EP1735898A2

EP1735898A2 - Electromagnetic coupler for transmitting electric power and a transmission device provided with said coupler

Info

Publication number: EP1735898A2
Application number: EP05739697A
Authority: EP
Inventors: Armando Fonseca; Paul Romagny
Original assignee: Renault SAS
Current assignee: Renault SAS
Priority date: 2004-04-01
Filing date: 2005-03-31
Publication date: 2006-12-27
Also published as: FR2868622A1; WO2005096476A3; WO2005096476A2; FR2868622B1

Abstract

The inventive electromagnetic coupler for transmitting electric power comprises two electric machines provided with two rotors connected to a first shaft (7) and to a second shaft (5), respectively, wherein the armatures of said electric machines are mounted between the two rotors. Said invention is characterised in that the windings (9) of the two armatures are common therefor, a first air gap (E1) is embodied between the part of one rotor and the part of the other rotor, a second air gap (E2) is embodied between the common windings (9) of the two armatures and the other part of one of rotors and an additional air gap (ES) is embodied between the windings (9) and said part of the other rotor in such a way that a main magnetic flux collected in the first air gap (E1) is combined with the magnetic flux from the second air gap (E2).

Description

Electromagnetic coupler for electrical power transmission and transmission device comprising such a coupler

The present invention relates to an electromagnetic coupler for power transmission and to a transmission device comprising such a coupler. The transmission of mechanical power between a source of motive movement and the element to be driven very often requires an adaptation of speed according to the operating regimes. This is particularly the case on motor vehicles where the internal combustion engine must be able to drive the wheels from a standstill to their maximum speed: the transmission then usually includes a coupling device allowing an at least temporary slip (clutch at friction, electromagnetic powder clutch, hydraulic torque converter, etc.) associated with a mechanical gear ratio with variable ratio (gearbox with discrete ratios, mechanical device with continuously variable ratio ...). In order to ensure this adaptation, there are also known, as an alternative to the mechanical arrangements, solutions with electric power transmission: firstly, the mechanical driving power is transformed into electric power by an electric generator machine, then it is again converted under mechanical form by an electric motor. The electronic controls of the generator and the engine then allow total decoupling in speed. It is possible to add to these solutions an electrical storage device (accumulator, etc.) which opens additional opportunities for managing energy flows, in particular with a view to saving fuel consumption or improving performance; (braking recovery, greater freedom of choice of the operating points of the power source according to performance criteria, transient add-on power, start of the combustion engine ...); this type of configuration is frequently called "serial hybridization". However, the serial electric transmission has some drawbacks, particularly in that it requires an intermediate double conversion of all of the energy: this cascade is costly in electronic design and the product of the conversion yields can call into question. 'energy benefit for certain operating modes; (for example, for stabilized speeds where the power source is well exploited with gearboxes). It will be noted that these drawbacks of the electrical transmission are less in the “electrical power derivation” transmission configurations where the flexibility of adapting electrical speeds is exploited in parallel with a mechanical power transmission channel; the electromechanical double conversion of power is then limited to a fraction of the power to be transmitted. Multi-mode transmission with continuous variator described in patent application FR-2 823 281 is for example of this type. It is in this context of electrical transmission that we can locate the "electromagnetic coupler". The principle can be understood from the block diagram of Figure 1 which shows a known electromagnetic coupler. The power source (heat engine, input axis of the electrical power bypass, etc.) rotates through the input shaft 7, the armature 1 of an electric machine M1. This armature 1 is supplied with current through a polyphase electronics 3 and is thus magnetically coupled through the air gap E1 with a second rotor 4. For this coupling, different types of connections can be used: the second rotor 4 can thus include for example magnets, an asynchronous squirrel cage, or even a reluctant toothing, the design of the rotating armature of the machine M1 being adapted accordingly. The second rotor 4 is linked to the output axis 5 of movement; it makes it possible to transmit the torque from the power source directly to it. Control of the rotating field by the electronics 3 associated with the armature 1 allows the speed differential between input and output to be adjusted as desired: depending on the sign of this shift, the armature will be generator or receiver. At synchronism, the motor source will be linked to the output as it would be by direct mechanical coupling, the armature then receiving only the electric power necessary for your magnetization; (under these conditions, the power supply is direct current if the machine M1 is of the synchronous type). The second rotor 4 is also coupled through a second air gap E2 with the armature 2 of an electric machine 2. It is consequently provided at this other air gap with the elements necessary for the corresponding magnetic coupling (magnets, cage asynchronous, ...). The armature 2 is fixed. Its current supply by polyphase electronics 6 generates on the intermediate rotor 4 a complementary torque (additive or subtractive) to that which already comes from the power source. It is understood that the coupler, like an electric transmission, makes it possible to adapt the transmission in speed and torque as desired and possibly to exploit the potential of an electrical energy storage. It can be used either to transmit all of the power as in a serial electric transmission, but also in an electric bypass power transmission. In addition to its power transmission function, it can provide an electricity generation or electric traction function. Compared to the electrical transmission with two separate machines, the electromagnetic coupler has several notable advantages, in particular the following: the dimensioning of the electronics, and therefore their cost, can be reduced significantly: indeed, the desired flexibility can be obtained by only dealing with the electronics with the sliding power between the two rotors, in the same way, the coupler has an efficiency advantage; (lower power passing through the electronics, low iron losses of the M1 machine which operates at low differential speed, or even at synchronism with the only magnetizing power), the principle integration of electric machines makes it possible to obtain compact embodiments.

