Forced magnetic short loop variable reluctance motor (VRM) and castelated VRM (CVRM) with negligible torque ripple

Abstract

This invention is a new type of Variable Reluctance Motor (VRM) where torque ripple and noise become negligible. The method for eliminating torque ripple is to combine 4 or 6 phased motor design with a new coil configuration which ensures that the magnetic field always move in a pattern that repeat as the different phases in the motor are energized, while also ensuring that magnetic forces are balanced out so net force that pull the shaft out of center always is zero. The design can be a Castellated VRM to produce a CVRM with the more torque and the same properties as the new VRM.

Description

[0001] Forced magnetic short loop Variable Reluctance Motor (VRM) and Castelated VRM (CVRM) with negligible torque ripple

[0002] 1 Technical field of invention

[0003] The presented invention is a variant of the variable reluctance motor (VRM) and castellated variable reluctance motor (CVRM). The reluctance motor is a class of electric motor where rotor is made from ferromagnetic materials and support materials only. If rotor comprise of coils, squirrel-cage or magnets, in addition to the ferromagnetic material, the motor will belong to other motor classes like permanent magnet motors, induction motors, slip ring motors etc. It is desirable to make rotor in reluctance motors of laminated steel, since any current in rotor will reduce motor efficiency.

[0004] Rotor in reluctance motors has teeth. Stator comprise of “electromagnets” that pull on the teeth. The electromagnets are turned on and off in sequence to create pull on rotor in the same directions at all positions. The explanation is a bit simplified, but it is useful to understand the principle. Unlike PM motors the electromagnets in a reluctance motor cannot push on rotor. They can only pull. This means that a reluctance motor need at least 3 phases for stator to be able to produce torque on rotor in all positions.

[0005] N omenclat ure

[0006] 0 Electrical angle

[0007] 0rnMechanical angle

[0008] I0 Peak current

[0009] 1 A Peak current density

[0010] U Voltage

[0011] T Torque

[0012] 2 Prior art

[0013] This is a presentation of prior art with focus on known variable reluctance motor (VRM) configurations. 2.1 Introduction to the reluctance motor

[0014] Reluctance motor uses the principle of magnetic reluctance to get the rotor moving. Rotor therefore consist of ferromagnetic materials and support mate- rials only. If rotor includes of coils, permanent magnets (PM) etc. the motor belongs in another electric motor class. Wiguna et. al. [1] and Sufei et. al. [2] have written recent review articles.

[0015] The simplest linear reluctance motor is an electromagnet, which can com- prise of a coil wound around an U-shaped magnetic core. This type electromag- net comes in many variants and is used for anything from lifting scrap metal to holding doors shot.

[0016] Rotating reluctance motors can be divided into synchronous reluctance mo- tors (SynRM) and variable reluctance motors (VRM). VRM are also called switch reluctance motor (SRM). In addition, there are some special reluctance motors like the segmented reluctance motor reviewed by Zhenyao et. al. [3] and investigated by Xiaodong et. al. [4] that fall between SynRM and VRM.

[0017] SynRM where discovered because rotor of some asynchronous motors would go synchronous at low load. Hubert [5] write: ”At some critical speed, the low- reluctance paths provided by salient poles causes squirrel-cage induction motors to snap into synchronism with the rotating flux of the stator. When this occurs, the slip is zero, induction motor action ceases, and the rotor is pulled around by simple magnetic attraction called reluctance torque.” In recent years the original squirrel-cage has been optimized into SynRM without rotor coils. In SynRM rotor follows the rotating magnetic field in stator. The magnetic field in rotor is therefore more or less constant. Most SynRM have distributed winding similar to most standard asynchronous motors, but research on SynRMs with concentrated windings are going on. Resent studies of SynRM are done by Jae-Kwang et. al. [6], Donaghy-Spargo et. al. [7] and Villet and Kamper [8].

[0018] In the VRM the coils can be considered electromagnets that are turned on in sequence and pull on different parts of rotor to create rotation. This action causes the magnetic field in rotor to change shape, direction and magnitude. The VRM is therefore more dependent on materials with low hysteresis losses than the SynRM.

[0019] The simplest commercially available rotating VRM that produce torque in all positions is shown in figure 1 and 2. This motor has 3 phases and is a 6-4 configuration meaning it has 6 stator teeth and 4 rotor teeth. You can multiply number of theeth with any positive integer to get other 3 phased configurations like 12-8, 18-12 etc. Quite a lot of research is going on to reduce torque ripple of 3 phased VRM by curve shaping of the current and other means. Yuan- feng et. al. [9] compare a 12-8 VRM with a 6 phased segmented reluctance motor. Torque ripple of Yuanfeng et. al. [9] 3 phased VRM is 97% which is no surprise, but he also have 47% ripple for the 6 phased segmented VRM, which indicate poor design. Yusuf et. al.

[0010] try to reduce noise of a 6-4 VRM with a leaf spring. Zhenyao et. al.

[0011] analyze a 12-8 VRM where rotor is suspended by coils. Ruizhen et. al.

[0012] has made something similar. Khodadadeh et. al.

[0013] investigate a bit special 6-10 radial flux 3 phased VRM with internal stator, which they claim has a torque ripple 63%. This is good for a 3 phased VRM but still quite a lot if you compare with other motors. Radmanesh and Fara- hani

[0014] has made a special 3 phased PM assisted 24-26 VRM with external rotor and segmented stator to avoid leek field between teeth in different phases. Vinh and Ha

[0015] has simulate a 3 phased 6-8 VRM with internal stator, where they try to reduce torque ripple. This is an example which show it is possi- ble to deviate from the standard 6-4 configurations. Wiguna et. al. [1] work on noise reduction for a standard 18-12 3 phased VRM. Chun et. al.

[0016] work on sensorless control for a 12-8 3 phased VRM. Geldhof et. al.

[0017] work on sensorless control for a 6-4 3 phased VRM. Soares et. al.

[0018] do a simulation on a 6-4 VRM. Han-Chieh et. al.

[0019] do thermal analysis on a 18-12 3 phased VRM. Tong et. al.

[0020] try to reduce torque ripple on a 3-phase 12-8 VRM. An et. al.

[0021] do vector torque control of a 6-4 VRM. CN101982931A

[0022] de- scribe a mix of a 6-2 and 6-4 VRM where two of the teeth in rotor are much thinner than the other. The motor will probably run, but it is not clear if torque ripple will be reduced as the inventors claim. US2018102722A1

[0023] describe a control device for a 6-4 and 18-12 three phased VRM for cars.

[0020] Some focus on VRM with more phases. Padmanaban et. al.

[0024] analyze 3 configurations of 8-6 axial flux VRM with two stators and one rotor. Torque ripple is 28.82% for the best motor and 37.10% for the worst, which is good for VRMs. Padmanaban et. al. can probably reduce the ripple further for all 3 configurations if they continue optimizing their design. The challenge with this motor is that stator and rotor comprise of several parts, some of them embedded in epoxy resin laminate. This makes it difficult to mount the parts together in a stiff enough structure with small enough misalignment to avoid ripple due to varying air gaps and mechanical deformation of the motor. Patil and Saxena

[0025] have investigated a 8-6 radial flux VRM. There are patents on 8-6 radial flux VRM. CN209767257U

[0026] show a 8-6 radial flux VRM where teeth in rotor are modified, presumable for the purpose of reducing torque ripple. The VRM described in CN209767257U

[0026] will suffer from long-loop / short-loop problems described later in this patent. Patent CN209767257U

[0026] also describe a way to couple the coils in a kind of star configuration where number of transistors and diodes in the drive are reduced to four, utilizing the fact that there is current in max two coils simultaneously. With regard to numbers of components, this drive solution is better than the on drive solution described here. This patent application will show that VRMs with 4 or 6 phases are the way to go.

