7th International Energy Conversion Engineering Conference 2-5 August 2009, Denver, Colorado AIAA 2009-4522 Development of Innovative Micro Power Converter Technologies at the Austrian Institute of Technology M. Keding 1, I. Vasiljevich 2, P. Dudzinski 3, J. Gerger 4 and M. Tajmar 5 AIT Austrian Institute of Technology, A-2444 Seibersdorf, Austria Waste heat is a primary source of energy loss in many aerospace and terrestrial applications. The Austrian Institute of Technology (AIT) is presently developing two different types of micro power converters, promising high efficiencies in their respective application areas. The first converter is based on an innovative thermoacoustic Stirling engine concept without moving parts. Such a maintenance-free engine system would be particularly suitable for advanced Stirling radioisotope space power systems. The second converter is based on microturbines to use exhaust-gases for improving the overall efficiency for a number of applications. This paper will summarize our efforts on micro power converter technologies. y 0 δ k λ ω η η c = half gap width = thermal penetration depth = wavelength = angular frequency = efficiency = efficiency relative to Carnot Nomenclature T I. Introduction he Space Propulsion & Advanced Concepts department at AIT is focusing on micro propulsion and energy systems for aerospace applications with spin-offs into terrestrial markets. Since a few years, our department has become active in the area of micro energy converters leveraging from our propulsion-related developments. A number of developments originally intended for our micro rocket engine promise innovative energy recovery solutions 1. One of the main activities in this area is the development of Stirling engines and microturbines for energy recovery especially in space-related as well as terrestrial applications. This paper will give an overview of our Stirling engine developments for energy recovery as well as our energy related developments around our microturbine. II. Nano-Stirling Engines The idea behind the development of nano-stirling engines is the integration of micro rocket engine technologies into heating, ventilation and air conditioning (HVAC), automotive and mobile sectors including auxiliary power supply, computer waste heat- and power recovery devices as well as radioisotope thermoelectric generators for deep space probes. 1 Team Leader Energy Systems, Space Propulsion & Advanced Concepts, Email: marcus.keding@ait.ac.at. 2 Senior Scientist, Space Propulsion & Advanced Concepts, Email: ivanhoe.vasiljevich@ait.ac.at 3 Research Scientist, Space Propulsion & Advanced Concepts, Email: piotr.dudzinski@ait.ac.at 4 Graduate Student, Space Propulsion & Advanced Concepts 5 Head of Space Propulsion & Advanced Concepts, Email: martin.tajmar@ait.ac.at Copyright 2009 by the, Inc. All rights reserved.
Solid-state thermoelectric converters have an efficiency of 3-5 percent and Stirling engines and turbines are usually only available in the 1 kw range and above. Nano-Stirling engines that go down to 100 W are currently developed with efficiencies of 25 %. However, this technology is still quite large, i.e. such an engine has the size of a small bottle 2. If a nano-stirling engine with significant lower size and/or higher efficiencies could be developed, it could open up a completely new market for energy savings. The waste heat from single computer chips could be recovered directly or nano-engines could be distributed around waste heat and exhaust sources like car engines to efficiently cool the engines and to produce power. That is of particular interest for hybrid vehicles. Also, nextgeneration power systems could be developed by combining the nano-converters with micro combustion units using micro rocket technologies and green fuel such as hydrogen peroxide (H 2 O 2 ). Energy densities much higher than present batteries could be achieved at similar sizes. These micro power sources are considered enabling technologies e.g. for UAV s (Unmanned Aerial Vehicles). In principle, such high efficiency power converters can therefore increase the efficiency of virtually all technologies including solar power, solar cells or heat pumps. AIT s focus within this project is the investigation of Stirling-based micro-technologies for application in HVAC and satellites. The following chapter will describe our closed cycle system based on the thermoacoustic Stirling engine concept without moving parts. A. Thermoacoustic Engine In a conventional heat engine, a working fluid, such as air or helium, is caused to expand by infusion of heat and the resulting change in pressure drives a piston down the pressure gradient so that power can be extracted from the engine, i.e. by connecting the piston to a power generator. In addition to a working piston a conventional Stirling engine utilizes a displacer piston to move the working fluid to the hot and cold heat exchangers. In the interest of miniaturization and reduction of complexity, a novel double acting β-type Stirling engine without any mechanical displacer pistons based on the calculations of Kumagai et al. 3 was designed, manufactured and tested. This means that the only remaining moving part is the working piston, which exhibited stable oscillation at 22 Hz and an amplitude of 16 mm at ambient