This invention relates to switched reluctance motors such as those described by TJE Miller, Oxford Press, 1993: 'Switched Reluctance Motor and Their Control', Page 1-80.

The concept of using a wheel type structure in a motor has been described in a number of papers, such as US4823039 issued to Lynch. However, the primary methods of exciting the motor in these cases have been similar to their 'cylindrical' counterparts. SUMMARY OF THE INVENTION

According to the invention in a first aspect there is provided an electric motor comprising a rotary shaft-mounted rotor member having a plurality of discretely spaced magnetisable elements disposed peripherally therearound extending co-axially with the axis of rotation of the shaft to form a ring, respectively opposite inner and outer stator winding pairs disposed radially on either side of the ring, each stator pair being co-operable when electrically energized with respectively adjacent magnetisable elements on or in the ring, whereby the ring and hence rotary shaft may be induced to rotate relative to said stator winding pairs upon energisation thereof. • The invention in this respect therefore preferably utilises passive iron members along the periphery of wheel rotor for torque production, with stator windings arranged radially on either side of the rotor. The main advantage of this structure is very high torque per unit iron mass can be developed, which makes the motor ideal for direct drive for e.g. a traction wheel of a vehicle.

Conveniently, in accordance with the first aspect of the invention the electric motor comprises a shaft-mounted non-metallic and/or non-magnetisable rotor member of generally annular shape in which the plurality of discrete magnetisable elements are disposed sequentially peripherally therearound in the form of teeth embedded within a non-magnetic base to form a ring. With this arrangement it is particularly convenient that the inner and outer stator winding pairs and associated stators are secured to the housing case for the motor, which housing case may itself be annular or otherwise in the form of a wheel in which the rotor member and hence rotary shaft on which it is mounted is rotatable relative to the housing.

However, it will be understood by those skilled in the art that the invention in another aspect may be easily adapted to a motor of non-rotary configuration such as a linear motor or other type of motor with a non-circular rotor. Accordingly, in a second aspect of the invention there is provided an electric motor comprising a moveable rotor member having a plurality of discrete magnetisable elements disposed therealong or therearound, which magnetisable elements are excitable and energisable by respectively opposite stator winding pairs on either side thereof, whereby the rotor member may be forced to move relative to said stator winding pairs when so energized. In a third aspect of the invention there is provided an electric motor comprising a non-magnetic rotor and a plurality of . discretely spaced magnetisable elements disposed at peripheral portions thereof, respectively opposite stator winding pairs arranged on either side of a line of movement of said magnetisable elements when said magnetisable elements are excited and

forced to move relative to said stator winding pairs when energized.

Accordingly, with these various embodiments of the invention, depending upon the type of motor concerned, the rotor can be generally circular, or generally non-circular, or may include discrete linkages, such as caterpillar linkages, such that the rotor describes a non-circular path upon excitation of the magnetisable elements by the respectively opposite stator winding pairs.

Alternatively, the rotor may be flexible, such as in the form of a ribbon or tape, in which the magnetisable elements are excitable by respectively opposite stator winding pairs on either side thereof to thereby force the rotor to move relative thereto when so energised.

The stator winding pairs can conveniently be mounted on generally U- shaped stator yokes and the rotor be in the form of a non-magnetic or noh- magnetisable base plate with a rim of non-magnetic but magnetisable elements in the form of teeth separated by non-magnetic and/or non-magnetisable ¹ ^ material in the form of tongues.

The base of the rotor can also conveniently be in the form of spokes, to thereby reduce weight.

The motor may conveniently include a housing case to which is fixed stator winding pairs and respective stators and through which housing runs a rotor shaft to which is secured a rotor having discrete magnetisable elements which travel between the stator winding pairs. The housing case conveniently includes a cover by which the motor is substantially enclosed, and the housing case and the cover are each preferably secured to respective portions of a motor drive shaft on bearings.

In a preferred embodiment, the adjacent respectively opposite inner and outer stator halves are integrally connected to form respective sets of integral inner stator halves and outer stator halves, which may conveniently each or both form a ring like structure. This may be achieved by the stator halves being connected by respective bridging webs to form respective sets of integral inner and/or outer stator halves, and bridging webs may further include means, such as apertures, for securing the stator assembly to e.g. the inner housing of the electric motor.

As will be appreciated by those skilled in the art, the electric motor may also be used as a generator when a suitable activation sequence is applied.

