**4.3 Bearings and rotor dynamics**

A UMGT is characterized by highest angular velocities (1 or 2 orders of magnitude higher than conventional ones) and by clearance between the components of the order of the

Here follow is reported a particular case of a built and tested ultra micro combustion chamber.

The 2.5 kW UDR1 UMTG is composed of a compressor, a compact air pre-heater (the pressurized air exiting from the compressor is heated by ΔT= 393 K), a combustion chamber and a turbine, that together compose the thermal section; a double effect (reversible) electrical engine for the electric part; a shaft and the bearings for the mechanical, and obviously all the auxiliaries components required, as valves, controllers, and ducts, including the management unit. For safety reasons the cylindrical fuel tanks are external to

Fig. 6. CAD rendering of the combustor with a detailed view of the pre-heater.

**Spark** 

The overall length of the assembled and modified combustion chamber is 2.2 cm, the outer

**A: fuel** 

**C: hot gases** 

A UMGT is characterized by highest angular velocities (1 or 2 orders of magnitude higher than conventional ones) and by clearance between the components of the order of the

**4.2 Other combustion chamber** 

**4.2.1 UDR1 UMGT combustion chamber** 

the metallic UMTG enclosure [Capata & Sciubba, 2010].

diameter 4.2 cm and the overall height is 13.0 cm.

Fig. 7. The tested prototype of the combustion chamber

**4.3 Bearings and rotor dynamics** 

**B: air inlet** 

micron. It is therefore from is easy to verify as the great part of the design development is dedicated to the bearings and to the rotor dynamics. The first reflection is, due to the high rotational speed (> 20000 rad/s), the device will work in supercritical conditions, with operational frequencies higher than the resonance ones and for this reason the study of the stability will be primary in the machine design. Clearly the bearings play a dominant role in the rotor dynamics as they must support axial and radial thrust of the rotor, dampen the oscillations and support the forces deriving by the accelerations of the rotary parts, without to count the electric forces and the operational pressures. It cannot certainly be neglected that the supports must be able to supporting the high turbine temperatures and the thermal gradient presents and, then, be able to work in any operational condition. At last the resolution of all these problems must be consider the actual technological limitations. Other devices MEMS with rotary parts currently exist, but the angular velocities are such to allow to using not-lubricated bearings. Moreover do not exist in commerce such "little bearings" to be used for these rotors. Honestly, we have to mentioned two manufacturers that guarantee dry bearings at 200000 rpm (20933 rad/s). Therefore, the academic study focuses on two possible solutions: electromagnetic or air bearings. The first solution introduces the possibility to take advantage of magnetic fields or electrical ones to support the rotors loads. The problem relatively to the magnetic fields resides in the impossibility to adopt ferromagnetic materials in the productive technologies chosen (recording laser and photolithography), moreover the Curie point is such configuration not to allow of using these systems to the temperatures previewed in turbine and would be therefore necessary to introduce a cooling system, with all the design constraints and possibilities connected. Studies have been carried on the electric fields, but the forces that are succeeded to produce are decidedly inferior to support the considered loads. To all these consideration we must add that the electromagnetic bearings are strongly unstable and required a feedback control systems, which would complicate the installation. The air bearings introduce numerous advantages, in particular in terms of constructive simplicity, of loads capability, and relative insensibility to the problem of the high temperatures. Currently these device are already used for medium-small dimensions turbomachines. At smaller scales the gas bearings are wide used in micro gyroscopic system since many years. Wanting to make scale considerations it can be said that, to parity of load conditions, the ability of these bearings grows up with the overall dimensions decrease, because the ratio volume/surface decreases and consequently the inertial load. Certainly the rotor dynamics results to be, partially, simpler (the introduced structure is more rigid, if compared to the conventional ones), thus allowing the approximation of the rigid body. The development of the bearings has to preview two different use: first, the bearings must support the radial loads, second, must support the axial loads. These bearings are composed from a cylindrical hinge in contact with a lubricating support. More efficient models exist, but considering the complexity of the system, the more reasonable choice is the first one. The lubricating fluid in our case is air in pressure and according to how it is injected, the bearings are distinguished in "hydrostatic" if the source is external and "hydrodynamic" if the support forces are directly generated by the spin of the disc (figure 8). Mixed configurations are possible (figure 9). Since the UMGT includes a compressor is possible to adopt both the configurations, and both solution are able to satisfying the loads requirements and temperatures stresses to these dimensions. The two types of bearings are characterized by various dynamic characteristics. In the hydrodynamic bearings the load ability grows up to increasing of the rotational speed, because the pressure of the film between disc and support increases with the spin. In theory, also for a hydrostatic bearing a mechanism of this type can be fed; we

