Views: 0 Author: Site Editor Publish Time: 2026-11-20 Origin: Site
Axial-flux motors are often described as a revolution that “flattens” the electric motor. Magnetic flux passes through the air gap in the axial direction, and the stator and rotor are like two discs placed face to face. The result is a flat shape, high torque density, and low rotational inertia. But “flat” is only the starting point. What really determines whether an axial-flux motor can be used, whether it works well, and whether it is worth mass-producing is how the stator and rotor are arranged in space.
Single-stator, single-rotor (SSSR), dual-stator, single-rotor (DSSR), and dual-rotor, single-stator (DRSS) are the three most basic “arrangements” in the axial-flux field. The differences among them are far more than “one more disc.” They directly affect axial magnetic pull, heat-dissipation paths, torque density, manufacturing yield, and cost structure. The first step in selection is not to find the “highest-performance” structure, but to identify the constraint line your application cannot compromise on.
SSSR is the most intuitive axial-flux configuration: one stator on one side, one rotor on the other, a single air gap, and magnetic flux that closes through the housing after crossing the air gap. Its advantages are almost “determined by the structure”: the fewest parts, the simplest assembly, an axial dimension that can be minimized to the extreme, and the lowest mass-production cost among the three. At low speed and light load, the magnetic circuit is short, and copper loss is also well controlled.
But SSSR has a physically unavoidable defect: a single-sided magnetic circuit inevitably produces a one-way axial magnetic pull. The stator continuously “pulls” the rotor to one side. Under the same power rating, this force can reach more than 30% of the rated force, and at high speeds the vibration amplitude is 20%–40% higher than that of symmetric structures. The consequences are accelerated bearing wear, rotor eccentricity, amplified noise, and a very low reliability ceiling for long-term operation.
Therefore, the reasonable boundary of SSSR is clear: low power, low speed, cost-sensitive applications that are not intended for continuous high-load operation. Small servos, household fans, and lightweight portable devices are its home ground. Force it into high-speed or continuous heavy-load conditions, and the cost savings will soon be eaten up by bearing replacement and vibration problems.
DSSR places the rotor between two stators, forming a symmetric dual-air-gap structure. This sandwich solves two things.
First, axial force self-balancing. The magnetic pull of the two stators on the middle rotor is equal in magnitude and opposite in direction, so it can theoretically be completely canceled. The rotor is no longer “dragged to one side,” bearing load returns to normal, and high-speed stability and overall service life are significantly improved.
Second, the heat-dissipation path is naturally friendly. Both stators are exposed to the outside, so heat from the stator windings can be conducted to the housing from both sides at the same time. Under air cooling, DSSR is already capable of supporting long periods of full-load operation, and its thermal margin is the largest among the three basic topologies. For applications such as industrial continuous operation, medium-power servos, and in-wheel drive—scenarios that “cannot stop”—this is more practically meaningful than peak torque figures.
The cost is on the cost side: two sets of stator cores and windings mean higher material cost than SSSR; the windings on both sides require synchronized wiring and precise alignment; the rotor is suspended between the two stators, and the tolerances for axial gap control are also tighter. The manufacturing yield of DSSR is approximately 92%–95%, lower than the 98%+ of SSSR but better than that of dual-rotor structures.
In one sentence, the selection logic of DSSR is: when “long-term stable operation” and “reliable heat dissipation” take priority over “extreme lightweight” and “extreme cost,” DSSR is the least likely of the three to go wrong.
DRSS places one stator between two rotors, and both faces of the stator contribute to torque production at the same time. This is the configuration with the highest magnetic-circuit utilization among the three. At the same volume, the torque density of DRSS is 15%–25% higher than that of DSSR and 2–3 times higher than that of conventional radial-flux motors. The well-known YASA (yokeless and segmented armature) structure is essentially a deeply optimized version of DRSS: it removes the stator yoke and retains only segmented teeth, greatly reducing stator weight and significantly lowering high-frequency iron loss.