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which:

Figure 1 is a side view of a conventional prior-art switched reluctance motor with 8 stator poles and 6 rotor poles, ^{• •} - Figure 2 is a side view of a switched reluctance motor according to an embodiment of the invention (only one stator shown),

Figure 3 is a cross sectional view of half of the motor of Figure 2,

Figure 4 is a perspective view of the rotor member for use with the motor of Figure 2, Figure 5 is a perspective view of the rotor member of Figure 4 with stators evenly distributed along its periphery,

Figure 6 is a perspective sectional view of the completely assembled synchronous reluctance motor of the invention within a fully sealed housing,

Figures 7 (a) - (c) show, respectively, side, cross sectional and developed

(rolled-out) views of a section of the rotor of Figure 4,

Figure 8 shows a developed diagram of a section of stator winding pairs with part of the rotor,

Figure 9 shows a developed diagram of the relative positions of all three upper (outer) stator windings relative to a section of rotor,

Figure 10 is a schematic view of a drive circuit for the synchronous electric motor,

Figure 11 is a schematic view of the magnetic circuit of the synchronous

electric motor, Figures 12 (a) - (c) show the respective relationship between the alignment profile, the inductance profile and the developed (rolled-out) structure of the stator and rotor assembly of the motor,

Figures 13 (a) - (d) show the activation sequence of the three stators of the motor of Figure 4 as the rotor moves, * ^;ϊ* Figures 14 (a) - (c) show the variation of magnetic flux distribution in three different alignment or overlapping positions,

Figures 15 (a) - (k) show finite element results of the magnetic flux variation when the rotor is excited and travels through 30°,

Figure 16 shows respective graphs giving a simulation of a typical start- up mission for the synchronous electric motor, and

Figure 17 is a view corresponding in part Figure 2 but showing a preferred embodiment of stator assembly. DESCRIPTION OF THE INVENTION

Referring firstly to Figure 1 there is shown generally at 1 a conventional

prior art switched reluctance motor (SRM) having an annular stator 2 with eight inwardly facing stator cores or poles 3, around each of which are respective pairs of windings or coils 4, 4', 5, 5\ 6, 6', 7, 7', 8, 8', 9, 9', 10, 10', 11 , 11 ', surrounding a rotor 12 mounted for rotation on a shaft 13, the rotor 12 having six rotor poles 14, 15, 16, 17, 18, 19. The invention described later with reference to Figures 2 et. seq. is also a motor of switched reluctance type, and therefore shares some important features of operation with the conventional switched reluctance motor (SRM) of a cylindrical structure shown in Figure 1 , which is a doubly salient, singly excited synchronous motor. The rotor 12 and stator 2 consist of stacked iron laminations with copper windings 4,4' - 11 ,11' on the poles 3 of the stator 2.

Because no permanent magnets or windings are on the rotor 12 all the torque developed in the SRM 1 is of reluctance nature. To excite the motor 1 , a power electronics inverter (not shown) energizes appropriate phases based on relative position of the rotor 13 and hence poles 14-19 to the stator core 3.

Torque is produced by the tendency of the motor 13 to move to a position where the inductance of an excited winding is maximised, and the reluctance minimised. This happens, for instance, when the rotor poles (14 and 17) are fully aligned with their respective stator poles 3 when stator winding pairs (4,4' and 8,8') are energized, as shown in Figure 1. Each winding usually consists of a number of electrically separate circuits or phases. These are usually excited separately in pairs or, less commonly, with more phases.

In operation of the motor 1 , each phase is usually excited when its inductance is increasing. Thus coils 4-4' and 8-8' should be fully unexcited and

coil 5-5' and 9-9' should start to excite for rotating in the direction shown arrowed. In generating electricity, the opposite is true. It is also to be noted that the direction of rotation depends on the switching sequence rather than direction of current flow in the coil. The excitation of a phase creates a magnetic field that attracts the nearest rotor pole (14, 15, 16, 17, 18 or 19) to an excited stator pole 3, in order to minimize the reluctance path through the rotor 12. Excitation is applied sequentially to the phase windings to achieve continuous rotation. Due to the concentrated windings and the manner the motor 1 is excited, the motor is replete with high torque ripples and is also acoustically noisy. These are the two well-known disadvantages of the conventional SRM shown in Figure 1.