If such electromagnetic couplers have already been described for a long time, for example in patent AU-5840173, their industrial introduction has been handicapped by two difficulties: the cooling of the rotor armature of the machine M1: this armature 1 indeed has windings difficult to cool because the heat exchange with this rotating part takes place through air (high thermal impedance) and the supply of efficient heat transfer fluids is obviously complicated by rotation. The supply of this armature through sliding electrical contacts 8: these contacts represent an integration constraint on the topological plane, in volume, in terms of compatibility with the physical environment and reliability; they also constitute a significant cost item.

The invention recently described in French patent application 04 00830 of January 29, 2004, an illustration of which is given in FIG. 2, makes it possible to resolve these two difficulties at the same time: the windings 9 of each phase of the armature 1 are formed in toroids and their magnetic coupling to the air gap E1 is achieved by systems of claws 10, 10a, as can be seen in the example of FIG. 3 for one of these windings, additional air gaps ES practiced in these magnetic circuits make it possible to mechanically separate the rotating claws 10 from the windings 9; thus, the windings in question become fixed, and therefore easy to cool effectively while the sliding contacts are eliminated. In this figure 3, the references 13 and 13a denote respectively the rotating surface and the fixed surface of the additional air gap ES. The present invention takes up these solutions of electromagnetic couplers without sliding contact and brings them new progress likely to further improve their efficiency and compactness. The invention thus relates to an electromagnetic coupler for electrical power transmission comprising two electrical machines, with two rotors respectively connected to a first and to a second shaft, the armatures of these two machines being arranged between the two rotors.

10 According to the invention, this electromagnetic coupler is characterized in that the windings of the two armatures are common to these two armatures, a first air gap being formed between a part of one of the rotors and a part of the other rotor, a second air gap being made between the common windings of the two armatures and another part of one of the rotors and an additional air gap being made between the windings and said

15 part of the other rotor, so that the main magnetic flux collected at the first air gap is composed with the magnetic flux from the second air gap. According to another aspect of the invention, the electrical power transmission device comprising an electromagnetic coupler for electrical power transmission comprising two electrical machines, with two rotors respectively connected

20. to a first shaft and to a second shaft, the armatures of these two machines being arranged between the two rotors is characterized in that the windings of the two armatures are common to these two armatures, a first air gap being produced between a part of one of the rotors and part of the other rotor, a second air gap being formed between the common windings of the two armatures and another part of one of the rotors and an air gap

25 additional being produced between the windings and said part of the other rotor, so that the main magnetic flux collected at the first air gap is composed with the magnetic flux from the second air gap and in that the windings are supplied by means of '' a single inverter by polyphase electric currents comprising superimposed components.

30 ^* This electromagnetic coupler preferably comprises several annular windings arranged one next to the other around a common axis, each of these windings being supplied by the single inverter. Each winding is between two magnetic elements each comprising a series of claws annularly arranged around the winding and directed towards the claws

