In-situ characterization of the size distribution of lubricant droplets in rotary compressor

Rotary compressors are widely used in split air conditioning units and have a significant global market share. However, the atomization of lubricant in the rotary compressor during discharge from the cylinder can cause negative effects on both heat transfer performance and reliability. Therefore, it is crucial to deepen the understanding of the behavior of atomized oil in the form of small droplets in order to improve the system's performance. In this study, lubricant droplets were analyzed at different locations, operating frequencies, and pressures using shadowgraph imaging. The droplets were extracted using the Canny operator and corrected using a Convolutional Neural Network, and their sizes were analyzed statistically. The results reveal that the droplet characteristic mean diameters increase with frequency and pressure, with larger sizes observed in the outer region above the rotor/stator and constant sizes in the radial direction at the discharge tube level. Additionally, the droplet's volume fraction increases with frequency, with higher values observed in the outer region above the rotor/stator and above the rotor/stator clearance at the discharge tube level. These findings provide valuable experimental data for setting up the boundary conditions used in computational fluid dynamics (CFD) simulations.