Papers by Federico Mazzelli
IIR eBooks, Jun 18, 2018
The use of two-phase flashing ejectors to improve the COP and capacity of CO2 vapor compression s... more The use of two-phase flashing ejectors to improve the COP and capacity of CO2 vapor compression systems has become a key research topic in recent years. Although many experimental investigations focused on both the ejector performance and overall energy efficiency have been carried out, further efforts devoted to significantly enhance this device are necessary. In the present paper, the authors experimentally investigated the performance of a two-phase ejector as its main parameters, such as pressure lift and entrainment ratio, are varied. The results are finally compared with a CFD model based on a commercial solver.
International Journal of Heat and Mass Transfer, Oct 1, 2020
Abstract Carbon dioxide (CO2) is an ideal low Global Warming Potential (GWP) working fluid, and h... more Abstract Carbon dioxide (CO2) is an ideal low Global Warming Potential (GWP) working fluid, and has the potential to increase the efficiency of heat pumps and other thermal systems. Further benefits can be harnessed by using microchannel components to realize compact systems. Saturated flow boiling of CO2 in a single microchannel with a hydraulic dimeter of 1.55 mm is investigated here. The effects of mass flux, quality, heat flux, and reduced pressure on evaporation are considered over the range 100
At a first glance, the nature of this thesis may be considered somehow "sui generis". By the auth... more At a first glance, the nature of this thesis may be considered somehow "sui generis". By the author's choice, its structure is not conceived as a report of the doctoral activities, but rather as a guide that can help the reader to appreciate the phenomena and critical points arising in the context of supersonic ejector research. The rationale for this choice has progressively appeared clear in the course of the doctoral period. During the last two decades, the literature on ejector refrigeration has become mostly self-referential. However, deep understanding of the physics governing both single-and two-phase high-speed phenomena requires a strong effort in terms of "broadening the research horizons". Work of high relevance to supersonic ejectors can be found in the fields of aerospace propulsion, wind tunnel, steam turbines and turbomachinery cavitation research areas. Consequently, the character of this dissertation is shaped by the firm belief that major advancements may only come by joining together the knowledge from all these different disciplines. Under this perspective, the main results of this three-year doctoral research should not only be considered the achievements in ejector chiller performances or the developments of analytical and numerical tools, but also the opening-up to different research fields and the elaboration of a logical structure where all the themes are rationally presented. The thesis is divided into three main parts. The first chapters of each part are devoted to literature reviews and in-depth investigation of theoretical concepts. The reader who is already familiar with these aspects may skip these sections and find novel results in the corresponding second chapters. Part I of the thesis introduces the reader to the basic ideas of ejector refrigeration. In the first chapter of this Part, the state of the art as well as the history and future perspective of this technology are discussed. Chapter 2 is devoted to the presentation of the results obtained with the industrial prototype developed by the University of Florence as well as to the development of rigorous tools for the analysis and optimization of ejectors and ejector cycles. Part II is concerned with the physics and modeling of singlephase supersonic ejectors. Specifically, Chapter 3 proposes a detailed examination of the phenomena occurring in the various ejector regions. In Chapter 4, the discussion concentrates on the problems related to the numerical and analytical modeling of single-phase ejectors. The knowledge of the single-phase aspects sets the basis to move on the more complex features of two-phase supersonic ejectors, explored in Part III. In Chapter 5, the physics of high-speed condensation is thoroughly investigated both from the point of view of macroscopic and microscopic behavior. Finally, chapter 6 concludes this work by focusing on the numerical modeling of condensing ejectors.
