A Variable-Reluctance Linear or Rotary Synchronous Electric Motor with Volumetric Development of Force

Technical Field The invention relates to a motor according to the introductory part of claim 1.

Background Art

There are many methods of making an electric motor, in other words of converting electrical energy to mechanical energy.

Except in electrostatic motors, which in any case form a very small part of the total, the force in an electric motor is developed by the interaction of magnetic fields with each other or with wires through which current passes.

The most common rotary electric motors may be grouped into a few major categories:

1) direct-current motors, with collectors and brushes, of the permanent magnet or wound type;

2) asynchronous induction motors;

3) synchronous motors, with permanent magnets or windings (modern "brushless" motors belong to this category) ; 4) variable-reluctance motors;

5) stepping motors, with permanent magnets or variable reluctance.

All these types of motor have a similar structure, namely: - the rotor and stator are separated by an air gap;

- a magnetic flux passes through the air gap; suitable distortions of the magnetic flux generate a lateral thrust in the rotor to provide the torque;

there is a system to continuously match the position of the flux to the position of the rotor in order to obtain a constant torque even while the rotor is moving. Evidently, each of these types of motor may be "opened out" and extended in a plane to form a linear motor.

The progress of technology and the requirement for increasingly high speeds of movement of parts of machines makes linear motors very attractive, but the high costs and low forces developed prevent their application on a wide scale.

To permit the understanding of the invention, which relates to a motor with high torque in the rotary form and high force in the linear form, the physical phenomena which cause the generation of the forces in electric motors will now be summarized.

For convenience, reference will be made to motors of linear form, without prejudice to the general nature of the description, since the same considerations apply to the rotary form.

Two forces, one parallel•to the flux (used in electromagnets) and one lateral (working force of the motor) act on the surfaces of ferromagnetic material separated by the air gap.

The force parallel to the flux (attractive force) is:

(1) Fp = s * B * B / (2 * μO) where Fp is the attractive force in newtons, s is the y surface in m , B is the magnetic induction in teslas, μO is the permeability of the gap.

If B = 1.5 tesla (easily obtainable with modern materials) , there is an attractive force of approximately 90 newtons/cm : an enormous force. The lateral force (working force) is apparently not easy to determine in a general way, since it seems to depend on the physical structure of the motor. In

fact, however, the phenomenon is identical for all types of motor, and the final results are identical: the lateral force is that resulting from the attraction of two partly facing magnetic poles separated by an air gap.

Fig. 1 shows schematically two magnetic poles 1 and 2 of height h, separated by an air gap with a thicknesf t, overlapping for a length x and linked by the lines of flux 3. If the distortions of the magnetic field due to the edge effect are disregarded (in other words, if the dimensions of the air gap are negligible with respect to the poles), the lateral force will be: (2) Fl = t * h * B * B / (2 * μO) where the units of measurement are the same as those in.formula (1) .

In a physical motor structure the situation becomes that shown in Fig. 2, where there is a sequence of facing poles 4 and 5 of height h, repeated at a constant pitch p, separated by an air gap t. The lateral force for each pole (where t is small) is again expressed by formula (2). The lateral force may be expressed as a function of the surface. For n poles, the force is: (3) Fn = n * t * h * B * B / (2 * μO ) and, since the total surface s is given by n*p*h,

(4) Fl = t/p * s * B * B / (2 * μO) Given (1) , then

(5) Fl = t/p * Fp = k * Fp where k = t/p.

This formula leads to interesting considerations.

The lateral force (used for constructing a motor) increases with the ratio between the air gap t and the pole pitch p: to construct a rotary motor with high torque or a linear motor with high force, the ratio k = t/p must be maximized, and therefore the air gap must be increased or the pole pitch must be decreased.

If each pole has its own electrical winding, the difficulties in increasing the ratio t/p are evident: the pole pitch must somehow also include the space required for the winding, a space which increases with the air gap, which requires a proportional magnetizing force and consequently a proportional excitation current.

The case in which more than one pole is controlled by a single winding is different: the situation may be as shown in Fig. 3, where the two electrical windings 8 magnetize the two general poles 6 which, owing to their geometry, concentrate the flux in a multiplicity of elementary poles 7.

In this case, the ratio t/p may be optimal to generate the maximum torque.

Electric motors may be grouped into two major categories: motors with one elementary pole for each winding; - motors with more than one elementary pole for each winding.

