Understanding the RSM (reluctance synchronous machine)

[CasualKilla]

Hi guys. I have managed to get a good grasp of the IM thanks to some very cleaver and friendly gentleman on this forum, but I am now struggling to understand the RSM operation.

 

I have an understanding of magnetic reluctance and how the motor is designed to have different reluctance depending on the orientation, but I am still unsure how this creates torque.

 

I have an idea on how the torque is generated at synchronous speed, but I think it may be flawed. The figures below show my logic.

 

Basically I know that flux does like not to change, so if the motor falls behind the stater field, it functions as a magnet for a brief period of time. This magnet is attracted to the rotating stater poles and thus the motor speeds back up to synchronous speed.

 

 

 

I have a feeling this logic may be slightly flawed and I also have no idea how the motor can create torque at anything but synchronous speed, so any help in this regard is appreciated.

 

Any links to videos, articles or easy to follow mathematical models would also be greatly appreciated. I am willing to put in time to get my theory down.

Edited March 16, 2015 by CasualKilla

[studiot]

As a start reluctance motors fall into two types.

 

The first type employes some form of permanent magnet.

 

The second type has its poles magnetised by induction.

Of this second type synchronous reluctance motors (eg old fashioned electric clocks) are run up to speed unmagnetised and then the poles are magnetised.

[CasualKilla]

As a start reluctance motors fall into two types.

 

The first type employes some form of permanent magnet.

 

The second type has its poles magnetised by induction.

Of this second type synchronous reluctance motors (eg old fashioned electric clocks) are run up to speed unmagnetised and then the poles are magnetised.

 

Yes I understand some types used permanent magnets to enhance the power factor. The PM torque is alot easier to understand, what is really getting me is these "induced poles" that occur from the difference in reluctance.

 

I am concerned with understanding a motor like the one shown below, though it may be easier to explain a 2 pole motor to start with.

 

Edited March 16, 2015 by CasualKilla

[Enthalpy]

A true reluctance machine runs only when synchronous. It needs electronics that provides current at each instant to the proper coils so a permanent torque results. This means that the current's phase and frequency are made as a function of the rotor's angle, and it begins at zero frequency to start the motor.

 

Though, many reluctance machines provide some path to induced currents, and then they achieve some torque despite the current's frequency doesn't match the mechanical frequency, just like in an asynchronous machine. These motors can start with a current at fixed frequency like 50Hz. Many centrifuges (for blood separation and so on) work that way.

 

Flux barriers would be so nice to have.... How do you achieve a good one? Magnetic leakage is a strong limit to every magnetic machine, and especially the reluctance machine is a poor answer to this, which limits the torque and power as compared with an induction machine. Type I superconductors are flux barriers but only against a very small induction, bringing nothing better than air.

[CasualKilla]

A true reluctance machine runs only when synchronous. It needs electronics that provides current at each instant to the proper coils so a permanent torque results. This means that the current's phase and frequency are made as a function of the rotor's angle, and it begins at zero frequency to start the motor.

 

Though, many reluctance machines provide some path to induced currents, and then they achieve some torque despite the current's frequency doesn't match the mechanical frequency, just like in an asynchronous machine. These motors can start with a current at fixed frequency like 50Hz. Many centrifuges (for blood separation and so on) work that way.

 

Flux barriers would be so nice to have.... How do you achieve a good one? Magnetic leakage is a strong limit to every magnetic machine, and especially the reluctance machine is a poor answer to this, which limits the torque and power as compared with an induction machine. Type I superconductors are flux barriers but only against a very small induction, bringing nothing better than air.

 

You can also add conductors into the flux barriers, or have the entire flux barriers made of conductive materials. I am working on this type of design to achieve a line-start RSM. But alas, I still have no idea how the RSM becomes a magnet or creates torque. If it was a PM machine it would be alot easier to understand.

