Theoretical and Experimental Activity on Ejector Refrigeration

Abstract

Ejector refrigeration has been studied at DIEF (Dipartimento di Ingegneria Industriale Firenze) since the ‘90s. Use of environmentally safe fluids (steam) was addressed. A two-stage prototype with cooling capacity 5kW was optimized and built. Later, the CRMC prescription for the design of the supersonic diffuser was focused. By a gradual reduction of the fluid velocity and a continuous profile, the CRMC design promises a reduction of the normal shock that usually develops in the mixing chamber. A second 40 kWf prototype was designed in 2010 for an industrial partner (Frigel Firenze s.p.a.). The design procedure used a thermodynamic code accounting for real gas behavior. This code gives a first design of the mixing chamber and diffuser according to the CRMC criterion. It also gives an estimation of the friction loss along the diffuser. A comparison between different operating fluids was performed and resulted in the selection of R245fa. A first design of the ejector was manufactured in carbon fiber. The primary nozzle is mounted on a movable support, in order to change its axial position with respect to the mixing chamber. In terms of COP, first results were below the values predicted by the simulation code. Meanwhile a numerical simulation was in progress with FLUENT. From the first CFD results it was decided that the diffuser throat had to be moved forward from the primary nozzle exit, in order to allow a complete mixing between the primary and secondary flows, and enlarged, the losses encountered in the mixing process being higher than expected and hence the fluid density lower. This produced a second ejector design, which was manufactured and tested in 2012, showing improved performance. These results suggested a third design, with a further lengthened diffuser, which has undergone a complete testing campaign, allowing validation of the CFD results. The activity performed till now suggests that ejector refrigeration plants have a robust operation and can be easily manufactured at relatively low cost with off-the-shelf components, a part from the ejector itself, which however represents a small fraction of the system cost. However, the COP is lower with respect to absorption refrigeration and hence needs substantial improvement through detailed thermodynamic and CFD design optimization.