Computational Modeling of Hermetic Reciprocating Compressors

Abstract

Detailed experimental measurements of a compressor (pressure, temperature, speed) are not easily available because of size, cost, and access difficulties. A comprehensive computer model is necessary to obtain pressure-volume diagrams for the compressor. The valve is a key component of a compressor as it determines both the efficiency and reliability of the compressor. The valves operate as a result of pressure differences between the ports and the gas chamber. Undesired vibrations and fatigue fracture of thin valves (0.2 mm steel sheet) have not been well understood and experimental difficulties do not allow a thorough analysis of the causes. Initial investigations and modifications to KIVA-3V showed this code can be used to model both gas and suction valve motion. Since the code was written for internal combustion engines, its adaptation to compressors proved to be a difficult task, particularly because of the orientation and complexity of valves. Extensive modifications were made to model an angularly moving valve system, however as we moved down from engine scale (10 cm) to compressor scale (1 cm), grid resolution became a challenge from time to time. Preliminary results showed that the design of the suction valve and the chamber impacts the pressure difference between the input port and the gas chamber. The higher the piston speed, the lower the pressure of the gas, making it perhaps more plausible to avoid premature valve closings at higher speeds. Even though it is still under investigation, valve location and the orientation of its loose end makes a difference to the pressure, suggesting that new designs may be necessary by manufacturers. Experimental data will be needed to validate our model and verify our observations. Future steps include adding a discharge valve and a fluid-solid interface (FSI) module to the modified KIVA-3V code in order to predict the motion of the valves as a result of their interaction with the chamber gas.