The motors are not really taking up a lot of useful space in the car and are not underpowered by any stretch so this tech needs to be price competitive. Like Hyundai said batteries are where we need to see advances.
Don't get me wrong it is interesting and I am glad to see advances but I don't see this as a terribly useful advance today.
I see it as a BIG advance for small drones. The same torque from smaller engine where every gram counts is a big win, you can replace that weight with more battery.
What counts there is power density, not torque density, isn't it? Quadcopters don't typically gear their motors down, instead preferring to pitch the rotors at shallow angles of attack.
So, better for racing drones, they will be able to change speed of propellers faster, they already use higher angles to readuce propeller speed (gyroscopic effects are killing maneuvring). Shallow angles are better for efficiency.
Getting rid of a gearbox that weighs 70kg seems like a significant weight saving. They also claim more efficiency which could result in smaller batteries or more range.
Well got to admit, the standard exaggerated imagery implies large battery savings where we really do not know how much it could imply and I bet it would be a big stretch to hit double digits.
They may save on reduction gearing but recently Porsche went the other way by adding in a two speed transmission so as to improve high speed efficiency. Electrics would tend towards requiring heavier duty transmissions common to diesels simply because of torque but a good application of multiple gears in a transmission could benefit EVs depending again, on application. We would need to see the amount of electricity being consumed at different load levels and compare the cost of a transmission to increased batteries as to which is best for it.
In practice, "battery savings" will translate to "we'll do more with the same battery life"; see also: smartphones who have gotten 1000x as fast over the past decade (or whatever) and still only manage a day of intense use on a battery.
The main use would be exactly what the article says - a higher output generator for when windmills operate at slow speeds. The car motor conceptualizations are purely there for click bait.
Cars are not where a motor like this will be applied. Exoskeletons, shoes, airplanes will all need lightweight, high power density, high torque, gear free motors.
Yes, they do and easy to screw up. Learn about first model of Tesla Roadster - it was quite speedy up to 160 km/h and then not speedy at all. All because of the wrong gearbox.
The Koenigsegg hybrid (so not strictly an EV, but can be operated as a pure EV) famously don't have a gearbox but a single, fixed reduction gear. There is a hydraulic coupling for the electrical motors to allow slow speeds. At higher speeds the coupling is mechanically locked to directly connect all engines to the wheels.
The Mitsubishi Outlander phev, which only directly powers the wheels with the ICE in specific modes, uses only a hydraulic clutch for that purpose. There might be a fixed gear reduction somewhere, but no true gearbox.
If by "EV" you mean "electric car", then they almost always do (and it's almost always a single fixed reduction, around 10:1).
But a lot of non-car electric vehicles (maybe the majority) use hub motors, where the wheel's angular velocity is the same as the motor's.
That's not the whole story, though, as a more illustrative number is the ratio between wheel radius and the motor's air gap radius, multiplied by whatever reduction gearing is used. I don't think there's a name for this concept, but you could borrow "gain ratio" from bicycles (https://sheldonbrown.com/gain.html).
For a hub motor the wheel pretty much has to be bigger than the motor for ground clearance (unless you're a monorail or something), so the gain ratio is less than one (maybe 0.5 for a skateboard and 0.25 for an ebike).
And practical wheel speeds tend to be a lot lower than even an inexpensive motor's top speed (not surprisingly, by about a 10:1 ratio).
Bottom line is a hub motor tends to be ~10 times larger than it would need to be on a peak torque basis, so fitting more torque into a smaller package (even at the expense of top speed) is a win. On the other hand solving the peak torque issue doesn't magically solve other issues that might crop up as you downsize (e.g. heat).
The key quote from the linked article seems to be "It’s essentially two concentric radial motors bookended by two axial ones."
My question is where's the cooling? You can put more magnetic field and torque into an ever smaller area, always true, but still going to have to dump heat out via some path. And the bigger the path for whatever is cooling, the lower the magnetic flux density and torque. So I'd be interested in hearing how the cooling system works. Its not like their copper is more conductive than everyone elses copper, LOL.
yetihehe suggests drone use, plenty of air movement there.