Ultra Micro Gas Turbines 19

"Jeffcott wheel": to low frequencies the rotor rotate around the geometric centre, while to at high frequencies rotates around the centre of mass. The dotted line represents just the loss of balance between geometric centre and centre of mass to which asymptotically stretches the rotor. For a hydrostatic bearing the critical frequency scale, simply, with the pressure and the viscous damping decreases to increasing of the speed. The amplitude of displacement of

Fig. 10. Comparison between the load abilities of several bearings [Van den Braembussche 1993]

The trend will be similar to that reported in Figure 12, but to increasing of the pressure the peak will be moved up towards right and, this suggests to cross the critical frequency at low pressure values and low speed to, then, increase the pressure and to harden the bearing. A factor that influences the relative displacement amplitude at the critical frequency and the instability phenomena is the balance loss of the rotor. Normally the conventional machines adopt a dynamic balance that substantially consists in measuring the loss of balance and to modify the mass locally, to compensate the displacement between geometric centre and centre of mass. In the micro machines such problem is less sensible (the loss of balance for a rotor of 4 millimetre is about 1-5 µm) because the manufacturing process uses Si mono crystals or wafer of material much homogenous. To this can be add the manufacturing precision of ~1 µm. That it renders geometry much uniform. On the other hand must pay great attention to the alignment of the wafer. The hydrostatic bearings have the advantage of being stable to all the speeds within a fixed range, on the contrary the hydrodynamics ones suffer to low speed, while they are perfectly stable to the high ones. A way to stabilize these last bearings is to apply a unidirectional force that presses and pushes the rotor

Fig. 11. Behaviour of an air micro bearing [Epstein 1997]

the rotor to the critical frequencies, then, increases with increasing of the pressure.

can imagine to support the gas in pressure through the compressor which will consequently regulate the pressure of the lubricating fluid. If the pressure of the fluid were maintained to constant the load ability would decrease gradually to increasing of the speed. In figure 10 is visible a comparison between the two types of bearings and a conventional one. In an hydrostatic bearings, pressure is maintained constant, while for an hydro dynamic one the load capability cargo depends on the speed and L/D ratio, where L is the length of the bearing while D is the diameter of the rotor. To decrease of this ratio the performances gets worse.

Fig. 8. Operational mode of a hydrostatic bearing (over) and hydrodynamic (under)

Fig. 9. Hybrid configuration

The fluid simply "sprayed" outside by the considered load, rather than to develop an adequate profile of pressures to balancing the external force. Unfortunately, in the laser manufacturing exist technological constraints, that does not allow to obtain elevated L/D ratio. Other parameters that influence the bearings design are the ratio between gap and the length of the bearing g/L, the mass of the rotor, the Reynolds number; while for the hydrodynamics bearings the load capability is dominated by the inverse of (g/D)5 ratio. Currently the DRIE techniques allow to realize aspect ratio of 30:1, but a value of 100:1 can be reached. This means that assuming gap of 10-20 µm (that is the minimum value can be adopted to avoid the rotor failure for contact) can be had lengths of the channel of lubrication of 300-1000 µm. The design of an hydrodynamics bearing would suggest to us to decrease the g/D ratio and, therefore, to choose rotors with possible greater diameter, but this contrasts with the requirement of maximizing relationship L/D. Such parameters will probably find outside the normal design range. As already said the rotor operational frequencies are much higher than the resonance ones, and to such frequencies it is possible to model the bearing as a complex of springs and dampers whose effect is given by the lubricating gas which is cause also of operational instability due to transverse forces. The gap it must be such to support a deviation of the rotor during the "crossing" of the critical frequencies without to hit the support. Figure 11 shows the MIT data (rotor diameter of 4 millimetre and gap of 12 µm). It can be noticed as the maximum displacement is lower than the maximum allowed. The data is in agreement with a famous analytical model, known as

can imagine to support the gas in pressure through the compressor which will consequently regulate the pressure of the lubricating fluid. If the pressure of the fluid were maintained to constant the load ability would decrease gradually to increasing of the speed. In figure 10 is visible a comparison between the two types of bearings and a conventional one. In an hydrostatic bearings, pressure is maintained constant, while for an hydro dynamic one the load capability cargo depends on the speed and L/D ratio, where L is the length of the bearing while D is the diameter of the rotor. To decrease of this ratio the performances gets

