Fabrication and testing of refrigeration using engine waste heat in Mechanical


The objective of this research project was to investigate the potential of developing an environmentally friendly pressure exchange (PE) refrigeration system. There are three basic refrigeration cycles that are suitable for domestic, commercial, and vehicular air conditioning: reverse Rankine cycle, absorption cycle, and ejector refrigeration. The most common conventional technology uses the reverse Rankine cycle, which provides excellent refrigeration but at a substantial cost because it is powered directly by mechanical energy taken from an electric motor or, in the vehicular application, from the internal combustion engine. If electricity is provided to the air conditioning system through fossil fuel generated electrical power plants, which often have efficiencies of less than 40 percent, and transmitted long distances through the power grid with consequent energy dissipation, the cost to the nation in terms of energy use and the emission of global warming effluents is enormous. Furthermore, reverse Rankine cycle refrigeration generally uses hydro-fluorocarbon based refrigerants, which deplete the earth’s ozone layer and further contribute to global warming. In the case of vehicular air conditioning, the reverse Rankine cycle uses valuable mechanical energy taken directly from the output shaft of the internal combustion engine. Because less than 40 percent of the fuel’s thermal energy is converted to mechanical energy and the reverse Rankine cycle uses this precious resource, vehicular air conditioning in its current form is a substantial contributor to energy consumption and to the emission of global warming effluents. Therefore, if a thermally based vehicular refrigeration cycle can capitalize on the 60 percent of the fuel’s thermal energy that is wasted, an enormous benefit would be derived by society in terms of energy saved and reduced emissions to the environment.
The absorption cycle is a thermally energized refrigeration system that is capable of using a vehicle’s waste heat. Modern technology has increased the coefficient of performance (COP) of absorption cycle units for large scale applications to reasonably high levels, particularly in applications where the absorption cycle is integrated with power generation. However, such systems are extremely complex, costly, and space consuming. At the small scale of vehicular air conditioning, the COP tends to decline, and absorption cycle refrigeration for vehicles tends to be expensive, heavy, space consuming, complicated, and generally impractical, as witnessed by its absence from the vehicular market.
Ejector refrigeration is a thermally energized refrigeration system, which has the added potential benefit of being capable of using environmentally friendly refrigerants, such as water. It has been used extensively for commercial air conditioning and refrigeration, particularly in applications where a source of waste heat is readily available. Several automotive manufacturers, such as Toyota, Renault, and Ford Visteon, have shown some interest in it. The advantages of ejector refrigeration systems are that they are very simple and have a low capital cost. For vehicular applications, studies have shown that by using appropriate heat exchangers, there is sufficient energy in the exhaust gas to provide ejector refrigeration for a typical vehicle. This technology can be scaled down to sizes appropriate for vehicles. The main disadvantage of conventional ejector refrigeration systems, however, is that they suffer from a low COP. The low COP is important even if there is abundant waste energy available, for example in vehicular applications, because all of the excess thermal energy used to power the system must be rejected in the condenser (Foa, 1960). If the COP is low, the size of the condenser becomes very large. This, in turn, increases the cost of the system and makes packaging difficult in vehicular applications.
Analysis of the steam-ejector refrigeration cycle reveals that the steam requirement and the COP of the cycle depend upon the efficiency of one key element: the ejector. Nearly a century of research and development on the steady-flow ejectors that have been used in refrigeration systems has brought us near to the pinnacle of this technology. The physical mechanism by which these steady-flow ejectors operate is the turbulent entrainment between the primary (driving) flow and the secondary (driven) flow. This entrainment mechanism is inherently dissipative of energy, and little can be done to improve it. Thus, after more than 100 years of research, conventional ejectors and ejector refrigeration using conventional ejectors probably has attained its ultimate level of performance, which is not adequate for society’s needs and never will be. Thus, even though there is an enormous societal need for an efficient, thermally energized air-conditioning/refrigeration system that uses environmentally benign refrigerants and is low cost, compact, and highly efficient, neither the absorption cycle nor the ejector refrigeration cycle can meet the needs. A comprehensive review of the literature reveals that there is no other technology that appears to be ready to fill the void.
A central hypothesis of our research under the current grant is that, if an ejector with a compressor efficiency on the order of 50 percent or better could be made, the COP of the ejector refrigeration system would increase dramatically, thereby making ejector refrigeration highly competitive with, or superior to, reverse Rankine cycle and absorption cycle systems in terms of efficiency, capital cost, size, weight, and benefit to the environment.