35 of the other element, the claws of one of the elements being disposed between the claws of the other element, these claws being arranged opposite magnets arranged annularly on an outer part of the rotor to produce the first air gap, the number of magnets being equal to that of the claws and the polarity of these magnets being successively reversed. In one embodiment, each coil annularly surrounds two magnetic elements each comprising a series of claws annularly arranged around the axis of the coil and directed towards the claws of the other element, the claws of one of the elements being disposed between the claws of the other element, these claws being arranged opposite magnets arranged on an inner part of the rotor to make the second air gap, the number of magnets being equal to that of the claws and the polarity of these magnets being successively reversed , each winding defining with said claw elements, a wafer. The wafers are angularly out of phase one with respect to the next by 2τr / (nι.p ₁ ) with respect to the first air gap and by 2π / (n ₂ .p ₂ ) with respect to the second air gap, the number n of windings being a common multiple of ni and n ₂ , ni being the number of phases corresponding to the air gap (E1) and n ₂ being the number of phases corresponding to the air gap (-≡2), pi being the number of pairs of poles on the first air gap and p ₂ the number of pairs of poles on the second air gap. The single inverter is adapted to generate in each wafer a first polyphase pulse current ωi phase shifted by 2τr / n. from one wafer to another and a second polyphase pulse current ω ₂ phase shifted by 2ττ / n ₂ from one wafer to another. Other features and advantages of the invention will become apparent throughout the description below. In the accompanying drawings, given by way of nonlimiting examples: FIG. 1 is a block diagram of an electromagnetic coupler according to the state of the art, illustrated here in a concentric arrangement of the two machines, FIG. 2 is a schematic diagram of an electromagnetic coupler without sliding contact with annular globalized windings immobilized by means of additional air gaps, FIG. 3 is an exploded view of an arrangement of magnetic claw circuit with additional air gaps allowing the immobilization of the winding, the Figure 4 is a variant of an electromagnetic coupler without sliding contact: the wound parts of the armatures 1 and 2 are arranged adjacent; in addition, they are both made with windings annulars with magnetic couplings by systems of claws _ rotary claws for the armature 1, fixed claws for the armature 2_, FIG. 5 is an arrangement according to the invention with pooling of the windings of the two armatures, - FIG. 5a is a diagram of the electronics used in the device according to FIG. 5, FIG. 6 is an exploded view of a winding according to the arrangement of FIG. 5 with its double system of claws, the double rotor is represented here in a schematic configuration with surface magnets, - Figure 7 is an equivalent diagram of the magnetic circuit of an exemplary embodiment with compound currents and flux passing through Figure 8 is an equivalent globalized diagram corresponding to that of Figure 7, Figure 9 shows examples of arrangements with compound currents allowing the cancellation of the powerful couples, FIG. 10 shows an example of adaptation to the invention of an asynchronous cage illustrated in perspe ctive on the inner part of the rotor, a non-magnetic space is provided between the magnetic circuits associated with each wafer, - Figure 11 shows another example of adaptation to the invention of an asynchronous cage: the perspective view does not show this time that half of the outer part of the rotor, the busbars have segments offset angularly to achieve the desired phase shift. To understand the invention more easily, it is useful to return first to one of the provisions already presented in French patent application 04 00830 of January 29, 2004. The principle of this arrangement is given in FIG. 4. The general operation is similar to that of the arrangement in FIG. 2. The shaft 7 called "motor" drives systems of claws 10 associated with the armature 1 and which ensure their coupling to the active parts of the air gap E1. The magnetic connection between these rotating claws 10 and the fixed annular windings 9 of the armature 1 takes place through additional smooth air gaps ES. This first armature 1 comprises several wafers placed side by side with an appropriate successive angular phase shift between claws and active parts of the rotor opposite, so as to constitute a polyphase system. This first electric machine, powered by an electronic inverter 3, allows you to transmit the torque from the motor shaft 7 to the motion output 5 with a positive or negative slip adjustable at will. The second armature, associated 2, associated with another electronic inverter 6 plays the role of a conventional electric machine and makes it possible to add or subtract torque from the rotor linked to the movement output. The sliding power of the first armature 1 can be used on the second armature 2 with possible exchanges with storage of electrical energy. The role of each air gap E1, E2, ES can be naturally reversed with an air gap E1 and its additional external air gap ES while the air gap E2 would become internal. Compared to that of FIG. 2, the arrangement of FIG. 4 however has two particularities: on the one hand, the fixed parts 11 of the armatures which carry the coils 9 are united in adjacent spaces, which is of interest for the 'mechanical integration and pooling of cooling circuits; on the other hand, the armature 2 is of the same type as the armature 1: its windings are annular, coupled to the air gap E2 by systems of claws 10 fixed in this case; it has the same number of pancakes. As will be seen now, this arrangement can be modified according to the present invention to obtain additional benefits to those already mentioned. An example of arrangement of the electromagnetic coupler according to the invention is illustrated in FIG. 5. The armatures 1 and 2 are inspired by the arrangement with circular windings 9 already represented in FIG. 4, with immobilization of the windings of the armature 1 by additional air gap ES. Unlike the configuration in Figure 4, however, i! there is only one coil 9 per wafer instead of 2, This coil 9 is common to the two armatures; the magnetic yokes 11 which previously separated the windings 9 of FIG. 4 have disappeared; the fixed magnetic circuits of the two armatures are pooled; the main flow collected at the air gap E1 is thus composed with that coming from the air gap E2. It will be said that the magnetic circuit of the fixed parts of the armatures 1 and 2 is "through flux".

