Typing this in a cafe right at this very moment, so let’s jump right in: LUT aka Lappeenranta University of Technology (or perhaps it’s LUT University today) has come up with an interesting synchronous reluctance motor (SynRM). They have made a 200 kW, 10 000 rpm prototype that they claim approaches the efficiency of a permanent magnet machine. Now, they are working on getting the prototype installed on the electrified rear axle of a scientific testbed truck, from University of Oulu.
Now, let’s get the obvious, always-quoted pros of a SynRM out of the way, just in case someone without that much background in motors happens upon this post. The motor is obviously magnet-free, and rare-earth free. That may or may not save some costs – they have a nasty tendency of popping up elsewhere. It does very obviously eliminate a tricky supply chain, meaning that the latest fit in the oval office is less likely to mess up your lead time estimated big time. In today’s geopolitical climate, before the actual climate kills us all thank you very much for that, I would call that a huge win.
Now, the trick.
Simply put, the motor has a bridgeless rotor.
Boom. Drop mic.
Bridgeless-ness obviously eliminates a significant leakage flux path, or flux going into the wrong way path. It improves the 
ratio, and everything else via downstream effects.
How LUT actually got rid of the bridges seems like a really neat trick. Deceivingly simple, and makes me angry that I didn’t myself figure out something so obvious in retrospect. Like all great ideas, that is. Here goes.
They have a laminated rotor, with non-magnetic disks / solid sheets (or plates, depending on how you call it) interspersed between the electric steel sheets doing the actual magnetic stuff. The illustration on the news article suggests a rather thin support per multiple electrical sheets. Or the magnetic pole-units being significantly thicker in the axial dimension compared to the support sheets – the article doesn’t really comment if the magnetic parts are laminated or not.
Non-laminated magnetics would obviously be simpler from the mechanical point of view, but I fear the magnetics get very eddy-y very fast.
For laminated sheets, I see a few options. We could face-bond the laminations together, or even weld them. Whatever the joining method, it would have to be robust enough to survive the bending-stress-thingy (as you can see, I’ve amped up on my mechanical engineering skills) coming from the centrifugal forces.
Alternatively, or in addition, we could of course have non-magnetic supports running axially through the length of the rotor. Now that we have the in-plane non-magnetic sheets taking some or all of the centrifugal stress, they wouldn’t need to be nearly as stiff as what you’d need with the more-obvious supports-only-at-endplates structure.
SynRMs in general
SynRMs suck for (peak) torque. There, I said it. That’s my personal observation, not based on a thorough scientific study or design space exploration or anything, and I would love to be proven (politely and convincingly and based on actual evidence – take heed, some individuals who I won’t deign to name) wrong on that respect. But, what I have seen and done supports it, and the theory too.
Here’s the basic gist. Torque comes from the interaction of a surface current density (equivalent concept, we rarely have a true surface layer though copper-coated induction machines get rather close) and the radial magnetic flux density, sharing the wave number and phase with it, in time and space both. Everything else creates either oscillating torque, or no torque at all.
Also, any surface current density creates a corresponding flux density. As this flux density is proportional to the line integral around the periphery, the flux density is offset by 90 electrical degrees from it, and inversely proportional to the wavelength (or pole count) and the airgap length. Thus, this flux wave does not produce torque when it interacts with the current that spawned it in the first place.
SynRMs are in a nasty slot in this case. They have the rotor modulating the aforementioned flux, with its saliency, to generate that needed offset. But, since the flux is generated by the stator winding, they generally prefer modest pole counts and short airgaps.
What I’m getting at is this. The flux density is constrained due to saturation, so we need a high current density to reach a high peak torque. Except, the physics being physics, this current density does create an unwanted flux density component too – we can’t modulate the flux density to have that ideal 90-degree offset with the current density. I think – again I would love to be proven wrong. So, we have a parasitic non-torque-producing flux density component there, and a rather significant one at that since we do want that short airgap and modest pole-count for good nominal performance and power factor. This flux component is doing nothing other than driving up the overall flux density (amplitude) and killing the torque.
PM machines are in a much happier place in this respect. They can and often do have large magnetic airgaps, and can drive the pole count high without killing the power factor (though the frequencies and all associated bits of nastiness get up, too).
Induction machines can be surprisingly good, too. We can have a huge surface current density component in the stator, and an equally huge density in the rotor opposing it. Now, a tiny offset between those to generates a tolerable airgap flux density. This modest flux interacts with the huge surface current, creating a large torque.
Obviously, the main limitation is the fact that the offset is defined by factors we can’t control after the machine has been manufactured. Perhaps a non-salient synchronous machine would then be a good candidate for ultimate peak torque density (other than pulsed inductive systems), if we were willing to sacrifice everything else for it.
But I digress.
Summary
A cool idea. Wish I’d come up with myself. I hope they make it.
Check out EMDtool - Electric Motor Design toolbox for Matlab.
Need help with electric motor design or design software? Let's get in touch - satisfaction guaranteed!
