Yuko Inatomi

In the phase analysis, accuracy can degrade when the intensity of bright and dark lines in interference fringes varies significantly spatially, or when the phase change Δφ is temporally discontinuous. In this study, we improved the interference fringe analysis method by focusing on a local region in phase analysis and classifying phase inter-frame differences δφ’ in phase unwrapping. The objective of this study is to clarify how to set the threshold for classifying δφ’ and the accuracy of the phase φ and phase change Δφ obtained by the proposed method. Both experimentally obtained in a microgravity and artificially generated interference fringes were analyzed using the proposed method. As a result, the obtained spatial distribution of Δφ had a few outliers. In addition, it was revealed that a reliable threshold for classifying δφ’ can be automatically determined by using the total amount of tolerable noise and estimated δφ’. The phase error and the phase change error were as small as 0.025π and 0.034π rad, respectively, which is less than one-fifth of the conventional method.

In this study, we proposed a moved-view analysis, a method for obtaining the Soret coefficient S_T by analyzing the interference fringe change during field-of-view movement in the steady state. This analysis was designed to solve the problem that the sign of the concentration gradient could also be reversed if the plot of the concentration distribution was only slightly shifted owing to the narrow field of view. The data obtained from the experiment conducted to measure S_T at the International Space Station were analyzed using moved-view analysis. For the moved-view analysis, the linearity of the concentration distribution induced by the Soret effect is larger than that for the method without field-of-view movement, and a more reliable S_T can be obtained. The analysis error that sometimes occurred when the laser wavelength switched led to the underestimation of the phase change Δϕ, resulting in generating the data with low linearity. This unreliable data should be removed.