At a first glance, the nature of this thesis may be considered somehow "sui generis". By the auth... more At a first glance, the nature of this thesis may be considered somehow "sui generis". By the author's choice, its structure is not conceived as a report of the doctoral activities, but rather as a guide that can help the reader to appreciate the phenomena and critical points arising in the context of supersonic ejector research. The rationale for this choice has progressively appeared clear in the course of the doctoral period. During the last two decades, the literature on ejector refrigeration has become mostly self-referential. However, deep understanding of the physics governing both single-and two-phase high-speed phenomena requires a strong effort in terms of "broadening the research horizons". Work of high relevance to supersonic ejectors can be found in the fields of aerospace propulsion, wind tunnel, steam turbines and turbomachinery cavitation research areas. Consequently, the character of this dissertation is shaped by the firm belief that major advancements may only come by joining together the knowledge from all these different disciplines. Under this perspective, the main results of this three-year doctoral research should not only be considered the achievements in ejector chiller performances or the developments of analytical and numerical tools, but also the opening-up to different research fields and the elaboration of a logical structure where all the themes are rationally presented. The thesis is divided into three main parts. The first chapters of each part are devoted to literature reviews and in-depth investigation of theoretical concepts. The reader who is already familiar with these aspects may skip these sections and find novel results in the corresponding second chapters. Part I of the thesis introduces the reader to the basic ideas of ejector refrigeration. In the first chapter of this Part, the state of the art as well as the history and future perspective of this technology are discussed. Chapter 2 is devoted to the presentation of the results obtained with the industrial prototype developed by the University of Florence as well as to the development of rigorous tools for the analysis and optimization of ejectors and ejector cycles. Part II is concerned with the physics and modeling of singlephase supersonic ejectors. Specifically, Chapter 3 proposes a detailed examination of the phenomena occurring in the various ejector regions. In Chapter 4, the discussion concentrates on the problems related to the numerical and analytical modeling of single-phase ejectors. The knowledge of the single-phase aspects sets the basis to move on the more complex features of two-phase supersonic ejectors, explored in Part III. In Chapter 5, the physics of high-speed condensation is thoroughly investigated both from the point of view of macroscopic and microscopic behavior. Finally, chapter 6 concludes this work by focusing on the numerical modeling of condensing ejectors.
Journal of Physics: Conference Series, 2021
In this paper, the test supersonic ejector with conjugate heat transfer in solid bodies has been ... more In this paper, the test supersonic ejector with conjugate heat transfer in solid bodies has been studied numerically. An extensive numerical campaign by means of open-source SU2 solver is performed to analyze the fluid dynamics of the ejector flowfield accounting for the heat conduction in solids. The fluid domain simulation is carried out by employing compressible RANS treatment whilst the heat distribution in solids is predicted by simultaneous solving the steady heat conduction equation. The working fluid is R245fa and all simulations are performed accounting for real gas properties of the refrigerant. Experimental data against numerical results comparison showed close agreement both in terms mass flow rates and static pressure distribution along the walls. Within the CFD trials, the most valuable flow parameters at a wall vicinity are compared: distribution across the boundary layer of the temperature and the turbulent kinetic energy specific dissipation rate, boundary layer dis...
Journal of Physics: Conference Series, 2020
Research activity on ejectors is ongoing at the University of Florence since the late nineties. T... more Research activity on ejectors is ongoing at the University of Florence since the late nineties. The most important achievement is a 40 kW ejector chiller designed according to the “CRMC” criterion. The experimentally validated CFD simulations have given some hints about some possible improvements, i.e. refine the surface finish of the ejector, study the effect of heat transfer and improve the final part of the diffuser, which in its present shape does not produce a measurable compression. The prototype has been recently filled with low-GWP refrigerant R1233zd, as a drop-in replacement of previously used R245fa. Both fluids are “dry-expanding” and hence significantly easier to model in CFD simulations. Synthetic low-GWP refrigerants may be an option for ejector chillers, due to their ability to reach below-zero temperature and high volumetric refrigerant capacity. Some lessons learned with synthetic refrigerants can be transferred to the project of a steam ejector chiller, which rema...
Energy Procedia, 2017
District heating networks are commonly addressed in the literature as one of the most effective s... more District heating networks are commonly addressed in the literature as one of the most effective solutions for decreasing the greenhouse gas emissions from the building sector. These systems require high investments which are returned through the heat sales. Due to the changed climate conditions and building renovation policies, heat demand in the future could decrease, prolonging the investment return period. The main scope of this paper is to assess the feasibility of using the heat demand-outdoor temperature function for heat demand forecast. The district of Alvalade, located in Lisbon (Portugal), was used as a case study. The district is consisted of 665 buildings that vary in both construction period and typology. Three weather scenarios (low, medium, high) and three district renovation scenarios were developed (shallow, intermediate, deep). To estimate the error, obtained heat demand values were compared with results from a dynamic heat demand model, previously developed and validated by the authors. The results showed that when only weather change is considered, the margin of error could be acceptable for some applications (the error in annual demand was lower than 20% for all weather scenarios considered). However, after introducing renovation scenarios, the error value increased up to 59.5% (depending on the weather and renovation scenarios combination considered). The value of slope coefficient increased on average within the range of 3.8% up to 8% per decade, that corresponds to the decrease in the number of heating hours of 22-139h during the heating season (depending on the combination of weather and renovation scenarios considered). On the other hand, function intercept increased for 7.8-12.7% per decade (depending on the coupled scenarios). The values suggested could be used to modify the function parameters for the scenarios considered, and improve the accuracy of heat demand estimations.