pressure using air as working fluid (Piston diameter = 16 mm; Displacer diameter = 18 mm; Displacer length = 115 mm). Having determined that this approach is viable, it is nevertheless clear that the moving piston severely limits the degree upon which the engine can be miniaturized further. If Stirling engines with higher power density and higher efficiency are the goal, modifications to the mechanism of power extraction, especially the piston, need to be researched. Since the piston is a massive component which oscillates back and forth, the speed with which it can move and extract power is limited by material and technological constraints. If a method were available which provided the function of a piston, namely the compression, displacement and expansion of a gas, but without the accompanying movement of a large mass in comparison to the displaced gas, heat engines with a significantly higher power density could be produced. Indeed, sound waves have the same functionality as a massive piston in the sense that they compress, displace and decompress gas, albeit at the speed of sound, which is significantly faster than any conventional piston can move. This in turn means that the process of extracting heat from a heat source and converting it into mechanical energy can go through a full cycle in much shorter time, so that the power output of the heat engine is increased. Apart from the benefit of faster power-cycles, replacing a conventional piston by high-amplitude sound waves means that there are no moving parts, such as sliding seals around the piston, which can wear out. Thermoacoustic engines (and their counterpart, thermoacoustic refrigerators) therefore consist of only very few parts, leading to a lightweight and inexpensive design with a long lifetime. While these attributes are advantageous in any machine, they are especially sought after in space applications, such as a power generator on satellites. The following chapters expound on the concept of thermoacoustic heat engines and the rationale for the chosen design of a short standingwave type. In the next step, we focus on the demonstrator hardware which uses single-piece heat exchangers manufactured by means of electric discharge machining (EDM) for maximum heat transfer to the active inner lamellae. Finally, the next steps in the development as well as planned applications are presented. 1. Concept The guiding principles of the design concept involve the following factors: high power density miniaturized design, for easy integration in existing hardware and waste-heat utilization maintenance-free, for long-term application in challenging environments, i.e. space Efficiency higher than 10 % of Carnot efficiency, in order to provide a superior alternative to thermoelectric generators.
The following Table 1 summarizes the design parameters: Table 1. Thermoacoustic engine design parameters. Property Symbol Value Gas He, N 2, air, Ar Pressure p 1-50 bar Length L 30 cm Resonator volume V < 1 l Acoustic Power max P max 100 W Efficiency relative to Carnot η c 10 % Though traveling-wave thermoacoustic engines in general display a higher efficiency than their standing-wave counterparts, they also exhibit more intricate phenomena, such as Gedeon- and Rayleigh- streaming 4, which require special fine-turned components to counter the deleterious effects. Because of this added level of complexity in designing traveling wave engines, and because the more simple standing-wave thermoacoustic engines are in fact used as 1 st stage in a more advanced cascade thermoacoustic engine 5, preference was given to the development of a standing-wave thermoacoustic engine. The main components of a thermoacoustic engine are a hot heat exchanger, HX an ambient heat exchanger, AX a stack of closely spaced material with a low thermal conductivity, ST a resonating cavity, RES The heat exchangers serve to transfer heat from an outside source to the gas in the engine and therefore consist of a material which has a very high thermal conductivity, such as copper-alloys. Furthermore, the interface area between the heat exchangers and the gas needs to be as large as possible in order to heat and cool the gas efficiently. The so-called stack is placed between the hot and cold heat exchangers so that a temperature gradient forms along its length. When the temperature gradient exceeds a certain critical threshold, spontaneous, resonant oscillations commence in the thermoacoustic engine which manifest themselves as extremely loud sound within the cavity. The sound waves which are generated by these means can easily reach intensities of 120-160 db which is comparable to the noise of a jackhammer and more. Luckily, this high-frequency noise is encapsulated in the sealed resonance cavity, so that the engine as a whole is quiet. The material of choice for the stack has a low thermal conductivity in order to minimize the heat transfer from hot to cold through the stack material. Most often the stack is made from stainless steel or nickel alloys, though glass or ceramics can be used as well. The resonance cavity, finally, determines the operating frequency of the engine. Neglecting viscous effects, the power density of a thermoacoustic engine scales directly with its operating frequency, so that, in general, high frequencies are favorable when aiming for a compact, yet powerful engine. There exist many methods of fabricating heat exchangers and stacks for thermoacoustic machines, such as using porous materials 6,7 woven steel meshes 8, bonded metal foils 9 or metal foils which have been rolled into a spiral 10 with interspersed spacers as well as other methods 11. 