Fig. 8. Operational mode of a hydrostatic bearing (over) and hydrodynamic (under)

The fluid simply "sprayed" outside by the considered load, rather than to develop an adequate profile of pressures to balancing the external force. Unfortunately, in the laser manufacturing exist technological constraints, that does not allow to obtain elevated L/D ratio. Other parameters that influence the bearings design are the ratio between gap and the length of the bearing g/L, the mass of the rotor, the Reynolds number; while for the hydrodynamics bearings the load capability is dominated by the inverse of (g/D)5 ratio. Currently the DRIE techniques allow to realize aspect ratio of 30:1, but a value of 100:1 can be reached. This means that assuming gap of 10-20 µm (that is the minimum value can be adopted to avoid the rotor failure for contact) can be had lengths of the channel of lubrication of 300-1000 µm. The design of an hydrodynamics bearing would suggest to us to decrease the g/D ratio and, therefore, to choose rotors with possible greater diameter, but this contrasts with the requirement of maximizing relationship L/D. Such parameters will probably find outside the normal design range. As already said the rotor operational frequencies are much higher than the resonance ones, and to such frequencies it is possible to model the bearing as a complex of springs and dampers whose effect is given by the lubricating gas which is cause also of operational instability due to transverse forces. The gap it must be such to support a deviation of the rotor during the "crossing" of the critical frequencies without to hit the support. Figure 11 shows the MIT data (rotor diameter of 4 millimetre and gap of 12 µm). It can be noticed as the maximum displacement is lower than the maximum allowed. The data is in agreement with a famous analytical model, known as

worse.

Fig. 9. Hybrid configuration

"Jeffcott wheel": to low frequencies the rotor rotate around the geometric centre, while to at high frequencies rotates around the centre of mass. The dotted line represents just the loss of balance between geometric centre and centre of mass to which asymptotically stretches the rotor. For a hydrostatic bearing the critical frequency scale, simply, with the pressure and the viscous damping decreases to increasing of the speed. The amplitude of displacement of the rotor to the critical frequencies, then, increases with increasing of the pressure.

Fig. 10. Comparison between the load abilities of several bearings [Van den Braembussche 1993]

Fig. 11. Behaviour of an air micro bearing [Epstein 1997]

The trend will be similar to that reported in Figure 12, but to increasing of the pressure the peak will be moved up towards right and, this suggests to cross the critical frequency at low pressure values and low speed to, then, increase the pressure and to harden the bearing. A factor that influences the relative displacement amplitude at the critical frequency and the instability phenomena is the balance loss of the rotor. Normally the conventional machines adopt a dynamic balance that substantially consists in measuring the loss of balance and to modify the mass locally, to compensate the displacement between geometric centre and centre of mass. In the micro machines such problem is less sensible (the loss of balance for a rotor of 4 millimetre is about 1-5 µm) because the manufacturing process uses Si mono crystals or wafer of material much homogenous. To this can be add the manufacturing precision of ~1 µm. That it renders geometry much uniform. On the other hand must pay great attention to the alignment of the wafer. The hydrostatic bearings have the advantage of being stable to all the speeds within a fixed range, on the contrary the hydrodynamics ones suffer to low speed, while they are perfectly stable to the high ones. A way to stabilize these last bearings is to apply a unidirectional force that presses and pushes the rotor

