Search

This thesis reports a two part investigation of single- and two-stage ejector systems in which the primary fluid is steam and the secondary fluid is air. The first part is an experimental investigation. The vacuum created by the ejector is strongly affected by the distance between the steam nozzle outlet and the diffuser throat section. The relation between this distance, which is called in this thesis "the nozzle optimum distance (Lop)", and the geometrical and operating parameters of the nozzle and the diffuser were investigated and forms the object of this part of the thesis. The second part is a theoretical approach. The exit Mach number for the nozzle was found by using the one-dimensional gas dynamic equations together with the first law of thermodynamics. Also a two-dimensional approach using the Method of Characteristics was used to find the exit Mach number and the characteristic net of the flow from the throat to the outlet of the nozzle. Two computer programmes were written on the basis of these two different theoretical techniques and the comparison between the results for the exit Mach number found to be 95% in agreement over the pressure range of the experimental work. A computer programme was also written using the Method of Characteristics to find the shape i.e. the characteristic net and the constant density lines within the flow of the steam jet leaving the nozzle and entering the diffuser. It is believed that the jet diameter at the point where it meets the diffuser wall, which is called in this work "the optimum jet diameter (Dop)", is strongly related to the nozzle optimum distance (Lop). When the characteristic net for the jet is drawn, its point of interception with the diffuser wall can be found and then (Dop) can be measured. This diameter (Dop) was then related to the ejector dimensional parameters and the ejector operating conditions; an equation was found to predict the optimum jet diameter from this equation (Dop)e. Then the predicted optimum nozzles distance (Lop)e was determined by using this computer program where the characteristic net meets the diffuser wall at the calculated optimum jet diameter (Dop)e. Finally, the experimentally determined value of the nozzle optimum distance (Lop) was compared to the theoretically predicted value, and the average error was found to be 1.23%.