[0021] Figure 3 show simplest possible 4 phased VRM configuration which is 8-6. Figure 4 show simplest possible 6 phased VRM configuration which is 12-10. A VRM with 10-8 configuration would be 5 phased. Again you can multiply the teeth configuration with any positive integer without changing number of phases.

[0022] Hubert [5] give in his book from 1991 a brief description of 4 phased VRM with 6 rotor teeth and 8 stator teeth (8-6 configuration). It is interesting that Hubert choose the 8-6 and not the 6-4 configuration for his example, since it in- dicates that it was common knowledge that the 4 phased 8-6 VRM configuration work better than the 3 phased 6-4 VRM configuration.

[0023] Other relevant prior art to this patent are patents NO332245B1

[0027] , N020150240A1

[0028] and N020210788A1

[0029] . The patents are treating a 3 phased radial flux CVRM and a 4-phased axial flux CVRM.

[0024] 2.2 Torque ripple in VRM with 3 to 6 phases

[0025] The variable reluctance motor (VRM) has a bad, maybe undeserved, reputa- tion for being noisy. Some explanation of why a 3-phased VRM of any type will have serious torque ripple in patent N020210788A1

[0029] . Here is a more comprehensive explanation.

[0026] A well-designed electric motor must run smooth without too much noise. If you take a well-designed 2 phased PM-motor, put constant current on phase A and twist the shaft with a torque wrench, you will measure a torque close to sinusoidal as function of angle. When you run the same motor, you supply a sinusoidal current to phase A. This current is also a function of rotor angle. This is two sinuses multiplied with each other, so it is reasonable to assume that the torque created by phase A is proportional to sin2θ while the motor run. It is also reasonable that the torque created by phase B is proportional to cos20. If we add torque from phase A and B, we get the total torque. Trigonometric relations say that sin2θ + cos2θ = 1 for all 0, meaning that the torque of a well-designed 2 phased PM-motor is close to constant for any electrical angle if the phases are supplied with sinusoidal current. Figure 5 show a plot of torque in both phases and the sum.

[0027] If we do the same exercise for a well-designed 3 phased PM-motor, we see that torque from phase A+B+C is proportional to sin2θ + sin2(θ + π / 3) + sin2(θ + 2π / 3) = 3 / 2 meaning that a well-designed 3 phased PM motor also has close to constant torque if supplied with 3 phased sinusoidal current. Figure 6 show a plot of torque in all 3 phases and the sum. You can do the torque wrench experiment with synchronous motors too, where the rotor coils are magnetized with current through slip rings and get the same result. This is why these motors can be designed to be quiet.

[0028] In a 3 phased VRM (e.g. figure 1) the rotor consist of only ferromagnetic materials and support materials. The coils are therefore only able to attract the iron, not repel it. This means that in figure 1 the current in coil A is max, while there is no current in phase B and C. Assuming sinusoidal current, the expression for the current in the coils is:

[0029] Phase A: l0(if(sin(θ + 0π / 3) > 0, sin(θ + 0?r / 3) , 0) Phase B: l0(if(sin(θ + 2π / 3) > 0, sin(θ + 2π / 3), 0) Phase C: l0(if(sin(θ + 4π / 3) > 0, sin(θ + 4π / 3), 0)

[0030] In figure 2 with given phase configuration and given equations for current, elec- trical position is 0 = 5π / 3. Relation between electrical and mechanical angle are θm= 40. Again, we assume torque a function of sin squared and get the torque curve shown in figure 7.

[0031] In figure 7, torque ripple is 100%, and since torque ripple translate to noise, the 3-phased VRM is noisy. Theoretical it is possible to shape the current, so the torque ripple is reduced. Practically this is difficult to archive because:

[0032] • You need good control of rotor position for the current curve shaping. This becomes more difficult as speed increases, due to the time needed to read sensors and do calculations.

[0033] • Compared with a sinusoidal current, you need considerable higher voltage to produce a shaped current with a H-bride. It is possible to design the 3 phased VRM to reduce the torque ripple by shaping the surface of the tooth so thickness of the air gap vary, but since the torque ripple in figure 7 is an underlying characteristics of the 3-phased VRM it is difficult to reduce the torque ripple a lot.

[0034] For the 4 phased VRM, the current in the phases are described by the following equations:

[0035] Phase A: l0(if(sin(θ + Oπ / 2) > 0, sin(θ + Oπ / 2), 0)

[0036] Phase B: l0(if(sin(θ + 1π / 2) > 0, sin(θ + 1π / 2) , 0) Phase C: l0(if(sin(θ + 2π / 2) > 0, sin(θ + 2π / 2), 0) Phase D: l0(if(sin(θ + 3π / 2) > 0, sin(θ + 3π / 2) , 0)

[0037] Again, we assume torque a function of sin squared and get the curve in figure 8. This shows that a 4 phased VRM can have zero torque ripple!

[0038] Figure 9 shows torque in 5 phased VRM. The ripple is a lot smaller than for 3 phased VRM, but torque ripple is still a fundamental property. After showing that 4 phased VRM can have zero torque ripple, 5 phases VRM is likely to stay a curiosity, because it will require a variable frequency drive (VFD) with 5 half H- bridges, and the only advantage the 5 phased VRM has over the 4 phased VRM is lack of ” long-loop” / ’’short-loop” symmetry flaw mentioned later in relation to figure 50 and 51.

[0039] Equations for 6 phased VRM are shown below.

[0040] Phase A: l0(if(sin(θ + 0π / 3) > 0, sin(θ + 0π / 3), 0) Phase B: l0(if(sin(θ + 1π / 3) > 0, sin(θ + 1π / 3), 0) Phase C: l0(if(sin(θ + 2π / 3) > 0, sin(θ + 2π / 3), 0) Phase D: l0(if(sin(θ + 3π / 3) > 0, sin(θ + 3π / 3) , 0) Phase E: l0(if(sin(θ + 4π / 3) > 0, sin(θ + 4π / 3), 0) Phase F: l0(if(sin(θ + 5π / 3) > 0, sin(θ + 5π / 3), 0)

[0041] VRM with 6 phases also have zero theoretical torque ripple. Figure 10 show torque in 6 phased VRM. The 6 phased VRM seems to be a curiosity, but there is a new way to attache a 4 and 6 phased VRM to a 2 or 3 phased VFD or a 3 phased grid, which makes these motors interesting. VRMs with with 7 or more phases will be curiosities because they need VFDs with 7 or more half H-bridges or 4 or more full H-bridges.

[0042] 2.3 Known configurations of 3 phased VRMs

[0043] When designing a rotating VRM 1-phased and 2-phased designs are impractical because they can’t produce torque in desired direction at all angular positions

[0044] This means that rotor must be attached to a flywheel if it shall be possible to run 1 and 2-phased VRM as a rotating motor. 3-phased VRM designs is therefore the first practical rotating designs.

