Hacker News new | past | comments | ask | show | jobs | submit login
Belgian startup Magnax has found a way to mass-produce the axial-flux motor (ieee.org)
378 points by sohkamyung 13 days ago | hide | past | web | favorite | 120 comments





I am very eager to find out what the price will be for these motors, this will have a huge impact in General Aviation. A banal 4 cylinder 100 HP petrol engine is ~$20,000 ("aviation price") and the weight is a bit over 60 kg, if my math is right a Magnax motor can be as light as 5 kg (even 10 kg is fabulous) but the cost needs to balance the huge price of batteries: the $20,000 petrol engine with 140 liter fuel tanks gives a flight time of almost 10 hours, batteries for that flight time would be insanely expensive and possibly quite heavy. But most people don't fly more than 5-6 hours, so a reduction to 6 hours would be acceptable for 95% of the people.

Another positive impact to GA is you don't need to adjust for mass and balance that much, batteries have a fixed center of gravity that does not change while discharging, while fuel tanks impact the plane's center of gravity as they go from full to empty. Less trimming, simpler flying.

And not the last, it is eliminating any problems with carburetors, mixture, air density or inverted flight: none of these affect an electrical motor, but it needs to be considered for the likes of Cessna where you need to adjust the (fuel/air) mixture depending on the altitude.

And it improves propeller efficiency, simplifying constant speed propeller operation. It can be easily fully automated.

All good news, but seeing it happen is a different story: it can take 3 to 5 years, more that a decade or never to reach mass availability and adoption in general aviation. It has the potential to make the market explode (in the positive way), but there is so much politics involved there are chances the status quo will be maintained. Let's see.


I’m absolutely ignorant on the topic of aircraft propulsion, but I’d always assumed that the diminishing mass of fuel that one was carrying around with oneself as the flight progresses is a key advantage that electric-propulsion aircraft can never hope to replicate (short of dropping battery packs with parachutes as they become exhausted of charge).

Am I absolutely wrong or is it not actually much of a factor?


You are right in the way you are thinking, but you don't have the full information: in a 600 kg plane the fuel is 100kg. The mass reduction does not make such a big impact, especially as you rarely top the tanks (you can carry more luggage or a heavier pilot & passenger) and you never empty it (30 minute reserve is usually raised to 1 hour reserve), so you have a 10% variation of the total mass of the plane on regular flights: no significant impact.

On the other side, large planes carry a lot more fuel as a percentage of the total mass, it really matters to them.


I hope to hear actual pilots chiming in, but from first principles:

It shouldn't be a large factor. Drag is independent of mass, and while lifting the fuel to cruising altitude takes energy, most of the flight should ideally be cruising.

The rocket equation doesn't apply; that's for an entirely different regime.


While drag is "independent" of mass, the lift required to maintain level flight is affected by weight, and drag is affected by lift. A heavier airplane will absolutely require more power to cruise.

If as you say most people fly for 5-6 hours wouldn’t we need the 10 hour range just to be safe? A 6 hour range when most people fly 5-6 hours is dangerously cutting it close.

They might need to reduce their flying time slightly, but you don't need hours and hours of reserve power unless you're flying long distances over the ocean (which requires specially certified equipment anyway). The FAA only requires 30 minutes of reserve fuel, or 45 minutes if flying at night. That is almost certainly enough to get you to the nearest usable airfield.

Also, bear in mind that it's much easier to accurately and reliably measure a battery's state of charge than it is to measure how many gallons of fuel are in an airplane's wing.


I agree that electric airplanes is a highly desirable goal.

I don't, however, get the point about propeller efficiency, as is it not about matching the pitch to the airspeed? Fixed-pitch propellers have a maximum efficiency at one particular airspeed (and maximum thrust at, in general, a different particular airspeed), regardless of what sort of motor is turning them.


Constant speed propellers are more efficient than fixed pitch; in order to keep the propeller rpm in the optimal point you have to adjust the engine gas control and the propeller pitch, not extremely complicated but not simple enough to be allowed for light sport planes - it is considered too complex for that.

