Car enthusiasts mourn the commoditization of propulsion. Once petrolheads would have chosen a BMW for its sonorous straight-six or a Mercedes-AMG for its thunderous V8. Now many believe that distinctiveness is rapidly diminishing. Electric cars might provide mad, silent thrust, but a common complaint is they are mostly indistinguishable for the character of their drivetrains.
Carmakers worry about this too. Their engineering DNA is less apparent in the EV age, leaving them more reliant on design, brand power, and other types of technology to differentiate their cars and keep their customers. There's no point trying to trump the competition on power when the quickest Teslas and Lucids already have far more than you can ever deploy on the public road. More isn't better when you already have too much.
But soon there'll be a choice again: between the conventional radial-flux motors that have powered almost every EV until now and something radically different.
Axial-flux motors won't necessarily offer more power, but they are so much lighter and smaller that their proponents say they have the potential to transform almost every other key measure of an EV's performance—and the entire architecture of a car designed around them.
By fitting axial flux motors into the wheels, the spaces in a car's body currently occupied by motors could be largely vacated, clearing the way for more batteries, people, or stuff, and permitting the sort of design exuberance that EVs have long promised but never quite delivered.
More importantly, this new design of motor might help address the growing public backlash against overweight, expensive EVs. They might reduce the weight of a typical EV by around 200 kilograms (440 pounds)—half in the motors themselves, and half from the mass-compounding effect which allows you to reduce the weight of other systems such as batteries and brakes as a result.
By sending mass into a virtuous downward spiral, carmakers could increase range, decrease cost, and perhaps even preserve the agile handling of lightweight cars, which enthusiasts also worry might disappear with the advent of the EV.
Flux Capacity
The principle isn't new. The axial-flux motor was first demonstrated by Michael Faraday in 1821, but in the intervening two centuries nobody had figured out how to mass-produce one reliably.
British academic Tim Woolmer, however, likes a challenge. He devoted his Oxford PhD to designing the optimum motor for an electric car. An axial-flux motor would make more sense than the almost ubiquitous and easily mass-manufactured radial flux design, he decided. But not only had his chosen design barely made it out of the lab in nearly 200 years, there simply wasn't a market for it when he started in 2005: GM's EV1 had long been canned, and the Tesla Roadster was still three years away.
In an axial-flux “pancake” motor, the stator (the stationary part of an electric motor) and rotors are discs, sitting alongside each other less than a millimeter apart, the flux flowing through the stator axially or parallel with the shaft, and acting on the permanent magnets in the rotors on either side to turn them.
In the “sausage roll” design of the familiar radial-flux motor, the cylindrical rotor (the sausage) passes through and is turned by the magnetic flux, which flows radially around the stator's copper windings (the pastry).
The advantages of axial flux motors are obvious when you see one on Tim's desk. It's similar in diameter to the Tesla motor sitting next to it, but in this case around one-sixth the length, a quarter of the weight at 12 kgs, and a quarter more powerful at 250 horsepower. It's a neat, resolved piece of engineering, and it makes the Tesla motor look like a science fair project with its messy, wasteful copper end-windings, hand-tied with Kevlar string bulging out of the stator.
Other advantages are less visible. With their large surface areas and hot bits not buried in the bowels of the motor behind multiple thermal interfaces, axial-flux motors can be cooled more easily. There's more surface area for magnetic interaction, and because most of it happens at the periphery of a much larger-diameter rotor, axial motors naturally produce more torque.
But figuring out how to mass-manufacture one had eluded everyone before Woolmer. Five weeks into his PhD he'd cracked it. His design eliminated the heavy “yoke,” which provided the base for the copper coils of the stator in conventional axial-flux designs, saving around 80 percent of the iron content, making his design even slimmer and lighter, and clearing space for a second rotor on the other side of the stator, increasing torque again.
Motors Like Jewelry
Woolmer's coils are like copper jewelry, edge-wound into triangular sections around soft magnetic composite cores and artfully arranged in the stator like the segments of an orange.
He figured out how to cool them directly with oil—impossible in a radial-flux design, and permitting his design to run continuously at 90 percent of its peak power compared to around 70 percent for a radial motor. This also maintains a consistent gap between stator and rotor despite the huge magnetic forces acting on them.