[0045] A lot of different VRM designs will be presented, so we start with a detailed description of a standard 3 phased VRM with 6 rotor coils and rotor with 4 teeth, (6-4 VRM configuration or just 6-4 for short) to introduce present jargon.

[0046] Figures 11 to 15 show a 3 phase 6-4 VRM. Figure 11 and 12 is the entire motor. Figure 13 shows the coils. Figure 14 shows the stator to the left and rotor to the right. Figure 15 shows one of the end plates. The two end plates 7,10 are almost equal, so only endplate 7 is shown in figure 15. 1 is stator iron, 2 is rotor iron, 3 is the shaft, 4 is coils, 5 are bolt holes for holding the end plates together, 6 are holes for mounting the motor onto whatever it is going to power, 7 is bottom end plate, 8 is coil insulation, 9 is bearings, 10 is top end plate, 11 is encoder, 12 is encoder shaft (part of shaft 3), 13 is output shaft (part of shaft 3) for connecting the motor to whatever it is going to power, 14 is the air gap which for VRM should be small, 15 is gap against end plate 10, 16 is rotor teeth (part of rotor iron 2), 17 is stator teeth (part of stator iron 1) and 18 is bolt for fastening the encoder. 19 is the edge of the encoder plate which give steering against the stator iron 1 and make the small air gap 14 possible. The edge 19 is a common solution used in many electric motors. 20 is the hole in end plate 10 for output shaft 13, 21 is space for bearing in the end plate 10 and 22 is space for the coil in end plate 10. In figure 11 the phases are marked A, B and C. Two and two coils belong to the same phase. In the motor in figure 11 to 15 the coils are concentrated meaning they are wound around a single stator tooth. Symbol 23 inside coils 5 indicates that current comes out of the plane. Symbol 24 inside coils 5 indicates that current go into the plane. 25 is stator back iron which is the part of the stator lamination 1 that connects the teeth 17. 26 is rotor back iron, which is the part of rotor lamination 2 that connects the teeth 16. When designing electric motors it is important that the back iron both in rotor and stator are thick enough so they are not completely saturated with magnetic field.

[0047] When discussing configurations, it is not necessary to show the motor in such detail as figure 11 to 15. It is enough to show the cross section in figure 16. All the parts and details in figure 11 to 15 are necessary to make a complete motor out of the motor presented in figure 16 and all later motor figures, but mostly all other parts than stator iron 1 or rotor iron 2 are omitted for clarity. Coils 4 are sometimes included to show electric current direction but electric current direction can also be deducted from the arrows 27,31 and the right hand rule.

[0048] In figure 16 only stator iron 1 and rotor iron 2 are shown. In addition, there are 3 arrows 27 which indicate the direction of the magnetic field through the rotor. The letters A, B and C mark the phases the coils wound around the rotor teeth 17 belong to. Here it is important to be aware that it is (usually) no point in changing the direction of the field in a VRM, meaning the current always goes the same way in the coils. The reason is that both a south pole and a north pole attract the rotor iron in the VRM the same way. Because the current in the coils always flow in the same direction the VFD used for powering a 3 phased VRM can be simplified to 3 half H-bridges instead of 3 full H-bridges. This saves components. Figure 16 means that when current is on in phase A the magnetic field will flow as indicated by arrow 25 in figure 17, When current is on in phase B, the magnetic field will flow as indicated in figure 18 and when current is on in phase C the magnetic field will flow as indicated as shown in figure 20. Figure 19 is the same as figure 18 but generated in ’’Comsol”. In figure 19 grey shows magnetic field strength. The lines in figure 19 are magnetic field lines. The arrows in figure 19 show field direction and strength.

[0049] The small arrow 26 in figures 17, 18 and 20 indicate rotation direction. Turn- ing on the phases in order A-B-C-A. . . will make the rotor rotate in clockwise direction. Turning on the phases in order A-C-B-A. . . will make the rotor ro- tate in counterclockwise direction. Please note that phase sequence and rotating direction vary with configuration.

[0050] When the rotor is in a position as shown in figure 21 both coil A and C are energized. The magnetic field will then flow as shown in figure 16 if the coils are connected to the VFD correctly. Here there is a VRM snag that can be difficult to discover. If phase A is connected the wrong way you get the connection shown in figure 22. Then, when both coils A and C are energized, the magnetic field will flow as shown in figure 23. This is longer loops than shown in figure 21. You notice the field in figure 23 goes behind the B coil from the C to A coil. This is the ’’short loop” and the ’long loop”. The ’Tong loop” will give more iron reluctance then the ’’short loop”. The motor still runs, but you get worse performance since you get more not useful reluctance in the stator back iron 25. You also get extra torque ripples because the field will take the ’’long loop” between phase A and B and phase A and C, while it takes the ’’short loop” (similar to the loop shown in figure 21) between phase B and C. With thick back iron and low back iron saturation, the effect is hardly noticeable. With high saturation and thin back iron (the way you have it in an optimized motor) the effect can be quite significant. In figure 24 torque in a 3 phased 6-4 VRM, a 4 phased 8-6 VRM and a 6 phased 12-10 VRM is simulated. You notice that the torque 3 phased motor repeat periodically, while the torque 4 and 6 VRM is random. This is because the field take the ’’long loop” in one phase in the 4 and 6 phased VRM.

[0051] 3-phased VRM configurations have n . 6 coils in the stator where n is a positive integer. Rotor must have a pair of teeth. The number of teeth must not be dividable by 6. A 6-2 VRM is shown in figure 25. Here you get torque in all positions. However, in the rotor position shown in the figure, coil A is turned off and only coil B is energized. It is ’’obvious” that the two rotor teeth are so far from the electromagnet in phase B (or C) that the torque in this position will be very small, meaning this is a poor design with high torque ripples. The 6-4 design shown in figure 16 is better, but also here the rotor teeth in phase B are not within proximity of the rotor teeth when current in phase A is turned off. In the 6-8 design shown in figure 26 there is proximity. The question is if figure 26 is actually the 12-8 design shown in figure 28 with 6 missing stator teeth. There is a flaw in the 12-8 VRM in figure 28 because it is impossible to draw alternating arrows marking phases all the way around the circle, because number of arrows is pair (6 arrows). You end up with at least two adjacent arrows 27 pointing in the same direction at least one place in the circle. This means the magnetic field takes the ’Tong loop” shown in figure 21 at least once with worse performance additional torque ripple as a result. It is possible to avoid the problem by using 3 full H-bridges and provide the coils with current in alternating direction, but then you need more components in the VFD. We include the 6-10 VRM in figure 27 to complete the 3-phased VRM’s with 6 coils. The 6-14 and 6-16 are possible motors, but for these motors the rotor teeth interact with the stator teeth only every second cycle, which is inefficient. In the 18-12 VRM shown in figure 29 the number of arrows is odd (9 arrows), and you avoid the ’’long loop” problem.

[0052] When adding teeth, you change the speed of the VRM. Table 1 shows speed for different configurations at 50Hz.

[0053] 2.4 Known type of configuration for 4 phased VRM

[0054] When it come to 4 phased designs the 4 phased 8-6 VRM shown in figure 36 seems to be the only variant which is currently investigated by e.g Padmana-

[0055] Table 1: VRM speed as function of configruation ban et. al.