The second category includes stepping motors, while the first category includes all other motors: it is now clearer why stepping motors provide higher torques on average for the same dimensions.

If the matter of whether a single pole or a group of elementary poles is provided for each winding is disregarded, the working force is always directly proportional to the dimension of the air gap traversed by the lines of force of the magnetic field.

The moving part of the motor is also subject to a force (Fp) parallel to the lines of flux of the magnetic field in the air gap and not participating in the generation of the torque (in a rotary motor) or of the total working force in a linear motor. Whereas in the rotary motor the forces Fp generated at each pole provide a theoretically zero resultant and one which

can in any case be easily discharged onto the rotation bearing of the rotor, in the linear motor the force Fp forms an attractive stress between the inductor and the armature which has to be discharged internally onto the supports which hold the armature. This represents a considerable technical problem which is difficult to solve and which drastically limits the working force which can be generated by a linear motor.

Objects of the Invention

The object of the present invention is to propose a new structure for an electric motor, whether rotary or linear, by means of which it is possible to obtain greater torque (in the case of rotary motors) or greater force (in the case of linear motors), without the problems stated above.

A further object of the present invention consists in the provision of a linear electric motor with which it is possible to obtain high working forces while simultaneously eliminating the attractive stresses between the inductor and armature.

A further object of the present invention consists in the provision of an electric motor with which it is possible to obtain higher torque or forces than those provided by conventional motors for the same excitation current, and consequently without increasing the dimensions of the motor.

Disclosure of the invention

These and other objects and advantages, which will be made clear by the following text to those skilled in the art, are achieved with the characteristics specified in claim 1.

The dependent claims relate to further developments and improvements of the invention.

Brief Description of the Drawings

The invention will be more clearly understood from the description and the attached drawings, which show a non-restrictive embodiment of the invention. In the drawing, Figs. 1 to 3 are simplified diagrams relating to the state of the art as described previously;

Fig. 4 is a plan view of a theoretical system on which the structure of the motor according to the present invention is based; Fig. 5 is a transverse section through V-V in Fig. 4;

Fig. 6 is an enlargement of Fig. 4; Fig. 7 is the graph of the force acting on the armature shown in Figs. 4 and 5 as a function of the displacement between the armature and inductor, when the inductor winding is supplied with direct current;

Fig. 7a is a series of diagrams showing the variation of the forces on the armature in a multiple- phase power supply; Fig. 8 is a plan view of an embodiment of a linear motor with three-phase supply;

Fig. 9 is a section through IX-IX in Fig. 8; Fig. 10 is a plan view of a further development of the structure shown in Fig. 8; Fig. 11 is a section through XI-XI in Fig. 10; Fig. 12 shows a further development of the embodiment shown in Fig. 11;

Fig. 13 shows a particular embodiment of one pole or phase of the motor; Fig. 14 shows a possible method of inserting a position transducer into the motor shown in Fig. 4;

Figs. 15a, 15b and 15c show three geometrical views of a possible embodiment of an armature. Figs. 15b and 15c being views through XVb-XVb and XVc-XVc in Fig. 15a; Figs. 16a and 16b show respectively a plan view and a section through XVI-XVI of a motor with a structure of the type shown in Fig. 10, with an armature made according to Fig. 15;

Fig. 17 shows an axial section of a rotary motor according to the invention;

Figs. 17a, 17b, and 17c show the disposition of the windings of the motor shown in Fig. 17; and Figs. 18 and 19 show a further embodiment.

Detailed Description of the Invention

In the following description, the part of the motor containing the excitation windings is called "fixed", and the remaining part is called "moving", but there is no reason why, in the practical embodiment, the moving and fixed parts should not be interchanged, only the relative motion of the two components being important for the purposes of operation. In its simplest form, one pole of the motor, complete with inductor and armature, is shown in plan in Fig. 4 and in section in Fig. 5.

The two polar expansions 9, which form the inductor, are excited by the winding 10 and are shaped so that they concentrate the flux generated in a multiplicity of elementary poles 9A provided on both expansions 9 and corresponding to each other.

A housing within which an armature 12 is disposed slidably as shown by the arrow X is formed ^' between the two polar expansions 9 of the inductor. A first air gap tl is formed between the armature 12 and the first

polar expansion 9, while a second air gap t2 is formed between the armature and the second polar expansion 9.