Where, in the arrangements described above, there were two inverters each supplying the polyphase windings specific to each armature, we now only use a single common inverter 3 which supplies the windings 9 with polyphase currents comprising two superposed components . This principle of supeφosition of currents will hereinafter be designated by the expression "control with compound currents". Orders of this type have already been described in other conditions in patents such as those of references US-6373160, US-6049152, EP-1089425. They will be presented more explicitly below in the context of the invention. As will be seen below, the choice of an arrangement with six wafers according to FIG. 5 corresponds to one of the possibilities of using the command with compound currents which makes it possible to get rid of parasitic torque ripples. The stator heights of FIG. 4 have been generally preserved in the representation of FIG. 5 to highlight the possible increase in section of the single winding 9 relative to each of the previous windings. In this arrangement, the storage of electrical energy is always optional. To facilitate understanding, FIG. 6 provides an exploded view indicative of a coil 9 common to the two armatures 1, 2 and of the double system of claws which is associated with it; the assembly is positioned opposite the movement output rotor. In this FIG. 6, the references 10, 10a designate the system of claws associated with the air gap E2, the references 12, 12a the system of rotating claws associated with the air gap E1, the references 13, 14 the surfaces facing the additional air gap ES and the references 15, 15a the external and internal parts of the movement output rotor which in this example is shown with a surface magnet arrangement. The references 16 and 17 designate the magnetic elements associated with the claws 12, 12a. Each of these groups of magnets is arranged on a ferromagnetic ring (respectively internal and external) which ensures the closure of the flux. This assumption will serve as a basis for the development of the presentation of your command with compound currents which follows, but, as has already been observed previously, many other embodiments are possible (magnets inserted, magnets buried, asynchronous, reluctance; in particular with transverse flow looping as indicated in French patent application 04 00830 of January 29, 2004. We will now describe the operation of a provision of this type supplied by compound currents.The number of claws of each air gap is equal to the number d '' magnets which are opposite; there are thus pi pairs of poles on - air gap E1, and p ₂ on the air gap E2, ^* The number n of wafers is chosen to be ur common multiple of n- and of n ₂ : n = k..nι {: n = k ₂ .π ₂ with ki and k ₂ integers.

At the air gap E1, the successive arrangement of the wafers comprises an angular phase shift of 2π / (n ₁ .p ₁ ); this phase shift can be obtained either by playing on the angular setting of the group of magnets associated with this wafer at the air gap E1, or at the level of the corresponding group of claws of the rotor of the power source. Thus, compared to the air gap E1, the system is electrically with neither phases. Likewise, at the air gap E2, the successive arrangement of the pancakes comprises an angular phase shift of 2π (n ₂ .p ₂ ); this phase shift can be obtained either by playing on the angular setting of the group of magnets associated with this wafer at the air gap E2, or at the level of the group of corresponding fixed claws. Thus, compared to the air gap E2, the system is electrically at n ₂ phases. and _! is the relative angular position of the rotor associated with the movement output compared to that of the rotor associated with the power source, whose angular position for the air gap El α ₂ is the angular position of the rotor associated with the movement output, therefore the angular position for the air gap E2. Thus, by noting Ωβ and Ω * the speeds of the motor input and the motion output:

Ω -Ω. --_ ₂ - ^ ι of Ω = dt ^~ dT

We will denote by ω - p _1. (Ω _{: s} -Ω _e ) and ₂ = p ₂ -Çl _s ^the electrical pulses associated respectively with the two air gaps. Θ _a ι; Θ _a2 and θ _b will be respectively the magnetic potentials of the magnets of the air gap El, of the air gap E2 and that of the coil (ie, its ampere turns). The inverter 3 is shown diagrammatically in FIG. 5a: it includes a number of arms corresponding to a common multiple of ni and n ₂ preferably the smallest common multiple. This number of arms 18 corresponds to the number of wafers, unless each polyphase system comprises several groups of identical phases: in this case, the windings of the same setting can be connected in parallel or in series. According to the known principle of switching switching of electronic components, and by using the angular information α-j from sensors placed for this purpose, the inverter can thus generate in each of the k. groups of pancakes with neither phases a system of polyphase pulsating currents ω- _. ; within a group, each current is successively phase shifted by 2π / n and the sum of the currents is zero. In the same way, the inverter can also generate in each of the ka groups of wafers with n ₂ phases a polyphase current system of pulsation ω ₂ ; within a group, each current is successively phase shifted by 2π / n ₂ and the sum of the currents is zero. The summation of the instructions makes it possible to obtain a superposition of the two polyphase systems, and a wafer i will be traversed by currents giving it a magnetic potential:

211 211 Θ ₆ ,. = Θ _M .sin (/vo.- + ç \ J) + ® _b2 .άn (p ₂ .a ₂ + φ ₂ i) «! "2 again, by replacing ni and n ₂ with their value as a function of n: Θ _fc - k ₂ .j) where Θ _b1 and Θb ₂ , <pι and φ _∑ are amplitudes and phasing adjustable by electronic control. We are now interested in the functioning of the magnetic circuit. Figure 7 gives an equivalent diagram of the magnetic circuit thus defined on a wafer. Like the illustrations used for electrical circuits, the circles represent the sources of magnetic potential and the rectangles represent the times. The circuit has only been partially represented: there are in fact as many sources Θ _a1 and Θa2 as there are magnets. In order not to unnecessarily complicate the presentation, we will consider this magnetic circuit as linear. The pemneances represented in gray form symbolize the escape routes; (leaks between prongs, leaks distributed over the winding). This diagram is globalized in Figure 8.

