Improved two-color LIF thermometry for gas–liquid system by optical flow algorithm

In this work, a two-color laser-induced fluorescence (LIF) thermometry approach for 2D (2 dimension) temperature field visualization in gas–liquid system is presented, which is still challenging in complex flow field changing rapidly with time and space. Our experiment shows that temperature deviation from two-color LIF image misalignment in few pixels scale can be the main source for temperature measurement error, especially at gas–liquid boundaries with high relative gradient value. In spatial modulated PL images with 0.2/pixel relative gradient value, temperature deviation estimated to be ~ 85.0 ℃ at temperature 25.0 ℃ under 1 pixel position, which is far from measurement error observed in uniformly PL intensity distributed images (6.0 ℃ maximum measurement error in 20.0–100.0 ℃). According to this problem, a nonlinear image registration method based on optical flow algorithm is proposed to enhance image registration accuracy. Relative ratio deviation is significantly reduced from maximum value 60 to < 10% in position deviation ranging from 1 to 15 pixels and corresponding temperature deviation value can be effectively reduced from maximum value ~ 75 ℃ to < 10.0 ℃ in the temperature ranging from 19.0 to 94.0 ℃. Finally, accurate 2D temperature distribution data and corresponding heat flux in liquid around hot steam bubble can be visualized by this convenient and low-cost thermometry method.