Ultra Micro Gas Turbines 21

would not be problematic for the correct operational mode of the device. Regarding the thrust bearings necessary to support axial stresses, both the hydrostatic configuration than the hydrodynamic one has been successfully tested and verified. In both cases the working point is sub critical, that is the exercise frequencies are under the critical ones, and this bring to an important design simplification. The hydrostatic bearings have a stationary behaviour and lead pressurized air in contact with the rotor through orifices circularly located inside the support. A value of about 2-5 atm (0.2÷0.5 MPa) are necessary to obtain adequate load capabilities and a good rigidity. The maximum rigidity is obtained when the pressure drop through the orifices equals the radial flow from the drainage edge of the bearing. The hydrodynamic bearings take advantage of the viscous resistance, exalted from a superficial envelope of spirals, to generate of the pressure gradient on the surface that increases towards the centre of the rotor. This mechanism self sustains and does not demand constructive complications, reducing the number of wafer necessary too. It is considered, moreover, that, while at the conventional scales, this type of bearings are less effective and not so used, at the small scales, thanks to the great increase of the superficial forces, regarding the volumetric ones, the load capability and the rigidity are similar to the ones of the high speeds hydrostatic bearings. This system introduces a problem of the minimal speed to generating a sufficient pressure to eliminate the rotor rubbing with the static parts: fortunately such minimal speed is of the order of 104 rpm [Epstein, Jacobson, Spakovszky] and this implies a malfunctioning limited to machine start up. The power dissipation of the two types of bearing has been calculated and verified by MIT institute and appears to be low. The necessity to obtain high accuracy in the bearings manufacturing, is in contrast with the demand of a device series production with an only one wafer. The plasma techniques, as well as all manufacturing techniques, introduce heterogeneous problems, that increase moving from the centre towards the periphery of the wafer. Regarding DRIE process, that is still in phase of maturation, seems that the manufacturing process carry to an inaccuracy in the recording depth of 3 µm on a diameter of 4 millimetre for the peripheral devices (the most penalized), than consist in a rotor loss of balance and, therefore, a total malfunctioning of the bearings. Actually MIT reports to have been tested this kind of bearings at a rotational

The most important efforts in the UMGT design are dedicated, undoubtedly, to the turbomachines, the combustion chamber and the bearings. But we have to consider that the others components that completing the device will not be simpler designed or less important, being the system, in its entirety, a good compromise between the several parts, a

The literature on the electric motors is extremely huge, but on the micro generators is insufficient. Seen the high temperature operational conditions, the necessity to integrate the components to avoid ulterior complications dictated by the bearings and the required power density, would be necessary an accurate study of the problem. Both magnetic and electrical motors are attractive as they are characterized by analogous power density, but the manufacturing difficulties correlated to the "not compatibility" of the ferromagnetic materials with the conventional micro technologies and the operational temperatures that

compromise between the obtainable good characteristics from every single piece.

speed of about 1.4 Mrpm [Epstein 2003]

**4.4.1 Electric generator and starter** 

**4.4 Auxiliaries** 

towards the wall of the support and, conventionally, the weight of the disc is used to generate such forces. The common method to quantify the decentralization necessary to create such forces is given by the eccentricity, ratio between the distance wall-rotor on medium gap (the 0 if the rotor is cantered, 1 if it touches the wall). The problem to the micro scales is that inertia is practically uninfluenced regarding the surface forces, this is favourable if we consider that the device demands orientation independence, but in terms of stability it has pushed to find such lay outs that take advantage of a heterogeneous distribution of the gas pressure to generate eccentricity. The zone behind the rotor is subdivided in two rooms that can be pressurized independently. Such system has been largely simulated and the result show that the rotor is stable for eccentricity values over 0.8- 0.9. A "not yet solved" problem is related to the axial flows that pass in the interstice between rotor and support of the bearing: considering dimensions of the gap these flows do not seem negligible neither in hydrodynamic neither in that hydrostatic one. The stability, as we said, strongly depends on the (g/D)5 ratio through a mass parameter:

$$\overline{\mathbf{M}} = \frac{\mathbf{m}\mathbf{p}}{72\mathbf{L}\mu^2} \left(\frac{\mathbf{g}}{\mathbf{D}}\right)^\flat \tag{23}$$

It can be noticed that is necessary to have high eccentricities to having a gap, locally, more little. In this manner mass parameters decrease and, as consequence, the dotted line in figure 12 will be lower. The conventional turbomachinery succeed to obtain a minimal eccentricities closed to 0.5, but to our scales we are limited in the choice of the shape ratio.