In the present description of the embodiment shown in Figs. 4 and 5 and in the subsequent Figs. 6 to 16, reference will be made to a motor of the linear type. All the considerations which will be discussed may, however, be applied to a rotary motor which is produced simply by "curving" the linear motor structure into a ring. The lines of flux 11 of the generated magnetic field pass twice through the air gap (through tl and t2) for each elementary pole 9A, and therefore contribute twice to the generation of the working force. The armature 12, movable in the direction X, consists of a multiplicity of elementary poles consisting of keepers 13, having the same pitch of the elementary poles 9A of the inductor 9, and held in position by a suitable insulating and non-magnetic support 14.

The armature 12 is subject to a force which tends to bring it to the minimum energy configuration, which is that in which the elementary poles 13 of the armature 12 are aligned with the elementary poles 9A of the inductor 9.

A major advantage of this structure, in the case of a linear motor, is that only lateral working forces, and not attractive forces, are generated between the moving part 9 and the fixed part 12, since the attractive forces are cancelled out on the moving part 12 because they are equal and opposing in the two air gaps tl and t2 traversed by the lines of flux 11. The attractive force is exerted only between the opposing poles of the inductor 9 and is easily withstood by the rigidity of the inductor.

If the system shown in Fig. 4 is supplied with a direct current, the resultant working force in the

direction X is an alternating cyclical function of the relative positions of the armature 12 and inductor 9, as shown in Fig. 7.

To facilitate the construction of a multiple- phase motor which provides a constant force as a function of the relative displacement of the inductor and armature, the function shown in Fig. 7 should preferably be as nearly rinusoidal as possible. For this purpose, it is advantageous to design the poles as shown in Fig. 6: the elementary pole 9A of the inductor 9 is half of the pole pitch p, while the elementary pole 13 of the armature (or the keeper 13) is one third of the pole pitch p. With this design, the second harmonic and all the even harmonics in the force diagram in Fig. 7 are canceled out, together with the third harmonic and all its multiples, resulting in a closer approximation to the sinusoid. The same result would be obtained by making the elementary poles of the inductor 1/3 and the elementary poles of the armature 1/2 of the pitch, but this does not appear to be convenient, since this would require a greater quantity of magnetic material, with consequently greater cost and weight, in the movable armature, which is probably longer than the inductor, since it has to extend over the whole path of the linear motor.

In constructing a motor, it is necessary to avoid the depletion of the force when the equilibrium position, in other words that of minimum energy, is reached. It is therefore necessary to provide a plurality of poles, to be supplied in sequence according to the relative positions of the inductor and armature.

An advantageous number of poles is three, since this simplifies the construction of the motor, minimizes the power supply and provides a theoretical

force which is completely constant in all relative positions of the inductor and armature.

The structure of the motor may become that shown in Figs. 8 and 9. The three-phase inductor consists of a ferromagnetic core 27 which terminates in three poles 17, 18 and 19, suitably shaped to concentrate the flux in the elementary poles 28 of pitch p.

The closure of the magnetic flux of each pole, which in Fig. 5 took place in the same pole, now takes place more conveniently through the other two poles and through the part of the inductor 23 opposite the core 27, exactly as in a three-phase E/I transformer. The excitation windings of the three phases, 20, 21 and 22, are wound on three insulating liners 28, 29 and 30, by a method completely identical to the technique of constructing a three-phase transformer.

The distance P between the poles and the pitch p of the elementary poles must be such that (6) P = p * (n + 1/3)

Thus, if (as shown in Fig. 8) the elementary poles of the pole 17 are perfectly aligned with the armature 24, the elementary poles of the pole 18 are out of phase by 1/3 of the pitch in advance and those of the pole 19 by 1/3 of the pitch in delay.

It is now possible to determine certain simple relations which give the force produced by the motor shown in Figs. 8 and 9 in a particular, highly advantageous, case of three-phase sinusoidal power supply.

It has been seen that, as a first approximation, the force produced by a single phase is of the sinusoidal type as a function of the displacement and is proportional to the square of magnetic induction B. The force produced by one pole or phase therefore has the form (7) fl = C * il ² * cos(2 * π * x/p)

where C is a suitable constant which depends on the construction of the motor, il is the instantaneous supply current, x is the relative displacement between the inductor and armature, and p is the pitch of the elementary poles 28.

The graph Gl in Fig. 7A shows the variation of the force fl generated by the first phase or pole 17 as a function of the relative displacement x for a constant excitation current related to the value of the current il.