Some of the quotes in the article sound very aerospace. Everyone is dancing around the concept without saying it, that the main advantage of removing or downsizing the gear train is along the lines of aerospace style "simplicate and add lightness" which is often much more expensive than it sounds. For example, imagine an electric general aviation airplane, you can't have problems man-rating the electric gearbox if your electric Cessna-172 doesn't even HAVE a gearbox between the electric motor and prop. Likewise for EVs, complete elimination of the gearbox would seem to eliminate all gearbox related maintenance and repairs. Honestly, the car manufacturers might not be amused at the concept of selling fewer more reliable cars. (Edited to add, also see car racing entertainment, if you enter an EV in a popular automobile race, you can't drop out of the race due to gearbox failure if you have no gearbox to fail... is that good or bad for ratings?)
Unlike helicopters, where they are used to drive the tail rotor. Even though dual disk helicopters don't need a tail rotor, they still need a gearbox to balance torque between the disks.
The aerospace suggestion would also lend itself to battery-powered passenger flight, which is starting to creep into the realm of feasibility.
But the difficult bit is proving and scaling the tech. It looks like they're sensibly starting small, with e-scooters. I'm reminded of Mazda and the Wankel engine somehow; a "unique" topology that remained niche.
> Honestly, the car manufacturers might not be amused at the concept of selling fewer more reliable cars.
I wouldn't worry about it at all. One EVs become truly mainstream, car manufacturers will start removing redundant components and reducing the amount and quality of input materials, until the reliability curve again reaches the shape they desire.
Honestly, the car manufacturers might not be amused at the concept of selling fewer more reliable cars
Competition takes care of that problem. If all but one of the carmakers try to avoid selling “fewer more reliable cars” then the remaining carmaker can sell “way more more reliable cars.”
This is what happened when Honda and Toyota came along as the US carmakers were making very unreliable cars.
Competition does not "take care of the problem". There are counteracting forces that act through competition, and competition is inherently only partially restorative. It reduces incentives but does not eliminate them. Companies will only abandon early obsolensence to the point that it is profitable, which depends almost only on the customers naivete.
Worse, that viewpoint ignores the obvious and numerous counterexamples... Everything from apple (semi-reasonably but certainly in their own interest) killing batteries in older phones to outright conspiracy like the lightbulb mafia.
> This is what happened when Honda and Toyota came along as the US carmakers were making very unreliable cars.
You will note that US automakers are still some of the least reliable in the world. Its also ridiculous to suggest that japanese cars outcompete because of reliability. It was cost. It was also power, size, and quality. In fact when the civic took off in the US it was still not that reliable. Toyota and honda didnt get their current reputation until the mid 90s.
> Furthermore, by using discrete rectangular coils inset into the stator poles, the HET needs 30% less copper than a motor of similar size. The design also eliminates end windings – lengths of copper that lie outside the stator in a typical motor, generating wasted magnetic field and heat.
So it sounds like less heat is part of the secret sauce.
Yep, I've also read the linked article and for me the summary is more like "HET motor has some interesting ideas, some difficulties, some claims are incorrect and it's probably a bit difficult to manufacture."
Interesting ideas- kind of, but the interesting ideas are not new. Dimensionality is one of the oldest formulated motor-related problems. Hybrid rotors are certainly an active area. The motor is a pretty novel and clever way to draw all these things together, but it's not the kind of thing to build a product around.
Some difficulties- these are fundamental issues. They are fundamentally unsolvable with the suggested architecture. Eg anything that adds cooling back in will conflict with the rotors or coils.
On manufacturability:
> Flux paths in the teeth seem very much 3-dimensional, which would require sintered materials to be used to limit eddy-current losses.
This really belies how much of a fucking nightmare this would be.
First, an unrelated area of development: grain-oriented electrical steels have 30% higher permeability (in one direction) than non-oriented steels. People are actively looking to ways to orient steel grains into the arches necessary to make a motor, because even small improvements in permeability can lead to smaller and more efficient motors. A 30% improvement would be incredible, and justify a large cost increase since the cost of motors is effectively determined by the electricity usage.
The permeability of sintered soft materials is at best ~1/10th that of the electrical steel used in motors. So... thats a huge issue. Its also much more expensive. Its also mechanically unsuitable for high torque motors, because it's weak and brittle. It also has much higher losses.
The shape of the stator would also require entirely new technologies to even make. Die-sintered powders are very precise in two axes- since you're pressing into a mould, you can rely on near-thousandth repeatable tolerances once you have accounted for shrinkage after sintering (which can take multiple tries to get right). That third axis is a real bitch. Powder will never be consistent; you cant apply hundreds of tonnes of pressing force and be precise to thousandths of an inch, and even if you can get the force repeatable the compressability of powders varies too much between batches due to grain shape and size, temperature, additives, humidity... And even if it was consistent, you cant design for that. You can only adjust a ten thousand dollar mould after the fact.