If you play with the propeller pitch, the force on the engine will make it change the rpm; if you change the engine rpm to the optimal of your propeller, you don't always have the best pitch, so you adjust it and this will impact the rpm. With an electrical motor the power delivery is much more robust, so you can adjust the pitch and then set the motor to the desired rpm and that's it.


> you can adjust the pitch and then set the motor to the desired rpm and that's it.

Normal operation of a constant-speed propeller involves setting the desired rpm via the pitch control, and then adjusting the power, via the throttle, according to your purpose (such as level cruise at a specific IAS, or climb at maximum sustainable power), with the propeller govenor maintaining the chosen rpm via pitch changes, and I do not see how electric power changes this relationship. I understand that electric power might simplify making a sort of FADEC for electric airplanes (so that there is only one power control), and that constant-speed propellers could eke out more duration from the batteries, but now you are adding the cost and complexity of a constant-speed prop. I am not sure that it would be justified in a light airplane having a cruise speed not much higher than its best climb speed.


Yes, for some light airplanes a constant speed propeller is not required, but if it is simple enough and you get sizable benefits (even 10% more range), then why not?

There are many planes in Europe that have the cruise speed much higher than the best climb speed. The best climb speed is around 100-120 km/h for light planes, while several plane models have the cruise speed over 200 km/h, up to 250 km/h. I am not aware of any plane with cruise speed below 200 km/h to use constant speed propellers, they have fixed pitch propellers, but the faster ones have in-flight adjustable pitch. I personally flew 7-8 different models of planes, only one with adjustable pitch, but they were all in the LSA class and const-speed is not used there.


There's a French startup that makes some pretty cool turbine range extenders for the general aviation market. Great power to weight ratios, and they come with the maintenance benefits of microturbines, which only have a single moving part. I can't remember exactly, and they don't have it in their specs, but I think they even use gasfoil bearings which means there is no need for oil lubrication.

https://www.turbotech-aero.com/index.html


Except right at this time, it will be a reduction to about 1 hour actually...

100 kg of fuel and 50 kg less on the engine weight allows for a lot of batteries. Not sure what is the flight time, but if someone can help with energy density we can calculate the exact number. I wildly guess it is more than 1 hours, but definitely less than 6 hours.

Later edit: I did some calculations, the result is less than 30 minutes.


Not as bad because of the higher efficiency of electric engine. But yeah, 1 hour or so.

One hour is quite enough for taxi-type and drone delivery.

Also, you could use structural batteries, which would substitute for part of the plane structure.


Don't forget about the people that lives near airports. They will love the silent engines. Maybe the areas will even live through a real estate boom.

Anyone else just impressed by the writing of the article itself? They made a very good effort of providing benefits over traditional radial flux electric motors and how theirs was difference as well as measurable improvement numbers.

I hope they succeed.


That's how IEEE Spectrum articles generally are: the actual inventors, not journalists, are doing some of the bigger articles.

Anyone can subscribe to the paper version and it's $40 for 12 issues international (or $20 in US): https://ieee.omeda.com/ieee/r-paid.do?p=WEBPAID


Actually, since you ask, while the subject matter is fascinating I had the opposite reaction and found the writing style a bit distracting. While the author clearly has a command of the field, I found myself needing to re-read several paragraphs because I couldn't differentiate when the author was referring to traditional EM designs and Magnox' new design. I also had a tough time maintaining a mental map of the design configurations he was comparing. More figures would've helped. There was one particular use of "nevertheless" that really through me off. Maybe it's just a difference in how we read (i.e. a me problem).

They are obviously making huge strides for a hardware startup if they're already field testing, so writing style opinion differences aside, mucho respect and I hope they succeed as well.


It reads like an ad to me. It's hard to figure out what the catch is, if any (and in this world of tradeoffs, there usually is a catch).

The catch is that radial motors are balanced magnetically and these are not. All that magnetic force is added to axial pull between the stator and rotor. That means they need a thrust bearing in addition to the guide bearings. This causes increased heat and mechanical losses and probably limits efficiency to somewhat less than a standard electric motor. Hi-E standard motors cap out around 98.8%, these will probably cap a point or two lower.