[0024] and Patil and Saxena

[0025] . If we go through possible configu- rations, the 8-2 VRM is possible, but obviously is a poor design. 8-4 does not work because 4 phased VRM with 8 coils can’t have a rotor where the number of teeth is dividable by 4. The 8-6 VRM has the normal configuration as estab- lished in figure 16 where the magnetic field go straight through rotor. So, do the 8-10, 8-14, 8-18 and 8-22 VRM. All the 4 phased VRMs suffer from the ’’long loop” problem because it will always be a pair number of magnet field arrows 27 as shown in figure 36 and figure 41. You avoid the ’’long loop” problem in 4 phased 12-9 configuration, but then you end up with two coils pointing out and one coils pointing in (or opposite) which is not permissible because all the field going out through two stator teeth have to return through one stator tooth. This also create imbalance.

[0056] ” Comsol” simulations where rotor in figure 36 are in different positions, are shown in figure 37, 38 and 39 In figure 37 phase A is energized. In figure 38 phase C and D are energized and we get the ’’short loop” configuration. In figure 39 phase D and A are energized and we get the ’’long loop” configuration. In figure 37, 38 and 39 grey shows magnetic field strength, lines are magnetic field lines and arrows show field direction and magnitude.

[0057] When we move to 16 teeth in stator and normal configuration as established in figure 16, 16-2, 16-4 and 16-6 VRM configuration are good but are obviously poor designs because many stator teeth are not interacting with rotor teeth each cycle. 16-8 is not permitted because a 4 phased motor can’t have a rotor where the number of teeth is dividable by 8.

[0058] The 16-10 VRM in figure 40 is questionable. It is possible to make it work with normal configuration in figure 16, but it seems to be some counteracting forces in some positions if you don’t turn on the current later than usual. It is not clear if it is possible to make the 16-10VRM work with the ’’forced short loop” configuration” . The 16-10 VRM configuration should be avoided.

[0059] The 16-12 design shown in figure 41 is good. 16-14 and 16-18 are bad. 16-20 is good. 16-22 is bad. 16-26 is bad. 16-28 is good but now we have rotor teeth that only interact with the stator ever second cycle. All the designs with 16 stator teeth suffer from the ’’long loop” problem.

[0060] 2.5 Known configurations for 6 phased VRM

[0061] The 12-10 VRM configuration is already mentioned in figure 4. It is also stated that you can double number of teeth to 24-20 to get another working known VRM configuration as established in figure 16. An 18-15 configuration is also possible where 3 coils from each phase are distributed evenly around stator.

[0062] For completeness some Comsol simulation are shown in figure 4 in figure figure 49, 50 and 51 to show ’’short loop” and ’’long loop” for 12-10 VRM.

[0063] 3 Description of new configurations for VRM

[0064] The purpose of the invention described in section 3 to 5 is to reduce torque ripples in Variable reluctance motors by eliminating the torque ripple caused by the magnetic field alternating between taking the “short loop” and “long loop” described in the introduction.

[0065] 3.1 New configurations for 3 phased VRM

[0066] The attentive reader has noticed that the 12-10 configuration shown in figure 30 so far has not been mentioned. The reason is that this configuration does not work if you have the same coil configuration as for the 12-8 configuration in figure 28. In figure 30 Phase A should obviously be energized, but if you look on the figure 30 the torque (indicated with arrow 30) on rotor teeth 16.8 and rotor teeth 16.9 has opposite direction and cancel each other out. The 12-14 VRM shown in figure 31 has the same problem. First at 12-16 you have a configuration that works. 12-20 works and 12-22 is bad. 12-26 is bad and 12-28 works, but for 12-28 the rotor teeth only interact with the stator teeth every second cycle, meaning the design is not optimal.

[0067] This means that not all rotors with pair number of teeth not dividable by 6 can be rotor in a 3 phased VRM with 12 coils with standard configuration of the coil. However it is possible to reconfigure the coils so the bad configura- tions work, and even work better than the standard configurations. This new configuration is the scoop of this patent.

[0068] Here is an explanation for how to make a working 12-14 configuration. First, you adjust the arrangement of the stator teeth, so two and two teeth 17.31 in figure 32 have the same angular distance between them as the rotor teeth 16. Each pair of teeth is a teeth pair. In figure 32 the two teeth pairs 17.8 and 17.9 belonging to phase A are marked. Then you assign the coils to new phases as shown in figure 32. Inspection of figure 32 shows that turning the phases on in sequence A-B-C-A. . . will make the rotor rotate in counterclockwise direction. Figure 33 shows the magnetic fields direction through rotor the same way as established in figure 29. The arrows 31 indicate direction of magnetic field to rotor in the same way as arrow 27. This configuration is a ’’forced short loop” configuration. Current direction in the coils can be found by applying the ’’right hand rule” on the arrow 31. The coils (4) belonging to the same phase must be connected accordingly. The motors in figure 32 and figure 33 are identical but arrows and numbers are divided between two figures for clarity.

[0069] Returning to figure 31, it become clear that it is unnecessary to adjust the angle between the stator teeth as done in figure 32 and 33. You can just rearrange the coils in the same pattern as figure 32. This is done in figure 34. The stator teeth is now numbered 17.32 to mark that the distance between the stator teeth 17.32 is different than the distance between the rotor teeth 16. In figure 34 the distance between all stator teeth 17.32 is equal. This is the configuration that gives most space for the coils. It is also possible to have a stator configuration that lay between figure 32 and figure 34. Both the configuration in figure 34 and configuration which lay between figure 32 and figure 34 will give a skewing effect which can be used to reduce torque ripple.

[0070] The ’’forced short loop” configuration with equal distance between the stator teeth is a new inventive step for VRM motors. The advantages of the VRM in figure 34 over the VRM in figure 32 are:

[0071] 1. Equal distance between the stator teeth gives more space for coils.

[0072] 2. The slight displacement between the stator teeth in phase A and the rotor teeth shown at the phase A turnoff point in figure 34 creates a skewing effect which can reduce torque ripples. (In a normal known VRM config- uration the stator and the rotor teeth would be aligned at this point.)

[0073] 3. The magnetic field from different phases no longer interfere. You always get the ’’short loop” and avoid torque variation through a cycle because ’’long loop” is no longer possible, simply because fields from different phases never mix.

[0074] 4. Force balance is maintained since any two coils on opposite side of stator are energized at the same time. 12-10, 12-22 and 12-26 configuration also work with adjacent coils from the same phase. Figure 35 shows the 12-10 VRM with adjacent coils from the same phase.

[0075] 3.2 New configurations for 4 phased VRM

[0076] Now that the ’’forced short loop” configuration is established for 3 phased motors it is easy to see that is is applicable to 4 phased motors too. Figure 42 shows a working 16-14 VRM configuration with 4-phases. Figure 43 shows a working 16-18 configuration. The stator teeth in figure 42 and 43 are numbered 17.42 for 4 phases, equal distance between the teeth and 4 stator teeth in each phase. The stator teeth are numbered 14.42 in both figures because both configurations have 4 phases and equal distance between the stator teeth 14.42. Teeth pairs 17.8 and 17.9 for phase A are also marked on both figures. 16-22 and 16-26 configuration are shown in figure 44 and 45, but because the difference in distance between stator teeth and rotor teeth in these two configurations is bigger than for the 16- 14 and 16-18 configuration it seems best to choose 16-14 or 16-18 when designing a 4 phased VRM with ’’forced short loop” configuration. Both the 16-14 and 16- 18 VRM configurations very interesting, because 4-phased VRMs can in theory, as deducted earlier, have zero torque ripple. In these configurations you also ensure ’’forced short loop”. This is a good starting point for designing a silent VRM. The 16-22 and 16-26 VRM configurations are less interesting, because of the larger mismatch between number of stator and rotor teeth give worse performance than 16-14 and 16-18.