[Enthalpy]

Once the reluctance motor runs synchronously with the mains, the added conductors don't prevent flux from passing the "barriers". Conductors only hinder the flux variations over time.

 

By the way, you can suppress the sketched "induction barriers". What you need are plain pole shoes with a big spacing between them, since distance is the only way to reduce the leaked flux. A look at existing designs maybe?

 

The torque comes from the reluctance that varies with the rotor's orientation in the stator's induction. You can compute it through the flux that the rotor's orientation permits; one numerical way is to check how many V*s at the stator's coils an angle variation of the rotor induces (either the gap defines the induction, or the saturation of the magnetic material); combined with the amps*turns at the stator, you get an energy per angle unit, which is a torque.

 

The whole difficulty is to evaluate the flux as a function of the angle, because the flux passes much through the air. This is a fundamental limit of the reluctance machine, which can't use big currents hence isn't very strong at a given size.

 

At an induction machine, stator and rotor currents define where the induction starts and stops. These can be - and are, at big machines - many times stronger than the current that would create the same induction over the gap or over a limited air distance. In addition, induction machines use inductions between -1T and +1T for instance, not just +0.5T and +1.5T for instance at a reluctance machine. All this combined makes the induction machine much more powerful.

[CasualKilla]

Once the reluctance motor runs synchronously with the mains, the added conductors don't prevent flux from passing the "barriers". Conductors only hinder the flux variations over time.

 

By the way, you can suppress the sketched "induction barriers". What you need are plain pole shoes with a big spacing between them, since distance is the only way to reduce the leaked flux. A look at existing designs maybe?

 

The torque comes from the reluctance that varies with the rotor's orientation in the stator's induction. You can compute it through the flux that the rotor's orientation permits; one numerical way is to check how many V*s at the stator's coils an angle variation of the rotor induces (either the gap defines the induction, or the saturation of the magnetic material); combined with the amps*turns at the stator, you get an energy per angle unit, which is a torque.

 

The whole difficulty is to evaluate the flux as a function of the angle, because the flux passes much through the air. This is a fundamental limit of the reluctance machine, which can't use big currents hence isn't very strong at a given size.

 

At an induction machine, stator and rotor currents define where the induction starts and stops. These can be - and are, at big machines - many times stronger than the current that would create the same induction over the gap or over a limited air distance. In addition, induction machines use inductions between -1T and +1T for instance, not just +0.5T and +1.5T for instance at a reluctance machine. All this combined makes the induction machine much more powerful.

Your comment on the conductors confused me. Conductors like copper and aluminium have approximately the same permeability of air, so it should not impede the RSM torque at synchronous speed. Also there is no current induced at synchronous speed, so they shouldn't cause any weird effects. Am i missing something?

 

That design is very similar to existing designs, except those use curved barriers. By suppress I assume you mean make the barriers thinner and less numerous?

 

Anyway, I will be running simulations to see the effect of increasing and decreasing the flux barrier thickness, but thank you for your insight, you may save me alot of time. it takes about 2 hours to run a simulation.

 

Here are two initial concepts I will be evaluating, I will use an existing design for the RSM component and experiment with adding the induction effects. I will also do a third design where I will fiddle with the RSM barriers myself, will probably looks something like this:

 

Existing RSM design as basis:

 

 

Custom RSM design:

 

 

I kept the RSM flux barriers simple to allow for easy variation of simulation parameters.

Edited March 20, 2015 by CasualKilla

[Enthalpy]

There you have a standard design of a reluctance motor

http://en.wikipedia.org/wiki/Reluctance_motor

variants exist, especially micro-stepper which have several teeth per pole, but these teeth join a bit farther from the gap. My opinion is that, for any significant power and torque, the "flux barriers" don't bring an advantage. Unless I miss something.

 

See the drawing a Wiki? There are wiiiiiide gaps between the poles, and one good reason is that the flux shall not cross them. TIny "Flux barriers" won't do that job.

