors were initially validated against literature data for stoichiometric methane/air flames. Subsequently they were demonstrated in a campaign of stoichiometric hydrogen/air and hydrogen-methane/air blends flame wall heat flux tests to determine the combined experimental repeatability of the rig and instrumentation setup as well as to provide, for the first time, data on wall heat flux of hydrogen/air flames. A cross-comparison of heat flux measurements among multiple sensors was performed. Excellent experimental repeatability was observed for the sensors and uncovered experimental variability not previously seen.