Multi-layer flat cable with layer permutations for making an electrical coil

Abstract

The invention relates to a multilayer flat cable (20) comprising several layers (20-1, 21-2, 21-3) of a conductive material arranged one above the other to form a stack of layers. The multilayer flat cable (20) has at least one permutation region in which the order of the layers in the stack changes. For at least one of the layers (21-3), the thickness of said layer (21-3) in the permutation region is greater than the thickness of said layer (21-3) outside the permutation region. The multilayer flat cable (20) can, in particular, be used to form the winding of a coil to make a transformer or an inductor in a high-frequency electrical converter. Figure for the abstract: Fig. 4

Description

Title of the invention: Multi-layer flat cable with permuted layers for the construction of an electrical coil. Field of the invention

[0001] The present invention relates to the field of coils used to make a transformer or an inductor in an electrical converter. In particular, the invention relates to a multi-layered flat cable intended to be wound to form a coil. The multi-layered flat cable is formed by stacking several superimposed layers of a conductive material. State of the art

[0002] High-frequency electrical converters are used in many industrial applications: isolated DC / DC converters (battery chargers for automobiles or aircraft), DC / DC or DC / AC converters with several arms in parallel through coupled inductors, power converters where a filtering inductor conducts a current with a high harmonic content at high frequency, etc. (the acronym DC stands for "Direct Current" in English, "courant continu" in French; the acronym AC stands for "Alternating Current" in French).

[0003] In this application, the term "high frequency" is used for frequencies above one or more tens of kilohertz. These frequencies can go up to one or more hundreds of kilohertz, or even one or more megahertz.

[0004] The coils used to make a transformer or an inductor for a high-frequency electrical converter are made so as to minimize Joule effect losses. At high frequencies, these losses are mainly due to eddy currents generated by the proximity effect and the skin effect.

[0005] The skin effect refers to the phenomenon whereby an electrical conductor (for example, a copper wire) experiences an increase in resistance when a high-frequency current is passed through it. The skin effect is due to the magnetic field generated inside the wire by the applied current. This magnetic field creates induced currents (eddy currents) that increasingly shift the current towards the outer surface of the wire's cross-section as the frequency increases. This leads to an increase in the wire's resistance at high frequencies and power dissipation through Joule heating. To mitigate these problems, it is known to replace the simple electrical wire with a Litz wire comprising several insulated twisted strands, each with a diameter less than the skin thickness at the considered working frequency.

[0006] The proximity effect refers to the phenomenon whereby a high-frequency current flowing in an electric wire generates a magnetic field which, in turn, generates eddy currents on nearby wires. This induces Joule losses on the nearby wires. Since the strands of a Litz wire are twisted (i.e., they are braided or stranded together), their position within the wire varies, which makes it possible to counteract the proximity effect with respect to another wire (each strand alternates between different positions, and this allows the current to be distributed evenly among all the strands).

[0007] However, the litz wire has several disadvantages: it is relatively expensive, it has a relatively high low frequency resistance (DC resistance or direct current resistance), and it has a poor space factor (relatively low winding window fill rate).

[0008] Another approach involves using a conductor formed by stacking several superimposed layers of a conductive material. This is referred to as a multilayer flat cable. In the present application, the term "flat" is used to define an object that is significantly wider than it is thick (for example, the width is at least five times greater than the thickness, or even at least ten times greater). Here again, to limit the skin effect and the proximity effect, it is necessary to ensure that each layer is associated with substantially the same amount of magnetic flux along the entire length of the winding. To this end, it is necessary to change the position of the layers in the stack (i.e., to swap the order of the layers in the stack) at various points along the cable (a point where the order of the layers is changed is called a "swap region").

[0009] To achieve such a stack of conductive material layers, it is known to use printed circuit board (PCB) technology with several copper layers and vias that connect the traces of two layers without making contact with the traces of other layers. However, prototyping and producing such a solution is quite expensive.

[0010] Another known solution, presented in particular in the document "Multi-Layer Barrel-Wound Foil Winding Design" by Nigam et al., consists of producing a multi-layered flat cable by superimposing several copper strips (each strip corresponding to a layer). At a region where the layers are swapped, a notch is cut in each layer, halfway across its width, and the layers are then joined to swap their positions. The notches can be cut manually, by photochemical etching, by die cutting, or by laser. The individual layers can also be insulated with adhesive tape, a powder coating, or a varnish applied by spraying or immersion. The exact location of the permutation regions takes into account the fact that the magnetic flux changes with each layer and also with each turn of the coil winding.

[0011] Figures 1 to 3 schematically represent a multi-layered flat cable 10 according to the prior art. The cable 10 is formed by a stack of two copper layers: a first layer 11-1 and a second layer 11-2.

[0012] As illustrated in [Fig. 1], each of the layers 11-1 and 11-2 takes the form of a ribbon whose thickness is very small compared to its width. [Fig. 1] schematically represents the change in position of the first layer 11-1 relative to the second layer 11-2 during a permutation region 12b. In a region 12a preceding the permutation region 12b, the first layer 11-1 is located below the second layer 11-2. During the permutation region 12b, the first layer 11-1 rises and the second layer 11-2 falls such that, in the region 12c following the permutation region 12b, the first layer 11-1 is located above the second layer 11-2.

[0013] Figure 2 schematically represents the effective cross-section of each of the layers 11-1 and 11-2 respectively in the region 12a preceding the permutation region 12b (part a) of the figure), in the permutation region 12b (part b) of the figure), and in the region 12c following the permutation region 12b (part c) of the figure). The "effective cross-section" of a layer is defined as the area occupied by the layer in a plane of cross-section to the layer. At the permutation region 12b, for each of the layers 11-1 and 11-2, the notch made in the layer results in a reduction of at least 50% of the layer's effective cross-section.