International Journal of Multiphase Flow, 2021
An impulse facility for analysis of water vapour nozzle flows using both shock tunnel and Ludwieg... more An impulse facility for analysis of water vapour nozzle flows using both shock tunnel and Ludwieg tube operating modes has been developed and tested at the University of Southern Queensland. Unique high-speed flow visualisation of the water vapour condensation shock has been acquired in the throat region of a nominally two-dimensional convergent-divergent nozzle with a 120 mm 2 throat area. This paper presents the facility performance and the time-resolved visualisation results for the nozzle flow, and the results are analysed with the aid of image post-processing tools and quasi-one-dimensional (Q1D) ther-modynamic calculations. The experiments have produced qualitative and quantitative data on the flow conditions at four different relative humidity values from 25% to 100%, and over a range of nozzle supply temperatures between 293 K and 343 K. The direct measurement of the pressures and wave speeds within the tunnel resulted in a good agreement with the ideal gas Q1D calculations, particularly for the first incident shocks and expansion waves. Differences progressively increased with the number of reflections , due to the effect of viscous dissipation and the non-ideal end-wall interaction caused by the presence of the nozzle inlet. Moreover, flow visualizations of the location and orientation of both the weak disturbance waves and the condensation shock enabled the investigation of the supersonic water vapour condensation within the nozzle of this impulse facility. Based on these measurements, the nozzle flow downstream of the condensation shock was demonstrated to have a lower Mach number relative to the dry nitrogen flow case. This phenomenon is due to the phase change heat release, which appeared to be more significant when the condensation shock intensity was higher. Although further development of the apparatus is possible, the impulse test-rig configuration that we have developed permits a relatively high flow rate of test gas for a short period of time, leading to significantly reduced operating expenses relative to a continuous flow facility with the same mass flow rate. It is therefore suggested that future studies further improve and test the impulse facility techniques for studying condensation phenomena as a potential alternative to high-power, continuous flow facilities.
Carbon dioxide (CO 2 ) is an ideal low Global Warming Potential (GWP) working fluid, and has the ... more Carbon dioxide (CO 2 ) is an ideal low Global Warming Potential (GWP) working fluid, and has the potential to increase the efficiency of heat pumps and other thermal systems. Further benefits can be harnessed by using microchannel components to realize compact systems. Saturated flow boiling of CO 2 in a single microchannel with a hydraulic dimeter of 1.55 mm is investigated here. The effects of mass flux, quality, heat flux, and reduced pressure on evaporation are considered over the range 100 < G < 500 kg m −2 s −1 ; P red = 0.70, 0.81; 15 < q < 72 kW m −2 and 0.15 < x v < 0.8. Heat transfer coefficients increase with an increase in reduced pressure and heat flux, an indication that nucleate boiling may be an important mechanism for the conditions tested. Good agreement of the data (absolute average deviation of 8.8%) was achieved with a correlation developed in this study for these conditions.
In this paper we present a set of experimental data acquired on a CO 2 flashing ejector designed ... more In this paper we present a set of experimental data acquired on a CO 2 flashing ejector designed for ex- pansion loss recovery applications. Experimental results are compared with CFD simulations obtained us- ing a Homogeneous Equilibrium Model (HEM) and a recently developed mixture model that treats both the liquid and vapour phases as compressible and metastable materials. The models are implemented within a commercial CFD software via user defined subroutines. Pro and cons of each of the two ap- proaches are critically discussed in order to understand the relative performances in terms of numerical stability and accuracy. From this comparison, it can be shown that while the HEM is faster and intrinsi- cally stable, the mixture model has the potential to better reproduce the experimental data, especially in terms of mass-flow rates predictions.
Energy, 2018
The present paper describes a novel CFD approach for the flashing of CO2 through nozzles and ejec... more The present paper describes a novel CFD approach for the flashing of CO2 through nozzles and ejectors. The novelty of the method is represented by the possibility of defining both the liquid and vapor phases as compressible materials. The properties of each phase are obtained via look-up tables calibrated against standard fluid libraries and are valid in the whole domain of interest, including the supercritical, subcritical and metastable regions. The model has been implemented within a commercial CFD solver and is completely general, i.e., it can be applied to any type of compressible multiphase flow. In the present study, the proposed approach has been validated against an experimental test-case available in literature.