2. Innovation The concept engine presented herein uses electric discharge machining (EDM) for the fabrication of the heat exchangers as well as for the stack. All components are of the parallel-plate type, because this approach yields a 10-20 % higher level of efficiency and power than other geometries according to numerical studies 10. The advantage of using EDM as manufacturing technique consists in its ability to produce a stack, or heat exchanger, from a single piece, so that the repeated thermal cycling during the lifetime of the engine does not induce internal stresses due to differing thermal expansion values of a composite part, such as an approach using sacrificial bonding layers or spacers. Because the structures in heat exchangers and plate-stacks are delicate by design in order to increase the active surface, any internal stresses will lead to large deformations in the thin plates. Because, however, the stack of plates poses a resistance to the gas flow, an inhomogeneous distribution of the plates with accompanying varying gap widths between them will affect the heat transfer between the gas particles and the material surface by a disproportionate amount, thus greatly affecting the efficiency. Indeed, this warping of plates after several thermal cycles is one of the main reasons why experimental engines with initially close tolerances
deteriorate over time. In addition to the improved resilience during thermal cycling, the approach of using heat exchangers made from a single piece improves the heat transfer from the outside to the inner plates since there is no junction between different materials or sub-components which may act as thermal barrier or resistance. The stack used for the thermoacoustic engine presented herein is shown in Fig. 1. In addition to the one-piece design of the critical components, as a further measure to reduce potential internal stresses, the plates have been halved along the centerline, so that each plate is free-standing from the inner surface of the heat exchangers and stack. In this manner, the plates have room to expand or contract horizontally in accordance with their local temperature, and do not have to yield in the vertical (warping) direction if short-term thermal gradients manifest themselves. This design therefore promises to maintain a homogeneous geometry and accordingly stable performance throughout a long lifetime and after repeated thermal cycling. Figure 1. Section view of the EDM-manufactured stack. The whole part of Fig. 1 is eroded out of a single block in order to minimize the risk of thermally induced deformations where different materials connect to each other. Furthermore, the plates are cut in the middle in order to further minimize thermal stresses. During initial testing, using a standard electrode diameter of 0.25 mm in the wire-edm processing yielded a gap width 2y o of 0.36 mm in Copper, so this value served as baseline for further calculations regarding the design. In the following discussion, it is assumed that the temperatures of the ambient and hot heat exchangers are 20 C and 400 C respectively. Using helium as working fluid and the given resonator length of 30 cm, the resonance frequency is calculated at 2150 Hz. When the thermally-induced oscillations have begun, a variable pressure component is superimposed over the mean pressure P m in the resonator chamber. A typical conservative value for the amplitude of the variable pressure component P A is 5 % of the mean pressure though also values of 10 % can be achieved. In the following we shall calculate with just 5 %. Since 5 % is assumed to be a realistic lower value for the magnitude of the pressure variation, the heat exchangers have been dimensioned with lengths of 10 mm, in order to accommodate larger pressure variations up to ~9 % as well. The potential downside of this extra length is that, if only low pressure amplitudes can be generated, the excess length adds to viscous losses and reduces the power output of the engine. The length of the stack is chosen to be double the length of the heat exchangers, 20 mm, and is to be placed at λ/10 from the end of the closest resonator to achieve a compromise between high power and small viscous losses in the stack. In a standing wave, the pressure and density oscillations of the gas are in phase and therefore do not contribute to acoustic power. When a thermal boundary condition is introduced in the standing wave, however, i.e. from a thin plate of a material with a heat capacity which is large compared to the gas, gas molecules which are roughly one thermal penetration depth away from the plate s surface experience a phase change in the range of 90, so that this layer of gas contributes to heat and work flux. The thermal penetration depth δ k is the distance which heat can travel through the gas during a characteristic time of 1/ω. Gas particles for which y 0 /δ k «1 or y 0 /δ k»1 do not contribute to heat and work flux appreciably.
