Computational Investigation of a Multiphase Turbo Expander for Heat Pumps and Refrigeration Cycles

Abstract

The design and development of an innovative Tesla style turbo expander for two-phase fluids is proposed, as a substitute for the lamination valve of a traditional Heat Pump cycle. Thereby enhancing the overall performance of the Heat Pump, by recovering mechanical work to offset the compressor requirements. The major challenge in such configurations is the reliable operation of the expander, when phase change occurs across it, from a purely liquid flow to a mainly vapour flow by volume with a dense cloud of liquid droplets. To investigate the phase change, a modelling approach is adopted which is routinely applied to modelling fuel-flashing in direct injection diesel engines, where the phase change deviates strongly from equilibrium. The Homogeneous Relaxation Model (HRM) is employed, which utilizes an Eulerian approach. The proposed computational model is firstly validated against experimental results available in the literature. A sensitivity analysis of the phase change model relaxation parameter is performed. It was found that a value 10 times lower than the published value gave closer agreement to the measured results. It is believed that this result is due to the roughened walls of the experiment, which would produce more nucleation sites for vapour bubble formation. This suggests that this model maybe is sensitive to the geometry of the turbine. Following this validation, the detailed flow profile in the proposed Tesla turbo-expander is investigated. Two different expander designs are considered in this project, one working with water [4,20] and the other with butane (R600). This study focuses particularly on the butane expander design. The expander performance is evaluated for rotational speeds up to 32’000 RPM. Results on the turbo-expander under investigation, showed that the presence of a dense cloud of liquid droplets produces a significant pressure drop across the turbine rotor, which increases with RPM, postponing the phase change. High volume-fraction of liquid was predicted to penetrate deeper inside the rotor above 16’000 RPM for the butane expander. The resulting lower liquid flow velocity relative to the rotor disk speed at the inlet of the rotor is predicted to significantly degrade the performance of the turbine at high rotational speeds. Decreasing the nozzle throat area improves the situation, by initiating the phase change further upstream and increasing the RPM operational range by 50%. Angling the nozzle radially inward by 10° was found to not have a great impact on the performance of the turbine. It was determined from this study that it is critical to predict correctly where the phase change starts, in order to accurately predict the performance of the turbine. Important is to remove as much liquid as possible from the flow, before it enters the rotor, to minimize the impact of the phase change on the turbine performance.