A B S T R A C T In the present paper, a numerical model for the simulation of wet-steam flows has... more A B S T R A C T In the present paper, a numerical model for the simulation of wet-steam flows has been developed and implemented within a commercial CFD code (ANSYS Fluent) via user defined functions. The scheme is based on a single-fluid approach and solves the transport equations for a homogeneous mixture coupled with conservation equations for the droplets number and liquid volume fraction. The model is validated against a steam nozzle test-case and then compared with experimental data from a steam ejector with a significant amount of generated liquid phase. The simulations show a good agreement both in terms of mass flow rates and pressure profile data. Some of the modeling assumptions are also reviewed and discussed.
Non-equilibrium condensation of steam occurs in many jet and turbomachinery devices, such as supe... more Non-equilibrium condensation of steam occurs in many jet and turbomachinery devices, such as supersonic nozzles, ejectors and across the last stages of steam turbines. Wet steam models are available in many commercial CFD codes and can represent the
metastable behaviour of the flow with reasonable accuracy. Unfortunately, the use of built-in models does not allow freedom in
the choice of model parameters and settings. In the present paper, a numerical model for the simulation of wet steam flow has been
developed and implemented within a commercial CFD code (ANSYS Fluent) via user defined functions. The scheme is based on
a single-fluid approach and solves the transport equation for a homogeneous mixture flow coupled with conservation equations for
the number of droplets and liquid mass fraction. The model is compared against a well-known steam nozzle test-case.
Supersonic steam ejectors are commonly used in many applications, for example suction of non-cond... more Supersonic steam ejectors are commonly used in many applications, for example suction of non-condensable gases in steam
power plants or heat-powered chillers. In the specific case of ejector chillers, CFD studies of the ejector are necessary for
optimization of this device because ejector performances have a direct impact on the COP of the chiller. The complex ejector
flow and the high Mach numbers make characterization of the phase change inside this component difficult. Metastability effects
also have to be accounted for and some CFD commercial software provide built-in wet-steam models for this purpose. A simpler
approach for numerical modeling of multiphase ejector is the Homogeneous Equilibrium Model (HEM) in which the phase
change occurs in equilibrium conditions, i.e. metastability is neglected. These second kinds of models are still important tools for
preliminary analysis of condensing ejectors. In the present paper a comparison between commercial software wet-steam models
and an in-house developed model based on HEM is presented.
Supersonic ejectors can be profitably used in supercritical CO2 vapor compression chiller to reco... more Supersonic ejectors can be profitably used in supercritical CO2 vapor compression chiller to recover the throttling losses. Various
configurations are possible that allow great benefits in terms of system efficiencies (COP can increase up to 30%). Although the application
and testing of ejectors is not particularly difficult in these systems, the study of the internal dynamics is an extremely complex
task because all phase-change phenomena inside the ejector occur at a high level of speed and compressibility. This work presents the
analysis of various approaches that could be exploited to tackle the analysis of high-speed, non-equilibrium flashing of CO2 inside a
supersonic ejector. The limits and advantages of some previous theoretical and numerical works are highlighted. CFD results, obtained
with an in-house developed numerical model, are discussed.
Supersonic ejectors are commonly used for many applications in the field of refrigeration. In man... more Supersonic ejectors are commonly used for many applications in the field of refrigeration. In many cases, condensation or evaporation occur, either within the motive or suction flow. A very important case of condensing ejector is the steam ejector, which can be used for suction of non-condensable gases in steam power plants or in heat-powered chillers. Design techniques for ejectors are still a matter of discussion.Ideal gas models are usually employed to obtain a first set of dimensions. However, in the case of steam ejectors, the ideal gas behavior is far from being physically consistent. Other design techniques are based on interpolations of experimental data and, although empirical, can provide suitable sizing of the main ejector dimensions. Finally, a still higher level of modeling is multiphase CFD. Wet steam models are available in some commercial CFD codes and should be able to represent the non-equilibrium behavior of the flow. All these levels of analysis are surveyed and compared in this paper. CFD simulations are performed to validate a 1D design procedure and to investigate the effects of non-equilibrium condensation within highly supersonic stream of pure steam.