The force produced by the other poles 18 and 19, which are mechanically out of phase by 1/3 of a rotation or by 2ττ/3 radians, will have the form (8) f2 = C * i2 ² * cos (-2 * π/3 + 2 * π * x/p) (9) f3 = C * i3 ² * cos (2 * ττ/3 + 2 * π * x/p) where i2 and i3 are the excitation currents of the corresponding windings.

The graphs G6 and G9 show these forces, again in the case of constant excitation currents, again related to the corresponding values of i2 and i3.

To obtain a unidirectional and usable force, the excitation currents must not be constant, but for each phase the excitation current must be dependent on the relative position x of the armature and inductor as indicated in Fig. 7A in graphs G2, G5 and G8.

For example, if phase 1 is analyzed, it will be seen that the absolute value of the current il is maximum (points A and B) when the force produced in the positive direction is maximum, and is zero (points C and D) when the force produced in the negative direction is maximum.

It should be noted that the direction of the current has no effect on the direction of the force produced, the force being proportional to the square of the current: graph G3 shows, for phase 1, the square of the excitation current il, represented by graph G2.

Graph G4, obtained by multiplying graphs G3 and Gl point by point, represents the force produced by phase 1.

Identical considerations apply to phases 2 and 3, and the resultant forces produced are shown by graphs G4, G7 and G10.

The sum F of the forces produced by the three phases, shown superimposed on graphs Gil, G12 and G13, is the horizontal straight line G14. In mathematical formulae, the currents shown by the graphs G2, G5 and G8 are expressed by the three following equations:

I * cos(rr * x/p) I * cos(2 * rτ/3 + π * x/p) I * cos(-2 * JT/3 + π * x/p) Given that

(13) (cos(y/2)) ² = (l+cos(y))/2 and substituting (10) in (7), the following transformations are obtained: fl = C*I ² /2*(l+cos(2*rτ*x/p) )*cos(2*7T*x/p) =

C*I ² /2*(cos(2*ττ*x/p) + (l+cos(4*rr*x/p) )/2) = C*I ² /2*(l/2 + cos(2*π*x/p) + cos(4*π*x/p) ) /2) and therefore

(14) fl = C*I ² /4 * (1 + 2*cos(2*7T*x/p) + cos(4*7T*x/p) ) It will be noted that this expression of the force of phase 1 (represented graphically by G4) contains a constant positive term C*I /4 and two sinusoidal terms with frequencies equal to and double that of the pole slip. By developing the calculations for phases 2 and 3. it is found that for both the constant term is identical, while the sinusoidal terms are identical to those of phase 1, but out of phase by 1/3 of a period positively or negatively. When the contributions of the various phases are added together, the sinusoidal terms, which form two perfect three-phase systems, give a zero sum, while

the three constant terms are added together. Finally, therefore, the total force developed is: (15) F = 3/4 * C * I ² which is represented graphically by the constant horizontal line G14.

It may be concluded that the three-phase power supply system provides a constant force perfectly smoothed with the variation of the relative position of the inductor and armature, at the cost of a modest 25% of loss with respect to the maximum force developed by a single phase.

Naturally, in the practical construction of the motor there will be residual modulations of the force, as is the case in other types of motor, due principally to the fact that the force developed by a single phase, supplied with direct current, is not perfectly sinusoidal as a function of the slip between the armature and inductor.

Another source of modulation of the force is the fact that the excitation current is not perfectly sinusoidal: with the three-phase system, the third harmonic is advantageously canceled from the current waveform.

In the case of a motor with a number of phases other than 3, the results are identical if the currents supplied to the various phases are kept in phase with the position of the movable armature with respect to the inductor, and are at a frequency half that of the slip of the poles of the movable armature past the inductor.

It is interesting to note that in the motor shown in Figs. 8 and 9 the attractive force through the air gap is again zero overall with respect to the armature, while it is exerted between the two parts 27 and 23 which form the inductor. This force is discharged onto the base 25 which can easily be

constructed with sufficient rigidity. Consequently there are no forces of reaction on the moving parts.

Figs. 8 and 9 show a motor which, for the sake of constructional simplicity, has excitation windings on side 27 of the inductor only, the opposite side 23 being a simple passive encloser of the flux. However, the excitation windings may be distributed on both sides of the motor, the 1-eeper 23 being replaced by an active inductor identical to 27. An important variant of the motor, which may multiply the force generated, requires the presence of a plurality of movable armatures (carrying elementary poles, in other words keepers) aligned with each other and having inserted between them fixed armatures carrying keepers, in other words elementary poles aligned with the elementary poles of the inductor.