[0077] There is a 4 phased 24-27 VRM with ’’forced short loop” configuration which is balanced and works, which is shown in figure 46. Notice that there is 3 teeth pairs 17.7, 17.8 and 17.9. All the stator teeth are numbered 17.43 in figure 46 to mark that there is 6 teeth in each phase and equal distance between the teeth. This shows you will find similar variants for VRMs with other number of phases if you multiply the teeth in the designs mentioned in this patent with 1.5. Note that numbers of coils in each phase in the motor in figure 46 is 6. 8,10,12.. coils in each phase is also possible, but it is better to make a forced short loop CVRM as mentioned later. 3.3 New configurations for 6 phased VRM

[0078] The first place to look for a ’’forced short loop” configuration would be 24-22 as shown in figure 47 and 24-26 as shown in figure 48. Here all the stator teeth are numbered 17.62. Teeth pairs belonging to phase A are numbered 17.8 and 17.9. It turns out that both configurations works in ’’forced short loop” configuration. It is possible to test other 6-phased configurations, but the pattern is established.

[0079] 3.4 A new way of running a 4 / 6 phased VRM with a 2 / 3 phased VFD

[0080] The normal way of running a 4 phased VRM is with a VFD with 4 half bridges as shown in figure 52. A 6 phased VRM will require a VFD with 6 half H- bridges. Half H-bride VFD for 4 / 6 phased VRM is not standard VFD’s, but there is a way to connect a 4 / 6 VRM to a standard 2 / 3 VFD.

[0081] The chart in figure 8 for 4 phased VRM and the chart in figure 5 for 2 phased PM-motor are almost equal. The same is true for figure 10 for 3 phased VRM and figure 6 for 3 phased PM-motor. This means a 4 phased and 6 phased VRM can be designed with similar torque ripple as a 2 and 3 phased PM-motor. Further it is worth noticing phase A and C in the 4 phased VRM never conduct current at the same time. The same is true for phase B and D. This means that if a 4 phased reluctance motor is fitted with a diodes A, B, C and D as shown in figure 53, you can control the motor with a double H-bridge for a 2 phased PM-motor. Diodes A, B, C and D can be an integrated part of the motor, meaning that 4 phased VRMs can be substitutes for 2 phased PM-motors in many applications, without doing any changes to the drive. Figure 54 show how how to connect a 6 phased VRM to a 3 phased drive.

[0082] With diodes, it should even be possible to attach the 6 phased VRM to a

[0083] 3 phased grid as a motor, presuming it is spun up to synchronous speed first. It is possible to run a reluctance motor as a generator, but only with a VFD were you run the motor as a brake. Without a VFD, spinning the rotor of a reluctance motor does not set up any varying magnetic field in the motor coils like it does in a PM-motor, so it will not work as a grid generator.

[0084] Since 4 and 6 phased VRM theoretically can be designed with zero torque ripple and can be attached to a 2 and 3 phased VFD there is good reason to investigate 4 and 6 phased VRM designs.

[0085] 4 Comsol simulations of old and new 3, 4 and 6 phased VRM

[0086] To build up under the torque ripple discussion, simulation results are presented for the simplest possible 3, 4 and 6 phased VRM shown in figure 2, 3 and 4. The air gap is slightly exaggerated. The coordinate system show dimensions in mm. In the figures letters are phases. 1 is stator iron, 2 is rotor iron, 3 is shaft, 4 coils, 14 is the air gap. In the figures, 14 point to a imaginary circle in the middle of the air gap used in Comsol to create a moving mesh.

[0087] All motors have the parameters in table 2. Number of turns per coil is adjusted according to available space. It is 48 turns for the 3 phased motor, 48 turns for the 4 phased motor and 32 turns for the 6 phased motor. Mark that RMS current is half the peak current in reluctance motors because the current is only half a sinus. With of teeth is set according to what looks reasonable. For the 3-phased motor the teeth with with regard to torque ripple is optimized.

[0088] Table 2: VRM design parameters

[0089] Here are some simulation results of the VRM in figure 2, 3 and 4.

[0090] Table 3: VRM performance parameters

[0091] Min torque is the torque the motor can overcome when it starts to rotate. In applications like actuators and robot arms this is the most important parameter. In the table 3 power is average torque multiplied with rotational speed, meaning that the 4 and 6 phased motor get a penalty because they rotate slower.

[0092] Figure 19, 37, 38, 39, 49, 50 and 51 shows magnetic field at different times in the simulated motors. Figure 24 shows simulated torque in the 3 reluctance motors. Table 3 show torque and other results from the simulations.

[0093] It is staggering how smooth torque you get for 4 and 6 phased VRM com- pared to the 3 phased VRM. Note that the 3 phased VRM in figure 24 does have torque pattern that repeats, while the torque curve for the 4 and 6 phased VRM is random. This has to do with the magnetic field taking the ’’long loop” or ’’short loop” depending of winch phases that are energized. It is possible to avoid the flaw by running current in both directions in each phase, but you don’t want to do it because you then need 4 full H-bridges instead of 4 half H-bridges or the 2 H-bridge and diode solution presented in figure 53. Thicker back iron in stator and rotor reduces the effect of the flaw.

[0094] The torque ripple (defined as (max torque — min torque) / (min torque)) of the 3 phased VRM is 343% in this simulation. This is the lowest torque ripple found, after optimizing teeth width. The torque ripple for the 3 phased VRM is even worse than the theoretical ripple shown in figure 7. Possible explanations are: • The teeth and gaps in the 6-4 configuration become wide since they must fill the entire circle. This gives bad overlap when current switch from one coil to the next. The 12-8 configuration would probably perform closer to the theoretical ripple of 100%.

[0095] • In this case the air gap in the 3 phased motor is too small. If you look at figure 2 the volume of the air gap between rotor teeth and stator teeth phase B is already large, and a little more rotation will not change the volume much. At the same position stator teeth phase A and rotor teeth is not yet near, meaning rotor has to rotate several degrees more before the teeth get close and you get a quick change in reluctance. This contributes to the torque ripple.

[0096] • The 6-4 configuration performs better with larger air gap. With 0.5 mm air gap the 3 phased motor has average torque of 2.4Nm min torque of l.lNm, max torque of 3.2Nm meaning the ripple has dropped to 181%. This is better but still bad.

[0097] To change from 6-4 to 12-8 configuration and shape the tip of the stator teeth to get a more suitable change or reluctance with rotor angel can reduce the torque ripple a lot, but it will be difficult to get it down to the low ripple you have for the 4 phased and 6 phased motor. If you manage, you probably end up with an average torque close to the minimum torque for a 3 phased VRM.

[0098] The ’’forced short loop” VRM configuration is simulated in figure 55 and 56 which is the 4-phased equivalent of figure 32 and 34. You then get a clean periodic torque within 360 electrical degrees as shown in figure 57, and not the mess shown in figure 24. This makes it a good starting point for further optimization. This optimizing can include:

[0099] • Shaping of teeth surface to reduce torque ripple.