Supersonic ejectors can be used in heat powered chillers to transfer mechanical energy between th... more Supersonic ejectors can be used in heat powered chillers to transfer mechanical energy between the motive and the inverse
cycle. Within the ejector, momentum is exchanged between a high speed flow produced by a primary nozzle and a slow
current coming from the chiller evaporator. Due to the supersonic regime of the primary flow, the mixing of the two streams
causes significant loss and impairs the system efficiency. Up to now, second law efficiency of ejector chillers is quite low
and optimization is highly needed. The fluid dynamics of the whole ejector involves turbulent mixing, shock trains and
complex wall flow, requiring CFD analyses for an adequate description. However, in order to attempt an optimization,
mathematically workable models are advantageous.
An analytical scheme that captures the basic features of the turbulent mixing zone is discussed here in view of a Constructal
design of an ejector chiller.
Numerical and experimental analyses are performed on a supersonic air ejector to evaluate the eff... more Numerical and experimental analyses are performed on a supersonic air ejector to evaluate the effectiveness
of commonly-used computational techniques when predicting ejector flow characteristics. Three
series of experimental curves at different operating conditions are compared with 2D and 3D simulations
using RANS, steady, wall-resolved models. Four different turbulence models are tested: k–e, k–e realizable,
k–w SST, and the stress–w Reynolds Stress Model. An extensive analysis is performed to interpret
the differences between numerical and experimental results. The results show that while differences
between turbulence models are typically small with respect to the prediction of global parameters such
as ejector inlet mass flow rates and Mass Entrainment Ratio (MER), the k–w SST model generally performs
best whereas e-based models are more accurate at low motive pressures. Good agreement is found
across all 2D and 3D models at on-design conditions. However, prediction at off-design conditions is only
acceptable with 3D models, making 3D simulations mandatory to correctly predict the critical pressure
and achieve reasonable results at off-design conditions. This may partly depend on the specific geometry
under consideration, which in the present study has a rectangular cross section with low aspect ratio.
Uploads
Papers by Federico Mazzelli
metastable behaviour of the flow with reasonable accuracy. Unfortunately, the use of built-in models does not allow freedom in
the choice of model parameters and settings. In the present paper, a numerical model for the simulation of wet steam flow has been
developed and implemented within a commercial CFD code (ANSYS Fluent) via user defined functions. The scheme is based on
a single-fluid approach and solves the transport equation for a homogeneous mixture flow coupled with conservation equations for
the number of droplets and liquid mass fraction. The model is compared against a well-known steam nozzle test-case.
power plants or heat-powered chillers. In the specific case of ejector chillers, CFD studies of the ejector are necessary for
optimization of this device because ejector performances have a direct impact on the COP of the chiller. The complex ejector
flow and the high Mach numbers make characterization of the phase change inside this component difficult. Metastability effects
also have to be accounted for and some CFD commercial software provide built-in wet-steam models for this purpose. A simpler
approach for numerical modeling of multiphase ejector is the Homogeneous Equilibrium Model (HEM) in which the phase
change occurs in equilibrium conditions, i.e. metastability is neglected. These second kinds of models are still important tools for
preliminary analysis of condensing ejectors. In the present paper a comparison between commercial software wet-steam models
and an in-house developed model based on HEM is presented.
configurations are possible that allow great benefits in terms of system efficiencies (COP can increase up to 30%). Although the application
and testing of ejectors is not particularly difficult in these systems, the study of the internal dynamics is an extremely complex
task because all phase-change phenomena inside the ejector occur at a high level of speed and compressibility. This work presents the
analysis of various approaches that could be exploited to tackle the analysis of high-speed, non-equilibrium flashing of CO2 inside a
supersonic ejector. The limits and advantages of some previous theoretical and numerical works are highlighted. CFD results, obtained
with an in-house developed numerical model, are discussed.
cycle. Within the ejector, momentum is exchanged between a high speed flow produced by a primary nozzle and a slow
current coming from the chiller evaporator. Due to the supersonic regime of the primary flow, the mixing of the two streams
causes significant loss and impairs the system efficiency. Up to now, second law efficiency of ejector chillers is quite low
and optimization is highly needed. The fluid dynamics of the whole ejector involves turbulent mixing, shock trains and
complex wall flow, requiring CFD analyses for an adequate description. However, in order to attempt an optimization,
mathematically workable models are advantageous.