Figs. 10 and 11 show a possible non-restrictive embodiment of this variant.

The two parts of the inductor 31 and 32 are identical to the preceding parts 27 and 23 in Fig. 8. However, there are two movable armatures 33 and 34, separated by a fixed armature 35.

The armatures 33 and 34 are movable in the longitudinal direction X and are integral with each other, being secured in a common support 41.

The fixed armature 35 is integral with the inductor, being secured in the support 40 which links the two parts 31 and 32 of the inductor.

The force acting on the movable armature in the direction X is, for a given flux, double that of the motor shown in Fig. 8, since there are twice as many air gaps: the line of flux 42 in Fig. 10 passes through four air gaps tl, t2, t3, and t4, and therefore generates a force four times greater than that which could be generated by a conventional motor in the same conditions.

The number of movable armatures (each of which generates force) may be increased until a magnetizing force is available to generate the desired magnetic induction through the sum of the air gaps. With a high number of movable armatures there may be an unfavourable reduction of the flux in the movable armatures farthest from the excitation winding, as indicated in Fig. 12. Only pr.rt of the flux generated by the pole 43 arrives at the opposite fixed pole 46, since with every passage through an air gap some of the lines of force undergo lateral leakage. This leakage of the flux also results in a certain attraction of the movable armature towards the energized part of the inductor. In this case it may be convenient to provide, as mentioned above, an excitation winding on each side of the motor, or even supplementary windings on the fixed intermediate armatures, to provide a more favourable form of the lines of force of the magnetic field and to eliminate the attraction of the movable armature towards the inductor.

With this motor structure, certain important considerations arise concerning the overall force developed. A pair (movable armature) + (fixed armature) has a total width of 2h, which must be proportional to the pitch p of the elementary poles. By way of example, it will be assumed that h — p, and also by way of example it will be assumed that the optimum value of t/p is 0.2.

The total force of the motor comprising n movable armatures then becomes, given formula (4),

(15) Fn = n * 0.2 * s * B * B / (2 * μ0)

The total volume of the space between the poles is

(16) v = 2 * h * n * s = 2 * p * n * s Therefore,

(17) Fn = 0.1 * v * B * B / (2 * μO) / p

It may therefore be concluded that the force developed overall by this type of motor is proportional to the volume and inversely proportional to the pole pitch: it is therefore convenient to reduce the pole pitch to obtain motors with greater force for a given volume.

A numerical example will give a better understanding of how this occurs. It is assumed that there is a three-phase inductor as shown in Fig. 8: the core is that of a three-phase transformer and the windings occupy the whole space available between the three columns 17, 18, and 19. It is assumed that the space to be provided for a winding is 800 mπr in total and that y the maximum current density is 3 A/mm ; this gives a resulting magnetizing force ni = 2400 At.

If the induction B to be obtained is 1.5 tesla, the maximum air gap must be: t = ni * μO / B = 2.01 mm

It is also assumed that the optimal t/p of formula (5) is 0.2; if the pitch p of the elementary poles is 10 mm, there may be only one air gap, and therefore the maximum total force in formula (4) will be:

F = t/p * s * B * B / (2 * μO) = 18 N/cm ²

If elementary poles with a pitch of 1 mm could be made, the air gap would be reduced to 0.2 mm, and therefore, with the same magnetizing force, 10 air gaps would be available. The maximum force would become 180 newtons per cπr of pole surface, while the inductor, and therefore its volume and weight, and the overall dimensions of the motor, remained identical.

It should be noted that, for a given speed of movement of the movable armature, the frequency of the supply to the inductor would increase by a factor of 10, and therefore the supply voltage would also have

to increase by a factor of 10 so that the energy balance would be retained.

Evidently, difficulties of construction increase when the pole pitch is decreased, since greater mechanical precision is required (the air gap becomes smaller), as well as better magnetic materials and better excitation current generators (the operating frequency becomes higher) .

The construction of high-frequency current generators is already part of the present art and forms no hindrance to the reduction of the pitch.

The mechanical precision appears to be obtainable at reasonable cost for pole pitches down to a few mm.

The operating frequency of the magnetic materials, however, must be evaluated more carefully.