[0100] • Adjusting of teeth thickness.

[0101] • Adjusting of back iron thickness

[0102] • Adjusting of teeth high for rotor.

[0103] • Adjusting of teeth high for stator

[0104] • Shaping the coil so it is easier to wind

[0105] Torque of the motor with even space between the coils is higher than for the motor with odd space between the coils, because the motor with even space necessarily has space for more turns in each coil. Also, the torque ripple for the motor with even space between the coils is smaller. It turns out that this is a way to introduce ’’skewing” for reluctance motors. Skewing can then be included in the optimizing.

[0106] 5 Optimized new type of 16-18 VRM

[0107] This motor project is motivated by the need of a motor with a big hole in center so the motor can be an integrated part of a gear solution for a MT400 actuator described in patent NO20220963

[0030] . This is why the shaft has a large diameter and the active part of the motor is thin. The coil is given a shape which should be easy to wind, and it is set of space for insulation for the coil. 5.1 Materials and methods 16-18 VRM

[0108] Table 4 show parameters for the optimized 16-18 VRM. Figure 58 show the geometry. The motor is optimized by numerical simulations in Comsol. Unless otherwise specified values in table 4 are input for the simulation results presented in figure 59 to 62.

[0109] Table 4: VRM design parameters for 16-18 VRM for MT400 actuator

[0110] 5.2 Results 16-18 VRM

[0111] For the design in figure 58 optimal teeth with is 10.1 mm at a peak current 16 A This might seem high, but please note that the ratio between RMS current and peak current is IQ = 2 * It is not the normal because the current in the coil is only a half sinus. The optimal motor had a min torque of 9.5Nm and a torque ripple of 5.7%. This is a low torque ripple for any type of motor. Torque ripple staid below 20% from teeth with 9.0 mm to 10.6 mm. At teeth with 5 mm torque ripple was 154%.

[0112] Electric motors sometimes run below or above rated torque. Also, when you design the air gap to 0.1 mm you will get typical ±0.04mm variation in the air gap if you manufactures a series of motors. Current and air gap obviously influence the torque, but due to saturation it can also influence the torque ripple. After optimizing the teeth with current and the air gap was varied. Simulated results for current density from 8^^ to 24 AL is shown in figure 59. Figure 59 shows what happens if the motor is run much above or below design current of 16A. Result for air gap between 0.06mm and 0.24mm for different current are shown in figure 60, figure 61 and figure 62.

[0113] The motor is optimized on torque ripple. It is room for some more optimizing on back iron thickness for both stator and rotor and teeth high in both rotor and stator to see if it is possible to squeeze some more torque out of the design without compromising the ripple. It is also possible to optimize on motors where stator and rotor teeth have different with. However the design looks reasonable and is already very good, so it is not likely that further optimizing will yield much. 5.3 Discussion 16-18 VRM

[0114] The torque ripple of the optimized 16-18 VRM is low for a wide range of teeth with, current and air gap. The motor is therefore likely to be quiet if run below, at and above rated torque when it is supplied with a sinusoidal current.

[0115] 6 The short loop castellated VRM (CVRM)

[0116] The purpose of the invention described in section 6 is to combine the CVRM technology described in patent NO332245B1

[0027] with the forced short loop technology described in this patent to archive negligible torque ripple for the CVRM too.

[0117] 6.1 Prior art, introduction

[0118] The CVRM seems to be an old idea that has not been properly investigated. Fitzgerald et al.

[0031] mention the CVRM in figure 8.8. First edition of this book is from 1952, and the CVRM is possible mentioned already here. Sargos et. al.

[0032] discuss possible configurations of the CVRM in an article from 1993. A CVRM shows up in patent US3972535A from 1976. Search in oria.no on “reluctance motors” give hit on 5654 articles. Search on “castellated reluctance” give two hits, which are CVRM patents

[0027] and

[0028] . “Castleated” as spelled by Fitzgerald et al.

[0031] give zero hits.

[0119] It is surprising that the CVRM has not been properly investigated before now. Simulations show it can be a serious competitor to the hybrid stepper motor. (It should be mentioned that the hybrid stepper motor is actually a pure PM-motor, because variable reluctance does not contribute to the torque in this motor. The proof is that if you take rotor out of a hybrid stepper motor, demagnetize the magnet(s) and put the rotor back into the motor, the motor won’t rotate when the coils receive current from the VFD.)

[0120] 6.2 New type of CVRM, materials and methods

[0121] The motor presented here is a low speed, high torque motor. Parameters for the motor is presented in table 5.

[0122] Optimization is done in Comsol. The motor presented in figure 63 and 64 is the optimal motor. It is low speed, high torque motors. In actuators you usual prefer high torque, low speed motors to reduce necessary gearing.

[0123] The base design for the CVRM in figure figure 63 and 64 is ’’forced short loop” CVRM with 16 coils in stator. This means that there is 16 major stator teeth 17.1. Each major stator teeth have 10 minor stator teeth denoted 17.1 and 17.2. Adjacent minor stator teeth 17.2 is on each side of the coil slot 4.1 for coils which belong to dififerent phases. Coil slot 4.2 is between coils that belong to the same phase. The distance between adjacent minor stator teeth 17.2 is different than the distance between other minor stator teeth 17.1.

[0124] 16 major stator teeth 17 with 10 minor stator teeth 17.1, 17.2 on each major stator teeth 17 and 162 teeth in rotor makes this a ’’forced short loop” 16 * 10 <- 162 CVRM. (It is the same as a 160-162 CVRM with 160 teeth in stator and 162 teeth in rotor, but instead here, since will be interpreted as a ’’minus” sign in an equation.) It is possible to make ’’forced short loop”

[0125] Table 5: CVRM design parameters

[0126] 16 * n t-> 16 * n + m CVRM where n is 2 or larger. With n = 1 and m = 2 you have a 16-18 VRM in figure 43. m must be pair, m can be negative, m can not be zero or dividable with number of phases. There are many possible configurations, but practically you always end up with choosing m = 2 because you want the distance between adjacent minor teeth 17.2 to be larger than the distance between other minor teeth 17.1, or it will be more difficult than necessary to wind the coils. ’’Forced short loop” CVRM configurations are:

[0127] phased) phased) phased) phased)

[0128] All VRMs mentioned earlier can be converted into forced short loop CVRM, but only ’’forced short loop” CVRM with 4 phases and 16 coils or 6 phases and 24 coils are likely to be commercially interesting. The 3 phased CVRM will have the same cogging as a 3 phased VRM and 5 phased CVRM have no advantage over 4 phased CVRM. Only very big motors will benefit from introducing more coils.

[0129] 6.3 New type of CVRM, results

[0130] The work on the CVRM is comprehensive. It is done optimization on all pa- rameters in table 5, and with this many parameters numbers of possible motors is huge. Many of the ideas presented earlier originates from this work. A list of the ideas is presented here.

[0131] • 1 mm teeth high is enough for CVRM. VRM also benefit from low rotor teeth high.

[0132] • It is smart to slant the edge of the minor teeth (17.1 and 17.2), so only the magnetic steel in top of the teeth is completely saturated. 60 degrees angle between the side edge of the teeth is good. You can do this for VRM too, but the benefit is less.