The magnetic coupling of the magnets with the claws is described by a set of permeances Λδ ₊ ι _or 2 and Λ _M "_ 2 variable with the position and which integrate the air permeance and the internal permeance of the magnet. We will admit that we can account for these variations at first approach by sinusoidal evolutions:

^Λ _im2 = = - cos (/? .Ûr _loa2 ) + - Λεuppi will be the permeance of an additional air gap. Under these conditions, the electromagnetic torque of a wafer at the air gap E1 is written: r - ^{l dK} ° ^ 1 ^l - --2a2 1 dk + bb Θ d _x da, 2 da + ^^ _. 2Θ _fll .2Θ _Λ2 + * -.2Θ _al .® _b + ^ A2Θ _α2 -Θ da _x d da _} and at the air gap E2: _c _ t _{ΛA [} 1 rfΛ _{a2n2 |} 1 d _{bb & 2} ^δ2 2 da 2 ^' d ₂ 2 da ₂ ^{' b}

with - u ,: mutual permeance between magnets ai and coil b, etc.

The terms affected by the coefficient ^Λ A correspond to the reluctant components. We will now evaluate each of the terms of these expressions of the couples. Preliminary remark: the magnets being "extinguished" (short-circuit), the groups of permeances gap-magnets identified by a banding on figure 8 respectively have an equivalent value of: Λ _{G / arΛ.} , ₊ _, - = Cp, .Λ _Λl + P _i A _δι _ in_series with p _l A _Si _ + p ₁ Λs _i ... _A _PιAs + P s _A. ^Λ * ^{bϋlL Jv} a Λ + _ι- - ^~ ~ ~

at • Λ - ^ '^ 2ma ^χ

""" ^l ^ Gr & s2 * _ - ~ r. ie a constant value.

Evaluation of the reluctant couples in ^dA 9 ^{lal dA "} ^ ^{2 dA} i _and ^d ^ - da ₂ da _x da _x d ₂ It follows from the preliminary remark that these reluctant couples are zero on each wafer:

Evaluation of reluctant couples in - ≈i ^ L _e - ≥ ---- i- da da ₂

The calculation of Λ _a ι _a ι leads to a formulation of the type: with: alaljnax 1 + - ^ Gnffesl + _l- thereforeΛ _alβlœflX <Λ _G „ _jr „, ₊ _

On a wafer, there is therefore a reluctant torque associated with the magnets in the air gap E1: 1 dh --lai (f21Θ Ot \ ² = ^ ,. Λ _alalmax .sin ( _l a _t ). Cos (, ûf ₁ ). (2β _Ω ) ^: 2. da _x let i_ ^rfΛ ------, (β _π ) ² = ^{P |, Λalal} ".sin (2p _l a _ï ). (2Θ _a ) ² 2. da 2 This couple on a wafer is drawing at twice the synchronous frequency of the air gap E1; it is proportional to the number of poles p _t ; the leaks tend to attenuate it. Its polyphase composition on all of the wafers gives, however, a zero result; (except for the particular case where n = 2 which is in fact similar to the single phase, with two windings in phase opposition). "Similarly, there is in the air gap E2 at each wafer a reluctant couple drawing from the associated magnets; tt is proportional to the number of poles p ₂ and the leaks tend to attenuate it again, the polyphase result is zero except for the case n = 2. Evaluation of the interaction couples between magnets of the two air _gaps (terms in ^d _a ^ _{2 ci} dA _ala2 - / " _! Da ₂

The calculation of Λ _a ι ₃₂ on the wafer i leads to: 2J-1 211 ^A _ala2 = ^Λ -. _{l 2} .n _* . - ^COS (V ^Ûf -. *, I) .COS (^ ₂ .α ₂ -λ ₂ -0 nn with, when the leak permeances can be neglected: _{Λ =}

The leakage permeances lead in practice to a reduction of this term.

Thus, the torque linked to the interaction between groups of magnets 1 and 2 is on the air gap E1 of the wafer i of: -M) either again: 211 + sin (p, .α- - » ₂ .α ₂ (^ -k ₂ ) i)) n The wafer in question is therefore subjected to the air gap E1 to a pulsating torque which has two components: one to the pulsation ω-ι + α- ₂ and the other to | e »ι-ω ₂ 1. However, with the exception of certain special cases such as that where the two air gaps have the same number of phases: nι = n ₂ . the polyphase resultants in © ι + ω ₂ and in | ωr-ω _∑ 1 are zero. This is the case in particular for the examples in the table in FIG. 9. Symmetrically, at the level of the air gap E2 of a wafer, there is a drawing torque with a component at ω-ι + <D ₂ and a other at | co -. 2. - Under the same number of phase conditions as above, the polyphase resultants also cancel out on this air gap E2: Evaluation of the interaction between magnets and coil (terms in - ^ - and ₇ ^a2b ; -— ^ and— -) da _x da ₂ da ₂ da _x 2T1 The calculation of Λ _ab leads to: an expression of the type: A. _alb - Λ _αlfcmax .cos {? ₁ .a: _] .k _v i) n 211 Similarly, the calculation of de _a2 b leads to: A. _a2b - Λ _Λ2ijnax .cos (? ₂ .û. ₂ k ₂ .i))

It is therefore the terms in - and - which translate the coupling of the coil with da _λ da ₂ your magnets; the terms in - —and - - do not produce a force da ₂ da _x To build a useful mean torque in the air gap E1, it is necessary to introduce a current component at the pulsation ω, synchronous with pi-α- .; conversely, the construction of a useful mean torque in the air gap E2 supposes a current component with the pulsation ω ₂ synchronous with p ₂ .α ₂ .