While the conventional SRM is simple in principle, it is also rather difficult to design and develop performance predictions. This is mainly due to the non¬ linear magnetic characteristics of the motor under normally saturated operation: *•> Thus, perhaps to a greater extent than in other types of motor, the structural design of the SRM requires careful attention. These, together with the two disadvantages referred to above, have made the SRM not widely accepted in industry.

Figures 2 - 6 show the mechanical structure of an SRM according to the first aspect of the invention which can provide a wheel-like SRM within a fully sealed casing (as shown in Figure 6) and is particularly suitable for high-torque applications, such as for a directly driven traction wheel.

Figure 2 schematically shows the mechanical structure of the motor in a first ^• aspect, where only one of the three sets of stator windings in this

embodiment is shown for clarity. The SRM, shown generally at 20, consists of one or more stators 21 (typically 3), each comprised of oppositely disposed inner stator 22 and outer stator 23 halves, each having respective pairs of stator cores 24 and respective stator winding pair or coils 25,25'. The inner 22 and outer 23 stator halves are disposed radially on either side of discretely spaced magnetisable rotor elements 26 disposed peripherally around and extending co- axially with a rotor shaft 27, the magnetisable rotor elements 26 being secured for rotation upon a flanged rotatable rotor member 28, typically of plastics construction, and being separated from each other by upstanding plastics spacer tongues 29 which, with the magnetisable rotor elements 26, collectively define a rim or ring forming part of the rotor member 28 which travels in a circular path around the axis of the rotor shaft 27 and between the inner and outer stator halves 22,23 (in a manner shown more clearly with reference to Figure 3), being separated therefrom and therebetween by a small air gap of -^f-* e.g. 0.3mm. This gap is small enough so as to enable efficient continuity in the magnet circuit, and large enough for reasonable ease of manufacture, when the motor (20) is in operation.

In Figure 3 it will be seen that the rotor shaft 27 is mounted on bearings 30,31 embedded within, respectively, a motor housing case 32 of generally annular shape and a motor housing case cover 33 securable thereto, the case

32 and cover 33 defining in combination a hollow housing in the form of a closed drum or wheel.

The rotor member 28 is fixed for rotation on the rotor shaft 27 so that the magnetisable elements 26, which are disposed peripherally therearound and

extend co-axially with the axis of rotation of the shaft 27, pass between respectively inner and outer stator halves 22,23 of each stator 21 , the stator halves 22,23 being secured to the motor housing case 32 by e.g. bolts (not shown). Each phase therefore includes one pair of windings disposed radially on either side of the rotor's rim.

Figure 3 also shows that the forces developed on the rotor member 28 will be balanced in the radial direction, thus producing virtually no radial stress on the rotor shaft 27.

Figure 4 shows a perspective view of the rotor member 28 and magnetisable elements 26 mounted on the rotor shaft 27, which rotor member

28 includes a number of spokes 34 which thereby provide a cost saving in terms of materials used and a weight saving for the SRM 20 while still maintaining mechanical integrity of the entire structure, particularly the radial position of the magnetisable elements 26 relative to the stator halves 22 and 23 shown in ,ix Figure 3.

Figure 5 shows the rotor member 28 with all three stator sets of inner and outer stator halves 22,23, spread radially evenly around the whole periphery of the rotor shaft 27. It should be noted that the three stator sets of stator halves 22,23 can also be positioned if preferred in just one half of the rotor assembly's periphery. The latter may result in smaller overall size of the wheel, and may be suitable for certain applications, whereas in the former case torque generation is spread more evenly around the rotor member 28 and thus there is relatively less overall mechanical stress. However resulting torque quality, in terms of torque ripples, is similar in both cases. It should also be noted that it is possible to have

six (or some other convenient number) pairs of stator halves evenly spread around the rotor's periphery, with the resultant torque being varied accordingly.

In a general preferred embodiment, the invention can be housed in a sealed wheel or drum-like enclosure as shown in Figure 6. This makes the invention look like, externally, a conventional disc motor such as the Lynch motor described in US4823039.