 

I had thought at superconductor time ago but the type I get overriden by a tiny induction already, and type II don't reject the flux. Normal conductors can block only a varying flux, which isn't the normal operation of a reluctance motor.

 

Anyway, I got disappointed by the reluctance motor because it's not faster than a permanent magnet motor. At identical rotor diameter it would, but at identical torque the permanent magnet motor can have a smaller diameter, and then it runs as quickly and is smaller, more efficient, silent and so on.

[CasualKilla]

There you have a standard design of a reluctance motor

http://en.wikipedia.org/wiki/Reluctance_motor

variants exist, especially micro-stepper which have several teeth per pole, but these teeth join a bit farther from the gap. My opinion is that, for any significant power and torque, the "flux barriers" don't bring an advantage. Unless I miss something.

 

See the drawing a Wiki? There are wiiiiiide gaps between the poles, and one good reason is that the flux shall not cross them. TIny "Flux barriers" won't do that job.

 

I had thought at superconductor time ago but the type I get overriden by a tiny induction already, and type II don't reject the flux. Normal conductors can block only a varying flux, which isn't the normal operation of a reluctance motor.

 

Anyway, I got disappointed by the reluctance motor because it's not faster than a permanent magnet motor. At identical rotor diameter it would, but at identical torque the permanent magnet motor can have a smaller diameter, and then it runs as quickly and is smaller, more efficient, silent and so on.

That is a switched reluctance motor, and it has some crazy stuff going on in the stator, stuff that I am not smart enough to even consider. The diagrams I posted are Reluctance Synchronous Machines, which use the same stator as an induction motor.

 

I had a discussion with one of my lecturers about this, and I asked him, why can you not just use some kind of cross for the rotor (4 pole). (This is the one of my original, now discarded designs below)

 

 

But the above design will not work in a IM stator. I am not 100% sure why, but I think it is because the flux is deliver through discrete teeth and in the above design all the flux entering into those massive air gaps will be wasted.

 

This is what a modern existing design looks like, courtesy of ABB:

 

Edited March 20, 2015 by CasualKilla

[Enthalpy]

I don't understand the benefit of such a design. I'm still convinced that simple teeth at the rotor, the rest plain steel, is at least as good. I can't even understand the picture, which seems edited. Do you have a link to ABB's document?

[CasualKilla]

I don't understand the benefit of such a design. I'm still convinced that simple teeth at the rotor, the rest plain steel, is at least as good. I can't even understand the picture, which seems edited. Do you have a link to ABB's document?

LOL, we are in the same boat bud, I don't know how it works, but it does. XD. You are still thinking about switched reluctance machines, which work differently on the stator side.

 

have a look at this video: https://www.youtube.com/watch?v=XuwV5dSvjh4. It is not very technical, but it will atleast prove to you it is a thing.

 

and here is a video by ABB: https://www.youtube.com/watch?v=69mtuom774E.

[Enthalpy]

Please provide a link to a technical description by ABB - not to videos of unknown origin.

Until I see that, I consider the picture as a fake. It was a picture of a squirrel cage motor, where someone has pasted some bizarre drawing.

Sorry, but either you try to pull my leg, or someone pulled yours.

I won't invest any time in that without a serious reference.

[CasualKilla]

Please provide a link to a technical description by ABB - not to videos of unknown origin.

Until I see that, I consider the picture as a fake. It was a picture of a squirrel cage motor, where someone has pasted some bizarre drawing.

Sorry, but either you try to pull my leg, or someone pulled yours.

I won't invest any time in that without a serious reference.

Please tell me your joking.. You mock me for posting easy puzzles, but you really proposing reluctance synchronous machines do not exist.

 

enjoy: http://www.abb.com/product/seitp322/c347a19269e8ad4ac125785b00217276.aspx?productLanguage=us&country=ZA

Edited March 22, 2015 by CasualKilla

[Enthalpy]

Seen it. I'll give it a thought.