Thus high eccentricities imply high accuracy in the circularity of the rotor parts and support, with the maximum shunting lines in the order of the µm, to avoid interferences. To adopt high eccentricities can be problematic in operating terms, because they produce a strong balancing losses. To obtain better performances could be the adoption of so-called "wave" bearings. This device is capable to dampening the whirl effects due to the excitation of the pressure perturbations generated by the particular geometry of the same bearings. This advantage does not involve some constructive complication and can be easy implemented through the photolithography techniques, however since the technological constraints define a minimal gap, the only way to create such geometry are to increase the interstices. From a performances analysis results that the load capability decreases, but the stability strongly increases and the minimal demanded load to stability operate (that means the minimal value to obtain the minimal stable eccentricity) decreases. Being the load capability of the bearings more sufficient than the required one, its decrease, in favour of the stability,

towards the wall of the support and, conventionally, the weight of the disc is used to generate such forces. The common method to quantify the decentralization necessary to create such forces is given by the eccentricity, ratio between the distance wall-rotor on medium gap (the 0 if the rotor is cantered, 1 if it touches the wall). The problem to the micro scales is that inertia is practically uninfluenced regarding the surface forces, this is favourable if we consider that the device demands orientation independence, but in terms of stability it has pushed to find such lay outs that take advantage of a heterogeneous distribution of the gas pressure to generate eccentricity. The zone behind the rotor is subdivided in two rooms that can be pressurized independently. Such system has been largely simulated and the result show that the rotor is stable for eccentricity values over 0.8- 0.9. A "not yet solved" problem is related to the axial flows that pass in the interstice between rotor and support of the bearing: considering dimensions of the gap these flows do not seem negligible neither in hydrodynamic neither in that hydrostatic one. The stability,

5

(23)

2 mp <sup>g</sup> <sup>M</sup> 72Lμ D <sup>=</sup>

It can be noticed that is necessary to have high eccentricities to having a gap, locally, more little. In this manner mass parameters decrease and, as consequence, the dotted line in figure 12 will be lower. The conventional turbomachinery succeed to obtain a minimal eccentricities closed to 0.5, but to our scales we are limited in the choice of the shape ratio.

Fig. 12. Gas bearing stability in function of the eccentricity and normalized rotational speed

Thus high eccentricities imply high accuracy in the circularity of the rotor parts and support, with the maximum shunting lines in the order of the µm, to avoid interferences. To adopt high eccentricities can be problematic in operating terms, because they produce a strong balancing losses. To obtain better performances could be the adoption of so-called "wave" bearings. This device is capable to dampening the whirl effects due to the excitation of the pressure perturbations generated by the particular geometry of the same bearings. This advantage does not involve some constructive complication and can be easy implemented through the photolithography techniques, however since the technological constraints define a minimal gap, the only way to create such geometry are to increase the interstices. From a performances analysis results that the load capability decreases, but the stability strongly increases and the minimal demanded load to stability operate (that means the minimal value to obtain the minimal stable eccentricity) decreases. Being the load capability of the bearings more sufficient than the required one, its decrease, in favour of the stability,

[Freccette 2003]

as we said, strongly depends on the (g/D)5 ratio through a mass parameter:

would not be problematic for the correct operational mode of the device. Regarding the thrust bearings necessary to support axial stresses, both the hydrostatic configuration than the hydrodynamic one has been successfully tested and verified. In both cases the working point is sub critical, that is the exercise frequencies are under the critical ones, and this bring to an important design simplification. The hydrostatic bearings have a stationary behaviour and lead pressurized air in contact with the rotor through orifices circularly located inside the support. A value of about 2-5 atm (0.2÷0.5 MPa) are necessary to obtain adequate load capabilities and a good rigidity. The maximum rigidity is obtained when the pressure drop through the orifices equals the radial flow from the drainage edge of the bearing. The hydrodynamic bearings take advantage of the viscous resistance, exalted from a superficial envelope of spirals, to generate of the pressure gradient on the surface that increases towards the centre of the rotor. This mechanism self sustains and does not demand constructive complications, reducing the number of wafer necessary too. It is considered, moreover, that, while at the conventional scales, this type of bearings are less effective and not so used, at the small scales, thanks to the great increase of the superficial forces, regarding the volumetric ones, the load capability and the rigidity are similar to the ones of the high speeds hydrostatic bearings. This system introduces a problem of the minimal speed to generating a sufficient pressure to eliminate the rotor rubbing with the static parts: fortunately such minimal speed is of the order of 104 rpm [Epstein, Jacobson, Spakovszky] and this implies a malfunctioning limited to machine start up. The power dissipation of the two types of bearing has been calculated and verified by MIT institute and appears to be low. The necessity to obtain high accuracy in the bearings manufacturing, is in contrast with the demand of a device series production with an only one wafer. The plasma techniques, as well as all manufacturing techniques, introduce heterogeneous problems, that increase moving from the centre towards the periphery of the wafer. Regarding DRIE process, that is still in phase of maturation, seems that the manufacturing process carry to an inaccuracy in the recording depth of 3 µm on a diameter of 4 millimetre for the peripheral devices (the most penalized), than consist in a rotor loss of balance and, therefore, a total malfunctioning of the bearings. Actually MIT reports to have been tested this kind of bearings at a rotational speed of about 1.4 Mrpm [Epstein 2003]