Woolmer founded YASA in 2009, before he'd finished his PhD, and his rethinking of Faraday's first principles now extends to more than 150 patents. The company name may sound exotic, but it stands for “yokeless and segmented armature,” the key engineering principles which still underly his designs. YASA's early years were mostly spent “making the machines to make the machines,” as Woolmer puts it, because no suppliers could build the tools to make something that had never been made before.
But the motors also progressed, by at least 20 percent each year in every important measure, Woolmer claims. His first prototype made just 20 kilowatts and weighed 20 kgs. By 2012, power had increased eightfold, and YASA had begun to supply prototype, racing or very low-volume road cars such as the Jaguar C-X75, Koenigsegg Regera, and the Lola-Drayson electric land speed record car.
In 2019, the Ferrari SF90 became the first standard-production car to get a YASA motor, followed by the 296GTB. The SF90's slim unit provides hybrid power to the rear axle, sitting perfectly between the combustion engine and gearbox. Lamborghini's new Revuelto hybrid uses two on the front axle to assist its V12. McLaren is also reported to be a customer.
Axial Aviation
Woolmer also founded Evolito to develop axial-flux motors for electric flight, where they have the potential to be even more transformative than in electric motoring. Three YASA motors powered the Rolls-Royce Spirit of Aviation experimental electric aircraft which set a series of speed records in 2021.
Since then, Evolito has been in discussions with most of the major aerospace companies, but Woolmer won't divulge anything further. Physics and economics work far differently here: Aviation clients are willing to pay €10,000 ($10,770) to save a kilogram. compared to €10 at the carmakers, Woolmer says, and the mass-compounding effects are greater too, with every kilo cut from a motor allowing up to eight to be cut elsewhere—potentially making an EVTOL design practically and commercially viable.
Evolito is developing motors with 50 kilowatts per kilogram—“and we have a realistic chance of getting there”—so around 4 kg, or one-third the mass of the SF90's motor for the same power with the use of exotic, expensive, lightweight materials.
For road cars, YASA is targeting motors of around 300 hp and 7 kg, less than a conventional brake disc, and allowing the motor to be placed in the wheel itself without the increase in unsprung mass ruining the ride or handling of the car.
Then the weight savings really start to multiply: no driveshafts, possibly no gearbox, and if the law is changed, no conventional friction brakes either—the electric motors doing as effective a job of shedding speed as creating it.
“You won't see that on the road for a number of years, but it's in the lab today,” says Woolmer. The UK government is convinced and has backed YASA's in-wheel motor project with £21.2 million ($26.5 million) of funding from its Advanced Propulsion Center last September, and a further £7.7 million in March.
"They thank you in automotive for saving weight, but it's not a game changer," Woolmer says. “But there is an inflection point where the motors become small and light enough that you can start to place them near the wheels, free up all that space, and drive new architectures. Even if our motor was more expensive, and I'm not saying it is, I think its packaging advantages would be such a compelling argument that cost isn't even a consideration.”
There are other companies attempting to commercialize axial-flux motors for use in EVs, but it’s a fraught and expensive business, and none has yet reached production or been included in a car. British start-up Saietta, which aimed to produce affordable, lightweight axial-flux motors for use in motorcycles and trikes, went into administration in March. Another, Magnax, is based in Belgium, but has not posted news on any business developments on its site for two years.
And not all of the attention YASA is attracting is welcome. "People have directly tried to copy aspects of the technology, and we have ended up in a few behind-the-scenes grapples," Woolmer says. "But because it's such an unexplored space, and we got there very early and have pursued the technology hard for 15 years, we've got some really strong, healthy intellectual property. Traditionally, problems are seen as a negative thing, but we think that if we hit a problem first, when we solve it we'll find some protectable IP. Someone else coming along can't directly copy. They're going to have to navigate a really nasty thicket of patents. And, anyway, there's no time to copy anyone in this market, because if you do, you're already five years too late."
Mercedes Makeover
Protecting that IP now falls to Mercedes-Benz, however. It wanted YASA's tech for its high-performance, all-electric AMG.EA platform, which will spawn its first cars late next year.
While YASA can scale up to produce motors for its low-volume supercar clients, building them in the six figures that Merc needs might have bankrupted it. So Mercedes bought the entire business—excluding Evolito—for an undisclosed sum in 2021.