[0014] Fig. 3 schematically represents a profile view of the cable 10 shown in Fig. 1.

[0015] Figures 2 and 3 in particular show that the thickness e of each of the layers 11-1 and 11-2 remains constant in the region 12a which precedes the permutation region 12b, in the permutation region 12b and in the region 12c which follows the permutation region 12b. Description of the invention

[0016] In the prior art solution described with reference to Figures 1 to 3, there are high losses in the permutation region because on the one hand the effective section of each conductor is reduced, and on the other hand the interface corresponding to the surface of the two layers directly opposite each other is particularly limited in the permutation region.

[0017] The present invention aims to remedy all or part of the drawbacks of the prior art, in particular those set out above.

[0018] To this end, and according to a first aspect, the present invention proposes a multi-layered flat cable comprising at least N layers of a conductive material, N being an integer at least equal to two. The flat cable has at least one region of layer permutation.

[0019] In a region preceding the permutation region, each layer forms a ribbon whose thickness is at least five times less than its width. The layers are arranged one above the other to form a stack of layers. Each layer occupies a rank r in the stack, r being an integer between one and N. The layer of rank one corresponds to the first layer of the stack, and a layer of rank r, for any value of r between two and N, is adjacent to the layer of rank r-1 in the stack.

[0020] During the permutation region, the order of the layers changes in such a way that the layer which occupies rank N in the region preceding the permutation region occupies rank one in a region following the permutation region, and the layer which occupies rank r in the region preceding the permutation region, for any value of r between one and Nl, occupies the layer of rank r+1 in the region following the permutation region.

[0021] For at least one of the layers, a thickness of said layer in the permutation region is greater than the thickness of said layer in the region preceding the permutation region.

[0022] The term "thickness of a layer" means the distance separating the top face and the bottom face of the layer in a direction perpendicular to at least one of the top and bottom faces (the top face of a layer is oriented towards the top of the stack, and the bottom face is oriented towards the bottom of the stack).

[0023] This increase in the layer thickness in the switching region helps to limit Joule effect losses at both low and high frequencies. The increased effective cross-section of the layer in the switching region reduces the DC resistance (Dr) of the layer. Furthermore, increasing the interface area between the layer and each of the other layers opposite it reduces the AC resistance (Ar) of the layer in the switching region.

[0024] In particular embodiments, the invention may further comprise one or more of the following features, taken individually or in all technically possible combinations.

[0025] In particular embodiments, the number N of layers is greater than or equal to three and for each of the first-rank layers, N1 and N, the thickness of said layer in the permutation region is greater than the thickness of said layer in the region that precedes the permutation region.

[0026] In particular embodiments, for at least one of the layers, a thickness of said layer in the permutation region undergoes an increase of at least 50% compared to the thickness of said layer in the region preceding the permutation region.

[0027] In particular embodiments, the number N of layers is equal to two, and an interface which separates the first layer and the second layer in a cross-section plane of the flat cable performs in the permutation region a rotation of 180 degrees around an axis orthogonal to said cross-section plane.

[0028] In particular embodiments, the number N of layers is equal to two, and a cross-sectional plane of the flat cable in the permutation region has a number of interfaces separating the first-rank layer and the second-rank layer at least equal to two.

[0029] In particular embodiments, the number N of layers is greater than or equal to three, and a cross-sectional plane of the flat cable in the permutation region has a number of interfaces separating the layer of rank N and the layer of rank Nl of at least four.

[0030] In particular embodiments, on at least a part of the flat cable which does not include a permutation region, the thickness of a layer of rank r, for any value of r between two and N, is greater than the thickness of the layer of rank r-1.

[0031] In particular embodiments, the N layers of conductive material are obtained by additive manufacturing.

[0032] In particular embodiments, the conductive material is copper or aluminum.

[0033] According to a second aspect, the present invention relates to a coil comprising a winding of a flat cable according to any one of the embodiments previously described.

[0034] In particular embodiments, the winding of the flat cable forms several turns around a core of ferromagnetic material and, for a part of a turn which does not include a permutation region, a layer of rank r, for any value of r between two and N, is closer to the core of ferromagnetic material than the layer of rank r-1, and the thickness of the layer of rank r is greater than the thickness of the layer of rank r-1.

[0035] According to a third aspect, the present invention relates to an electrical converter comprising a coil according to any one of the embodiments previously described. Presentation of the figures

[0036] The invention will be better understood upon reading the following description, given by way of non-limiting example, and made with reference to Figures 1 to 20, which represent:

[0037] [Fig-1] a schematic representation of a multi-layer flat cable according to art prior,

[0038] [Fig.2] a schematic representation of the effective section of each of the layers of the cable of [Fig.1],

[0039] [Fig.3] a schematic representation of a profile view of the cable of [Fig.1],

[0040] [Fig.4] a schematic representation of a first embodiment of a cable multi-layer flat bar according to the invention,

[0041] [Fig.5] a schematic representation of the effective section of each of the layers of the cable of [Fig.4],

[0042] [Fig.6] a schematic representation of a profile view of the cable of [Fig.4],

[0043] [Fig.7] a schematic representation of a second embodiment of a cable multi-layer flat bar according to the invention,

[0044] [Fig.8] a schematic representation of a profile view of the cable of [Fig.7],

