A novel approach for reducing uncertainties in soot volume fraction estimates from light extinction measurements

Abstract

Studies on the combustion of isolated droplets are crucial for gaining insights into the mechanisms of soot formation and destruction in practical combustion devices involving liquid fuels. The light extinction (LE) method is arguably the most popular technique for quantifying soot in laboratory flames and has been used frequently for investigating droplet flames. Fluctuation in the intensity of the background image is one of the primary sources of uncertainties in measuring soot volume fractions during transient combustion events using the LE technique. In the commonly adopted approach, referred to in this paper as the averaged background image (ABI) method, multiple frames of the background image are acquired prior to combustion, and the mean value of intensity computed from these frames is assumed as the best estimate of the true value of the background signal. In this work, a novel post-processing method, referred to as the extrapolated background image (EBI) method is presented, wherein the best estimate of the true value of background intensity for each pixel at a given time instant is calculated based on the measured value of intensity at a reference point in the image, outside the sooty region. Results are presented for a droplet combustion event, analyzed using the ABI and EBI techniques, for various fuels. As compared to the ABI method, the implementation of the EBI method was found to reduce the root mean square errors (RMSE's) by a factor of approximately three, helping significantly to reduce measurement uncertainties, especially for fuels with low sooting propensities.