Since the pressure shall remain variable in the range between 1 and 50 bar, it is important to check whether the condition that y 0 /δ k remains close to 1 is true throughout the envisioned pressure range. Figure 2 shows the value of y 0 /δ k as function of the pressure and demonstrates that the ratio lies in a range which will indeed lead to a net heat and work flux due to the aforementioned phase shift. In order to effect a large net heat and work flux in a standing wave engine, the ratio should be close to 1. In practice, values of 2-4 are typically used in standing-wave-type thermoacoustic devices. Figure 2. Plot of the ratio y 0 /δ k between half-gap width y 0 and the thermal penetration depth δ k as function of helium pressure in bar for the chosen design. The expected efficiency of the presented setup is shown in Fig. 3 as percent of Carnot s efficiency and as function of the temperature of the hot heat exchanger. In order to start the thermoacoustic oscillations a typical temperature difference between ambient and hot heat exchanger of 200-400 C is necessary, depending on the quality of resonance and viscous losses. From the graph, one can deduce that for a temperature of 400 C, the efficiency of the standing wave thermoacoustic engine is in the range of 13 %, which is compatible with the design goals presented earlier. Figure 3. Percent of Carnot s efficiency as function of the temperature in C of the hot heat exchanger, assuming a constant temperature of 20 C on the ambient side. The efficiency of the engine increases as the temperature on the hot side is lowered, but the maximum power of the engine decreases accordingly, so, in practice, a trade-off between power and efficiency needs to be made.
3. Breadboard Demonstrator Hardware The entire thermoacoustic engine design contains merely 7 parts (not counting screws and instrumentation), so that assembly of the manufactured parts is fairly straightforward. An overview of the engine is shown in Fig. 4. The hot heat exchanger is drawn in red, the ambient heat exchanger in blue, the stack is green and the resonators are drawn in gray. The overall length of the device is 34 cm, with a resonance cavity of 30 cm. The resonance frequency when helium is used as working fluid is 2150 Hz and the maximum expected acoustic power output lies in the range of 100 W. At a later stage, this generated acoustic power will be used to drive an alternator in order to produce electric power. At present, no power transducers have been included yet, so that the level of generated acoustic power will be used as measure of the potential power output. Figure 4. CAD-model of the small-scale thermoacoustic engine developed at AIT/SPA (overall length of 34 cm). The resonator cavities (gray) were lathed from stainless steel and the resonator adjacent to the hot heat exchanger (red) was thickened near the interface in order to retain its mechanical stability at elevated temperatures and pressures, since softening of the material is one of the prime concerns at very high temperatures (~750 C). The larger area of contact to the resonator poses a drain on the generated heat, so that efficiency may suffer somewhat, but the considerably lower thermal conductivity of stainless steel compared to copper limits the impact. Both heat exchangers were made from CuCr1Zr (ElmedurX), a copper alloy which displays greater strength at elevated temperatures than industrial copper, yet retaining its exceptionally high thermal conductivity. The outer side of the hot heat exchanger is prepared to hold a heater-coil of NiCr-wire, but can also be heated with an open flame or by solar concentrators. Vice versa, the ambient heat exchanger is designed to be cooled by a steady, circulating water flow through the cooling fins, but may also be operated in air-cooled mode if a smaller amount of cooling power is required. The stack (see Fig. 5), finally, consists of Inconel718, a NiCr-alloy with superior material properties at high temperatures and low thermal conductivity in order to achieve good thermal separation between the two opposing heat exchangers. Figure 5. EDM-shaped hot side heat exchanger and stack (outer diameter of heat exchanger = 64 mm).