And that z dimension is the most critical, since any asymmetry will cause uneven forces on the axle bearings. In a normal motor you can trivially be off by several hundredths of an inch because the axial force is negligible. In this design a thou will cause a large force imbalance, so each motor will need to be a precision part. Even variations inside the rotor will cause large vibrations, so sintering + heat treating will be a nightmare. Not to mention that the magnets will have to be matched individually, both for size/thickness and permeability tolerance.
If my boss wanted me to produce this thing, I'd quit.
The article is diplomatic, but not ambiguous. For instance:
> Cooling issues: can enough cooling air be forced inside the rotors?
> Cooling issues, again. The primary cooling path for more traditional motors is usually conduction through the stator yoke, to the frame, to either air or coolant fluid. For the HET, this path seems almost completely missing.
> [four lines later] Assuming the cooling performance remains unchanged, the reduced winding losses could then be cashed in by increasing the current density, and thus torque, torque-per-mass, and bringing losses back to the original level.
This may read as if there are good ideas here, but make no mistake- the implication is that these problems are nigh-unworkable. The possibilities of minor increases are locked behind immense problems.
There is a parallel with combustion engines here. The ideal combustion cylinder is spherical: it ensures more even burning and minimizes the area heat from escape from a given volume. You could make an engine with spherical combustors instead of cylindrical and reap that improvement. That's basically what toroidal motors try to do- use more surface area per unit of rotor volume (although due to the nature of air gaps[not discussed in the article], it's not obvious that this is actually beneficial). In both cases, you're cutting off your nose to spite your face.
Finally got the time to read through that link. Thanks!
The difficulty I had with that article is that I'm only familiar with how electric motors work at a high level. (I'm a software engineer.) An electrical engineer, or someone who's spent more time than me studying electric motors will be able to understand your link much better than I can understand it.
Thus, it's difficult for me to really understand the author's tone. Is it an honest evaluation of a promising technology where the author is pointing out lots of potential limitations? Is the author just plain biased? I just don't know enough about the field to make such a judgement. The conclusion, at the end, though, implies that HETs are very promising, but not ready for mainstream usage.
Anyway, for me, the questions basically come down to: How much will an HET car cost to compared to the "traditional" electric motor + reduction gear + differential + inverter + extra batteries. (Apologies if I got a term wrong.) Furthermore, is there a niche where the HET is going to be useful, like performance vehicles, or vehicles where the extra traction of dual motors is worth the extra cost, like pickup trucks?
Assuming the cost is a "slam dunk," or the extra cost is worth it for some niche, this is very promising indeed.
Reminds me of another "revolutionary new motor design" from a few years ago that promised greater efficiency than any other motor ever built. That one had an out-of-context praise blurb from an MIT professor, which is clearly better than an out-of-context praise blurb from a UT professor. At best, this is a solution in search of a problem. And until it's tested and certified by an independent party, it's not even that. The whole company reeks of dumb-money snake oil and I'm putting this in the same category as Juicero.
Not all new motor design promises are trash though.
The brushless DC outrunner motor, which used to be a niche design due to the complexity of the control circuits and expense of permanent magnets now seems to be used in nearly everything new. Most electric cars are moving towards it, electric bikes, hoverboards and escooters, and it's used in every drone, some electric powertools, etc.
I think it will be in next-gen trains, and will become the 'standard' motor to replace current uses of AC synchronous motors in factories, mostly due to its ability to brake and regen power when stopping, meaning it gets less hot.
So I was wondering whether "Hunstable" is some new technical term like "h-unstable", or some kind of wordplay on "unstable". Turns out it's just the surname of the creators.
47 comments
[ 3.3 ms ] story [ 108 ms ] threadDon't get me wrong it is interesting and I am glad to see advances but I don't see this as a terribly useful advance today.
They may save on reduction gearing but recently Porsche went the other way by adding in a two speed transmission so as to improve high speed efficiency. Electrics would tend towards requiring heavier duty transmissions common to diesels simply because of torque but a good application of multiple gears in a transmission could benefit EVs depending again, on application. We would need to see the amount of electricity being consumed at different load levels and compare the cost of a transmission to increased batteries as to which is best for it.