The characteristics of the most common magnetic materials are:

TYPE Bmax Permeability fmax (Hz)

(tesla) (μr) sheet iron 1.8 >5,000 400-1,000 ferrite 0.35 >3,000 >20,000 pressed 1.8 <300 >20,000 powder

The operating frequency to which the magnetic materials are subjected differs substantially between linear and rotary motors.

It is assumed, as a typical but non-restrictive example of the embodiment of the motor, that a pole pitch of 4 mm is required. Given that the electrical frequency is half the frequency at which the movable poles slip mechanically over the fixed poles, three examples can be calculated:

Linear motor, maximum speed of movement 60 m/min. : electrical frequency = 125 Hz Rotary motor, rotor diameter 200 mm (314 poles), maximum speed of rotation 6000 r.p.m. : electrical frequency = 15,700 Hz

Rotary motor, as above, but with a maximum speed of rotation of 120 r.p.m.: electrical frequency = 314 Hz.

It may be concluded that sheet iron is optimal for linear motors and slow rotary motors, but not suitable for fast rotary motors.

Ferrite is unsuitable for making elementary poles, ov.άng to the low saturation induction; since the force is proportional to the square of induction on the poles, the force would be approximately 25 times lower than that obtainable with sheet iron or pressed powder.

Pressed ferromagnetic powder has a low relative permeability and therefore requires a higher excitation current. For movable and fixed armatures, which close the lines of flux over a limited path, the percentage increase of magnetizing force is minimal and therefore tolerable, while for the inductor and the opposing keeper there may be a significant percentage increase.

However, it is interesting to note that induction in iron is high near the elementary poles, but that (Fig. 10) its value is approximately halved inside the inductor 31 and the closer 32, since, as a first approximation, the flux is present only at the ends of the poles which are half full and half empty.

Moreover, there is no reason why the section of the pole inside the winding should not be greater than the active terminal section, to obtain a distribution of the flux over a greater section and consequently a further reduction of the mean induction. Indeed, a variant of the invention provides for the poles to be shaped as shown in Fig. 13.

If the inevitable leakage is disregarded, the magnetic flux remains constant through the movable armature 50, the concentrator 49 and the pole 48 which is excited by the winding 47: the mean induction

inside the pole 48 is therefore equal to the maximum induction on the ends of the elementary poles multiplied by the ratio between the overall surface of the elementary poles and the surface of the pole 48. If the inductor 49 is made from magnetic powder, its greater section proportionally reduces the magnetic reluctance and consequently the greater magnetizing current required.

A differential use of materials is also possible, with pressed magnetic powder being used for the concentrator 49 and armatures 50, and ferrite for the inductor 48.

By using ferrite or pressed powder or a combination of the two, fast rotary motors, with electrical supply frequencies possibly exceeding 20 kHz, may be made according to the invention.

Certain important aspects of the practical construction will now be disclosed.

It has been seen (formula 17) that the reduction of the pitch of the elementary poles is convenient because it enables a greater force to be obtained per unit volume.

By reducing the pitch a higher cost is incurred for making the high number of keepers which form the elementary poles, which also require greater precision if the pitch is reduced.

It is also necessary to reduce proportionately the thickness of the armatures (2h in Fig. 12) which can bend laterally more easily; the simultaneous reduction of the air gap increases the risk of lateral friction between the moving parts and the fixed parts, with the risk of serious damage if the poles become interlocked.

The invention therefore also provides a possible form of construction which provides optimal geometrical repeatability, strength, and economy of production. This form of construction is shown in

Fig. 15, in which three orthogonal sections of an armature are shown by way of example.

The armature basically consists of an insulating non-magnetic support 55, which contains a series of transverse apertures 54 which have the exact dimensions of the elementary poles of the motor; the apertures will therefore have different dimensions and positions according to what has been disclosed above, depending on whether there is a movable armature or a fixed armature.

Packs of plates 56, formed from magnetic sheet iron of the type normally used for motors and transformers, are fitted precisely inside these apertures. These packs form the transverse keepers. Two thin sheets 57 of insulating non-magnetic material are fixed laterally to the armature and keep the plates in position.

This structure may be formed with the techniques and materials normally used for making printed circuits; the insulating material may be resin-glass fibre laminate, and the two thin lateral sheets, also made of resin-glass fibre laminate, may be glued as in the case of a multiple-layer circuit.

The packs 56 may be made of ferrite or pressed powder as well as of plates of the transformer type. In the case of pressed powder, the powder may be pressed directly in the apertures present in the resin-glass fibre laminate, with higher economy and precision of mounting. The structure of a linear motor made by this technique is shown schematically in Fig. 16.