If we thus assume that, by an appropriate electronic control, we generate on each wafer i: 2TT 2IT ®b, ^≈ Θ -.-. Sin α. +. "V) + Θ ₆₂ .sin (/ ^? ₂ - <ar-- + Ψι -) then it appears in the air gap E1 of the wafer i a couple: * ι ^J ) + _b2 .sm (p ₂ .a ₂ + φ ₂ J. ₂ .-)) which can be put in the form: f ^ -S..2Θ _flt .Θ _” ≈ - P ^ J2.Θ _Λ . _bl {Θ <cosf _the -∞B & p ^ + φ _x --- "da 2 n ιτ 2π Θ + _62. (COSQ ?, .u., -p ₂ .a ₂ -φ _2. (£, -k ₂ ) j) - case (p _y .a _l + p ₂ .a + φ _2. (A, + k ₂ ) .i)) nn

Therefore, at the level of the air gap E1 on a wafer: the interaction coil-group 1 of magnets results in a continuous useful component, and three components drawing respectively at o. _1t ωι + ω _2l and | ω.-ω. ! . The result is zero for nι> 2. The other two, drawing components also have zero results except in special cases already mentioned; they are zero in particular for the examples of FIG. 9. Symmetrically, a similar result is obtained in the space-r E2. aπ -? -?.?. î5-? y J. ^e .Ç.

Finally, taking into account the result of the -couples under the cancellation conditions of the drawing components: on the air gap E1:

on the air gap E2: _{Csτ =} ΛA - * -. 2. @ _A2 θ _c2 .cosffl ₂

The increase in torques with the number of poles, within the limit of increasing parasitic effects linked to leaks, is a natural effect of structures with globalized armature: the multiplication of poles does not generate any stress on the winding section. In steady state, the torque of the first drive is adjusted to balance that of the power source by playing on Θ _M .cosφι. The torque on the output rotor is then adjusted by playing on the torque of the second air gap by Θ _b2 .cosφ ₂ . The arrangement according to the invention which has just been described with compound current control makes it possible to obtain the desired function of coupler. Its comparison to its analogue with separate coils goes beyond the scope of this presentation, but we can however note qualitatively the following points: - your magnets and your associated flow looping heads are crossed by pulsing flux components: to guard against development of eddy currents within them, it is desirable that these magnets are with high internal electrical resistivity or fractionated into elements of short length isolated from each other; the constitution of the cylinder heads must likewise be adapted to the variable flows (lamination, "iron powders" ...). This observation naturally applies to transpositions to configurations with different magnet arrangements, asynchronous or variable reluctance realizations. - as in the separate armature arrangements, the question of parasitic couplings between neighboring wafers neglected in the first approach above must be taken in consideration: as already observed for the arrangements with separate armatures, as an alternative to the spacing of the wafers, it may be preferable to make annular magnetic cuts in the median spaces between wafers at the level of the external and internal cylinder heads of the rotor Release. - an oversizing of the magnets is necessary: indeed, the factor ™ ^Ά .2.Θ of proportionality of the useful torque to θbj 2 corresponds to a magnetizing flux; we would find a coefficient of the same kind in the case of separate windings. Compared to the latter, with comparable geometric dimensions, this factor degrades due to the elongation of the magnetic path due to the flow-through structure, which implies an increase in current or dimensions. Precautions relating to the risk of demagnetization of magnets which operate regularly in opposition also go in this direction; the leakage permeances correspond to a dimensioning optimization parameter Naturally, the question of the demagnetization limit does not arise in asynchronous or reluctance realizations; the elongation of the magnetic path resulting from the series of air gaps affects only the magnetizing components provided by the coil.

- in return, significant reductions in Joule losses are possible; this is an important point for improving the efficiency of the coupler and pushing further the progress made in its thermal: in fact: with similar geometry, the magnetic potentials Θ _M and Θ _b2 required to produce the torques are substantially preserved. However, to accommodate the single winding, there is a section corresponding to the sum of the sections of the separate reference windings to which the space freed up by the removal of the cylinder heads is potentially added; we can thus roughly consider that your cross-section and the volume of copper of the single coil compared to one of the previous ones have been multiplied by k> 2. If the reference current density was j in each of the separate windings, the densities ji and j ₂ of the compound currents are now each of the order of j / k; except in particular cases where the pulsations e ^ and ω ₂ are linked, the Joule losses associated with ji and j ₂ are simply additive: h ² + J2 ² ); (P being the resistivity of the conductor and V _Ctl its overall volume); this means that the overall Joule losses are then divided by k> 2. - your losses can be reduced in the electronic components, from where another progress on the output and the cost associated with the dimensioning: Indeed: if one considers that the losses in the electronic components are largely related to the passage of the current to across a waste voltage (IGBT transistors, freewheeling diodes of the bridge arms) and that this fraction of the losses is roughly expressed in the form: il), it comes in the case of separate windings: Global losses≈ V _d * average (| h.sinω .tl) + V _d ^* average (| l ₂ .sin o ₂ tl); in the typical operations of the coupler in transmission without electrical supply, the power of machine 1 is similar to that of machine 2 and this under identical voltages except for parasitic drops; we will write: l-Ha≈l; from where:

Global losses≈ V _d * l * (mean (| sinω-itl) + mean (| sinω ₂ t |)). In the case of the command in compound currents, the same reasoning leads to: Losses globates = V _d *! * (Average (| sinωιt + sin <a ₂ tl)). Numerical estimates over a time horizon of a few periods show an advantage of the command in compound currents on these losses of the order of 35%, (except very special cases of the type e> ι = ω ₂ ).