I don’t have a firm grasp of the article, but isn’t the force balanced in this case since the two rotors, linked together, sandwich the stator? If I understood correctly, this would mean the thrust bearing would only feel the imbalance between the two sides.

See: https://m.youtube.com/watch?feature=youtu.be&v=BTuCshX5bc0&t...


In general, yes, though the article mentions one in-hub motor with a single rotor, for reduced size or weight. I guess an in-hub motor, which also acts as the wheel's axle, has to deal with much larger axial forces, from cornering, than those generated by the motor, so the latter are moot.

Also, during operation, is not each magnet in the motor being simultaneously pulled by the pole it is moving towards, and pushed by the pole it is moving away from? If so then, at any given point in the rotation, the axial force would be the net of the two effects.


Agreed, this seems to be an incorrect conclusion. Maybe ew6082 just looked at the asymmetrical figure without reading the text.

Fair enough. They said that a double rotor version has been developed, but all their prototypes and marketing material are single rotor. We can probably assume that single rotor is all that would fit on an in-wheel design but there are other ways to skin the cat.

They would still need to include a light thrust bearing on the double rotor design since you can never assume perfectly balanced air gap and one side or the other will dominate, but the losses would be much lower with that setup.


For vehicles (cars, planes, etc.) if they can mass-produce at 97-98% efficiency and the power-to-weight ratio they are claiming to achieve (15 kW/kg) then it's still a huge improvement.

You will find on enquiry, as I have, that under continuous load the power density is a LOT lower.

They're claiming higher efficiency because of less iron loss. Does it balance out?


"ABSTRACT: The application of permanent magnet machines as direct-drive generators for wind turbines is considered. The Axial-Flux Permanent-Magnet (AFPM) machine with slots is compared to the Radial-Flux Permanent-Magnet (RFPM) machine, for this application. Using a computer-based design model, the two machine types are optimized with respect to the lowest Cost/Torque. Design optimization for maximum Torque/Volume is also investigated. The calculations show that the RFPM machine has lower Cost/Torque than the AFPM machine with slots, and lower torque/volume than the AFPM machine with slots.

...

Conclusion... ...Based on the results presented herein, the RFPM machine still appears as a better choice for a direct-drive generator, than the AFPM machine with slots"


Oh, it's definitely an ad. But it's the good kind of ad, the one that informs you about the existence and advantages of a product that you hadn't heard about. Contrast this throwback style of ad with the modern style of ad, which emotionally engineers you into subconsciously wanting a product that you don't actually need. I would love it if more ads looked like Magnax's.

Yes, you could call it an ad, but it reads as much more down-to-earth and straightforward than what most people think of when they think of advertising. They're simply describing what they built, why it was difficult, and why they think it's important, with a minimum of marketing fluff.

One thing I remember is some friends I met with a warehouse full of crazy projects they had worked on. They had a few axial flux motors they were really excited about that they had custom made. They seemed to think it was a superior technology, though I can’t recall the specifics.

Heat dissipation.

Bingo... Of course there's plenty of apps where that doesn't matter... we'll see them implemented in "bursty" automobiles a long time before we see them implemented in "continuous" railroad traction motors, for example.

Its an interesting time to be alive. I'm not THAT old and when I was a kid a generic NEMA 50 HP industrial electric motor took a crane or at least a large engine hoist to lift, more than 500 pounds for sure, but now there are high end automotive motors that I can easily do bicep curls with (like 40 pounds)

VFDs are kinda like that too. What used to be a large rack is now a little tiny DIN unit that handles 10x the power.


This is very interesting. And it is very good news that companies are finding higher efficiencies in mass produce-able electric motors. It means that electric cars will become more affordable and practical quicker.

But I do hope they do not proceed with the in wheel concept. In wheel motors make the ride not only bumpier but more dangerous. This is because the wheel with more un-sprung weight bounces more and tends to lose contact with the road more, which in turn makes the car more likely to spin out of control, miss a turn, etc.