An analytical scheme that captures the basic features of the turbulent mixing zone is discussed here in view of a Constructal
design of an ejector chiller.
of commonly-used computational techniques when predicting ejector flow characteristics. Three
series of experimental curves at different operating conditions are compared with 2D and 3D simulations
using RANS, steady, wall-resolved models. Four different turbulence models are tested: k–e, k–e realizable,
k–w SST, and the stress–w Reynolds Stress Model. An extensive analysis is performed to interpret
the differences between numerical and experimental results. The results show that while differences
between turbulence models are typically small with respect to the prediction of global parameters such
as ejector inlet mass flow rates and Mass Entrainment Ratio (MER), the k–w SST model generally performs
best whereas e-based models are more accurate at low motive pressures. Good agreement is found
across all 2D and 3D models at on-design conditions. However, prediction at off-design conditions is only
acceptable with 3D models, making 3D simulations mandatory to correctly predict the critical pressure
and achieve reasonable results at off-design conditions. This may partly depend on the specific geometry
under consideration, which in the present study has a rectangular cross section with low aspect ratio.
metastable behaviour of the flow with reasonable accuracy. Unfortunately, the use of built-in models does not allow freedom in
the choice of model parameters and settings. In the present paper, a numerical model for the simulation of wet steam flow has been
developed and implemented within a commercial CFD code (ANSYS Fluent) via user defined functions. The scheme is based on
a single-fluid approach and solves the transport equation for a homogeneous mixture flow coupled with conservation equations for
the number of droplets and liquid mass fraction. The model is compared against a well-known steam nozzle test-case.
power plants or heat-powered chillers. In the specific case of ejector chillers, CFD studies of the ejector are necessary for
optimization of this device because ejector performances have a direct impact on the COP of the chiller. The complex ejector
flow and the high Mach numbers make characterization of the phase change inside this component difficult. Metastability effects
also have to be accounted for and some CFD commercial software provide built-in wet-steam models for this purpose. A simpler
approach for numerical modeling of multiphase ejector is the Homogeneous Equilibrium Model (HEM) in which the phase
change occurs in equilibrium conditions, i.e. metastability is neglected. These second kinds of models are still important tools for
preliminary analysis of condensing ejectors. In the present paper a comparison between commercial software wet-steam models
and an in-house developed model based on HEM is presented.
configurations are possible that allow great benefits in terms of system efficiencies (COP can increase up to 30%). Although the application
and testing of ejectors is not particularly difficult in these systems, the study of the internal dynamics is an extremely complex
task because all phase-change phenomena inside the ejector occur at a high level of speed and compressibility. This work presents the
analysis of various approaches that could be exploited to tackle the analysis of high-speed, non-equilibrium flashing of CO2 inside a
supersonic ejector. The limits and advantages of some previous theoretical and numerical works are highlighted. CFD results, obtained
with an in-house developed numerical model, are discussed.
cycle. Within the ejector, momentum is exchanged between a high speed flow produced by a primary nozzle and a slow
current coming from the chiller evaporator. Due to the supersonic regime of the primary flow, the mixing of the two streams
causes significant loss and impairs the system efficiency. Up to now, second law efficiency of ejector chillers is quite low
and optimization is highly needed. The fluid dynamics of the whole ejector involves turbulent mixing, shock trains and
complex wall flow, requiring CFD analyses for an adequate description. However, in order to attempt an optimization,
mathematically workable models are advantageous.
An analytical scheme that captures the basic features of the turbulent mixing zone is discussed here in view of a Constructal
design of an ejector chiller.
of commonly-used computational techniques when predicting ejector flow characteristics. Three
series of experimental curves at different operating conditions are compared with 2D and 3D simulations
using RANS, steady, wall-resolved models. Four different turbulence models are tested: k–e, k–e realizable,
k–w SST, and the stress–w Reynolds Stress Model. An extensive analysis is performed to interpret
the differences between numerical and experimental results. The results show that while differences
between turbulence models are typically small with respect to the prediction of global parameters such
as ejector inlet mass flow rates and Mass Entrainment Ratio (MER), the k–w SST model generally performs
best whereas e-based models are more accurate at low motive pressures. Good agreement is found
across all 2D and 3D models at on-design conditions. However, prediction at off-design conditions is only
acceptable with 3D models, making 3D simulations mandatory to correctly predict the critical pressure
and achieve reasonable results at off-design conditions. This may partly depend on the specific geometry
under consideration, which in the present study has a rectangular cross section with low aspect ratio.