Merc showed off the Vision One Eleven concept using YASA motors last year, and it is now building a factory in Berlin to make the motors for AMG at scale, and maybe maintain some of the distinctiveness which Merc's commitment to full electrification threatened to erase.
"That is precisely the reason we bought the company,: Mercedes-Benz CEO Ola Kallenius tells WIRED. “From the start, AMG always had the powertrain as a signature feature. Transport yourself 60 years later into the electric era: Why should we think any different? You go back to the recipe again.”
"You've been at YASA. Look at how compact that unit is. It's almost ridiculous. That power-to-weight ratio really can only be achieved with the axial-flux design. Marry that to our performance battery, and you're back to the original AMG recipe, but with fundamentally different technologies that I believe will be relevant for the AMG customer. And they know they're getting something that nobody else gets."
Well, nobody except Lamborghini, Ferrari, and McLaren buyers, though the association with those marques will hardly harm AMG, and their more extreme demands might drive improvements in YASA tech, which will filter down to mere AMG customers.
Lucid Alternatives
But not everyone is sold on axial-flux tech. Although their torque density is hard to argue with, some question the wisdom of moving motors into wheels, possibly increasing unsprung mass, duplicating ancillaries, and exposing the motors to damage.
Radial-flux motors have also been making rapid advances in design and power density, and Aston Martin has chosen not to follow its supercar rivals down the axial route, instead giving Lucid a stake in its business in return for access to the American firm's propulsion systems, which are probably now the benchmark for radial-flux tech.
Disassembled, a Lucid motor is as beautiful a piece of engineering as a YASA. The copper field coils are shaped from a single stretch of wire and pressed into the stator eight-deep in processes unique to Lucid, eschewing the hundreds of resistance-inducing welds of a Porsche motor and Tesla's messy end-windings.
At 500 kW from a 31.4-kg motor, or 16 kW per kg, the Lucid unit has more than twice the power density of the Porsche motor, and more than two and a half times that of even the Tesla “Raven” motor with its carbon-fiber rotor.
Rimac's radial motors boast similar numbers, their power boosted by their ability to turn at 24,000 rpm, up from 15,000 just six years ago.
"We never want to be complacent about our technology," says Woolmer. “Radial machines have improved massively. We talk about this three-to-one advantage in torque density that axial motors have, but if we'd stood still for the last five years, that would've been edged away to not very much.”
Lucid's VP of powertrain, Emad Dlala, acknowledges that axial motors have their place, but doesn't see Lucid making one. “In hybrid applications where their dimensions are an advantage, in aviation and maybe for brands like Ferrari where cost is less relevant, sure,” he says. “But what's the definition of the best electric motor? We already have a design that gives us 500 kW and over 200 mph. We need to step back and think about what the customer needs. We think that's making EVs more affordable, and I don't think axial-flux motors can be made cost-effectively in the hundreds of thousands, or millions.”
Mate Rimac is complimentary about Merc's purchase, but he also says he'll stick to radial. “Our current strategy is based on radial-flux motors, and as we go to higher volumes we see the scalability of radial-flux motors as an advantage,” he tells WIRED. “Axial-flux motors have their advantages and applications as well, but YASA seems best positioned in that field, so we don't see the necessity for us to go there. I think that Mercedes did a great move to acquire YASA. Those motors will give Mercedes-AMG a competitive advantage. But hypercars are a different animal. They need even more power and torque—and there, in my view, radial-flux motors are currently better positioned.”
Woolmer might disagree. YASA motors might produce up to 800 hp one day, he says, and while they're not currently working out how to make them by the million, it's on the radar.
He now has more than 400 staff in the UK, and the capacity to make up to 70,000 high-end motors for YASA's supercar clients at his two plants near Oxford, just a short electric drive from where he first had the idea.
Remaining independent and agile within the Mercedes-Benz machine might be another challenge, but Ola Kallenius says he's committed to protecting YASA: particularly if it gives the public a reason to buy his cars rather than his rivals'.
"That's the dream, right?" says Woolmer. “A little Y on the back of an AMG. But we also need to offer a step-change in vehicle architecture, and I hope the rest of the world will fall in with us and Mercedes-Benz as we take this idea into mass production.”