Electric cars, being still relatively new, cannot afford to get a reputation for danger.


Don't worry, hub motors are so completely bad in highway-capable vehicles that even the most oblivious person would notice while on a test drive. They won't get a chance to sell.

Hub motors make excellent sense in indoor forklifts, which are often electric anyway. The reduced weight is not so important there, because the forklift body needs to weigh enough, but moving to the wheels means there is more room for batteries. Size matters, in forklifts, which need to maneuver in tight spaces.

Trucks and trains are often not sprung, and can benefit from lighter, smaller,more efficient hub motors. Cooling is easier when the whole motor is moving through the air.


"The reduced weight is not so important there, because the forklift body needs to weigh enough"

More than the body needs to weigh enough, you need a proper center of gravity via weight distribution or you're simply going to tilt the thing over and wreck everything when you lift an object (source: have multiple sit/stand/clamp lift certifications.)

"Size matters, in forklifts, which need to maneuver in tight spaces."

That's why we made the reach forklift, which is operated while standing up. Half the footprint of a traditional forklift with the ability to reach out across a shelf to pick up a pallet.

Where this will make even more sense is for driving conveyor belts on assembly lines. You'll save more power there in a manufacturing warehouse environment than you typically will switching your entire forklift fleet over to these new motors.


This is a very good comment

While an expensive option, if the advantages of an in-wheel motor were sufficient for a given application, active/magnetorheological suspension could largely compensate for it.

http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.678...

https://www.clearmotion.com/

https://en.wikipedia.org/wiki/MagneRide

https://www.youtube.com/watch?v=hiaMpG4Y640

Further, on newer vehicles with the trend of 20inch+ rims, that is already 60-90+lbs (27-40+kg) of unsprung weight per wheel with tire. There is certainly room there for improvement to offset at least some of the motor weight.

Would have to know how much their in-wheel models would actually weigh though to determine beyond speculation if it could be reasonably compensated for in a highway vehicle.


Might work out if it’s paired with active suspension.

The idea with motors at the wheel seems to be that the moving parts are all concentrated there. No mechanical distribution. Compartmentalization. But that is not incompatible with active suspension systems.


What would be the advantage of doing it that way instead of just putting the motor on the car frame and connecting it to the wheel with a CV axle? From what I could dig up, mechanical losses in a CV axle are on the order of 1%.

An active suspension requires power, too. And a heavier wheel would presumably require a more power-hungry active suspension than a lighter wheel, so you're not getting something for nothing in terms of efficiency.

I guess it is a matter of numbers, though. A CV axle also gives a wheel inertia, as it has mass and moves when the wheel does (partially). If you can make an insanely light motor, maybe you could make it so light that it's actually contributes less inertia than a CV axle would, in which case go ahead.


The small diameter of the CV axle makes the inertia small compared to the rest of the wheel, tire, brake assembly.

Doesn't active suspension need to be powered itself? I skimmed the article, and the main advantage they cite is more efficient energy use. That may be gone if you have to spend extra energy elsewhere.

maybe motor-wheels are bad for normal cars but maybe it's still a neat devices for simple setups, carts, or whatever rover that doesn't go super fast.

They said they've achieved 15kW/kg, so a powertrain of 120kW(~160BHP) would add around 2kg to each wheel - that's as much as two different rims can differ in weight.

>But I do hope they do not proceed with the in wheel concept. In wheel motors make the ride not only bumpier but more dangerous. This is because the wheel with more un-sprung weight bounces more and tends to lose contact with the road more, which in turn makes the car more likely to spin out of control, miss a turn, etc.

>Electric cars, being still relatively new, cannot afford to get a reputation for danger.