[0133] • The idea of ’’skewing” presented in figure figure 56 originates from the CVRM optimization. In the first design, all teeth in phase B in figure 64 was straight above a the rotor teeth, but in later designs Min torque go up and ripple go down if the stator teeth 16.3 are slightly shifted in the right way compared to the rotor teeth 16.3 as shown in figure 65. If you study this figure carefully you see the shift in the form of different overlap between stator teeth 17.3 and rotor teeth 16.3 in position Pl and P2 in the figure 65. Simulations show that the solution where adjacent teeth between coil A-A and coil A-B has the same distance give a good solution with regard to torque ripple. The parameter ’’Skew” that controlled devi- ation between distance in regular stator teeth relative to distance between rotor teeth then became obsolete, and it was only necessary to optimized on ’’Gapskew” which controls distance between adjacent teeth on each side of the coil gap 4.2.

[0134] • The side wall between two adjacent coil gaps must be parallel. If these side walls are not parallel, the copper wire will slide during winding, making it nearly impossible to wind the coil with good fill factor.

[0135] • You need some space (parameter ’’CoilDepth”) between the air gap and the coil so the magnetic field can reach the outer teeth under each coil.

[0136] Figure 66 shows torque at different coil current for air gap between stator and rotor of 0.1 mm. The wire diameter in these simulations is 0.7 mm, so you have to multiply current with ca 2 to get current density.

[0137]

[0138] 6.4 New type of CVRM, discussion

[0139] The CVRM motor is optimized for a current of 6 A which equals 12.2 A / mm2and an air gap of 0.1 mm. At this current and air gap torque is 18Nm and torque ripple is 12%. The CVRM can reach 25Nm without too much ripple and can be pushed above 30Nm if noise is acceptable.

[0140] Due to saturation torque is only partially linear with current. This is shown in figure 66. CVRM motor also seems to require some current to set up a magnetic field, because torque with IA current in the coils is only 0.65Nm for the CVRM. This torque increase to 19Nm for 6A, and this is definitely not linear. Simulations give the relation in table 6.

[0141] Table 6: CVRM, Ig, U and T at 15RPM

[0142] 6.5 New type of CVRM, conclusion

[0143] Both the 16-18VRM and the CVRM are good motors with regard to torque ripple. The CVRM can provide a torque 2-3 times higher than the 16-18VRM for investigated size. It is likely that this ratio will increase for bigger sized motors and go against 2 for smaller sized motors. This because there is a practical minimum size of the teeth meaning that you have to reduce number of teeth in smaller CVRMs. The investigated CVRM will require 9 times higher frequency of the current to rotate at the same speed as the investigated 16- 18VRM.

[0144] Both motors are good motors that can work with a two phased power if the motor is fitted with diodes according to figure 53.

[0145] Simulations axe only done with silicon steel NGO 50PN250 which is a low loss steel for high frequency, which saturate at relatively low magnetic field. Choosing e.g. NGO 50PN800 which saturate at a high magnetic field will give more torque but also more loss at high frequency.

[0146] It is simulated torque as function of air gaps for both the 16-18VRM and the CVRM. Figure 62 show the 16-18VRM can handle an air gap of up to 0.24 mm without too much loss in torque. Figure 68 show the CVRM is very sensitive to large air gap. For the CVRM it is important that the air gap is 0.05 to 0.1 mm. References

[0147] [1] Wiguna, C. A., Furqani, J. F., and Chiba, A., “Improved Current Pro- file Selection for Noise Reduction of Switched Reluctance Motor at Middle Speed Considering Back EMF,” IEEE Transactions on industry applications, Vol. 57:5, October 2021.

[0148] [2] Li, S., Zhang, S., Habetler, T. G., and Harley, R. G., “Modeling, Design Optimization, and Applications of Switched Reluctance Machines - A Re- view,” IEEE Transactions on industry applications, Vol. 55:3, June 2019.

[0149] [3] Zhenyao, X., Tao, L., Fengge, Z., Yue, Z., Dong-Hee, L., and Jin- Woo, A., “A Review on Segmented Switched Reluctance Motors,” Energies, Vol. 15, 2022.

[0150] [4] Xiaodong, S., Kaikai, D., Gang, L., Youguang, G., and Jianguo, Z., “Study on Segmented-Rotor Switched Reluctance Motors With Different Rotor Pole Numbers for BSG System of Hybrid Electric Vehicles,” IEEE Transactions on vehicular technology, Vol. 68:6, June 2019.

[0151] [5] Hubert, C. I., Electric Machines, Theory, Opperation, Applications, Adjustment, and Controll, Prentice-Hall, 1991.

[0152] [6] Jae-Kwang, L., Dong-Hoon, J., Jongsuk, L., Ki-Doek, L., and Leel, J., “A Study on the Synchronous Reluctance Motor Design for High Torque by Using RSM,” IEEE Transactions of magnetics, Vol. 54, March 2018.

[0153] [7] Donaghy-Spargol, C. M., Mecrow, B. C., , and Widmer, J. D., “Electro- magnetic Analysis of a Synchronous Reluctance Motor With Single-Tooth Windings,” IEEE Transactions of magnetics, Vol. 53, November 2017.

[0154] [8] Villet, W. T. and Hamper, M. J., “Variable-Gear EV Reluctance Synchronous Motor Drives — An Evaluation of Rotor Structures for Position-Sensorless Control,” IEEE Transactions on industrial electronics, Vol. 61:10, October 2014.

[0155] [9] Yuanfeng, L., Croonen, J., Frikha, M. A., Baghdadi, M. E., and Hegazy, O., “A Comprehensive Performance Comparison between Segmental and Con- ventional Switched Reluctance Machines with Boost and Standard Con- verters,” Energies, Vol. 16, 2023.

[0156]

[0010] Yasa, Y., Sozer, Y., and Garip, M., “Acoustic Noise Mitigation of Switched Reluctance Machines With Leaf Springs,” IEEE Transaction of industrial electronics, Vol. 15, February 2023.

[0157]

[0011] Zhenyao, X., Jialu, D., Zhijun, F., Huijun, W., and Fengge, Z., “Character- istics Analysis of a Novel 12 / 8 Double Stator Bearingless Switched Reluc- tance Motor,” Journal of Electrical Engineering and Technology, Vol. 18, October 2023, pp. 347-357.

[0158]

[0012] Ruizhen, C., Yi, H., Xilian, W., Xinyu, F., and Zhuliang, H., “Torque analytical model and self-decoupling characteristics analysis of a bearingless switched reluctance motor,” IET Power Electronics, Vol. 14, March 2021, pp. 1149-1156.

[0013] Khodadadeh, 0., CheshmehBeigi, H. M., and Hossein, M. M., “Design, simulation and optimisation of a novel low ripple outer-rotor switched re- luctance machine for variable speed application,” International Journal of Electronics, February 2023.

[0159]

[0014] Radmaneshl, H. and Farahani, E. F., “Performance evaluation of a new modular split-tooth permanent magnet-assisted switched reluctance mo- tor,” IET Electric Power Applications, November 2022.

[0160]

[0015] Vinh, V. Q. V. and Ha, V. T., “Improved Torque Ripple of Switched Re- luctance Motors using Sliding Mode Control for Electric Vehicles,” Engineering Technology & Applied Science Research, Vol. 13:1, January 2023, pp. 10140-10144.