For vehicular applications, such a system would increase vehicular energy efficiency by using waste heat. It would be low in cost, small in size and weight, and would use environmentally friendly refrigerants, such as steam. It would reduce substantially the amount of global warming effluents emitted by vehicles in the summer, and it would take over the vehicular market rapidly.
In commercial and domestic air conditioning applications, because ejector refrigeration is thermally based, the air conditioning system would be amenable to providing air conditioning by burning natural gas on site. This could provide overall improvements in energy efficiency by eliminating electrical transmission losses and losses from power plant machinery, and also would provide reductions in the release of global warming effluents (CO2) by virtue of the use of energy derived from low carbon natural gas (CH4) rather than energy from high carbon crude oil or coal burned at power plants.
The inspiration for this EPA-supported program was the idea that there is a mechanism for direct flow induction, other than turbulent mixing, that does not rely on highly dissipative turbulent mixing, but rather on the thermodynamically reversible work of interface pressure forces through a mechanism known as pressure exchange (PE). PE was defined by Foa (1960) as “any process whereby a body of fluid is compressed by pressure forces that are exerted on it by another body of fluid which is expanding.” Unlike turbulent mixing, PE relies on the work done by pressure forces exerted by an expanding energetic fluid on a compressed, less energetic fluid across the interface between them. Such energy transfer can be accomplished only in non-steady flow, because pressure forces against stationary interfaces can do no work. Because, in general, the PE process is non-dissipative, an ejector-like device using this operating principle can be highly efficient. Although the overarching goal of this research program was to investigate the potential of obtaining an environmentally friendly PE refrigeration system, the assumption has been that development of such a refrigeration system would be a straightforward engineering exercise if a novel, high-efficiency, ejector-like device could be obtained. The emphasis of our research program, therefore, has been on trying to obtain a fundamental understanding of the PE process in various flow conditions, both subsonic and supersonic, and trying to obtain information needed to design a replacement for a conventional ejector that: (1) is suitable for ejector refrigeration; (2) is of substantially higher efficiency; and (3) can bring on a new era of thermally driven, environmentally friendly, high efficiency refrigeration.
Our literature surveys, before and during our grant period, have shown that there has been a considerable amount of research on two types of PE devices. The first well-known PE device is the “wave-rotor,” which is a fairly complex device that attempts to use PE by filling a relatively long tube with low-energy fluid and then suddenly opening a valve and allowing high-energy fluid to enter. The expansion of the high-energy (primary) fluid compresses and drives the low-energy (secondary) fluid. To achieve these benefits, the wave rotor has a series of tubes arranged on the periphery of a rotor. As the tubes rotate, their relative, instantaneous position causes ports on either end of the tubes to open and close, allowing any one of high-energy primary fluid, low-energy secondary fluid, de-energized primary fluid, or energized secondary fluid to enter or leave the tubes. The device is ingenious and created much excitement in the 1950s and 1960s. Intensive research was done by Brown Boveri, who developed turbochargers, which have been and continue to be used on a number of vehicles. The National Aeronautics and Space Administration (NASA) has been very interested for a number of years in using wave rotors as topping cycles for improving the efficiency of gas turbine engines. Although wave rotors have achieved a good measure of success, they never have attained the high efficiencies sought through PE because of the prevalence of three mechanisms of dissipation, which are unavoidable and inherent in the wave rotor. They are: (1) throttling losses caused by the opening and closing of the ports; (2) friction losses, which result from the rotating dynamic seals around the ports; and (3) the presence of strong normal shock waves. It appears to be well accepted in the field of wave rotors that these dissipation mechanisms are inherent and limit the maximum attainable compressor efficiency of the device. Commercial applications of wave rotors find niches, not as a result of their high efficiencies, but by virtue of other attributes, such as the elimination of “turbo-lag” in superchargers and their ability to directly receive very high temperature products of combustion.