I'm really sick of this "anything that measurably effects handling in a negative way is a non-starter because safety" attitude that's become increasingly pervasive. The fact of the matter is that there is a range of handling that is "acceptable" and anything in it is fair game. Nobody is expecting a Dodge Journey or a Spark EV to handle like a Porsche. At 15kwh/kg these motors will be fine for vehicles with average power to weight ratios that do not place a massive emphasis on performance (i.e all the "appliance" vehicles that purists hate). For low power levels these motors may weigh on the same order as the parts you need to transmit power to the wheels in a conventional applications so there might not even be any net gain in some applications.


If their numbers bear out then 120kW would only add 8kg of unsprung weight. (160hp 18lbs)

That's not an 'only'. Unsprung weight is very precious.

There is obviously a threshold, that where the unsprung part of the transmission is heavier than the motor itself. That's why brakes are in the wheel and not inboard working through the axle (which would have to be massively beefed up to handle the enormous braking power). Do they reach that threshold? Probably not, but it's worth noting that they are getting closer and that as a consequence application-specific preference shifts.

Perhaps the an in hub motor could be integrated with braking and a KERS system.

Maximising regenerative braking is always a goal with BEV.

If there was a design tradeoff that allowed to somehow massively sacrifice efficiency for very low low specific mass and extreme heat tolerance, we might see the discs replaced with in-wheel "harvesting brakes" in addition to inboard high efficiency cruise motors.

But #1 I doubt that kind of tradeoff exists, at least not in sufficient magnitude, and #2 it would probably be quite uneconomic to exploit, lots of lower hanging fruits to pick first.


Off topic, but there are a few cars with inboard brakes. The Hummer H1 is one.

Because it has portals at the knuckle and finding a way to put brakes on that without massive rims would result in some janky packaging compromise. Also with the vehicle weight, wheel and tire size they run having the brakes benefit from the portal reduction sure doesn't hurt.

Then why did wheels grow from 15' to 21' on most cars in the past 25 years? I'm pretty sure that hurts unsprung weight significantly...

Bigger wheels reduce the need for suspension to keep the wheel in contact with the road, so there may be a tradeoff that makes sense being made in this case.

It's pretty much a wash. The reason wheels are bigger is for better handling as well as space for larger disc brakes. The combined total weight of brake disc, tire, hub and rim is roughly a constant due to the use of lighter materials.

It's not a wash. A common cast aluminum 20-inch wheel weighs a whole lot more than a common cast 16-inch wheel. Also, only performance vehicles get fitted with larger brakes. Many vehicles have discs that would fit fine under a much smaller wheel because they are designed with the smaller wheel as standard, and the larger wheel as an option.

Also, a larger rotor weighs more than a smaller one.

Combined, a large wheel and tire will require more energy to move and to stop, and will add as 75 pounds or more to the weight of the vehicle. This has negative impact on handling and fuel economy.

The only reason why the manufacturers add a large wheel option is because people want them.


The alternative to a cast alloy wheel is not another smaller cast alloy wheel but a steel one.

Larger wheel vs larger rim is not the same thing. A larger wheel has a larger weight but smaller rolling resistance. A larger rim on the same sized wheel has better handling characteristics and more room for larger discs, whether it is on a performance vehicle or a regular one doesn't really matter, plenty of mid-sized sedans of today are yesterdays performance vehicles.

There are outside exceptions (any SUV, ridiculously sized wheels on many American cars and people driving trucks when a normal passenger vehicle would be more appropriate). Those do have worse handling and fuel economy, but they're the exception, not the rule.


I'm going to defend my "only" as at 2kg per wheel it's about 10% of the weight of tires and wheels (presumably disc brakes are also unsprung. so total increase would be less than 10%) of a (stock) BMW 3 series with 16" aluminum alloy wheels[1].

[edit]

Disc brakes add about 9.5kg on the front[2] (maybe half that on the rear?), so a single disc brake rotor weighs more than the entire weight of all 4 motors.

Obviously this is using Magnax's numbers for very large motors and assuming they scale down, which is silly, but clearly there is some weight at which it would make sense, particularly for a non-high-performance vehicle (1 motor per wheel has lots of advantages for handling, but AFAICT this is not unique to in-wheel motors).