### **4.4 Auxiliaries**

The most important efforts in the UMGT design are dedicated, undoubtedly, to the turbomachines, the combustion chamber and the bearings. But we have to consider that the others components that completing the device will not be simpler designed or less important, being the system, in its entirety, a good compromise between the several parts, a compromise between the obtainable good characteristics from every single piece.

### **4.4.1 Electric generator and starter**

The literature on the electric motors is extremely huge, but on the micro generators is insufficient. Seen the high temperature operational conditions, the necessity to integrate the components to avoid ulterior complications dictated by the bearings and the required power density, would be necessary an accurate study of the problem. Both magnetic and electrical motors are attractive as they are characterized by analogous power density, but the manufacturing difficulties correlated to the "not compatibility" of the ferromagnetic materials with the conventional micro technologies and the operational temperatures that

Ultra Micro Gas Turbines 23

The problems connected to the choice of the materials for a UMGT are referred, partially, to those of conventional machine, in terms of mechanical stresses constraints, operational temperature and manufacturing processes. The materials, generally, satisfy some characteristics penalizing some others, therefore, it is indispensable in the preliminary design to find a good compromise, and not always, the better choice is univocal. It is considered, moreover, that according to manufacturing, the same material can introduce different characteristics. The properties of greater interest for the studied machine are the material absolute specific strength and its resistance to the thermal shocks at high temperatures, the creep and the oxidation, and its derangement characteristics under fatigue cycle. Considering the manufacturing processes and the device dimensions, the first choice for a device MEMS falls on the silicon, thanks the great maturity of the material productive technologies. The silicon also having of the good properties of specific strength and to thermal shock, presents a ductility and a strong "flauge" over the 550°C. This means that I can not use this material in the turbine blade manufacturing. On the contrary, the Silicon alloys appear more interesting, but the manufacturing of these materials is not still mature as well as the silicon ones. The metallic alloys are not able to operate at the demanded high temperatures without cooling or covering, that could create constructive complications. Moreover, the metallic manufacturing does not have a good accuracy. Other advanced ceramic materials, characterized by higher operational temperatures, have insufficient mechanical characteristics to the high temperatures. Adding the fragile behaviour of the ceramic materials, the manufacturing process results problematic and, at the moment, MEMS techniques are accurate only for Silicon and its alloy. Figure 13 put in evidence that the silicon alloys, in this case alumina, can be used at the high temperature, but the Al2O3 alloy has lower thermal conductivity and an higher thermal expansion. These characteristics renders the alloy particularly subject to thermal shocks and deformations. Also regarding hardness and elasticity modulus the SiC and Si3N4 have higher resistance performances, in particular to bending, and it can be noticed that, also having lower values to low temperatures respect to, for example, the zirconium, such value is maintained approximately constant to the high temperatures and is higher than the other materials. In figure 14 the SiC mechanical properties are maintained, substantially, unchanged with the temperature. The weak of the silicon alloy, like all ceramic materials, is the fragile behaviour, that to large-scale has delayed theirs uses. In large scale components, in fact, the inner imperfections can be in such amounts and largeness that is sufficient the effect of a small extra-solicitations to "prime" the propagation of crack, especially in these materials where the reticulum plans sliding is extremely reduced. For the UMGT order of magnitude, a single piece is composed by a low grains number, so the problem of the structure inner defects is more controllable. The piece surface-volume ratio grows, and will be, therefore, necessary to put greater attention to the superficial defects, prime points for the propagation of crack. From this first comparison, SiC and Si3N4, seem to be the more appropriate materials, thanks to the better behaviour at the high temperatures. There is Recently, adding some elements which boron, aluminium, yttrium and/or relative oxides, meaningful improvements of the mechanical properties have been verified. As previously said, the characteristics of these

**5. Materials** 

materials depend on the manufacturing process.

