Development of combustion gas measurement systems for micro-rocket torch using a diode laser sensor

Abstract

Much noise from the combustion oscillations and more are included in the measurement data for combustion experiments of micro-rocket torch. To develop the combustion gas measurement system for the micro-rocket torch using tunable diode laser absorption spectroscopy (TDLAS) method, we develop the denoising algorithms using the Savitzky-Golay filter and K-fold cross validation for improved noise-rejection capabilities. First, a validation study has been conducted on the gas measurement system using denoising algorithms in the case of atmospheric measurements. The atmospheric measurement results using denoising algorithms are in exceptionally good agreement with atmospheric conditions. Therefore, this result indicated that this algorithm gave satisfactory accuracies in the case of atmospheric measurements. Next, we investigated the combustion gas measurements for micro-rocket torch using the TDLAS method and demonstrated the usefulness of the denoising algorithms. From the results of this study, it is obvious that a significant accuracy improvement was obtained in the majority of absorbance spectrum cases with use of the denoising algorithms. Furthermore, the tendency of time-series torch gas temperature using TDLAS method is similar to the time-series estimated values.