[0045] [Fig.9] a schematic representation of a third embodiment of a cable multi-layer flat bar according to the invention,

[0046] [Fig. 10] a schematic representation of a generalization of the three embodiments illustrated in Figures 4 to 9,

[0047] [Fig. 11] a schematic representation of another example for the embodiments illustrated in Figures 5 and 7 (without splitting a layer in the permutation region),

[0048] [Fig. 12] a schematic representation of another example for the embodiment illustrated in [Fig. 9] (without division of a layer in the permutation region),

[0049] [Fig. 13] a schematic representation of another example for the embodiments illustrated in Figures 5 and 7 (division of two of the layers into three parts in the permutation region),

[0050] [Fig. 14] a schematic representation of a fourth embodiment of a multi-layer flat cable according to the invention,

[0051] [Fig. 15] a schematic representation of the effective cross-section of each of the layers of the cable of [Fig. 14],

[0052] [Fig. 16] a schematic representation of a profile view of the cable of [Fig. 14],

[0053] [Fig. 17] another representation of the fourth embodiment illustrated in the [Fig.14]

[0054] [Fig. 18] a schematic representation of an example of the embodiment of a coil comprising a winding of a flat cable obtained according to the embodiment illustrated in Figures 14 to 17,

[0055] [Fig. 19] a schematic representation of the effective cross-section of each layer of a multi-layer flat cable according to the invention with a variation in the thickness of each layer,

[0056] [Fig.20] a schematic representation of a fifth embodiment of a multi-layer flat cable according to the invention,

[0057] [Fig.21] a schematic representation of a sixth embodiment of a cable multi-layer flat bar according to the invention,

[0058] [Fig.22] a schematic representation of the sixth embodiment illustrated in [Fig.21] in the case where each permutation of two layers is carried out as illustrated in [Fig. 14],

[0059] [Fig.23] a schematic representation of an example of an embodiment of a coil comprising a winding of a flat cable obtained according to the embodiment illustrated in [Fig.22].

[0060] In these figures, identical reference numerals from one figure to another designate identical or analogous elements. For clarity, the elements shown are not necessarily to the same scale, unless otherwise stated.

[0061] It should be noted that Figures 1 to 3 have already been described above in the presentation of the prior art.

[0062] Detailed description of an embodiment of the invention

[0063] As explained above, the invention relates in particular to a multi-layered flat cable used, for example, to connect and / or wind a transformer coil or an inductor coupled in a high-frequency converter. The multi-layered flat cable is designed to minimize Joule effect losses (high-frequency losses due to eddy currents generated by the proximity effect and the skin effect, as well as low-frequency losses).

[0064] Figures 4 to 6 schematically represent a first example of an embodiment of a multi-layer flat cable 20 according to the invention. The cable 20 is formed by a stack of three layers 21-1, 21-2 and 21-3 of an electrically conductive material (for example copper or aluminum) arranged one above the other.

[0065] Part a) of [Fig. 4] represents the stacking of the three layers 21-1, 21-2, and 21-3 before a layer permutation occurs. In the example considered and illustrated in Figures 4 to 6, in the region 22a preceding the permutation region 22b, the first layer 21-1 is located at the bottom of the stack, the second layer 21-2 is located in the middle of the stack, and the third layer 21-3 is located at the top of the stack. Part b) of [Fig. 4] illustrates the layer permutation at the level of the permutation region 22b. In permutation region 22b, the first layer 21-1 and the second layer 21-2 move upwards while the third layer 21-3 moves downwards such that, in region 22c following permutation region 22b, the first layer 21-1 is located in the middle of the stack, the second layer 21-2 is located at the top of the stack, and the third layer 21-3 is located at the bottom of the stack. The position of layers 21-1, 21-2, and 21-3 in region 22c following permutation region 22b is shown in part c) of [Fig. 4].

[0066] Fig. 5 schematically represents the effective section of each of the layers 21-1, 21-2 and 21-3 respectively in the region 22a which precedes the permutation region 22b (part a) of the figure), in the permutation region 22b (part b) of the figure), and in the region 22c which follows the permutation region 22b (part c) of the figure).

[0067] Fig. 6 schematically represents a profile view of cable 20 shown in Fig. 4.

[0068] As illustrated in Figures 4 and 5, in the permutation region 22b, the first layer 21-1 and the second 21-2 each split into two parts located on either side of the third layer 21-3. However, it would also be conceivable that none of the layers 21-1, 21-2 and 21-3 split in the permutation region, as is the case for example in the example illustrated in [Fig. 1 1].

[0069] As illustrated in Figures 5 and 6, for the cable 20 according to the invention (and contrary to the prior art described above with reference to Figures 1 to 3) the thickness of at least one of the layers increases in the permutation region 22b. Indeed, in the example considered, at least at one point in the permutation region 22b, the thickness e3b of the third layer 21-3 is significantly greater than the thickness e of said layer 21-3 in the region 22a which precedes the permutation region 22b.

[0070] Increasing the thickness of layer 21-3 in the permutation region 22b offers several advantages over the prior art. Firstly, it increases the effective cross-section of layer 21-3 in the permutation region 22b. This reduces the DC resistance (Dr) of layer 21-3 in the permutation region 22b, thus limiting Joule losses at low frequencies. It also limits Joule losses at high frequencies because, in the permutation region 22b, the AC resistance (Ar) of layer 21-3 is lower when its thickness is greater (unlike what is observed outside the permutation region).This reduction in the alternating current resistance (RAc) in the permutation region 22b is the result of the increase in the interface area between layer 21-3 and each of layers 21-1 and 21-2 (increase in the area where layers 21-1 and 21-2 are directly opposite layer 21-3 in the permutation region 22b). Such arrangements lead to a reduction in high-frequency losses.