[Enthalpy]

It's getting clearer.

 

- Something indeed happened in the world of electric motors when I was looking in an other direction.

- Most "pictures" at Abb's doc are edited, more or less heavily (one is completely botched, intentionally I'd say). Do not rely on them to infer dimensions.

- Explanations that look convincing, but available in German only:

http://de.wikipedia.org/wiki/Synchron-Reluktanzmotor

- Have a look at Vagati's patent too.
http://www.google.com/patents/US5818140

 

It really is a reluctance motor. To my understanding, it has the properties and performances of any reluctance motor, BUT Alfredo Vagati managed to let it run smoothly by subdividing the pole shoes and the gaps. The division shall have no simple ratio with the stator's teeth. Interestingly, the divided gaps are now narrow, but they nearly add up to define the maximum reluctance of the rotor, so this maximum reluctance isn't lost when subdividing the pole shoes and gaps.

 

Check with your stamping workers what minimum metal width they want. Abb's "pictures" lie there, it's nearly certain. Beware, at the rotor's rim, this has consequences on the motor's performance.

 

"Flux barriers" must be wording of the patent, but that's misleading. Alas, we don't have barriers against the induction - a basic difficulty of magnetic design, especially harmful at reluctance motors.

 

Varied approaches can evaluate the torque. I suggest again to compute the flux variation, multiply by the current around the best positions - and as usual, the number of poles, an equivalent value for the non-uniform induction, AND not forget how currents add in a three-phase stator, not confound peak, mean and rms currents, and so on. The less simple part is that the stator's induction is phased (tilted) with respect to the rotor. This needs magnetic leaks small enough, which is possible on a small machine.

 

Now that a reluctance motor can run smoothly, its advantage at moderate power machines is that the rotor losses are small. Then, at identical losses, it can be smaller than a squirrel cage machine. Not smaller than a permanent magnets machine, but cheaper. And for MW-GW machines, I don't see it compete, because the leaks prevent the reluctance machine work, at least if reasonably sized.

[CasualKilla]

It's getting clearer.

 

- Something indeed happened in the world of electric motors when I was looking in an other direction.

- Most "pictures" at Abb's doc are edited, more or less heavily (one is completely botched, intentionally I'd say). Do not rely on them to infer dimensions.

- Explanations that look convincing, but available in German only:

http://de.wikipedia.org/wiki/Synchron-Reluktanzmotor

- Have a look at Vagati's patent too.

http://www.google.com/patents/US5818140

 

It really is a reluctance motor. To my understanding, it has the properties and performances of any reluctance motor, BUT Alfredo Vagati managed to let it run smoothly by subdividing the pole shoes and the gaps. The division shall have no simple ratio with the stator's teeth. Interestingly, the divided gaps are now narrow, but they nearly add up to define the maximum reluctance of the rotor, so this maximum reluctance isn't lost when subdividing the pole shoes and gaps.

 

Check with your stamping workers what minimum metal width they want. Abb's "pictures" lie there, it's nearly certain. Beware, at the rotor's rim, this has consequences on the motor's performance.

 

"Flux barriers" must be wording of the patent, but that's misleading. Alas, we don't have barriers against the induction - a basic difficulty of magnetic design, especially harmful at reluctance motors.

 

Varied approaches can evaluate the torque. I suggest again to compute the flux variation, multiply by the current around the best positions - and as usual, the number of poles, an equivalent value for the non-uniform induction, AND not forget how currents add in a three-phase stator, not confound peak, mean and rms currents, and so on. The less simple part is that the stator's induction is phased (tilted) with respect to the rotor. This needs magnetic leaks small enough, which is possible on a small machine.

 

Now that a reluctance motor can run smoothly, its advantage at moderate power machines is that the rotor losses are small. Then, at identical losses, it can be smaller than a squirrel cage machine. Not smaller than a permanent magnets machine, but cheaper. And for MW-GW machines, I don't see it compete, because the leaks prevent the reluctance machine work, at least if reasonably sized.

