In the modern aerospace industry, the importance of a lubrication system is self-evident for aero-engines, and the aero-engine bearing chamber return line is an even more challenging environment, as it involves a complex two-phase flow. The designer of the scavenge pipe needs to have an accurate understanding of the flow conditions in the scavenge pipe. This paper establishes a visual scavenge pipe test system. The flow direction was vertical flow, the test temperature was 370 k, and a high-speed camera was used to take high-definition flow photographs, which can observe the three main flow types: bubble flow, slug flow, and annular flow. Code program was created to analyze many pictures taken to obtain the apparent flow rate and perimeter of bubbles in the pipe under different flow conditions and to explore the gas–liquid two-phase flow in the scavenge pipe. A support vector machine (SVM) was used for data regression prediction, and the converted velocities of the gas–oil phases were inputted as eigenvalues to obtain the predicted values of bubble-flow velocity. The bubble-flow analysis prediction model established in this paper has a good prediction effect with root mean square error RMSE = 0.0172, which can more objectively and accurately describe the bubble-flow characteristics in the scavenge pipe.
Read full abstract