[0071] Figures 7 and 8 schematically represent a second embodiment of a multi-layer flat cable 20 according to the invention. The cable 20 of the second embodiment is similar to that of the first embodiment described with reference to Figures 4 to 6, except that the thickness of each of the layers 21-1, 21-2 and 21-3 increases in the permutation region 22b (whereas only the thickness of layer 21-3 increases in the permutation region 22b in the first embodiment described with reference to Figures 4 to 6).

[0072] Indeed, it can be observed in part b) of [Fig. 5] that, in the permutation region 22b, the raising of the first layer 21-1 leaves a free space below the first layer 21-1. Also, the change in shape of the third layer 21-3 leaves a free space above the second layer. As illustrated in part b) of [Fig. 7], for the second embodiment, in the permutation region 22b, the thickness eib of the first layer and the thickness e2b of the second layer 21-2 are increased to fill these free spaces.

[0073] Fig. 8 schematically represents a profile view of cable 20 shown in Fig. 7.

[0074] In the second embodiment illustrated in Figures 7 and 8, each of the layers 21-1, 21-2, and 21-3 has a thickness e in the region 22a preceding the permutation region 22b (and also in the region 22c following the permutation region 22b). Conversely, at least at one location in the permutation region 22b, the layers 21-1, 21-2, and 21-3 have thicknesses eib, e2b, and e3b, respectively, greater than e (eib > e; e2b > e; e3b > e). Increasing the thickness of the layers 21-1 and 21-2 in the permutation region 22b further reduces Joule heating losses.

[0075] Figure 9 schematically represents a third example of an embodiment of a multi-layer flat cable 20 according to the invention, for which the cable 20 comprises two layers 21-1 and 21-2. The thickness of each of the layers 21-1 and 21-2 increases in the permutation region 22b.

[0076] It should be noted that it would also be possible, in the permutation region 22b, to divide one of the layers 21-1 and 21-2 in two. For example, and as illustrated in [Fig. 12], the first layer 21-1 can be divided in two and placed on either side of the second layer 21-2 in the permutation region.

[0077] Figure 10 schematically represents a generalization of the three embodiments described with reference to Figures 4 to 9.

[0078] Figure 10 schematically represents a flat cable 20 comprising at least N layers 21-1 to 21-N of a conductive material (for example, copper, aluminum, or other). The number N is an integer at least equal to two. The number N represents the number of layers involved in a permutation (therefore nothing would prevent us from considering a cable 20 with an integer number P of layers greater than N (P > N), see in particular below the description made with reference to [Fig.20]).

[0079] The cable 20 has at least one layer permutation region. Part a) of [Fig. 10] shows the effective cross-sections of layers 21-1 to 21-N in a region preceding the permutation region. Part b) of [Fig. 10] shows the effective cross-sections of layers 21-1 to 21-N at a location within the permutation region. Part c) of [Fig. 10] shows the effective cross-sections of layers 21-1 to 21-N in a region following the permutation region. For each layer 21-1 to 21-N, there is continuity of material in the region preceding the permutation region, in the permutation region, and in the region following the permutation region.

[0080] As illustrated in part a) of [Fig. 10], in the region preceding the permutation region, each layer forms a ribbon whose thickness is at least five times less than its width, or even at least ten times less. Thus, and as illustrated in part a) of [Fig. 10], a cross-section of a layer 21-1 to 21-N, for example, takes the form of a rectangle whose width is at least five times less than its width.

[0081] In the region preceding the permutation region, layers 21-1 to 21-N are arranged one above the other to form a stack of layers. Each layer occupies a rank r in the stack, r being an integer between one and N. The layer of rank one corresponds to the first 21-1 layer of the stack (it is located at the bottom of the stack in the example shown in [Fig. 10]). The layer of rank N corresponds to the last 21-N layer of the stack (it is located at the top of the stack in the example shown in [Fig. 10]). A layer of rank r, for any value of r between two and N, is adjacent to the layer of rank r-1 in the stack (in the example shown in [Fig. 10], this means that the layer of rank r is located above the layer of rank r-1).

[0082] In the permutation region, the order of the layers changes such that the layer which occupies rank N before the permutation region occupies rank one after the permutation region, and the layer which occupies rank r before the permutation region, for any value of r between one and Nl, occupies the layer of rank r+1 after the permutation region (it is considered that a layer of rank r before the permutation region keeps this rank throughout the permutation region and becomes the layer of rank r+1 after the permutation region).

[0083] For at least one of the layers 21-1 to 21-N, the thickness of said layer at at least one location in the permutation region is greater than the thickness of said layer before the permutation region. The thickness of a layer corresponds to the thickness of the effective cross-section of the layer, that is, the thickness of the surface occupied by the layer in a cross-sectional plane to the layer. A cross-sectional plane to a layer is a plane orthogonal to at least one of the top and bottom faces of the layer (the top face of a layer is oriented towards the top of the stack, and the bottom face is oriented towards the bottom of the stack).

[0084] It should be noted, as can be seen in [Fig. 10] (and also in Figures 4 to 9 described above), that in a cable 20 according to the invention, the total surface area occupied by the cable in a cross-section plane of the cable does not increase in the permutation region.