Thank you for your input enthalpy, I really appreciate you taking the time to give a well thought out answer. I just have to query one thing you said though, "same properties and performances of any reluctance motor". This design differs from a switched reluctance machine quite heavily, since the stator configuration is very different. This machine uses the same stator as the induction machine.

 

Why do you think separated flux carriers are used, instead of one large cross like a switch reluctance machine, is the the due fact that flux is delivered from discrete stater teeth, rather than the perfectly uniform rotating field we like to imagine the stator creates.

Edited March 25, 2015 by CasualKilla

[Enthalpy]

It is my understanding, stated by the inventor's patent as well, that the improvement is a smoother operation. In others aspect, I believe the performance is pretty much the same, whether the rotor laminations are solid or now slotted.

 

What Alfredo Vagati has nicely seen is that several thinner slots add their reluctance as if they were the traditional thicker spacing between the pole shoes, so they keep its performance, well done.

 

And then, a well-chosen number of thinner slots and shoes (narrower than one slot) permit a smoother operation. This number shall not equal the number of slots at the stator; quite the opposite, it must differ, and the patent gives some rules of choice.

 

The idea behind smooth operation is that the iron teeth at the stator and at the rotor do not meet and leave an other all at the same time, which would be the case with equal numbers. As far as possible, different pairs of stator+rotor teeth shall meet and leave at angles regularly spread over one rotation.

 

You know the Vernier scale at calipers? On the third picture there

http://en.wikipedia.org/wiki/Vernier_scale

the mover has 10 divisions on the length of 9 stator divisions. This spreads the coincidences of stator and mover divisions regularly over one period. Same trick; now at the motor, the cogging torques of teeth pairs don't add because they appear at different angles.

 

This one idea wasn't new for other motors. Salient poles DC motors can have 3 or 5 pole shoes at the rotor for 2 poles at the stator, to cog a little less. Squirrel cage motors also have a number of rotor bars in complicated relation with the number of pole shoes at the stator. A first attempt would be numbers relatively prime at the stator and rotor, but electromagnetism imposes some restrictions in the choice.

 

The patent tells which numbers are good.

http://www.google.com/patents/US5818140

[CasualKilla]

It is my understanding, stated by the inventor's patent as well, that the improvement is a smoother operation. In others aspect, I believe the performance is pretty much the same, whether the rotor laminations are solid or now slotted.

 

What Alfredo Vagati has nicely seen is that several thinner slots add their reluctance as if they were the traditional thicker spacing between the pole shoes, so they keep its performance, well done.

 

And then, a well-chosen number of thinner slots and shoes (narrower than one slot) permit a smoother operation. This number shall not equal the number of slots at the stator; quite the opposite, it must differ, and the patent gives some rules of choice.

 

The idea behind smooth operation is that the iron teeth at the stator and at the rotor do not meet and leave an other all at the same time, which would be the case with equal numbers. As far as possible, different pairs of stator+rotor teeth shall meet and leave at angles regularly spread over one rotation.

 

You know the Vernier scale at calipers? On the third picture there

http://en.wikipedia.org/wiki/Vernier_scale

the mover has 10 divisions on the length of 9 stator divisions. This spreads the coincidences of stator and mover divisions regularly over one period. Same trick; now at the motor, the cogging torques of teeth pairs don't add because they appear at different angles.

 

This one idea wasn't new for other motors. Salient poles DC motors can have 3 or 5 pole shoes at the rotor for 2 poles at the stator, to cog a little less. Squirrel cage motors also have a number of rotor bars in complicated relation with the number of pole shoes at the stator. A first attempt would be numbers relatively prime at the stator and rotor, but electromagnetism imposes some restrictions in the choice.