The second well-known PE device is the “crypto-steady pressure-exchange thrust augmenter,” which has been described by Foa (1960). Foa sought to avoid the throttling losses inherent in non-steady flow devices with the hope of attaining the high efficiencies theoretically possible through PE. Foa observed that PE can be created without valving and ports by the use of what he termed a “crypto-steady” flow (i.e., a flow that is steady in a certain moving frame of reference but is non-steady relative to the laboratory). He demonstrated that such flows can be obtained by the use of a free-spinning rotor having canted jets of primary fluid. In such a flow, the flow appears steady relative to the rotor but is non-steady relative to the laboratory. Relative to the laboratory, the primary fluid jets emanating from the rotor form a helical pattern whereby the secondary fluid becomes entrapped in the interstices of the helices and, by the use of an appropriate shroud, is forced into a duct or channel. Thus, work is done by the expanding primary fluid on the compressed secondary fluid by the pressure forces acting across the helical boundary between the two fluids. Much work was done on this concept for aircraft and marine propulsion. All of this work, however, was for applications in which the pressure rise was very small but the ratio of secondary to primary fluid mass flow rates was very high. Under these conditions, high thrust augmentations were obtained, and the value of PE in thrust augmentation was verified. The refrigeration application, however, called for very high pressure rises with small mass flow ratios.
Our research program started by attempting to use Foa’s crypto-steady PE concept under flow conditions appropriate to refrigeration. Armed with the knowledge that crypto-steady PE has the potential of providing society with a highly efficient means of compressing a low-energy fluid through direct contact with a relatively high-energy fluid, thereby avoiding the complexity and the dissipation associated with intervening machinery inherent in conventional compressors, our overarching goal was to determine what configuration such a device might have. To achieve this goal, it was essential to develop an understanding of the underlying physics that control and limit the performance of such a device. It must be understood that, prior to this research program, there never has been an attempt to obtain such a device for high-pressure rise applications, such as refrigeration. The only previous work was done by researchers in the propulsive thrust augmentation area, having an entirely different performance requirement (low-pressure rise/high mass flow ratio.)
Summary/Accomplishments (Outputs/Outcomes):

The following activities were central to the program. The overview of the activity will be described first, and the specific activity will be described next:
Aerodynamic Invention of PE Flow Induction Devices. We are particularly proud of the novel technologies devised during the course of our grant. In this activity, several inventions were conceived, and several U.S. patents were obtained. Because there was no prior guidance from the technical literature as to what should constitute the optimal flow configuration, a wide range of possible aerodynamic flows providing PE were investigated. Consideration was given to aerodynamic and thermodynamic efficiency, mechanical design issues, compactness, and manufacturability. Both radial flow and axial flow systems were investigated. Furthermore, crypto-steady flow patterns were generated by the use of rotors having imbedded nozzles (as was suggested by Foa) and by the use of fixed nozzles providing fluid that was deflected by rotating vanes (invented under this grant). Both subsonic and supersonic configurations were studied. Our supersonic flow configurations revealed substantial advantages in terms of solving sealing and bearing problems that plagued previous subsonic designs, despite the introduction of a new loss mechanism (shock waves). The following are some specific innovations that emerged from the grant.
  • Ejector refrigeration system using PE technology.
  • PE Ejector with radial flow rotating nozzle (subsonic and supersonic) with vaneless diffuser.
  • PE Ejector with balanced thrust—symmetric primary/secondary inlet radial flow rotating nozzle with vaneless diffuser.
  • Supersonic PE Ejector with wedge-vane free-spinning rotor and axial flow.
  • Supersonic PE Compressor-Expander for use as a turbo-charger for internal combustion engines and fuel cell pressurization.
  • Fuel Cell Pressurization System.
  • Supersonic Ramp-Vane PE Ejector.
Simulation of Flow Behavior. Using various commercial computational fluid dynamics (CFD) codes, the PE process in both elementary models and complex geometries were studied in depth. It was found that the state-of-the-art in commercial CFD codes is adequate to model the complex flow structure, and this was used as an analytic tool to develop a deeper understanding of the PE process and its exploitation in real machinery. Because the essence of achieving our objective of attaining a new level of ejector performance is the minimization of entropy generation through all mechanisms, the simulations were invaluable in identifying loss mechanisms that might lead to deterioration in performance. Specific activities included:
  • One-dimensional PE in shock tube.
  • Two-dimensional PE with a supersonic flat plate.
  • Validation simulations with cone flow, with and without coaxial supersonic flow, to evaluate the ability of the commercial codes to resolve supersonic mixing layers accurately, shock waves, expansion fans, and other structure in three dimensions.
