59 Kilowatts Per Kilogram: How YASA's Pancake Motor Finally Killed the Sausage

Nikola Tesla patented the first axial flux electric motor in 1889. It took 126 years for one to reach a production car, the 1,500-horsepower Koenigsegg Regera in 2015, and another eleven for someone to build a factory around the concept. On June 9, 2026, Mercedes launched series production of YASA's axial flux motor at Berlin-Marienfelde, the plant where Karl Benz's company has been making things since 1902. The motor will debut in the next Mercedes-AMG GT 4-door coupe.

Sounds incremental, but it is not. Nearly every electric vehicle on the road today uses a radial flux motor: every Tesla, every Taycan, every Hyundai Ioniq, every Rivian. Electrifying the highest-performance segment demands a fundamentally different electromagnetic architecture, and YASA's design is the first to survive the jump from hypercar exotica to volume manufacturing, which means understanding the physics of magnetic flux orientation is no longer optional for anyone who cares about where electric cars are actually going.

Sausages and Pancakes

A radial flux motor looks like a sausage. A cylindrical rotor spins inside a cylindrical stator, and the magnetic field lines run perpendicular to the central shaft, radiating outward from center to circumference. Torque is generated by the interaction between permanent magnets on the rotor and electromagnets on the stator, and in this geometry, torque scales linearly with the motor's length but only modestly with diameter.

This is the architecture that won. Radial flux machines are simple to manufacture using stacked steel laminations stamped from flat sheet stock, the tooling is well understood, the supply chains are mature, and heat dissipation is straightforward because the stator sits on the outside where it contacts a cooled housing. Tesla's Model 3 permanent magnet motor produces about 211 kW from 32 kg, roughly 6.5 kW/kg, which was world-class in 2017.

An axial flux motor looks like a pancake. In YASA's configuration, two disc-shaped rotors bracket a central stator, all three components sharing roughly the same diameter, and the magnetic field lines run parallel to the shaft instead of perpendicular. Torque in this topology is proportional to the rotor diameter squared. Punishing advantage. A moderate increase in disc diameter yields a disproportionate torque gain because you are leveraging the moment arm of the entire rotor face rather than just its length, and doubling the active magnetic surfaces with dual permanent-magnet rotors creates a short, symmetrical flux path that reduces losses in the magnetic circuit.

YASA's latest prototype motor generates 750 kW of peak power, which is 1,005 horsepower from 12.7 kilograms, a unit roughly the weight of three gallons of milk that fits on a dinner plate. Resulting power density: 59 kW/kg, an unofficial world record according to YASA, and approximately triple what the best radial flux designs from any manufacturer can achieve today.

Continuous output sits in the 350 to 400 kW range: 469 to 536 horsepower sustained without thermal derating. Matters enormously for track use, where a motor that peaks at 1,000 hp but fades to 600 after three laps is a party trick, not a powertrain.

Why Axial Flux Lost the First Round

Manufacturing killed it. Radial flux stators are made from laminations: thin steel sheets stamped into rings and stacked like poker chips. Each lamination is electrically isolated from its neighbors by a thin oxide layer, which suppresses eddy currents that would otherwise waste energy as heat, using a fast process built on commodity steel and automation refined over a century of industrial motor production.

You cannot laminate an axial flux stator. Flux runs parallel to the shaft, which means any effective lamination plane would need to be oriented radially rather than axially, and that geometry does not lend itself to stamping and stacking flat sheets into a coherent magnetic circuit no matter how creatively you arrange the tooling. For decades, this manufacturing incompatibility kept axial flux motors in laboratories, academic papers, and the occasional aerospace prototype where cost was irrelevant.

Tim Woolmer cracked it.

The Yokeless Segmented Armature

Woolmer was five weeks into his PhD at Oxford when the insight arrived: remove the stator yoke entirely and split what remains into individual segments. In a conventional motor, the stator yoke is the continuous ring of iron or steel that forms both the structural backbone and the magnetic return path for all the coil-carrying teeth, connecting them into a single monolithic ring that is heavy, lossy even in laminated form, and in an axial flux topology represents a disproportionate fraction of the machine's total mass because its diameter must match the rotor discs.

YASA's architecture eliminates the yoke completely, and the name itself encodes the solution: Yokeless And Segmented Armature. What replaces the continuous ring is a set of discrete pole pieces, each one individually wound with copper coils and held in position by a non-magnetic structural carrier, with no continuous magnetic ring connecting them and each segment handling its own flux path independently.

