Non-equilibrium spontaneous condensation flow in cryogenic turbo-expander based on mean streamline off-design method

Abstract

In a modern cryogenic liquefaction system, the turbo-expander is used to operate in the two-phase region so that a high cooling capacity can be provided. Compared with the expansion process in a conventional over-heated turbo-expander, non-equilibrium condensation occurs in the duct when the turbo-expander operates in the two-phase region, which varies the expansion process. In our previous works, the simulation by CFX can predict the nucleation process controlled by gaseous sub-cooling well, and the outlet liquid fraction has a good agreement with our experimental data. As the non-equilibrium condensation simulated by using CFX consumes much time and massive computing resource, it is not feasible to simulate the two-phase expansion in the turbo-expander in a wide off-design range. In this study, we investigate the non-equilibrium spontaneous condensation along the mean streamline of a cryogenic turbo-expander passage by using an off-design computational code compiled by Matlab. The two-phase off-design computational code is developed from our previous overheated off-design code. The calculation function of thermodynamic parameters on the mean streamline and non-equilibrium condensation module are integrated with this code. The variations of thermodynamic parameters on the mean streamline in a non-equilibrium condensation process are simulated. Nucleation onset, droplet growth/vaporization process, liquid fraction distribution are revealed. Droplet developing region is divided according to the droplet critical and mean radius. Simulation results of five two-phase cases are compared and analyzed, and different non-equilibrium condensation features and liquid fraction distributions are found. At last, isentropic efficiency and outlet liquid fraction are used to evaluate two-phase expansion performance, and it is found that the performance results from the simulation in five cases agree with experimental data well.