The entire device is assembled without any welding, so that all connecting parts are equipped with a groove in which a gold-wire is embedded to serve as pressure seal. The parts are held together by 6 long threaded rods running from one resonator half to the other; their function is to apply even pressure on the heat exchangers and the stack which are sandwiched between the resonators as seen in Fig. 6. Figure 6. Standing-wave thermoacoustic engine with installed heat exchangers, stack and resonators with an overall length of 34 cm. The electric heating wire is of the type Thermocoax SEI 20/100, with a hot wire length of 1 m and an outer diameter of 2 mm, which is wound 6 times around the hot heat exchanger. The heater wire has been embedded in Ceramabond 865, consisting of aluminum nitride in order to provide good thermal contact to the heat exchanger. The piezoresistive pressure sensor is a Keller Series 23/25 with an absolute pressure range of 1-50 bar and a maximum frequency range of 5 khz, so that the oscillations with an expected frequency of 2 khz can be resolved. The temperature was measured at the outer surface of the heat exchangers with type K thermocouples which have been bonded to the copper with a ceramic glue with high thermal conductivity (Ceramabond 865). 4. Test Results The small-scale thermoacoustic demonstrator was operated with the gases helium, nitrogen and argon, in order to study the effect of different molecular weights on the working parameters, such as pressure amplitude, power and operating frequency. Regardless of the working fluid, the best performance of the engine was observed for pressures at or below 10 bar and the temperature gradient necessary to activate the oscillations was in the range of 330-380 C as shown in Fig. 7.
Figure 7. Temperature difference between hot and cold heat exchanger during operation. Using helium as working fluid, the operating frequency was measured at 1800 Hz. The pressure amplitude of the sound wave was only 3 % of the mean pressure, which hints at major dissipation processes at these high operating frequencies. In contrast, the sound frequencies when using nitrogen or argon were 620 Hz and 560 Hz respectively, resulting in a pressure amplitude of 9 % of the mean pressure as shown in Fig. 8. Figure 8. Summary of test results of the small-scale thermoacoustic prime mover using different working fluids.
Despite the comparatively low pressure amplitude of the oscillations when using helium as working fluid, the device generated an acoustic power of 14.5 W due to the high operating frequency. With nitrogen and argon, the acoustic power was 19.4 W and 20.4 W respectively. Considering the electric heating power of 440 W, these values translate into efficiencies of 6.1 %, 8.6 % and 8.1 % relative to Carnot. While the latter two values after initial testing are reasonably close to the envisioned efficiency of 10 %, it is clear that the obtained efficiency for helium is subpar. The likely reasons for this behavior, along with methods of improvement are the topic of the next chapter. 5. Discussion Earlier it was mentioned that the distribution of gap widths in the stack plays an important role in the efficiency, so that an inspection of the eroded plates was performed. Measurement of the gap width by means of a scanning electron microscope yielded the following distribution: Table 2. Analysis of gap widths and plate thicknesses of the heat exchanger. Overall Left half Right half Plate thickness [µm] 292 ± 9 297 ± 8 287 ± 7 Gap width [µm] 388 ± 39 382 ± 22 395 ± 50 Group 1 Group 2 351 ± 19 438 ± 27 The analysis of the distances in the stack shows that the plate thickness is very uniform and close to the mark of 0.3 mm. The overall gap width is close to 0.39 mm, but with a significantly larger standard deviation. Because the plates in the stack have been separated by a central gap in order to minimize the danger of warping, a closer analysis regarding the symmetry of the left and right halves was made. It was discovered that in the right half of the plates in the stack the gap width displays an unusually large standard deviation and that two equally represented sub-groups can be identified. It turns out that every second gap is significantly larger (by ~100 µm) than the adjacent smaller gaps, but that the overall gap width is close to the value on the left half. This asymmetry is likely a side-effect of the transition of the EDM-wire from one row to another. Figure 9 shows a close-up of plates of the Inconel-stack along with the measured distances. Figure 9. Close-up of plates of the Inconel-stack. Inspection of the gap widths after thermal cycling of the heat exchangers and stack demonstrated that the gap widths had not changed and that no warping of plates took place. Thus the concept of fabricating the parts out of a single piece and slitting the filigree plates along the middle is successfully validated. The systematic variations in gap width, however, deteriorated the overall efficiency of the device, so that future versions of the stack and heat exchangers will require higher accuracy during discharge machining. The efficiency and power of the engine will improve considerably when the gap width in the stack is distributed evenly throughout the entire area.