  • Simulation of radial flow/rotating nozzle PE ejector having a vaneless diffuser. Subsonic and supersonic jet primary flows into a subsonic secondary flow were investigated.
  • Simulation of two-sided primary/secondary inlet, balanced thrust radial flow/rotating jet PE ejector with vaneless diffuser.
  • Simulation of supersonic primary axial flow wedge-vane rotor.
  • Preliminary simulations of supersonic primary, supersonic secondary ramp-vane rotor having axial inlet and radial outlet.
  • Study of the benefits of free-spinning operation and the flow physics of off-design behavior.
  • Effects of primary/secondary temperature differential.
  • Effects of under-expanded or over-expanded primary nozzles.
  • Effects of different primary/secondary fluid molecular weights.
Experimental Diagnostics. During the course of the grant, many test rigs were designed and fabricated. Some of these experiments were conducted on complete ejectors, and some were conducted on subsystems or on experimental devices suggested by analytical models. Some experiments were used to validate CFD codes. Detailed internal flow diagnostics were performed using schlieren flow visualization and laser anemometry, and extensive instrumentation was used to determine overall ejector performance. For most of the experimental work, air was used as the working fluid, primarily because of the relative ease of using available diagnostic tools. Experiments with steam, however, were conducted by industrial collaborators. These tests were important in demonstrating real-world physical problems facing the technology and led to innovations that avoided these difficulties. In particular, bearing and seal technologies were explored; this ultimately will provide needed information when commercialization is considered.
The following lists several of the activities undertaken in the experimental diagnostics portion of our research program:
  • Schlieren flow visualization of supersonic flows over various wedge-vane configurations. Studies of mixing layer development, shock formation, and effects of under and over expansion.
  • Fabrication and diagnostics of PE Ejector with radial flow rotating nozzle (subsonic and supersonic) with vaneless diffuser.
  • Fabrication and diagnostics of PE Ejector with balanced thrust—symmetric primary/secondary inlet radial flow rotating nozzle with vaneless diffuser.
  • Fabrication and diagnostics of Supersonic PE Ejector with wedge-vane free-spinning rotor and axial flow.
  • Experimental studies of PE in a shock tube.
  • Laser anemometry diagnostics of flow patterns in a PE Ejector.
  • Development of gas bearings for a PE Ejector.
  • Evaluation and use of compliant foil bearings for use in a PE Ejector.
  • Baseline experimental diagnostics of a conventional ejector.
  • Development of motorized test rig for studying PE in three dimensions.
  • Design, fabrication, and testing of an ejector refrigeration system.
  • Experimental evaluation of a wedge-vane PE ejector refrigeration system (in collaboration with JFE Engineering Corporation).
Refrigeration System Analyses. The underlying assumption of our research was that if a breakthrough could be obtained in ejector efficiency, then, automatically, a corresponding breakthrough in the COP of an ejector refrigeration system would occur. Nevertheless, comparative studies were conducted to determine goals on ejector efficiency to make ejector refrigeration competitive with absorption cycle and reverse Rankine cycle refrigeration systems in terms of COP. In addition, studies were conducted to determine how the design of sub-components of the ejector refrigeration system would vary in size and configuration if the ejector efficiency were improved. This is a crucial aspect of using ejector refrigeration in automotive applications. Specific activities undertaken in this area include the following:
  • Comparative study between PE ejector refrigeration and absorption cycle refrigeration to establish goals for ejector efficiency to enable ejector refrigeration to provide comparable performance (albeit at much lower capital cost).
  • Design of a PE ejector refrigeration system.
  • A study on the correlation between ejector efficiency and condenser size in an ejector refrigeration system.
  • Design of a steam experimental facility for studying PE ejector refrigeration.
The overarching accomplishment of this research program has been to expand the knowledge base of PE energy transfer and the various ways in which it may be implemented practically. The benefits and obstacles have been elucidated, and the technology has been brought to a level of understanding whereby implementation in real devices is within grasp. PE has the potential to be a transformative technology in several energy-intensive areas of great importance to society, including refrigeration, internal combustion engines, fuel cells, water desalinization, and others.
In its purest form, PE offers a non-dissipative flow induction process that is highly efficient, extremely compact, and low in capital cost. All of these benefits are derived from the fact that PE is a direct fluid-fluid mode of energy exchange that eliminates all intermediary processes and hardware that are present in conventional modes of energy transfer.