1: https://www.tirerack.com/wheels/tech/techpage.jsp?techid=108

2: http://www.dbausa.com/brake-discs-is-lighter-better/


I remember this being mentioned in car related articles once in a while, would love to learn more. Any sources you can recommend?

I got the impression that the wheel motors were just for testing. They'll leave the car making to someone else for sure.

Electric cars are pretty heavy thanks to the batteries, so it's not like having a bit more weight in the wheels is going to drastically change the unsprung:sprung weight ratio for the worse.

In a light vehicle, it's very important to minimize unsprung weight for the reasons you've listed. Not so much with a heavy vehicle.


If they're right that it's much lighter than an equivalent motor, maybe in-hub motors wouldn't be all that bad.

Watching the video that was linked in another thread, it made me think that maybe the hub could be stationary, and the tire just spins around the outside. That could cut down a lot on wind resistance.


"Electric cars, being still relatively new, cannot afford to get a reputation for danger."

The Lohner-Porsche was a hit at the World's Fair in 1900 and was in development before that. I mean I guess a concept 120 years old is still new enough to get a bad reputation...


> "If yokeless axial-flux machines replaced only a fraction of the older machines, we would save our customers some money and make the planet more livable while we’re at it."

The engineering feat is impressive and exciting, but this conclusion doesn't account for the Jevons paradox. In itself, a technical progress that improves the efficiency with which a resource is used doesn't lower the demand for this resource, but instead increases it. The solution to a more livable planet is political, and this is why it's so difficult to find.


Jevons is not universal. Power expended on illumination has been decreasing as people switch to LED lighting. Enough light is enough.

I don't think they're trying to reduce the amount of electricity generated/used. Rather, they're trying to help electrical out-compete fossil fuels.

If you make it cheaper to use a less polluting resource than a more polluting one, this can still work. Improved solar technology, for instance.

Yes, but the point GP makes is that you need carrots and sticks. One or the other alone will not be enough.

He's claiming that, I guess, but he doesn't actually provide any support. It seems totally possible, depending on the technological specifics, to require zero political coordination.

Off topic.

Perhaps if every person embraces the politician within themselves.


Here is a video of one in operation with some techno music:

https://www.youtube.com/watch?v=rGu7XDapR58


That's cool, but I was hoping to find a video detailing the difference in design between axial and radial motors. You know, with exploded 3d models and such. And also with techno music.


Thanks for an actual answer to my flippant comment. Actually, I was hoping to see an explanation of why the different design leads to smaller, more efficient motors. I understand that the flux direction is different, but I don't understand how that enables the improvements.

That techno music makes me feel very empowered.

I'm sorry to say but that is not techno music at all.

What is it called?

A concise and technical summary from their blog:

https://www.magnax.com/magnax-blog/axial-flux-vs-radial-flux...


> "One of our designs has a peak power density of around 15 kilowatts per kilogram. Compare that with today’s motors, such as the one in the all-electric BMW i3, which delivers a peak power density of 3 kW/kg—or just one-fifth as much. And the Magnax machine is also more efficient."

It will be interesting to see how Moore's law will materialize in EVs. Understanding that the majority of time and effort spent in the automobile industry has been aimed towards combustion engine design as opposed to electric motors, it will be equally interesting to follow the efficiencies to be discovered in the imminent future, taking into consideration how blazing quick they already are today.


15kw is about 9hp/lb. If you can get that maybe you can design a wheel motor that doesn't suck. Two 75 hp motors would weigh about 8 lbs each.

No gear box, differential or CV joints. Potentially really cheep.


The most I understand about motors is how they typically work and that innovations are possibly more lucrative than battery innovations; this seems like a major innovation. I must be missing something? What's the healthy skepticism here?

Radial flux motors also work great. Whether one or the other is better depends on your application.

1) I don't consider this a breakthrough. 2) Magnax's motors are a really sensible design (like Emrax's) and I think they'll do very well.

Radial flux motors can achieve just as high of specific power and are already pretty cheap and available. I think it has taken a while for BLDC motors to become cheaply available at high specific powers except for R/C applications, and there wasn't really any fundamental reason why, but it's better late than never.