As noted above, the rotor member 28 is a slotted or spoked ring structure made up of rigid and non-magnetic plastic type materials such as grey p.v.c. or carbon fibre. The rotor member 28 provides secure positions for 12 discretely positioned magnetisable iron elements or segments 26, which are fastened inside each of 12 correspondingly shaped slots 35 between non-magnetisable (e.g. plastics) spacer tongues 29 along the ring-like periphery 36 of the rotor member 28. The rotor member 28 and the 12 fixed iron magnetisable elements 26 therefore form a seamless ling 36 with no protrusions. For analysis purpose, ««• the rotor structure can be seen as making up of 12 identical functional segments, one of each is shown in Figure 7(a). Each functional segment has an overall angular length of 30 degrees. It is made up of an iron segment 26 of 15 degrees, and a non-magnetic segment 29 of 15 degrees. The 15-degree iron segments will create, as will be seen in the next section, an 'overlap' zone of 5 degrees during rotation. The overlap zones are very critical to the torque characteristic of the motor. No overlapping usually results in some non-torque zones. If the rotor lies in the non-torque zones, no torque can be generated regardless of the input power. Neither can there be too much overlapping since the torque generated by each phase could be acting against each other more

easily.

In Figure 7(b) the 'rolled-out ¹ form of one phase of a winding pair is shown. The stator core for each phase is made up of an outer (upper) and inner (lower) part. By connecting the windings in series and having the same number of conductor turns wound on each core, a similar level of magnetic flux will be generated on both sides of the rotor. Thus, as shown in Figure 8, the cross- sectional area of windings are virtually equal for the outer and inner stator cores. Similarly, the outer iron and inner iron stator cores are also virtually equal in size. The advantage of this stator design is that the magnetic forces generated in the radial direction will be cancelled out. Thus, null or minimum radial stresses will be exerted on the rotor-member, which could otherwise be serious for large wheel size. The torque is generated by the magnetic forces in the tangential direction. It is also to be noted that the stator can be conveniently adapted to be used as an arm-rest for mobility aids applications. As ^~ also noted above, the stator consists of three sets of separated winding pairs mounted on individual iron cores. This is illustrated by Figure 9, in which only the upper (outer) stator members are shown for clarity. Thus, in this embodiment the machine is regarded as a 3-phase motor. Theory of Operation Although the basic operation theory is similar to conventional SRMs as discussed with reference to Figure 1, the invention operates ^' more like a linear motor, and there are also some features specific to the motor of the invention.

Figures 10 and 11 respectively show the basic electric drive circuit and magnetic circuit for a phase winding. The electric drive circuit in Figure 10 is a

typical per phase power inverter circuit for the SRM. When power switches (SWu and SWl) are turned on with the same switching signal, current will flow through the stator winding (inductance) L from the supply Vdc via the power switches SWu, SWl, producing motoring torque and storing magnetic energy in L. Supply Vdc is the DC link power supply. When the power switches are off, current will continue to flow via diodes (Di and D ₂ ) back to the supply. Since current flows against the supply, it will decrease to zero before end of the cycle. During this period, stored magnetic energy returns to the supply. However, this apparently simple operation is made very complex due to the variation of the value of inductance, the variation being due to the excitation current level / and the position of the moving iron rotor, as illustrated in Figure 11. It should be noted that this variation is true for SRMs in general, not just for the motor of the invention. The variation of L is also highly dependent on the magnetic materials used. Complex control methods are often needed to ensure the correct profile of excitation current to produce the required torque level.

Figures 12(a) - (c) show the relationship between the variation of inductance and the degree of relative alignment between the stator and rotor members. In practice the inductance L has a complex relationship with excitation current and relative position of the rotor to the stator. To explain this, however, a low excitation current level has been used to obtain a linear, near-idealised, inductance profile-. The alignment profiles for phase A, phase B and phase C are shown in Figure 12(a). The alignment profile is linear and of triangular shape. This can be explained when the profile for A is referred to the relative position of the stator and rotor of phase A, shown in the developed (rolled-out) diagram in

Figure 12(c). When point p (on the rotor) is directly under point a (on the stator), full alignment occurs for phase A, corresponding to zero angle at the alignment profile graph. When p moves under b, full misalignment occurs, corresponding to the apex of the alignment profile graph. The alignment angle decreases to zero again (fully aligned) when p is directly under c. The profile for phases B and

C are shown accordingly.