[0085] The different layers 21-1 to 21-N are separated from each other by an insulator. This can be a simple layer of air, a powdered insulating material applied by spraying, or a varnish-like insulating material applied by immersion. The thickness of the interface between two adjacent layers depends on the manufacturing method (based on the type of insulator used and the electrical potential between the opposing layers). In some cases, the thickness of the interface between two adjacent layers can be significantly less than the thickness of each of the two layers. In other cases, the thickness of the interface between two adjacent layers can be of the same order of magnitude as, or greater than, the thickness of each of the two layers. The interface between two layers before the permutation region maintains essentially the same thickness in the permutation region and after the permutation region (if the two layers are still adjacent after the permutation region).This ensures that a minimum distance is maintained between two adjacent layers along the entire length of the cable.

[0086] As illustrated in [Fig. 10] (and as explained above with reference to Figures 7 and 8), when the number N of layers is greater than or equal to three, it is advantageous for the thickness of each of the first-rank layers, N1 and N, to increase in the permutation region. In other words, for each of the first-rank layers, N1 and N, the thickness of said layer at least at one point in the permutation region is greater than the thickness of said layer before the permutation region.

[0087] In the example illustrated in [Fig. 10], the thickness of the layer of rank one increases progressively during the permutation region until it doubles compared to its thickness before the permutation region (increase in thickness of at least 100%); the thickness of the layer of rank Nl, on the other hand, doubles from the beginning of the permutation region before decreasing progressively during the permutation region; the thickness of the layer of rank N, on the other hand, has a thickness almost throughout the permutation region equal to more than N times its thickness before the permutation region (increase in thickness of at least Nxl00%).

[0088] There are an infinite number of possible geometries for producing a multi-layered flat cable according to the invention, that is to say, a cable in which the thickness of at least one of the layers increases in the permutation region. However, it is preferable to ensure that the thickness of said layer in the permutation region increases by at least 50% compared to the thickness of said layer before the permutation region.

[0089] The N layers of electrically conductive material can be obtained, in particular, by additive manufacturing (3D printing), for example using powder bed fusion technologies (LBM for "Laser Beam Melting" or EBM for "Electron Beam Melting") or sintering technologies (MBJ for "Metal Binder Jetting"). Additive manufacturing makes it possible to design parts that were previously impossible to produce. Indeed, while some examples of embodiments described in this application (in particular those described with reference to Figures 9 and 11) could be produced by conventional (non-additive) manufacturing technologies, the majority of them have complex geometric shapes that are better produced using additive manufacturing. For fusion processes, support techniques allow the parts to be held in place during additive manufacturing.This support can be removed mechanically or by a pressure injection operation of the insulating material.

[0090] Figures 9 and 12, on the one hand, and Figures 7, 11, and 13, on the other hand, illustrate different possible arrangements of the layers in the permutation region for cases where N is equal to two (Figures 9 and 12) or three (Figures 7, 11, and 13). In each of these figures, part a) of the figure schematically represents the effective cross-section of the layers in the region preceding the permutation region; part b) of the figure represents the effective cross-section of the layers in the permutation region; and part c) of the figure represents the effective cross-section of the layers in the region following the permutation region.

[0091] In the example illustrated in part b) of [Fig. 9], in the permutation region, a single interface 23 separates the first layer 21-1 from the second layer 21-2. In the example illustrated in part b) of [Fig. 12], in the permutation region, there are two interfaces 23 separating the first layer 21-1 from the second layer 21-2 (indeed, in this example, the first layer 21-1 splits in two on either side of the second layer 21-2 in the permutation region). It would also be conceivable to have more than two interfaces 23 separating the first layer 21-1 from the second layer 21-2. For example, if in the permutation region the first layer 21-1 split into three and the second layer 21-2 split into two, then there would be four interfaces 23 separating the first layer 21-1 from the second layer 21-2.

[0092] Similarly, and as illustrated in Figures 7, 11, and 13, when the number N of layers is greater than or equal to three, the number of interfaces 23 separating the layer of rank N and the layer of rank N1 varies according to the number D of parts into which the layer of rank N is divided in the permutation region. The number of interfaces 23 separating the layer of rank N and a lower-rank layer is then equal to 2xD. In the example illustrated in [Fig. 13], for which N is equal to three, the layer 21-3 of rank N is divided into two parts, and there are therefore four interfaces 23 separating the layer 21-3 of rank N and the layer 21-2 of rank N1 (there are also four interfaces separating the layer 21-3 of rank N and the layer 21-1 of rank one).

[0093] It can be advantageous to increase the number of interfaces 23 separating the Nth-rank layer and a lower-rank layer in the permutation region 22b. This effectively increases the surface area where these layers are directly adjacent to each other in the permutation region, leading to a reduction in high-frequency losses. However, dividing a layer into several parts can lead to an increase in low-frequency losses because the effective cross-section of each part is smaller, and this can lead to an increase in the DC resistance (Dr) for the layer in question. Therefore, a compromise must be made. The best geometry to use depends on the application (operating frequencies, layer dimensions, interface thickness between two adjacent layers, etc.).To determine the most optimal cable geometry, it is possible to use a computer-implemented analysis and simulation method called the "finite element method" (FEM).

[0094] Figures 14 to 17 illustrate a preferred embodiment of a cable 20 according to the invention with the number N of layers involved in a permutation equal to two.