The inductor 58, the windings 59 and the terminal keeper 60 are similar to the preceding ones, but the concentration of the flux in the multiplicity of the elementary poles is provided by the structures 61. These structures are similar to those of the fixed armatures 62, with the only difference that the

protective sheet 63 is not present on the side of the inductor or its terminal short-circuit, to avoid an unnecessary air gap.

Since the keepers are housed precisely in the cut-outs formed in the insulating support, there is an optimal alignment, which becomes more important as the poles are brought closer together.

The movable armatures are also constructed by the same technique. The precise mounting of the movable and fixed armatures is assisted by the spacers 65, which force the armatures into suitable positions to obtain the requisite air gap.

With this structure the air gap is partly filled by the covering sheets, which also make the surfaces of the poles smooth and uniform. The reduced freedom of lateral oscillation is particularly useful in the case of thin armatures, which have lower rigidity. Furthermore, any occasional friction is less likely to cause damage.

In the preceding text, reference has been made to linear motors, with the proviso that theoretically the same considerations are applicable to rotary motors which may be formed by a simple "bending into a ring" of the linear structures illustrated hitherto.

Evidently, the considerations put forward hitherto in relation to the working forces developed in linear motors must be understood in relation to the working torque in rotary motors. An example of a rotary motor is shown in Fig. 17, in which certain details of the armatures which are identical to those of the linear motor are not shown.

The armatures are in disk form: the fixed armatures have their keepers grouped in three poles (Fig. 17A) disposed at 120 degrees to each other, while the movable ones (Fig. 17B) have them disposed uniformly over their whole circumferences.

The inductor (section in Fig. 17C) consists of three columns 68 on which the three windings 69 act, while the two terminal toroids 70 provide the magnetic circuit closure. To prevent parasitic currents and consequent electrical losses, these terminal toroids may consist of a winding of thin magnetic sheet.

Possible variants of the rotary motor may have 6 or more poles (providing better distribution of the forces and consequently a smoother motion) or the closure of the pole flux in each pole (as in Fig. 5), or windings on both sides of the inductor.

An important final component of the motor is the position sensing system, which is used to keep the three-phase supply current correctly phased according to the position of the movable armatures, and which may also be used for position control.

For this purpose, use may be made of linear or rotary sensing systems available on the market (encoders, optical lines, etc.) or, advantageously, an inductive position transducer with a passive scale of the type described and illustrated in Italian patent application 9383a/90, corresponding to US Patent No. 5233294, the content of which is incorporated in the present description. In the last case, the passive scale of the sensing system consists of the movable armature itself, whose keepers repeated with a regular pitch are sensed by the inductive proximity sensor integral with the fixed part. Fig. 14 shows an example of a possible disposition of the proximity sensor 51, facing the movable armature 52, in the space available between the poles of the inductor 53.

The example in Figs. 8 and 9 is an illustration of a motor in which are provided three poles, each consisting of a corresponding pole expansion 17, 18,

19, on which the corresponding winding is completely wound.

In order to optimize the use of space and consequently to generate a greater force (with the same overall dimensions) , and also to make the force exerted on the armature more uniform, it is possible to provide an embodiment similar to that of three- phase asynchronous motors, where the windings are not concentrated on three separate pole expansions, but are suitably distributed over the whole extension (linear or annular) of the motor.

Figs; 18 and 19 show schematically one embodiment of this type. The numerals 27 and 28 again indicate the elementary poles of the armature and of the inductor respectively. The elementary poles 27 are equidistant from each other, as are the elementary poles 28. The windings (only indicated in Fig. 18), distributed in a way similar to that provided, for example, in three-phase asynchronous motors, are disposed in the cavities formed between every two elementary poles 28 of the inductor. For each pole, the armature has two elementary poles 27 less than the inductor (or vice versa) , to ensure the generation of a thrust on the armature at every instant. Fig. 19 shows highly schematically the distribution of the windings of the three phases in the case of a linear motor with three poles. The diagram in Fig. 19 shows that the distribution of the windings is not uniform, as in rotary three-phase asynchronous motors, owing to the fact that the armature and inductor are developed in a plane instead of being annular. This causes the occurrence of edge effects, the incidence of which may be reduced simply by elongating the inductor and then repeating the disposition illustrated in Figs. 18 and 19 a sufficient number of times.