This is where the material choice becomes critical. Those pole pieces cannot be made from stacked laminations for the same geometric reasons that killed laminated axial flux stators in general. Woolmer turned to Soft Magnetic Composite, a material consisting of iron powder particles individually coated with a thin electrical insulation layer and then compressed under extremely high pressure into solid three-dimensional shapes.

SMC has two properties that make it transformative for this application. First, the insulated particle structure suppresses eddy currents in all three spatial dimensions, not just the two that laminations address. Conventional laminations block eddy currents only in the plane perpendicular to the lamination stack direction; currents can still circulate within each sheet. SMC kills them everywhere because the insulation surrounds every individual grain. Second, SMC can be pressed into complex three-dimensional geometries using powder metallurgy tooling, which means the pole pieces can be shaped to optimize the magnetic flux path in ways that would be impossible with stamped steel.

Weight savings are stark, a six-to-one ratio: where a conventional motor stator might contain 30 kg of iron and steel yoke material, a comparable YASA design uses approximately 5 kg of SMC pole pieces to generate equivalent power and torque. Woolmer's segmented architecture, combined with the flat copper windings that wrap each pole piece, creates a stator roughly one-third the mass and one-third the axial length of its radial flux equivalent.

Cooling the Flat Copper

Thermal management in electric motors is usually a compromise between where heat is generated and where it can be removed. In a radial flux machine, stator windings sit inside slots cut into the lamination stack, surrounded by iron on three sides, while oil or water-glycol coolant flows through channels in the motor housing, contacting only the outer stator surface. Heat generated deep inside those slots must conduct through layers of insulation, copper, and iron before reaching the coolant. Deeper winding, worse thermal path.

YASA's segmented stator eliminates the buried-copper problem entirely, because the flat copper windings wrap around individual pole pieces that sit in the open gap between the two rotors, and YASA routes oil coolant directly across the winding surfaces so that every square millimeter of resistive-heat-generating copper is in direct contact with the cooling medium, a thermal architecture that would be geometrically impossible in a conventional slotted stator where the windings are imprisoned behind iron walls.

Continuous power rating is where this pays off. Peak output is limited by magnetic saturation and the inverter's current capacity, but continuous output is limited by thermal equilibrium: how much power the motor can deliver indefinitely without its windings exceeding their temperature rating. YASA's direct oil cooling allows 350 to 400 kW continuously, roughly half its peak, a continuous-to-peak ratio that significantly exceeds what most radial flux machines achieve because those lamination slots create a thermal bottleneck that forces steeper derating under sustained load.

The AMG GT XX as Proof of Concept

Before committing to mass production, Mercedes validated the motors as publicly as possible. Three YASA axial flux motors in the Concept AMG GT XX, producing a combined output of approximately 1,000 kW, were sent to the Nardo high-speed ring in southern Italy in August 2025 and held at a sustained 300 km/h for 7.5 consecutive days, covering 40,075 kilometers, a distance equal to the circumference of the Earth, with stops only for charging at 850 kW.

Nardo was not about speed, because speed is easy and any electric motor can hit 300 km/h briefly. Holding it for 5,300 kilometers per day, day after day, with desert-level ambient temperatures and the full aerodynamic drag of a four-door sedan fighting the motors constantly, is a sustained thermal endurance test that exposes every weakness in the cooling system, the magnetic circuit, and the bearing assemblies long before any marketing team writes the press release. Dozens of EV endurance records fell.

That car used three motors, and the production AMG GT 4-door reportedly uses a similar multi-motor layout, though Mercedes has not confirmed the exact configuration. What matters more than the specific vehicle application is what happened next: a factory.

Berlin-Marienfelde: From Combustion to Flux

Mercedes chose its oldest continuously operating plant for axial flux production, a facility making powertrain components since well before anyone at the company imagined electric cars, and retooled it with a production target of up to 100,000 YASA motors per year.

A hundred thousand motors per year is significant because it marks the transition from low-volume supercar supply to genuine industrial scale. YASA previously supplied Ferrari for the SF90 Stradale and the 296 series, Lamborghini for the Temerario, Koenigsegg, and McLaren, but combined volume across those customers measured in the low thousands annually. Berlin-Marienfelde, supplemented by the Yarnton factory near Oxford that opened May 2025 with capacity for 25,000 units per year, quadruples the production ceiling overnight.