In the real world, entropy is generated by a variety of mechanisms. The success or failure of PE technology to meet society’s needs hinges on determining what these mechanisms are and learning how to manage or control them. This research program, through both simulations and experiments, has sought to gain such an understanding and to apply this knowledge to the design of practical machinery.
As an historical illustrative example of a somewhat analogous technology, consider the gas turbine. It was invented by John Barber in 1791, but was not implemented in a practical machine until 1939. As late as 1941, a distinguished National Academy of Sciences committee on gas turbines, whose members included Theodore von Karman, concluded that “In its present state… the gas turbine could hardly be considered a feasible application to airplanes...”The ultimate success of the gas turbine engine hinged on learning to compress and expand compressible fluids with a minimum amount of entropy generation. It took many years and much research to attain acceptable levels of performance. Crypto-steady PE similarly is a technology with great potential, but much research is needed to learn how to manage and minimize the generation of entropy at various steps in the process and to understand the aerodynamic, thermodynamic, and mechanical advantages and disadvantages of many possible configurations. This grant served as a pioneering effort towards making PE technology serve future society. Although this grant was focused on the application of PE to environmentally friendly refrigeration, the technology will have equal importance in several other energy-intensive applications, such as fuel cell pressurization, hydrogen reforming, water desalinization, internal combustion engine turbo-charging, topping cycles, and many others.
The following are some specific findings:
  • If a PE ejector could be developed having an ejector efficiency above approximately 50 percent, ejector refrigeration would attain a COP comparable with absorption cycle refrigeration.
  • If an ejector refrigeration system were developed having a COP comparable to a commercially available absorption cycle system, it would be considerably more compact, be more environmentally friendly (i.e., using only water as the refrigerant), and have a substantially lower capital cost per ton of refrigeration provided.
  • PE ejector refrigeration is an excellent candidate for vehicular air conditioning using waste heat. Our research has shown that there is more than enough heat rejected from an automotive engine to provide adequate air conditioning for the vehicle.
  • The impact of using waste heat to power vehicular air conditioning on the release of global warming effluents is substantial if applied on a nationwide basis. Our studies, however, have shown that the only candidates available for such an application are ejector refrigeration and absorption cycle refrigeration. The former currently is too inefficient, and the latter does not scale well to the vehicular range and is far too bulky, heavy, and costly for this application. A rule of thumb in the industry is that absorption cycle refrigeration is competitive for 20-ton units and higher—far larger than that required in the vehicular environment.
  • One of the major obstacles to using any thermally energized refrigeration system in a vehicle for air conditioning is that the condenser must reject to the atmosphere the heat provided to energize the system in addition to the heat removed from the air conditioned space. It must, therefore, be considerably larger, which presents cost and packaging difficulties. As the COP of the system increases, the heat rejection requirement, and correspondingly the condenser size, is reduced. In the case of ejector refrigeration, there is a direct correlation between COP and ejector efficiency. As stated earlier, ejector refrigeration systems are well known to have low COPs because of the low ejector efficiency. Hence, it follows that if the ejector efficiency is low, a requirement for an excessively large condenser results. This was confirmed by conversations with industrial contacts in the auto industry. Our analyses have shown that if the ejector efficiency can be increased, a substantial reduction in the physical size of the condenser of the refrigeration system is possible. Thus, if ejector efficiency by the use of PE technology can be increased substantially, the feasibility of using waste heat-energized ejector refrigeration would become practical for vehicles. If such a system using waste heat were adopted widely, there would be a substantial impact on the quantity of global warming effluents released by automobiles in the summer.
  • One of the great advantages of ejector refrigeration is its suitability for use with low molecular weight refrigerants (e.g., water), which are environmentally benign; however, ejector refrigeration also can be effective with the use of conventional refrigerants.
  • PE ejectors provide momentum and energy transfer through two processes which occur essentially in series. First, PE takes place very rapidly after the primary and secondary fluids come into contact. Second, turbulent mixing occurs and tends to equalize any differences in energy or momentum that exist after PE. A conventional ejector depends entirely on turbulent mixing. The advantage of the PE ejector is that, depending on the design, a high proportion of the energy exchange process can occur through reversible means, thereby reducing the total production of entropy in comparison to the conventional ejector.
  • PE Ejectors based on the rotating nozzle concept, as shown in Figure 1, can be aerodynamically and thermodynamically efficient, particularly when used in a radially outward flow configuration. This mode is amenable to use with a radial vaneless diffuser to recover the high kinetic energy of the fluid following PE.