[0161]

[0016] Chun, G., Fanyu, M., Zhiyue, Y., Ronghai, Q., Zicheng, L., and Jikai, S., “Online Calibration of Sensorless Position Estimation for Switched Re- luctance Motors With Parametric Uncertainties,” IEEE Transactions on power electronics, Vol. 35:11, November 2020.

[0162]

[0017] Geldhof, K. R., Van den Bossche, A. P. M., and Melkebeek, J. A., “Rotor-Position Estimation of Switched Reluctance Motors Based on Damped Voltage Resonance,” IEEE Transactions on industrial electronics, Vol. 57:9, September 2010.

[0163]

[0018] Soares, F. and Branco, P. J. C., “Simulation of a 6 / 4 Switched Reluctance Motor Based on Matlab / Simulink Environment,” IEEE Transactions on aerospace and electronic systems, Vol. 37:3, July 2001.

[0164]

[0019] Han-Chieh, C., Jer-Huan, J., Wei-Mon, Y., and Ruei-Bin, S., “Thermal performance analysis of a 30 kW switched reluctance motor,” International Journal of Heat and Mass Transfer, Vol. 114, June 2017, pp. 145-154.

[0165]

[0020] Tong, B., Heng, N., Long, C., Libing, J., and Rongge, Y., “Torque ripple reduction for switched reluctance motors using global optimization algo- rithm,” Journal of Power Electronics, Vol. 22, August 2022, pp. 1897-1907.

[0166]

[0021] An, L., Dong, J., Jianbo, S., Zicheng, L., and Christopher, H. T. L., “Unified Vector Torque Control for Reluctance Motor With Different Coil Pitch,” IEEE Transactions on industrial electronics, Vol. 70:6, June 2023.

[0167]

[0022] Baoquan, K., Dagang, X., and Xihui, Y., “Multiphase magnetic resistance motor with pulse torque compensation structure,” Patent CN101982931A, Mar. 2011.

[0168]

[0023] Deguchi, J., Funahashi, M., Yoshizue, K., Hoshi, N., Abe, T., Kobayashi, Y., and Murakami, Y., “Control device for switched reluctance motor,” Patent US2018102722A1, Oct. 2016.

[0169]

[0024] Padmanaban, L. A. and Saravanan, P., “Design, analysis and compari- son of switched reluctance motors for electric vehicle application,” Journal for Control, Measurement, Electronics, Computing and Communications, Vol. 64:2, November 2023, pp. 239-247.

[0025] Patil, S. V. and Saxena, R., “SR Motor T-I Characteristics Performance Simulation Validation through Experimental Results,” Cybernetics and Systems, January 2023.

[0170]

[0026] She, Z., “Switched reluctance motor,” Patent CN209767257U, Jul. 2019.

[0027] Brennvall, J. E., “Three phased balanced or unbalanced asymmetric reluc- tance motor,” Patent NO332245B1, Aug. 2012.

[0171]

[0028] Brennvall, J. E., “Coil slot for Castellated Variable Reluctance Motor (CVRM),” Patent N020150240A1, Feb. 2015.

[0172]

[0029] Brennvall, J. E., “Multi plate reluctance motor,” Patent N020210788A1, Dec. 2020.

[0173]

[0030] Brennvall, J. E., “Actuator for a Valve,” Patent application NO20220963, Sep. 2022.

[0174]

[0031] Fitzgearald, A. E., Kingsley, Jr., C., and Umans, S. D., Electric Machinery Sixth edition, McGraw-Hill, 2003.

[0032] Sargos, F., Zaskalicky, P., and Gudefin, E., “Generalized theory of the structures of reluctance step motors,” Groupe de Recherches en Electrotech- nique et Electronique de Nancy, CNRS URA D1438, ENSEM, 54516 Van- doeuvre Cedex, France, 10 1993.

Claims

Claims1. Variable reluctance motor (VRM) with N phases where N > 3, and with M teeth pairs (17.7, 17.8, 17.9) where M > 2, where the VRM comprises a rotor iron (2) with rotor teeth 16, coils (4) and a stator iron (1) com- prising stator teeth (17.31, 17.32, 17.42, 17.43, 17.62, 17.1) and a shaft (3) characterized by that the adjacent teeth is grouped in teeth pairs (17.7, 17.8, 17.9) which have coils (4) belonging to the same phase wound around them arranged so the magnetic field go in a loop inside each teeth pair (17.7, 17.8, 17.9) as indicated by arrow (31), and the teeth pairs (17.7, 17.8, 17.9) are distributed evenly around the stator iron (1) for all phases so net radial forces on the shaft (3) is minimized.

2. VRM according to claim 1, where N = 3 and M = 2 creating a VRM comprising 3 phases and 12 coils (4) wound around stator teeth (17.31, 17.32) characterized by that two coils (4) from the same phase are placed around adjacent teeth pair (17.8) and that the other teeth pair (17.9) is placed on the opposite side of the stator (1) relative to the first teeth pair(17.8) for all phases.

3. VRM according to claim 1, where N = 4 and M = 2 creating a VRM comprising 4 phases and 16 coils (4) wound around stator teeth (17.42) characterized by that two coils (4) from the same phase are placed around adjacent teeth pair (17.8) and that the other teeth pair (17.9) is placed on the opposite side of the stator (1) relative to the first teeth pair(17.8) for all phases.

4. VRM according to claim 1, where N = 4 and M = 3 creating a VRM comprising 4 phases and 24 coils (4) wound around stator teeth (17.43) characterized by that two coils (4) from the same phase are placed around adjacent teeth pair (17.7) and that the teeth pairs (17.7, 17.8, 17.9) are placed 120° apart for all phases.

5. VRM according to claim 1, where N = 6 and M = 2 creating a VRM comprising 6 phases and 24 coils (4) wound around stator teeth (17.62) characterized by that two coils (4) from the same phase are placed around adjacent teeth pair (17.8) and that the other teeth pair (17.9) is placed on the opposite side of the stator (1) relative to the first teeth pair(17.8) for all phases.

6. VRM according one of the claims 1, 3 and 4 characterized by that a diode (32) is included in each phase, and two and two phases are connected in parallel so the current will go through one phase when the current is positive and through the other phase when the current is negative, so 4 phased VRMs can be run with a standard 2-phased VFD or be connected to a 2 phased grid.

7. VRM according to claim 1 and 6 characterized by that a diode (33) is included in each phase in the same way as in claim 6 so the 6 phased VRMs can be run with a standard 3-phased VFD or be connected to a 3 phased grid.

8. VRM according to claim 1-5 characterized by that the distance between the stator teeth (17.32, 17.42, 17.43, 17.62) in teeth pairs (17.7, 17.8, 17.9) have different internal distance (angle) than the internal distance (angle) between the rotor teeth (16) to get more space for the coils (4) and to achieve a skewing effect which can reduce torque ripple.

9. VRM according to claim 1-8 characterized by that the VRM is a CVRM where comprising 2 or more minor stator teeth (17.2) on each (major) sta- tor tooth 17.1 where the stator teeth 17.1 have configurations as described according to one of the claims 1-5 and the rotor teeth (16.2) have approx- imately the same size as the minor stator teeth (17.2).

10. VRM according to claim 1-5 and 9 characterized by skewing of the teeth defined as the stator minor teeth 17.3 having a slightly different internal distance (angle) compared to the internal distance (angle) between the rotor teeth 16.3 to achieve a skewing effect which can reduce torque ripple.

Images