Michael Schiebe, Mercedes's production chief, framed the launch as "turning a groundbreaking innovation for electric mobility into industrial reality," language that acknowledges the uncomfortable truth that axial flux motors have been "groundbreaking" for decades while remaining commercially irrelevant, a gap between laboratory power density records and actual production lines that has killed more electric motor startups than any technical limitation ever did.

The Cascading Weight Thesis

YASA's argument for axial flux extends well beyond the motor itself. Replace a 32 kg radial flux motor with a 12.7 kg axial flux unit producing equal or greater output and the 19.3 kg direct saving is obvious, but that initial reduction compounds through the vehicle architecture in ways that are easy to miss and impossible to ignore once you trace the chain.

A lighter motor needs less structural support, which means lighter motor mounts and subframe brackets. Less rotating mass and less total vehicle mass reduce braking energy requirements, which permits smaller and lighter brake components. Reduced braking energy also means less kinetic energy is converted to heat during deceleration, and in a regenerative braking system, the motor recovers a higher fraction of available energy, improving effective range. Better range from the same battery chemistry means the battery can be physically smaller for a given range target, saving additional mass and volume. Lighter batteries need less structural reinforcement in the floor pan, which saves more mass still.

YASA estimates the total cascading weight reduction at approximately 200 kg per vehicle: roughly half from the motors themselves and half from the downstream savings in batteries, brakes, structures, and thermal management systems. For context, 200 kg is approximately the weight of two adult passengers, and removing it from an EV improves acceleration, braking, range, and handling simultaneously with no trade-offs.

What Axial Flux Cannot Do Yet

Axial flux motors are not a universal replacement for radial flux machines, at least not today. YASA's design excels at high torque density within a compact diameter, which makes it ideal for performance applications, hybrid integration where the motor must fit between an engine and a gearbox, and in-wheel motor concepts where the flat pancake geometry fits naturally within the width of a wheel hub.

At very high rotational speeds, the pancake architecture faces structural challenges because the rotor discs, significantly wider than those in a radial flux machine, experience higher centrifugal forces at equivalent RPM, and the permanent magnets bonded to those discs must resist shear and peel forces without shifting or delaminating, which limits practical maximum speed compared to the long, narrow, carbon-fiber-sleeved rotors used in machines like Tesla's that can exceed 20,000 RPM.

Cost is still an obstacle. SMC stator segments require specialized powder metallurgy tooling and pressing equipment that the radial flux supply chain does not possess. Economics improve with volume, which is precisely why the Berlin-Marienfelde commitment matters, but YASA's motors will initially appear only in AMG-tier vehicles where the price premium is absorbed by buyers paying six figures for a car and who care more about lap times than unit costs.

SMC also carries an inherent disadvantage in permeability compared to grain-oriented electrical steel laminations. Its ability to suppress three-dimensional eddy currents comes at the cost of lower peak magnetic performance per unit volume, meaning the magnetic circuit design must be more carefully optimized to compensate. YASA's proprietary pole piece geometry and the dual-rotor flux path address this, but it remains a constraint that shapes the design envelope.

Where This Goes

Three of the last four hybrid supercars to launch used YASA motors: the Ferrari 296, the Lamborghini Temerario, and the McLaren W1. Rolls-Royce's Spirit of Innovation, currently the world's fastest electric aircraft at 559.9 km/h, ran three YASA motors to drive its propeller. Jaguar set a maritime electric speed record of 142.6 km/h in England's Lake District with YASA propulsion. Land, sea, air. Anywhere power-to-weight ratio determines outcomes, these motors show up.

Mercedes buying YASA outright in 2021 was the signal that axial flux had graduated from boutique supplier to strategic asset, and building a 100,000-unit factory confirmed it. AMG GT 4-door will be the first mass-produced car to run on axial flux motors, but not the last.

YASA's motors use no exotic materials beyond the neodymium magnets common to all permanent-magnet motor designs. SMC stator material is iron powder, one of the most abundant metallic elements on Earth, pressed and sintered using industrial processes that scale without rare-earth supply chain risk. Woolmer has emphasized this repeatedly. A motor architecture that depends on materials controlled by a single geopolitical supplier is not a long-term solution for global automotive production regardless of how impressive its dyno numbers are.

The sausage had a good run. A hundred and thirty-seven years after Tesla's first axial flux patent, the pancake has a factory.