  • The free-spinning rotor speed is the speed at which no torque is introduced to the fluid by the rotor. In this state, the rotating nozzles on the rotor serve only to create moving interfaces across which PE takes place between primary and secondary fluids, and no energy is added to the fluid by the rotor. We have determined that the design goal should be to approach the free-spinning speed.
  • By canting the nozzles and by providing frictionless bearings, the rotor can be self-driving at the free spinning speed. However, bearing friction works to reduce the speed below the free spinning value. A motorized drive of the rotor to make up for bearing friction can be provided but is unnecessary if the bearings are close to frictionless.
  • It was determined that very high speeds are desirable. A survey of related technology and stress analyses on the rotors revealed that operational speeds on the order of 150,000 rpm are reasonable.
  • Because the rotor does not do work on the fluid, the radial loading is very light, which makes the rotor amenable to use with gas bearings. Gas bearings can meet the low friction requirements of PE. Plain gas bearings, however, were shown to develop instabilities at high rotational speeds.
  • Collaboration with Mohawk Innovative Technology, Inc., demonstrated that compliant foil bearings can be designed to meet the low friction requirements and produce stable operation. Such bearings recently have been used in turbomachinery applications at speeds exceeding 250,000 rpm.
  • For a rotating nozzle rotor PE Ejector, providing frictionless seals proved very challenging. As seen in Figure 1, a high-pressure differential across the rotor requires very stringent sealing on the periphery of the rotor, which is hard to achieve without excessive friction.
  • For a rotating nozzle rotor PE Ejector, because of the high-pressure differential across the rotor, thrust management proved very challenging. The combination of high axial thrust and high rotational speeds produced very stringent demands on the thrust bearings, which we never were able to overcome.
  • Although CFD simulations showed that the radial flow ejector of Figure 1 has excellent potential, the mechanical difficulties proved overwhelming for us to attain good performance in our testing. We have continued, nevertheless, to study the flow computationally where sealing and thrust management can be “turned off.” It is possible that, in the industrial environment where there is great expertise in bearing and seal design coupled with superior fabrication resources, these problems may be amenable to solution.
  • Thrust management can be alleviated by the use of a double inlet/matched pressure configuration, whereby both primary and secondary fluids are introduced on opposite sides of a plane of symmetry, and the discharge is radially outward. A test rig was built on this principle but was found to be unsatisfactory because the sealing problem was doubled, thus introducing more friction.
  • To overcome the combined problems of dynamic sealing and thrust management, a revolutionary idea emerged: the use of supersonic flow to create interfaces for PE. The design is shown in Figure 2. It is seen that the primary flow emerges supersonically from a stationary nozzle and impinges on the conical forebody of a free-spinning rotor having wedge-shaped vanes. There are no dynamic seals and, because of the high kinetic energy of the fluid, the thrust forces against the rotor are reduced drastically in comparison to the rotating nozzle PE ejector shown in Figure 1. The rotor is preceded by a cone, which generates a weak oblique shock wave. Other weak oblique shocks are created over the wedge vanes. Because the vanes are canted, the rotor is self-driving and does not do work on the fluid—it merely creates rotating interfaces between which primary and secondary fluids can exchange energy.
  • Schlieren flow visualization, Figure 3, and numerical simulations revealed that the shape of the wedge-shaped vanes has a great influence on the flow pattern surrounding it, in which the PE interfaces are formed. It was found that by introducing curvature selectively, the strength of shock waves could be ameliorated.
  • Both computational simulations and experimental results confirmed the basic hypothesis that supersonic flow patterns can be established through which PE takes place. Figure 4 shows how the PE process is enhanced by increasing the rotational speed.
  • For both subsonic rotating jet and supersonic rotating vane configurations, it was found that if the rotor is free spinning (i.e., no torque introduced through the rotor) and, prior to PE, neither primary fluid nor secondary fluid has angular momentum, then after PE takes place through the mutual deflection of both primary and secondary flows to a common orientation in the rotor frame of reference, in the laboratory frame of reference, the de-energized primary fluid acquires negative angular momentum, while the energized secondary fluid acquires positive angular momentum.