SiC presents an optimal material specific strength, a good resistance to the thermal shocks, thanks to its relative high thermal conductivity. In Table 3 the characteristics of the SiC

**5.1 Silicon carbide (SIC)** 

reduce the properties of the same materials, have pushed to consider, in first analysis, the adoption of electrical devices. The power density scale with the square of the force of the electric field, with the frequency and the rotational speed. Numerous possible configurations for an electrical motor-generator exist; the first choice made by the MIT researchers is an induction machine, because this type of machine does not require a direct contact between the electrical device and the rotor neither the exact knowledge of the position of the rotor [Epstein 2003]. The rotor is composed by a layer of 5-20 µm of good insulator covered by a thin layer of a low conductor (high superficial resistance), while the stator is composed by a series of radial electrodes conductors supported by an insulating layer. The rotating electrical potential is imposed by the external electronics on the stator electrodes and the rotating electric field generates a distribution of charges on the rotor, which is mechanically driven. According to the relative phase between the motion of the charges on the rotor and the statoric field, adjustable from the outside, the device will operate as generator, motor or brake. The torque increases with the square of the force of the electric field and the frequency, but the maximum allowable force is regulated by the dimension of the gap. For the air the maximum is obtained for interstitial values of little micron, so that such machine can theoretically achieve an higher power density than the conventional device with analogous configuration. The frequency is regulated by the external electronic systems and by manufacturing device of the electrodes. Currently 300 volts and a frequency of 1-2 MHz is the maximum value obtained with a 6 mm rotor to producing 10 W with a gap of 3µm and a number of poles much high (beyond 100). To maximize the output power is necessary that the space between rotor and stator is of the same order of the statoric electrodes pitch, that means few µm, but with this type of gap the losses for viscous resistance, considering the machine rotational speed, are extremely high and represent the main source of losses. It is necessary to find a compromise between power density and efficiency. The viscous losses unfortunately represent more than the half of the total ones and limit the efficiency to 40-50%. A magnetic induction machine has the advantage of having less poles and optimal distances rotor-stator very wider (≈10 µm), that produce higher efficiency, around 60%. An ulterior advantage of these devices is that they operate at low frequencies and voltages, and are simpler to build. Currently the greater problems is the thermal resistance of the materials (actually 500 K, but material to working to 800 K have been currently studied) and the rotational speed in conflict with the necessity to insert on the rotor surface a iron layer of about 100 µm. Recently, thanks to new technologies of implantation of the ferromagnetic materials on silicon wafer, a good magnetic induction machine to with interesting efficiency and performances has been realized. As already said, to succeed to adopt a magnetic induction machine respect to analogous electrical devices, implies a greater efficiency, tolerances less limiting and higher manufacturing simplicity. An example of a magnetic induction machine is shown in the figure. The rotor diameter is 10 mm. The machine consists of a two phases stator and 8 poles and of a annular rotor [Epstein 2003]. The electromechanical conversion of the energy is achieved by the interaction of the magnetic field that evolve in the interstice rotor-stator with the current induced in the rotor the displacement of the magnetic wave. The stator consists of two phases composed by planar copper spirals, insert in three-dimensional blocks of vertically laminated ferromagnetic material, all supported by a silicon chassis. The ferromagnetic nucleus has an "onion" configuration, the sheets forms concentric rings, approximately 30 µm thick. This particular shape serves to reduce the induced current losses. The rotor is composed by 2 ferromagnetic ring with a thickness of 250 µm and 2 mm wide covered by a copper layer of 20 µm. The copper is extended over the external beam of the rotor for an ulterior millimetre, to exalt the induced current generation and to increase the maximum torque.
