Paper presented to the 10th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics, Florida, 14-16 July 2014.
In the paper the new method for the identification of two-phase flow patterns in a minichannel has been proposed. The two-phase flow (water-air) occurring in rectangular minichannel (3x3 mm) has been analysed. It was tested whether the change of the two-phase flow pattern was accompanied by changing the dynamics of two-phase flow. The signal from laser - phototransistor sensor, recorded during the experiment, was analysed. The proposed method was based on the recurrence plot and the principal component analysis.
The thirteen coefficients describing the dynamics of considered data (results of recurrence quantification analysis) have been treated as a set of variables describing the dynamics of two phase flow patterns. The principal component analysis (PCA) has been used to obtain the set of independent variables, which characterise the dynamics of system under consideration. Obtained results show that the first two components explain the 91% of input data variation. It has been shown that in the plots of two first components the points are grouped in the four separated clusters, which correspond to different flow patterns (two types of bubbly flow, and two types of slug flow - small and long slugs).