The main reason for the underperformance during testing with helium as working fluid is the surface roughness of the plates in the heat exchangers and the stack. Figure 10 shows a detail view of the surface of such a plate. As the gas oscillates back and forth across this jagged surface, a significant portion of energy is expended to overcome the forces of friction and resulting turbulence at large speeds, such as in the case of helium with a resonance frequency of 1.8 khz. While the surface roughness also affects nitrogen and argon, the dissipative effect is less pronounced because the heavier gases exhibit a much lower resonance frequency in the range of 0.6 khz. This explains their comparatively better performance during testing, despite their otherwise inferior properties compared to helium. As consequence, additional manufacturing steps will be introduced in order to ensure a smoother surface finish on the plates of the stack and heat exchangers. The reduction in dissipative losses will result in higher efficiency, larger power and a smaller temperature gradient in order to effectuate the initial oscillations. Figure 10. Surface roughness of single-pass eroded Inconel plates. The breadboard demonstrator hardware successfully proved that it is possible to build a small-scale, highfrequency thermoacoustic prime mover with heat exchangers and a stack which are resistant to warping during repeated thermal cycling and which works with a multitude of gases at various pressures. During testing it became evident that the length of the heat exchangers (10 mm) befitted the oscillatory displacement amplitude of nitrogen and argon much better than helium. In effect, the heat exchangers were twice as long as would have been ideal for helium gas at these frequencies, so that the extra surface merely added to viscous losses while the gas was shuttled back and forth. Future versions of such small-scale thermoacoustic prime movers will therefore focus on a specific working fluid and optimize the design accordingly in order to achieve higher efficiency and power. Finally, in order to achieve acoustic power in the range of 100 W while retaining the current outer dimensions of the device, a smaller gap width in the stack is necessary so that a larger portion of the gas contributes to sound amplification at high mean pressures. 6. Further Development and Application The thermoacoustic engine presented herein is the first step in a series of miniaturized engines and serves to demonstrate the working principle as well as to validate the concept of a stack and heat exchangers which are highly resistant to repeated thermal cycling. The experience and data gathered within this project will be used to further miniaturize the design, to improve the gas-flow properties within the heat exchangers and to increase the efficiency as outlined in the previous chapter. If the system can be made sufficiently small and can operate at very small temperature differences, the thermoacoustic device can be applied to extract power from computer chips without any moving parts. Another application is a highly efficient Stirling cycle power converter for space applications. Operating a probe on the surface of a planet will require a lightweight, reliable, nuclear-powered energy source capable of functioning in the dense atmosphere at high temperatures (e.g. 460 C for the Venus). A multi-faceted mission study for such a planet was completed at NASA Glenn Research Center in December 2003 12 that resulted in the conceptual design of a helium-charged, kinematic Stirling converter, which is powered by radioisotope-based, General Purpose Heat Source (GPHS) modules. The converter is configured to cool sensors and generate 100 W of electrical power with an estimated total efficiency of 23.4 %.The converter rejects waste heat at a hot sink temperature of 500 C. Using a fully thermoacoustic engine without any moving parts and vibrations, as described herein, may provide a feasible alternative for missions in which a lower system complexity and a high level of scalability is desirable.
III. Microturbine Expander The development of micro propulsion systems for space applications including micro-rocket engines is the main expertise in our department. It was natural to extend our efforts towards the development of micro power generators, since the power density in such microcombustion chambers is very large. A microturbine was developed to produce the power for the pumps of a rocket engine earlier in a previous project 13. The successful testing of the microturbine triggered our interest towards micro power converters in general since they may be used to increase the overall energy efficiency for a number of applications including compression and absorption heat pumps, air conditioning units, thermal storage devices or even building elements. This chapter will give an overview of our microturbine development in the past as well as the development of the current 3 rd generation turbine for application towards HVAC. 1. Turbine Development History Two types of turbines with different rotor diameters were investigated for the first generation. The smaller one (10 mm diameter) were manufactured using a LIGA process (Lithography, Electroplating, and Molding) in combination with EDM process (Electrical discharge machining) and the bigger one (23 mm diameter) has been machined with a high-speed milling machine using aluminum as material for the turbine. The two turbine blades have different optimum rotational speed, 80,000 for the bigger one instead of the 250,000 for the smaller turbine. The achieved results after manufacturing and running tests of the first turbine generation showed a low power output due to sealing problems, friction losses, stray fields of the coupling magnets and inefficient magnetic coupling between the turbine and the generator. Several decisions were made in order to improve the system. Figure 11 shows the progress between the first generation turbine (65 mm height and 40 mm length) and the second generation turbine (55 mm height and 35 mm diameter) 1. First generation Second generation Gas Outlet Inlet Inlet Gas Inlet and Turbine Blade Bearings Leakage 1 Leakage 1 Leakage Leakage 2 2 Magnetic Coupling Nozzle Generator Figure 11. First generation turbine (left) and second generation turbine (right). The second generation turbine allowed higher power outputs at lower rotational speeds of 10,000 to 15,000 rpm due to the reduced leakages by redesigning the gas inlet and outlet and the reduced dynamic loads by using two bearings instead of one. Several improvements on the third generation were done after precise consideration and analysis of 2 nd generation turbine tests. This includes the use of stronger magnetic coupling, the new design of the turbine rotor blades, the use of a stronger generator and the redesign of the inlet and outlet. 