 

The patent tells which numbers are good.

http://www.google.com/patents/US5818140

Oh ok I see what you mean by smoother operation now, if flux carrier = flux teeth you could have the flux being completely blocked at a certain orientation. Uneven numbers allow flux to enter the rotor into the flux carriers at any orientation.

 

I believe this is *similar to the reason we apply a slight slant to induction motors, otherwise you have large torque when aligned with the stator teeth and smaller torque when no aligned. Similarly, you don't want conductor bars = stator teeth.

 

 

One thing I must mention is that a rotor design like (above) has a mathematical limit of 3 for the saliency ratio (can't find the proof on the web, but it is in HANDBOOK OF ELECTRIC MOTORS Second Edition, Revised and Expanded), whereas a design with separated flux carriers can achieve much higher saliency ratios (more than 10)

 

The torque and PF in these machines dependent on the saliency ratio. Derivation of these equations involve implementing the dq axis transformation.

 

http://en.wikipedia.org/wiki/Dqo_transformation#Geometric_Interpretation

 

Which coverts the stator winding into 2 dc (WHATTT??) currents. The Ld and Lq is defined and the salency ratio = Ld/Lq.

 

So, the multi-carrier design can achieve a much higher torque, the smooth operation is just a nice bonus.

 

Trouble is I cant intuitively figure how the multi-carrier design achieves better saliency, and the maths is beyond me.

Edited March 25, 2015 by CasualKilla

[Enthalpy]

I have the strongest doubts that Vagati's design increases the torque. By the way, his patent doesn't even suggest that. But you believe what you want.

[CasualKilla]

I have the strongest doubts that Vagati's design increases the torque. By the way, his patent doesn't even suggest that. But you believe what you want.

I similarly have my doubts, based on simple intuition, but sometimes we must put that aside. This is not a question of belief, I am reading the material and trying to make sense of it. I understand you are just trying to help, and I appreciate that, but from here it looks like you are the one clinging to beliefs...

 

I have PMed you an extract of the handbook, perhaps the Maths makes sense to you.

[Enthalpy]

Maybe... I consider that the torque results from the flux variation per unit angle and from the stator current (summed over several phases, accounting properly the poles, and so on).

 

My understanding is that Vagati's rotor offers the same flux variation as the more classical rotor, narrower but plan, and that the change in flux or reluctance happens over the same angle. Then the torque must be essentially the same as for a plain rotor.

 

The air separation between iron pertaining to different poles is essentially the same, only split in several steps; at least there is no drawback here. It means that the same maximum stator current can be used before the flux neglects the rotor's iron. Though, loss considerations may impose a lower limit to the stator current.

 

This would lead to the comparison: a small reluctance motor can be more compact than a squirrel cage thanks to lower rotor loss, and Vagati's design reduces the cogging of a reluctance motor.

----------

 

Improving...

 

The flux varies more quickly than one pole separation. It takes about one slit separation. Though, only a few slits contribute efficiently at any time, so this gain isn't huge.

 

Then, only one air separation prevents flux leaks. This is why Abb offers only small motors (<400kVA) of this design.

 

Designing the magnetic circuit so the flux varies quickly with the angle was already known, with pole shoes having several teeth each, but (1) it needed a higher drive frequency (2) cogging was still strong.

 

Though, how much torque it creates and how big the stator losses are isn't obvious to compare qualitatively with other designs.

[CasualKilla]

Maybe... I consider that the torque results from the flux variation per unit angle and from the stator current (summed over several phases, accounting properly the poles, and so on).

 

My understanding is that Vagati's rotor offers the same flux variation as the more classical rotor, narrower but plan, and that the change in flux or reluctance happens over the same angle. Then the torque must be essentially the same as for a plain rotor.

 

The air separation between iron pertaining to different poles is essentially the same, only split in several steps; at least there is no drawback here. It means that the same maximum stator current can be used before the flux neglects the rotor's iron. Though, loss considerations may impose a lower limit to the stator current.