To clarify what has been said about the possibility of embodiments according to the invention using active parts asynchronous to the rotor, FIG. 10 gives an example of adaptation of an asynchronous cage on the air gap E1. In this FIG. 10, the references 25, 26 designate short-circuit rings of the asynchronous cage, the reference 27 of the parallel conducting bars, the reference 28 of the annular surfaces of the ferromagnetic circuit included between the bars 27 and annular spaces 29 not magnetic. The reference 30 designates the magnetic yoke. It is assumed here that the required phase shift between successive wafers is obtained by an angular offset at the level of the consecutive claw systems. The conductive bars 27 disposed at regular intervals at the periphery of the rotor are thus substantially rectilinear and parallel to the longitudinal axis. Depending on the shape of the claws and the space between them, it may or may not be desirable to give these bars 27 an inclination relative to their reference direction, as is often done in the usual asynchronous machines for smoothing the phenomena drawing associated with the notching of the stator. The ends of the bars 27 are electrically connected to each other on each side of the rotor by a conductive ring 25, according to the usual principle of asynchronous cages. A first feature relating to the electrical insulation of the conductive bars 27 is however to be noted for this cage. The parasitic electrical paths between conductive bars must indeed be avoided: each of the segments of a bar which is in the air gap of a wafer is the seat of two electromotive force components associated respectively with the two systems of compound currents; the assembly operates on the summation of these fems on all of the pancakes; thus for example, the parasitic polyphase component intended for the other rotor leads to zero summation on all the segments of each bar. If intermediate currents in the looping through the end rings can develop, they will cause losses. For this reason, the bars must be isolated from each other here along their length. This insulation can be obtained naturally if the ferromagnetic material used is not a good electrical conductor (in the case of iron powders). For the same reason, ferromagnetic material cannot be solid if it is electrically conductive; iron powders or stacks of magnetic sheets will therefore be used for example. A second feature concerns your non-magnetic spaces 29 which are formed between the magnetic circuits associated with the different wafers: these spaces are visible in FIG. 10. As has already been seen, tts constitute an alternative to the spacing of the claw systems for limit magnetic couplings by leaks between wafers. Protrusions provided on the bars 27 can act as shims between the ferromagnetic elements thus fractionated. ^" Figure 11 shows an alternative embodiment of an asynchronous cage of an electromagnetic coupler according to the invention. It shows in section the external part of the rotor. In this figure, the reference 31 represents the path of a conductive bar with its six staircase segments. The reference 32 represents the cylinder head. The references 33 and 34 show the two short-circuit end rings. The reference 35 designates the non-magnetic spaces. We find in Figure 11, the general principle which comes to be described with bars 31 electrically insulated along their length and electrically connected at their ends by short-circuit rings 33, 34. There are also non-magnetic spaces 35 between wafers intended for decoupling. the conductive bars 31 appear to consist of a set of segments delimited by the borders between successive wafers; these seg each is essentially straight and parallel to the longitudinal axis, but between them a successive angular offset that can contribute partially or completely to ensuring the required phase shift between wafers at this gap. Electrical continuity between the segments of a bar 31 is ensured at the borders between wafers by connections which in principle take the form of arcs of a circle in the plane perpendicular to the longitudinal axis. These connections can be used as wedges in non-magnetic spaces. As already noted above for the intermediate rotor, the bar segments may have an inclination relative to their reference position, and the principle aliasing between the segments may be greatly reduced, or even masked. This embodiment in which the phase shift is carried out with the rotor makes it possible to freely choose the relative angular positioning between wafers of the claw systems, for example on criteria of minimisatioπ of the leakage permeances between wafers. Furthermore, in terms of phase shift, it is also possible to play on the order of the pancakes. The asynchronous cages can be produced by various methods: conductive bars 27, 31 made of copper can for example be attached and welded in situ to their end rings 25, 26; 33, 34. It is also possible from the start to make a complete cage, for example of cast aluminum, to which the elements of sectored magnetic circuits have been attached. It is also possible, in the case of using iron powders, to consider pressing the magnetic material on the cage. The mechanical strength of these assemblies can be obtained by bonding, overmolding, shrinking, etc. solutions. In summary, the invention applies to an electromagnetic coupler without sliding contact as described in French patent application 04 00830 of January 29, 2004, that is to say the windings of the rotary armature are made fixed in an arrangement "with centralized armature" and "additional air gaps" with distribution of the alternating flow in the air gap by systems of claws. According to the invention, the magnetic circuits of the two armatures are pooled, the parts of the stator becoming "through flow" and each wafer no longer comprising a single coil common to the two armatures. The control is carried out by a single inverter which supplies the windings with compound currents. The active parts of the rotor associated with the movement output are processed to accept the alternating flows of the non-synchronous component, avoiding the development of parasitic currents in the magnetic circuit itself (lamination, iron powders), in any magnets (materials with suitable electrical resistivity, fragmentation in isolated elements) or in intermediate current paths along the length of the asynchronous cage bars (electrical insulation of the bars along their length, use of a non-conductive ferromagnetic material). The above description also shows how to make non-magnetic spaces at the rotor to limit the undesirable coupling between wafers. Finally, we note the special adaptation of the asynchronous cages which makes it possible to partially or totally treat at the rotor level the angular shifts required for your phase shifts between slabs. Several embodiments already described in French patent application 04 00830 of January 29, 2004 can be used here, in particular in terms of hooping or variable reluctance arrangement with transverse flow looping, they have not been included in this description their transposition according to the present invention goes without saying.