2. 3 rd Generation Turbine Development for HVAC AIT has successfully developed an application case for use of microturbine technology in HVAC systems in a former study. The chosen application case is a microturbine boosted transcritical CO 2 heat pump cycle for domestic applications like tap water heating or high temperature room heating for retrofitted buildings. For this application case a thermodynamic study of a CO 2 cycle was done in order to get a design specification for the microturbine which is used in our turbine design 14.
Associated with the low flow rate of working fluid in CO 2 heat pumps, two types of turbine geometry were considered. The first type is the 90 inward radial turbine and the second type is a turbine based on the Pelton turbine. The solution with a micro Pelton turbine seems to be favourable because of the high pressure differences between inlet and outlet (max. 70 bar) and the low flow rate of working fluid (0.02 kg/s). The Institute for Thermodynamics and Energy Conversion of the Vienna University of Technology started designing layouts of rotor blades and stator blades/nozzle for these requirements. One big challenge for the layout was the fact that the working medium changes its state from liquid to wet steam during the expansion process. Due to these facts and further manufacturing reasons, the design will be based on the so-called Turgo -principle, which means a half- Pelton turbine. The turbine blades shown in Fig. 12 were made with a laser milling process. Figure 12. Half-Pelton wheel. Figure 13 shows the design (left) and the prototype (right) of our 3 rd generation turbine. The application of this system is as a throttle valve replacement in CO 2 heat pumps. An increase in efficiency (> 10 %) and/or production of electricity (20 100 W) seemed to be possible with this turbine. Figure 13. 3 rd generation prototype of a microturbine for energy recovery. 14 The results of the dedicated testing of the turbine under laboratory conditions and the optimisation of some components are shown in Fig. 14. A maximum electrical power of 52 Watts was reached with a starting pressure of 70 bar at 32,500 rpm in gas operation mode. A maximum electrical power output of 36 W was reached with a gaswater mixture of 70 bar at 23,000 rpm. The lower power output could be explained with higher friction losses generated by water which flows between the turbine blades-housing box and inside of the magnetic coupler housing.
Figure 14. Power output in gas operation mode (left) and in gas-water operation mode (right). 14 The next steps will be the substitution of the generator with a fully integrated power system without friction losses and the testing of the system with CO 2. Another application for this converter is a high power refillable energy cell based on H 2 O 2 microcombustion described in Chapter II. Such a system could have additional features such as its immediate refill capability within 10 seconds, the supply of internal heat e.g. for high altitude UAV s and the supply with clean water. In summary, the development of our microturbine is a good example for technology transfer from space to earth. In fact, the lessons learned by developing microturbines for HVAC could be helpful for developing such turbines for space or aerospace applications too. IV. Conclusion A compact thermoacoustic engine with a theoretic power capability of 100 W and an efficiency of more than 10 % relative to Carnot s efficiency was presented in detail. The main advantages of this approach include simplicity of design high power density self-starting properties no moving parts which can wear out long lifetime, especially in maintenance-free environments, such as space. In the course of the development of the thermoacoustic components, the topic of warping of the fragile plates in the stack and heat exchangers from thermal cycling was especially addressed by using EDM to manufacture the parts from a single piece and by adding a tension inhibiting cut along the length of the plates. The presented small-scale thermoacoustic prime mover has been shown to work with the working fluids helium, nitrogen and argon and has demonstrated an acoustic power output of 20 W at a maximum efficiency of 8.6 % relative to Carnot. Future developments regarding the thermoacoustic activities include an improved design for higher efficiency and power as well as a transition to ever smaller dimensions, so that heat from micro-circuitry as well as highly localized heat sources may be utilized. Furthermore, the developed standing wave engine may be used in a more efficient combined standing- & traveling-wave engine. The design of a laboratory model microturbine for energy recovery in CO 2 heat pumps has been presented in the third chapter. The main component of the system and its performance has been tested and the results seem to be very promising. Investigating the influence of operational parameters (mass flow rate and system pressure), a significant improvement of the rotational speed and hence the power output should be obtained. Additionally, an advanced version of the turbine system is in preparation, which will feature an improved design with lower weight and volume and increased efficiency.
Acknowledgments The work described in this paper was performed for the Austrian Institute of Technology. We also wish to thank M. Monsberger and I. Malenkovic (AIT / Sustainable Thermal Energy Systems) for their support in HVAC systems and R. Willinger and K. Käfer (Vienna University of Technology/Institute for Thermodynamics and Energy Conversion) for their support in microturbine blades design. References 1 Keding, M., Tajmar, M., Dudzinski, P. and Reissner, A., "Development of Innovative Hydrogen and Micro Energy Solutions at the Austrian Research Centers", AIAA 6th International Energy Conversion Engineering Conference, IECEC, Cleveland/Ohio, AIAA-2008-5643, 2008 2 Wood, J. G., Carroll, C. and Penswick, L. B., "Advanced 80 W e Stirling Convertor Development Progress", Space Technology and Applications International Forum (STAIF 2005), Albuquerque, 2005 3 Kumagai, K., Yamasaki, H., "Performance Prediction of Linear Stirling Power Generator with Two Displacers", AIAA 6th International Energy Conversion Engineering Conference, IECEC, Cleveland/Ohio, AIAA-2008-5643, 2008 4 Backhaus, S, Swift, G. W., "A thermoacoustic-stirling heat-engine: Detailed study", Journal of the Acoustical Society of America 107 (6) 3148-3166, June 2000 5 Gardner, D. L., Swift, G. W., "A cascade thermoacoustic engine", Journal of the Acoustical Society of America 114 (4), Pt 1,1905-1919, October 2003 6 Arnott, W. Pat, Bass, Henry, E., Raspet, Richard, "General formulation of thermoacoustics for stacks having arbitrarily shaped pore cross sections", Journal of the Acoustical Society of America 90 (6), 3228-3237, December 1991 7 Arnott, W. Patrick, Sabatier, James M., Raspet, Richard, "Sound propagation in capillary-tube-type porous media with small pores in the capillary walls", Journal of the Acoustical Society of America 90 (6), 3299-3306, December 1991 8 Swift, G.W., "Analysis and performance of a large thermoacoustic engine", Journal of the Acoustical Society of America 92 (3), 1551-1563, September 1992 9 Backhaus, S., Swift, G.W, "Fabrication and use of parallel plate regenerators in thermoacoustic engines", Proceedings of IECEC 01, 2001 10 Swift, G.W. "Thermoacoustics: A unifying perspective for some engines and refrigerators", Journal of the Acoustical Society of America, 2002 11 Swift, G.W., Keolian, R.M., "Thermoacoustics in pin-array stacks", Journal of the Acoustical Society of America Vol. 94, No 2, Pt 1, 941-943, August 1993 12 Thieme, L. G. and Schreiber, J. G., "NASA GRC Stirling Technology Development Overview", Space Technology and Applications International Forum (STAIF 2003), Albuquerque, 2003 13 Scharlemann, C., Schiebl, M., Marhold, K., Tajmar, M., Miotti, P., Guraya, C., Seco, F., Kappenstein, C., Batonneau, Y., Brahmi, R., and Lang, M., "Test of a Turbo-Pump Fed Miniature Rocket Engine", AIAA Joint Propulsion Conference, AIAA- 2006-4551, 2006 14 Keding, M., Dudzinski, P., Tajmar, M., Willinger, R., Käfer, K., "Development of a µ-scale Turbine Expander for Energy Recovery", ASME Turbo Expo 2009: Power for Land, Sea and Air, GT 2009, Orlando, Florida, 2009