 

This would lead to the comparison: a small reluctance motor can be more compact than a squirrel cage thanks to lower rotor loss, and Vagati's design reduces the cogging of a reluctance motor.

----------

 

Improving...

 

The flux varies more quickly than one pole separation. It takes about one slit separation. Though, only a few slits contribute efficiently at any time, so this gain isn't huge.

 

Then, only one air separation prevents flux leaks. This is why Abb offers only small motors (<400kVA) of this design.

 

Designing the magnetic circuit so the flux varies quickly with the angle was already known, with pole shoes having several teeth each, but (1) it needed a higher drive frequency (2) cogging was still strong.

 

Though, how much torque it creates and how big the stator losses are isn't obvious to compare qualitatively with other designs.

When I begin simulating I will create an optimized single carrier rotor and compare it with multi-carrier or Vagati design. That handbook seems to indicate I will be able to get 2/3 times the torque for the same frequency and voltage stator design, but if you say I should be skeptical, I will, a little skepticism never hurts.

 

Do you have any idea what the meaning is of the Iq and Id current from the dq transform. Could I visualize these as DC electromagnets position in the q and d directions? If yes, why are there 2 currents? I thought the stator could be visualized as a single rotating magnet, which would corresponds to one one current in d-axis.

Edited March 28, 2015 by CasualKilla

[CasualKilla]

So few months later now, and I still don't fully understand this type of motor, but I have gained sufficient understanding to create and optimize a line start synchronous reluctance machine (LS-SynRM) within simulation software. Here is my finalized optimised geometry if anyone is curious. Stimulation results show better performance (effeciency and power factor) than an inductance motor using the same stater and for the same load, so it seems like a good result.

 

Edited August 9, 2015 by CasualKilla

[Enthalpy]

Nice to read you again!

 

Where do your laminations hold together? Right below the aluminium bars? Usual designs have small bridges across the so-called flux barriers.

 

I haven't invested the necessary time in this motor. It can be that, at the limited stator current usual for kW motors, the induction can't cross the flux barriers, and then more barriers create more induction transitions and a bigger torque. I attempted it in the MW range years ago (for a slow wind turbine generator) but there, the full pole distance was needed to limit the flux leaks, so splitting this distance brought nothing.

 

Without conduction losses at the rotor, the efficiency may improve, and also, it eases the more difficult rotor cooling.

[CasualKilla]

Nice to read you again!

 

Where do your laminations hold together? Right below the aluminium bars? Usual designs have small bridges across the so-called flux barriers.

 

I haven't invested the necessary time in this motor. It can be that, at the limited stator current usual for kW motors, the induction can't cross the flux barriers, and then more barriers create more induction transitions and a bigger torque. I attempted it in the MW range years ago (for a slow wind turbine generator) but there, the full pole distance was needed to limit the flux leaks, so splitting this distance brought nothing.

 

Without conduction losses at the rotor, the efficiency may improve, and also, it eases the more difficult rotor cooling.

 

Hi Enthalpy!

 

Yes, I also have some concern about the mechanical strength of the rotor, but I think the squirrel cage will add some mechanical strength in the radial direction (normal designs do not have the squirrel cage), otherwise I hope there is enough time to make a second prototype. I will examine what performance drop is expected if I add some radial ribs.

 

This is effectively a synchronous motor and the induction bars are used purely to start the motor on a constant 3-phase voltage supply. So the induction bars are only active for the first 2 seconds as the motor accelerates to steady state. This means a motor controller is not needed to start-up the motor as with normal synchronous machines. The application of this motor would be constant operation at a constant speed and torque such as a fan.

Here is a comparison of the starting responses of my synchronous reluctance motor compared with a induction machine in the same stater and for the same line voltage. Note, the synchronous speed is 1500 rpm (50Hz 4 poles).

 

Edited August 11, 2015 by CasualKilla