Personally, the less parts the better.
https://en.wikipedia.org/wiki/Tesla_Roadster_(2008)#Transmis...
The Koenigsegg hybrid (so not strictly an EV, but can be operated as a pure EV) famously don't have a gearbox but a single, fixed reduction gear. There is a hydraulic coupling for the electrical motors to allow slow speeds. At higher speeds the coupling is mechanically locked to directly connect all engines to the wheels.
https://en.wikipedia.org/wiki/Koenigsegg_Regera
https://www.koenigsegg.com/car/regera/
As mentioned, Taycan and the preprod Roadster are exceptions, and those are also the two exceptions I know.
But a lot of non-car electric vehicles (maybe the majority) use hub motors, where the wheel's angular velocity is the same as the motor's.
That's not the whole story, though, as a more illustrative number is the ratio between wheel radius and the motor's air gap radius, multiplied by whatever reduction gearing is used. I don't think there's a name for this concept, but you could borrow "gain ratio" from bicycles (https://sheldonbrown.com/gain.html).
For a hub motor the wheel pretty much has to be bigger than the motor for ground clearance (unless you're a monorail or something), so the gain ratio is less than one (maybe 0.5 for a skateboard and 0.25 for an ebike).
And practical wheel speeds tend to be a lot lower than even an inexpensive motor's top speed (not surprisingly, by about a 10:1 ratio).
Bottom line is a hub motor tends to be ~10 times larger than it would need to be on a peak torque basis, so fitting more torque into a smaller package (even at the expense of top speed) is a win. On the other hand solving the peak torque issue doesn't magically solve other issues that might crop up as you downsize (e.g. heat).
My question is where's the cooling? You can put more magnetic field and torque into an ever smaller area, always true, but still going to have to dump heat out via some path. And the bigger the path for whatever is cooling, the lower the magnetic flux density and torque. So I'd be interested in hearing how the cooling system works. Its not like their copper is more conductive than everyone elses copper, LOL.
yetihehe suggests drone use, plenty of air movement there.
Some of the quotes in the article sound very aerospace. Everyone is dancing around the concept without saying it, that the main advantage of removing or downsizing the gear train is along the lines of aerospace style "simplicate and add lightness" which is often much more expensive than it sounds. For example, imagine an electric general aviation airplane, you can't have problems man-rating the electric gearbox if your electric Cessna-172 doesn't even HAVE a gearbox between the electric motor and prop. Likewise for EVs, complete elimination of the gearbox would seem to eliminate all gearbox related maintenance and repairs. Honestly, the car manufacturers might not be amused at the concept of selling fewer more reliable cars. (Edited to add, also see car racing entertainment, if you enter an EV in a popular automobile race, you can't drop out of the race due to gearbox failure if you have no gearbox to fail... is that good or bad for ratings?)
https://en.wikipedia.org/wiki/Geared_turbofan
But the difficult bit is proving and scaling the tech. It looks like they're sensibly starting small, with e-scooters. I'm reminded of Mazda and the Wankel engine somehow; a "unique" topology that remained niche.
I wouldn't worry about it at all. One EVs become truly mainstream, car manufacturers will start removing redundant components and reducing the amount and quality of input materials, until the reliability curve again reaches the shape they desire.
Competition takes care of that problem. If all but one of the carmakers try to avoid selling “fewer more reliable cars” then the remaining carmaker can sell “way more more reliable cars.”
This is what happened when Honda and Toyota came along as the US carmakers were making very unreliable cars.
Worse, that viewpoint ignores the obvious and numerous counterexamples... Everything from apple (semi-reasonably but certainly in their own interest) killing batteries in older phones to outright conspiracy like the lightbulb mafia.
> This is what happened when Honda and Toyota came along as the US carmakers were making very unreliable cars.
You will note that US automakers are still some of the least reliable in the world. Its also ridiculous to suggest that japanese cars outcompete because of reliability. It was cost. It was also power, size, and quality. In fact when the civic took off in the US it was still not that reliable. Toyota and honda didnt get their current reputation until the mid 90s.
> Furthermore, by using discrete rectangular coils inset into the stator poles, the HET needs 30% less copper than a motor of similar size. The design also eliminates end windings – lengths of copper that lie outside the stator in a typical motor, generating wasted magnetic field and heat.
So it sounds like less heat is part of the secret sauce.
Some difficulties- these are fundamental issues. They are fundamentally unsolvable with the suggested architecture. Eg anything that adds cooling back in will conflict with the rotors or coils.