But the “performance ceiling” of DRSS comes with a steep rise in engineering complexity.
Heat dissipation is the first hurdle for DRSS. The stator is sandwiched between two rotors, so its direct contact area with the external environment is actually the smallest. It does not have the natural heat-dissipation advantage of DSSR, where “both stators are exposed.” High-performance operation almost necessarily depends on closed liquid-cooling oil channels or immersion cooling to “draw” heat out from inside the stator. Under air cooling, the thermal disadvantage of DRSS directly limits continuous power output.
Manufacturing precision is the second hurdle. The two rotors must be strictly synchronized and aligned, and dynamic balancing is complicated. Although the yokeless segmented stator improves winding yield, the difficulty of controlling splicing precision, coaxiality, and air-gap uniformity is extremely high. The manufacturing yield of conventional DRSS is around 90%, lower than that of DSSR. In addition, permanent-magnet retention and centrifugal risk for large-diameter thin-disc rotors at high speed are also issues that DRSS must specifically address.
The application scenarios of DRSS are therefore very clear: occasions that have a rigid demand for power density and torque density and are willing to pay for cooling systems and precision manufacturing—EV main drives, aerospace actuators, high-performance robot joints, and eVTOL propulsion systems. It is not a “general solution,” but a “targeted solution under high constraints.”
SSSR | DSSR | DRSS / YASA | |
Axial magnetic pull | Severely unbalanced; high bearing load | Symmetrically canceled; good stability | Symmetrically canceled; good stability |
Thermal conditions | Single-sided cooling; moderate | Both stators exposed; continuous operation possible with air cooling | Stator sandwiched; heavily dependent on liquid cooling |
Torque density | Baseline | Good | 15%–25% higher than DSSR |
Cost level | Lowest | Moderate (15%–25% higher than SSSR) | Higher (liquid cooling + precision manufacturing) |
Manufacturing yield (reference) | Above 98% | 92%–95% | Around 90% |
Core constraint | Not suitable for high-speed/continuous heavy load | Dual-stator alignment and gap control | Cooling system and dual-rotor synchronization accuracy |
Typical applications | Low-power servos, fans, light-load portable equipment | Industrial continuous operation, medium-power servos, in-wheel drive | EV main drive, aerospace, high-performance robotics |
Data source:
If the three topologies are put back into real engineering decisions, the selection path can be narrowed down to three questions:
Question 1: Is cost the first constraint?
If the answer is “yes,” and the operating conditions are mainly low-speed, light-load, and intermittent, SSSR is a reasonable starting point. Its shortcomings can be managed under low-stress conditions, and the cost advantage brought by its minimalist structure is real.
Question 2: Does the application require long-term continuous operation, and are cooling conditions limited?
If the equipment “cannot stop” and cooling means are limited, the dual exposed stators of DSSR provide system-level thermal margin. The extra cost of one set of stators is exchanged for improvements in bearing life, winding insulation life, and overall equipment availability. In industrial continuous operation and commercial in-wheel drive, this is often a worthwhile trade.
Question 3: Is power density a non-negotiable hard?
If the application is extremely sensitive to weight and volume—for example, an electric vehicle that is extremely particular about unsprung mass, or an aircraft where every gram affects range—then the high torque density of DRSS/YASA is an option that cannot be avoided. But the premise is: the budget for the cooling system and precision assembly must be in place at the same time. The performance ceiling of DRSS is bought with liquid-cooling circuits and micron-level air-gap control; omit either one, and the high torque density will remain only in datasheet parameters.
There is no “generational dominance” among the three topologies. The minimalism of SSSR, the balance of DSSR, and the density of DRSS each correspond to a clear set of physical constraints and engineering trade-offs. The essence of selection is to first honestly answer “what my application can least accept,” and then match the structure that least touches the red line. The charm of axial-flux motors lies in being “flat,” but what truly makes them run is always a clear understanding of constraints.