Claims

1. Electromagnetic coupler for electrical power transmission comprising two electrical machines, with two rotors respectively connected to a first shaft (7) and to a second shaft (5), the armatures of these two machines being disposed between the two rotors, characterized in that the windings (9) of the two armatures are common to these two armatures, a first air gap (E1) being produced between a part of one of the rotors and a part of the other rotor, a second air gap (E2) being produced between the common windings (9) of the two armatures and another part of one of the rotors and an additional air gap (ES) being produced between the windings (9) and said part of the other rotor, so that the main magnetic flux collected at the first air gap (E1) is composed with the magnetic flux from the second air gap (E2) and in that the coils (9) are supplied by means of a single inverter (3) by electric currents polyphas és with superimposed components.

2. Electrical power transmission device comprising an electromagnetic coupler for electrical power transmission comprising two electrical machines, with two rotors respectively connected to a first shaft (7) and to a second shaft (5), the armatures of these two machines being disposed between the two rotors, characterized in that the windings (9) of the two armatures are common to these two armatures, a first air gap (E1) being produced between a part of one of the rotors and a part of the other rotor , a second air gap (E2) being produced between the common windings (9) of the two armatures and another part of one of the rotors and an additional air gap (ES) being produced between the windings (9) and said part of the 'other rotor, so that the main magnetic flux collected at the first air gap (E1) is composed with the magnetic flux from the second air gap (E2) and in that the windings (9) are fed to the into a single inverter (3) by polyphase electric currents comprising superimposed components.

3. Transmission device according to claim 2, characterized in that it comprises several annular coils (9) arranged one beside the other around a common axis, each of these coils (9) being supplied by the single inverter (3).

4. Transmission device according to claim 3, characterized in that each winding (9) is between two magnetic elements (13, 14) each comprising a series of claws (10, 10a) arranged annularly around the winding (9) and directed towards the claws (10, 10a) of the other element, the claws (10) of one of the elements being arranged between the claws (10a) of the other element, these claws (10, 10a) being arranged in look of magnets (15) arranged annularly on an outer part of the rotor to make the first air gap (E1), the number of magnets (15) being equal to that of the claws (10, 10a) and the polarity of these magnets ( 15) being successively reversed.

5. Transmission device according to claim 4, characterized in that each winding (9) annularly surrounds two magnetic elements (16, 17) each comprising a series of claws (12, 12a) arranged annularly around the axis of the winding (9) and directed towards the claws (12, 12a) of the other element, the claws (12) of one of the elements being disposed between the claws (12a) of the other element, these claws (12, 12a ) being arranged opposite magnets (15a) arranged on an inner part of the rotor to make the second air gap (E2), the number of magnets (15a) being equal to that of the claws (12, 12a) and the polarity of these magnets (15a) being successively reversed, each winding (9) defining with said elements (16, 17) with claws (12, 12a), a wafer.

6. Transmission device according to claims 4 and 5, characterized in that the wafers are angularly phase shifted with respect to the following by 2π / n .p ₁ with respect to the first air gap (E1) and by 2ττ / n ₂ .p ₂ with respect to the second air gap (E2), the number n of windings (9) being a common multiple of ni and n _2l nor being the number of phases corresponding to the air gap (E1) and n ₂ being the number phases corresponding to the air gap (E2), pi being the number of pairs of poles on the first air gap (p1) and p ₂ the number of pairs of poles on the second air gap (E2).

7. Transmission device according to your claim 6, characterized in that the single inverter (3) is adapted to generate in each wafer a first polyphase pulse current ω-i phase-shifted by 2rr / nι from one wafer to the other and a second polyphase pulse current ω ₂ phase shifted by 2ττ / n ₂ from one wafer to another. Electromagnetic coupler for electrical power transmission and transmission device comprising such a coupler.