Supersonic ejectors are used in a wide range of applications such as compression of refrigerants in cooling systems, pumping of volatile fluids, or vacuum generation. The objective of the present paper is to mesh and simulate, in an OpenFOAM environment with an open access implicit density-based solver HiSA, the physics of the vacuum ejector, and, later, compare the results with experimental measurements. In order to achieve this a 2D axisymmetric mesh made by hexahedral cells has been created. Steady solutions have been obtained, with prescribed total pressure in primary and secondary inlets. Secondary total pressure ranges from 1 to around 0.2 bar in which the secondary flow is zero. Numerical results are compared with experimental measurement, with two flowmeter sizes for small flow rate accuracy. Two regimes are encountered. In supercritic regime the secondary is chocked and sonic flow is reached in the second nozzle. In subcritic regime, the secondary flow is subsonic. The agreement is good, although simulation tends to slightly overestimate flow rate for large values region.

A vacuum supersonic ejector is an indispensable pneumatic device placed in nearly all industrial production lines. This device, also called a zero-secondary flow ejector, is characterized by the maximum entrained flow and the minimum secondary pressure. Numerical simulations were carried out by means of the CFD toolbox OpenFOAM v8 and its solver HiSA, which uses the AUSM+up upwind scheme. A single-factor analysis of eight parameters was performed to find how the ejector’s performance was enhanced or decreased, while other parameters were fixed. Four parameters were subject to further analysis to find the geometry that improves the standalone performance of the ejector. The mixing chamber length is the parameter that most improves its performance; alone it leads to a 10% improvement. A multi-factor analysis, based on a fractional factorial design, is carried out with the four relevant parameters. Results indicate that the multi-factor analysis enhances the performance of the ejector by 10.4% and the mixing chamber length is the factor that most influences the improvement. Although a multi-factor design improves the performance, no significant relevance has been detected with respect to the mixing chamber length improvement alone. The improved performance of this device leads to a reduction in operating time and, as a consequence, results in significant energy savings.