Complexity of two-phase flow dynamics using the recurrence and high speed video analysis

Abstract

During flow boiling in a system with small (mini/micro) channels, several instabilities may occur at the same time, which overlap each other  such a phenomenon complicates the analysis of boiling dynamics. The above mentioned processes cause that the fluctuation of recorded signals occur on various time scales. Although many criteria for the stability of two-phase flows are available, their practical application is limited (they need many recorded parameter of two phase flow). Methods which we are looking for should allow flow pattern identification based on a small number (or single) recorded signals. The paper presents a new approach to the recurrence plot method combined with Principal Component Analysis and Self-Organ-izing Map analysis. The single signal of pressure drop oscillations has been analyzed and used for flow pattern identification. New method of correlation analysis of flow patterns on video frames has been presented and used for flow pattern identifica-tion. The obtained results show that pressure drop oscillations and high speed video contain enough information about flow pattern for flow pattern identification.