[0095] Part a) of [Fig. 14] represents the stacking of a first layer 21-1 and a second layer 21-2 before a permutation of the layers takes place. In the example considered and illustrated in Figures 15 to 16, in the region 22a preceding the permutation region 22b, the first layer 21-1 is located below and the second layer 21-2 is located above. Part b) of [Fig. 14] illustrates the permutation of the layers at the permutation region 22b. In the permutation region 22b, the positions of the first layer 21-1 and the second layer 21-2 are exchanged such that, in the region 22c following the permutation region 22b, the first layer 21-1 is located above and the second layer 21-2 is located below. The positions of the layers 21-1 and 21-2 in the region 22c following the permutation region 22b are shown in part c) of [Fig. 14].

[0096] Figure 15 schematically represents the effective section of each of the layers 21-1 and 21-2 respectively in the region 22a which precedes the permutation region 22b (part a) of the figure), at a location in the permutation region 22b (part b) of the figure), and in the region 22c which follows the permutation region 22b (part c) of the figure).

[0097] Fig. 16 schematically represents a profile view of cable 20 shown in Fig. 14.

[0098] Parts bl), b2), b3), and b4) of [Fig. 17] schematically represent the cross-section of each of the layers 21-1 and 21-2 in a cross-sectional plane of the cable 20 at four different locations in the permutation region 22b. Parts a) and c) of [Fig. 17] correspond respectively to parts a) and c) of [Fig. 14]. As illustrated in [Fig. 17], in the permutation region, the interface 23 which separates the first layer 21-1 and the second layer 21-2 in a cross-sectional plane of the cable undergoes a continuous rotation of 180° around an axis orthogonal to this plane (an axis passing through the center of the cable cross-section in the cross-sectional plane).

[0099] Thus, during the permutation region, the effective cross-section of the first layer 21-1 takes on different forms. The effective cross-section of the first layer 21-1 initially has the shape of a rectangle (as illustrated in part a) of [Fig. 17]). This rectangle then transforms into a right trapezoid whose longer base is located on the left lateral face of the cable 20. The length of the longer base of this trapezoid gradually increases while the length of its shorter base decreases until it forms a right triangle (as illustrated in part b1) of [Fig. 17]). The effective cross-section of the first layer 21-1 then takes the shape of a right trapezoid whose longer base is located on the underside of the cable 20 (as illustrated in part b2) of [Fig. 17]). The length of the larger base of this trapezoid gradually decreases as the length of its smaller base increases, until it forms a rectangle.The effective cross-section of the first layer 21-1 then takes the form of a right trapezoid whose longer base is located on the upper face of the cable 20 (as illustrated in part b3 of [Fig. 17]). The length of the longer base of this trapezoid gradually increases as the length of its shorter base decreases, until it forms a right triangle (as illustrated in part b4 of [Fig. 17]). The effective cross-section of the first layer 21-1 then takes the form of a right trapezoid whose longer base is located on the left lateral face of the cable 20. The length of the longer base of this trapezoid gradually decreases as the length of its shorter base increases, until it forms a rectangle (as illustrated in part c of [Fig. 17]).During the permutation region, the effective cross-section of the second layer 21-2 is the symmetrical reflection of the effective cross-section of the first layer 21-1 with respect to the center of the effective cross-section of cable 20.

[0100] Figures 15 and 16 highlight the fact that the thickness of each of the layers 21-1 and 21-2 increases in the permutation region 22b. Indeed, in the example Considering that, at least at one location in the permutation region 22b, the thickness e2b of the second layer 21-2 is more than twice the thickness e of said layer 21-2 in the region 22a preceding the permutation region 22b. Increasing the thickness of layers 21-1 and 21-2 in the permutation region 22b reduces Joule effect losses (at both low and high frequencies).

[0101] For the preferred embodiment described in Figures 14 to 17, the interface surface between two opposing layers is optimized in the permutation region. This optimizes the reduction of Joule losses at high frequencies. Furthermore, the change in shape of the layer's cross-section is relatively smooth (the transition from cross-section in part a) of [Fig. 17] to cross-section in part b2) of [Fig. 17] is smoother than the transition from cross-section in part a) of [Fig. 9] to cross-section in part b) of [Fig. 9]). This optimizes the reduction of Joule losses at low frequencies.

[0102] As explained previously, swapping the different layers at different locations along the cable 20 helps to limit the skin effect and the proximity effect. However, for this to be effective, the different layers must be subjected to the same field averaged over their length. When the cable 20 is used as a linear support for an electrical connection, the swap regions are, for example, positioned at multiples of the ratio between the length of the cable 20 and the number of swapped layers (see, for example, Section III of the document "Strategies to Reduce Copper Losses in Connections of Medium Frequency High-Current", B. Cougo et al.).When cable 20 is used to wind a coil, the positions of the permutation regions along cable 20 can be defined, in particular, according to the number of layers permuted, the length of a turn in the winding, the number of turns in the winding, etc. The optimal positions of the permutation regions can be calculated using analytical methods, for example, a method based on Dowell's equations or a FEM-type method.

[0103] Figure 18 schematically represents an example of an embodiment of a coil 30 comprising a winding of a flat cable 20 obtained according to the embodiment previously described with reference to Figures 14 to 17. In the example considered and illustrated in Figure 18, the first layer 21-1 and the second layer 21-2 exchange their position twice for each turn of the winding. In other words, there are two permutation regions 22b at each turn of the winding.

[0104] The coil 30 shown in [Fig.18] can in particular be used to make a transformer or a coupled inductance intended to be integrated into a high-frequency electrical converter.