(I make high performance electric motors, including both radial and axial flux ones)


Emrax sells this technology for years now. This isn't new, and I couldn't find a datasheet to meaningfully compare them with what else is available on the market.

I do know of axial flux motors that couldn't have had iron teeth, but I think it was a single-side design with a metal backplate to support the resin-soaked stator windings. I recall a nice discussion about them using GaN power electronics to reach very high efficiencies. Those machines were still hand made last year, however...


You can, of course, make an ironless axial flux motor with the resin and wire itself acting as a structural composite. Two-sided. I think commonly were used for solar cars (i.e. for student competitions).

The prototype I touched was for/part of a 200kW pair (I don't recall if each or total) for a demonstrator plane...

I don't see anything to be skeptical about; they seem to have done a lot of work and research.

One curiosity for me is whether putting the motor in the wheel will work well. It's important to keep sprung weight as low as possible for suspension reasons. Then again, many (most?) electrical vehicles will be run on smooth roads.


These motors are light. If you integrate them with the wheel to the point where you mount the tire to it instead of to the rim, you should be close to current mass, possibly breaking even (depending on how much of the axle/differential is unsprung mass, and how heavy the flexible torque transmission ends up being).

I already have potholes destroying my rims, I'm not looking forward to the day a pothole wrecks one of my engines!

Even if it does what it says, it’s a minor improvement over existing motors once you factor in the rest of the car. The mass of the motor matters but much less than the mass of the batteries.

1) you're right, for every "revolutionary" technology innovation we hear about that actually turns out being important, there are several that don't, but... 2) automobiles were not, until recently, an area that received a lot of effort at major innovation. Credit electric (and hybrid electric) cars with redefining the amount of change that seemed practical to consider.

As far as I know electric motors have very high theoretical and also practical efficiencies. Something like 95%+ theoretically, not sure about the practical value, but it's most likely higher than diesel (~40%) or gasoline (less than 40%, probably a bit higher than 35%?).

The batteries are the big pain point. A battery that's 50% lighter, has 50% more capacity and costs the same would trigger a technological revolution.


> "Motors and motor systems account for approximately 53 percent of global electricity consumption. We estimate that improving the efficiency of all the world’s motors by just 1 percent would reduce the motors’ power consumption by 94.5 terawatt-hours and shrink their carbon dioxide footprint by the equivalent of 60 million metric tons."

I find this surprising. Perhaps a big chunk of this is air conditioning, which is technically using a motor to run a compressor?


Imagine a factory. All those moving things, power tools, conveyor belts, cranes, pumps, etc. are driven by electric motors.

And refrigerators?

slight tangent...

Interesting thing to consider is that reverse cycle airconditioners are notoriously inefficient compared to dedicated cooling or heating units..

Reverse cycle units can use anything from 2 times to 5 times the power of a dedicated unit on cooling


At least 2 times and as much as 5 times seems surprisingly bad.

Do you have a source or a concrete example for that?

The most plausible explanation I can think of for what you're saying is that you're comparing an evaporative cooling system to a heat pump based one, in which case the 'reverse cycle' aspect is not the thing causing the inefficiency.


How much does motor weight matter in a modern long-range EV, vs battery weight?

An AC motor for a car might be in the 100-200 pound range or so. DC motors tend to be a little heavier. Along with the motor, you'd need a motor controller and cabling.

Example conversion motors: https://www.evwest.com/catalog/index.php?cPath=8

A general rule-of-thumb is that it takes about 100 watt-hours to move a thousand pounds of car one mile. I just started on a project to convert a Mazda RX-8 to electric. That's a 3,000 pound car (before the conversion), and I'm planning to use a Netgain Hyper9 AC motor, which is 120 pounds, which is pretty small compared to the rest of the car (or compared to the original rotary engine). Battery weight is the bigger concern; if I can keep the pack under 400 pounds I'm doing pretty good, and I'll be lucky if that gets me 100 miles of range.