The second set of graphs in Figure 12(b) show the inductance profile for the respective phases. Figure 12(c) shows a rolled-out diagram of the relative alignment of the stator and rotor as a reference. For illustration, the inductance profiles shown are generated by a very low excitation current (e.g. 0.1A) to ensure magnetic saturation does not occur in the iron cores. Thus there is a linear relationship between the excitation current and the level of magnetic flux generated, resulting in a nearly-linear inductance profile. It is to be noted that the inductance profile is to some ^' extent an 'inverted' version of the alignment profile. -* ^• This is because for a given phase, maximum inductance occurs when the rotor is fully aligned with the stator (alignment angle = 0°). However, in practice, the very low excitation current will not generate enough torque. Most machines in fact operate at close to the magnetic saturation point to fully utilize the magnetic material. At more usual excitation levels of over 10A for this motor, the inductance profiles could be more non-linear and rather different from the one shown here.

Figures 13(a) - (d) show the switching or activiation sequence for this embodiment of wheel SRM, again using a developed diagram. It shows only the upper stator windings and the rotor in a 'rolled-out' form for ease of illustration.

The stator windings are each separated by 25°. Thus, in Figure 13(a), with phase A fully aligned with the rotor, phase B has a 'forward' overlap zone of 5° with the rotor, and phase C has a 'backward' overlap zone of 5°. A 'forward' overlap will generate a forward torque (to the right) when excited, and a 'backward' overlap will generate retarding torque. The choice of these overlaps is important in determining the torque characteristics of the motor, and in particular minimising torque ripples.

The sequence of diagrams in Figures 13(a) - (d) aim to illustrate graphically the switching sequence, or more precisely the activation sequence, as will be seen, of the stator windings required to develop a smooth and unidirectional torque to move the rotor to the right. In the top diagram, excitation current flowing in stator winding A must be reduced to zero as soon as possible, as further flow of current will result in pulling the rotor to the left, i.e. a negative torque to the direction of rotation. Winding A is therefore finishing-activation. At w the same time, winding B is in its full activation state as it is the only winding that can produce the main positive torque. However, it does not necessarily mean that during this activation period winding B is fully turned on all the time. The power switches must control the current into winding B at a certain level to develop the appropriate torque. The activation period is therefore not always a switched-on period, but one in which high frequency on-off switching occurs.

Also, during this activation period for winding B, winding C is in the non- activation period, when it cannot develop any positive torque. In the non- activation period, the power switches in the respective phase will be off all the time.

The activation sequence in the following Figures for different rotor travelled angles can be explained in a similar way. It is important to note that the control of the starting and finishing of the activation periods may, in practice, be very dependent on the driver circuit. A 'weak' driver may not be able to control these periods precisely, leading to degraded torque capability.

However, it will be understood by those skilled in the art that when the aforementioned activation sequence is reversed in relation to the rotor movement, the invention can then be readily operated as a generator.

Finite element (FE) analysis allows flux and torque to be very precisely determined for any given geometry of rotor and stator, and has been used for such analysis of the invention. Figures 14(a) - (c) shows the variation of flux distribution for three alignment (or overlapping) positions. The density of the lines of flux is an indication of the flux level and therefore the magnitude of the force generated. Figures 15(a) - (k) show the FE results and the activation* *r sequence when the rotor is excited to travel for 30 mechanical degrees.

Figure 16 shows the simulation for the performance of a wheel SRM. The rotor's diameter was 25cm and with a thickness of 2cm. The motor starts a load of 120kg at 5° uphill, with a terrain having a rolling coefficient of 0.02. The simulation shows the dynamic change of various parameters during a start-up. It . shows that the torque ripple is un-typically low for a SRM.

Figure 17 shows a preferred construction of stator assembly 21a comprising integral sets of inner stator halves 22a and outer stator halves 23a in each of which the respective stator poles 24a, 24b are connected together and thus made integral by bridging webs 37, 38 for, respectively, the inner stator and

outer stator windings (not shown). Centrally disposed between each of the bridging webs 37, 38 are respective apertures 39, 40 to allow fixing of the sets of stator halves 22a, 23a to e.g. the inside of the housing 32 (shown in Figure 6) such that the entire stator assembly 21a is fixed against rotation. This is a particularly convenient arrangement since it increases mechanical stability of the entire structure and there are fewer individual components even though, as compared to the arrangement shown in Figure 2, there are twice as many stator winding pairs (6 instead of 3). In addition, the effective reduction in integral stator components to just two, the inner and outer sets of stator halves 22a, 23a, makes it easier to manufacture, and to ensure accurate radial alignment relative to the rotor 20 to thereby facilitate the use of a small air gap therebetween within the machining tolerances of these components and thereby maximise the efficient continuity in the magnetic circuit.