On manufacturability:
> Flux paths in the teeth seem very much 3-dimensional, which would require sintered materials to be used to limit eddy-current losses.
This really belies how much of a fucking nightmare this would be.
First, an unrelated area of development: grain-oriented electrical steels have 30% higher permeability (in one direction) than non-oriented steels. People are actively looking to ways to orient steel grains into the arches necessary to make a motor, because even small improvements in permeability can lead to smaller and more efficient motors. A 30% improvement would be incredible, and justify a large cost increase since the cost of motors is effectively determined by the electricity usage.
The permeability of sintered soft materials is at best ~1/10th that of the electrical steel used in motors. So... thats a huge issue. Its also much more expensive. Its also mechanically unsuitable for high torque motors, because it's weak and brittle. It also has much higher losses.
The shape of the stator would also require entirely new technologies to even make. Die-sintered powders are very precise in two axes- since you're pressing into a mould, you can rely on near-thousandth repeatable tolerances once you have accounted for shrinkage after sintering (which can take multiple tries to get right). That third axis is a real bitch. Powder will never be consistent; you cant apply hundreds of tonnes of pressing force and be precise to thousandths of an inch, and even if you can get the force repeatable the compressability of powders varies too much between batches due to grain shape and size, temperature, additives, humidity... And even if it was consistent, you cant design for that. You can only adjust a ten thousand dollar mould after the fact.
And that z dimension is the most critical, since any asymmetry will cause uneven forces on the axle bearings. In a normal motor you can trivially be off by several hundredths of an inch because the axial force is negligible. In this design a thou will cause a large force imbalance, so each motor will need to be a precision part. Even variations inside the rotor will cause large vibrations, so sintering + heat treating will be a nightmare. Not to mention that the magnets will have to be matched individually, both for size/thickness and permeability tolerance.
If my boss wanted me to produce this thing, I'd quit.
> Cooling issues: can enough cooling air be forced inside the rotors?
> Cooling issues, again. The primary cooling path for more traditional motors is usually conduction through the stator yoke, to the frame, to either air or coolant fluid. For the HET, this path seems almost completely missing.
> [four lines later] Assuming the cooling performance remains unchanged, the reduced winding losses could then be cashed in by increasing the current density, and thus torque, torque-per-mass, and bringing losses back to the original level.
This may read as if there are good ideas here, but make no mistake- the implication is that these problems are nigh-unworkable. The possibilities of minor increases are locked behind immense problems.
There is a parallel with combustion engines here. The ideal combustion cylinder is spherical: it ensures more even burning and minimizes the area heat from escape from a given volume. You could make an engine with spherical combustors instead of cylindrical and reap that improvement. That's basically what toroidal motors try to do- use more surface area per unit of rotor volume (although due to the nature of air gaps[not discussed in the article], it's not obvious that this is actually beneficial). In both cases, you're cutting off your nose to spite your face.
The difficulty I had with that article is that I'm only familiar with how electric motors work at a high level. (I'm a software engineer.) An electrical engineer, or someone who's spent more time than me studying electric motors will be able to understand your link much better than I can understand it.
Thus, it's difficult for me to really understand the author's tone. Is it an honest evaluation of a promising technology where the author is pointing out lots of potential limitations? Is the author just plain biased? I just don't know enough about the field to make such a judgement. The conclusion, at the end, though, implies that HETs are very promising, but not ready for mainstream usage.
Anyway, for me, the questions basically come down to: How much will an HET car cost to compared to the "traditional" electric motor + reduction gear + differential + inverter + extra batteries. (Apologies if I got a term wrong.) Furthermore, is there a niche where the HET is going to be useful, like performance vehicles, or vehicles where the extra traction of dual motors is worth the extra cost, like pickup trucks?
Assuming the cost is a "slam dunk," or the extra cost is worth it for some niche, this is very promising indeed.
Until then total snake oil crying for attention.
The brushless DC outrunner motor, which used to be a niche design due to the complexity of the control circuits and expense of permanent magnets now seems to be used in nearly everything new. Most electric cars are moving towards it, electric bikes, hoverboards and escooters, and it's used in every drone, some electric powertools, etc.
I think it will be in next-gen trains, and will become the 'standard' motor to replace current uses of AC synchronous motors in factories, mostly due to its ability to brake and regen power when stopping, meaning it gets less hot.