According to Magnax no-load losses reduce the effective range of the vehicle by 10% to 20% (iron close to rotating permanent magnets equals braking when there is no load) Magnax claims 85% less iron losses due to the low amount of iron.

Awesome design, makes me want to put one together just to see it work. :)

One concern: I saw some designs with the rotor outside the stator, physically mounted to a wheel. I'm not a mechanical engineer, but as a guy who considers himself not quite dumb enough to fail at everything...

I hope they build those motors tough. Typical electric motor mounts create a situation where the only real stresses on the motor are rotational, along the axis of rotation, or in the case of a pulley / gear-mount setup, you are usually limited to 1-2 additional push/pull axis. But on a wheel? That wheel is going to want to rotate and move on all axis, which now needs to be supported by a motor, which also includes permanent electric magnets.

Awesome lighter motor, awesome that we can sandwich the rotor between stators (more layers more torque!).


My first thought is that Toyota needs to license this for their hybrids, and mount it on/as the flywheel. Using something like this as either MG1 or both MG1 and MG2 would seriously minify their hybrid drivetrain.

Why hybrids, Why not just go for full electric?

Because hybrids are the next step until EVs reach the junction of affordable and practical for the average household. Toyota knows this, and is playing it safe until there's a mass-market demand, when they'll step in with the most practical and boring EV possible.

There was a lot of ifs, mays and hopefullys in that article, but maybe it's just written in a strange way since it does say they have actually found a way to mass produce this thing. It would be great if true.

One of the first questions that popped into my mind was how would this motor with less mass (less structure) handle 5 times the power/mass of traditional electric motors which are already pretty power full and sometimes break from strain.


> One of our designs has a peak power density of around 15 kilowatts per kilogram. Compare that with today’s motors, such as the one in the all-electric BMW i3, which delivers a peak power density of 3 kW/kg—or just one-fifth as much. And the Magnax machine is also more efficient.

>We believe that we can scale the design to whatever size carmakers (and other customers) may demand.

Count the weasel words. I'm at 5.


Even at 3kw/kg, an electric motors blows ICE out of the water.

Not my point, and an electric motor isn't going to do much without a battery.

Complex issue. With this innovation they can rethink the whole axel/brakes/rims/hub motor system and perhaps come up with more novel solutions. E.g. Brakes/hubmotor integration to remove the unsprung weight penalty? Rim/hubmotor integration? It's an exciting time.

> In tests at the University of Ghent on the first prototype, our yokeless axial-flux motor reached efficiencies from 91 to 96 percent. And that was just the prototype.

Wouldn't one expect a prototype to be better than a mass produced unit? It gets individual care and can be optimized for the test.


Nah. Prototypes have lots of compromises - wrong materials, crude circuits etc. Maybe bolted to a plywood table top etc. The purpose of a prototype is proof-of-concept and measurement/recording. Not a simulation of real deployment.

It will probably undergo several revisions and optimizations before mass production. Prototyping is usually pretty early in the process.

An electric plane with a motor like this, fueled with hydrogen like Toyota Mirai... seems like it can improve the situation with range?

Looks lovely for drones.

You get by without the fancy iron teeth mechanical support they invented as long as you remain small enough. The main issue is miniaturizing magnet manufacture without per-magnet costs staying fixed.

I judge by COTS pricing, but the $/kW falling on magnets increases a lot below 5~15kW electrical.


Yeah, if you have really tiny rare earth magnets, their cost-per-weight (for a given magnetic energy product) goes up a lot. Just finicky to deal with sintering a bunch of small things, I suppose.

But when will we see a motor powered by magneto-reluctance and capacitive diractance instead of the relative motions of conductors and fluxes?

Seriously, though, this article is clear and well-written. I love it when I see people solve problems we didn't know we had. I'd figured we'd basically found the global optimum on electrical motors decades ago and were basically stuck there by physics now, with advances happening elsewhere in the powertrain. This development is like CPU vendors figuring out how to write 3x more single-core performance out of the same thermal budget after our own decade of stagnation in this area.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: