One is emerging from perhaps the most deliberate and least colourful engineering fields of all: gas turbine engineering. Gas turbines are internal combustion engines, like the ones that drive cars, except that they use a rotating shaft or rotor instead of pistons "reciprocating" in cylinders. This makes their operation smooth and steady, which lowers maintenance costs and increases reliability. Though they became practical only sixty years ago, today gas turbines are one of the keystone technologies of the civilization. As jet engines, they deliver most of our air transport, while stationary gas turbines are responsible for an increasing fraction of our electrical power generation.Partly because of this critical role, gas turbine engineers tend to innovate one tiny step at a time. In a field where liability exposures and development costs both can run into nine and ten figures, any kind of sweeping enthusiasm makes people nervous. Still, that doesn't mean engineers can't dream on their own time. In the spring of 1994, when a MIT turbine engineer named Alan Epstein found himself sitting in a jury pool, he started to think about what it would take to build the smallest possible jet engine. He concluded that in theory the device could be shrunk a lot, perhaps to the size of a collar button.
If you attached a microgenerator to the turbine, essentially creating a tiny power plant, the combination would act like a battery, making power at twenty to fifty times the rate of anything you could get at the hardware store. (Because there is much more energy per gram in burning hydrocarbons than in the electrochemicals that usually go in batteries.) Depending on how much fuel came with the turbine, a laptop might run for months on a single charge; a cellphone, for half a year. Given the insatiable appetite our portable gizmos have for batteries, the microturbine project suddenly became very interesting. The U.S. Army, which badly wants to reduce the weight carried by their "soldier systems", agreed to write the checks.
By 1995 the microturbine project was humming along. Unlike a conventional gas turbine design job, where each member is a world-class expert on one (but only one) phase of the process, all the researchers on this project were starting from the same place: how to make engines less than a hundreth the size of a conventional turbine design. For instance, for a gas turbine to work well, the tips of its rotors have to turn at about the speed of sound, or five hundred meters a second. The smaller the diameter of a turbine, the faster the rotor has to spin to move its tips at that speed. A conventional jet engine can get there with a few tens of thousands of revolutions a minute. The microturbine had to do much better: closer to two million rpm, or twenty thousand revolutions a second.
This awe-inspiring number raised all kinds of questions. For one: How was the rotor going to be attached The usual solution to this problem would be some sort of bearing, but what material could handle that level of abuse And even if such a substance existed, how would you make the bearings or keep them in place Eventually, after many failures, the team discovered clever ways for the rotor to use its blistering speed to lift itself up during operation, essentially making it fly in place, so that no material bearings were needed. The project required such innovations constantly, radical ideas too new for anyone to be expert on them.
Over the next seven years the project made amazing progress, considering that designing a conventional jet engine usually takes five years. Today actual working models exist, though the microturbine is not quite ready to be handed over to a manufacturer. (One of the remaining problems is exactly how to cool the exhaust to a level comfortable for consumer use .The success of the microturbine project has inspired a whole R&D sector in micropower devices. The Defense Department alone is funding well over a dozen projects, from microfuel cells and micropiston engines to microrockets. The University of Wisconsin is even looking at a micronuclear reactor. (One of the attractions is that tiny jet engines deliver ten times the thrust per unit weight of a conventional turbine, which means the huge cost airplanes now pay to haul their engines around might be radically reduced.)
It is influenced by fluid and structural mechanics, and by material, electrical and Thermal power systems encompass multitude of technical disciplines. The architecture of the overall system is determined by thermodynamics while the design of the systemâ„¢s components fabrication concerns, the physical constraints on the design of the mechanical and electrical components are often different at micro scale than at more familiar sizes so that the optimal component and system designs are different as well. Most thermodynamic systems in common use today are variations of the Brayton (air), Rankine (vapour, Otto, or Diesel cycles. The Brayton power cycle (gas turbine) was selected for the initial investigation based on relative considerations of power density, simplicity of fabrication, ease of initial demonstration, ultimate efficiency, and thermal anisotropy. A conventional, macroscopic gas turbine engine consists of a compressor, a combustion chamber, and turbine (driven by the combustion exhaust) that powers the compressor, and can drive machinery such as an electric generator. The residual enthalpy in the exhaust stream provides thrust. A macro scale gas turbine with a meter diameter air intake generates power on the order of 100 MW. Thus, tens of watts would be produced when such a device is scaled to millimeter size if the power per unit of airflow is maintained. When based on rotating machinery, such power density requires (1) combustor exit temperatures of 1300-1700 K; (2) rotor peripheral speeds of 300-600 m/s and thus rotating structures centrifugally stressed to several hundred MPa (the power density of both fluid and electrical machines scales with the square of the speed, as does the rotor material centrifugal stress); low friction bearings; high geometric tolerances and tight clearances between rotating and static parts; and thermal isolation of the hot and cold sections. These thermodynamic considerations are no different at micro- than at macroscale. But, the physics influencing the design of the components does change with scale, so that the optimal detailed designs can be quite different. Examples include the viscous forces in the fluid (larger at microscale), usable strength of materials (larger), surface area to volume ratios (larger), chemical reaction times (invariant), realizable electric field strength (higher), and manufacturing constraints (planar geometries).
There are many thermodynamic and architectural design choices in a device as complex as a gas turbine engine. These involve trade-offs among fabrication difficulty, structural design, heat transfer, fluid mechanics, and electrical performance. Given that the primary goal is to demonstrate â€œ that a high power density MEMS heat engine physically reliable, the design philosophy adopted is that the first engine will be as simple as possible, trading performance for simplicity. For Example, the addition of a heat exchanger transferring heat from the turbine exhaust to the compressor discharge fluid (a recuperated cycle) offers many benefits including reduced fuel consumption and relaxed turbo machinery performance requirements, but it introduces additional design and fabrication complexity. Thus, the baseline design is a simple cycle gas turbine generator. While this engine is the simplest of gas turbines, it is an extremely complex and sophisticated MEMS device. Arriving at a satisfactory design requires heavy dependence on simulation of the mechanical, thermo fluid, and electrical behavior to achieve the required levels of component performance and integration .The baseline engine design is illustrated in Figure 1.The engine consists of a supersonic radial flow compressor and turbine connected by a hollow shaft. Gaseous H2 fuel is injected at the compressor exit and mixes with air as it flows radially outward to the flame holders. The combustor discharges radially inward to the turbine whose exhaust turns 90 degrees to exit the engine nozzle. A thin film electric induction starter-generator is mounted on a shroud over the compressor blades and is cooled by compressor discharge air. Cooling air is also used to thermally isolate the compressor from the combustor and turbine. The rotor is supported on air bearings. The following sections briefly discuss component design considerations.
MATERIALS AND MECHANICAL DESIGN
Conventionally sized engines, constructed from titanium and heavily cooled nickel and cobalt-based super alloys, are stress-limited in the rotating components. Nonmetallic such as silicon (Si), silicon carbide (SiC), and silicon nitride (Si3N4) offer substantial improvement in strength-to-density ratio and temperature capability, but large parts with acceptable properties have proven difficult to manufacture from these materials. However, they are readily available in essentially flaw-free form for micro scale fabrication so that significantly superior material performance is available for micro-heat engines than can now be realized in conventionally- sized devices. In addition, because of the small length scales required here, material which are unsuitable for a large heat engine due to thermal shock considerations (e.g. aluminum oxide), would be usable in a micro engine given a fabrication technology . Silicon is suitable for the compressor (600 K) but cannot operate at the combustor discharge temperature needed (1300-1700 K) without cooling. SiC can operate uncooled but SiC fabrication technology is much less developed than that for Si. The baseline design assumes uncooled SiC for simplicity but a cooled Si design is also under study. The individual components are being demonstrated in Si while SiC manufacturing technology is being developed. Since the properties of such materials are a strongly influenced by the details of their fabrication, material testing is an integral part of this program .g/sec implies airfoil and passage heights on the order of 200-300 microns as in figure 2.Deep reactive ion etching was used to produce the turbine shown in Figure 3, which has a 4mm rotor diameter and 200 micron span blades. The rim of the 300-micron thick disk serves as a journal bearing. This unit is a rotor dynamics test piece. With the addition of a generator on the back surface of the disc, it becomes an 80-watt turbine generator. Also, using only known process steps, a strawman process simulation yields wafers of completed engines, including a freely turning rotor, without additional assembly. It is a complex and aggressive process requiring 7 aligned wafer bonds, 20 lithography steps, and the deposition of 9 thin film layers.
TURBO MACHINERY AND FLUID MECHANICS
Considerations of engine thermodynamic efficiency, combustor performance, and turbine viscous losses suggest that compressor pressure ratio should be relatively high. Since both the pressure ratio and the centrifugal stress in the rotor scale with the square of the peripheral Mach number, the pressure ratio per stage of compression is set by the allowable material stress. Material property values in the literature are consistent with a 500 m/s rotor tip speed, which was therefore adopted as a baseline. A 4:1 pressure ratio compressor has been designed to operate at this speed. Current fabrication technology largely restricts complex curvatures to in plane, which inhibits the use of the high degree of three-dimensionality typically employed in centrifugal turbomachinery to improve efficiency and reduce material stresses. However, the usable material strength is higher at microscale. Also, this flow regime is unusual in that it is supersonic (Mach 1.4) but laminar (Reynolds number 20,000). Three-dimensional fluid calculations suggest that this machine should achieve an adiabatic efficiency of about 70%. To facilitate detailed measurement of the turbomachinery fluid mechanics, a 75:1 geometrically scaled up test rig has been built. It operates at the same Mach and Reynolds numbers as the microturbomachinery.
Air breathing combustion requires fuel injection (and evaporation if a liquid), fuel-air mixing, and chemical reaction of the mixed reactants. The time required for these processes (the combustor residence time) sets the combustor volume. In large engines, the residence time is typically 5-10 ms. Most of this is for fuel mixing; chemical reaction times are a few hundred microseconds or less. In order to expedite the engine development process, hydrogen was selected as the baseline engineâ„¢s fuel. Hydrogen offers rapid mixing and chemical reaction times, and flammability over a wide range of fuel-to-air ratios. By operating at a low fuel-to-air ratio, the peak combustor temperature can be reduced to levels compatible with uncooled SiC construction (1600 K), eliminating the requirement for the complicated cooling geometries needed on large engines. A combustor with the geometry of Figure 1 has been built and tested. It has demonstrated the predicted levels of performance over a wide range of temperatures and mixture ratios. The data agree with numerical simulations that suggest that complete combustion can still be achieved with a factor of two reductions in combustor volume . Work is now beginning on a hydrocarbon fueled catalytic combustor
BEARINGS AND ROTOR DYNAMICS
Low friction bearings are required to support the rotor against fluid and electrical forces, rotor dynamics, and externally applied accelerations while operating at speeds of over two million rpm. Gas film, electrical, and hybrid gas electrical bearing concepts were examined. Gas bearings were selected for the baseline engine based on superior load bearing capability and relative ease of fabrication. A journal bearing supports the radial loads and thrust plates support the axial loads. The physical regime that the microgas bearings operate in is unusual in several regards: the peripheral speed of the bearing is transonic so compressibility effects are important; the ratio of inertial to viscous forces (Reynolds number) is high; the surface area of the bearing is very large compared to the mass of the rotor; and the journal length-to-diameter ratio is quite low. The net effect of these influences is a journal bearing well outside existing theory and empirical design practice. Magnitude higher than the critical frequency (spring-mass damper equivalent) of the rotating system. Sub critical operation would require submicron-operating clearances, which are difficult to fabricate and incur viscous losses greater than the engine power output. The design adopted uses a ten-micron journal gap to reduce losses to a few watts but is linearly unstable at some speeds. Numerical simulations indicated, however, that this design would operate satisfactorily in a nonlinear limit cycle. Turbine-driven rotor dynamic test rigs have been constructed both at 1:1 microscale (Figure 3) and at 26:1 macroscale (to facilitate detailed instrumentation). Preliminary data confirm that the rotor does operate in a stable limit cycle. As a precaution, an electric damper is being designed to augment the bearing stability should it prove desirable.
A motor-generator starts the gas turbine and produces the electrical power output. Integrating the motor-generator within the engine offers the advantages of mechanical simplicity since no additional bearings or structure are required over that needed for the engine and cooling air is available. Either electric or magnetic machines could be used. Here, an electric machine was chosen due to considerations of power density, ease of microfabrication, and high-temperature and high-speed operation. The baseline design is a 180-pole planar electric induction machine mounted on the shroud of the compressor rotor. Simulations suggest that such a machine can produce on the order of 20-40 watts with an electrical efficiency in excess of 80%. The major source of loss in the machine is viscous drag in the rotor-stator gap.
In a first phase of the project, the problem has been scaled down to a turbine powered by compressed air. Compressor, combustion chamber, and generator have been left out and will be addressed in a later phase. The micro turbine is a single-stage axial impulse turbine. Expansion of the gas takes place in the stationary nozzles and not between the rotor blades. This type of turbine has been chosen because of its simple construction.
Figure 1 shows an exploded view and an assembly of the microturbine design. The compressed air enters via a standard pneumatic connector (1) and expands over the stationary nozzles (3) where it is deflected in a direction tangential to the turbine rotor (5). After the air has passed the rotor blades, it leaves the device through the openings in the outlet disc (6). Screwing the pneumatic connector in the housing (8) presses the stationary nozzle disc against a shoulder in the housing. The rotor blades, wheel and axis are one monolithic part. The rotor is supported by two ball bearings (4), one mounted in the stationary nozzle disc and one mounted in the outlet disc. The outlet disc is locked in the housing by a circlip (7).
Figure 2: Microturbine design.
The diameter of the turbine rotor is 10 mm. The housing has a diameter of 15 mm and is 25 mm long. All parts, except pneumatic connector and circlip, are made of stainless steel. The nozzles are designed for subsonic flow, so have a converging cross-section. Sonic speed is reached for a relative supply pressure of 1 bar. The exit losses are minimal when turbine is designed for a u/c1 ratio of 0.5, with u the circumferential speed and c1 the absolute speed at the nozzle exit. At 1 bar, c1 reaches sonic speed resulting in optimal turbine speed of 420,000 rpm. As this is too high for the bearings, the turbine has been designed for a u/c1 ratio of 0.25, and is operated below its optimal speed of 210,000 rpm.
The different parts of the turbine are produced by turning and EDM .The nozzle disc and rotor are the most complex parts. In a first step, their cylindrical surfaces are machined on a lathe. In a second step, the nozzles and blades are created by die-sinking EDM as illustrated for the rotor in figure 3. The rotor is clamped in a rotary head, which is indexed with steps of 30Ã‚Âº. A prismatic copper electrode with a cross-section having the shape of the air channels between the blades is sunk into the turbine wheel by EDM. The electrode is produced by wire-EDM. One of the problems during the production of the turbine blades is electrode wear. This wear is difficult to predict and not uniform across the electrode. This problem has been solved by cutting away the lower edge of the electrode by wire-EDM at regular intervals. As the electrode is prismatic, the shape after shortening remains the same. Figure 4 shows a subassembly of nozzle disc, rotor, and bearings.
Fig 3: Machining of the rotor blades by EDM. Fig4:Subassembly of nozzle disc, turbine rotor, and bearings.
Torque and power of the turbine have been tested up to a speed of 100,000 rpm. For this purpose, a 30 mm diameter brass wheel has been fixed to the turbine axis. An optical sensor measures the rotation of the wheel in a contact less way: two vanes on the wheel interrupt the optical path of a photo sensor. The turbine is tested by switching on the pressure and accelerating the turbine until it reaches its maximum speed. The torque is then derived from the acceleration and the moment of inertia of the wheel and turbine rotor. As the turbine passes through the whole speed range, acceleration, torque and power are know as a function of speed.
When the turbine is rotating at full speed, the pressure is switched off and a new measurement is done while the turbine slows down. This gives the friction torque as a function of speed. Friction mainly occurs between the wheel with vanes and the surrounding air. The friction torque and power are added to the results of the acceleration test to obtain the total torque and power of the turbine.
Fig 5 and 6 show torque and mechanical power as a function of speed for different supply pressures up to 1 bar. The maximum torque and power are respectively 3.7 Nmm and 28 W. The dashed lines represent the friction losses determined with the deceleration test.
Figure 5: Torque as a function of speed and supply pressure.
Figure 6: Mechanical power of the turbine.
At 1 bar, the turbine consumes 8 Nm3/h of compressed air, which corresponds to a power consumption of 152 W when assuming an ideal isentropic expansion. This means that the mechanical efficiency of the turbine lies around 18 %. Figure 7 shows the turbine efficiency as a function of speed for different supply pressures.
The Ëœdipsâ„¢ in the characteristics at high speed are caused by the measurement method as they always occur at the maximal speed, even for different loads and pressures. In reality, power and efficiency increase further with speed to reach their maxima theoretically at 210,000 rpm (for 1 bar). These speeds can be reached using a smaller load.
Figure 7: Efficiency of the turbine (compressed air to mechanical power).
To measure the electrical power output of the system, the generator is connected to a variable 3-phase load consisting of 3 potentiometers (range 2 kW, 10 turns). In contrast with the mechanical tests, the electrical tests are performed at constant speed. The speed of the turbine, which is measured from the frequency of the generator voltage, is controlled by varying the load. Figure 8 shows the electrical power measured for different supply pressures and speeds. At a pressure of 1 bar, the maximal electrical power is 16 W and is reached at a speed of 100,000 rpm. Measurements show that the airflow and input power depend only on the supply pressure and not on speed or load. Therefore, the input power is the same as in the mechanical test at 1 bar, i.e. 152 W. Figure 9 shows the total efficiency (compressed air to electricity) as a function of speed and for different supply pressures. The maximal total efficiency is 10.5 % and is reached at a speed of 100,000 rpm.
Figure 8: Electrical power generated by the total system (turbine plus generator).
Figure 9: Total efficiency (compressed air to electricity).
The energy flow and the different losses are illustrated in the Sankey diagram shown in figure 10. The diagram is generated for a supply pressure of 1 bar and a speed of 100,000 rpm. This corresponds to the working point at which the maximal electrical power and maximal total efficiency are reached. Input power, mechanical power, electrical power and the combination of ventilation losses (6) and bearing friction (7) are measured values. This last value (6 + 7) is obtained with a deceleration test of the turbine without generator and without external load. The loss associated with the leak flow around the turbine wheel (2) and the exit losses (8) are calculated from the known air speeds. The expansion losses (1), incidence losses (4) and blade profile losses (5) are calculated using friction and loss coefficients known from large turbines and may be less accurate. The generator losses (10) are derived from the manufacturerâ„¢s data sheets. The obstruction losses (3) and the losses in the coupling (9) are derived as the difference between the calculated and measured values.
The major losses are the blade profile losses and the exit losses. The large blade profile losses can be explained by the increased friction in miniature systems (large surface-to-volume ratio and low Reynolds numbers). The high exit losses can be explained by the low u/c1 ratio (0.25 instead of 0.5 in the optimal case). Additionally, the turbine operates below its optimal speed because the ball bearings limit the speed. Both factors result in higher air speeds at the turbine exit, and thus higher exit losses.
Figure 10: Sankey diagram for a supply pressure of 1 bar and a speed of 100,000 rpm.
CHALLENGES IN DESIGN AND FABRICATION
The Microturbine presents challenges in the mechanical and electrical engineering disciplines of fluid dynamics, structural mechanics, bearing and rotor dynamics, combustion and electrical machinery design. Then comes difficulties involved in fabrication, heat transfer, structural design and electrical performance .The challenges also includes the need of small bearings and in manufacturing components .the turbine blades may come across with hydrogen burning, so it should be ceramic blade with micro ion etching .The challenges also there for cooling the systems. There will be a chance of high centrifugal stress, which will effect the life of micro turbine during high speed of rotation. Here there is the chance of heat losses due to the high surface to volume ratio, which will effects the efficiency and performance of microturbine.
APPLICATIONS OF MICROTURBINE
Microturbines are suited to meet the energy needs of small users such as schools, apartments, restaurants, offices and small businesses. The Microturbine coupled with solid oxide fuel cell can be used in supermarkets, factories, and military developing nations. This can be applied for heating, drying, cooling, desalination and several others. They include space vehicles, electronic devices, unmanned aircrafts. The microturbines can be used in remote areas. This is because of small size. Microturbines are used when a high quality, energy density energy is needed.
Microturbines and miniature thermal devices pose unique challenges and opportunities for combustion in small volume. The principal difficulties are associated with limited residual time and heat transfer losses due to high surface to volume ratio. This paper addresses a preliminary analysis of Microturbine .The microturbine is in early stages of pre-production and is still in the developmental phase .The coupling of microturbine with a high temperature fuel cell (SOFC â€œ solid oxide fuel cell) is one of them .If the waste heat is used the overall fuel utilization efficiency can be increased. Major features, parameters and performance of the microturbine are discussed here. Fully understanding these and identifying the solutions, it is key to the future establishing of an optimum overall system. In the case of the microturbine changes will be minor as they enter production on a large scale within the next year or so, there is an extensive efforts are expanded to reduce unit cost .It is reasonable to project that a high performance and cost effective hybrid plant, with high reliability, will be ready for commercial service in the middle of the first decade of the twenty century
The first goal is to increase the efficiency of the turbine, mainly by decreasing the exit losses. This can be reached in two ways: introducing air bearings, which allow much higher speeds, or decreasing the speed by using a multiple-stage design. In the long term, a compressor and a combustion chamber will be added to finally come to a micro-generator
1. A.F. Massardo and C.F.McDonald, Microturbine for high-efficiency electrical generation, Transactions of ASME for gas turbine and power, January 2002.
2. E.Utrainen and B.Sunden,Evaluation of the heat transfer surfaces for Microturbine Recuperator, Transactions of ASME for Gas turbine and power, July 2002.
3. Lue Frrchette, Stuart A.Jacobson, Kenneth S.Breuer, Fedric F.Ehric, Reza ghodssi, Ravi khanna, Martin A.Schmidt and Alan H.Epstein,Demonstration of microfabricated high speed turbine supported on gas bearings, solid- state Sensor and actuator workshop, Hilton Head, June 4-8,200