  • The fact that after PE, but prior to substantial mixing, primary and secondary fluids have components of angular momentum in the axial direction that are of opposite sign offers a possibility of separating the primary and secondary fluids after energy exchange has taken place. This finding was of particular importance because being able to separate the fluids after energy exchange provides an opportunity to develop a new class of compressor-expanders with application far beyond that of ejectors. Such applications include turbochargers for internal combustion engines, expanders for air cycle refrigeration, compressor-expanders for fuel cell pressurization, and the like. Pursuing this concept, we patented a PE compressor-expander, which potentially is very important for the above-cited applications. A sketch of our invention is shown in Figure 5 in a form designed for automotive supercharging.
  • The compressor-expander of Figure 5 could be used as an environmentally friendly air-cycle refrigeration device providing cold air. If high-pressure air contained in a reservoir at room temperature is applied to the supersonic nozzle of Figure 5 (primary fluid), the device could be designed to undergo nearly isentropic expansion to atmospheric pressure. Under these conditions, the primary air would exit the compressor expander at a very low temperature, which is dependent on the design parameters. If the low-energy secondary fluid were similarly introduced at atmospheric temperature and pressure, it would undergo nearly isentropic compression and would, therefore, exit the device at high pressure and temperature. Thus, the device is very amenable to serve as a heat pump capable of providing either heating or cooling, depending on which exit flow one applies to environmental control. Because this device requires the availability of a high-pressure air supply, it would be very suitable for aerospace applications, where compressor bleed air is readily available from the gas turbine engines.
  • We became aware that our invention of a novel PE compressor-expander could find important application in our nation’s vehicular fuel cell program. In 2002, President Bush made proton-exchange membrane (PEM) fuel cell vehicles a centerpiece of his FreedomCar program. A highly efficient, compact, and low-cost compressor expander is needed desperately in the development of PEM fuel cells for vehicular propulsion applications because the power density of a fuel cell increases with pressure, and an efficient pressurization technology for this purpose is not available. Incorporating our PE compressor-expander, we invented and patented a new fuel cell pressurization system, which is shown in Figure 6. This system not only can provide efficient pressurization to the fuel cell, but it can be used to assist in the fuel cell moisture management as well as to provide compressed air to an onboard steam reformer.
  • Our computational fluid dynamic simulations revealed that in the supersonic PE ejector, the growth of the mixing layer separating the primary and secondary flows can interfere with the PE process. Furthermore, it was observed that it is advantageous for PE to be initiated as close to the apex of the conical forebody of the rotor as possible. Based on this finding, we conceived a new double-cone apex design and applied for a patent. The new design is shown in Figure 7.
Figure 1. Radial Flow Rotating Nozzle PE Ejector
Figure 2. Schlieren Photos of Mach 2.5 Nozzle Flow Over Various Vane Configurations
Figure 3. Supersonic Wedge-Vane Rotor PE Ejector
Figure 4. Simulation Showing Effect of Rotating Speed on Flow Induction Process
Figure 5. PE Compressor-Expander
Figure 6. Fuel Cell Pressurization and Moisture Management System Using PE Technology
Figure 7. Double Cone Forebody Supersonic PE Ejector With Radial Diffuser Discharge

Our original hypothesis was that the well-known steam jet refrigeration, which has been long discarded for domestic and transportation applications because of its low COP, could become competitive or surpass state-of-the-art reverse Rankine Cycle and absorption cycle systems if a highly efficient ejector based on thermodynamically reversible PE could be obtained. Success in this endeavor would provide the nation with an enormous payoff in terms of reduction in global warming effluents and in displacing environmentally harmful refrigerants with the use of steam. Although this project has not succeeded in developing such an ejector, we have identified the many difficulties in rotating jet type PE ejectors and have invented the supersonic PE ejector, which overcomes many of these difficulties but introduces new obstacles. This program has been very helpful in gaining an understanding of this totally new technology, which we created under EPA support and hope to bring to wide commercialization in the not too distant future. We will continue to seek support to make this happen. In terms of the potential environmental impact, it also is important to note that if we were to succeed in gaining the quantum leap in ejector efficiency that we are seeking through effective use of supersonic PE, the impact on a wide range of energy-intensive technologies beyond refrigeration would be enormous. They include turbocharging for internal combustion engines, fuel cell pressurization, water desalinization, and many other areas. We believe that this EPA grant was instrumental in creating this new technology, and we believe that the payoff to the nation and the world will be significant in the years to come.
Next Post »