A New Void Fraction Measurement Method for Gas-Liquid Two-Phase Flow in Small Channels

Area of Science:

• Fluid Dynamics
• Measurement Science
• Machine Learning Applications

Background:

• Accurate void fraction measurement is crucial for understanding gas-liquid two-phase flow dynamics.
• Conventional methods often struggle with flow pattern variations and high costs.
• Small channel flows present unique challenges for measurement techniques.

Purpose of the Study:

• To develop a novel, cost-effective void fraction measurement method for gas-liquid two-phase flow in small channels.
• To overcome the limitations of existing techniques, particularly their sensitivity to flow patterns.
• To validate the proposed method's effectiveness and accuracy across various flow regimes.

Main Methods:

• Utilized a laser diode and a 12x6 photodiode array sensor.
• Employed machine learning techniques, including Fisher Discriminant Analysis (FDA) for flow pattern identification.
• Applied Support Vector Machine (SVM) for void fraction modeling based on identified flow patterns.

Main Results:

• Successfully developed and validated a void fraction measurement system for small channels.
• The proposed method accurately measured void fraction across bubble, slug, stratified, and annular flow patterns.
• Achieved satisfactory measurement accuracy, demonstrating the system's effectiveness.

Conclusions:

• The developed low-cost laser diode system offers a viable and accurate solution for void fraction measurement in gas-liquid two-phase flow.
• The machine learning-based approach effectively mitigates the impact of flow pattern variations.
• This research provides a cost-effective and robust alternative to conventional measurement systems.