[0105] In the embodiments described above, it has been considered (for the sake of (simplification) that outside a permutation region all layers 21-1 to 21-N had the same thickness. However, it may be advantageous to define a different optimal thickness for each of the layers 21-1 to 21-N. For example, if, for at least a portion of the cable outside a permutation region, some layers experience a greater magnetic flux than others, then it is advantageous for the layers experiencing a greater magnetic flux to have a thinner profile than the layers experiencing a lesser magnetic flux.This can be the case, for example, for a 20-strand cable wound in turns around a ferromagnetic core to form a coil of a high-frequency transformer or a coupled inductor: the layers furthest from the core, i.e., the layers located towards the center of the winding window, experience a greater magnetic flux than the layers closest to the core. Thus, for a portion of a turn that does not include a permutation region, if a layer of rank r, for any value of r between 2 and N, is closer to the ferromagnetic core than the layer of rank r-1, then it is advantageous for the thickness of the layer of rank r to be greater than the thickness of the layer of rank r-1.

[0106] This is schematically illustrated by [Fig. 19], which represents the effective cross-section of N layers 21-1 to 21-N of a flat cable 20 according to the invention in a region preceding a permutation region. As illustrated in [Fig. 19], the thickness eai of the first layer 21-1 is less than the thickness e2a of the second layer 21-2 (eia < e2a); the thickness e2a of the second layer 21-2 is less than the thickness e3a of the third layer 21-3 (e2a < e3a); and so on. In general, for a given rank r in the layer stacking, with r between 2 and N, the thickness of the layer at rank r is less than the thickness of the layer at rank r-1. In this example, for the part of the cable considered, it is assumed that the lower the rank of a layer in the stack, the more this layer experiences a significant magnetic flux.

[0107] In the embodiments described above, all the layers of the cable 20 change position within a permutation region. However, it would also be possible to consider the case where only a subset of the layers forming the cable 20 change position within a permutation region. In other words, the number N represents the number of layers involved in a permutation, and one can consider the case of a cable 20 having an integer number P of layers with P greater than N (P > N). From one permutation region to another, different layers can be permuted.

[0108] This is schematically illustrated by [Fig. 20], which schematically represents a particular embodiment of a flat 20 cable according to the invention. The flat 20 cable comprises four layers 21-1 to 21-4 (P = 4), but a permutation does not imply each time two layers (N = 2). In other words, the different layers are swapped two by two.

[0109] In the example illustrated in [Fig.20], there is a successive alternation between a permutation region for which there are two permutations (a first permutation between the layer of rank one and the layer of rank two, and a second permutation between the layer of rank three and the layer of rank four), and a permutation region for which there is only one permutation between the layer of rank two and the layer of rank three.

[0110] Thus, in the example illustrated in [Fig. 20], in a first permutation region 22b-1, there is, on the one hand, a first permutation between the first layer (this is the first layer 21-1) and the second layer (this is the second layer 21-2), and on the other hand, a second permutation between the third layer (this is the third layer 21-3) and the fourth layer (this is the fourth layer 21-4). In a second permutation region 22b-2, there is a permutation between the second layer (this is the first layer 21-1) and the third layer (this is the fourth layer 21-4).In a third permutation region 22b-3, there is a first permutation between the first layer (the second layer 21-2) and the second layer (the fourth layer 21-4), and a second permutation between the third layer (the first layer 21-1) and the fourth layer (the third layer 21-3). In a fourth permutation region 22b-4, there is a permutation between the second layer (the second layer 21-2) and the third layer (the third layer 21-3). And so on until the initial order of the layers is restored after an eighth permutation region 22b-8.

[0111] Other permutation schemes could also be considered. For example, and as illustrated in [Fig.21], for a cable comprising four layers (P = 4) and layer permutations two by two (N = 2), one could successively repeat a sequence of three permutation regions 22b-1, 22b-2 and 22b-3 in which the first layer and the second layer exchange their positions in a first permutation region 22b-1, the second layer and the third layer exchange their positions in a second permutation region 22b-2, the third layer and the fourth layer exchange their positions in a third permutation region 22b-3.

[0112] In these different examples, each permutation of two layers can be carried out according to any one of the embodiments described above, in particular those described with reference to figures 9, 12 or 14.

[0113] By way of example, [Fig. 22] illustrates the embodiment described in [Fig. 21] in the case where each permutation of two layers is carried out as described in reference to [Fig.14]. Parts a) to d) of [Fig.21] illustrate the permutation of the first layer 21-1 (which thus goes from rank one to rank two) with the second layer 21-2 (which thus goes from rank two to rank one) in the first permutation region 22b-1; parts d) to g) of [Fig.21] illustrate the permutation of the first layer 21-1 (which thus goes from rank two to rank three) with the third layer 21-3 (which thus goes from rank three to rank two) in the second permutation region 22b-2; parts g) to j) of [Fig.21] illustrate the permutation of the first layer 21-1 (which thus goes from rank three to rank four) with the fourth layer 21-4 (which thus goes from rank four to rank three) in the third permutation region 22b-3. As illustrated in parts a) and j) of the [Fig.

[22] , after the sequence of the three permutation regions 22b-1, 22b-2 and 22b-3, the first layer 21-1 has moved from rank one to rank four, the second layer 21-2 has moved from rank two to rank one, the third layer 21-3 has moved from rank three to rank two and the fourth layer 21-2 has moved from rank four to rank three. This sequence of permutations can then be repeated further along cable 20.

[0114] Fig. 23 schematically represents an example of an embodiment of a coil comprising a winding of a flat cable obtained according to the embodiment previously described with reference to Figures 21 and 22. In the example considered and illustrated in Fig. 23, a sequence of three permutation regions is repeated four times at each turn of the winding.

