Flow identification of vertical upward two-phase flow using the return map of the void fraction signal in the nonlinear phase space

The impedance signals of the void fraction in two-phase flow have nonlinear characteristics. In the present paper, a new index considering the nonlinear chaotic character is developed to extend the conventional probability density function(PDF) method. The average mutual information method and sensitivity study have been employed to construct the portrait of attractor in te 2 dimensional phase space. The specific bifurcations of each flow patterns were attained by return map and 3D PDF of reconstructed attractor. Applying the new index to the data from the vertical two phase flow in the 25.3 mm and 50.8 mm ID pipes gave a good performance in identification of the discrete bubbly flow, the cap-bubbly flow, the slug flow, the churn flow and the annular flow. The present results showed that the nonlinear chaos analysis is capable of providing more objective index in flow pattern identification