[0115] The inventors performed simulations to compare the Joule losses observed respectively for a prior art coil A using PCB technology, a prior art coil B formed by winding a flat cable comprising two copper layers (without permutation of the layers), a coil C according to the invention formed by winding a flat cable according to the embodiment described with reference to Figures 14 to 17 (two layers with permutation), a coil D according to the invention formed by winding a flat cable according to the embodiment described with reference to Figures 7 and 8 (three layers with permutation), a coil E according to the invention formed by winding a flat cable according to the embodiment described with reference to Figures 7 and 8 (three layers with permutation), with further variation in the thickness of each layer. The results are listed in the table below:

[0116] [Tables 1] Copper coil losses (per unit) A 1.7 B 1.0 C 0.65 D 0.53 E 0.47

[0117] For these simulations, the operating frequency is 200 kHz. For coils B to E, the flat cable has a width of 10 mm. For coils B to D, each copper layer has a thickness of 0.54 mm and two opposing layers are insulated from each other by a 0.1 mm thick air layer.

[0118] Coil C exhibits a reduction in Joule effect losses of 35% compared to coil B and of nearly 62% compared to coil A. Coil D exhibits a reduction in Joule effect losses of 47% compared to coil B and of nearly 69% compared to coil A. Coil E exhibits a reduction in Joule effect losses of 53% compared to coil B and of approximately 72% compared to coil A.

[0119] The foregoing description clearly illustrates that, through its various features and their advantages, the present invention achieves the stated objectives. In particular, the permutation of the layers with an increase in the effective cross-section of at least one of the layers in the permutation region makes it possible to reduce Joule heating losses. The invention also proposes optimizing the surface areas of two layers directly opposite each other in the permutation region, and optimizing the thickness of each layer in the stack according to its position relative to the magnetic flux, which further contributes to reducing Joule heating losses.

[0120] It should be noted that the implementation and realization methods considered above have been described as non-limiting examples, and that other variants are therefore conceivable.

Claims

Demands

1. Flat cable (20) having at least N layers (21-1 to 21-N) of a conductive material, N being an integer at least equal to two, said flat cable (20) having at least one permutation region (22b) of the layers; in a region (22a) preceding the permutation region (22b), each layer forms a ribbon whose thickness is at least five times less than its width, the layers (21-1 to 21-N) are arranged one above the other to form a stack of layers, each layer occupies a rank r in the stack, r being an integer between one and N, the layer of rank one corresponds to a first layer of the stack and a layer of rank r, for any value of r between two and N, is adjacent to the layer of rank r-1 in the stack;during the permutation region (22b), the order of the layers changes such that the layer which occupies rank N in the region (22a) preceding the permutation region (22b) occupies rank one in a region (22c) following the permutation region (22b), and the layer which occupies rank r in the region (22a) preceding the permutation region (22b), for any value of r between one and N1, occupies the layer of rank r+1 in the region (22c) following the permutation region (22b); said flat cable (20) being characterized in that, for at least one of the layers (21-1 to 21-N), a thickness of said layer in the permutation region (22b) is greater than the thickness of said layer in the region (22a) preceding the permutation region (22b).

2. Flat cable (20) according to claim 1 wherein the number N of layers (21-1 to 21-N) is greater than or equal to three and for each of the rank one layers, N1 and N, the thickness of said layer in the permutation region (22b) is greater than the thickness of said layer in the region (22a) which precedes the permutation region (22b).

3. Flat cable (20) according to any one of claims 1 to 2 wherein, for at least one of the layers (21-1 to 21-N), a thickness of said layer in the permutation region (22b) undergoes an increase of at least 50% compared to the thickness of said layer in the region (22a) preceding the permutation region (22b).

4. Flat cable (20) according to claim 1 wherein the number N of layers (21-1, 21-2) is equal to two, and an interface (23) which separates the rank one layer and the rank two layer in a cross-section plane of the flat cable (20) performs in the permutation region (22b) a rotation of 180 degrees around an axis orthogonal to said cross-section plane.

5. Flat cable (20) according to claim 1 wherein the number N of layers (21-1, 21-2) is equal to two, and wherein a cross-sectional plane of the flat cable (20) in the permutation region (22b) has a number of interfaces (23) separating the first-rank layer and the second-rank layer at least equal to two.

6. Flat cable (20) according to any one of claims 1 to 3 wherein the number N of layers (21-1 to 21-N) is greater than or equal to three, and wherein a cross-sectional plane of the flat cable (20) in the permutation region (22b) has a number of interfaces (23) separating the layer of rank N and the layer of rank Nl of at least four.

7. Flat cable (20) according to any one of claims 1 to 6 wherein, on at least a part of the flat cable (20) which does not include a permutation region, the thickness of a layer of rank r, for any value of r between two and N, is greater than the thickness of the layer of rank r-1.

8. Flat cable (20) according to any one of claims 1 to 7 wherein the N layers (21-1 to 21-N) of conductive material are obtained by additive manufacturing.

9. Flat cable (20) according to any one of claims 1 to 8 wherein the conductive material is copper or aluminium.

10. Coil (30) comprising a winding of a flat cable (20) according to any one of claims 1 to 9.

11. Coil (30) according to claim 10 wherein the winding of the flat cable (20) forms several turns around a core of ferromagnetic material and, for a part of a turn which does not include a permutation region (22b), a layer of rank r, for any value of r between two and N, is closer to the core of ferromagnetic material than the layer of rank r-1, and the thickness of the layer of rank r is greater than the thickness of the layer of rank r-1.

12. Electrical converter comprising a coil (30) according to any one of claims 10 to 11.

Images